The Ageing Brain: The Neurobiology and Neuropsychiatry of Ageing

THE AGEING BRAIN

THE AGEING BRAIN THE NEUROBIOLOGY AND NEUROPSYCHIATRY OF AGEING

Edited by

PERMINDER S. SACHDEV

This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Copyright © 2003 Swets & Zeitlinger B.V., Lisse, The Netherlands All rights reserved. No part of this publication or the information contained herein may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, by photocopying, recording or otherwise, without written prior permission from the publishers. Although all care is taken to ensure the integrity and quality of this publication and the information herein, no responsibility is assumed by the publishers nor the author for any damage to property or persons as a result of operation or use of this publication and/or the information contained herein. Published by: Swets & Zeitlinger Publishers www.szp.swets.nl ISBN 0-203-97097-7 Master e-book ISBN

ISBN 90 265 1943 5 (Print Edition)

Contents

ACKNOWLEDGEMENTS

ix

Section I

1

Introduction

CHAPTER 1: THE AGEING BRAIN Perminder S Sachdev

3

CHAPTER 2

POPULATION AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS: A FIFTH EPIDEMIOLOGIC TRANSITION

11

G Anthony Broe Section II

Characteristics of the ageing brain

CHAPTER 3 STRUCTURAL Jillian J Kril

CHANGES IN THE AGEING HUMAN BRAIN

CHAPTER 4 STRUCTURAL NEUROIMAGING OF Jeffrey CL Looi and Perminder S Sachdev

THE AGEING BRAIN

NEUROPHYSIOLOGICAL, SENSORY AND MOTOR CHANGES WITH AGEING

33 35 49

CHAPTER 5

63

Stephen R Lord and Rebecca St George CHAPTER 6 COGNITIVE CHANGES AND THE Helen Christensen and Rajeev Kumar

AGEING BRAIN

75

AGEING OF THE HUMAN BRAIN AS STUDIED BY FUNCTIONAL NEUROIMAGING

97

CHAPTER 7

Julian N Trollor and Perminder S Sachdev

vi

CONTENTS

CHAPTER 8 NEUROENDOCRINE George A Smythe

ASPECTS OF BRAIN AGEING

CHAPTER 9 CEREBROVASCULAR SYSTEM AND THE Valendai K Srikanth and Geoffrey A Donnan Section III

AGEING BRAIN

Factors influencing brain ageing

CHAPTER 10 THE

MOLECULAR BASIS OF AND FRONTOTEMPORAL DEMENTIA

139 153

171

ALZHEIMER’S

DISEASE

173

John BJ Kwok and Peter R Schofield CHAPTER 11 OXIDATIVE

AND FREE RADICAL MECHANISMS IN BRAIN

187

AGEING

Judy de Haan, Rocco C Iannello, Peter J Crack, Paul Hertzog and Ismail Kola CHAPTER 12 THE

ROLE OF NUTRITIONAL FACTORS IN COGNITIVE

205

AGEING

Janet Bryan CHAPTER 13 THE BRAIN Peter W Schofield Section IV

223

RESERVE HYPOTHESIS

Clinical interface

CHAPTER 14 WILL Carol Brayne

241

WE ALL DEMENT IF WE LIVE LONG ENOUGH?

CHAPTER 15 DETECTING ALZHEIMER’S PRE-SYMPTOMATIC STAGE Gary W Small

DISEASE AT THE

259

CHAPTER 16 PARKINSONISM AND AGEING John GL Morris, Mariese A Hely and Glenda M Halliday CHAPTER 17 AGE VARIATION IN THE PREVALENCE ARE STUDY FINDINGS MEANINGFUL? John Snowdon CHAPTER 18 VASCULAR DEMENTIA Perminder S Sachdev

243

OF

275

DEPRESSION: 283 299

vii

CONTENTS

CHAPTER 19 CONCLUSION Perminder S Sachdev

323

CONTRIBUTORS

327

SUBJECT AUTHOR

INDEX INDEX

ADDRESS LIST

333

Acknowledgements

The seed for this book was sown with the formation of The Ageing Brain Program at the University of New South Wales in 1998, and the early sprout appeared in 2000 at the International Conference on the Ageing Brain held at the Scientia, University of New South Wales, Sydney. The book is, of course, more than the Conference, and its diverse foliage is the dedicated work of many scientists and scholars. I am extremely grateful to all the authors for that extra effort that made each of the chapters a significant contribution. The editing of a book is a labour of love that demands doggedness and compulsive persistence. The latter qualities were brought to this work with measured good humour by Angela Russell, who undertook the tasks of editing and compiling. If she sometimes annoyed the contributors with her deadlines and diligent proofreading, the final manuscript will more than compensate for it. She was assisted in this task by the quiet and behind-the-scenes contribution of Wanda Schinke. The flair of Joanna Christie was an important determinant of the success of the Conference. The planning of this book, and its intellectual content, were influenced by many colleagues to whom I am extremely grateful. I would like to make particular mention of Sam Aroni, Henry Brodaty, Tony Broe, Felicia Huppert, Jeffrey Looi, Gary Small, Julian Trollor, Michael Valenzuela and Xing Li Wang who were generous with their suggestions and time. I found, in the publishers of this book, a rather indulgent group of professionals, led by Arnout Jacobs, who let many deadlines go past with little more than gentle reminders. For the undisturbed small hours of the morning that the writing took me into, I am in debt of my family — my beautiful wife Jagdeep and our lovely daughters Nupur and Sonal. They have provided the environment which has continued to nurture me through all my academic travails. My research into neuropsychiatric disorders of the elderly has been supported generously by the University of New South Wales and the National Health

x

ACKNOWLEDGEMENTS

and Medical Research Council of Australia. Additional support has been provided by the Rebecca Cooper Foundation, the Brain Foundation, the Fairfax Foundation and Pfizer Inc. None of these organizations has any vested interest in the intellectual content of the book or any commercial interest in it. Perminder S Sachdev

SECTION I INTRODUCTION

Chapter 1 THE AGEING BRAIN Perminder S. Sachdev

According to Hesiod, a Greek philosopher, the history of mankind could be divided into five epochs. The first was the Age of Gold in which mortals never aged and peace and happiness were pervasive. This was followed by the Age of Silver in which childhood lasted a hundred years but adulthood was transient. The third age, in which Hesiod lived, was the Bronze Age, which was a time of greed, corruption, injustice and violence. When this ended, Zeus created the Heroic Age in which the world was populated by demigods. Then came the Iron Age, in which we now live. Hesiod wisely predicted that this would be the age of violence, the love of profit, and an increasingly decadent lifestyle. Hesiod predicted that Zeus would be particularly incensed by the lack of honour shown to the elderly by the young, and by children not repaying their parents for the nurturance they received. Zeus would then create a new and idyllic Age. A hallmark of our Age is also the belief that future Ages are of our own making. Few would disagree that the salient characteristic of a golden age of the future would be eternal youthfulness, or at least youthfulness until the time of delayed but sudden death. This may explain our preoccupation with ageing. The images of ageing we confront on a daily basis are contrasting in nature. For those of us who are in mid-life, ageing represents a relentless erosion of our vitality. There are the obvious reminders in the greying hair, the balding scalp, the slight stiffness in the joints, and the small lapses of memory. The death notices in the newspaper become noticeable. A sudden dread fills our hearts as we witness our parents succumb to the travails of senescence. Yet we still hope to age like fine wine, accumulating Talmudic wisdom with our years. Individuals, like the late Madame Jean Calment of France, remind us that we could be living independently well into the 12th decade of our lives.1 We wonder if the tools of modern biology will uncover the mysteries of ageing and help control, if not reverse, it. We look with wonderment at the genome project and the vast worldwide army of biomedical scientists. Our dread is inter-mixed with awe and expectation.

4

THE AGEING BRAIN

More than any other organ of the body, we are concerned with the ageing of the brain. The ageing brain must be considered a special case within the domain of ageing. While age-related changes in the brain in general parallel those of the body, there are important exceptions. Brain diseases that are usually regarded as concomitants of old age, such as Alzheimer’s disease (AD) and Parkinson’s disease (PD) are sometimes seen in the young, and most elderly individuals manage to evade them. It is not uncommon to see a very active mind in a frail old body. Our examination of the ageing of the brain must therefore occur in the context of a neurobiological understanding. The ageing brain is a special case also for social reasons. An epidemic of dementia is upon us, and governments in the developed world are preoccupied with the impact this will have on the society of the future. As Professor Broe states later in this book, we are in the Age of Neurodegenerative Disorders, and insights into the mechanisms of ageing are urgently needed. . In the face of this dread, we must take heart in the pace of neuroscientific research, which has been breathless indeed. Let us consider a few examples. Most studies both in vivo and post mortem, suggest shrinkage of the adult brain as it ages, with a reported reduction of about 5% in brain weight per decade after the age of 40 years.2 This change is not uniform, however, with the prefrontal regions being affected more than the temporal and parietal neocortex. In the subcortical regions, the neostriatum atrophies moderately with age, while the globus pallidum and thalamus are relatively spared.3 What we do not understand are the reasons for this variation. Why is the substantia nigra, for example, more susceptible to age-related degeneration than the thalamus? What are the determinants of hippocampal degeneration seen in ageing brains? Questions such as this may be the keys to understanding the physiological processes involved in brain ageing. It is quite likely that the mechanisms of neural ageing are the same as for the rest of the body. There must also be important differences. A large part of the human genome is involved in brain development, suggesting that a great complexity must be fathomed. It was thought for many years that the changes in brain volume seen in ageing were a consequence of age-related neuronal loss.4 This notion was so well accepted that it had entered lay parlance. Recent studies using better stereological methods have shown that this may not be true, and in fact most brain regions do not suffer an age-related neuronal loss.5 If there is a sparing of the total number of cortical neurones, what is the basis of loss of cortical volume with ageing? The hippocampus has been studied extensively to understand this, and it has been shown that its functional organization is altered with ageing. This is related to alterations in connectivity, because of reductions in dendrites and synapses. In both rodents and humans, changes have been reported in dendritic arbor, spines and synapse morphology that could impact on the function of hippocampal circuits but would not be reflected as neuronal loss.6 This is functionally important as the most cognitively impaired aged rats demonstrate the greatest degree of abnormality.

THE AGEING BRAIN

5

The number of synapses can be judged by the density of receptors in the molecular layer as has been shown for the glutamate NMDA receptor which plays a critical role in mechanisms of plasticity comprising the cellular basis of learning and memory.7 The physiological implications of this change with ageing have been studied from many perspectives, one of which is long-term potentiation (LTP). This is a functional change in synaptic transmission secondary to neuronal stimulation, and has a role in memory functioning. The stimulation necessary to induce peak LTP, and the maximal potentiated response attained, are the same in young and old brains, but LTP decays to the prepotentiated baseline levels more rapidly in aged subjects — a possible reason for the “forgetting” experienced by the elderly. Many aspects of synaptic transmission are unaffected by age and there may even be compensatory changes. Functional imaging studies show that aged brains are less efficient in the processing of information, tending to recruit more extensive networks of neurones.8 This may, however, be a correctable change, since the brain is known to retain much of its plasticity despite age, and presents a potential for intervention. The cognitive changes associated with ageing are the subject of intense research. It is reassuring that crystallised intelligence remains intact with age, although some cognitive abilities do show a gradual reduction. Ageing causes a decline in information-processing resources, such as working memory capacity, attentional regulation and processing speed. Ageing results in greater intra-individual and inter-individual variation in performance. The change in these resources must be understood from a neurobiological perspective. It is interesting that the onset of the decline is relatively early in life, and to some extent parallels the decline in physical and sensory functions. The relationship with structural brain change is far from perfect, and can only be demonstrated in those with a pathological degree of impairment. It may be also be interesting to examine the cognitive decline using computational theories of neuronal function.9 Does the change in resources lead to deficient neuromodulation and increased neural noise, i.e. haphazard activation during neuronal information processing? There is evidence that mental representations in the elderly are less distinctive. Events that happen in the course of a day are less distinctly remembered by older individuals, suggesting that they may be processing information less elaborately. Neuro-imaging studies show that the elderly are more likely to activate both hemispheres for tasks that are lateralised in young individuals.10 One proposed mechanism for the increased neuronal noise with age is reduced neuronal responsivity due to a declining dopaminergic modulation.9 There are obviously other possibilities, which prompt for a multi-level approach to the problem of cognitive ageing. When one examines the risk factors for cognitive decline with age, one is confronted with the question: how much of the change is because of pathology in the brain? This is an issue not easy to settle. An ageing brain accumulates pathology that may be due to cerebrovascular disease or systemic diseases with their secondary brain effects. This may account for some of

6

THE AGEING BRAIN

the age-related deficits that are likely to be misattributed to ageing-related changes. As an example, brains of elderly individuals frequently show hyperintense signals on T2-weighted magnetic resonance imaging (MRI), which has been the subject of hundreds of studies.11 Are these findings always indicative of pathology, or can they represent normative ageing-related changes? What role do these “lesions” play in the development of cognitive change and psychiatric disorders in late life? It is necessary to pose such questions to understand the nature of the “normal” ageing process. We have seen impressive advances in our understanding of neurodegenerative diseases in the last two decades, and are now at the threshold of effective treatments. Much work however remains to be done. Epidemiological studies have revealed many risk factors that have to be explained in terms of pathophysiological processes, and major gaps in this understanding remain. Genetic factors still explain only a minority of cases of Alzheimer’s disease (AD). The great divide between neurodegeneration and vascular pathology is no longer an unbridgeable gulf, but the physiological basis for this association is yet to be understood. The argument whether AD is an extreme form of ageing is unresolved, and some do believe, as Carol Brayne argues in this book, that if we live long enough, all of us will succumb to AD. We know a great deal about other causes of brain degeneration, but the reasons why one brain with fronto-temporal degeneration produces Pick bodies and not another remain elusive. Why is it that a particular region of the neocortex is preferentially affected in some dementias such as semantic dementia, progressive aphasia, etc.? This is a field in which the contributions of clinicians, epidemiologists and neuroscientists have gone hand in hand, and the future lies in the continued cooperation between disciplines. If we are to influence the ageing process, it is necessary that we understand the underlying molecular mechanisms. The frontier of the biochemistry of ageing, although yet to be conquered, is witness to many raging battles. As a consequence, terms like free radicals, heat shock proteins and nerve growth factors have become household words. Some of these theories have inextricably linked the brain with the rest of the body. As an example, the role of corticosteroids in the stress response, and its influence on brain structures has provided a link between psychology and biology.12 Brain regions that are important for learning and memory processes are particularly sensitive to stress hormones. The hippocampus has a high concentration of adrenal steroid receptors. Stress can thereby impair memory acutely; and chronic or repeated stress can lead to atrophy of dendrites and reduced neuronal connections. If prolonged this change can become irreversible and loss of neurones results. It has also been shown that early stress, such as prolonged separation of rat pups from their mothers, may lead to a chronic over-reactivity to stress in these animals. This may result in accelerated brain ageing. Other hormones such as oestrogen, growth hormone, melatonin, testosterone and dehydroepiandrosterone are being examined for their role in reversing some aspects of ageing. Growth or trophic factors abet these.

THE AGEING BRAIN

7

The brain is intricately linked with the immune system, and age-related changes to the immune system have been of special interest to neurobiologists. The function of T-cells and their ability to proliferate declines with age. The T-cells produce powerful chemicals known as lymphokines, which mobilise mediators of the immune response. The effects of age on these lymphokines are variable, with rise in some and fall in others. It is not known how this may be linked to neuronal function. In this age of genomics, some of the causes of brain ageing are being sought in genetic factors. Most of the progress in neurogenetics has been in discovering genes for various neurological diseases, including those that affect the elderly. There have been exciting developments in AD, with the discovery of three genes that cause early-onset AD. However, this accounts for the disorder in but a small proportion of AD patients. The discovery of the tau gene in fronto-temporal dementia has raised the question of the relationships between the different genes, and what pathways may be shared in neurodegenerative disorders. The pace of this research is likely to increase as animal models are established. Recent research has shown that the expression levels of many genes related to neuronal signalling, plasticity and structure are altered with ageing.13 For example, the expression of certain proteases, such as prolyl oligopeptidase and caspase-6, is up-regulated in the aged brain. These proteases play essential roles in regulating neuropeptide metabolism, amyloid precursor protein processing, and neuronal apoptosis, and are likely contributors to brain ageing. It is interesting that some of these changes in gene expression can be reversed by environmental enrichment,14 providing hope for intervention. The mapping of the human genome, and the recognition that a large number of genes are involved in brain development, has opened up exciting opportunities for understanding the molecular basis of brain ageing. It would be important to find out if there are a few major genes that determine ageing, or is it the result of the cumulative effect of changes in many genes? Is ageing the result of defects in DNA repair that gradually accumulate? Are there some genetic modifications that can delay, if not stop or reverse the processes of brain ageing? Does dietary restriction delay ageing through genetic factors?15 Attempts have recently been made to apply gene transfer technology to protect neurones from death following neurological insults.16 It is conceivable that gene therapy in the future may be able to protect the nervous system from ageing. Transgenic intervention could be in order to over-express a particular gene to protect against decline of its product in old age, or gene therapy could target a discretely damaging event highly likely to occur in the elderly. Technologies for the delivery of genes into neurones to maintain function and protect against injury are being developed. Genomics is likely to be complemented by the newly developed science of Proteomics,17 as there are many more proteins in the human body than can be accounted for by the number of genes recognised on the genome. The nearly 30,000 genes have a complement of nearly 300,000 proteins, and each

8

THE AGEING BRAIN

of the 200 cell types in the body has a different set of proteins. The proteins produced by the cells at any particular time are moderated according to the biological needs of the body, and are influenced by the disease process present. Proteomics therefore permits the identification of proteins associated with particular diseases. This will assist diagnosis, as is already the case in AD, and speed the development of new treatments. It also offers the exciting possibility that drugs may be tailored the to individual patient, opening up an era of personalised medicine. The treatments of neuropsychiatric disorders of the elderly are likely to look very different in the future, as suggested by the above developments. In many respects, the future is already here. Depressive disorders have recently seen the introduction of two novel treatments: transcranial magnetic stimulation and vagus nerve stimulation. Parkinson’s disease patients worldwide are benefiting from deep brain stimulation. Targeted drugs are increasingly being developed for specific receptors. Stem cells promise to open up a new era in therapy. In fact, recent findings suggest that a decline in the numbers and plasticity of stem cells may contribute to ageing itself.18 It is likely that methods will be developed to tweak the stem cells already in the brain, or introduce new ones, to replace lost or dysfunctional cell populations. The above developments reveal the rapidity with which new information is being acquired and old orthodoxies challenged. However, a glorious ageless society is not upon us yet. In the medium-term, our goals as a society must be limited. We can start by emphasising the positive aspects of old age. Some people may find this a difficult concept to grasp, and yet for thousands of years, societies have valued age and even venerated it. The wisdom of old age is difficult to quantify, but recent research showed that on a rational choice task, 70 year old subjects performed much more consistently than those 50 years younger.19 With the inevitable ageing of our populations, we have little choice but to make old age productive, healthy and enjoyable. Much of this will be achieved through social and political change and not medical advances. An increasing number of healthy older people can make a significant contribution to the lives of younger generations. Age can help temper and direct the energy of the young. Medical science does not, in the near future, hope to conquer ageing. It can have a more modest goal, however, in delaying the onset of late-life dysfunction. Old age is characterised by an array of ageing-related diseases, which include cardiovascular disease, dementia, sensory deficits, Parkinson’s disease, diabetes, osteoporosis and incontinence. The mere delaying of the onset of some of these will have a major public health impact. For instance, a delaying of the onset of Alzheimer’s disease by five years will halve the prevalence of the disorder. It is this promise that is prompting a burgeoning industry of health promotion. Low-fat labels, cholesterol free diets, folic acid supplementation, aspirin prophylaxis, anti-oxidants and organic foods are more than passing fads. In this rush toward a healthy old brain, it is difficult to separate established scientific facts from overvalued ideas. A few messages

THE AGEING BRAIN

9

do seem to have sufficient empirical basis. We can protect the brain somewhat if we control hypertension early and effectively, and attend to other cerebrovascular risk factors such as diabetes, smoking, high cholesterol and obesity. We should aim for moderation in our use of alcohol, and perhaps try to restrict it to red wine, while we refrain from using illicit substances. Whether we will benefit from using anti-oxidants or anti-inflammatory drugs remains to be established. The use of folic acid supplementation to reduce serum homocysteine levels is again not established as an epidemiological health measure20 but is increasingly popular. Also without sufficient scientific backing, the use of a daily multivitamin tablet that does not exceed the RDA of its components makes sense for most adults, given the greater likelihood of benefit than harm and the low cost.21 The use of vitamin E at 400 IU per day in middle and old age by those at risk of vascular disease can also be recommended.21 Stress, no doubt, is bad for the body and the brain, and has been linked with psychiatric and cognitive disorders, and we should unequivocally recommend stress-reduction strategies to our patients. The action of stress on neurones is most probably through the glucocorticoid cascade. This response can be modified by environmental manipulation as early as in the neonatal period,22 and continuing on into later life. It is interesting that this manipulation of the adrenocortical axis can safely and effectively be brought about by a psychologist.22 The promotion of other hormones, such as growth hormone, melatonin, DHEA, pregnenolone, testosterone, oestrogen and progesterone, as elixirs of youth is without unambiguous scientific basis. A study from Boston23 showed that exercise can strengthen muscles, improve mobility, and reduce frailty even among 90-year-old individuals. The same may be true for the brain, which harbours a significant potential for plastic change well into old age.24 Another analogy to be drawn with muscles is that the brain has a reserve than can be influenced by mental activity, and serves to protect the individual from age-related changes.25 A number of studies have reported that higher educational and occupational levels, mental activity and high intellectual performance are protective factors for dementia. These findings, and those relating to nutritional factors and stress, promote an agenda for the future that is hopeful, and suggest interventions at the population level that should begin now without awaiting breakthroughs in the understanding of molecular processes. It is important to take this message to decision-makers if we are to influence the future of an ageing society. References 1. 2.

Ritchie K. Mental status examination of an exceptional case of longevity — J.C. aged 118 years. Br J Psychiatry. 1995; 166:229–235. Kemper T. Neuroanatomical and neuropathological changes during aging and in dementia. In: Albert M, Knoepfel J, editors. Clinical neurology of aging. New York: Oxford University Press, 1994; 3–67.

10

THE AGEING BRAIN

3. Trollor J, Valenzuela M. Brain ageing in the new millenium. ANZ J Psychiatry. 2001; 35:788–805. 4. Brody H. Structural changes in the aging nervous system. Interdiscip Topics Gerontol. 1970; 7:9–21. 5. Wickelgren I. Is hippocampal cell death a myth? Science. 1996; 271:1229– 1230. 6. Hamrick J, Sullivan P, Scheff S. Estimation of possible age-related changes in synaptic density in the hippocampal CA1 stratum radiatum. Soc Neurosci Abstr. 1998; 24:783. 7. Gazzaley AH, Siegel SJ, Kordower JH, Mufson EJ, Morrison JH. Circuit-specific alterations of N-methyl-D-aspartate receptor subunit 1 in the dentate gyrus of aged monkeys. Proc Natl Acad Sci USA. 1996; 93:3121–3125. 8. Almkvist O. Functional brain imaging as a looking glass into the degraded brain: reviewing evidence from Alzheimer disease in relation to normal aging. Acta Psychol. 2000; 105: 255–277. 9. Li S-C, Lindenberger U, Sikstrom S. Aging cognition: from neuromodulation to representation. Trends Cog Sci. 2001; 5:479–486. 10. Cabeza R, McIntosh AR, Tulving E, Nyberg L, Grady CL. Age-related differences in effective neural connectivity during encoding and recall. NeuroReport. 1997; 8:3479–3483. 11. Pantoni L, Garcia J. Pathogenesis of leukoaraiosis: A review. Stroke. 1997; 28: 652–659. 12. Sapolsky R. Stress, the aging brain, and mechanisms of neuronal death. Boston: MIT Press, 1992. 13. Jiang CH, Tsien JZ, Schultz PG, Hu YH. The effects of aging on gene expression in the hypothalamus and cortex of mice. Proc Nat Acad Sci USA. 2001; 98: 1930–1934. 14. Rampon C, Jiang CH, Dong H, Tang YP, Lockart DJ, Schultz PG, Tsien JZ, Hu YH. Effects of environmental enrichment on gene expression in the brain. Pro Nat Acad Sci USA. 2000; 97:12880–12884. 15. Weindruch R, Walford RL. The retardation of aging and disease by dietary restriction. Springfield, IL: Charles C Thomas, 1988. 16. Ogle WO, Sapolsky RM. Gene therapy and the aging nervous system. Mech Ageing Dev. 2001; 122:1555–1563. 17. Banks RE, Dunn MJ, Hochstrasser DF, Sanchez JC, Blackstock W, Pappin DJ, Selby PJ. Proteomics: new perspectives, new biomedical opportunities. Lancet. 2000; 356:1749–1756. 18. Rao MS, Mattson MP. Stem cells and aging: expanding the possibilities. Mech Ageing Dev. 2001; 122:713–734. 19. Tentori K, Osherson D, Hasher L, May C. Wisdom and aging: irrational preference in college students but not older adults. Cognition. 2001; 81:B87–96. 20. Diaz-Arrastia R. Homocysteine and neurologic disease. Arch Neurol. 2000; 57: 1422–1427. 21. Willett WC, Stampfer MJ. Clinical Practice. What vitamins should I be taking, doctor? New Eng J Med. 2001; 345:1819–1824. 22. Seligman M. Learned optimism. New York: Alfred Knopf, 1991. 23. Levine S. Plasma-free corticosteroid response to electric shock in rats stimulated in infancy. Science. 1962; 135:795–798. 24. Anstey K. How important is mental activity in old age? Austr Psychol. 1999; 34: 128–131. 25. Schofield P. Alzheimer’s disease and brain reserve. Australas J Ageing. 1999; 18: 10–14.

Chapter 2 POPULATION AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS: A FIFTH EPIDEMIOLOGIC TRANSITION G Anthony Broe

Introduction Rapid population ageing, or a rising percentage of older people in the population, was a 20th century phenomenon in developed countries and is now affecting most of the world’s populations, the poor as well as the rich. Countries as disparate as Australia, Iran, Thailand and Tunisia are approaching or achieving below replacement fertility levels.1 The immediate cause of world population ageing was fertility decline; this followed the reduction in infant deaths due to infectious diseases from the 19th century onwards and their substitution by deaths due to adult onset degenerative diseases in the first half of the 20th century. The infectious diarrhoeas, influenza and tuberculosis, in children and young people, were gradually replaced by cardiovascular and lung diseases at older ages. Omran,2 in 1971, referred to this shift in disease patterns as the “Epidemiologic Transition”; and he described three disease transitions occurring in developed countries up to the mid-20th century. This chapter examines population ageing and life span in relation to further disease transitions and changing causes of mortality and morbidity later

12

THE AGEING BRAIN

in the 20th century. A progressive delay in age of onset, and a decline in mortality from the systemic degenerative diseases (such as cardiovascular and lung diseases) was described as the fourth transition of “delayed degenerative diseases” by Olshansky and Ault in 1986.3 We are now seeing yet another substitution of mortality due to later onset neurodegenerative disorders, such as dementia and Parkinson’s disease (PD).4;5 It is predicted that the neurodegenerative disorders will gradually replace the systemic degenerative disorders as the major causes of both death and morbidity in the 21st century. Population Ageing Population ageing is the product of three factors: birth rate, infant mortality and life span. During the 19th century, and the first half of the 20th, these three demographic factors were largely determined by major external or environmental assaults due to infectious diseases, malnutrition and trauma. Reductions in these risk factors resulted in improvements in maternal and child health and increased infant survival. This was followed by declining fertility and a decreased birth rate leading towards zero population growth; hence the almost instantaneous ageing of Western populations in the first half of the 20th century. Population ageing is one area of human ageing that cannot be claimed by the geneticists as their responsibility. So far it is primarily environmental. The History of Ageing Group in Cambridge examined birth and death records in five parishes in England between 1541 and 1981 to produce an

1541

1751

1921

1981

Figure 1. The proportion of elderly in the English population, 1541-1981 (adapted from Laslett,6).

AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS

13

accurate projection of the number of over –60s in the English population during a period of 440 years6 (Fig. 1). For almost 400 of those 440 years the over –60s fluctuated at around 8% of the population. A consistent rise above 8% did not occur until the 1920s or a mere 80 years ago when rapid population ageing in developed countries commenced. The Epidemiologic Transition Theory The epidemiologic transition theory 2,3,7 explains these population changes by major shifts in health and disease patterns, and in recorded causes of death, as societies change or “modernise” with resultant improvement in social, economic and health factors. The base or first stage of the epidemiologic transition, graphically described by Omran2 as “The Age of Pestilence and Famine,” was characterised by very high death rates due to pandemic infections, trauma, poverty and malnutrition. The major killers during this long pre-industrial era were the ubiquitous diarrhoeas, influenza, tuberculosis and pneumonia, as well as epidemics such as bubonic plague and small pox. The major death toll occurred in infants and young children, with low average life span and a low and static percentage of older people in the population. The next stage of the epidemiologic transition followed the scientific revolution and the start of the industrial revolution in England and Europe in the 18th and 19th centuries. Concomitant social and economic changes gradually brought greater wealth, better nutrition, less crowding, better education, better hygiene, healthier mothers and stronger infants despite the social upheavals of industrialisation and urbanisation. Omran’s second epidemiologic stage, “The Age of Receding Pandemics,” with decreasing infant mortality and an ongoing high birth rate, resulted in more infant survivors and an overall younger population in England up to the mid-19th century. Declining fertility then lead to progressive population ageing, with a shift in causes of death to the later-onset systemic degenerative diseases (particularly cardiovascular and lung disease) and a shift in mortality from the young to the old. These shifts heralded Omran’s third epidemiologic stage “The Age Table 1.

The Epidemiologic Transition Theory (Western model). • The Age of pestilence and famine • The Age of receding pandemics • The Age of degenerative diseases • The Age of delayed degenerative diseases

Omran (1971)2 and Olshansky & Ault (1986)3

14

THE AGEING BRAIN

of Degenerative and Man-made Diseases”. In retrospect, this stage represents a major achievement for the ageing survivors of an era of fatal infectious diseases, rather than a “man-made epidemic” of modern life, as it has often been painted. At the time of publication of his general theory of epidemiologic transition and mortality change in 1971, Omran and other demographers were predicting that average human life span would not progress beyond 70 years, then considered to be the biological as well as the biblical limit. Omran himself believed his third stage of degenerative diseases would be the completion of the epidemiologic transition. However mortality rates at older ages have continued to decline and average life expectancy at birth has continued to increase worldwide. These changes lead to the description of a fourth stage of the epidemiologic transition “The Age of Delayed Degenerative Diseases”, by Olshansky and Ault in 1986.3 This stage recognized the rapid decline in mortality due to chronic systemic diseases from the 1960s, particularly a decline in cardiovascular disease and stroke in developed countries. There was a delay in the ages at which these potentially fatal systemic diseases tended to kill, with rapid improvement in life expectancy concentrated among the population at advanced ages; a phenomenon described as “the ageing of the aged.” This ongoing decline in mortality has been attributed to new public health measures, including changes in major risk factors for systemic degenerative diseases such as smoking, diet and exercise, as well as advances in medical technology and drugs.3,8 However Olshansky3,9 has predicted that increases in life expectancy due to the prevention or delay of the known systemic degenerative diseases would not increase average life expectancy at birth much beyond 85 years. Lifespan and Compression of Morbidity The epidemiologic transition theory has focused on mortality with only implicit reference to morbidity, defined as the length and quality of survival in the presence of age-associated disease or disability.7 Description of the phenomenon of “ageing of the aged”, with recognition of significant increases in average life span beyond seven decades, brought increased attention to the concepts of “healthy ageing” or “successful ageing” with emphasis on the duration of disability-free survival, rather than longevity per se. James Fries10 outlined his theory of “Compression of Morbidity” in 1980. Based on a human life span of around 85 years, Fries predicted that chronic systemic disease, and consequent disability, would be delayed and compressed to the end of life by ongoing changes in life style and risk factors such as reduction in smoking, improved diet and more exercise. Recent data support this association.11 However, with further increases in human life span, it remains possible that morbidity is simply being delayed to later decades of life rather than

AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS

15

Figure 2. The theory of “compression of morbidity” as outlined by James Fries.10 Fries predicted that chronic systemic disease and associated disability would be delayed and compressed to the end of life by ongoing changes in lifestyle and risk factors.

compressed, i.e. from the “young-old” to the “old-old” or from the 70s to the 80s and 90s. A shift to a new epidemiologic transition of later-onset neurodegenerative disorders would result in additional causes of morbidity as well as mortality at advanced ages. While morbidity related to chronic systemic diseases appears to be declining, or being compressed to the end of life, the morbidity related to neurodegenerative disease, particularly the dementias, is increasing with advanced old age. Health related factors (reduction in smoking, improved diet and more exercise, etc.) reducing mortality and morbidity due to chronic systemic diseases have not been demonstrated to be protective against the chronic neurodegenerative diseases. Other protective factors for these diseases may be found, however, particularly those related to early brain growth and development, and to intellectual ability and education in early life. Average human life span in developed countries is now approaching the “natural limit” of 85 years described by Fries10 and predicted by Olshansky9 on the basis of possible cures for the major (systemic) degenerative diseases: cardiovascular diseases, lung diseases, and cancer. More extreme longevity remains possible, with average human life span going beyond the predicted 85 years and up to 100 or more years, but only if new causes of mortality decline are determined and modified, additional to those responsible for the

16

THE AGEING BRAIN

decline in later onset systemic diseases, or new factors determining longevity are identified. Important factors for further increases in longevity are likely to be those related to brain development and to lifelong improvements in cognitive and behavioural capacity. The worldwide nature of rapid population ageing12 and the timing of improvements in old age survival from the 1950s (before major advances in the “new” public health) suggest that general social and biodemographic factors, as well as health factors, are producing the improvement in the survival of the “old-old”. Vaupel’s group13 have demonstrated a substantial increase in human survival, commencing in the 1950s, with mortality data showing unpredicted and unexplained decline in mortality in those over 80 years of age accompanied by a marked rise in absolute numbers of the “old-old”. Their data indicate that the population of centenarians in developed countries has doubled every decade since 1960, mostly as a result of increases in survival after 80 years of age. This improvement in late-life survival is primarily nongenetic. It is largely determined by early life factors and experiences, which influence late life survival attributes, rather than by current conditions or risk factors operating in late life. Individual life span is seen as a product of internal (including genetic) defence mechanisms or survival attributes and external assaults on those defence mechanisms. External assaults, such as childhood infections, malnutrition or trauma, may overwhelm internal defences and lead to rapid or early death, as was common in the 19th century. However, the survivors of these external assaults in early life may improve their internal defence mechanisms (survival attributes) and lengthen their subsequent life spans in old age. Vaupel’s group13 have also shown that death rates decelerate with advanced age in multiple species: humans, medflies, wasps, drosophila, nematodes and yeast cells. From the combined data, they postulate that mortality decline with advancing age is a property of many complex systems. It appears to be related to a cohort “survivor effect” transmitted through individual fixed survival attributes, in an environment with markedly reduced external assaults compared to previous centuries. This late life “survivor effect” will apply to half the population in developed countries, as average survival reaches 80 years of age. Human Lifespan and the Brain In terms of human life span, it is proposed that brain function responsible for the human capacities for learning, cognition, insight and social knowledge, is one determinant of longevity in human populations.13-15 Socio-economic status, educational level, and mental ability or intelligence are closely linked. A cohort effect of increasing fluid intelligence, as measured by psychometric tests of verbal reasoning, spatial orientation and inductive reasoning, has been demonstrated over the 20th century in data that span the period from 1889 to 1966.16 This cohort effect, which has been attributed to improvements in

AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS

17

education, parallels observed changes in both education and longevity over the same period. Although no causal links have previously been suggested, it is arguable that improvements in education and fluid intelligence are in part responsible for increases in longevity. Socioeconomic status in childhood has been associated with mortality from a number of illnesses17;18 and educational level contributes to differences in mortality.19;20 Higher mental ability on test performance also correlates with better educational and occupational outcomes.21 However, until recent longitudinal observations on Scottish school children,22 there were few well-studied links between mental ability and mortality. These included an Australian Vietnam Veteran study23 and the Canberra24 and Rotterdam25 studies on older populations. The larger Scottish study was carried out on a cohort of 2792 school children in Aberdeen, with mental ability assessed at 11 years of age (in 1932) and survival determined 65 years later on 80% of the sample. It concluded that childhood mental ability was a significant factor among the variables that predicted age at death and hence longevity. The effects of IQ are difficult to separate from the effects of social class and education. It can be argued, however, that better brain development, whether it be in utero or in infancy and childhood, is likely to be an important survival attribute in 21st century society with its greatly reduced level of major external assaults and physical risk factors compared to previous centuries. It can also be argued that average human life span is likely to go beyond the 85 years which Olshansky has predicted on the basis of projected reductions in mortality from chronic systemic diseases.9 The important determinants of both mortality and morbidity in the “old-old”, in the knowledge-based societies of the 21st century, are likely to be better brain function on the one hand, and the neurodegenerative diseases associated with brain ageing on the other. The Neurodegenerative Diseases The important late-onset neurodegenerative diseases, for the determination of mortality and morbidity data in the older population, are dementia and Parkinson’s disease. Despite their very high prevalence in the “old-old”, the neurodegenerative diseases are, in general, poorly defined and diagnosed compared to the common systemic degenerative diseases (heart disease, stroke, chronic lung disease and cancer) and there is a high current level of under-ascertainment of neurodegenerative disease mortality.5 The late-onset neurodegenerative diseases include the dementias (Alzheimer’s disease [AD], dementia with Lewy bodies [DLB] and fronto-temporal dementia [FTD]) as well as Parkinson’s disease [PD] and motor-neuron disease [MND]). The commonest cause of visual loss in older people, age-related macular degeneration (ARMD), can be classified as a neurodegenerative disease, as can the almost universal age-related sensori-neural deafness. MND or amyotrophic

18

THE AGEING BRAIN

lateral sclerosis is the third commonest well-defined neurodegenerative disease of ageing. The term is also used for a host of less common familial and/ or sporadic neurological diseases of unknown cause, including progressive supranuclear palsy (PSP), cortico-basal degeneration (CBD) and the spinocerebellar atrophies (SCA), many of which appear to be age-related.26 As a class, the neurodegenerative diseases are primary neuronal disorders, i.e. not secondary to known vascular, malignant or toxic causes. Their defining feature is selective neuronal loss in a pattern that tends to be specific to each disease. Many neurodegenerative diseases (AD, PD, MND) manifest as a more common late onset sporadic form, which increases exponentially in incidence with advancing age over 70 years, and rare early onset dominantly inherited forms of what appear to be the same disease process. A number of neurodegenerative diseases are characterized by the accumulation, over many decades, of abnormal gene products in the brain; these proteins have variable associations with the selective patterns of neuronal loss observed in each disease. They include the accumulation of β-amyloid and tau in Alzheimer’s disease, forms of synuclein in PD and DLB, and forms of tau in FTD and other less common neurodegenerative diseases (PSP, CBD). The role of these proteins remains poorly understood in the pathogenesis of the specific diseases and particularly their late onset forms. Detailed study of the early onset familial forms is, however, providing significant insights into the role of some specific gene products. β-Amyloid in particular clearly plays an important role in the commonest age-related neurodegenerative disease, Alzheimer’s disease, and appears to have an additional role in brain ageing. Furthermore, β-amyloid, in conjunction with the evolution of the apolipoprotein alleles in humans, may have an association with basic evolutionary processes determining ageing and longevity.27 Neurodegenerative Diseases and Mortality The first four stages of the epidemiologic transition have been defined by mortality data using life expectancy and survival curves, combined with mortality data on specific causes of death.2,3 The focus has been on deaths from infectious diseases, and from the rapidly fatal systemic diseases, in particular mortality data for heart disease, stroke, lung disease and cancer,3 which are the commonest recorded causes of death in developed countries. They are diseases that tend to have well defined fatal outcomes, and mortality data for these disease categories is likely to be accurate. Clinical diagnosis is also likely to be accurate for the other systemic diseases listed among the 10 common causes of death in developed countries including endocrine, gastrointestinal and genito-urinary causes. Age-standardised mortality, as well as morbidity, for most of these systemic disease categories has been shown to be declining over the 20th century in developed countries, including Australia, with the most recent decline occurring in cancer deaths.28-30 The major exceptions

AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS

19

to this pattern of declining mortality from the Australian data are late-onset neurological diseases (over 75 years) and later onset “mental health disorders” (over 85 years). While the quality of mortality data for both infectious and systemic degenerative diseases is relatively good, the same is not true for mortality data for the neurodegenerative diseases, where under-ascertainment is a significant problem.5,31 Although dementia is the commonest neurodegenerative disease, its prevalence and incidence had not been well defined up until the last decades of the 20th century, particularly in the “old-old”. It is now clear that the prevalence and incidence of dementia rise exponentially with age at least up to 90 years.32,33 Alzheimer’s disease is the commonest dementia, and shows a doubling of incidence every five years from 65 to 90 years of age.33 There has been controversy as to whether dementia incidence plateaus off over 90 years of age, with two meta-analyses producing conflicting results.33,34 However a decline in AD incidence over 90 years, in men, is now supported by the recent EURODEM project35 and in both men (over 93 years) and women (over 97 years) by the Cache County study.36 Diagnosis of dementia during life remains difficult in old age and a number of studies demonstrate the lack of recognition of dementia in the community by family informant37 medical practitioners38 and nurses.39 Death certificates tend to record the “acute” systemic cause of death and mortality from death certificate data is estimated to be as low as 15% for dementia.5 Finally, studies suggest that Vascular dementia (VaD), a dementia of systemic cause, does not rise as rapidly with age as AD40,41 and that mixed dementia (AD and VaD) is common.42,43The inclusion of VaD in mortality and morbidity data for neurodegenerative diseases is often difficult to avoid, but should not greatly distort the results.5 Parkinson’s disease (PD), the second commonest neurodegenerative disease, is also poorly defined and diagnosed in older people, in whom atypical forms of the disease are more common.44 Until recently, idiopathic PD was stated to decline in incidence with ageing over 75 years in studies based on inappropriate methodology.45 It is now clear that its prevalence and incidence continue to rise with advanced ageing.26,42,46 This rising age-related prevalence may not be well recorded in mortality data. Only 25% of decedents with PD have the disease listed on their death certificates.5 Finally, in terms of the quality of mortality data, neurodegenerative disorders are commonly mixed in the “old-old”.42,43 Furthermore multiple preclinical syndromes commonly co-exist in older people and have been shown to predict subsequent dementia;47 these include cognitive or memory impairment (not reaching criteria for AD), motor slowing (not reaching criteria for PD) and evidence of vasculopathy. Because of multiple pathology in the “old-old”, the neurodegenerative disorders outlined commonly present as multi-factorial “Geriatric Syndromes”, rather than as specific neurological diseases amenable to specific diagnoses on death certificates. Many of the “Geriatric Syndromes” have a high mortality rate including: “immobility” with underlying parkinsonism and dementia; “instability and falls” with

20

THE AGEING BRAIN

underlying impairments of balance gait and vision; “delirium” with underlying frontal system impairments and dementia; and “aspiration pneumonia” due to underlying brain and oesophageal-motility disorders. However, the underlying causal diagnoses rarely appear on death certificates. Despite these potential problems with certification of deaths due to neurodegenerative diseases, a number of recent studies have shown increasing mortality from the three major neurodegenerative diseases: dementia,5,31,48 Parkinson’s disease49,50 and MND.51 Few studies have been able to compare mortality data for these neurodegenerative diseases with mortality data for common systemic diseases. Lilienfeld and Perl5 used US Census Bureau population estimates to project the annual death rate from three neurodegenerative diseases (dementia, PD and MND) and from six comparison systemic diseases (liver cirrhosis, colon cancer, lung cancer, cancer of the female breast, multiple sclerosis, and malignant melanoma) over the period between 1990 and 2040. The US National Center for Health Statistics routinely collects individual death certificates for all US residents. To determine death rates they used data for deaths in which the underlying cause was dementia, PD or MND (and the six comparison diseases) for the years 1985–1988. Assuming that the US disease-age-gender-race-specific death rates for these years remained constant over the period between 1990 and 2040, they found that neurodegenerative disease mortality increased by 119–231%, depending on the population model used. For the “middle” population growth model the increase was 166%, with the major component being deaths due to dementia. The increases in mortality for the six comparison diseases ranged from 52% (multiple sclerosis) to 130% (colon cancer). A number of factors make it likely that these projections for neurodegenerative disease mortality are underestimates including under-ascertainment on death certificates (for the reasons outlined above) and the conservative nature of the US Census Bureau estimates of population ageing. Furthermore the comparison with cancer deaths is with a category of systemic disease in which mortality is either still rising or static or showing the slowest falls, in comparison with other common systemic diseases, such as cardiovascular and lung disease, in which mortality is declining rapidly.28;30 Based on this review of the limited mortality, life expectancy and survival data, deaths from most systemic degenerative diseases continue to decline and are being replaced by deaths from even later onset neurodegenerative diseases, as part of a new disease transition. Neurodegenerative Diseases and Morbidity Overall, the quality of data to examine and compare morbidity and disability by disease cause across populations, has been less accurate than mortality data, with few reliable data on morbidity available.7 However it is increasingly important to define and measure morbidity, taking into account the

AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS

21

delayed onset, slower course and reduced mortality associated with the chronic systemic diseases over the past 50 years, the concomitant rise of the age-related neurodegenerative diseases, and the current controversies about compression of morbidity. This is particularly the case in comparing the burden of chronic systemic disease with that of chronic neurodegenerative disease, as it is the latter that is likely to rise in the 21st century with further population ageing. Many studies have shown rising age-related incidence, prevalence and morbidity in individual neurodegenerative disorders, such as dementia32,33,52 and PD.26,46,50 More general population based studies using Australian Bureau of Statistics (ABS) or US Census Data53,54 commonly rely on selfreport instruments to identify diseases and compare causes of disability, and have not shown this trend. However such instruments may not be sensitive to the impact of neurodegenerative diseases on disability, since cognition and/or insight are commonly impaired.55 This has been confirmed for the ABS disability instrument in a study comparing different disability measures, given to both respondents and informants.56 Few studies have used detailed clinical assessments in the field or compared incidence, prevalence or morbidity data, or disability rates, due to neurodegenerative diseases as a group with those rates due to the common systemic diseases. The Kilsyth Study in Scotland,40,57,58 completed in early 1970, examined the prevalence of major chronic disorders in the elderly and compared disability and dependence due to systemic and neurological causes. The study involved three community-living random samples, comprising 808 people, 65 years and over, examined by physicians experienced in geriatric medicine.58 It demonstrated that the prevalence of disability for IADL (defined as the inability to live at home without domestic help) increased from 12% at 65–69 years to over 80% at the age of 85 years. It showed that neurological disorders — dementia, balance/gait disorder, stroke and parkinsonism in that order — were the commonest cause of disability in 48% of subjects. Neurological and functional psychiatric disorders together contributed to 70% of disability compared with cardio-respiratory (38%), joint disease (24%) obesity (16%) and vision (11%). Neurological disorders, in 93% of cases (particularly dementia in 77% of cases), were by far the greatest contributor to the more severe category of dependence in ADL (defined as impairment in personal care) followed by joint disease (30%), cardio-respiratory (18%), vision (15%) and obesity (11%). Subsequent studies suggest that chronic systemic diseases have been delayed and compressed over the decades since the Kilsyth Study was completed in 1970,3,9-11 and the morbidity of specific neurodegenerative disorders, such as dementia32,52 and PD,26,46 has risen exponentially with advanced ageing of the population. The only known study in the subsequent two decades comparing prevalence and disability data in older people, for a range of common chronic systemic disorders with neurodegenerative disorders, is the Sydney Older Persons Study (1991 to 2002). This longitudinal study comprised two

22

THE AGEING BRAIN

Figure 3. The age distribution, by gender, in the Sydney Older Persons Study, Wave 1 (Waite et al., 1997)42.

random samples of 647 community-living people aged 75 to 98 years, with equal numbers of men and women. 59,42,43,47,56,60-62 The study provides prevalence data, on the same neurological and systemic disorders as the Kilsyth Study, in 522 subjects who agreed to a detailed medical, neurological, psychometric and disability assessment by a physician experienced in geriatric medicine.42 The 392 survivors examined at Wave 2 of the study, three years later, provided incidence data. The numbers assessed were modest, but the prevalence and incidence of both systemic and neurodegenerative disorders at this age range is very high, enabling examination of the data to look at the concept of a new epidemiologic transition in terms of causes of morbidity in an ageing population. The six major chronic systemic disorders measured included five shown to be the causes of disability from the Kilsyth study (heart disease, stroke, respiratory disease, arthritis and obesity), with the addition of peripheral vascular disease. The six neurodegenerative disorders measured included dementia, Parkinson’s disease, visual impairment, and a disorder of gait and balance, which had been identified as significant neurodegenerative causes of disability in the Kilsyth Study. Gait and balance disorder was further divided into two clinical components measured as motor slowing (gait slowing in subjects not reaching standard criteria for Parkinson’s disease) and gait ataxia (impaired heel-toe gait performance). Mild cognitive impairment (in subjects not reaching standard criteria for dementia) was included as the sixth neurodegenerative disorder measured. The data presented in Figures 4 and 5 are given in three- year age bands using smoothed estimates from a logistic regression model.

AGEING, HUMAN LIFESPAN AND NEURODEGENERATIVE DISORDERS

23

Figure 4. The prevalence of chronic systemic diseases, in three year age bands combined for males and females, in the Sydney Older Persons Study, Wave 1 using smoothed estimates from a logistic regression model (Waite et al., 1997)42. (N=522; *p L assymetry with age.

Standardised ROI template. Counts normalised to whole brain.

Regional CBF decreased with age in most cortical regions; minimally so in ocipital and cerebellar ROI’s, maximally in frontal ROI’s. Regression analysis indicated non-linear model; broken stick model with breakpoint median 36.6 yrs across regions. Increase in white matter CBF noted.

Standardised ROI template. Counts normalised to cerebellum and whole brain.

Subjects divided into young 50 Cerebellum normalisation: global decrease in rCBF with age. Regional decrease in left superior temporal, left basal ganglia, and bilateral frontal, occipital and parietal ROI’s. Whole slice normalisation: regional decrease in occipital, right frontal and increases in right thalamus, right anterior cingulate and bilateral posterior cingulate. Hemispheric perfusion difference R>L unaffected by age.

108

THE AGEING BRAIN

trapped intraneuronally by metabolic reduction, thus being distributed in the brain in proportion to rCBF. One of the main advantages of 99mTcHMPAO over other SPECT radiotracers is that it remains intraneuronally for at least six hours after intravenous administration, thus allowing images to be acquired for some hours after the injection, without loss of resolution. 99mTc-HMPAO has favourable dosimetry, allowing improvement in image quality at higher doses. An important development in SPECT technology has been the availability of high-resolution dedicated neuro-SPECT scanners, significantly narrowing the gap between SPECT and PET in achievable spatial resolution. Modern SPECT studies using 99mTc-HMPAO are listed in Table 2. These studies contrast in some respects to the earlier Xenon studies. With the exception of the study by Krausz et al.,17 in which detail was omitted, assessment of subjects has been sufficiently detailed to exclude confounding illnesses. Improved spatial resolution has enabled more detailed region of interest (ROI) analysis. Approaches to this analysis have involved either manually tracing individualised regions onto subjects’ scans18,19 or overlaying of templates with predefined anatomical or geometric regions.17,20,21 The technological and methodological advances are reflected in results which challenge the notion of age-dependent decline in global CBF, with a number of studies reporting a negative finding.18,20 Where age effects on laterality of CBF have been explored, preservation of R–L asymmetry has been reported.17,19 The most consistent regional effect of age has been the demonstration of a reduction in frontal CBF.17-19,21 Decline in temporal CBF18–20 has also been noted as an age effect. An interesting finding, reported in one study, was that age-dependent decline in rCBF was non-linear, with more reduction occurring up until middle age, and negligible further reduction thereafter.21 If this were confirmed, there would be significant implications for the age range included in future studies. A significant limitation of these studies is that SPECT measures represent relative rather than absolute CBF values. In obtaining this relative value, normalisation is frequently obtained to either whole brain or cerebellum. Discrepancies in the way in which the ageing process affects the cerebellum compared to other brain regions may introduce some variability in findings, and this has been noted in one study in which normalisation to both denominators was used.17 Resting 15O PET studies of ageing The methodological details and results of key 15O-PET studies of normal ageing are summarised in Table 3. PET investigations of cerebral correlates of brain ageing share similar problems to that of SPECT. Earlier studies featured samples drawn from hospital populations 22,23 or suffered from poorly described or inadequate screening processes.22,24 Early studies were performed with equipment offering poor spatial resolution.22–26 Despite recent improvement in a number of these methodological issues, the heterogeneity of the methods of analysis makes comparison of studies

CORRELATES OF BRAIN AGEING

109

difficult, and limits the conclusions able to be drawn. The majority of studies have used regions of interest (ROI) analysis, relying on poorly represented anatomical definition on PET scans themselves to allow placement of standardised or geometric ROIs. Supprisingly, only a small number of studies27,28 have sought to co-register the MRI and PET images. In the study by Bentourkia et al.,27 however, the co-registration method (visual interpolation of the two images) lacked the sophistication of other contemporary ROI analyses. Only one 15O PET study has used a semiautomated statistical analysis package — Statistical Parametric Mapping (SPM; Wellcome Department of Behavioural Neurology, Institute of Neurology, London) — for analysis. The issue of correction of CBF and CMRO2 measurements for partial volume effects has not been addressed in the majority of studies. Two studies have used CT scan based methods to correct CBF and CMRO2 measures for the effect of global atrophy.25,29 One study utilised an MRI-based method to correct for partial volume effects for individual regions,28 offering a more rigorous approach to partial volume correction. Studies of global CBF and CMRO2 Although there are inconsistencies in the results from these studies with respect to global changes in CMRO2, CBF and OEF with age, this is not unexpected given the methodological disparities. A number of studies document a global decline in CMRO2 with age.23,24,26,29–31 The magnitude of this effect is relatively small, with various estimates of 0.3%23 0.37%,27 0.5%26 and 0.6% per annum.29 The failure of one study to find evidence of reduction in CMRO2 with age deserves mention. Itoh et al25, showed linear decline with age in a measure of cerebral atrophy (cerebrocranial index) but no decline in either CBF or CMRO2. However, the age range of subjects was narrow, and weighted toward middle age and older subjects between 50 and 85 years. In view of the proposed non-linear decline in CBF with age proposed by Mozley et al.,21 it is possible that the failure of this study to detect global reduction in CBF or CMRO2 was related to the restricted age range of subjects included in this study. Modest decline in global CBF with age has been a common but less consistent finding (see Table 3). In addition to the negative finding by Itoh et al.25 with respect to both CBF and CMRO2, two studies demonstrated a discrepancy between age effects on global CMRO2 and CBF.30,31 In these two studies, CBF reduction was not seen with age, raising the possibility of an uncoupling between CBF and CMRO2 as a factor of age.31 However, as hypothesised by Yamaguchi et al.,30 such a discrepancy may be artifactual and reflective of other age-associated changes known to affect CBF measurement such as reduction in haematocrit and increase in PaCO2 with age. A less consistent finding in 15O-PET studies is that of changes in oxygen extraction ratio (OER) with age. A number of studies have failed to find such a relationship.22,29,30 In the study by Pantano et al.,23 a small but statistically insignificant increase in grey matter OER of 7% was found between young

110

THE AGEING BRAIN

Table 3. Summary of Major Resting

15O

PET Studies of Ageing.

Author

Sample Screening

N (sex ratio)

Age Range (years)

Eye/Ear

Equipment/ Technique

LebrunGrandie et al. (1983)22

s

19 (14M, 5F)

19–76

EC/EU

ECAT, measured attenuation. 15O Steady State Inhalation Res: n

Lenzi et al. (1981)24

sss

27 (NS)

Age not specified. Divided into 2 groups: Young < 50 (n=16) Old >50 (n=11)

NS/NS

Equipment NS 15O Steady State Inhalation Res: NS

Pantano et al. (1984)23

s

27 (19M, 8F)

19–76 Divided into 2 groups: Young: av age 38 Old : av age 63

EC/EU

ECAT II 15O Steady State Inhalation Res: n

Yamaguchi et al. (1986)30

sss

22 (17M, 5F)

26–64 Divided into 2 groups: Young: av age 35.7 Old: av age 57.6

EO/EU

HEADTOME III 15O Steady State Inhalation Res: n n n

Itoh et al. (1990)25

sss

28 50–85 (17M, 11F)

NS/NS

ECAT II 15O Steady State Inhalation Res: n

Leenders et al. (1990)26

sss

30 22–82 (18M, 16F)

EC/EO

ECAT II 15O Steady State Inhalation Res: n

Burns & Tyrrell (1992)101

sss

14 (6M, 8F)

51–85

NS/NS

ECAT/931/08/12 15O Steady State Inhalation Res: NS

Takada et al. (1992)31

sss

32 27–67 (15M, 17F)

EC/EU

Equipment: NS 15O Steady State Inhalation Res: NS

Martin et al. (1991)32

sss

30 30–85 (15M, 15F)

EC/EU

ECAT/931/08/12 15O Steady State Inhalation Res: n n n

u

u

uu

u

u

uu

uu

uuu

uu

111

CORRELATES OF BRAIN AGEING

Method of Analysis Procedure

Results

Circular ROI’s 10mm diamater manually placed over CBF image at point of highest CBF in lobe.

Selected rCBF decrease with age temporosylvian, medial frontal, medial occipital regions. No correlation between regional CMRO2 or OER and age.

NS

rCBF & CMRO2 decline with age, most pronounced in visual cortex & insula. Increased OER with age.

Circular ROI’s traced on CBF image and copied onto CMRO2 & OEF images Normalised to mean of all ROI’s

rCBF and CMRO2 declined by 18% and 17% respectively in grey matter. Largest decrease in frontal, parieto-occipital and temporosylvian regions. Non-significant increase in OEF in grey matter. No difference in white matter between 2 groups.

Circular ROI’s placed with CT superimposed for guidance

Mean left hemisphere CMRO2 significantly lower in older age group. Significant correlation between CMRO2 and age demonstrated. No correlation between age and rCBF, or OEF.

ROI measuring 3 x 7 pixels CT brain used to calculate cerebrocranial index (CCI)

CCI decreased linearly with age (Correlation –0.23) CBF and CMRO2 unchanged with age and not influenced by CCI

Geometric ROI (rectangular for cortical regions, circular for other regions), individually placed.

Decrease in CMRO2, CBF and CBV with age (approx 0.5% per year), both in grey and white matter. Mostly non-significant increase in OER with age.

Stereotactic co-ordinates based on Talairach atlas used to obtain CMRO2 values

Decrease in CMRO2 in parietal lobe. In other regions, this failed to reach significance.

ROI, method unstated

Decrease in mean CMRO2 with age and in specific ROI’s (bilateral putame, left supratemporal, left infrafrontal and left parietal corticies). Decreased CBF in left superior temporal cortex only.

SPM

Mean CBF unchanged with age. RegionaL decrease in CBF in cingulate, parahippocampal gyri, superior temporal gyri, medial frontal gyri, parietal cortex bilaterally and in left insular and inferior frontal gyrus. Table 3 continues

112

THE AGEING BRAIN

Table 3. Continued. Author

Sample Screening

N (sex ratio)

Age Range (years)

Eye/Ear

Equipment/ Technique

Marchal et al. (1992)29

sss

25 (NS)

20–68

EC/EU

LETI TTV03 15O Steady State Inhalation Res n n n

Eustache et al. (1995)102

sss

25 20–68 (14M, 11F)

EC/EU

LETI TTV03 15O Steady State Inhalation Res: n n n

Bentourkia et al. (2000)27

sss

20 (13M, 7F)

21–75 Divided into 2 groups: Young: aged 21-36 Old: aged 55–75

EC/EU ECAT EXACT-HR Subjects Both 15O H2O and asked to 18FDG injection “avoid Res: n n n focussing their mind on anything”

Meltzer et al. (2000)28

sss

27 (9M, 18F)

19–76

EC/EU

uuu

uuu

uuu

uu

Siemens 951R/31 15O H O injection 2 Res: n n.

s, ss, sss sample drawn from increasingly optimal source (s= hospitalised subject; ss outpatient clinic; sss volunteer) u, uu, uuu increasingly rigorous screening (u history only, or unstated; uu history & physical examination +/- laboratory tests; uuu history, physical examination & CT/MRI brain or neuropsychological testing) Res: Spatial Resolution at Full-width half-maximum (FWHM) n = spatial resolution ≥ 15mm; n n = spatial resolution 8.6-14 mm; n n n = spatial resolution ≤ 8.5mm NS: not specified EO: eyes open; EC: eyes closed; EU: ears unplugged; EP: ears plugged

CORRELATES OF BRAIN AGEING

113

Method of Analysis Procedure

Results

ROI, circular, 14 pixel diameter CT scans given atropy rating on 4 point scale

Decrease in CMRO2 whole cortex and in multiple cortical gyri (24/31) with age (approx –6% per decade). Effect on whole cortex independent of cerebral atrophy. Decrease in CBF in 10/31 cortical gyri with age. No change in OEF with age. No decrease in CMRO2 or CBF in white matter or deep grey matter structures.

CT & PET images co-registered (method undefined). Circular ROI 14mm diameter manually placed. Normalisation of ROI values to cerebellum.

Global decline in CMRO2 with age. rCMRO2 values negatively correlated with age in all neocortical regions and left thalamus.

MRI and PET images anatomically matched by visual interpolation ROI’s manually drawn on MRI and adjusted for FDG study. Transferred onto 15O study

Global decline in CBF with age (0.37% per year) Decline in CBF was greatest in frontal regions and least in Occipital cortex. Preserved coupling of rCBF and rCMRGlu with age

MRI & PET coregistered using using automated computerised algorithm. ROI’s traced on MRI & transferred to PET. MRI based partial volume correction method.

Prior to partial volume correction: negative correlation of mean CBF with age. Regional statistically significant results in medial orbitofrontal, lateral orbitofrontal, lateral temporal regions. After partial volume correction: mean CBF no longer significantly correlated with age. Loss of regionally significant results except for medial orbitofrontal region.

114

THE AGEING BRAIN

subjects (mean age 38 years) and older subjects (mean age 63). A non-significant increase of 0.35% per year in grey matter OEF was observed by Leenders et al.26 The reported increase in OER by Lenzi et al.24 was modest and not examined for statistical significance. In general, these results suggest that OER may increase slightly with age, and may simply be a reflection of reciprocal decreases in CBF. Regional changes in CMRO2 and CBF with age have been examined in a number of studies. An obvious discrepancy between white matter regions and grey matter regions has emerged, with white matter CMRO2 and CBF being largely unaffected by age.22,23,29,30 The most consistent regional effect of age in cortical grey matter is that of decline in CBF and CMRO2 in selected frontal and temporal regions.22,23,27,29,31,32 Reports of regional declines in cortical CBF or CMRO2 have been predominantly bilateral, although some studies have noted a left-sided31,32 or right-sided emphasis29 for particular regions. An asymmetric left hemispheric decline in CMRO2 with age was noted in one study.30 Resting FDG studies of ageing Evaluation of the effect of ageing on global CMRglu has been undertaken in a number of studies, with conflicting results (see Table 4). In an early study in which the assessment methods for subject selection were not detailed, Kuhl et al.,33 reported a decline in global CMRglu of 0.43% per year. A number of studies have subsequently reported a global decrease in CMRglu ranging from 0.21%34 to 0.6% per year.35 However, the reported decline in CMRglu is uncertain. The screening of aged subjects appears to have been inadequate in a number of studies in which global declines in CMRglu has been described.33,34 Although most studies excluded those with established hypertension, other cerebrovascular risk factors were often not assessed. Several studies have used either CT scan36,37 or MRI scan27 as part of the screening process, but this represents the exception rather than the rule. The effect of cerebral atrophy on global CMRglu has been evaluated in a small number of studies. Schlageter et al.37 employed an automated segmentation technique to measure CSF volume on CT scan of the brains of their subjects. Cerebral atrophy so measured was negatively correlated with CMRglu but only explained 13% of the variance. No age effect on CMRglu was apparent in this study. Another study initially revealed a decline in CMRglu with age, which was no longer statistically significant after the effect of cerebral atrophy (as measured by semiquantitative ratings of MRI scans) was partialled out38. In summary, there are methodologically sound studies demonstrating both decline in CMRglu with age35,39 and no change with age.36,37 The variability of these findings indicates that perhaps a modest effect of age on global CMRglu is being inconsistently determined as a result of methodological and statistical differences between studies. The comparison of FDG studies examining regional changes in CMRglu with age is difficult owing to the different methods used to examine regional

CORRELATES OF BRAIN AGEING

115

effects. The majority of studies have employed ROI analysis. The spectrum of ROI methods employed in these studies include: the poorly standardised procedure of tracing irregular regions directly onto the FDG study;40 the use of the subject’s CT scan of the brain36,37 or an atlas41 as an anatomical guide for ROI definition; the placing of standard geometric ROIs38,39,42 directly onto the PET image and the tracing of regions onto individual’s MRI scan of the brain and transferring onto the PET image.27 The recent use of standardised packages such as SPM has enabled exploratory comparison across the whole brain and enables easy comparison of results across studies using reference to Talairach co-ordinates. Despite the difference in methods of analysis for regional effects, only a small number of studies have failed to find a significant regional effect of age on CMRglu.36,40–42 A number of factors including poor resolution of the older scanners and less sophisticated ROI analyses may explain many of these negative findings. A distinct pattern of age related CMRglu decline has emerged which is reminiscent of CBF and CMRO2 studies in ageing. The most consistent pattern of reduction in CMRglu with age is that of frontal reduction27,33,34,35,38,43–45 including that of anterior cingulate. 35,43 Other regions showing reduction in CMRglu with age include specific anterior, posterior and lateral temporal regions35,39,43,45 and parietal cortex.33,34,39,45 The effect of ageing on deep grey matter nuclei is occasionally reported as being a decline, and in the study by Bentourkia et al., 27 the greatest CMRglu reduction was observed in the right striatum. The effect of ageing on CMRglu in the cerebellum, occipital cortex and white matter appears to be minimal. Methodological Issues in Resting Studies Defining healthy ageing One of the key difficulties encountered in any study of the effects of ageing is that of the confounding effects of pre-existing disease. Reducing the likelihood that such confounding factors will influence results requires adoption of a narrow definition of “normal ageing”, thereby excluding those with risk factors for cerebral disease. Although for the sake of sample uniformity conservative inclusion criteria may be desirable, there is a risk that findings from any such study will reflect those of “elite ageing” rather than “normal ageing”. Such results are less likely to be generalisable to the majority of the aged population. Until more is known about the impact of common agerelated conditions on functional imaging findings, it would seem prudent to select a conservative sample to minimise potential confounding effects of systemic disorders on the results. The adequacy of the screening process prior to enrolment of subjects can be questioned in many of the studies reviewed in Tables 1–4. Appropriate clinician review including detailed history, physical examination, laboratory evaluation and neuropsychological assessment,

116

THE AGEING BRAIN

Table 4. Summary of Major Resting FDG Studies of Ageing. Author

Sample Screening

N (sex ratio)

Kuhl et al. (1982)33

sss

Eye/Ear

Equipment/ Technique

40 18–78 (17M, 23F)

EO/EU

ECAT II Res: n

De Leon et al. (1984)42

NS

37 (NS)

Divided into 2 groups: Young: av age 26.1 Old: av age 66.6

EC/EU

PET III Res: n

Hawkins et al.

NS

8 (7M, 1F)

18–68

NS/NS

NeuroECAT Res: n n

Duara et al (1983)41

sss

21M

21–83

EC/EP

ECAT II Res: n

Duara et al. (1984)36

sss

40M

21–83

EC/EP

ECAT II Res: n

Horwitz et al. (1986)103

sss

30M

Divided into 2 groups Young: 20–32 Old: 64–83

EC/EP

ECAT II Res: n

Schlageter et al. (1987)37

sss

49M

21–83

EC/EP

ECAT II Res: n

Yoshii et al. (1988)38

sss

76 21–84 (39M, 37F)

EC/EU

PETT V Res: n

Hoffman et al. (1988)45

sss

36 21–74 (22M, 14F)

NS/NS

NS Res: NS

NS

NS

Age Range (years)

uu

uuu

NS

uuu

uu

uu

117

CORRELATES OF BRAIN AGEING

Method of Analysis Procedure

Results

Not stated

Global decline in rCMRGlu with age (0.43% per year). Regional decline in rCMRGlu evidenced by reduction in metabolic ration of superior frontal cortex to parietal cortex.

Fudicial markers on PET and CT brain used to match CT and PET slices. Geometric ROI’s placed on CT and transferred to PET image.

No difference in global or regional rCMRGlu between young and elderly subjects.

Irregular individual ROI’s Method unspecified.

No change in rCMRGlu with age.

Individual ROI’s defined on PET using atlas guide.

No significant change in global or regional rCMRGlu for white or grey matter with age.

Individual ROI’s defined on PET using atlas and patient’s CT as a guide.

Study was an extension of above study (Duara et al 1983). Replicated above results.

Individual ROI’s defined on PET image with reference to brain atlas. Partial correlation coefficiants determined between each pair of 59 regions.

No change in mean CMRGlu between groups. Elderly subjects demonstrated a reduced number of significant correlations between pairs of regions noted. This reduction was especially noted in frontal-parietal and parietal-parietal correlations.

Individual ROI’s defined on PET using individual’s CT as a guide. CSF volume measured using automated segmentation on CT scan.

CSF volume, ie cerebral atrophy was negatively correlated with CMRGlu but accounted for no more than 13% of variance. No effect of age on CMRGlu, even after correction for cerebral atrophy.

Geometric ROI’s placed automatically over PET image and manually adjusted to overlay specific regions. MRI scans of 58 subjects rated semiquantiatively for atrophy.

Divided into those with or without 1 or more risk factors for thromboembolic stroke. Significant lower mean CMRGlu in aged subjects compared to young. Age effect non-significant when effects of cerebral volume and atrophy partialed out. Atrophy accounted for 8.3% of variance of mean CMRGlu. CMRGlu not affected by presence of risk factors for stroke.

ROI analysis, method unspecified.

RCMRGlu reduction with age in gyrus rectus, orbital gyri, inferior frontal gyri, medial prefrontal cortex, insula, superior parietal lobule & globus pallidus. Table 4 continues

118

THE AGEING BRAIN

Table 4. Continued. Salmon et al. (1991)104

sss

DeSanti et al. (1995)105

sss

Moeller et al. (1996)34

sss

25 (NS)

Divided into 2 groups: Young: av age 25.8 Old: av age 60.1

NS/NS

NeuroEcat Res: n n

72M

Divided into 2 groups: Young: av age 27.5 Old: av age 67.6

EO/EU

Siemens CTI-931 Res: n n n

uu

uu

NS

Group 1: Group 1: 21–90 130 (62M, 68F) Group 2: Group 2: 24–77 20 (10M, 10F)

Murphy et al. (1996)39

Group 1: Group 1: EC/EP Scanditronix PC 1024-7B Res: n n n Group 2: Group 2: EO/EU Scanditronix Superpett 3000 Res: n n n

120 21–91 (55M, 65F)

EC/EP

Scanditronix PC 1024 7B Res: n n n

Petitsss Taboué et al. (1998)35 uuu

24 (15M, 9F)

EC/EP

LETI TTV03 Res: n n n

Garraux et al. (1999)43

sss

EC/EU

uu

43 19–75 (25M, 18F) Divided into 2 groups: Young: 19–28 Old: 47-75

Siemens CTI 951 R 16/31 Res: n n n

Bentourkia et al. (2000)27

sss

20 (13M, 7F)

EC/EU ECAT EXACT-HR Subjects Res: n n n asked to “avoid focussing their mind on anything”

sss uu

uuu

20–70

21–75 Divided into 2 groups: Young 21–36 Old 55-75

s, ss, sss sample drawn from increasingly optimal source (s= hospitalised subject; ss outpatient clinic; sss volunteer) u, uu, uuu increasingly rigorous screening (u history only, or unstated; uu history & physical examination +/- laboratory tests;

CORRELATES OF BRAIN AGEING

119

ROI analysis, method unspecified. Absolute values & ratios normalised to mean cortical values both examined.

Global CMRGlu unchanged with age. Absolute rCMRGlu values unchanged with age. Normalised rCMRGlu reduced in frontal lobes and increased in cerebellum with age.

Manually traced ROIs, some anatomical & some geometric.

Absolute rCMRGlu values reduced with age in frontal and temporal lobes. Dorsolateral frontal region demonstrated stronger relationship with age than orbitofrontal region, decreasing 2.6 and 2.2umoles/100g/min per decade respectively.

ROI method not described. Used scaled Subprofle Model (SSM) to examine regional covariation with age. Group 2:

Both groups, global CMRGlu decreased significantly with age (0.21% per year). Regional declines in frontal regions observed in group 1 only. In group 1, relative decrease in frontal rCMRGlu was associated with covariate relative increases in parietooccipital association areas, basal ganglia, brainstem and cerebellum.

Circular ROI template superimposed on each patients PET.

Mean CMRGlu decreased with age. Regional CMRGlu declined in frontal, temporal and parietal ROI’s, with assymetry of this decline noted (parietal L>R decline, frontal R>L decline). L>R assymetry in frontal rCMRGlu was significantly less in women compared with men.

SPM analysis.

Global decline in rCMRGlu with age (6% per decade). Regional decline in rCMRGlu in most areas except occipital cortex and right cerebellum. Most marked age related decline seen bilaterally in perisylvian temporoparietal and anterior temporal regions, insula, inferior and postero-lateral frontal region, anterior cingulate, head of caudate, anterior thalamus.

SPM analysis.

Frontal CMRGlu decreased in elderly subjects in bilateral dorsolateral prefrontal and medial prefrontal areas including anterior cingulate, left lateral premotor area, Broca’s area and left insular, right superior temporal gyrus.

MRI and PET images anatomically matched by visual interpolation. ROI’s manually drawn on MRI and adjusted for FDG study. Transferred onto 15O study.

Global decline in rCMRGlc (0.34% per year). Decline in rCMRGlc greatest in right striatum and least in cerebellum. Preserved coupling of rCBF and rCMRGlc with age.

uuu history, physical examination & CT/MRI brain or neuropsychological testing). Res: Spatial Resolution at Full-width half-maximum (FWHM) n = spatial resolution ≥ 15mm; n n = spatial resolution 8.6-14 mm; n n n = spatial resolution ≤ 8.5mm NS: not specified. EO: eyes open; EC: eyes closed; EU: ears unplugged; EP: ears plugged.

120

THE AGEING BRAIN

is required to minimise the chance of inclusion of subjects with disorders affecting cerebral function. The resting state The majority of SPECT and PET studies of ageing rely on evaluation of “resting” rCBF or rCMRglu. The notion of what represents an adequate “resting state” has varied significantly across studies. The common measures of ensuring reduced sensory input by plugging ears and patching eyes are not always adopted. Limited attempt has been made to reduce the effects of cognitive processing. Mental processing occurring during scanning may vary substantially between subjects and is likely to be influenced by factors such as anxiety provoked by the procedure, mood at time of scanning, etc. This uncontrolled mental activity may in turn significantly influence rCBF or rCMRglu. An approach to potentially overcome this confound and minimise inter-subject variability is the introduction of a standardised cognitive task. Accounting for cerebral atrophy The comparison of ageing and youthful brains brings with it the challenge of taking account of the effects of structural changes of ageing on functional data. The most important issue is the effect of brain atrophy. As the resolution of both PET and SPECT fall significantly short of structural imaging techniques such as MRI, functional data will be influenced by the effect of atrophy due to partial-volume averaging. If such an effect is not taken into account, functional data can considerably underestimate rCBF and rCMRglu in the elderly, resulting in misinterpretation of results. This effect may account for some of the positive findings regarding reductions in rCBF and rGMRgl in aged compared to young subjects. There have been several different methods used to attempt to correct for this source of error. Simple techniques include correction using an atrophy score derived from visual ratings of structural scans, correction by using global measures of atrophy such as VBR and more complicated correction by using a MRI-based segmentation of images and generation of regional correction factors. Meltzer et al.28 attempted to address this issue with surprising results. Their study used manually traced regions of interest on MRI that were transferred to PET images using an automated image registration algorithm. An MRI-based partial volume correction coefficient was applied to the CBF data. This correction coefficient was derived for individual regions after segmentation of the MRI into brain and cerebrospinal fluid, creation of a binary data set and smoothing of MRI data to approximate resolution of the PET scan. After corrections were applied, most age-associated changes in cerebral blood flow failed to reach significance. The exception was that CBF continued to demonstrate a statistically significant negative correlation with age in the mesial orbitofrontal region. Such a finding suggests that previous studies may have overestimated age-related reduction in CBF and CMRO2. However, as prefrontal atrophy is about twice that found in the temporal or parietal neo-

CORRELATES OF BRAIN AGEING

121

cortex44 atrophy effects are unlikely to account for all age effects on CBF or CMRO2 noted in previous studies, uncorrected for atrophy or partial volume effects. These findings challenge the notion of a large age-specific decrease in brain blood flow and metabolism, and underscore the importance of correction for possible confounding factors. Summary of resting SPECT and PET studies Methodological discrepancies between studies limit the conclusions that can be drawn from this body of literature. However, a general pattern is appreciable which both SPECT and PET studies of CBF and PET studies of CMRglu share. With respect to global effects of the ageing process, a modest effect is demonstrable in some studies. Regional effects appear to mirror known patterns of age-related pathological change, including regional effects of atrophy. Although atrophy may partially account for regional CBF, CMRO2 and CMRglu effects observed with age, it does not appear to be the sole arbiter of these changes. Regional effects demonstrated by resting functional imaging studies add support to theories of cognitive ageing which espouse decline in frontal lobe function as a mediator of age-related change. Resting PET and SPECT Studies in “At Risk” Groups The major focus so far has been on defining the changes associated with “healthy ageing”. The rigorous screening of subjects and exclusion of those with risk factors for conditions such as cerebrovascular disease means that such studies may be poorly representative of the general population. A limited but expanding literature is exploring influence of age-associated phenomena such as cerebrovascular disease and mild cognitive impairment on functional imaging parameters. Vascular risk factors A few studies of healthy ageing have evaluated the effects of cerebrovascular risk factors on CBF. Measuring with the nitrous oxide technique, earlier studies46 failed to demonstrate an effect of hypertension alone on CBF or CMRO2. However, in the presence of systemic arteriosclerosis or frank cerebrovascular disease, both parameters declined and cerebrovascular resistance increased. Using 133Xe inhalation SPECT, Naritomi et al.15 divided their elderly group into those with and those without risk factors for stroke. Although an agerelated effect was observed on rCBF, the presence of cerebrovascular risk factors did not alter rCBF. In an FDG PET study of ageing, Yoshii et al.38 divided their sample of 76 subjects into those with and without risk factors for thrombo-embolic stroke. No appreciable effect on CMRglu was noted in the presence of stroke risk factors. In a study of 60 aged individuals ranging from 65 to 84 years, Claus et al.47 evaluated the effect of cerebrovascular risk factors, quantified indicators of atherosclerosis and cerebral atrophy on

122

THE AGEING BRAIN

CBF as measured by 99mTc-HMPAO. An age-related decline in rCBF in temporal and parietal cortex was observed with age, and this relationship was preserved after adjustment for the effect of cerebrovascular risk factors. The relationship between age and rCBF disappeared after correction for fibrinogen levels and measures of carotid atherosclerosis. The authors proposed that age-related declines in CBF might be related to atherosclerosis rather than to vascular risk factors per se. Hypertension 133Xe inhalation was used in a 36-month prospective study evaluating the effect of antihypertensive treatment on CBF in 12 individuals with mild hypertension at baseline.48 As a group, an overall increase in CBF was demonstrable at 6, 12 and 24 months after initiation of antihypertensive treatment. This difference was no longer appreciable by 36 months. Four subjects developed overt signs of cerebrovascular disease over the course of follow-up. The CBF values of these four individuals showed decline from the 24-month evaluation onward, whereas the asymptomatic group continued to demonstrate improved CBF values throughout the study period. In a cross-sectional study, Nobili et al.49 showed that hypertensives who were neurologically asymptomatic, especially when untreated, had focal or diffuse cerebral hypometabolism. More recently, utilising FDG PET, Salerno et al.50 compared a group of 17 elderly hypertensives compared with 25 age-matched non-hypertensive controls. In the hypertensive group, a significant reduction in FGD uptake was demonstrable in regions supplied by basal ganglia perforating arteries and at the middle cerebral/anterior cerebral artery watershed. Although data are limited, it appears that the presence of hypertension alone has a modest but appreciable influence on CBF and CMRglu. This effect may be regional, affecting areas most vulnerable to the effects of ischaemia such as long perforating vessels and “watershed” regions. White matter hyperintensities (WMHs) Hyperintensities on T2-weighted MRI brain imaging are common in the white matter of elderly individuals. The functional significance of these has been a matter of dispute and extensive investigation. In a study of 51 healthy individuals between 19 and 91 years, De Carli et al.51 found that when WMHs comprised >0.5% of intracranial volume, they were associated with cognitive deficits and reduced frontal lobe blood flow on PET scanning. In another study52 it was shown that cortical metabolic dysfunction was related to ischemic subcortical lesions, both lacunar infarcts and non-infarction WMHs, in patients with vascular dementia. Metabolism in the frontal cortex may be particularly dependent on pathologic alterations of subcortical nuclei. In a more recent study of 231 individuals53 without overt neurological disease, the most striking relationship was that observed between periventricular white matter hyperintensities (PVHs) measured by semiquantitative analysis and CMRglu measured by region of interest analysis. CMRglu values showed

CORRELATES OF BRAIN AGEING

123

a progressive reduction with increasing grades of PVHs. The relationship between CMRglu and severity of PVHs was significant for multiple cortical and subcortical regions. A less strong relationship was observed between CMRglu and deep white matter or basal ganglia hyperintensities. The relationships between MRI hyperintensities, cerebrovascular risk factors and functional imaging measures awaits further exploration, but promises to be important in understanding the determinants of age-related decline in CBF and CMRglu. Groups at risk of dementia Functional imaging techniques have been used to evaluate individuals who are at risk of age-related disease processes such as AD. The most informative findings have come from those studies in which asymptomatic subjects at risk of AD are compared with healthy age-matched controls. In a study of 24 asymptomatic first degree relatives from familial AD pedigrees, Kennedy et al.54 demonstrated deficits in global and regional CMRglu compared with age matched control subjects not at risk of AD. The regional pattern of CMRglu abnormality was similar to, but less severe than that of affected familial AD controls. In a PET study of cognitively normal subjects with family history of AD,55 the effect of apolipoprotein A (ApoE ε4) was determined by comparing 22 ApoE ε4-negative individuals with 11 subjects homozygous for ApoE ε4. Subjects in both groups performed equally well on neuropsychological measures. Global CMRglu was not different between groups. Those subjects who were homozygous for ApoE ε4 demonstrated reduced rCMRglu in posterior cingulate, parietal, temporal and prefrontal regions compared to ApoE ε4-negative controls. A further study from the same group56 evaluated 10 cognitively normal individuals who were heterozygous for ApoE ε4 and 15 ApoE ε4-negative individuals with a family history of AD. Subjects were followed longitudinally and underwent FDG PET at baseline and two-year follow-up. Cognitive deterioration was not observed over the study period. However, greater CMRglu declines were observed in the ApoE ε4 heterozygous subjects in temporal, posterior cingulate and prefrontal cortex as well as parahippocampal gyrus and thalamus, in comparison to ApoE ε4-negative subjects. Taken together, these preliminary studies of ApoE ε4 positive individuals suggest that functional imaging changes are seen which parallel those of AD rather than normal ageing. These results underscore the potential of ApoE status as a confounding factor in studies of normal ageing. The screening of aged individuals for ApoE status prior to inclusion in functional imaging studies of healthy ageing is therefore highly recommended. This topic is described in greater detailed in Chapter 15. Cognitive impairment Those with mild cognitive impairment (MCI) may be at risk of developing AD, with estimates of conversion varying from 10% to 15% per year.57 The degree to which functional imaging may assist in the prediction of those at

124

THE AGEING BRAIN

risk of conversion, has yet to be properly explored. In preliminary reports, significant temporo-parietal abnormalities have been noted with SPECT in those with MCI.58,59 In a PET study, Small et al.60 evaluated 12 cognitively normal subjects with age associated memory impairment (AAMI) and positive family history of AD, who were heterozygous for ApoE ε4 genotype and compared them to a similar group of 19 ApoE ε4-negative individuals. CMRglu was significantly lower in both right and left parietal regions in the ApoE ε4-positive group. In addition, left-right parietal asymmetry was significantly higher in those at risk of AD and who were ApoE ε4-positive, than those without ApoE ε4. These results suggest that functional imaging may be of future utility in identification of subjects at risk of later cognitive decline. However, at present, the positive predictive value of SPECT and PET in those with MCI or AAMI is questionable.59 Activation Studies in Ageing Introduction In healthy young subjects, motor, cognitive and pharmacological challenges have been shown to result in discrete changes in rCBF (using SPECT), rCMRglu or rCMRO2 (using PET) and BOLD signal using fMRI. These techniques are now being applied to ageing and age-associated dysfunction, with a particular focus on cognitive activation. A range of cognitive and non-cognitive tasks has proved suitable for probing functional integrity in aged subjects. Activation procedures have included motoric,61 photic,62 working memory,63 visual attention,64 frontal/executive,65 word identification,66 spatial orientation,67 visual encoding and retrieval68 and verbal encoding and retrieval69 tasks. Intuitively, many of these tasks have been chosen on the assumption that some age-dependent decline occurs within that particular cognitive domain. There are several aims of this exercise. Firstly, there is a possibility that the activation process may unmask changes not appreciable at rest. Secondly, there is a desire to study the functional neuroanatomy of the ageing brain, and to examine functional neuroimaging correlates of age-related neuropsychological changes. Thirdly, the studies aim to identify individuals at increased risk of age-related neuropsychiatric diseases. Potential advantages of activation tasks One advantage of introducing a cognitive task during scanning is the reduction in intra-subject variability of functional data. As mental activity and sensory stimulation have cerebral metabolic and blood flow correlates, standardisation of sensory, motor and cognitive activity via an activation procedure will minimise these potential confounds. The influence of these variables over reliability of cerebral metabolic measurements has been demonstrated by Duara et al.70 who re-scanned nine normal subjects at rest and seven normal subjects during an activation task. Resting scans were performed with eyes

CORRELATES OF BRAIN AGEING

125

closed and covered and ears open. Activation scans involved viewing a series of coloured pictures projected onto a screen, and depressing a foot pedal with the right or left foot, depending on whether the subject did, or did not, like a picture. Within subject variability for normal subjects for the repeated scans performed at rest was marked, with correlation coefficient for mean rCMRglu of 0.001, indicating virtually no correlation. However, in repeated activated scans in normal subjects, there was a correlation coefficient of 0.703. Metabolic ratios for specific regions also varied considerably between resting scans, but less so between activated scans. A study by Skolnick et al. 71 used 133Xe inhalation to measure rCBF in elderly normal subjects at rest and during verbal and spatial tasks during two scans, separated by an average of nine weeks. Global CBF was reduced in the repeated baseline scans, but this reduction was not evident in the repeated activated scans. Regional CBF was consistent between the two activated scans. These results suggest that measurement of CBF and CMRglu is reproducible and reliable during activated states and may be less so at rest. Motor and sensory stimulation In a simple reaction time task with BOLD fMRI, D’Esposito et al.72 compared 32 young subjects (mean age 22.9 years) and 20 elderly subjects (mean age 71.3 years). They found a reduced number of suprathreshold voxels and lower signal-to-noise ratio (SNR) in the elderly, but no significant group differences in the shape of the haemodynamic response. The authors suggested that an age effect on the haemodynamic coupling between neural activity and BOLD signal change may need to be taken into account in interpretation of age effects demonstrated by functional imaging techniques. Calautti et al.61 used 15O PET to investigate age effects in a cued thumb-to-index finger opposition task. An over-activation in the superior frontal cortex in aged relative to young subjects was noted, which, in the authors’ opinion, could not be fully explained by differences in resting CBF in this area between young and old subjects. In a BOLD contrast fMRI experiment, Ross et al.62 examined the response to photic stimulation in a group of 9 healthy elderly subjects (mean age 71 years) and healthy young subjects (mean age 24 years). The mean BOLD signal response to photic stimulation in the visual cortex was significantly reduced in aged subjects. There was a trend for elderly subjects, without atrophy, to demonstrate less profound reduction than those with atrophy. During presentation of checkerboard stimuli to 11 young subjects (mean age 23 years) and 11 elderly subjects (mean age 66 years), Huettel et al.73 noted an earlier peak in the haemodynamic response, smaller spatial extent of activation and lower SNR in the elderly subjects. The amplitude, form and refractory properties of the haemodynamic response were similar across groups, and the authors concluded that the smaller spatial extent of activation was attributable to the lower SNR in the elderly subjects. In a study of 12 young and 14 older subjects, Levine et al.74 examined CBF correlates during passive viewing of

126

THE AGEING BRAIN

black and white formed and unformed textures. Subjects were scanned in two states: during viewing of random (formed and unformed) textures or formed textures. Random versus formed texture viewing was associated with activation outside the ventral occipitotemporal pathway (predominantly in left anterior cingulate, medial, middle and superior frontal gyri) in the older subjects. Comparison with the younger group demonstrated reduced activation in the older subjects of left precuneus, middle temporal and posterior cingulate and of the right parahippocampal gyrus. Taken together, these results suggest an age-related change in processing of visual stimuli, which may represent a decrement in the efficiency of visual processing with age. Grady et al. have examined age-related changes in object and spatial visual processing in a series of 15O PET experiments. An initial study contrasted a face-matching task with a spatial location-matching task.75 Results from this study suggested that functional separation of dorsal (occipitoparietal) versus ventral (occipitotemporal) pathways subserving spatial relations and object discrimination respectively was apparent in both aged and young subjects. However, functional separation of these two systems appeared less distinct in aged individuals. A further study replicated these findings,76 and in addition demonstrated greater activation by older adults during face matching in bilateral dorsolateral prefrontal cortex, fusiform gyrus, inferior frontal gyrus and left insula as well as left middle temporal gyrus. During location matching, older subjects again activated a more widespread network in bilateral prefrontal cortex, bilateral fusiform gyri as well as left occipitotemporal cortex and inferior parietal cortex. Such changes were taken to represent reduced processing efficiency of prestriate occipital cortex with age and hence increased utilisation of additional networks in order to compensate this reduction in efficiency. Similar results were obtained in a 15O PET study evaluating the effects of age on selective and divided visual attention by Madden et al.64 During the divided attention task only, subjects activated a network of regions including occipitotemporal, occipitoparietal and prefrontal regions. Activation patterns in younger subjects were relatively greater in the occipitotemporal pathway and for the older subjects greater in prefrontal regions. A further study by Grady et al.77 evaluated the effect of age on a task of degraded and nondegraded face perception. Results similar to the earlier study76 were obtained for the non-degraded face matching. Analysis of the activation patterns from the degraded face-matching task revealed that different regions were positively correlated with task performance in the old compared with the young subjects. In the old subjects, activity in the posterior occipital cortex, thalamus and hippocampus showed positive correlation with task performance. In the younger subjects this correlation occurred in the fusiform gyrus, suggesting that brain networks subserving success in this task differed between young and old subjects. Using a short-term visual memory task in which subjects discriminated pairs of vertical sinusoidal gratings of differing spatial frequency, McIntosh

CORRELATES OF BRAIN AGEING

127

et al.78 determined age-related differences in activated networks. Older and younger subjects performed equally well on the task. Although a common pattern of activation was seen in many regions (occipital, temporal and inferior prefrontal cortex and caudate), older subjects were observed to activate additional distinct regions (medial temporal and dorsolateral prefrontal cortices). Activity in these additional regions was related to task performance in the older subjects, suggesting a role for these additionally activated networks in maintenance of performance. Working memory In the first report of ageing effects on visual working memory, Grady et al.63 used 15O PET to evaluate rCBF during a delayed match to sample task. Independent of age, a common pattern of activation was noted with delay, including increased rCBF in left anterior prefrontal and decreased rCBF in the ventral extrastriate cortex. Less activation was seen in the right ventrolateral prefrontal cortex, and greater activation was seen in left dorsolateral prefrontal cortex in the aged group. A subsequent 15O PET study utilising two verbal working memory tasks79 demonstrated similar networks of activation in older and younger subjects. However, younger subjects showed more right dorsolateral prefrontal activation during one task and older subjects demonstrated greater left dorsolateral prefrontal activation during another working memory task. Taken together, these two studies support the notion that increased activation observed in the older subjects is a reflection of compensatory strategies required to overcome cognitive inefficiency in working memory occurring with age. Complex tasks Age effects have been examined during more complex tasks such as card sorting. An 15O PET study by Nagahama et al.80 utilised a modified card sorting task. This study revealed age effects of reduced ability to activate left dorsolateral prefrontal cortex, left inferior parietal lobule, left striate and prestriate cortex, bilateral precuneus, left occipital cortex and left cerebellum. In aged subjects a negative correlation between perseverative errors and activation was noted in several of these regions including left dorsolateral prefrontal cortex. A subsequent 15O PET utilising the Wisconsin Card Sort Test (WCST) and Raven’s Progressive Matrices (RPM) for activation65 has demonstrated age specific reductions in activated networks. Age-related reduction in activation was seen in the dorsolateral prefrontal cortex with WCST and in the inferolateral temporal cortex with RPM. In addition, aged subjects activated areas that were normally suppressed in younger individuals. These areas included right parahippocampal gyrus with WCST and polar and medial portions of the prefrontal cortex in both WCST and RPM. Reduced activation in key areas normally subserving these more complex neuropsychological functions appears linked with decrements on task performance. However, it remains unclear whether enhanced activation in regions normally suppressed

128

THE AGEING BRAIN

by younger individuals represents use of alternative cognitive strategies or inefficiency of operating networks subserving these cognitive tasks.65 Madden et al.66 investigated the effect of age on a visual word identification task using 15O PET in 10 young and 10 older subjects. Activity representing passive encoding of letter strings was observed in left frontal, striate and inferior temporal cortex, with more prominent activations observed in frontal and temporal regions in younger subjects. Several decreases in rCBF were also observed, with these being greater in the older subjects in the right superior frontal region and greater in the younger subjects for the right anterior thalamus, right posterior cingulate and insula. The activity associated with the semantic retrieval component of the task (distinguishing word from pronounceable non-word) also revealed an age effect. More prominent left occipital activation was observed in younger subjects, and a more prominent deactivation was observed in left superior frontal, anterior cingulate and lateral aspect of inferior temporal gyrus. Memory tasks Encoding Correlates of age-related declines in encoding ability for visual material has been examined in a study by Grady et al.68 using 15O PET in 10 young subjects (mean age 25 and 10 older subjects (mean age 69). In this study, older individuals activated the left ventral temporal cortex with encoding of faces but failed to activate the network seen in younger individuals (anterior cingulate, left prefrontal cortex, left temporal cortex and right medial temporal lobe including hippocampus). The poor activation in older subjects was attributed to failure of the encoding process. In a more complex task, encoding of face/name pairs was examined in a small 15O PET study.81 Activation during encoding was seen in bilateral occipital association areas, extending into parietal lobes bilaterally. No activation was seen in the hippocampus on either side. Reduced rCBF was observed in right temporal, frontal and anterior cingulate regions and in the left superior temporal gyrus. No age effect was identified, although this may have been attributable to the small sample size. Studies examining encoding of verbal material have also shown some age-effects. During the encoding phase of a word pair task, Cabeza et al.69 demonstrated that younger subjects had higher activation in the left prefrontal and occipito-temporal regions than older subjects. Somewhat in contrast to this study are the activation effects noted by Madden et al.82 During an encoding task involving a living/non-living judgement of nouns, no significant activation was observed in the younger subjects relative to a baseline task. However, in the aged group, activation was observed in bilateral prefrontal cortex, left thalamus, fusiform gyrus and parahippocampal gyrus. Direct contrast between young and older subjects revealed a significant difference for thalamus only. Reductions during the encoding task were not seen in young

CORRELATES OF BRAIN AGEING

129

subjects but were seen in several regions in the elderly group, including left anterior cingulate and right inferior parietal lobule. Direct comparison of the two age groups failed to reveal significant differences in deactivation patterns. The demonstration of age-related contrasts in patterns of activation during encoding in this study stands in some contrast to the studies of Grady et al.68 and Cabeza et al.,69 and raises the possibility that age related differences in rCBF during encoding may vary across tasks rather than as a function of age per se.82 Retrieval A number of studies have examined the effects of age on functional imaging correlates of recall of verbal material. However, the comparability of results is limited by the widely differing paradigms employed. Cabeza et al.69 used a word pair task to study recognition and cued recall with 15O PET. Effects of these “retrieval” tasks revealed that whilst younger subjects primarily demonstrated a unilateral (right) sided effect for retrieval of verbal material, older subjects demonstrated a more bilateral pattern of frontal activation. Retrieval was examined in the noun recognition experiment undertaken by Madden et al.82 Young subjects appeared to activate a network including right prefrontal cortex, left middle frontal gyrus and left thalamus, whereas older subjects activated bilateral prefrontal cortex, inferior parietal lobule, left inferior temporal cortex and left cerebellum. Direct comparisons revealed statistically significant increase in activation for younger subjects in the thalamus only, and in prefrontal regions in the elderly. Deactivation was also examined and, although deactivated regions were different in young and older subjects, no statistically significant differences were apparent. Using a word stem completion paradigm, Backman et al.83 explored 15O PET correlates of a word stem completion task, incorporating baseline, priming and recall components. Activity attributed to the recall component of the task in both young and elderly included bilateral increases in prefrontal cortex and anterior cingulate. Somewhat unexpectedly, older subjects activated perirhinal cortex bilaterally. The two explanations posited for this medial temporal lobe activity were that either optimal cued recall in elderly subjects involved use of strategic search strategies not utilised by younger subjects or that older subjects were continuing to encode and consolidate the information even during the cued recall task. The medial temporal lobe activation nevertheless stands in some contrast to the majority of studies of recall in both aged and young subjects. In a large 18FDG study of 70 subjects, Hazlett et al.84 examined cerebral metabolic activity during performance of a serial verbal learning task, without reference to a resting condition. Good performance was associated with higher metabolic activity in the frontal lobes for in the younger subjects and in the occipital lobes for the older subjects. An age related decrement in cerebral metabolism in the frontal lobes was observed, which remained significant after correction for cerebral atrophy. The shift in metabolic activity away

130

THE AGEING BRAIN

from anterior patterns of activation in the elderly was taken to indicate reallocation of networks invoked by the task. Comparisons of these results with 15O PET studies are made difficult by the combination of encoding and recall components in this experiment. Finally, Cabeza et al.85 examined age effects on activation during a verbal retrieval task, evaluating both ability to recognise a previously studied word from a distractor (content recognition), and ability to determine which of two words had been presented more recently in the study list (recall of temporal order). For content recognition, activation was observed in ventromedial temporal regions in both groups, suggesting a limited effect of age on strategic retrieval processes. However, age effects included reduced activation of right prefrontal regions and enhanced activation of left prefrontal cortex. This effect was thought to represent a compensatory process enacting semantic processing in the elderly. An alternative explanation, that of increasing left prefrontal activity with increasing demands of the task, was not supported by performance data (which indicated equal performance across age groups) or by comparison of hard versus easy blocks included in the experiment. For the temporal order task, an increase was seen in the right prefrontal cortex of young, but not older subjects, an effect that may reflect general age effects on frontal function or regionally specific effects of age on temporal-order retrieval. Summary of activation studies A number of studies have evaluated the effect of age on cerebral activation. There is evidence, from studies employing simple motor and sensory paradigms, of an age-related decline in cerebral activation, an effect that may have its basis in dysfunction in pathways subserving sensory and motor functions. For cognitive processes, the effect of ageing on patterns of activation is partly dependent on the task under consideration. A unifying observation is that during many cognitive tasks, elderly subjects activate brain networks that are similar to those activated by young subjects. However, the extent of this activation is reduced in elderly subjects. With some exceptions, this statement can be made of tasks involving visual processing, short-term visual memory, visual and verbal working memory, verbal recall and complex cognitive tasks such as card sorting. Reduced ability to activate certain brain regions has also been noted as a feature of ageing, and is often seen in association with reduced performance of the task under consideration. Activation of a number of regions additional to those seen in younger subjects has also been observed. In encoding and recall tasks, a more bilateral pattern of prefrontal activation has generally been observed. Alterations in the pattern of prefrontal activation are of interest, given the propensity of these regions to structural change with age. Enhanced activation in prefrontal regions may be an attempt to compensate for reduced functionality of this brain region with age, but this is a simple explanation that requires more study. In particular, the relationship between prefrontal activation and

CORRELATES OF BRAIN AGEING

131

success during task performance requires further evaluation. Activation of additional networks outside the prefrontal cortex has also been observed. At times, this extra-frontal activation has been associated with maintenance of performance. This suggests that functional activation of additional regions may reflect engagement of different cognitive strategies, perhaps by way of compensation for age related inefficiencies in processing. In addition to age differences in activation, an age effect on deactivation of various regions has been observed, with older subjects showing less strong deactivation in usually deactivated networks, along with additional deactivation in other areas. Deactivation may be a way in which optimisation of cognitive performance occurs in health, thus suggesting inability to streamline performance with age. A more detailed analysis of deactivation with respect to task performance will allow the potential implication of this age-related change in deactivation pattern to be elucidated. Magnetic Resonance Spectroscopy (MRS) Much of the work related to the effect of ageing on MRS-defined brain metabolites has focused on the neurodevelopmental period 86,87 or young adulthood.87 The studies of the elderly are limited by: small numbers, crosssectional design, limited sampling of brain tissue, different methods and discrepant findings that warrant further studies before definitive conclusions can be made. The relative concentration of N-acetylaspartate (NAA), which is the major marker of neurones, has been the focus of some investigations. Results have been variable, with reports of a reduction in NAA with age in the basal ganglia88 and the grey and white matter89–91 in some, but not all92,93 earlier studies. More recent studies94–96 have reported no reduction in NAA in the elderly, although this is still an inconsistent finding.97 These discrepancies could be related to methodological differences and the age range studied. Since MRS does not yield absolute values, metabolites are quantified in reference to an internal standard. The commonly used standard is total creatine (Cr), but this is not invariant with age, thereby reducing the value of the NAA/Cr ratio as a measure of age-related change. The other reference used is MRI-visible water content, which accounts for >95% of tissue water in the brain.98 While one study reported no change in the water content of the brain with ageing,98 another study reported a significant reduction in brain water with ageing.94 Total creatine (Cr), a marker of the energetics of neurones and glia, has been reported to decrease with age in the basal ganglia88 and increase with age in the frontal white matter,93 frontal gray matter,94 parietal white matter96 and both grey and white,95 but some studies have reported no change.93,96 Choline-containing compounds (Cho) have been shown to increase with age in some studies,89,94,95 decrease according to others88,93,97 and remain

132

THE AGEING BRAIN

unchanged in some studies.90,96 The heterogeneity of the above findings suggests that this field is still in its infancy and much work remains to be done. The potential of MRS has therefore not been fully exploited in this field. MRS studies in the healthy elderly are important if a normative database is to be developed for MRS to be used as a diagnostic investigation for neuropsychiatric disorders in the elderly, in particular dementia. Conclusions The increasing sophistication of functional imaging studies, particularly those involving activation techniques, is extending our understanding of how the brain is affected by the ageing process. Although many methodological pitfalls exist, this body of literature has enabled an appreciation of more subtle age effects than were initially thought to exist. At rest, age results in small, regionally specific declines in CBF and CMRglu, maximal in the frontal regions, which mirrors structural and neuropsychological alterations observed with advancing years. The differing patterns of prefrontal activation observed with age across a variety of cognitive activation tasks further highlights a key role for the frontal lobes in mediation of the age related cognitive changes seen in health. Specific but contrasting changes in functional integrity are beginning to be demonstrated in association with risk factors for age-related diseases such as AD and cerebrovascular disease. There are many areas in which functional imaging techniques may contribute to the further unravelling of secrets of the ageing brain. A potential role exists for these technologies in assessing the effects of interventions designed to modify or delay the ageing process. Early identification of those at risk of age-related neuropsychiatric disease might also be a realistic future role for these tools. References 1. 2.

3. 4. 5. 6.

Baltes P, Staudinger U, Lindenberger U. Lifespan psychology: Theory and application to intellectual functioning. Annu Rev Psych. 1999; 50:471–507. Raz N, Gunning-Dixon F, Head D, Dupuis J, McQuain J, Briggs SD, Loken WJ, Thornton AE, Acker JD. Selective aging of human cerebral cortex observed in vivo: Differential vulnerability of the prefrontal gray matter. Cereb Cortex. 1997; 7:268–282. Raz N, Gunning-Dixon F, Head D, Dupuis J, Acker J. Neuroanatomical correlates of cognitive aging: Evidence from structural MRI. Neuropsychology. 1998; 12: 95–114. O’Donnell K, Rapp P, Hof P. Preservation of prefrontal cortical volume in behaviorally characterized aged macaque monkeys. Exp Neurol. 1999; 160:300–310. Schuckit MA. An introduction and overview to clinical applications of neuroSPECT in psychiatry. J Clin Psychiat. 1992; 53 Suppl:3-6. Kety SS, Schmidt CF. Nitrous oxide method for the quantitative determination of cerebral blood flow in man: Theory, procedure and normal values. J Clin Invest. 1948; 27:476–483.

CORRELATES OF BRAIN AGEING

133

7. Kety SS. Human cerebral blood flow and oxygen consumption as related to aging. J Chron Dis. 1956; 3(5):478–486. 8. Melamed E, Lavy S, Bentin S, Cooper G, Rinot Y. Reduction in regional cerebral blood flow during normal aging in man. Stroke. 1980; 11:31–35. 9. Yamaguchi T, Hatazawa J, Kubota K, Abe Y, Fujiwara T, Matsuzawa T. Correlations between regional cerebral blood flow and age-related brain atrophy: a quantitative study with computed tomography and the xenon-133 inhalation method. J Amer Geriatr Soc. 1983; 31:412–416. 10. Matsuda H, Maeda T, Yamada M, Gui LX, Tonami N, Hisada K. Age-matched normal values and topographic maps for regional cerebral blood flow measurements by Xe-133 inhalation. Stroke. 1984; 15(2):336–342. 11. Takeda S, Matsuzawa T, Matsui H. Age-related changes in regional cerebral blood flow and brain volume in healthy subjects. J Amer Geriatr Soc. 1988; 36(4): 293–297. 12. Tsuda Y, Hartmann A. Changes in hyperfrontality of cerebral blood flow and carbon dioxide reactivity with age. Stroke. 1989; 20:1667–1673. 13. Hagstadius S, Risberg J. Regional cerebral blood flow characteristics and variations with age in resting normal subjects. Brain Cognition. 1989; 10:28–43. 14. Iwata K, Harano H. Regional cerebral blood flow changes in aging. Acta Radiol — Suppl. 1986; 369:440–443. 15. Naritomi H, Meyer JS, Sakai F, Yamaguchi F, Shaw T. Effects of advancing age on regional cerebral blood flow. Studies in normal subjects and subjects with risk factors for atherothrombotic stroke. Arch Neurol. 1979; 36:410–416. 16. Zemcov A, Barclay L, Blass JP. Regional decline of cerebral blood flow with age in cognitively intact subjects. Neurobiol Aging. 1984; 5:1–6. 17. Krausz Y, Bonne O, Gorfine M, Karger H, Lerer B, Chisin R. Age-related changes in brain perfusion of normal subjects detected by 99mTc-HMPAO SPECT. Neuroradiology. 1998; 40:428–434. 18. Waldemar G, Hasselbalch SG, Andersen AR, Delecluse F, Petersen P, Johnsen A, Paulson OB. 99mTc-d,l-HMPAO and SPECT of the brain in normal aging. J Cerebr Blood F Me. 1991; 11:508–521. 19. Catafau AM, Lomena FJ, Pavia J, Parellada E, Bernardo M, Setoain J, Tolosa E. Regional cerebral blood flow pattern in normal young and aged volunteers: A 99mTc-HMPAO SPET study. Eur J Nucl Med. 1996; 23:1329–1337. 20. Swartz JR, Lesser IM, Boone KB, Miller BL, Mena I. Cerebral blood flow changes in normal aging — SPECT Measurements. Int J Geriatr Psych. 1995; 10:437– 446. 21. Mozley PD, Sadek AM, Alavi A, Gur RC, Muenz LR, Bunow BJ, Kim HJ, Stecker MH, Jolles P, Newberg A. Effects of aging on the cerebral distribution of technetium-99m hexamethylpropylene amine oxime in healthy humans. Eur J Nucl Med. 1997; 24:754–761. 22. Lebrun-Grandie P, Baron JC, Soussaline F, Loch’h C, Sastre J, Bousser MG. Coupling between regional blood flow and oxygen utilization in the normal human brain. A study with positron tomography and oxygen 15. Arch Neurol. 1983; 40: 230–236. 23. Pantano P, Baron JC, Lebrun-Grandie P, Duquesnoy N, Bousser MG, Comar D. Regional cerebral blood flow and oxygen consumption in human aging. Stroke. 1984; 15:635–641. 24. Lenzi GL, Frackowiak RS, Jones T, Heather JD, Lammertsma AA, Rhodes CG, Pozzilli C. CMRO2 and CBF by the oxygen-15 inhalation technique. Results in normal volunteers and cerebrovascular patients. Eur Neurol. 1981; 20:285– 290. 25. Itoh M, Hatazawa J, Miyazawa H, Matsui H, Meguro K, Yanai K, Kubota K, Watanuki S, Ido T, Matsuzawa T. Stability of cerebral blood flow and oxygen metabolism during normal aging. Gerontology. 1990; 36:43–48.

134

THE AGEING BRAIN

26. Leenders KL, Perani D, Lammertsma AA, Heather JD, Buckingham P, Healy MJ, Gibbs JM, Wise RJ, Hatazawa J, Herold S. Cerebral blood flow, blood volume and oxygen utilization. Normal values and effect of age. Brain. 1990; 113 (Pt 1): 27–47. 27. Bentourkia M, Bol A, Ivanoiu A, Labar D, Sibomana M, Coppens A, Michel C, Cosnard G, De Volder AG. Comparison of regional cerebral blood flow and glucose metabolism in the normal brain: effect of aging. J Neurol Sci. 2000; 181: 19–28. 28. Meltzer CC, Cantwell MN, Greer PJ, Ben Eliezer D, Smith G, Frank G, Kaye WH, Houck PR, Price JC. Does cerebral blood flow decline in healthy aging? A PET study with partial-volume correction. J Nucl Med. 2000; 41):1842–1848. 29. Marchal G, Rioux P, Petit-Taboue MC, Sette G, Travere JM, Le Poec C, Courtheoux P, Derlon JM, Baron JC. Regional cerebral oxygen consumption, blood flow, and blood volume in healthy human aging. Arch Neurol. 1992; 49:1013–1020. 30. Yamaguchi T, Kanno I, Uemura K, Shishido F, Inugami A, Ogawa T, Murakami M, Suzuki K. Reduction in regional cerebral metabolic rate of oxygen during human aging. Stroke. 1986; 17:1220–1228. 31. Takada H, Nagata K, Hirata Y, Satoh Y, Watahiki Y, Sugawara J, Yokoyama E, Kondoh Y, Sishido F, Inugami A. Age-related decline of cerebral oxygen metabolism in normal population detected with positron emission tomography. Neurol Res. 1992; 14 (2 Suppl):128–131. 32. Martin AJ, Friston KJ, Colebatch JG, Frackowiak RS. Decreases in regional cerebral blood flow with normal aging. J Cerebr Blood F Met. 1991; 11:684–689. 33. Kuhl DE, Metter EJ, Riege WH, Phelps ME. Effects of human aging on patterns of local cerebral glucose utilization determined by the [18F]fluorodeoxyglucose method. J Cerebr Blood F Met. 1982; 2:163–171. 34. Moeller JR, Ishikawa T, Dhawan V, Spetsieris P, Mandel F, Alexander GE, Grady C, Pietrini P, Eidelberg D. The metabolic topography of normal aging. J Cerebr Blood F Met. 1996; 16:385–398. 35. Petit-Taboue MC, Landeau B, Desson J, Desranges B, Baron J. Effects of healthy aging on the regional cerebral metabolic rate of glucose assessed with statistical parametric mapping. Neuroimage. 1998; 7:176–184. 36. Duara R, Grady C, Haxby J, Ingvar D, Sokoloff L, Margolin RA, Manning RG, Cutler NR, Rapoport SI. Human brain glucose utilization and cognitive function in relation to age. Ann Neurol. 1984; 16(6):703–713. 37. Schlageter NL, Horwitz B, Creasey H, Carson R, Duara R, Berg GW, Rapaport SI. Relation of measured brain glucose utilisation and cerebral atrophy in man. J Neurol Neurosur Ps. 1987; 50:779–785. 38. Yoshii F, Barker WW, Chang JY, Loewenstein D, Apicella A, Smith D, Boothe T, Ginsberg MD, Pascal S, Duara R. Sensitivity of cerebral glucose metabolism to age, gender, brain volume, brain atrophy, and cerebrovascular risk factors. J Cerebr Blood F Met. 1988; 8:654–661. 39. Murphy DG, DeCarli C, McIntosh AR, Daly E, Mentis MJ, Pietrini P, Szczepanik J, Schapiro MB, Grady CL, Horwitz B, Rapoport SI. Sex differences in human brain morphometry and metabolism: an in vivo quantitative magnetic resonance imaging and positron emission tomography study on the effect of aging. Arch Gen Psychiat. 1996; 53(7):585–594. 40. Hawkins RA, Mazziotta JC, Phelps ME, Huang SC, Kuhl DE, Carson RE, Metter EJ, Riege WH. Cerebral glucose metabolism as a function of age in man: influence of the rate constants in the fluorodeoxyglucose method. J Cerebr Blood F Met. 1983; 3:250–253. 41. Duara R, Margolin RA, Robertson-Tchabo EA, London ED, Schwartz M, Renfrew JW, Koziarz BJ, Sundaram M, Grady C, Moore AM. Cerebral glucose utilization, as measured with positron emission tomography in 21 resting healthy men between the ages of 21 and 83 years. Brain. 1983; 106:761–775.

CORRELATES OF BRAIN AGEING

135

42. de Leon MJ, George AE, Ferris SH, Christman DR, Fowler JS, Gentes CI, Brodie J, Reisberg B, Wolf AP. Positron emission tomography and computed tomography assessments of the aging human brain. J Comput Assist Tomo. 1984; 8:88–94. 43. Garraux G, Salmon E, Degueldre C, Lemaire C, Laureys S, Franck G. Comparison of impaired subcortico-frontal metabolic networks in normal aging, subcortico-frontal dementia, and cortical frontal demential. Neuroimage. 1999; 10: 149–162. 44. Murphy D, DeCarli C, McIntosh A, Daly E, Mentis M, Pietrini P, et al. Agerelated differences in volumes of subcortical nuclei, brain matter, and cerebrospinal fluid in healthy men as measured with magnetic resonance imaging (MRI). Arch Gen Psychiat. 1996; 53:585–594. 45. Hoffman JM, Guze BH, Hawk TC. Cerebral glucose metabolism in normal individuals: effects of aging, sex and handedness. Neurology. 1988;38 (suppl 1): 167. 46. Shenkin HA, Novak P, Goluboff B, Soffe AM, Bortin L. The effects of aging, arteriosclerosis, and hypertension upon the cerebral circulation. J Clin Invest. 1953; 32:459–465. 47. Claus JJ, Breteler MMB, Hasan D, Krenning EP, Bots ML, Grobbee DE, Van Swieten JC, Van Harskamp F, Hoffman A. Regional cerebral blood flow and cerebrovascular risk factors in the elderly population. Neurobiol Aging. 1998; 19:57–64. 48. Meyer JS, Rogers RL, Mortel KF. Prospective analysis of long term control of mild hypertension on cerebral blood flow. Stroke. 1985; 16:985–990. 49. Nobili F, Rodriguez G, Marenco S, De Carli F, Gambaro M, Castello C, Pontremoli R, Rosadini G. Regional cerebral blood flow in chronic hypertension: a correlative study. Stroke. 1993; 24:1148–1152. 50. Salerno JA, Mentis MD, Grady CJ, Rapoport SI, Schapiro MB. Positron emission tomographic studies of brain function in older men with chronic essential hypertension. J Am Geriatr Soc. 2001; 40. 51. De Carli C, Murphy DGM, Tranh M, Grady CL, Haxby JV, Gillette JA, Salerno JA, Gonzales-Aviles A, Horwitz B, Rapaport SI. The effect of white matter hyperintensity volume on brain structure, cognitive performance, and cerebral metabolism of glucose in 51 healthy adults. Neurology. 1995; 45:2077–2084. 52. Sultzer DL, Mahler ME, Cummings JL, Van Gorp WG, Hinkin CH, Brown C. Cortical abnormalities associated with subcortical leisons in vascular dementia: clinical and positron emission tomographic findings. Arch Neurol. 1995; 52:773–780. 53. Takahashi W, Takagi S, Ide M, Shohtsu A, Shinohara Y. Reduced cerebral glucose metabolism in subjects with incidental hyperintensities on magnetic resonance imaging. J Neurol Sci. 2000; 176:21–27. 54. Kennedy A, Frackowiak R, Newman S, Bloomfield PM, Seaward J, Rogues P, Lewington G, Cunningham VJ, Rosser MN. Deficits in cerebral glucose metabolism demonstrated by positron emission tomography in individuals at risk of familial Alzheimer’s disease. Neurosci Lett. 1995; 186:1720. 55. Reiman E, Caselli R, Yun L, Chen K, Bandy D, Minoshima S, Thibodeau SN, Osborne D. Preclinical evidence of Alzheimer’s disease in persons homozygous for the E4 allele for apolipoprotein E. N Engl J Med. 1996; 334:752–758. 56. Reiman EM, Caselli RJ, Chen K, Alexander GE, Bandy D, Frost J. Declining brain activity in cognitively normal apolipoprotein E varepsilon 4 heterozygotes: A foundation for using positron emission tomography to efficiently test treatments to prevent Alzheimer’s disease. Proc Nat Ac Sci USA. 2001; 98:3334–3339. 57. Ritchie K, Touchon J. Mild cognitive impairment: conceptual basis and current nosological status. [see comments]. Lancet. 2000; 355 (9199):225–228. 58. Celsis P, Agniel A, Cardebat D, Demonet JF, Ousset PJ, Puel M. Age related cognitive decline: a clinical entity? A longitudinal study of cerebral blood flow and memory performance. J Neurol Neurosur Ps. 1997; 62:601–608.

136

THE AGEING BRAIN

59. McKelvey R, Bergman H, Stern J, Rush C, Zahirney G, Chertkow H. Lack of prognostic significance of SPECT abnormalities in non-demented elderly subjects with memory loss. Can J Neurol Sci. 1999; 26:23–28. 60. Small GW, La Rue A, Komo S, Kaplan A, Mandelkern MA. Predictors of cognitive change in middle-aged and older adults with memory loss. Amer J Psych. 1995; 152:1757–1764. 61. Calautti C, Serrati C, Baron JC. Effects of age on brain activation during auditory-cued thumb-to-index opposition — A positron emission tomography study. Stroke. 2001; 32:139–146. 62. Ross MH, Yurgelun-Todd DA, Renshaw PF, Maas LC, Mendelson JH, Mello NK, Gohen BM, Levin JM. Age-related reduction in functional MRI response to photic stimulation. Neurology. 1997; 48:173–176. 63. Grady CL, McIntosh AR, Bookstein F, Horwitz B, Rapoport SI, Haxby JV. Agerelated changes in regional cerebral blood flow during working memory for faces. Neuroimage. 1998; 8:409–425. 64. Madden DJ, Turkington TG, Provenzale JM, Hawk TC, Hoffman JM. Selective and divided visual attention - age-related changes in regional cerebral blood flow measured by (H2O)-O-15 pet. Hum Brain Mapp. 1997; 5:389–409. 65. Esposito G, Kirkby BS, Van Horn JD, Ellmore TM, Berman KF. Context-dependent, neural system-specific neurophysiological concomitants of ageing: mapping PET correlates during cognitive activation. Brain. 1999; 122:963–979. 66. Madden DJ, Turkington TG, Coleman RE, Provenzale JM, DeGrado TR, Hoffman JM. Adult age differences in regional cerebral blood flow during visual world identification: evidence from H215O PET. Neuroimage. 1996; 3:127–142. 67. Gur RC, Gur RE, Obrist WD, Skolnick BE, Reivich M. Age and regional cerebral blood flow at rest and during cognitive activity. Arch Gen Psychiat. 1987; 44: 617–621. 68. Grady CL, McIntosh AR, Horwitz B, Maisog JM, Ungerleider LG, Mentis MJ, Pietrini P, Schapiro MB, Haxby JV. Age-related reductions in human recognition memory due to impaired encoding. Science. 1995; 269 (5221):218–221. 69. Cabeza R, Grady CL, Nyberg L, McIntosh AR, Tulving E, Kapur S, Jennings JM, Houle S, Craik FI. Age-related differences in neural activity during memory encoding and retrieval: a positron emission tomography study. J Neurosci. 1997; 17:391–400. 70. Duara R, Gross-Glenn K, Barker WW, Chang JY, Apicella A, Loewenstein D, Boothe T. Behavioral activation and the variability of cerebral glucose metabolic measurements. J Cerebr Blood F Met. 1987; 7:266–271. 71. Skolnick BE, Gur RC, Stern MB, Hurtig HI. Reliability of regional cerebral blood flow activation to cognitive tasks in elderly normal subjects. J Cerebr Blood F Met. 1993; 13:448–453. 72. D’Esposito M, Zarahn E, Aguirre GK, Rypma B. The effect of normal aging on the coupling of neural activity to the bold hemodynamic response. Neuroimage. 1999; 10:6–14. 73. Huettel SA, Singerman JD, McCarthy G. The effects of aging upon the hemodynamic response measured by functional MRI. Neuroimage. 2001; 13:161– 175. 74. Levine BK, Beason-Held LL, Purpura KP, Aronchick DM, Optican LM, Alexander GE, Horwitz B, Rapoport SI, Schapiro MB. Age-related differences in visual perception: a PET study. Neurobiol Aging. 2000; 21:577–584. 75. Grady CL, Haxby JV, Horwitz B, Schapiro MB, Rapoport SI, Ungerleider LG, Mishkin M, Carson RE, Herscovitch P. Dissociation of object and spatial vision in human extrastriate cortex: Age-related changes in activation of regional cerebral blood flow measured with [15O] water and positron emission tomography. J Cognitive Neurosci. 1992; Vol 4:23–34.

CORRELATES OF BRAIN AGEING

137

76. Grady CL, Maisog JM, Horwitz B, Ungerleider LG, Mentis MJ, Salerno JA, Pietrini P, Wagner E, Haxby JV. Age-related changes in cortical blood flow activation during visual processing of faces and location. J Neurosci. 1994; 14: 1450–1462. 77. Grady CL, McIntosh AR, Horwitz B, Rapoport SI. Age-related changes in the neural correlates of degraded and non-degraded face processing. Cognitive Neuropsych. 2000; 17:165–186. 78. McIntosh AR, Sekuler AB, Penpeci C, Rajah MN, Grady CL, Sekuler R, Bennett PJ. Recruitment of unique neural systems to suport visual memory in normal aging. Curr Biol. 1999; 9:1275–1278. 79. Haut MW, Kuwabara H, Leach S, Callahan T. Age-related changes in neural activation during working memory performance. Aging Neuropsychol C. 2000; 7:119–129. 80. Nagahama Y, Fukuyama H, Yamauchi H, Katsumi Y, Magata Y, Shibasaki H, Kimura J. Age-related changes in cerebral blood flow activation during a Card Sorting Test. Exp Brain Res. 1997; 114:571–577. 81. Herholz K, Ehlen P, Kessler J, Strotmann T, Kalbe E, Markowitsch HJ. Learning face-name associations and the effect of age and performance: a PET activation study. Neuropsychologia. 2001; 39:643–650. 82. Madden DJ, Turkington TG, Provenzale JM, Denny LL, Hawk TC, Gottlob LR, Coleman RE. Adult age differences in the functional neuroanatomy of verbal recognition memory. Hum Brain Mapp. 1999; 7:115–135. 83. Backman L, Almkvist O, Andersson J, Nordberg A, Winblad B, Reineck R, Langstrom B. Brain activation in young and older adults during implicit and explicit retrieval. J Cognitive Neurosci. 1997; 9:391. 84. Hazlett EA, Buchsbaum MS, Mohs RC, Spiegel-Cohen J, Wei TC, Azueta R, Haznedar MM, Singer MB, Shihabuddin L, Luu-Hsia C, Harvey PD. Age-related shift in brain region activity during successful memory performance. Neurobiol Aging. 1998; 19:437–445. 85. Cabeza R, Anderson ND, Houle S, Mangels JA, Nyberg L. Age-related differences in neural activity during item and temporal-order memory retrieval: a positron emission tomography study. J Cognitive Neurosci. 2000; 12:197–206. 86. Grachev ID, Apkarian AV. Aging alters regional multichemical profile of the human brain: an in vivo 1H-MRS study of young versus middle-aged subjects. J Neurochem. 2001; 76:582–593. 87. Kadota T, Horinouchi T, Kuroda C. Development and aging of the cerebrum: assessment with proton MR spectroscopy. Amer J Neuroradiol. 2001; 22:128–135. 88. Charles HC, Lazeyras F, Krishman KR, Boyko OB, Patterson LJ, Doraiswamy PM, McDonald WM. Proton Spectroscopy of human brain: effects of age and sex. Prog Neuro-Psychoph. 1994; 18:995–1004. 89. Bruhn H, Stoppe G, Merboldt KD, Michaelis T, Hanicke W, Frahm J. Cerebral metabolite alterations in normal aging and Alzheimer’s dementia (Abstract). Proc Soc Magn Reson Med. 1992; 1:752. 90. Christiansen P, Toft P, Larsson HBW, Stubgaard M, Henriksen O. The Concentration of N-acetyl aspartate, creatine + phosphocreatine, and choline in different parts of the brain in adulthood and senium. Magn Reson Imaging. 1993; 11: 799–806. 91. Lim KO, Spielman DM. NAA in cortical gray matter with applications for measuring changes due to aging. Magn Reson Med. 1997; 37:372–377. 92. Kreis R, Ernst T, Ross BD. Absolute quantitation of water and metabolites in the human brain. II. Metabolite concentratons. J Magn Reson Imaging. 1993; 102: 9–19. 93. Soher BJ, van Zijl PCM, Duyn JH, Barker PB. Quantitative proton MR spectroscopic imaging of the human brain. Magn Reson Med. 1996; 35:356–363.

138

THE AGEING BRAIN

94. Chang L, Ernst T, Poland RE, Jenden DJ. In vivo proton magnetic resonance spectroscopy of the normal aging human brain. Life Sci. 1996; 58:2049–2056. 95. Pfefferbaum A, Adalsteinsson E, Spielman D, Sullivan EV, Lim KO. In Vivo Spectroscopic Quantification of the N-acetyl moiety, creatine, and choline from large volumes of brain gray and white matter: Effects of normal aging. Magn Reson Med. 1999; 41:276–284. 96. Saunders DE, Howe FA, van den Boogaart A, Griffiths JR, Brown MM. Aging of the adult human brain: in vivo quantitation of metabolite content with proton magnetic resonance spectroscopy. J Magn Reson Imaging. 1999; 9:711–716. 97. Angelie E, Bonmartin A, Boudraa A, Gonnaud P-M, Mallet J-J, Sappey-Marinier D. Regional differences and metabolic changes in normal aging of the human brain: proton MR spectroscopic imaging study. Amer J Neuroradiol. 2001; 22: 119–127. 98. Christiansen PB, Toft P, Gideon ER, Danielsen PR, Henriksen O. MR-visible water content in human brain: A proton MRS study. Magn Reson Imaging. 1994; 12:1237–1244.

Chapter 8 NEUROENDOCRINE ASPECTS OF BRAIN AGEING George A Smythe

Introduction Regulatory function of the human (and animal) body depends on the integrated and co-ordinated activity of two major control systems: the endocrine system and the nervous system. That there is a close interrelationship between the function of the mind and endocrine hormone secretion is an old concept of western medicine that came from findings such as the association of depression with dysfunction of the hypothalamic–pituitary–adrenal (HPA) axis.1 The field of neuroendocrinology arose out of observations pointing to significant influences being exerted by hormones, and related peptides, on the brain and vice versa. Research into this “neuroendocrine” hypothesis accelerated from the early 1970s as the technologies became increasingly available for the isolation, characterization and measurement of neurotransmitters, hypothalamic peptides, pituitary and other endocrine hormones. Emerging techniques are opening up new ways of examining brain chemistry; proton magnetic spectroscopy, for example, has recently been used to show the marked reorganization of brain chemical networks that occurs with normal ageing.2 Neuroendocrine interactions are critically important in normal human development. The role of the brain, especially its neurotransmitters and hypothalamic peptides, in control of the pituitary, thyroid, thymus, adrenals, pancreas and gonads has been extensively documented from early to adult development phases. Changes to these neuroendocrine systems post-maturity and in the elderly are less well defined. The question arises as to whether there are neuroendocrine factors which have roles in maintaining “healthy ageing,”

140

THE AGEING BRAIN

dysfunction of which may result in premature ageing and neurone loss. With ageing there are notable declines in both mental and physical function that may be mediated, in part at least, by known age-related changes in endocrine function and feedback to the brain. These include: • Cognitive function • Reproductive function • Muscle mass and strength • Cardiac performance • Immune function Ageing and Endocrine Relationships The major age-related changes in endocrine hormones include declines in circulating levels and responses of pituitary growth hormone3–6 and gonadal steroid hormones7–16, adrenal dehydroepiandrostenedione (DHEA) 17-24, and insulin-like growth factor-1 (IGF-1). 13,25–28 On the other hand, it is significant that secretion of pituitary ACTH and adrenal glucocorticoids (in contrast to DHEA) trend upward with ageing.24,29-39 These latter findings are consistent with significant changes to HPA function with ageing; evidence

Figure 1. Neuoendocrine aspects of ageing: Brain–body neuroendocrine feedback and putative age-related changes to selected brain and endocrine hormones.

NEUROENDOCRINE ASPECTS OF BRAIN AGEING

141

of cortisol excess raises questions about the role, in ageing, of stress and its effect on cognition and hippocampal neurons40-46. In Figure 1 feedback of neural signals to the periphery and from the peripheral target hormones (and products such as growth hormone-derived IGF) to the pituitary and brain are indicated by the large shaded arrows. The putative direction of change of neuroendocrine effectors with ageing are indicated by the small arrows inside the boxes. Note 1, hypothalamic corticotropin releasing hormone (CRH)42,47 and somatotropin release-inhibiting factor (somatostatin, SRIF) increase;48 growth hormone releasing hormone (GHRH) is decreased49,50 but downtrends in gonadotropin releasing hormone (GnRH) show sexual dimorphism with changes being more evident in females (data from animal studies — see below). Note 2, consistent with the hypothalamic changes, there is evidence that pituitary secretion of ACTH is increased and growth hormone (GH) is reduced but the data is less clear-cut in the case of the pituitary gonadotropins where, again, there is evidence of sexual dimorphism and significant variation between changes in luteinizing hormone (LH) versus those in follicle stimulating hormone (FSH).51 Note 3, as a consequence of reduced GH secretion, the product of its action on the liver and other tissues,52 insulin-like growth factor-1 (IGF-1) is reduced. Note 4, the bulk of evidence is consistent with increased adrenal glucocorticoid production whereas the adrenal androgenic steroid DHEA and its sulphate is reduced.44 Here there is sexual dimorphism with women showing greater changes than men with ageing.53 Note 5, estrogen levels decline following menopause in women and testosterone levels are reduced in men (and women) with ageing.51 Ageing, Stress, and the Hypothalamic–Pituitary–Adrenal (HPA) Axis Evidence of altered HPA function with ageing comes from both animal and human studies. Sapolsky and co-workers showed in the rat there is a significant age-related increase in corticosterone production and that the ability of animals to “turn off” stress-induced glucocorticoid release is impaired.54,55 The apparent age-related failure of glucocorticoid excess to exert normal feedback inhibition on central, hypothalamic, or pituitary receptors has attracted considerable research.24,33,34,40,42, 46,56–60 Normally, a principle brain response to stress is markedly increased activity of noradrenergic neuronal activity. 61 This increased noradrenergic drive acts on CRH neurons at the level of the paraventricular nucleus of the hypothalamus and CRH is released into the pituitary portal circulation to stimulate pituitary ACTH release (see Figure 2). Figure 2 also indicates the feedback of glucocorticoids at the level of the anterior pituitary and certain brain centres to inhibit further CRF and ACTH release. Figure 2 summarizes neuroendocrine control of the HPA axis and glucocorticoid feedback. Included in this diagram are inputs to and from the hippocampus, amygdala and the bed nucleus of the stria terminalis (BNST)

142

THE AGEING BRAIN

Figure 2. The hypothalamic-pituitary-adrenal-axis. Neuroendocrine control of ACTH release and central sites of glucocorticoid feedback.

all of which can mediate inhibition or stimulation of CRH release.42,62,63 Using modern analytical methods, the hypothalamic neuronal activity of norepinephrine (NE) and serotonin (5-HT) at terminals arising from cells in the brainstem (locus coeruleus) and midbrain (Raphe) can be assessed by measuring the transmitters and their primary neuronal metabolites (DHPG and 5-HIAA, respectively).61,64 The “glucocorticoid cascade” hypothesis of ageing and hippocampal damage A number of animal studies have shown an age-related chronic increase in corticosterone and an apparent failure of glucocorticoid negative feedback. These data, taken with evidence that excess glucocorticoid levels caused damage to hippocampal neurones that are involved in cognition, led Sapolsky et al. to propose the glucocorticoid cascade hypothesis of ageing and hippocampal damage.65 In ageing man there is a decline in cognitive function and several groups have proposed that this may be due to cortisol excess in accord with Sapolsky’s hypothesis.40,41,57,66,67 Not all studies support the hypothesis in man68,69 and Angelucci, in questioning earlier conclusions, has suggested that changes seen represent an adaptive response to a CNS neurodegenerative inflammatory process.69 It should be noted that, in man, age-related eleva-

NEUROENDOCRINE ASPECTS OF BRAIN AGEING

143

Figure 3. Effects of age on the HPA axis in male Wistar rats. Means± SEM are shown, n=6 per group.

tions of cortisol are not always significant and there are sex differences.44,68 However most evidence does point to a failure of glucocorticoid feedback with ageing but the level at which this failure occurs is not clear.29,31,35,36,70 The author has investigated the status of hypothalamic noradrenergic neuronal activity, circulating ACTH and corticosterone in young (2 months old) compared with old (20–24 months old) rats. The results of this unpublished study are shown in Figure 3. Consistent with published results47 the data show a significant increase in serum concentrations of ACTH and a nonsignificant increase in serum corticosterone. The novel finding here is that of a significant decrease in hypothalamic neuronal activity (DHPG/NE)61 in the old rats. This is discordant with the HPA relationships in normal animals subjected to stress where there is a positive relationship between ACTH and medial basal hypothalamus (mbh) DHPG/NE ratio.61 These data indicate that in the rat, with respect to the HPA axis, ageing is not akin to stress and that the failure of feedback inhibition does not seem to occur at the level of the hypothalamus. Neuroendocrine Changes with the Menopause Ovulatory failure is one of the earliest events of ageing and is one that clearly involves neuroendocrine mechanisms. The central nervous system has been described as the pacemaker of reproductive senescence.71 The consequences of menopause for women relate not only to the loss of reproductive ability but the loss of ovarian follicular estrogen and the decline of circulating estrogen levels. In a major review, Wise et al.72 have highlighted many findings with respect to neurotrophic and neuroprotective properties of estrogen. These findings reinforce the ideas underlying estrogen replacement therapy,73,74 for which there are negative as well as positive issues to be considered.75 The reported benefits of peri- and post-menopausal estrogen replacement include maintenance of normal bone and mineral metabolism,76,77 memory and cognition,78–82 decreased sympathetic nervous system activity, 83,84

144

THE AGEING BRAIN

decreased cortisol and HPA axis response.85-87 Improved IGF-1 production is seen with transdermal estrogen administration compared with the oral route;25 the transdermal route of estrogen administration is also associated with improved insulin sensitivity and glucose metabolism.88 Improved memory and attention in postmenopausal women with Alzheimer’s disease has been reported following estrogen administration.79 However, not all investigations have demonstrated a protective effect of estrogen treatment against declines in cognitive function or stress reponses in older non-demented women.89,90 Two recent randomized, double-blind trials of estrogen treatment in women with Alzheimer’s disease (AD) failed to demonstrate any cognitive or functional improvement following estogen.91,92 While it is clear from animal and clinical studies that estrogen acts in the brain via a number of established estrogen receptors as a neuromodulatory and neuroprotective hormone,93-95 it does not appear to act to restore existing neural damage. The Age-Related Decline in Growth Hormone and IGF-1 Human growth hormone secretion (hGH) and IGF-1 production undergo significant declines with ageing.96 The age-related decline in hGH with ageing is exemplified by the comprehensive studies of van Cauter and colleagues into sleep-related hGH release.39 Which neuroendocrine factors either mediate or are affected by the age-related changes have not been established, but several possibilities do arise (cf. Figure 1). These include: i) a primary pituitary defect, ii) increased hypothalamic release of SRIF to inhibit pituitary GH release, iii) reduced hypothalamic release of GHRH, and iv) altered activity of the hypothalamic neurotransmitters that mediate release of SRIF and/or GHRH — either primarily or as a consequence of reduced negative feedback. The bulk of research data from both man and animals mitigates against a primary pituitary defect and favours reduced GHRH activity being a major contributor to the age-related decline in GH.6,49,97–102 Some evidence also supports the possibility that there may also be increased SRIF activity with ageing.49,50,103 Alterations in the activities of these peptides may, in turn, be mediated by their controlling neurotransmitters. The monoaminergic control of GHRH and SRIF has been a contentious issue for many years with both catecholamines and serotonin being proposed as the primary mediators of hypothalamic GHRH release.104–107 Direct measurement of hypothalamic monoamine neurotransmitter activity in the rat is consistent with serotonin neuronal activity being a major activator of GHRH release.105,106,108 When considered with the reduced activity of GHRH noted above, it is conceivable that there may be an age-related decline in central serotonin function and support for this is found in human studies demonstrating reduced brain serotonin activity and binding sites.109–112 If serotoninergic systems in the brain are reduced with ageing then this would not appear to be due to reduced GH

NEUROENDOCRINE ASPECTS OF BRAIN AGEING

145

negative feedback. In the rat at least, GH lack (hypophysectomy) is associated with increased hypothalamic serotonin neuronal activity which is reduced to normal by exogenous GH administration.105 GH and IGF-1 decline with ageing and, taken with the ready availability of recombinant GH, there has been considerable interest in the role of GH replacement therapy to treat this so-called “somatopause”.113–119 This concept is not without controversy and Morley120 has questioned whether, in this context, “GH is a fountain of youth or a death hormone?” The general consensus points to the need for carefully controlled studies.113,115,119,121–123 Androgens and Aging: the “Adrenopause” LH secretion in men is regulated by release of GnRH from the pituitary and negative feedback of circulating testosterone. With ageing, this neuroendocrine control is altered as LH levels increase at the rate of 1.9% per year,124 serum testosterone levels decline at the rate of about 0.4% per year123 and DHEA decreases at a faster rate of 3.1%.124 On the basis of a relative “hypogonadism of ageing” in older men, hormone replacement with testosterone and DHEA have been investigated. In general, testosterone replacement in older hypogonadal males have indicated positive effects on muscle mass, bone mineral density and fat mass but it is too early to establish its true efficacy51 or whether it has any effect in improving neural function. In developing humans, DHEA and its sulphate (DHEAS) are the most abundant steroids but their levels decline after adrenarche and with ageing. DHEAS treatment in animals has been reported to have memory-enhancing properties and replacement in ageing has been proposed.125, 126 Longitudinal studies in elderly men and women have failed to show any association between DHEA levels and cognitive performance127 and caution has been advised in relation to its use as a hormone replacement in ageing.123, 128 Summary Neuroendocrine systems change with ageing. The HPA axis is altered, particularly with respect to glucocorticoid responses and feedback to the pituitary and brain. These changes are proposed to alter hippocampal neurons and cognition. Menopause is associated with marked changes as the decline in estrogen levels reaching the brain take effect. The case for estrogen replacement is strong. Estrogen replacement appears able to prevent (but not repair) brain neuronal damage but specific brain actions at the hippocampus and other regions require further study. The ageing male undergoes a so-called “adrenopause” with falling androgens and brain feedback but effects of exogenous androgen treatment is less well studied than that of the estrogen deficient aged female. Age-related declines of GH and IGF-1 are associated with

146

THE AGEING BRAIN

physical and metabolic deterioration and, at the level of the brain, declining serotoninergic systems. The change in this neurotransmitter may reflect a primary failure in GHRH stimulation with ageing. While many of the neuroendocrine changes seen with ageing are reminiscent of those seen with chronic stress, there is no clear evidence that ageing equates with stress. No doubt, however, stress can accelerate or exacerbate the effects of ageing. Acknowledgements The author wishes to express his sincere thanks to Mr Ray Williams for his inspirational support and for generously providing equipment for our work. I also wish to thank all of my colleagues in the BMSF for their constructive assistance. References 1. Carroll B, Mendels J. Neuroendocrine regulation in affective disorder. In: Sachar E, editor. Hormones, behavior, and psychopathology. New York: Raven Press, 1976:193–224. 2. Grachev ID, Apkarian AV. Chemical network of the living human brain. Evidence of reorganization with aging. Brain Res Cogn Brain Res. 2001; 11:185–197. 3. Gil-Ad I, Gurewitz R, Marcovici O, Rosenfeld J, Laron Z. Effect of aging on human plasma growth hormone response to clonidine. Mech Ageing Dev. 1984; 27:97–100. 4. Muggeo M, Fedele D, Tiengo A, Molinari M, Crepaldi G. Human growth hormone and cortisol response to insulin stimulation in aging. J Gerontol. 1975; 30: 546–551. 5. Rudman D, Kutner MH, Rogers CM, Lubin MF, Fleming GA, Bain RP. Impaired growth hormone secretion in the adult population: relation to age and adiposity. J Clin Invest. 1981; 67:1361–1369. 6. Russell-Aulet M, Dimaraki EV, Jaffe CA, DeMott-Friberg R, Barkan AL. Agingrelated growth hormone (GH) decrease is a selective hypothalamic GH-releasing hormone pulse amplitude mediated phenomenon. J Gerontol A Biol Sci Med Sci. 2001; 56:M124–129. 7. Wise PM, Smith MJ, Dubal DB, Wilson ME, Krajnak KM, Rosewell KL. Neuroendocrine influences and repercussions of the menopause. Endocr Rev. 1999; 20:243–248. 8. Jiroutek MR, Chen MH, Johnston CC, Longcope C. Changes in reproductive hormones and sex hormone-binding globulin in a group of postmenopausal women measured over 10 years. Menopause. 1998; 5:90–94. 9. Santoro N, Banwell T, Tortoriello D, Lieman H, Adel T, Skurnick J. Effects of aging and gonadal failure on the hypothalamic-pituitary axis in women. Am J Obstet Gynecol. 1998; 178:732–741. 10. Coulam CB. Age, Estrogens, and the psyche. Clin Obstet Gynecol. 1981; 24: 219–229. 11. Basaria S, Dobs AS. Hypogonadism and androgen replacement therapy in elderly men. Am J Med. 2001; 110:563–572.

NEUROENDOCRINE ASPECTS OF BRAIN AGEING

147

12. Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR. Longitudinal effects of aging on serum total and free testosterone levels in healthy men. Baltimore Longitudinal Study of Aging. J Clin Endocr Metab. 2001; 86:724–731. 13. Leifke E, Gorenoi V, Wichers C, Von Zur Muhlen A, Von Buren E, Brabant G. Age-related changes of serum sex hormones, insulin-like growth factor-1 and sex-hormone binding globulin levels in men: cross-sectional data from a healthy male cohort. Clin Endocrinol (Oxf). 2000; 53:689–695. 14. Nankin HR, Calkins JH. Decreased bioavailable testosterone in aging normal and impotent men. J Clin Endocr Metab. 1986; 63:1418–1420. 15. Vermeulen A. Gonadal senescence in the male. J Endocr Invest. 1985; 8:93–98. 16. Noth RH, Mazzaferri EL. Age and the endocrine system. Clin Geriatr Med. 1985; 1:223–250. 17. Ohashi M, Kato K, Nawata H, Ibayashi H. Adrenocortical responsiveness to graded ACTH infusions in normal young and elderly human subjects. Gerontology. 1986; 32:43–51. 18. Pavlov EP, Harman SM, Chrousos GP, Loriaux DL, Blackman MR. Responses of plasma adrenocorticotropin, cortisol, and dehydroepiandrosterone to ovine corticotropin-releasing hormone in healthy aging men. J Clin Endocr Metab. 1986; 62:767–772. 19. Baulieu EE. Studies on dehydroepiandrosterone (DHEA) and its sulphate during aging. CR Acad Sci III. 1995; 318:7–11. 20. Hornsby PJ. Biosynthesis of DHEAS by the human adrenal cortex and its agerelated decline. Ann NY Acad Sci. 1995; 774:29–46. 21. Hinson JP, Raven PW. DHEA deficiency syndrome: a new term for old age? J Endocrinol. 1999; 163:1–5. 22. Parker CR, Jr. Dehydroepiandrosterone and dehydroepiandrosterone sulfate production in the human adrenal during development and aging. Steroids. 1999; 64:640–647. 23. Parker CR, Jr., Slayden SM, Azziz R, et al. Effects of aging on adrenal function in the human: responsiveness and sensitivity of adrenal androgens and cortisol to adrenocorticotropin in premenopausal and postmenopausal women. J Clin Endocr Metab. 2000; 85:48–54. 24. Ferrari E, Cravello L, Muzzoni B, et al. Age-related changes of the hypothalamicpituitary-adrenal axis: pathophysiological correlates. Eur J Endocrinol. 2001; 144:319–329. 25. Ho KK, O’Sullivan AJ, Weissberger AJ, Kelly JJ. Sex steroid regulation of growth hormone secretion and action. Horm Res. 1996; 45:67–73. 26. Toogood AA, Shalet SM. Ageing and growth hormone status. Baillieres Clin Endocrinol Metab. 1998; 12:281–296. 27. Aleman A, de Vries WR, de Haan EH, Verhaar HJ, Samson MM, Koppeschaar HP. Age-sensitive cognitive function, growth hormone and insulin-like growth factor 1 plasma levels in healthy older men. Neuropsychobiology. 2000; 41: 73–78. 28. Ravaglia G, Forti P, Maioli F, et al. Body composition, sex steroids, IGF-1, and bone mineral status in aging men. J Gerontol A Biol Sci Med Sci. 2000; 55: M516–521. 29. Born J, Ditschuneit I, Schreiber M, Dodt C, Fehm HL. Effects of age and gender on pituitary-adrenocortical responsiveness in humans. Eur J Endocrinol. 1995; 132:705–711. 30. Copinschi G, Van Cauter E. Effects of ageing on modulation of hormonal secretions by sleep and circadian rhythmicity. Horm Res. 1995; 43:20–24. 31. Ferrari E, Magri F, Dori D, et al. Neuroendocrine correlates of the aging brain in humans. Neuroendocrinology. 1995; 61:464–470.

148

THE AGEING BRAIN

32. Raskind MA, Peskind ER, Pascualy M, et al. The effects of normal aging on cortisol and adrenocorticotropin responses to hypertonic saline infusion. Psychoneuroendocrinology. 1995; 20:637–644. 33. Deuschle M, Gotthardt U, Schweiger U, et al. With aging in humans the activity of the hypothalamus-pituitary-adrenal system increases and its diurnal amplitude flattens. Life Sci. 1997; 61:2239–2246. 34. Magri F, Locatelli M, Balza G, et al. Changes in endocrine circadian rhythms as markers of physiological and pathological brain aging. Chronobiol Int. 1997; 14: 385–396. 35. Wilkinson CW, Peskind ER, Raskind MA. Decreased hypothalamic-pituitaryadrenal axis sensitivity to cortisol feedback inhibition in human aging. Neuroendocrinology. 1997; 65:79–90. 36. Boscaro M, Paoletta A, Scarpa E, et al. Age-related changes in glucocorticoid fast feedback inhibition of adrenocorticotropin in man. J Clin Endocr Metab. 1998; 83:1380–1383. 37. Luisi S, Tonetti A, Bernardi F, et al. Effect of acute corticotropin releasing factor on pituitary-adrenocortical responsiveness in elderly women and men. J Endocrinol Invest. 1998; 21:449–453. 38. Giordano R, Arvat E, Maccagno B, et al. Corticotroph and adrenal responsiveness to hCRH, hexarelin and ACTH in young and elderly subjects. J Endocrinol Invest. 1999; 22:82. 39. Van Cauter E, Leproult R, Plat L. Age-related changes in slow wave sleep and REM sleep and relationship with growth hormone and cortisol levels in healthy men. J Amer Med Assoc. 2000; 284:861–868. 40. Lupien SJ, Nair NP, Briere S, et al. Increased cortisol levels and impaired cognition in human aging: implication for depression and dementia in later life. Rev Neurosci. 1999; 10:117–139. 41. McEwen BS. Stress and the aging hippocampus. Front Neuroendocrin. 1999; 20: 49–70. 42. Pedersen WA, Wan R, Mattson MP. Impact of aging on stress-responsive neuroendocrine systems. Mech Ageing Dev. 2001; 122:963–983. 43. McEwen B, Chao H, Spencer R, Brinton R, Macisaac L, Harrelson A. Corticosteroid receptors in brain: relationship of receptors to effects in stress and aging. Ann NY Acad Sci. 1987; 512:394–401. 44. Yen SS, Laughlin GA. Aging and the adrenal cortex. Exp Gerontol. 1998; 33: 897–910. 45. Bremner JD, Narayan M. The effects of stress on memory and the hippocampus throughout the life cycle: implications for childhood development and aging. Dev Psychopathol. 1998; 10:871–885. 46. Gotthardt U, Schweiger U, Fahrenberg J, Lauer CJ, Holsboer F, Heuser I. Cortisol, ACTH, and cardiovascular response to a cognitive challenge paradigm in aging and depression. Am J Physiol. 1995; 268:R865–873. 47. Herman JP, Larson BR, Speert DB, Seasholtz AF. Hypothalamo-pituitary-adrenocortical dysregulation in aging F344/Brown-Norway F1 hybrid rats. Neurobiol Aging. 2001; 22:323–332. 48. Ghigo E, Arvat E, Gianotti L, et al. Human aging and the GH-IGF-I axis. J Pediatr Endocr Met. 1996; 9:271–278. 49. Soule SG, Macfarlane P, Levitt NS, Millar RP. Contribution of growth hormonereleasing hormone and somatostatin to decreased growth hormone secretion in elderly men. S Afr Med J. 2001; 91:254–260. 50. Arvat E, Ceda GP, Di Vito L, et al. Age-related variations in the neuroendocrine control, more than impaired receptor sensitivity, cause the reduction in the GHreleasing activity of GHRPs in human aging. Pituitary. 1998; 1:51–58. 51. Morley JE. Androgens and aging. Maturitas. 2001; 38:61–71.

NEUROENDOCRINE ASPECTS OF BRAIN AGEING

149

52. Le Roith D, Scavo L, Butler A. What is the role of circulating IGF-I? Trends Endocrin Met. 2001; 12:48–52. 53. Laughlin GA, Barrett-Connor E. Sexual dimorphism in the influence of advanced aging on adrenal hormone levels: the Rancho Bernardo Study. J Clin Endocr Met. 2000; 85:3561–3568. 54. Sapolsky RM, Krey LC, McEwen BS. The adrenocortical stress-response in the aged male rat: impairment of recovery from stress. Exp Gerontol. 1983; 18: 55–64. 55. Sapolsky RM. Glucocorticoids, stress, and their adverse neurological effects: relevance to aging. Exp Gerontol. 1999; 34:721–732. 56. Cizza G, Gold PW, Chrousos GP. Aging is associated in the 344/N Fischer rat with decreased stress responsivity of central and peripheral catecholaminergic systems and impairment of the hypothalamic-pituitary-adrenal axis. Ann NY Acad Sci. 1995; 771:491–511. 57. Wang PS, Lo MJ, Kau MM. Glucocorticoids and aging. J Formos Med Assoc. 1997; 96:792–801. 58. De Kloet ER, Sutanto W, Rots N, et al. Plasticity and function of brain corticosteroid receptors during aging. Acta Endocrinol–(Cop.) 1991; 125:65–72. 59. Heuser IJ, Gotthardt U, Schweiger U, et al. Age-associated changes of pituitaryadrenocortical hormone regulation in humans: importance of gender. Neurobiol Aging. 1994; 15:227–231. 60. Dodt C, Theine KJ, Uthgenannt D, Born J, Fehm HL. Basal secretory activity of the hypothalamo-pituitary-adrenocortical axis is enhanced in healthy elderly. An assessment during undisturbed night-time sleep. Eur J Endocrinol. 1994; 131: 443–450. 61. Smythe GA, Bradshaw JE, Vining RF. Hypothalamic monoamine control of stress-induced adrenocorticotropin release in the rat. Endocrinology. 1983; 113: 1062–1071. 62. Feldman S, Conforti N, Weidenfeld J. Limbic pathways and hypothalamic neurotransmitters mediating adrenocortical responses to neural stimuli. Neurosci Biobehav Rev. 1995; 19:235–240. 63. Lee Y, Davis M. Role of the hippocampus, the bed nucleus of the stria terminalis, and the amygdala in the excitatory effect of corticotropin-releasing hormone on the acoustic startle reflex. J Neurosci. 1997; 17:6434–6446. 64. Smythe GA, Brandstater JF, Lazarus L. Serotoninergic control of rat growth hormone secretion. Neuroendocrinology. 1975; 17:245–257. 65. Sapolsky RM, Krey LC, McEwen BS. The neuroendocrinology of stress and aging: the glucocorticoid cascade hypothesis. Endocrinol Rev. 1986; 7:284–301. 66. Martignoni E, Costa A, Sinforiani E, et al. The brain as a target for adrenocortical steroids: cognitive implications. Psychoneuroendocrinology. 1992; 17:343–354. 67. O’Brien JT. The ‘glucocorticoid cascade’ hypothesis in man: prolonged stress may cause permanent brain damage. Br J Psychiat. 1997; 170:199–201. 68. Kudielka BM, Schmidt-Reinwald AK, Hellhammer DH, Schurmeyer T, Kirschbaum C. Psychosocial stress and HPA functioning: no evidence for a reduced resilience in healthy elderly men. Stress. 2000; 3:229–240. 69. Angelucci L. The glucocorticoid hormone: from pedestal to dust and back. Eur J Pharmacol. 2000; 405:139–147. 70. Veldhuis JD, Iranmanesh A, Samojlik E, Urban RJ. Differential sex steroid negative feedback regulation of pulsatile follicle-stimulating hormone secretion in healthy older men: deconvolution analysis and steady-state sex-steroid hormone infusions in frequently sampled healthy older individuals. J Clin Endocr Met. 1997; 82:1248–1254. 71. Wise PM. Neuroendocrine modulation of the “menopause”: insights into the aging brain. Am J Physiol. 1999; 277:E965–970.

150

THE AGEING BRAIN

72. Wise PM, Dubal DB, Wilson ME, Rau SW, Liu Y. Estrogens: trophic and protective factors in the adult brain. Front Neuroendocrin. 2001; 22:33–66. 73. Lichtman R. Perimenopausal hormone replacement therapy. Review of the literature. J Nurse Midwifery. 1991; 36:30–48. 74. Bjorntorp P. Neuroendocrine ageing. J Intern Med. 1995; 238:401–404. 75. Nerhood RC. Making a decision about ERT/HRT. Evidence to consider in initiating and continuing protective therapy. Postgrad Med J. 2001; 109:167–170, 173–174, 178. 76. Kulak CA, Bilezikian JP. Osteoporosis: preventive strategies. Int J Fertil Womens M. 1998; 43:56–64. 77. Shiflett S, Cooke CE. Osteoporosis: a focus on treatment. Maryland State Med J. 1997; 46:303–307. 78. Verghese J, Kuslansky G, Katz MJ, et al. Cognitive performance in surgically menopausal women on estrogen. Neurology. 2000; 55:872–874. 79. Erkkola R. Female menopause, hormone replacement therapy, and cognitive processes. Maturitas. 1996; 23:S27–30. 80. Asthana S, Craft S, Baker LD, et al. Cognitive and neuroendocrine response to transdermal estrogen in postmenopausal women with Alzheimer’s disease: results of a placebo-controlled, double-blind, pilot study. Psychoneuroendocrinology. 1999; 24:657–677. 81. LeBlanc ES, Janowsky J, Chan BK, Nelson HD. Hormone replacement therapy and cognition: systematic review and meta-analysis. J Amer Med Assoc. 2001; 285:1489–1499. 82. Rice MM, Graves AB, McCurry SM, et al. Postmenopausal estrogen and estrogenprogestin use and 2-year rate of cognitive change in a cohort of older Japanese American women: The Kame Project. Arch Intern Med. 2000; 160:1641–1649. 83. Menozzi R, Cagnacci A, Zanni AL, Bondi M, Volpe A, Del Rio G. Sympathoadrenal response of postmenopausal women prior and during prolonged administration of estradiol. Maturitas. 2000; 34:275–281. 84. Ceresini G, Freddi M, Izzo S, et al. Post-menopausal estrogen supplementation only partially blunts the sympathoadrenal response to mental stress. J Endocrinol Invest. 1999; 22:72–73. 85. Prinz P, Bailey S, Moe K, Wilkinson C, Scanlan J. Urinary free cortisol and sleep under baseline and stressed conditions in healthy senior women: effects of estrogen replacement therapy. J Sleep Res. 2001; 10:19–26. 86. Lindheim SR, Legro RS, Bernstein L, et al. Behavioral stress responses in premenopausal and postmenopausal women and the effects of estrogen. Am J Obstet Gynecol. 1992; 167:1831–1836. 87. Kudielka BM, Schmidt-Reinwald AK, Hellhammer DH, Kirschbaum C. Psychological and endocrine responses to psychosocial stress and dexamethasone/ corticotropin-releasing hormone in healthy postmenopausal women and young controls: the impact of age and a two-week estradiol treatment. Neuroendocrinology. 1999; 70:422–430. 88. O’Sullivan AJ, Ho KK. A comparison of the effects of oral and transdermal estrogen replacement on insulin sensitivity in postmenopausal women. J Clin Endocr Met. 1995; 80:1783–1788. 89. Matthews KA, Flory JD, Owens JF, Harris KF, Berga SL. Influence of estrogen replacement therapy on cardiovascular responses to stress of healthy postmenopausal women. Psychophysiology. 2001; 38:391–398. 90. Matthews K, Cauley J, Yaffe K, Zmuda JM. Estrogen replacement therapy and cognitive decline in older community women. J Am Geriatr Soc. 1999; 47:518–523. 91. Mulnard RA, Cotman CW, Kawas C, et al. Estrogen replacement therapy for treatment of mild to moderate Alzheimer disease: a randomized controlled trial. Alzheimer’s Disease Cooperative Study. J Amer Med Assoc. 2000; 283:1007– 1015.

NEUROENDOCRINE ASPECTS OF BRAIN AGEING

151

92. Henderson VW, Paganini-Hill A, Miller BL, et al. Estrogen for Alzheimer’s disease in women: randomized, double-blind, placebo-controlled trial. Neurology. 2000; 54:295–301. 93. Garcia-Segura LM, Azcoitia I, DonCarlos LL. Neuroprotection by estradiol. Prog Neurobiol. 2001; 63:29–60. 94. Shughrue PJ, Merchenthaler I. Evidence for novel estrogen binding sites in the rat hippocampus. Neuroscience. 2000; 99:605–612. 95. Dubal DB, Zhu H, Yu J, et al. Estrogen receptor alpha, not beta, is a critical link in estradiol-mediated protection against brain injury. Proc Natl Acad Sci USA. 2001; 98:1952–1957. 96. Corpas E, Harman SM, Blackman MR. Human growth hormone and human aging. Endocr Rev. 1993; 14:20–39. 97. Corpas E, Harman SM, Pineyro MA, Roberson R, Blackman MR. Growth hormone (GH)-releasing hormone-(1–29) twice daily reverses the decreased GH and insulin-like growth factor-I levels in old men. J Clin Endocrinol Metab. 1992; 75:530–535. 98. Merriam GR, Buchner DM, Prinz PN, Schwartz RS, Vitiello MV. Potential applications of GH secretagogs in the evaluation and treatment of the agerelated decline in growth hormone secretion. Endocrine. 1997; 7:49–52. 99. Guldner J, Schier T, Friess E, Colla M, Holsboer F, Steiger A. Reduced efficacy of growth hormone-releasing hormone in modulating sleep endocrine activity in the elderly. Neurobiol Aging. 1997; 18:491–495. 100. degli Uberti EC, Ambrosio MR, Cella SG, et al. Defective hypothalamic growth hormone (GH)-releasing hormone activity may contribute to declining GH secretion with age in man. J Clin Endocrin Metab. 1997; 82:2885–2888. 101. Mulligan T, Jaen-Vinuales A, Godschalk M, Iranmanesh A, Veldhuis JD. Synthetic somatostatin analog (octreotide) suppresses daytime growth hormone secretion equivalently in young and older men: preserved pituitary responsiveness to somatostatin’s inhibition in aging. J Am Geriatr Soc. 1999; 47: 1422–1424. 102. Thorner MO, Chapman IM, Gaylinn BD, Pezzoli SS, Hartman ML. Growth hormone-releasing hormone and growth hormone-releasing peptide as therapeutic agents to enhance growth hormone secretion in disease and aging. Recent Prog Horm Res. 1997; 52:215–244; (discussion) 244–246. 103. Marcell TJ, Wiswell RA, Hawkins SA, Tarpenning KM. Age-related blunting of growth hormone secretion during exercise may not be soley due to increased somatostatin tone. Metabolism. 1999; 48:665–670. 104. Muller EE. Some aspects of the neurotransmitter control of anterior pituitary function. Pharmacol Res. 1989; 21:75–85. 105. Smythe GA, Duncan MW, Bradshaw JE, Cai WY. Serotoninergic control of growth hormone secretion: hypothalamic dopamine, norepinephrine, and serotonin levels and metabolism in three hyposomatotropic rat models and in normal rats. Endocrinology. 1982; 110:376–383. 106. Smythe GA, Gleeson RM, Stead BH. Stimulation of the hypothalamic-pituitaryadrenal axis and inhibition of growth hormone release via increased central noradrenaline neuronal activity by urethane anaesthesia in the rat: blockade by clonidine. Aust J Biol Sci. 1987; 40:91–96. 107. Conway S, Richardson L, Speciale S, Moherek R, Mauceri H, Krulich L. Interaction between norepinephrine and serotonin in the neuroendocrine control of growth hormone release in the rat. Endocrinology. 1990; 126:1022–1030. 108. Gotoh M, Hirooka Y, Tajima T, Iguchi A, Smythe GA. Adrenocorticotropin and growth hormone secretions after intracerebroventricular administration of neostigmine in rats: their relationships to hypothalamic monoaminergic neuronal activities. Brain Res. 1994; 659:259–262.

152

THE AGEING BRAIN

109. Lerer B, Gelfin Y, Shapira B. Neuroendocrine evidence for age-related decline in central serotonergic function. Neuropsychopharmacology. 1999; 21:321–322. 110. Kakiuchi T, Nishiyama S, Sato K, Ohba H, Nakanishi S, Tsukada H. Age-related reduction of [11C]MDL100,907 binding to central 5-HT(2A) receptors: PET study in the conscious monkey brain. Brain Res. 2000; 883:135–142. 111. van Dyck CH, Malison RT, Seibyl JP, et al. Age-related decline in central serotonin transporter availability with [(123)I]beta-CIT SPECT. Neurobiol Aging. 2000; 21:497–501. 112. Meltzer CC, Smith G, DeKosky ST, et al. Serotonin in aging, late-life depression, and Alzheimer’s disease: the emerging role of functional imaging. Neuropsychopharmacology. 1998; 18:407–430. 113. Lamberts SW. The somatopause: to treat or not to treat? Horm Res. 2000; 53: 42–43. 114. Savine R, Sonksen P. Growth hormone – hormone replacement for the somatopause? Horm Res. 2000; 53:37–41. 115. Cummings DE, Merriam GR. Age-related changes in growth hormone secretion: should the somatopause be treated? Semin Reprod Endocr 1999; 17:311–325. 116. Hoffman AR, Ceda GP. Should we treat the somatopause? J Endocr Invest. 1999; 22:4–6. 117. Lieberman SA, Hoffman AR. The somatopause: should growth hormone deficiency in older people Be treated? Clin Geriatr Med. 1997; 13:671–684. 118. Hoffman AR, Lieberman SA, Butterfield G, et al. Functional consequences of the somatopause and its treatment. Endocrine. 1997; 7:73–76. 119. Sonsken P. Growth hormone and the somatopause. Growth Horm IGF Res. 1999; 9:1–2. 120. Morley JE. Growth hormone: fountain of youth or death hormone? J Am Geriatr Soc. 1999; 47:1475–1476. 121. von Werder K. The somatopause is no indication for growth hormone therapy. J Endocrinol Invest. 1999; 22:137–141. 122. Toogood AA, Shalet SM. Conflicts with the somatopause. Growth Horm IGF Res. 1998; 8:47–54. 123. Janssens H, Vanderschueren DM. Endocrinological aspects of aging in men: is hormone replacement of benefit? Eur J Obstet Gyn R B. 2000; 92:7–12. 124. Gray A, Feldman HA, McKinlay JB, Longcope C. Age, disease, and changing sex hormone levels in middle-aged men: results of the Massachusetts Male Aging Study. J Clin Endocr Metab. 1991; 73:1016–1025. 125. Baulieu EE. Dehydroepiandrosterone (DHEA): a fountain of youth? J Clin Endocrin Metab. 1996; 81:3147–3151. 126. Majewska MD. Neuronal actions of dehydroepiandrosterone. Possible roles in brain development, aging, memory, and affect. Ann NY Acad Sci. 1995; 774: 111–120. 127. Carlson LE, Sherwin BB. Relationships among cortisol (CRT), dehydroepiandrosterone-sulfate (DHEAS), and memory in a longitudinal study of healthy elderly men and women. Neurobiol Aging. 1999; 20:315–324. 128. Steel N. Dehydro-epiandrosterone and ageing. Age Ageing. 1999; 28:89–91.

Chapter 9 CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN Velandai K. Srikanth and Geoffrey A. Donnan*

Introduction It is inevitable that the vascular system of the brain undergoes change with increasing age. The primary purpose of this chapter is to review the existing literature with respect to the impact of ageing on anatomical and physiological aspects of the cerebrovascular system. The links and postulated mechanisms between ageing, cerebrovascular changes and disease states will also be discussed, as this is of primary interest to the clinician. Ageing and Cerebrovascular Anatomy Ageing and cerebral macrovasculature Ageing and Arterial Changes The arterial tree supplying the brain is comprised of the extracranial large arteries and the intracranial anastomotic network (The Circle of Willis), together with the progressively smaller vessels that penetrate brain tissue and supply specific areas of the brain. The defining feature of these groups of blood vessels is that they contain smooth muscle and are considered to be responsible for maintaining adequate cerebral blood flow. *To whom correspondence should be addressed.

154

THE AGEING BRAIN

Arterial changes in ageing closely resemble changes seen in the vascular tree elsewhere. The changes most often observed with ageing include intimal thickening, medial fibrosis and loss of elasticity for the larger arterioles and arteries. These changes may occur at a slower rate in cerebral arteries than in peripheral arteries such as the radial artery or the coronary artery.1 The magnitude of these changes is observed to increase with each decade from the age of 55.2 Smaller intraparenchymal vessels often tend to show tortuosity, loops and kinks.3 It has been postulated that these loops and kinks may have been mistaken for the so-called Charcot-Bouchard aneurysms, given that current pathological techniques are superior in delineating their characteristics. Ageing, disease and arterial changes Atherosclerosis is best considered as an ageing-related disease state that is almost ubiquitously present in older humans. The longer one lives, the chance of developing atherosclerotic disease increases. However, it is unclear as to how much the pathogenesis of atherosclerosis is dependent on a biological ageing process. Age-related cellular changes in the arterial wall might contribute in part to the development of the atherosclerotic plaque over time. A number of putative factors may lead to this including decreased ability of the endothelium to repair in the presence of injury, altered control of vascular smooth muscle proliferation and the interaction between smooth muscle and circulating lipoproteins. However, increasing age is only one of many important risk factors for the development of atherosclerosis. Changes in other cardiovascular risk factors invariably occur with ageing, and make a major contribution to the accelerated formation of atherosclerotic plaque. Occlusive disease due to atherosclerosis is more frequent at the origin of the internal carotid artery, the carotid siphon, the proximal middle cerebral artery, the proximal anterior carotid artery and the proximal basilar artery. Artery-to-artery embolization occurs as an end result of severe atherosclerosis of these vessels leading to well-known stroke syndromes. The increasing incidence of stroke with age is thus a function of an ageing vascular system with both intrinsic and extrinsic biological mechanisms at play. Another important effect of age-related arterial changes and disease such as atherosclerosis is a loss of elasticity and distensibility of vessels. This may lead to decreased arterial compliance predisposing the affected individual to systolic hypertensive disease. The potentially serious effects of systolic hypertension on the elderly brain include stroke, white matter disease and cognitive decline. Ageing and cerebral microvasculature Although the larger vessels described previously are responsible for maintaining overall cerebral blood flow, the cerebral microvasculature is chiefly responsible for the important function of providing active nutrient substrate for the neuron. There is a good correlation between the extent of this capillary network and the activity of functioning neurons in the cerebral cortex.4

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

155

These microvessels have structural properties that provide the basis for the Blood–Brain Barrier (BBB). The BBB is comprised of a continuous endothelium lacking fenestrations, thus acting as a very selective barrier to bloodborne substances reaching the central nervous system. Pinocytotic vesicles are found in very low numbers in the endothelium indicating that the BBB is comparatively less permeable than other organ endothelium. The BBB allows the transport of lipid-soluble and water-soluble substances either by passive (diffusion) or active transport. The cellular components of the BBB possess specialized carrier processes to provide adequate transport of nutrients, hormones and neurotransmitter peptides. Ageing and blood–brain barrier A number of changes in the Blood–Brain Barrier with the ageing process in animal and human models have been described. Thickening of the basement membrane in the ageing rat 5 together with other changes including loss of capillary endothelial cells and elongation of remaining endothelial cells 6 have been observed. A reduction of mitochondrial numbers has also been described in the endothelial cells of the BBB in the monkey but not the rat, leading the authors to postulate that some BBB changes with ageing may be species specific.6,7 In the human brain, examination of biopsy material showed thinning of white matter capillaries with age possibly related to loss of pericytes and thinning of endothelial cytoplasm.8 There is no evidence to date in humans that endothelial mitochondrial numbers decline with age or that significant changes occur in membrane permeability characteristics such as junctional gaps and pinocytotic vesicle density. In the rat model, alterations in BBB transport function include a decrease in BBB choline transport with ageing and decreased brain glucose influx.6 However there is no conclusive evidence in the literature that age-related permeability of the BBB is altered significantly in the absence of neurological or vascular disease. Although most reports have concentrated on changes in the afferent microvasculature (arterioles and capillaries) in ageing, few have described specific age-related changes in the efferent microvasculature (venules). These include the observation of a non-inflammatory mural thickening of the venular system due to collagenosis, predominantly in the periventricular region.9 Ageing, disease and cerebral microvasculature The cerebral microvasculature has been implicated in diseases such as Alzheimer’s disease (AD), cerebral amyloid angiopathy and potential disease states such as leukoaraiosis (deep white matter change). However, the role of the microvessels and the BBB in these disorders is a matter of ongoing research and controversy. It remains difficult to separate out the effects of intrinsic ageing and disease on microvascular changes. Alzheimer’s disease increases in prevalence with every decade of life in the elderly population. Intense debate and research continues in the search for

156

THE AGEING BRAIN

the aetiological factors responsible for the development of the disease. It is tempting to consider that age-related changes to cerebral microvessels may have some role to play in the pathogenesis of AD. It has been proposed that a breakdown of BBB may be an essential step in the pathogenesis of AD.10 Several investigators have hypothesised that microvascular injury secondary to amyloid deposition leads to leakage of substances that may cause neuronal injury.11–14 However, other investigators have demonstrated a lack of BBB alteration in either AD or non-demented controls by immunohistochemical.15 Similarly, in a post-mortem immunohistochemical study of AD compared to vascular dementia, age-matched controls, other neurodegenerative disorders and young controls, albumin leakage in the neuropil was demonstrated in all groups with no statistically significant difference between groups.16 These authors were also unable to consistently detect significant amounts of other serum proteins in the neuropil such as IgG and complement C3c. They concluded that alteration in BBB was neither a primary nor a consistent event in AD. In summary, in spite of ongoing interest in the field, it is still undecided whether BBB alterations are causally related to AD or merely an epiphenomenon. Amyloid Angiopathy: The microvasculature of the ageing brain is susceptible to the development of a specific disease termed cerebral amyloid angiopathy (CAA). This is characterized by the deposition of fibrillar amyloid material in the arteriolar media, gradually replacing the smooth muscle component of the arteriole. This leads to a weakened arterial wall that is susceptible to BBB dysfunction, as well as rupture leading to lobar haemorrhages. The deposition of vascular amyloid is extensive in cases of Alzheimer’s dementia, but is also associated with other conditions such as radiation injury and vasculitis. Amyloid deposition in CAA may lead to small brain infarcts due to microvascular occlusion, and thus may contribute to a dementing syndrome.17 The study of CAA in familial forms of the disease may provide important insights into the mechanistic links between ageing, microvascular pathology and disease states of the brain. Leukoaraiosis is the radiological finding of white matter changes that are commonly visible in CT scans or MRI of brains of elderly people. These changes may contribute to a small extent to age-related decline in intellectual function.18 Some researchers have proposed that these changes may be related to disease processes such as chronic hypertension, diabetes mellitus or even AD.19 They hypothesize that these disease processes may lead to disruption of the BBB, leading to protein leakage into white matter. Others implicate “periventricular venous collagenosis” in the pathogenesis of this condition on finding that venous changes were correlated with the presence of white matter hyperintensity on magnetic resonance imaging, whereas only mild agerelated change was observed in the afferent micro-vessels.20 Their findings led them to hypothesise that deep venous occlusion as an age-related process may lead to a long-term increase in vascular resistance and insidious oedema in surrounding tissue, contributing in part to the development of white matter

157

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

lesions. However, these findings have not yet been duplicated in other studies. Investigators who have used gadolinium-MRI and diffusion MRI techniques have provided conflicting results regarding a role for the BBB in pathogenesis of white matter lesions, although the methods used are qualitatively different.21,22 Ageing and Cerebrovascular Physiology Cerebral Blood Flow and metabolism — physiological concepts Cerebral Blood Flow (CBF) Great advances have been made into the research of CBF due to refinement in existing techniques of measurement including positron emission tomography (PET), 133Xenon inhalation computed tomography (Xenon-CT) and single positron emission computed tomography (SPECT), and the development of new techniques such as perfusion/diffusion magnetic resonance imaging (DWI/PWI MR), magnetic resonance spectroscopy (MRS) and transcranial Doppler. Determinants of Cerebral Blood Flow Average CBF is around 60 ml/min/100 g of brain tissue in the resting adult under physiological conditions with the cerebral grey matter receiving the bulk of supply. Gender differences have been described in global CBF,

Intracranial Pressure

artery

Arteriolar bed

Auto-regulation

vein

Synaptic Activity

Figure 1. Schematic summary of factors involved in regulating Cerebral Blood Flow (CBF).

158

THE AGEING BRAIN

with women of pre-menopausal age showing higher levels than similar aged men.23,24 This difference begins to disappear after the fourth decade of life, suggesting a possible role for oestrogen in augmenting CBF. CBF is directly related to cerebral perfusion pressure and inversely related to cerebral vascular resistance (Figure 1). Regional CBF is also determined by the level of synaptic activity and consequently regional cerebral metabolism (refer section on cerebral metabolism). Other factors playing an important role in regulation of CBF include intracranial pressure (ICP) and blood viscosity.25 Cerebral perfusion pressure Cerebral perfusion pressure is the difference between arterial inflow pressure and venous outflow pressure and represents the driving pressure for CBF. The mean perfusion pressure in humans varies between 50 to 150 mm Hg (autoregulatory range) without affecting cerebral blood flow due to the phenomenon of cerebral autoregulation (Figure 2).

CBF (ml/100 g/min)

Cerebral vascular resistance Cerebral vascular resistance (CVR) is primarily a function of the vessel radius such that even small changes in luminal diameter (via constriction or dilatation) can have major effects on resistance. The larger extracranial vessels and the intracranial pial vessels contribute to about half the total cerebrovascular resistance. CVR is controlled largely by cerebral autoregulatory mechanisms such that an increase in CVR leads to a reduction in CBF. Vascular resistance also depends to a certain extent on the viscosity of blood. The factors

Figure 2. Idealized cerebral pressure–flow curve.

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

159

that may play an important role in determining blood viscosity include shear (flow) rate, red cell rheology, platelet aggregability, plasma protein (fibrinogen) levels. An increase in blood viscosity leads to reduction in CBF.26 Reduction in blood flow may lead to further increase in viscosity leading to a vicious cycle of events. Cerebral autoregulation Cerebral autoregulation is achieved by the tight coupling of cerebral perfusion pressure and the diameter responses of the arteriolar system in an attempt to stabilise CBF in the presence of fluctuations in systemic arterial pressure. Within the autoregulatory range, CBF remains constant due to cerebral vasoconstriction when perfusion pressure increases as a result of systemic arterial hypertension. Neural, metabolic, myogenic and endothelial mechanisms have been postulated to explain the phenomenon of autoregulation.27 It appears that the autoregulation of pial vessels may be predominantly due to neurogenic control whereas the intracerebral vessels may well be under the influence of local metabolic phenomena such as changes in arterial CO 2 tension. Autoregulatory control of cerebral pial vessels may be under the influence of adrenergic and trigeminovascular systems although this is the subject of further study.28,29 Parasympathetic innervation of cerebral blood vessels has assumed greater importance in the last decade, contributing to neural vasodilation by means of nitric oxide, acetylcholine, glutamate and other agonists. However, investigators using animal models have not provided evidence for a major role for this system in resting and autoregulatory control of CBF. Smaller intracerebral vessels are extremely sensitive to metabolic influences such as arterial pCO2 tension. CBF may change by up to 5% with every mmHg change in arterial pCO2 within physiological ranges (30–60 mmHg). Arterial response to CO2 may be lessened in situations such as hypotension or cerebral ischaemia.30 The effect of CO2 is mediated by the autonomic supply of the vessels and perivascular cerebrospinal fluid pH. Arterial pO2 and Hydrogen ion also play a role in CBF regulation. The myogenic theory holds that changes in transmural pressure may directly affect the tone of the vascular smooth muscle leading to vasoconstriction or vasodilation in response to increase or decrease in systemic arterial pressure respectively.27,31 More recently, investigators have shown that such smooth muscle response to pressure changes may be dependent on the presence of intact endothelium.32 This may be linked to the release of contractile substances from the endothelium rather than a tonic inhibition of endothelium-derived relaxing factor (EDRF or Nitric Oxide).33 Intracranial pressure Intracranial pressure (ICP) is an important determinant of cerebral perfusion. A rise in ICP leads to compression of intracerebral vessels causing a decrease in perfusion. Conversely, a fall in ICP leads to an increase in cerebral per-

160

THE AGEING BRAIN

fusion. This mechanism appears to be important in maintaining relatively constant cerebral perfusion in states of positive and negative “g” and physiological states of straining. Extreme rises in ICP (> 30 mm Hg) can lead to stimulation of the brainstem vasomotor centre, leading to reflex increase in systemic arterial pressure in order to maintain adequate cerebral perfusion. Cerebral metabolism The brain has a high demand for oxygen and glucose in order to maintain optimal neuronal function even at rest. There are a number of unique features regarding the activity and metabolism of the brain. More than a century ago, it was postulated that cerebral metabolism was closely coupled with the level of pre-synaptic activity in the brain.34 Results of studies using various techniques appear largely to support this hypothesis.35–38 This coupling may (hypothetically) occur as a result of a glutamate-mediated uptake of glucose into peri-synaptic astrocytes leading to lactate production. Lactate can then be used as energy substrate in the pre-synaptic vesicle. Cerebral metabolism and CBF are also closely coupled with functional neuronal activity as an intermediate link, suggesting that cerebral metabolism may be an important determinant of CBF.34,39,40 This coupling may be mediated by a number of vasodilatory metabolites produced by the neuron (Vasoactive Intestinal Peptide, Nitric Oxide etc) which serve to increase regional CBF by acting on local small vessels as well as upstream resistance vessels.41,42 The coupling mechanism between neuronal activity and cerebral metabolism may be altered in disease states such as cerebral ischaemia. Ageing and cerebrovascular haemodynamics Controversy exists regarding the effect of “normal ageing” on physiological parameters such as CBF and cerebral metabolism. As with other aspects of “normal ageing” and disease, the difficulty in establishing a clear relationship between ageing and cerebrovascular physiology has been partly due to the lack of a clear definition of “normal ageing”. Most studies examining this relationship have been cross-sectional in nature (rather than longitudinal) comparing younger to older age groups. Cross-sectional approaches lead to difficulty in differentiating between genuine age-related change and cohort effects unrelated to age. A great deal of uncertainty still exists about the actual mechanisms involved in the perceived changes in CBF and cerebral metabolism. Ageing and CBF Studies of ageing and CBF have been mostly cross-sectional in nature. “Normal ageing” in most of these studies refers to subjects who are relatively free of vascular disease or risk factors for vascular disease. The results of most of these studies, utilizing different techniques, seem to point towards a reduction in mean CBF values with increasing age. These reductions have not only been demonstrated in older age groups and may begin in the third or fourth dec-

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

Table 1.

161

Cross-sectional studies of cerebral blood flow and perfusion in human ageing.

Investigators

Year

Technique

Reduction of CBF with Ageing Kety44 1956 Wang et al.46 1975 Obrist48 1979 De Koninck et al.50 1977 Meyer et al.52 1978 Yamaguchi et al.54 1979 Naritomi et al.56 1979 Thomas et al.58 1979 Yamamoto et al.60 1980 Melamed et al.62 1980 Davis et al.64 1983 Iwata et al.66 1986 Zemcov et al.67 1984 Gur et al.23 1987 Hagstadius et al.68 1989 Leenders et al.69 1990 Martin et al.70 1991 Markus et al.71 1993 Krausz et al.72 1998 Bentourkia et al.73 2000 Scheel et al.74 2000

Nitrous oxide 133 Xenon inhalation 133 Xenon inhalation 133 Xenon intracarotid 133 Xenon inhalation 133 Xenon inhalation 133 Xenon inhalation 133 Xenon intravenous 133 Xenon inhalation 133 Xenon inhalation 133 Xenon inhalation 131 Xenon CT 133 Xenon inhalation 133 Xenon inhalation 133 Xenon inhalation P.E.T. P.E.T. 99mTc-HMPAO SPECT 99mTc-HMPAO SPECT P.E.T. Colour duplex sonography

No Reduction of CBF with Ageing Shieve and Wilson45 1953 Shenkin et al.47 1953 Gordan et al.49 1956 Lassen et al.51 1960 Dastur et al.53 1963 Sokoloff et al.55 1966 Aizawa et al.57 1961 Gottstein et al.59 1979 Waldemar61 1991 Takada et al.63 1992 Meltzer et al.65 2000

Nitrous oxide Nitrous oxide Nitrous oxide 85Kr Nitrous oxide Nitrous oxide Nitrous oxide Nitrous oxide 99mTc-HMPAO SPECT P.E.T. P.E.T.

P.E.T. – positron emission tomography; 99mTc-HMPAO SPECT – 99mTechnetiumhexamethylpropyleneamine oxime single photon emission computed tomography.

ade of life. However there are a significant number of cross-sectional studies with contradictory reports, with some evidence to support either a lack of or a non-significant decline in CBF with increasing age (Table 1). This controversy probably highlights the difficulty with inherent selection bias in using cross-sectional designs, different subject populations and differing techniques of measurement of CBF. In studies where CBF reduction with age has been demonstrated, the reductions are more likely to involve cerebral grey matter

162

THE AGEING BRAIN

with an annual estimated decline of approximately 0.5 ml/min/100 g of brain tissue. Evidence from current literature does not seem to support reduction in white matter flow with healthy ageing. Few longitudinal data exist with regards to CBF in ageing, presumably due to the complexity and labour-intensive nature of CBF studies. A limited analysis of CBF data in eight subjects over a 11-year period showed some reduction in CBF in the absence of significant change in mean arterial blood pressure, with a concomitant increase in cerebrovascular resistance.43 In a large cohort of healthy volunteers, volunteers with vascular risks, and patients with stroke or TIA, both cross-sectional and longitudinal analyses revealed significant reduction in gray matter flow with age in all groups, with more decline demonstrated in groups with risk factors or stroke/TIA.24 The estimated decline in cerebral gray matter in this study was slightly higher than in cross-sectional studies, estimated at approximately 1.0 ml/min/100 g of brain tissue. Ageing and cerebral metabolism Data regarding changes in the rate of cerebral metabolism with human ageing are even less conclusive. Most of these data have been derived from positron emission tomography studies (PET) measuring cerebral metabolic rate for oxygen (CMRO2) and cerebral metabolic rate of glucose (CMRG) in resting state (Table 2). Some investigators hold that CMRO2 declines in healthy ageing in parallel with decline in CBF. However a number of researchers have failed to demonstrate reduction in mean CMRO2 levels with age. Similar controversy exists for CMRG, with contradictory reports in the literature regarding changes in glucose metabolism in “normal” ageing. The physiological state of ‘coupling’ of cerebral glucose metabolism and CBF is thought to be largely maintained with increasing age in the absence of disease. Ageing, cerebrovascular reactivity and autoregulation Cerebrovascular reactivity in response to changes in systemic arterial pressure or to other parameters such as arterial CO2 and arterial O2 levels has been investigated in only a few studies. Researchers examining CBF change with posture showed some increase in the “dysautoregulation index” (defined as CBF decrease per unit fall in effective perfusion pressure on head-up tilt). 88,89 Others have demonstrated a decline in CO2 reactivity with ageing using quantitative CBF techniques.45,54,60 However, other groups using either quantitative CBF techniques64 or transcranial doppler methods90–92 have failed to demonstrate significant change in cerebral autoregulation with healthy ageing. Ageing, disease and cerebrovascular haemodynamics The more “common” form of ageing is one that is accompanied by the development of disease as compared to the less common “healthy” (supranormal) form of ageing. Studies of cerebrovascular haemodynamics have mostly been performed in the latter group in order to gain insight into “normal”processes. However, it is the study of the former group that is more relevant to the

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

Table 2.

163

Cross-sectional studies of cerebral metabolism in human ageing.

Investigator

Year

Method

Comment

Reduced Cerebral Metabolism with Ageing Sokoloff et al.75 Kuhl et al.76 Pantano et al.77

1975 1984 1984

Dastur78

1985

Yamaguchi et al.54 1986 Leenders et al.69

1990

Marchal et al.79

1992

Takada et al.63 De Santi et al.80

1992 1995

Eberling et al.81 1995 Bentourkia et al.73 2000

Nitrous oxide Reduction in CMRO2, CMRG and CBF 18FDG PET Reduction in whole brain mean CMRG H215O_PET Non-linear reduction in mean gray CBF and CMRO2; white matter unchanged 18FDG PET Reduction in CMRG; no reduction in CBF and CMRO2 H215O_PET Reduction in CMRO2; CBF variable and less age-dependent H215O_PET Coupled reduction in CMRO2 and CBF in pure gray and white matter H215O_PET Reduction in CMRO2 and CBV in gray matter; CBF variable H215O_PET Reduction in CMRO2 but not CBF 18FDG PET Reduction in CMRG in frontal and temporal lobes 18FDG PET Reduction in CMRG in temporal cortex 18FDG PET Coupled reduction in CBF and CMRG

Cerebral Metabolism Unchanged with Ageing Duara et al.82

1984

18FDG

Cutler et al.83 De Leon et al.84

1986 1987

18FDG PET 11CDG PET

Horwitz et al.85

1987

18FDG

Schlageter et al.86 1987 Burns et al.87 1992

PET

PET

18FDG PET H215O_PET

No reduction in mean or regional CMRG; resting CMRG poorly correlated with cognitive tests No reduction in mean CMRG No reduction in absolute regional CMRG No reduction in mean CMRG; age associated reduction in regional cerebral functional interaction No reduction of global CMRG Trend for reduction in CMRO2 only in parietal lobe; effect less significant with advancing age

CMRO2: cerebral metabolic rate for oxygen; PET: positron emission tomography; 18FDG: 18-Fluoro-deoxyglucose; CMRG: cerebral metabolic rate for glucose; H215O: 15-oxygen labelled water; CBF: cerebral blood flow; 11CDG: 11-carbondeoxyglucose; CBV: cerebral blood volume.

majority of the ageing population and provides the opportunity to examine the link between pathological vascular changes and a variety of disease states. Certain vascular risk factors such as hypertension, diabetes, atrial fibrillation and hypercholesterolaemia are common in the elderly population. The incidence of disease states such as stroke and dementia increases with age. The questions then arise as to whether a causal link may exist between vascular

164

THE AGEING BRAIN

risk and disease, and whether age plays a part in determining the expression of such disease. Dementia (especially Alzheimer’s dementia) is a prototype disease of the elderly that is the subject of intense interest and study with regards to putative vascular aetiologies. Epidemiological evidence exists from population-based studies describing associations of vascular risk factors with prevalent dementia.93,94 The results of studies of cerebral blood flow and metabolism performed in people with dementia have indicated reductions in regional cerebral metabolism with concomitant reduction in regional CBF.24,78,95–97 These observations have led to the debate whether the reduction in cerebral perfusion (putatively as a result of vascular risk) plays an aetiological role in the development of dementia,98 or whether CBF reduction is a consequence of neuronal damage and the vascular disease merely an epiphenomenon.96 Age by itself remains a crucial risk factor in the development of dementia.99 Neuronal loss, oxidative stress, reduction in vascular reserve and impaired repair mechanisms may all play a role in reducing the reserve of the ageing brain thus leaving it vulnerable to injury and disease. Summary A wide variety of changes are seen in the human cerebrovascular system on a structural and functional basis as one gets older. However, there is a significant amount of disagreement about the true nature of age-related vascular changes in the brain. This is partly due to the increasing complexity in defining the boundaries of ‘normal’ ageing. The limits of human ageing are being pushed further with every passing decade of medical scientific progress. As people get to live longer, they become exposed to a larger number of vascular risks, predisposing them to accrue more vascular change. A more fruitful approach towards unravelling this problem may be to identify the links between such vascular risks, ageing and disease. This is likely to become a reality with the increasingly sophisticated research methods now available. The ultimate goal of such research is to develop ways to prevent and treat important age-related neurological disorders. References 1. 2. 3.

Bouissou H, Emery MC, Sorbara R. Age related changes of the middle cerebral artery and a comparison with the radial and coronary artery. Angiology. 1975; 26(3):257–68. Fang HCH. Observations on aging characteristics of cerebral blood vessels, macroscopic and microscopic features. In: Terry RD, Gershon, S., editors. Neurobiology of aging. New York: Raven Press, 1976; 155–166. Spangler KM, Challa VR, Moody DM, Bell MA. Arteriolar tortuosity of the white matter in aging and hypertension. A microradiographic study. J Neuropath Exp Neurol. 1994; 53:22–26.

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

165

4. Mann DM, Eaves NR, Marcyniuk B, Yates PO. Quantitative changes in cerebral cortical microvasculature in ageing and dementia. Neurobiol Aging. 1986; 7: 321–330. 5. Donahue J, Pappas G.D. The fine structure of capillaries in the cerebral cortex of the rat at various stages of development. Am J Anat. 1961; 108:331–348. 6. Mooradian AD. Effect of aging on the blood-brain barrier. Neurobiol Aging. 1988; 9:31–39. 7. Burns EM, Kruckeberg TW, Comerford LE, Buschmann MT. Thinning of capillary walls and declining numbers of endothelial mitochondria in the cerebral cortex of the aging primate, Macaca nemestrina. J Gerontol. 1979; 34:642–650. 8. Stewart PA, Magliocco M, Hayakawa K, Farrell CL, Del Maestro RF, Girvin J, Kaufmann JC, Vinters HV, Gilbert J. A quantitative analysis of blood-brain barrier ultrastructure in the aging human. Microvasc Res. 1987; 33:270–282. 9. Moody DM, Brown WR, Challa VR, Ghazi-Birry HS, Reboussin DM. Cerebral microvascular alterations in aging, leukoaraiosis, and Alzheimer’s disease. Ann NY Acad Sci. 1997; 826:103–116. 10. Hardy JA, Mann DM, Wester P, Winblad B. An integrative hypothesis concerning the pathogenesis and progression of Alzheimer’s disease. Neurobiol Aging. 1986; 7:489–502. 11. Wisniewski HM, Vorbrodt AW, Wegiel J. Amyloid angiopathy and blood-brain barrier changes in Alzheimer’s disease. Ann NY Acad Sci. 1997; 826:161–172. 12. Alafuzoff I, Adolfsson R, Grundke-Iqbal I, Winblad B. Blood-brain barrier in Alzheimer dementia and in non-demented elderly. An immunocytochemical study. Acta Neuropathol. 1987; 73:160–166. 13. Kalaria RN. The blood-brain barrier and cerebrovascular pathology in Alzheimer’s disease. Ann NY Acad Sci .1999; 893:113–125. 14. Halliday G, Robinson SR, Shepherd C, Kril J. Alzheimer’s disease and inflammation: a review of cellular and therapeutic mechanisms. Clin Exp Pharmacol P. 2000; 27:1–8. 15. Rozemuller JM, Eikelenboom P, Kamphorst W, Stam FC. Lack of evidence for dysfunction of the blood-brain barrier in Alzheimer’s disease: an immunohistochemical study. Neurobiol Aging. 1988; 9:383–391. 16. Munoz DG, Erkinjuntti T, Gaytan-Garcia S, Hachinski V. Serum protein leakage in Alzheimer’s disease revisited. Ann NY Acad Sci. 1997; 826:173–189. 17. Vinters HV, Natte R, Maat-Schieman ML, van Duinen SG, Hegeman-Kleinn I, Welling-Graafland C, Haan J, Roos RA. Secondary microvascular degeneration in amyloid angiopathy of patients with hereditary cerebral hemorrhage with amyloidosis, Dutch type (HCHWA- D). Acta Neuropathol. 1998; 95:235–244. 18. Garde E, Mortensen EL, Krabbe K, Rostrup E, Larsson HB. Relation between age-related decline in intelligence and cerebral white- matter hyperintensities in healthy octogenarians: a longitudinal study. Lancet. 2000; 356(9230):628–634. 19. Pantoni L, Garcia JH. Cognitive impairment and cellular/vascular changes in the cerebral white matter. Ann NY Acad Sci. 1997; 826:92–102. 20. Moody DM, Brown WR, Challa VR, Anderson RL. Periventricular venous collagenosis: association with leukoaraiosis. Radiology. 1995; 194:469–476. 21. Bronge L, Wahlund LO. White matter lesions in dementia: an MRI study on blood-brain barrier dysfunction. Dement Geriatr Cogn. 2000; 11:263–267. 22. Werring DJ, Brassat D, Droogan AG, Clark CA, Symms MR, Barker GJ, MacManus DG, Thompson AJ, Miller DH. The pathogenesis of lesions and normalappearing white matter changes in multiple sclerosis: a serial diffusion MRI study. Brain. 2000; 123:1667–1676. 23. Gur RC, Gur RE, Obrist WD, Skolnick BE, Reivich M. Age and regional cerebral blood flow at rest and during cognitive activity. Arch Gen Psychiat. 1987; 44: 617–621.

166

THE AGEING BRAIN

24. Shaw TG, Mortel KF, Meyer JS, Rogers RL, Hardenberg J, Cutaia MM. Cerebral blood flow changes in benign aging and cerebrovascular disease. Neurology. 1984; 34:855–862. 25. Thomas DJ. Whole blood viscosity and cerebral blood flow. Stroke. 1982; 13: 285–287. 26. Humphrey PR, Du Boulay GH, Marshall J, Pearson TC, Russell RW, Symon L, Wetherley-Mein G, Zilkha E. Cerebral blood-flow and viscosity in relative polycythaemia. Lancet. 1979; 2(8148):873–877. 27. Paulson OB, Strandgaard S, Edvinsson L. Cerebral autoregulation. Cerebrovas Brain Met. 1990; 2:161–192. 28. Edvinsson L, Owman C, Siesjo B. Physiological role of cerebrovascular sympathetic nerves in the autoregulation of cerebral blood flow. Brain Res. 1976; 117: 519–523. 29. Edvinsson L, Degueurce A, Duverger D, MacKenzie ET, Scatton B. Central serotonergic nerves project to the pial vessels of the brain. Nature. 1983; 306(5938):55–57. 30. Hossmann KA. Treatment of experimental cerebral ischemia. J Cerebr Blood Met. 1982; 2:275–297. 31. Bayliss WM. On the local reaction of the arterial wall to changes of internal pressure. J Physiol. 1902:220–231. 32. Harder DR. Pressure-induced myogenic activation of cat cerebral arteries is dependent on intact endothelium. Circ Res. 1987; 60:102–107. 33. Harder DR, Sanchez-Ferrer C, Kauser K, Stekiel WJ, Rubanyi GM. Pressure releases a transferable endothelial contractile factor in cat cerebral arteries. Circ Res. 1989; 65:193–198. 34. Roy CS, Sherrington, C.S. On the regulation of the blood supply of the brain. J Physiol .1890; 11:85–108. 35. Swanson RA, Morton MM, Sagar SM, Sharp FR. Sensory stimulation induces local cerebral glycogenolysis: demonstration by autoradiography. Neuroscience. 1992; 51:451-461. 36. Jueptner M, Weiller C. Review: does measurement of regional cerebral blood flow reflect synaptic activity? Implications for PET and fMRI. Neuroimage. 1995; 2: 148–156. 37. Sibson NR, Dhankhar A, Mason GF, Rothman DL, Behar KL, Shulman RG. Stoichiometric coupling of brain glucose metabolism and glutamatergic neuronal activity. Proc Natl Acad Sci USA. 1998; 95:316-321. 38. Gerrits RJ, Raczynski C, Greene AS, Stein EA. Regional cerebral blood flow responses to variable frequency whisker stimulation: an autoradiographic analysis. Brain Res. 2000; 864:205–212. 39. Baron JC, Lebrun-Grandie P, Collard P, Crouzel C, Mestelan G, Bousser MG. Noninvasive measurement of blood flow, oxygen consumption, and glucose utilization in the same brain regions in man by positron emission tomography: concise communication. J Nucl Med. 1982; 23:391–399. 40. Baron JC, Rougemont D, Soussaline F, Bustany P, Crouzel C, Bousser MG, Comar D. Local interrelationships of cerebral oxygen consumption and glucose utilization in normal subjects and in ischemic stroke patients: a positron tomography study. J Cerebr Blood F Met. 1984; 4:140–149. 41. Akgoren N, Dalgaard P, Lauritzen M. Cerebral blood flow increases evoked by electrical stimulation of rat cerebellar cortex: relation to excitatory synaptic activity and nitric oxide synthesis. Brain Res. 1996; 710:204–214. 42. Ngai AC, Ko KR, Morii S, Winn HR. Effect of sciatic nerve stimulation on pial arterioles in rats. Am J Physiol. 1988; 254:H133–139. 43. Libow LS. Cerebral and electroencephalographic changes in elderly men. Rockville, US Dep. of Health, Education and Welfare, National Institute of Mental Health, 1971. Report No: (HSM) 71–9037.

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

167

44. Kety SS. Human cerebral blood flow and oxygen consumption as related to aging. Res Publ Assoc Res N. 1956; 35:31–35. 45. Schieve JF, Wilson, WP. The influence of age, anesthesia and cerebral arteriosclerosis on cerebral vascular reactivity to CO2. Am J Med. 1953; 15:171–174. 46. Wang HS, Busse EW. Correlates of regional blood flow in elderly community residents. In: Harper M, Jennett, B, Miller, D, Rowan, J, editors. Blood flow and metabolism in the brain. London: Churchill Livingstone, 1975; 17–18. 47. Shenkin HA, Novak, P., Golobuff, B., Soffe, A.M., Bortin, L. The effects of aging arteriosclerosis, and hypertension upon the cerebral circulation. J Clin Invest. 1953; 32:459–465. 48. Obrist WD. Cerebral circulatory changes in normal aging and dementia. In: Bayer Symposium VII;Brain function in old age. New York: Springer-Verlag, 1979; 278–287. 49. Gordan GS. Influence of steroids on cerebral metabolism in man. Recent Prog Horm Res. 1956; 12:153–174. 50. De Koninck WJ, Calay, R., Hongne, JC. CBF in elderly with chronic cerebral involvement. Acta Neurol Scand. 1977; (Suppl) 64:412–413. 51. Lassen NA, Feinberg, I., Lane, M.H. Bilateral studies of cerebral oxygen uptake in young and aged normal subjects and in patients with organic dementia. J Clin Invest. 1960; 39:491–500. 52. Meyer JS, Ishihara N, Deshmukh VD, Naritomi H, Sakai F, Hsu MC, Pollack P. Improved method for noninvasive measurement of regional cerebral blood flow by 133Xenon inhalation. Part I: description of method and normal values obtained in healthy volunteers. Stroke. 1978; 9:195–205. 53. Dastur DK, Lane MH, Hansen, DB. Effects of aging on cerebral circulation and metabolism in man. Washington DC: US Government Printing Office, 1963. USPHS publication no. 986. 54. Yamaguchi T, Kanno I, Uemura K, Shishido F, Inugami A, Ogawa T, Murakami M, Suzuki K. Reduction in regional cerebral metabolic rate of oxygen during human aging. Stroke. 1986; 17:1220–1228. 55. Sokoloff L. Cerebral circulatory and metabolic changes associated with aging. Res Publ Assoc Res N. 1966; 41:237–254. 56. Naritomi H, Meyer JS, Sakai F, Yamaguchi F, Shaw T. Effects of advancing age on regional cerebral blood flow. Studies in normal subjects and subjects with risk factors for atherothrombotic stroke. Arch Neurol. 1979; 36:410–416. 57. Aizawa T, Tazaki, Y., Gotoh, F. Cerebral circulation in cerebrovascular disease. World Neurol. 1961; 2:635–45. 58. Thomas DJ, Zilkha E, Redmond S, Du Boulay GH, Marshall J, Russell RW, Symon L. An intravenous 133xenon clearance technique for measuring cerebral blood flow. J Neurol Sci. 1979; 40:53–63. 59. Gottstein U, Held, K. Effects of aging on cerebral circulation and metabolism in man. Acta Neurol Scand. 1979; (Suppl) 72:54–55. 60. Yamamoto M, Meyer JS, Sakai F, Yamaguchi F. Aging and cerebral vasodilator responses to hypercarbia: responses in normal aging and in persons with risk factors for stroke. Arch Neurol. 1980; 37:489–496. 61. Waldemar G, Hasselbalch SG, Andersen AR, Delecluse F, Petersen P, Johnsen A, Paulson OB. 99mTc-d,l-HMPAO and SPECT of the brain in normal aging. J Cerebr Blood F Met. 1991; 11:508–521. 62. Melamed E, Lavy S, Bentin S, Cooper G, Rinot Y. Reduction in regional cerebral blood flow during normal aging in man. Stroke. 1980; 11:31–35. 63. Takada H, Nagata K, Hirata Y, Satoh Y, Watahiki Y, Sugawara J, Yokoyama E, Kondoh Y, Shishido F, Inugami A. Age-related decline of cerebral oxygen metabolism in normal population detected with positron emission tomography. Neurol Res. 1992; 14:128–131.

168

THE AGEING BRAIN

64. Davis SM, Ackerman RH, Correia JA, Alpert NM, Chang J, Buonanno F, Kelley RE, Rosner B, Taveras JM. Cerebral blood flow and cerebrovascular CO2 reactivity in stroke-age normal controls. Neurology. 1983; 33:391–399. 65. Meltzer CC, Cantwell MN, Greer PJ, Ben-Eliezer D, Smith G, Frank G, Kaye WH, Houck PR, Price JC. Does cerebral blood flow decline in healthy aging? A PET study with partial-volume correction. J Nucl Med. 2000; 41:1842–1848. 66. Iwata K, Harano H. Regional cerebral blood flow changes in aging. Acta Radiol Suppl. 1986; 369:440–443. 67. Zemcov A, Barclay L, Blass JP. Regional decline of cerebral blood flow with age in cognitively intact subjects. Neurobiol Aging. 1984; 5:1–6. 68. Hagstadius S, Risberg J. Regional cerebral blood flow characteristics and variations with age in resting normal subjects. Brain Cogn. 1989; 10:28–43. 69. Leenders KL, Perani D, Lammertsma AA, Heather JD, Buckingham P, Healy MJ, Gibbs JM, Wise RJ, Hatazawa J, Herold S. Cerebral blood flow, blood volume and oxygen utilization. Normal values and effect of age. Brain. 1990; 113:27–47. 70. Martin AJ, Friston KJ, Colebatch JG, Frackowiak RS. Decreases in regional cerebral blood flow with normal aging. J Cerebr Blood F Met. 1991; 11:684–689. 71. Markus HS, Ring H, Kouris K, Costa DC. Alterations in regional cerebral blood flow, with increased temporal interhemispheric asymmetries, in the normal elderly: an HMPAO SPECT study. Nucl Med Commun. 1993; 14:628–633. 72. Krausz Y, Bonne O, Gorfine M, Karger H, Lerer B, Chisin R. Age-related changes in brain perfusion of normal subjects detected by 99mTc-HMPAO SPECT. Neuroradiology. 1998; 40):428–434. 73. Bentourkia M, Bol A, Ivanoiu A, Labar D, Sibomana M, Coppens A, Michel C, Cosnard G, De Volder AG. Comparison of regional cerebral blood flow and glucose metabolism in the normal brain: effect of aging. J Neurol Sci. 2000; 181: 19–28. 74. Scheel P, Ruge C, Petruch UR, Schoning M. Color duplex measurement of cerebral blood flow volume in healthy adults. Stroke. 2000; 31:147–150. 75. Sokoloff L. Cerebral circulation and metabolism in the aged. Psychopharmacol Bull. 1975; 11:45-46. 76. Kuhl DE, Metter EJ, Riege WH, Hawkins RA. The effect of normal aging on patterns of local cerebral glucose utilization. Ann Neurol. 1984; 15(Suppl): S133–137. 77. Pantano P, Baron JC, Lebrun-Grandie P, Duquesnoy N, Bousser MG, Comar D. Regional cerebral blood flow and oxygen consumption in human aging. Stroke. 1984; 15:635–641. 78. Dastur DK. Cerebral blood flow and metabolism in normal human aging, pathological aging, and senile dementia. J Cerebr Blood F Met. 1985; 5:1–9. 79. Marchal G, Rioux P, Petit-Taboue MC, Sette G, Travere JM, Le Poec C, Courtheoux P, Derlon JM, Baron JC. Regional cerebral oxygen consumption, blood flow, and blood volume in healthy human aging. Arch Neurol. 1992; 49:1013–1020. 80. De Santi S, de Leon MJ, Convit A, Tarshish C, Rusinek H, Tsui WH, Sinaiko E, Wang GJ, Bartlet E, Volkow N. Age-related changes in brain: II. Positron emission tomography of frontal and temporal lobe glucose metabolism in normal subjects. Psychiatr Quart. 1995; 66:357–370. 81. Eberling JL, Nordahl TE, Kusubov N, Reed BR, Budinger TF, Jagust WJ. Reduced temporal lobe glucose metabolism in aging. J Neuroimaging. 1995; 5:178–182. 82. Duara R, Grady C, Haxby J, Ingvar D, Sokoloff L, Margolin RA, Manning RG, Cutler NR, Rapoport SI. Human brain glucose utilization and cognitive function in relation to age. Ann Neurol. 1984; 16:703–713. 83. Cutler NR. Cerebral metabolism as measured with positron emission tomography (PET) and [18F] 2-deoxy-D-glucose: healthy aging, Alzheimer’s disease and Down syndrome. Prog Neuro-psychoph. 1986; 10:309–321.

CEREBROVASCULAR SYSTEM AND THE AGEING BRAIN

169

84. de Leon MJ, George AE, Tomanelli J, Christman D, Kluger A, Miller J, Ferris SH, Fowler J, Brodie JD, van Gelder P. Positron emission tomography studies of normal aging: a replication of PET III and 18-FDG using PET VI and 11-CDG. Neurobiol Aging. 1987; 8:319–323. 85. Horwitz B. Brain metabolism and blood flow during aging. Electroen Clin Neuro Suppl. 1987; 39:396–402. 86. Schlageter NL, Horwitz B, Creasey H, Carson R, Duara R, Berg GW, Rapoport SI. Relation of measured brain glucose utilisation and cerebral atrophy in man. J Neurol Neurosur Ps. 1987; 50:779-785. 87. Burns A, Tyrrell P. Association of age with regional cerebral oxygen utilization: a positron emission tomography study. Age Ageing. 1992; 21:316–320. 88. Shinohara Y, Takagi S, Kobatake K. Effect of aging on CBF and autoregulation in normal subjects and CVD patients. Monogr Neural Sci. 1984; 11:204–209. 89. Oiwa K, Shimazu K, Tamura N, Hienuki M, Kim HT, Yamamoto T, Hamaguchi K. Effect of aging on cerebral blood flow autoregulation--with special reference to the role of the prostaglandins. Monogr Neural Sci. 1984; 11:210–215. 90. Kastrup A, Dichgans J, Niemeier M, Schabet M. Changes of cerebrovascular CO2 reactivity during normal aging. Stroke. 1998; 29:1311–1314. 91. Carey BJ, Eames PJ, Blake MJ, Panerai RB, Potter JF. Dynamic cerebral autoregulation is unaffected by aging. Stroke. 2000; 31(12):2895–2900. 92. Lipsitz LA, Mukai S, Hamner J, Gagnon M, Babikian V. Dynamic regulation of middle cerebral artery blood flow velocity in aging and hypertension. Stroke. 2000; 31(8):1897–1903. 93. Hofman A, Ott A, Breteler MM, Bots ML, Slooter AJ, van Harskamp F, van Duijn CN, Van Broeckhoven C, Grobbee DE. Atherosclerosis, apolipoprotein E, and prevalence of dementia and Alzheimer’s disease in the Rotterdam Study. Lancet. 1997; 349(9046):151–154. 94. Elwood PC, Pickering J, Gallacher JE. Cognitive function and blood rheology: results from the Caerphilly cohort of older men. Age Ageing. 2001; 30: 135–139. 95. Benson DF, Kuhl DE, Hawkins RA, Phelps ME, Cummings JL, Tsai SY. The fluorodeoxyglucose 18F scan in Alzheimer’s disease and multi- infarct dementia. Arch Neurol. 1983; 40:711–714. 96. Jagust WJ, Eberling JL, Reed BR, Mathis CA, Budinger TF. Clinical studies of cerebral blood flow in Alzheimer’s disease. Ann NY Acad Sci. 1997; 826:254–262. 97. Tohgi H, Yonezawa H, Takahashi S, Sato N, Kato E, Kudo M, Hatano K, Sasaki T. Cerebral blood flow and oxygen metabolism in senile dementia of Alzheimer’s type and vascular dementia with deep white matter changes. Neuroradiology. 1998; 40:131–137. 98. de la Torre JC, Stefano GB. Evidence that Alzheimer’s disease is a microvascular disorder: the role of constitutive nitric oxide. Brain Res Rev. 2000; 34: 119–136. 99. Munoz DG, Feldman H. Causes of Alzheimer’s disease. Cmaj. 2000; 162: 65–72.

SECTION III FACTORS INFLUENCING BRAIN AGEING

Chapter 10 THE MOLECULAR BASIS OF ALZHEIMER’S DISEASE AND FRONTOTEMPORAL DEMENTIA John B.J. Kwok and Peter R. Schofield*

Introduction Alzheimer’s disease (AD) was thought to be an intractable disorder. Yet, genetic analyses have successfully uncovered three genes, the amyloid precursor protein (APP) gene, presenilin-1 (PS-1) and the presenilin-2 (PS-2) gene which can cause early-onset Alzheimer’s disease. Functional analysis of these genes and gene mutations has highlighted the importance of the amyloid cascade hypothesis to our understanding of the disease process. Moreover, the correlation of mutations in the AD genes with specific clinical outcomes and variant neuropathology has allowed us to detect the existence of modifying factors which alter the course of the disease. More recently, the tau gene has been identified as the causative agent for another form of dementia, fronto-temporal dementia. The challenge now is to determine the enigmatic relationship between tau and the AD genes and to determine whether there is a common neurodegenerative mechanism.

*To whom correspondence should be addressed.

174

THE AGEING BRAIN

A Common Affliction Alzheimer’s disease (AD) is a devastating affliction of the brain. A patient will suffer an irreversible deterioration of intellectual abilities involving memory loss, impairment of judgement and reasoning, as well as personality changes in later stages. Ultimately, the condition is fatal due to failure of physical function.1 AD is the most common cause of senile dementia, accounting for 50% of dementia cases. The disease will strike an estimated one in ten persons over the age of 65 years and increases to nearly one in two of those over 80.2 Despite intense basic scientific and pharmacological investigations, there are still no truly effective therapeutic drugs for AD. With such an emotional and financial cost to patients, and to a rapidly ageing society, there is a pressing need for greater understanding of this disease. AD is distinct from other forms of dementia by key pathological features in the brain. Firstly, there are a large number of senile plaques in the extracellular spaces between neurones. The plaques are spherical deposits that consist of central cores of amyloid beta (Aβ) fibrils, surrounded by degenerated neurites and glial cells.2 The Aβ peptide consists of a sequence of hydrophobic amino acids of 39 to 43 amino acids in length.3,4 The peptide is derived from proteolytic cleavage of a larger multidomain glycoprotein, the amyloid precursor protein (APP).5 Secondly, there are neurofibrillary tangles (NFTs) found within neurones. The major components of NFTs are paired helical filaments, which in turn are composed of hyperphosphorylated form of tau, a microtubule associated protein2. Finally, there is extensive neuronal loss in the cerebral cortex and hippocampus, which is directly responsible for the cognitive decline.2 Amyloid Cascade The exact contribution of each neuropathological feature to the clinical symptoms of AD is unclear.6 However, several studies have suggested that the Aβ deposits can be directly neurotoxic, in part through the generation of free radicals6 whose effects can be attenuated by the addition of antioxidants.6 Other studies suggest that Aβ can disrupt ionic homeostasis and lead to severe effects on cellular processes and induction of neuronal cell death6. The amyloid cascade hypothesis postulates that the deposition of Aβ is the central causative event in AD and that the NFTs, cell death and dementia follow as a direct result of this deposition7 as shown in Figure 1. The amyloid cascade hypothesis predicts that mutations in genes, which lead to the overproduction of APP or subsequent mismetabolism, would underlie the genetic basis of AD.

ALZHEIMER’S DISEASE AND FRONTOTEMPORAL DEMENTIA

175

Figure 1. Amyloid Cascade hypothesis suggests the aberrant metabolism of APP molecule to form the longer peptide isoform, Aβ1-42 and greater deposition of senile plaques. This may arise as a result of age-related factors or genetic mutations. Neurofibrillary tangles and neuronal cell death are secondary events to the initial amyloid production and deposition.

Amyloid Precursor Protein (APP) Gene The Aβ peptide is cleaved from APP via a series of proteolytic steps mediated by enzymes called secretases (Figure 2). Cleavage of APP with the β- and γ-secretase will generate an intact Aβ peptide. Cleavage by the α-secretase within the peptide sequence will prevent the formation of Aβ.5 All the mutations appear to cluster within or adjacent to the sequence which encode Aβ peptide as shown in Figure 2. Each mutation has an effect, either on the metabolism of APP or the nature of the Aβ sequence itself, but ultimately all mutations increase the rate of amyloid deposition.8 For example, the Swedish double mutation results in increased secretion of the normal 40 amino-acid peptide (Aβ1-40) and a longer 42 amino-acid isoform of Aβ (Aβ1-42), most probably by enhancing β-secretase activity. The cerebral angiopathy with amyloidosis (CAA) mutation has been shown to diminish α-secretase activity, thus increasing the secretion of both forms of intact Aβs. Finally, there are a series of mutations which cluster around the γ-secretase site (Figure 2). These mutations include the London mutation at codon 717 (Val to Ile),

176

THE AGEING BRAIN

Figure 2. Schematic diagram of APP. The protein is a multi-domain cell-surface molecule. The Aβ region (grey box and circles) contains part of the transmembrane domain and part of the extracellular domain of APP. Secretase cleavage sites are indicated by open arrows. Familial mutations that cause AD are indicated.

which alters the conformation of the γ-secretase recognition site so that APP is preferentially cleaved to produce the Aβ1-42 isoform.9 Presenilin Genes A major locus responsible for early onset AD (EOAD), presenilin-1 (PS-1), was shown to map to the long arm of chromosome 14.10 Together with its homologue, presenilin-2 (PS-2), on chromosome 1,11 pathogenic mutations in

ALZHEIMER’S DISEASE AND FRONTOTEMPORAL DEMENTIA

177

Figure 3. Schematic diagram of PS-1. The protein has eight hydrophobic transmembrane spanning domains and a large hydrophilic loop. The protein is cleaved within the hydrophilic loop by an unknown protease (presenilinase) and a caspase. The majority of mutations detected in EOAD pedigrees and cases are missense mutations (mutant amino indicated in grey). One special class of mutations which deletes exon 9 (boundaries indicated by open arrows) is associated with variable neuropathology and differing clinical presentations.

these genes account for approximately 50% of EOAD cases.12 As shown in Figure 3, presenilin-1 is predicted to code for a novel transmembrane protein with up to nine potential hydrophobic domains.5 Over 70 point mutations, and splice-site mutations which results in the deletion of a small portion of the protein, have been identified in the presenilin genes.13 These mutations span all putative domains of the presenilin protein, including every potential transmembrane domain as shown in Figure 3. Biochemical analyses indicate all mutations result in the elevated secretion of the amyloidogenic Aβ1–42 isoform.14 Transgenic mice and clonal cell lines which express mutant forms of the presenilin genes have elevated production of Aβ1–42 compared with wildtype constructs.15 Moreover, the elevated secretion of Aβ1-42 peptide was also found in plasma and fibroblasts cultured from AD patients with known PS-1 mutations.16 The mechanism of presenilin-induced Aβ1-42 elevation is still being debated.17 Several lines of evidence suggest that PS-1 is a diaspartyl

178

THE AGEING BRAIN

protease which serves as the γ-secretase in the processing of APP. Mutation of two conserved aspartate residues at position 257 and 385 of PS-1, which face the predicted catalytic site, abrogated γ-secretase actvity.18 Moreover, both photoactivated and transition state analogues of γ-secretase inhibitors appear to bind directly to PS-1.19,20 AD Variants — Cotton Wool Plaques and Spastic Paraparesis The hallmark feature of AD is the presence of a large number of neuritic plaques and neurofibrillary tangles (NFTs). Mutations in the PS-1 gene are normally associated with severe neuropathology. Typically, there are large numbers of diffuse as well as cored, neuritic plaques, which are deposited in the cerebral cortex.21 Moreover, there is intense tau pathology, with neuritic dystrophy and NFTs.22,23 However, variations exist in AD neuropathology which include the morphology of senile plaques and the level of tau pathology. Three mutations in PS-1, a deletion of exon 9 (PS-1Δexon9),24,25 a double amino acid deletion (ΔI83/ΔM84),26 and a misssense mutation (P436Q)27 have been associated with a variant, “cotton wool” plaque pathology. As shown in Figure 4, a brain from an affected individual carrying the (PS1Δexon9) mutations had extensive deposition of large, spherical plaques that lacked distinctive cores and neuritic dystrophy. Biochemical analysis of cells transfected with PS-1 cDNAs carrying either of the three mutations secrete exceptionally high levels of Aβ1-4227 and this was suggested to be the molecular basis for PS-1 induced variant plaques. The main clinical symptoms of AD are initial memory deficits and progressive loss of higher cognitive function. However, variant forms of AD have been reported which manifest other neurological disorders. Spastic paraparesis (SP), or progressive spasticity of the lower limbs, frequently occurs on a hereditary background and a number of reports have described SP in families with dementia.13,24,27,28 Mutations reported in familial AD with SP have been confined to PS-1 with the majority of mutations consisting of deletions of exon 9.25,27,29 In the Finn2 pedigree, which has a PS-1Δexon9 mutation,24 10 of 14 individuals with dementia also had SP. Examination of the brains of three subjects, two with and one without SP, revealed many cotton wool plaques in all three cases, together with NFTs and pronounced congophilic angiopathy.24 This led to the suggestion there was an association between the variant plaques and SP clinical presentation. Further investigations into the pattern of inheritance of the dementia and SP phenotypes within another branch of the Finnish pedigree,30 and in an Australian pedigree25 (Figure 4a), suggests that the SP phenotype is due to the inheritance of an unlinked genetic locus acting in concert with the PS-1 mutation. A characteristic of the EOAD pedigrees with PS-1 mutations is that the mutation is usually fully penetrant and that there is a narrow range of age of onset of the disease within family members.31 However, there are excep-

ALZHEIMER’S DISEASE AND FRONTOTEMPORAL DEMENTIA

179

Figure 4. Spastic paraparesis and AD in an Australian EOAD pedigree. (A) Variable presentation of clinical phenotypes. Left-half black symbols indicate dementia without spastic paraparesis, right-half black symbols indicate spastic paraparesis without apparant dementia and filled symbols indicates spastic paraparesis with dementia. Haplotype analysis using four microsatellite markers flanking the PS-1 gene reveals a common disease haplotype (open box) detected in the affected individuals. (B) Variable neuropathology in EOAD pedigree as detected by antibodies against Aβ (upper panel) or tau (lower panel). Cotton wool plaques are distinguished by their lack of a distinctive amyloid core and neuritic dystrophy which were mostly detected in the individual with SP (III:9).

tions. A pedigree with the H163T mutation has been reported which has a non-penetrant individual as well as a wide range of age of onset.32 This supports the existence of genes and/or environmental factors that may modulate the expression of the AD phenotype. Analysis of large numbers of sib pairs affected with AD33 and ascertainment of the risk of relatives of aged, nondemented probands to develop AD,34 have indicated that there are genetic factors which are protective against AD or modify the age of onset of the disease. One such factor is the apolipoprotein E (ApoE) gene on chromosome 19. Inheritance of the ApoE ε4 allele has been shown to increase the risk of late-onset AD (LOAD) and results in an earlier age of onset in EOAD pedigrees with heritable mutations in the APP gene.35 Conversely, inheritance of the ApoE ε2 allele delays the onset of EOAD and LOAD cases.36 Another pos-

180

THE AGEING BRAIN

sible factor is the butyrylcholinesterase (BChE) gene on chromosome 3, which may be protective against LOAD.37 The clinical phenotype in our subjects with SP is unusual in that dementia onset appears to be delayed compared to affected individuals who presented with dementia only, since three of the four individuals who developed SP remained dementia-free for up to ten years.25 Thus, the study of variant forms of AD has been important in suggesting the existence of phenotypic modifier gene(s) act in concert with specific PS-1 mutations to alter the clinical presentation of the disease. Frontotemporal Dementia and the tau Gene Frontotemporal dementias (FTD) represent a significant group of degenerative disorders that overlap in their clinical and neuropathological descriptions. These include seemingly separate disorders such as pallido-ponto-nigral

Figure 5. Mutations in tau detected in FTDP-17. The largest tau isoform is shown within residues from alternatively spliced exons 2,3 and 10. Gray boxes represent each of the four microtubule binding domains. The stem loop structure of the 5’ splice donor site of exon 10 is drawn above the tau isoform (not to scale). Exonic sequence is given in upper case and intronic sequence is in lower case letters.

ALZHEIMER’S DISEASE AND FRONTOTEMPORAL DEMENTIA

181

Figure 6. Exon trapping analysis of tau mutations. (A) Schematic diagram showing the exon trap procedure. Genomic DNA containing exon 10 of the tau gene is subcloned into the pSPL3 exon trap vector between two vector splice sites. The recombinant construct is transfected into cells whereby the vector promoter drives expression of chimaric RNA. In vivo splicing will generate either mRNA which includes or excludes exon 10 sequence. The exon trap products are then detected by RT-PCR. (B) Gel electrophoresis of RT-PCR products from exon trapping assays in COS-7 cells. The splicing in of exon 10 yeilds a 270 bp product, while the absence results in a 177bp product. The normal tau exon 10 (wt) shows the presence of both exon 10-positive (E10+) and exon 10-negative (E10-) RT-PCR products. The +16 and S305S mutations result in a marked increase in E10+ products. The +19 and 29 mutations result in a marked decrease in E10+ products.

182

THE AGEING BRAIN

degeneration, familial multiple system tauopathy, familial progressive subcortical gliosis, progressive supranuclear palsy and disinhibition-dementiaparkinsonism-amyotrophy complex. However, as a group, these disorders share many clinical and neuropathological similarities and the tau gene has been genetically linked to a number of such pedigrees which have now been defined as Frontotemporal Dementia with Parkinsonism (FTDP-17).38 Tau is a microtubule-associated protein that is involved in the neuronal cytoskeleton, in particular, the assembly and stability of microtubules. Alternative splicing of exon 10 in the carboxy-terminal half of the protein result in tau protein which contains either three or four of the microtubule-binding repeat motifs (3 repeat tau and 4 repeat tau respectively). In 1998, tau gene mutations were shown to be causal for FTDP-17.39–41 Tau mutations can be functionally divided into two groups as shown in Figure 5. Firstly, missense mutations such as P301L and V337M have been shown to result in mutant forms of tau with decreased affinity for binding to mictotubules. The second group of exonic or intronic mutations alter the efficiency of splicing of exon 10.42 Intronic mutations (+3, +11, +12, +13, +14 and +16) identified in the 5’ splice donor site of intron 10 cause an increase in splicing of exon 10 by disrupting a stem-loop structure (Figure 5). The effect of a mutation on splicing can be assayed using the in vitro exon trapping assay which analyses the ability of the splice sites flanking an exon of interest to be spliced onto an exon trap vector’s reporter splice donor and reporter site43 (Figure 6A). As shown in Figure 6B, the presence of the +16 mutation results in an increase in exon trap products which contains the spliced exon 10 sequence compared with wildtype sequence. A Common Biochemical Pathway? One of the most crucial questions in AD research is the enigmatic relationship between senile plaques and NFTs. Studies have shown a direct causal relationship between the two diagnostic features of AD, in which Aβ appears to induce the hyperphosphorylation of tau. For example, when transgenic mice, which overexpress mutant tau (Pro301Leu) are injected with Aβ1-42 fibrils, NFTs are induced.44 Moreover, there are mutations in the tau gene that causes tau polymerisation and NFTs, but are associated with FTD rather than AD. Conversely, all mutations in the AD genes give rise to the development of tau pathology. These studies support the amyloid cascade hypothesis which states that the production and deposition of Aβ is the primary event in AD aetiology, which then cause the secondary formation of NFTs. There has been a spirited debate within the AD research field as to the relevance of Aβ1-42 and the amyloid cascade hypothesis to AD aetiology. Yet, it is still possible that there is a more fundamental neurodegenerative mechanism at work. There have been tantalising studies which suggest that there is a common cytotoxic mechanism involved in both AD and FTD. A recent study

ALZHEIMER’S DISEASE AND FRONTOTEMPORAL DEMENTIA

183

has shown that the apoptotic pathway involving the caspases is activated in AD brains.45 In vitro studies using cells transfected with mutant presenilin genes or exposed to Aβ peptides demonstrated that the cells undergo apoptosis.46,47 Similar experiments have shown that cells transfected with mutant tau cDNAs undergo the same process.48 It is hoped that the insights into the fundamental pathological mechanisms of the AD and FTD genes can be applied to the design of an effective prophylactic or therapeutic strategy for all forms of dementia. Knowing the specific molecules involved in this process, which has been achieved through genetic studies, is a crucial first step in the process. Acknowledgements Supported by a Department of Veterans Affairs Research Grant 9937441 and the National Health and Medical Research Council Block Grant and Research Fellowship 993050 and Unit Grant 983302. References 1. Iqbal K, ed. Research advances in Alzheimer’s Disease and related disorders. Chicester: John Wiley and Sons, 1995. 2. Clark RF, Goate AM. Molecular genetics of Alzheimer’s disease. Arch Neurol. 1993; 50:1164–1172. 3. Masters CL, Simms G, Weinman NA, Multhaup G, McDonald BL, Beyreuther K. Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc Natl Acad Sci USA. 1985; 82:4245–4249. 4. Selkoe DJ, Abraham CR, Podlisny MB, Duffy LK. Isolation of low-molecularweight proteins from amyloid plaque fibers in Alzheimer’s disease. J Neurochem. 1986; 146:1820–1834. 5. Price DL, Sisodia SS. Mutant genes in familial Alzheimer’s disease and transgenic models. Annu Rev Neurosci. 1998; 21:479–505. 6. Drouet B, Pincon-Raymond M, Chambaz J, Pillot T. Molecular basis of Alzheimer’s disease. Cell Mol Life Sci. 2000; 57:705–715. 7. Hardy JA, Higgins GA. Alzheimer’s disease: the amyloid cascade hypothesis. Science. 1992; 256:184–185. 8. Hardy J. Amyloid, the presenilins and Alzheimer’s disease. Trends Neurosci. 1997; 20:154–159. 9. Lichtenthaler SF, Wang R, Grimm H, Uljon SN, Masters CL, Beyreuther K. Mechanism of the cleavage specificity of Alzheimer’s disease gamma-secretase identified by phenylalanine-scanning mutagenesis of the transmembrane domain of the amyloid precursor protein. Proc Natl Acad Sci USA. 1999; 96:3053–3058. 10. Sherrington R, Rogaev EI, Liang Y, et al. Cloning of a gene bearing missense mutations in early-onset familial Alzheimer’s disease. Nature. 1995; 375:754– 760. 11. Levy-Lahad E, Wijsman EM, Nemens E, et al. A familial Alzheimer’s disease locus on chromosome 1. Science. 1995; 269:970–973. 12. Kwok JBJ, Taddei K, Hallupp M, et al. Martins RN. Two novel (M233T and R278T) presenilin-1 mutations in early-onset Alzheimer’s disease pedigrees and preliminary evidence for association of presenilin-1 mutations with a novel phenotype. NeuroReport. 1997; 8:1537–1542.

184

THE AGEING BRAIN

13. Hutton M. Presenilin mutation database. Available from: URL: http:// www.alzforum.org. 14. Murayama O, Tomita T, Nihonmatsu N, et al. Enhancement of amyloid beta 42 secretion by 28 different presenilin 1 mutations of familial Alzheimer’s disease. Neurosci Lett. 1999; 265: 61–63. 15. Citron M, Westaway D, Xia W, et al. Mutant presenilins of Alzheimer’s disease increase production of 42-residue amyloid beta-protein in both transfected cells and transgenic mice. Nat Med. 1997; 3:67–72. 16. Scheuner D, Eckman C, Jensen M, et al. Secreted amyloid beta-protein similar to that in the senile plaques of Alzheimer’s disease is increased in vivo by the presenilin 1 and 2 and APP mutations linked to familial Alzheimer’s disease. Nat Med. 1996; 2:864–870. 17. Sisodia SS, Annaert W, Kim S-H, De Strooper B. Gamma-secretase:never more enigmatic. Trends Neurosci. 2001; 24 (Suppl): 2–6. 18. Wolfe MS, Xia W, Ostaszewski BL, Diehl TS, Kimberly WT, Selkoe DJ. Two transmembrane aspartates in presenilin-1 required for presenilin endoproteolysis and gamma-secretase activity. Nature. 1999; 398:513–517. 19. Li YM, Xu M, Lai MT, et al. Photoactivated gamma-secretase inhibitors directed to the active site covalently label presenilin 1. Nature. 2000; 405:689–694. 20. Esler WP, Kimberly WT, Ostaszewski BL et al. Transition-state analogue inhibitors of gamma-secretase bind directly to presenilin-1. Nat Cell Biol. 2000; 2: 428–434. 21. Lemere CA, Lopera F, Kosik KS, et al. The E280A presenilin 1 Alzheimer mutation produces increased A beta 42 deposition and severe cerebellar pathology. Nat Med. 1996; 2:1146–1150. 22. Smith MJ, Gardner RJ, Knight MA, et al. Early-onset Alzheimer’s disease caused by a novel mutation at codon 219 of the presenilin-1 gene. NeuroReport. 1999; 10: 503–507. 23. Singleton AB, Hall R, Ballard CG, et al. Pathology of early-onset Alzheimer’s disease cases bearing the Thr113-114ins presenilin-1 mutation. Brain. 2000; 123: 2467–2474. 24. Crook R, Verkkoniemi A, Perez-Tur J, et al. A variant of Alzheimer’s disease with spastic paraparesis and unusual plaques due to deletion of exon 9 of presenilin 1. Nat Med. 1998; 4:452–455. 25. Smith MJ, Kwok JBJ, McLean CA, et al. Variable phenotype of Alzheimer’s disease with spastic paraparesis. Ann Neurol. 2001; 49:125–129. 26. Steiner H, Revesz T, Neumann M, et al. A pathogenic presenilin-1 deletion causes abberrant Abeta 42 production in the absence of congophilic amyloid plaques. J Biol Chem. 2001; 276:7233–7239. 27. Houlden H, Baker M, McGowan E, et al. Variant Alzheimer’s disease with spastic paraparesis and cotton wool plaques is caused by PS-1 mutations that lead to exceptionally high amyloid-beta concentrations. Ann Neurol. 2000; 48: 806–808. 28. Sato S, Kamino K, Miki T, et al. Splicing mutation of presenilin-1 gene for earlyonset familial Alzheimer’s disease. Hum Mutat. 1998; (Suppl 1): 91–94. 29. Prihar G, Verkkoniem A, Perez-Tur J, et al. Alzheimer disease PS-1 exon 9 deletion defined. Nat Med. 1999; 5:1090. 30. Hiltunen M, Helisalmi S, Mannermaa A, et al. Identification of a novel 4.6-kb genomic deletion in presenilin-1 gene which results in exclusion of exon 9 in a Finnish early onset Alzheimer’s disease family: an Alu core sequence-stimulated recombination. Eur J Hum Genet. 2000; 8:259–266. 31. Rohan de Silva HA, Patel AJ. Presenilins and early-onset familial Alzheimer’s disease. Neuroreport. 1997; 8:i–xii.

ALZHEIMER’S DISEASE AND FRONTOTEMPORAL DEMENTIA

185

32. Axelman K, Basun H, Lannfelt L. Wide range of disease onset in a family with Alzheimer disease and a His163Tyr mutation in the presenilin-1 gene. Arch Neurol. 1998; 55:698–702. 33. Tunstall N, Owen MJ, Williams J, et al. Familial influence on variation in age of onset and behavioural phenotype in Alzheimer’s disease. Brit J Psychiat. 2000; 176:156–159. 34. Silverman JM, Smith CJ, Marin DB, et al. Identifying families with likely genetic protective factors against Alzheimer disease. Am J Hum Genet. 1999; 64:832– 838. 35. Chartier-Harlin MC, Parfitt M, Legrain S, et al. Apolipoprotein E, epsilon 4 allele as a major risk factor for sporadic early and late-onset forms of Alzheimer’s disease: analysis of the 19q13.2 chromosomal region. Hum Mol Genet. 1994; 3: 569–574. 36. Corder EH, Saunders AM, Risch NJ, et al. Protective effect of apolipoprotein E type 2 allele for late onset Alzheimer disease. Nat Genet. 1994; 7:180–184. 37. Laws SM, Taddei K, Fisher C, et al. Evidence that the butylcholinesterase K variant can protect against late-onset Alzheimer’s disease. Alzheimers Rep. 1999; 2: 219–223. 38. Foster NL, Wilhelmsen K, Sima AA, Jones MZ, D’Amato CJ, Gilman S. Frontotemporal dementia and parkinsonism linked to chromosome 17: a consensus conference. Conference Participants. Ann Neurol. 1997; 41:706–715. 39. Spillantini MG, Goedert M. Tau mutations in familial frontotemporal dementia. Brain. 2000; 123:857–859. 40. Hutton M, Lendon CL, Rizzu P, et al. Association of missense and 5’-splicesite mutations in tau with the inherited dementia FTDP-17. Nature. 1998; 393: 702–705. 41. Poorkaj P, Bird TD, Wijsman E, et al. Tau is a candidate gene for chromosome 17 frontotemporal dementia. Ann Neurol. 1998; 43:815–825. 42. Spillantini MG, Murrell JR, Goedert M, Farlow MR, Klug A, Ghetti B. Mutation in the tau gene in familial multiple system tauopathy with presenile dementia. Proc Natl Acad Sci USA. 1998; 95:7737–7741. 43. Stanford PM, Halliday GM, Brooks WS, et al. Progressive supranuclear palsy pathology caused by a novel silent mutation in exon 10 of the tau gene: expansion of the disease phenotype caused by tau gene mutations. Brain. 2000; 123: 880–893. 44. Gotz J, Chen F, van Dorpe J, Nitsch RM. Formation of neurofibrillary tangles in P301l tau transgenic mice induced by Abeta 42 fibrils. Science. 2001; 293: 1491–1495. 45. Marx J. Neuroscience. New leads on the ‘how’ of Alzheimer’s. Science. 2001; 293: 2192–2194. 46. Wolozin B, Iwasaki K, Vito P, et al. Participation of presenilin 2 in apoptosis: enhanced basal activity conferred by an Alzheimer mutation. Science. 1996; 274: 1710–1713. 47. Guo Q, Sopher BL, Furukawa K, et al. Alzheimer’s presenilin mutation sensitizes neural cells to apoptosis induced by trophic factor withdrawal and amyloid beta-peptide: involvement of calcium and oxyradicals. J Neurosci. 1997; 17: 4212–4222. 48. Furukawa K, D’Souza I, Crudder CH, et al. Pro-apoptotic effects of tau mutations in chromosome 17 frontotemporal dementia and parkinsonism. NeuroReport. 2000; 11:57–60.

Chapter 11 OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING Judy B. de Haan*, Rocco C. Iannello, Peter J. Crack, Paul Hertzog, and Ismail Kola

Introduction In this chapter, we discuss the relationship between increased oxidative stress and cellular ageing, with particular emphasis on the physiological and pathological changes associated with ageing of the brain. In this regard, focus will be on the role of the antioxidant enzymes during cellular ageing, and the consequences of an altered antioxidant balance. We will highlight the role that antioxidant genes play in the regulation of senescent-like changes both in in vitro and in vivo models, and how perturbations of the antioxidant pathways may lead to clinical outcomes associated with ageing, e.g. neurodegenerative diseases such as Parkinson’s and Alzheimer’s disease. Furthermore, this chapter will illustrate how an altered antioxidant ratio as a direct consequence of the over-expression of the antioxidant gene, Sod1, leads to ageing changes associated with the Down syndrome (DS) phenotype. In this manner it is hoped to emphasize the importance of the antioxidant genes in the regulation of redox status during cellular ageing, and how perturbations of redox balance may have pathological consequences associated with ageing. Molecular oxygen: a paradox for aerobic organisms The existence of reactive oxygen species (ROS) within cells is an unavoidable consequence of both oxidative metabolism and exposure to environmental *To whom correspondence should be addressed.

188

THE AGEING BRAIN

stresses such as radiation, air pollutants and herbicides. During oxidative metabolism, oxygen is reduced to water via reactive intermediates that include the superoxide radical (.O2-), hydrogen peroxide (H2O2), and hydroxyl radical (.OH). Irrespective of the mode of radical generation, ROS cause cellular damage through interactions with macromolecules, resulting in mutations in DNA (both mitochondrial1 and nuclear DNA), destruction of protein structure and function, and peroxidation of membrane lipids.2 Together with non-enzymatic antioxidants (e.g. ascorbate, glutathione, αtocopherol), aerobic organisms have evolved highly efficient enzymatic antioxidant defences to overcome the problems of oxidative stress. These include the superoxide dismutases (Sod), glutathione peroxidase (Gpx) and catalase enzymes. Superoxide dismutases function in the first step of the antioxidant pathway (Figure 1) where .O2- is converted to H2O2, while Gpx and Cat are independently involved in the neutralisation of H2O2 to water in a second step. This is a finely tuned process and a balance exists within cells between the first (Sod) and second steps (Gpx and/or Cat) of the antioxidant pathway. Indeed it has been postulated that perturbations of this balance could affect cell function, since a shift in favour of H2O2 (the intermediate product) could result in Fenton-type reactions with transition metals, resulting in even more noxious .OH radicals.3 Once formed, these species quickly interact with molecules in their immediate vicinity, particularly lipids, causing large-scale macromolecular damage.4

Figure 1. Two-step Antioxidant Pathway. Superoxide radicals generated during oxidative metabolism, are neutralised to water via a two-step process involving superoxide dismutase (Sod) in a first step, and both or either glutathione peroxidase (Gpx) and catalase (Cat) in a second step. Fenton-type reactions occur when an imbalance in this pathway favors the build-up of hydrogen peroxide (H2O2), resulting in peroxidation of molecules such as lipids. Adapted from Groner et al.4

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

189

Compared with other organs, the brain is most vulnerable to ROS-induced damage. This is in part due to: (1) the large generation of ROS during oxidative phosphorylation as a consequence of the high rate of oxygen consumption by the brain; (2) high levels of iron in some brain regions which catalyse the generation of ROS via Fenton-type reactions; and (3) the brain is lipid rich, particularly in polyunsaturated fatty acids that are known targets of lipid peroxidation. Thus the levels of cellular antioxidant enzymes become important during the ageing process, since antioxidants are able to limit ROS reactions that would otherwise damage macromolecules in an unabated fashion. Indeed, it is conceivable that continued damage to macromolecules could translate at a higher level into cellular and/or organ dysfunction and ultimately cellular ageing and/or death. Antioxidants and Lipid Peroxidation during Brain Ageing To gain a better understanding of the role played by antioxidants in limiting damage during brain ageing, a number of studies have focussed on the activity of these antioxidants during cellular ageing. However, the literature is controversial with respect to the activities of the major antioxidants, often due to limitations in the methodologies. Often no distinction is made between the different isoforms of the superoxide dismutases, such that Sod1 (the cytosolic and most abundant isoform) and Sod2 (the mitochondrial isoform) are assayed as one. This has resulted in reports of either an increase,5 a decrease6 or no change7 in the Sod activity in ageing murine or rat brains. Even when Sod1 and Sod2 are assayed independently, Sod1 has been reported to increase,8 remain constant,9 or decrease10,11 with ageing in mouse or rat brains. Likewise, Gpx1 (the most abundant cytosolic and mitochondrial isoform) activity has been reported to increase5,11 or remain relatively unchanged6,7 in ageing murine or rat brains. As a marker of oxidative damage in the ageing brain, the levels of lipid peroxidation are often assayed. Here again, the results have been controversial, with reports of either an increase11–13 or a decrease14 occurring. Indeed, very few studies simultaneously investigate the activities of Sod1, Gpx1, catalase and the lipid peroxidation status, which becomes important if discrepancies such as differences in sex, species, strain and age are to be eliminated. We have examined the levels of the three major antioxidant enzymes (Sod1, Gpx1 and Cat) and the levels of lipid peroxidation in a range of ageing murine brains. We show that Sod1 mRNA levels and activity are significantly increased in murine brains during the ageing process. Similarly, Sod1 mRNA levels and activity are increased in other murine organs with age. These include the ageing liver, lung, kidney, heart, ovary, and bone.15–17 An analysis of the Gpx1 profile showed that most organs adapt to the increased Sod1 activity by upregulating Gpx1 and/or Cat, at both the mRNA and activity level with advanced age. However and most importantly, the brain failed

190

THE AGEING BRAIN

to show an increase in either Gpx1 or Cat mRNA or activity with increasing age. This implies that the ageing brain has an altered Sod1 to Gpx1 and Cat ratio. As already suggested, an altered ratio can lead to increased H2O2 and .OH, which in turn can damage macromolecules such as lipids. Indeed the ageing brain showed significantly increased levels of lipid peroxidation, while those organs that adapted to the increased Sod1 levels by upregulating Gpx1 and/or catalase showed reduced peroxidative damage.15,17 Our data therefore suggest that an altered antioxidant balance may result in peroxidative damage to biologically important molecules in the ageing brain. Regulation of the antioxidant genes during ageing becomes important, since we are able to show that upregulation of the Sod1 gene in murine brains is transcriptionally regulated. Why the Gpx1 gene is not upregulated during murine brain ageing, but is upregulated in other ageing murine organs, is not yet understood. It may be the need for increased levels of the intermediate H2O2, to perform other regulatory functions in the brain during the lifetime of the individual, that out-ways the damaging peroxidative changes seen in ageing brains. In a recent study, providing the first profile of gene regulation at the molecular level in ageing murine brains, Lee et al. 18 analysed the expression patterns of 6,347 genes using oligonucleotide array analysis. They were able to show that ageing of the murine neocortex and cerebellum resulted in a gene-expression profile indicative of increased oxidative stress, an altered inflammatory response, and reduced neurotrophic support. These data provide good evidence at a global level that oxidative stress and inflammatory processes, the latter known to involve ROS, are increased in the ageing brain and that these processes are regulated at the gene level during brain ageing. Lessons from Gene Targeting Experiments The above studies, which imply a role for an altered antioxidant balance, increased oxidative stress and peroxidative damage during cellular ageing, although informative, are still limited since they are of a correlative nature. To overcome these limitations, and to directly prove a role for ROS in cellular ageing, researchers have used molecular techniques to address these issues. In vitro models of cellular senescence Various studies have shown that over-expression of the human Sod1 gene in human Hela cells, mouse-L cells and rat PC12 cells, leads to increased lipid peroxidation as well as structural and functional alterations of lipid membranes.4,19 We have extended these studies by investigating the effects of Sod1 overexpression in murine NIH 3T3 cells with respect to cellular senescence.20 Two types of cell clones were isolated; those overexpressing both Sod1 and Gpx1 which were termed “adapted cell clones”, and those that failed

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

191

to upregulate Gpx1 in response to increased Sod1 levels, which were termed “non-adapted cell clones”. The adapted cells were morphologically and biochemically indistinguishable from the parental lines from which they were derived. An important observation was that the non-adapted cells looked and behaved like senescent cells in culture. They exhibited a greater cell-surface area and had a larger nuclear and cellular volume. These cells also grew more slowly, as assessed by both growth rate and 3H-dThymidine incorporation. Furthermore, expression levels of Cip1, which is a well-characterized marker of cellular senescence, were elevated in all non-adapted cells investigated.20 Since H2O2 is the intermediate between the two steps of the antioxidant pathway, we measured both intra-and extracellular H2O2 levels of parental, adapted and non-adapted cells. Hydrogen peroxide levels were not significantly different in adapted and parental cells. However, both intra- and extracellular H2O2 levels were significantly increased in non-adapted cells compared with both adapted and parental cells. Thus by altering the Sod1 to Gpx1 ratio in favour of H2O2 formation, we were able to demonstrate senescence-like changes, and we postulated that it was the overproduction of H2O2, either directly or via Fenton-like reactions to produce .OH radicals, that was mediating the senescence-like effects. Indeed, we were able to mimic these changes in both NIH 3T3 cells and primary cultures after the addition of H2O2.20,21 We addressed the issue of an altered redox balance and senescence-like changes in two further culture systems where ROS production is elevated as a consequence of a change in antioxidant gene expression. First we analysed cultured fibroblasts derived from mice that were genetically manipulated to have a null mutation for the Gpx1 gene. Indeed, fibroblasts lacking any functional Gpx1 (Gpx1-/- cells) exhibited features of senescence when compared with control cells, and were more susceptible to H2O2 -induced apoptosis than controls. Furthermore, Gpx1-/- neurons demonstrated decreased viability after exposure to H2O2 compared with controls.22 Second, we showed that cell lines derived from Down syndrome aborted conceptuses (where the Sod1 to Gpx1 ratio is increased as a consequence of three copies of the Sod1 gene) exhibit senescence-like characteristics, namely they grew more slowly, incorporated less 3H-dThymidine and expressed higher levels of the senescence marker, Cip1. Indeed, premature ageing is one of the phenotypes associated with individuals with DS (see Down syndrome section below). From these studies, it is evident that an altered antioxidant balance, either as a consequence of the overexpression of Sod1 or a reduction in the activity of Gpx1, leads to senescence-like changes in cultured cells. In vivo models of cellular senescence The most striking data that an altered antioxidant ratio is involved in the genesis of ageing comes from Yarom et al.,23 who show that mice transgenic for Sod1, develop morphological and biochemical changes at neuromus-

192

THE AGEING BRAIN

cular junctions of tongue muscles (namely, withdrawal and destruction of some terminal axons and the development of multiple small terminals), which are similar to those seen in tongue muscles of ageing mice and rats.24 Furthermore, these changes are similar to those seen in tongue muscles of individuals with DS.25,26 A subsequent study has also demonstrated premature ageing changes in the incidence, length and number of nerve branch-points in Sod1 transgenic hind-limb motor-neuron terminals. Again these changes are analogous to those seen in ageing mice and rat muscles of the hindlimb.27 The results may also explain how over-expression of Sod1 affects motor neurons in individuals with DS, resulting in the impairment of central motoric coordination and generalized hypotonia of joints. These authors also report ageing changes in thymocyte populations isolated from Sod1-transgenic animals,28 and the accumulation of the age-related pigment, lipofuscin, in the myofibers of Sod1transgenic animals.27 Importantly, Ceballos et al.29 demonstrate increased lipid peroxidation in the brains of Sod1-transgenic mice. Taken together, the above data strongly suggest that altered redox status, as a consequence of Sod1 over-expression, leads to accelerated ageing changes in these models. The role of Free Radical Mechanisms in Diseases of the Ageing Brain Parkinson’s disease Parkinson’s disease (PD) is characterized by the progressive loss of pigmented dopamine-containing neurons in the substantia nigra pars compacta of the brain. The mechanisms involved in the specific targeting of these cells have received much attention over the past number of years, since an understanding of these process(es) may facilitate drug design to either reduce or limit such damage. One theory is that individuals with PD have a defective antioxidant system that is incapable of removing harmful ROS generated during the oxidation of dopamine. Monoamineoxidase catalyses the oxidation of dopamine via the reactive intermediate, 6-hydroxy-dopamine.30 It is during this process that .O2and H2O2 are generated. The Sod1 activity has been reported to increase in the dopamine-containing neurons of Parkinsonian brains, possibly as a consequence of the increased .O - flux.31 Interestingly, it is within these same cell-types that Ceballos et al.32 2 demonstrate increased Sod1 activity in aged brains (with a mean age of 83 yrs). Cellular damage would be limited if second-step antioxidants were increased concomitantly. However, Gpx1 activity has been reported to remain unchanged or is reduced in the substantia nigra of individuals with PD.33 Furthermore, no difference in either catalase or glutathione reductase is seen in the substantia nigra or basal nucleus of Parkinsonian brains compared with normal brains.34 Interestingly, Sian et al.35 have shown that levels of reduced glutathione are decreased by approximately 40–50% in the substantia nigra, thus limiting the efficient removal of hydrogen peroxide by Gpx in these Parkinsonian

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

193

brains (reduced glutathione is required as a cofactor by Gpx). Furthermore, the potential for the formation of harmful .OH radicals also exists, since this region exhibits an increase in iron content.36 Thus from in vivo studies it would appear that there is a significant shift in the balance of antioxidant enzymes in favour of increased H2O2 and .OH production in these brains, which may account for the increased cell damage seen in this region of Parkinsonian brains. Indeed lipid peroxidation is elevated in nigral tissue of post mortem brains from Parkinsonian individuals compared with control brains.37,38 Further strong evidence for an altered antioxidant balance contributing to PD pathology comes from recent data of Klivenyi et al.39 who demonstrate that Gpx1 knockout mice are more susceptible to 1-methyl-4-phenyl1,2,5,6-tetrahydropyridine (MPTP)–induced PD. Administration of MPTP, a mitochondrial toxin and known inducer of oxidative stress, resulted in significantly greater depletions of dopamine in Gpx1 deficient mice compared with control mice. These data strongly suggest a neuroprotective role for Gpx1 in the prevention or reduction of PD-like symptoms after challenge with mitochondrial toxins. In support of this notion, Bensadoun et al.40 have shown that mice transgenic for Gpx1 are less susceptible to the toxic effects of 6hydroxydopamine, which is known to produce H2O2 and superoxide radicals. Furthermore, Klivenyi et al.39 suggest that Gpx1 protects against neurotoxicity by detoxifying harmful peroxynitrite radicals (the latter are formed through the reaction of nitric oxide and superoxide radicals), since inhibitors of neuronal nitric oxide synthase, enzymes that generate NO, block MPTP neurotoxicity. However, it should be emphasised that any genetic defect(s) in free radical scavenging enzymes appear to be compensated for under physiological conditions, since we22 and others39,41 show no pathology of Gpx1 knockout mice under physiological conditions. These defects translate into PD pathology only when Gpx1 knockout mice are exposed to certain environmental factors or toxins. These toxins may even be produced endogenously. Naoi and Maruyama42 demonstrate degeneration of dopaminergic neurons by NM(R)Sal, an endogenous MPTP-like neurotoxin. This mechanism may also hold true for individuals susceptible to PD, i.e. an altered antioxidant balance, possibly due to a reduction in second-step antioxidants such as Gpx1, and exposure to an environmental/endogenous toxin. A possible therapeutic target for PD is therefore the upregulation of the glutathione/glutathione peroxidase system, and in this regard, glutathione precusors such as N-acetyl-cysteine (NAC) have been considered.43 Stroke Stroke is the leading cause of long-term disability in adults and ranks as the third leading cause of death after heart disease and cancer. Approximately 80% of all strokes are ischaemic, that is, due to a reduction of blood flow to certain brain regions caused by blockage of a vessel. This results in oxygen deprivation to those regions normally supplied by the occluded blood vessel. Blood flow back into the occluded region (reperfusion) is accompanied by

194

THE AGEING BRAIN

the production of ROS at an enhanced rate. The increased ROS production is thought to trigger certain molecular pathways leading to necrosis, apoptosis and neuroinflammation, resulting in subsequent neuronal loss and serious cognitive and/or motor disturbances.44 Thus the role played by antioxidants in the removal of these harmful species becomes extremely important in limiting the neuronal damage post-ischemia. Use has been made of transgenic and knockout mice to address the issue of the role of antioxidant genes in the pathogenesis of stroke. For example, following mid cerebral artery occlusion and subsequent reperfusion, the infarct volumes of Sod1 knockout mice are significantly increased, 45,46 implying a protective role for Sod1 against ROS such as .O2-. Conversely, the infarct volume of Sod1 transgenic mice are reduced compared with control mice.47 This is somewhat surprising since overexpression of Sod1 should lead to increased levels of H2O2 and in the absence of an adaptive rise by second-step antioxidants, increased H2O2 levels are known to be cytotoxic. Thus the mechanism by which Sod1 over-expression confers protection in these mice is unclear. It may well be that the Gpx1 gene is upregulated in stroke-related situations in an adaptive response to the increased levels of H2O2. In further trying to tease out the role of the various antioxidants in neuroprotection, Weisbrot-Lefkowitz et al. 48 were able to show that mice transgenic for Gpx1 show a greater level of protection against ischaemiareperfusion damage than Sod1 transgenic mice. These results imply that H2O2 and/or hydroxyl radicals (formed from H2O2 in the Fenton reaction) are more neurotoxic than superoxide radicals and therefore second-step antioxidants such as Gpx1 play a far greater neuroprotective role than the superoxide dismutases. This becomes particularly important when designing strategies for drug therapy. In agreement with this notion, recent studies in our laboratory have shown that Gpx1 knockout mice are more susceptible to ischaemia-reperfusion injury than controls. We show that the severity of the infarct volume is significantly increased in Gpx1-/- mice compared with controls. This increase also correlated with an increase in caspase-3 activation and an elevation in the level of apoptosis in neural cells. 49 Our results suggest that Gpx1 plays an important role in the protection of neural cells against the elevated oxidative stress that accompanies ischemia/reperfusion injury. Amyotrophic lateral sclerosis (AML) Amyotrophic lateral sclerosis (AML) is a progressive disorder of motor neurons found in the cortical regions of the brain, brain stem and spinal cord. Muscular wasting, weakness and fasciculations, spasticity and hyperreflexia characterise the disease. From the time of onset, patients with ALS survive a mean period of 3–4 years. Ninety percent of ALS cases occur as a sporadic event, while the remaining 10% are inherited as an autosomal dominant trait, with high penetrance after the sixth decade. Whether the disease occurs sporadically or is inherited, the clinical features in most instances are similar.

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

195

The Sod1 gene has been identified as one of the key players associated with familial forms of amyotrophic lateral sclerosis (FALS).50 The study of Rosen et al.50 demonstrated that single base mutations in the coding region of the Sod1 gene are associated with FALS (different mutations are detected in different families). These mutations resulted in amino-acid substitutions in regions of the Sod1 enzyme that are highly conserved amongst organisms, suggesting that these sites are important for Sod1 function. Transgenic mice expressing mutated forms of Sod1 have provided some of the most informative data regarding the mechanisms involved in human ALS. Notably, Gurney et al.51 were able to demonstrate ALS-like symptoms in mice that over-express a human Sod1 mutation, namely progressive paralysis and motor neuron loss in the spinal cord and brain stem, thus providing proof that altered Sod1 function can lead to neurodegenerative changes.52 Furthermore, it has been shown that mutations in Sod1 result in a dominant gain-of-function that is peroxidase-like, resulting in the increased formation of .OH radicals.53,54 Recently, Cha et al.55 have shown that neuronal nitric oxide synthetase (nNOS) expression is enhanced in mutant Sod1 transgenic mice, particularly within astrocytes, implying a role for NO in the genesis of neurotoxic injury. NO together with superoxide radicals are known to generate highly noxious peroxynitrate radicals. Furthermore, mutant Sod1 has been shown to facilitate peroxynitrite-mediated nitration of proteins in mutant Sod1 transgenic mice.56 Pathogenesis in the transgenic mouse model of FALS has recently been proposed to occur via a two-step sequential process, in which damage is mediated by free radicals which accumulate to a threshold, triggering catastrophic motor neuron loss through glutamate-mediated excitotoxic mechanisms.57 Evidence in support of this hypothesis comes from therapeutic studies with antioxidants and inhibitors of glutamatergic neurotransmission. Feeding of mutant Sod1 transgenic mice with vitamin E and selenium (selenium is an essential cofactor of Gpx1) delayed onset of the disease, and strength and mobility were transiently improved compared with non-supplemented Sod1 mutant mice. Administering riluzole and gabapentin, two drugs that reduce presynaptic glutamate release or biosynthsis, improved survival of the mutant Sod1 mice.58 It is probably correct to assume that both sporadic and familial forms of ALS trigger a common ROSmediated pathway of motor neuron death that is .OH and/or peroxynitratemediated. Alzheimer’s disease Alzheimer’s disease (AD) affects 7% of the population over 65 years of age and is characterized by slow progressive intellectual decline and personality deterioration. Autopsy studies show intra-neuronal fibrillary tangles and neuritic plaques. The latter are spherical extracellular cores of β-amyloid protein surrounded by degenerating nerve-cell processes. Both plaques and tangles occur throughout the cerebral cortex of the brain and consist of bundles of uniform proteins that appear as paired helical filaments on electron microscopic examination.59

196

THE AGEING BRAIN

There is now growing evidence that amyloid β-peptide (Aβ) and oxidative stress are implicated in the pathogenesis of AD.60,61 It has been shown that Aß is over-produced in the brains of patients with AD and that Aβ peptides can be toxic to neuronal cells through a mechanism that involves H2O2.62 Indeed, oxidative stress actually exacerbates Aβ aggregation,63 while the deposition of Aβ in turn, increases intraneuronal generation of ROS.64 In this manner, a cyclical response is established with continued deposition of Aβ. In vitro studies have also shown marked oxidative injury, including lipid peroxidation, protein carbonyl formation, mitochondrial DNA damage and the induction of stress related alterations that include ubiquination of cytoskeletal proteins.65 Furthermore, the amino acid hydroxyproline that is not normally a constituent of cytoplasmic protein in the brain, was identified as an integral part of paired helical filament proteins in AD brains. This led Zemlan et al.59 to propose that these modified amino acids arise due to non-enzymatic hydroxylation of proline residues, presumably arising from .OH radicals. If the hypothesis that oxidative stress contributes to the pathology of AD is correct, then H2O2 and/or .OH could be elevated in Alzheimer’s disease as a consequence of an alteration in antioxidant balance. Indeed, Sod1 levels are elevated in AD brains. In particular, post-mortem analysis of AD brains showed that large pyramidal neurons of the hippocampus contained higher amounts of Sod1 mRNA and protein than control brains.66 It is these cells that are particularly vulnerable to degenerative processes in AD. In addition, Sod1 activity is increased by 30% in fibroblasts of familial AD patients compared with normal controls.59 Also lending support to this notion are the data from amyloid precursor protein (APP)-transgenic mice that show elevated Sod1 levels in brain regions, especially around Aβ deposits.65 These authors also demonstrate an increase in heme-oxygenase1, a marker of oxidative stress, around most amyloid deposits in their APPtransgenic mice. Further evidence that oxidative damage may contribute to cellular damage in AD brains comes from an analysis of lipid peroxidation in these brains. Indeed, the two regions most susceptible to neurodegeneration in AD, the temporal and parietal cortex, showed elevated levels of lipid peroxidation, whilst the least affected areas, namely the occipital cortex and cerebellum, showed no elevation when compared with control brains.67 Overexpression of the Antioxidant Gene Sod1, Leads to Ageing Changes Associated with the Down Syndrome Phenotype Accelerated ageing changes have been observed in individuals with Down syndrome. In particular, the rapid or early onset of ageing is evident visually as premature greying or loss of hair. Detailed biochemical analysis has revealed that individuals with DS show a decline in immune responsiveness similar to that seen in older people. Furthermore, alterations in cyclic nucleotide metabolism have been noted in lymphocytes from individuals with DS, which

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

197

is comparable with what occurs in ageing human lymphocytes. Granulovacuolar degeneration of neurons and the appearance of Alzheimer’s disease pathology, amyloidosis, hypogonadism and degenerative vascular disease have also been noted in DS individuals.68,69 A perturbation in the ratio of Sod1 relative to Gpx1 and/or catalase is of particular relevance in DS. Importantly, the Sod1 gene is located on human Ch21, and in 95% of cases the entire chromosome is triplicated in DS individuals. Indeed, elevated Sod1 activity has been observed in tissues such as red blood cells, platelets, fibroblasts and lymphocytes of DS individuals.70 Furthermore, we show that the Sod1 to Gpx1 ratio is increased in five fetal DS organs.16 These include the fetal brain, liver, thymus, lung and heart. Our results are in agreement with Brooksbank and Balazs71 who demonstrate increased Sod1 activity, unaltered Gpx activity, and increased lipid peroxidation in DS fetal brains. However an altered antioxidant balance may not occur in all DS organs and tissues, since a compensatory increase in Gpx1 activity occurs in erythrocytes and lymphoid cells of DS individuals.72 Interestingly, catalase activity has been shown to be unaffected in DS erythrocytes.73 Thus it appears that the ratio of the major antioxidant enzymes is shifted in favour of H2O2 production in the majority of organs, and may in this manner contribute to the age-related pathologies associated with DS.74 Evidence in favour of this hypothesis comes from data of Buscioglio and Yankner75 who demonstrate increased ROS formation in primary cultures of DS fetal neurons. Glycoxidation and in particular, the accumulation of advanced glycation end products (AGEs), are also enhanced in DS fetal brains.76 These results suggest that brain oxidative stress occurs very early in the life of DS individuals, and may contribute to some of the pathologies associated with DS, such as the premature ageing and neurodegenerative AD-like pathology. Of particular interest has been the age-associated neuritic plaque formation analogous to that found in AD.77 Post-mortem examination of DS brains over the age of 35 years, almost invariably show pathology similar to that seen in AD. Interestingly, the areas most affected in DS brains parallel those affected in individuals with AD. Evidence has already been presented that ROS may contribute to the pathology of AD. Indeed, the gene dosage increase in Sod1 activity in DS may contribute in a similar fashion to the ADlike pathology. However, in this instance it likely to be the concordant effect of increased Sod1 activity and increased Aβ production that is responsible for the AD-like pathology in DS, since the gene coding for APP is also localized to human chromosome 21 and is over-expressed in DS.78 Conclusion This chapter has described the role of the antioxidant genes and their gene-products in the regulation of cellular ageing. It has focussed primarily on their role in brain ageing since this organ is particularly vulnerable to peroxidative insults.

198

THE AGEING BRAIN

Figure 2. An altered antioxidant balance may have pathological consequences. An imbalance in the major antioxidants can result in the build-up of noxious radicals, predisposing and/or contributing to various pathologies, e.g. (i) neuropathological outcomes such as stroke, Familial Lateral Sclerosis (FALS), Parkinson’s and Alzheimer’s disease, (ii) Down syndrome as a consequence of the overexpression of SOD1, a chromosome 21 gene, and (iii) the ageing process per se, which in turn may predispose to various pathologies. Furthermore inflammation often accompanies these pathologies e.g. AD, PD and Down syndrome, often as a consequence of radical-mediated induction of transcription factors such as NF-κB, leading to upregulation of TFN-α, IL-1 and IL-6. This in turn generates more radicals. An increase in both SOD1 and APP may contribute to AD-like pathology in DS.

It has shown that an altered Sod1 to Gpx1 and catalase ratio exists in ageing murine brains and that this altered ratio is accompanied by an increase in lipid peroxidation. It has highlighted the importance of maintaining a redox balance in cells and that a perturbation in first to second step antioxidant enzymes can affect cell function, leading to senescence-like changes. Furthermore, evidence was presented that altered redox states exist in various pathologies associated with ageing (Figure 2). Cumulative evidence now strongly suggests the existence of a molecular basis for ageing, with the regulation of the antioxidant genes playing an important role in this regard. Current thinking supports the hypothesis of Sohal and Allen79 who extended the free radical theory of ageing (which was based on the production of ROS in an uncontrolled fashion during aerobic metabolism), to

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

199

include controlled genetic changes induced by ROS during ageing. It is only by clearly defining the molecular basis of senescence-like changes during normal cellular ageing and in pathologies associated with ageing, that one can hope to design drugs to reduce or ameliorate these detrimental ROS-induced effects. References 1. Richter C. Oxidative damage to mitochondrial DNA and its relationship to ageing. Int J Biochem Cell Biol. 1995; 27:647–653. 2. Halliwell B, Gutteridge JMC. In: Free radicals in biology and medicine, Oxford, Clarendon Press, 1985. 3. Imlay JA, Chin SM, Linn S. Toxic DNA damage by hydrogen peroxide through the Fenton reaction in vivo and in vitro. Science. 1988; 240:640–642. 4. Groner Y, Elroy-Stein O, Avraham KB, Yarom R, Schickler M, Knobler H, Rotman G. Down syndrome clinical symptoms are manifested in transfected cells and transgenic mice overexpressing the human Cu/Zn-superoxide dismutase gene. J Physiology-Paris. 1990; 84:53–77. 5. Tayarani I, Cloez I, Clement M, Bourre JM. Antioxidant enzymes and related trace elements in ageing brain capillaries and choroid plexus. J Neurochem. 1989; 53:817–824. 6. Benzi G, Pastoris O, Villa RF. Changes induced by ageing and drug treatment on cerebral enzymatic antioxidant system. Neurochem Res. 1988; 13:467–478. 7. Cand F, Verdetti J. Superoxide dismutase, glutathione peroxidase, catalase, and lipid peroxidation in the major organs of the ageing rats. Free Radical Bio Med. 1989; 7:59–63. 8. Bracco F, Burlina AP, Malesani R, Rigo A, Battistin L. Free-radical related enzymes in the ageing brain. In: Bes A, et al. editors. Senile dementias: Early detection, Libbey, 1986; 293–297. 9. Kurobe N, Suzuki F, Kato K, Sato T. Sensitive immunoassay of rat Cu/Zn superoxide dismutase: concentrations in the brain, liver, and kidney are not affected by ageing. Biomed Res. 1990; 11:187–194. 10. Mariucci G, Ambrosini MV, Colarieti L, Bruschelli G. Differential changes in Cu, Zn and Mn superoxide dismutase activity in developing rat brain and liver. Experientia 1990; 46:753–755. 11. Sahoo A, Chainy GBN. Alterations in the activities of cerebral antioxidant enzymes of rat are related to ageing. Int J Dev Neurosci. 1997; 15:939–948. 12. Mizuno Y, Ohta K. Regional distributions of thiobarbituric acid-reactive products, activities of enzymes regulating the metabolism of oxygen free radicals, and some of the related enzymes in adult and aged rat brains. J Neurochem. 1986; 46:1344–1352. 13. Sawada M, Carlson JC. Changes in superoxide radical and lipid peroxide formation in the brain, heart and liver during the lifetime of the rat. Mech Ageing Dev. 1987; 41:125–137. 14. Boehme DH, Kosecki R, Stern F, Marks N. Lipoperoxidation in human and rat brain tissue: development and regional studies. Brain Res. 1977; 136:11–21. 15. de Haan JB, Newman JD, Kola I. Cu/Zn superoxide dismutase mRNA and enzyme activity, and susceptibility to lipid peroxidation, increases with ageing in murine brains. Mol Brain Res. 1992; 13:179–186. 16. de Haan JB, Wolvetang E, Cristiano F, Iannello R, Kelner M, Kola I. Reactive oxygen species and their contribution to pathology in Down syndrome. Adv Pharmacol. 1997; 38:379–402.

200

THE AGEING BRAIN

17. Cristiano F, de Haan JB, Iannello I, Kola I. Changes in the levels of enzymes which modulate the antioxidant balance occur during ageing and correlate with cellular damage. Mech Ageing Dev. 1995; 80:93–105. 18. Lee C-K, Weindruch R, Prolla TA. Gene-expression profile of the ageing brain in mice Nat Genet. 2000; 25:294–297. 19. Elroy-Stein O, Bernstein Y, Groner Y. Overproduction of human Cu/Zn-superoxide dismutase in transfected cells: extenuation of paraquat-mediated cytotoxicity and enhancement of lipid peroxidation. EMBO J. 1986; 5:615–622. 20. de Haan JB, Cristiano F, Iannello R, Kelner M, Kola I. Elevation in the ratio of Cu/Zn-superoxide dismutase to glutathione peroxidase leads to cellular senescence and this effect is mediated by H2O2. Hum Mol Genet. 1996; 5:283–292. 21. Bladier C, Wolvetang EJ, Hutchinson P, de Haan JB, Kola I. Response of a primary human fibroblast cell line to H 2O2: Senescence-like growth-arrest or apoptosis? Cell Growth Differ. 1997; 8:589–598. 22. de Haan JB, Bladier C, Griffiths P, Kelner M, O’Shea RD, Cheung NS, Bronson RT, Silvestro M.J, Wild S, Zheng SS, Beart PM, Hertzog PJ, Kola I. Mice with a homozygous null mutation for the most abundant glutathione peroxidase Gpx1, show increased susceptibility to the oxidative stress-inducing agents paraquat and hydrogen peroxide. J Biol Chem. 1998; 273:22528–22536. 23. Yarom R, Sapoznikov D, Havivi Y, Avraham KB, Schickler M, Groner Y. Premature ageing changes in neuromuscular junctions of transgenic mice with an extra human CuZnSOD gene: a model for tongue pathology in Down’s Syndrome. J Neurol Sci. 1988; 88:41–53. 24. Fahim MA, Robbins N. Ultrastructural studies of young and old mouse-neuromuscular junctions. J Neurocytol. 1982; 11:641–656. 25. Yarom R, Sherman Y, Sagher U, Peled IJ, Wexler MR. Elevated concentrations of elements and abnormalities of neuromuscular junctions in tongue muscles of Down’s syndrome. J Neurol Sci. 1987; 79:315–326. 26. Epstein CJ, Avraham KB, Lovett M, Smith S, Elroy-Stein O, Rotman G, Bry C, Groner Y. Transgenic mice with increased Cu/Zn-superoxide dismutase activity: Animal model of dosage effects in Down syndrome. Proc Natl Acad Sci USA. 1987; 84:8044–8048. 27. Avraham KB, Sugarman H, Rotshenker S, Groner Y. Down’s syndrome: morphological remodelling and increased complexity in the neuromuscular junction of transgenic CuZn-superoxide dismutase mice. J Neurocytol. 1991; 20:208–215. 28. Peled-Kamar M, Lotem J, Okon E, Sachs L, Groner Y. Thymic abnormalities and enhanced apoptosis of thymocytes and bone marrow cells in transgenic mice over-expressing Cu/Zn-superoxide dismutase: implications for Down syndrome. EMBO J. 1995; 14:4985–4993. 29. Ceballos I, Nicole A, Briand P, Grimber G, Delacourte A, Flament S, Blouin JL, Thevenin M, Kamoun P, Sinet PM. Expression of human Cu-Zn superoxide dismutase gene in transgenic mice: model for gene dosage effect in Down syndrome. Free Radical Res Com. 1991; 12-13:581–589. 30. Cohen G. The pathobiology of Parkinson’s disease: biochemical aspects of dopamine neuron senescence. J Neural Transm Supp. 1983; 19:89–103. 31. Saggu H, Cooksey J, Dexter D, Wells FR, Lees A, Jenner P, Marsden CD. A selective increase in particulate superoxide dismutase activity in parkinsonian substantia nigra. J Neurochem. 1989; 53:692–697. 32. Ceballos I, Lafron M, Javoy-Agid F, Hirsch E, Sinet PM, Agid Y. Superoxide dismutase and Parkinson’s disease. Lancet. 1990; 335:1035–1036. 33. Kish SJ, Morito C, Hornykiewicz O. Glutathione peroxidase activity in Parkinson’s disease brain. Neurosci Lett. 1985; 58:343–346. 34. Marttila RJ, Lorentz H, Rinne UK. Oxygen toxicity protecting enzymes in Parkinson’s disease. J Neurol Sci. 1988; 86:321–331.

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

201

35. Sian J, Dexter DT, Lees AJ, Daniel S, Agid Y, Javoy-Agid F, Jenner P, Marsden CD. Alterations in glutathione levels in Parkinson’s disease and other neurodegenerative disorders affecting basal ganglia. Ann Neurol. 1994; 36:348–355. 36. Hirsch EC, Faucheux BA. Iron metabolism and Parkinson’s disease. Movement Disord. 1998; 13:39–45. 37. Dexter DT, Carter G, Agid F, Agid Y, Lees AJ, Jenner P, Marsden CD. Lipid peroxidation as cause of nigral cell death in Parkinson’s disease. Lancet. 1986; 11:639–640. 38. Jenner P, Dexter DT, Sian J, Schapira AH, Marsden CD. Oxidative stress as a cause of nigral cell death in Parkinson’s disease and incidental Lewy body disease. Ann Neurol. 1992; 32:S82–87. 39. Klivenyi P, Andreassen OA, Ferrante RJ, Dedeoglu A, Mueller G, Lancelot E, Bogdanov M, Andersen JK, Jiang D, Beal MF. Mice deficient in cellular glutathione peroxidase show increased vulnerability to malonate, 3-nitropropionic acid, and 1-methyl-4-phenyl-1,2,5,6-tetrahydropyridine. J Neurosci. 2000; 20:1–7. 40. Bensadoun JC, Mirochnitchenko O, Inouye M, Aebischer P, Zurn AD. Attenuation of 6-OHDA-induced neurotoxicity in glutathione peroxidase transgenic mice. Eur J Neurosci. 1998; 10:3231–3236. 41. Ho YS, Magnenat JL, Bronson RT, Cao J, Gargano M, Sugawara M, Frank CD. Mice deficient in cellular glutathione peroxidase develop normally and show no increased sensitivity to hyperoxia. J Biol Chem. 1997; 272:16644–16651. 42. Naoi M, Maruyama W. Cell death of dopamine neurons in ageing and Parkinson’s disease. Mech Ageing Dev. 1999; 111:175–188. 43. Martinez M, Martinez N, Hernandez AI, Ferrandiz ML. Hypothesis: can N-acetylcysteine be beneficial in Parkinson’s disease. Life Sci. 1999; 64:1253–1257. 44. Dirnagl U, Iadecola C, Moskowitz MA. Pathobiology of ischemic stroke: an integrated view. Trends Neurosci. 1999; 22:391–397. 45. Kondo T, Reaume AG, Huang TT, Carlson E, Murakami K, Chen SF, Hoffman EK, Scott RW, Epstein CJ, Chan PH. Reduction of CuZn-superoxide dismutase activity exacerbates neuronal cell injury and edema formation after transient focal cerebral ischemia. J Neurosci. 1997; 17:4180–4189. 46. Murakami K, Kondo T, Kawase M, Li Y, Sato S, Chen SF, Chan PH. Mitochondrial susceptibility to oxidative stress exacerbates cerebral infarction that follows permanent focal cerebral ischemia in mutant mice with manganese superoxide dismutase deficiency. J Neurosci. 1998; 18:205–213. 47. Yang G, Chan PH, Chen J, Carlson E, Chen SF, Weinstein P, Epstein CJ, Kamii H. Human copper-zinc superoxide dismutase transgenic mice are highly resistant to reperfusion injury after focal cerebral ischemia. Stroke. 1994; 25:165–170. 48. Weisbrot-Lefkowitz M, Reuhl K, Perry B, Chan PH, Inouye M, Mirochnitchenko O. Overexpression of human glutathione peroxidase protects transgenic mice against focal cerebral ischemia/reperfusion damage. Brain Res Mol Brain Res. 1998; 53:333–338. 49. Crack PJ, Taylor JM, Flentjar NJ, de Haan J, Hertzog P, Iannello RC, Kola I. Increased infarct size and exacerbated apoptosis in the glutathione peroxidase1 (Gpx-1) knockout mouse brain in response to ischemia/reperfusion injury. J Neurochem. 2001; 78:1389–1399. 50. Rosen DR, Siddique T, Patterson D, Figlewicz DA, Sapp P, Hentati A, Donaldson D, Goto J, O’Regan JP, Deng HX, Rahmani Z, Krizus A, McKenna-Yasek D, Cayabyab A, Gaston SM, Berger R, Tanzi RE, Halperin JJ, Hertzfeldt B, Van den Bergh R, Hung WY, Bird T, Deng G, Mulder DW, Smyth C, Laing NG, Soriano E, Pericak-Vance MA, Haines J, Rouleau GA, Gusella JS, Horvitz HR, Brown RH Jr. Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature. 1993; 362:59–62.

202

THE AGEING BRAIN

51. Gurney ME, Pu H, Chiu AY, Dal Canto MC, Polchow CY, Alexander DD, Caliendo J, Hentati A, Kwon YW, Deng HX. Motor neuron degeneration in mice that express a human Cu, Zn superoxide dismutase mutation. Science. 1994; 264: 1772–1775. 52. Dal Canto MC, Gurney ME. A low expressor line of transgenic mice carrying a mutant human Cu,Zn superoxide dismutase (SOD1) gene develops pathological changes that most closely resemble those in human amyotrophic lateral sclerosis. Acta Neuropathol. 1997; 93:537–550. 53. Gurney ME, Cutting FB, Zhai P, Andrus PK, Hall ED. Pathogenic mechanisms in Familial Amyotrophic Lateral Sclerosis due to mutation of Cu, Zn superoxide dismutase. Pathol Biol. 1996; 44:51–56. 54. Liu R, Althaus JS, Ellerbrock BR, Becker DA, Gurney ME. Enhanced oxygen radical production in a transgenic mouse model of familial amyotrophic lateral sclerosis. Ann Neurol. 1998; 44:763–770. 55. Cha CI, Kim JM, Shin DH, Kim YS, Kim J, Gurney ME, Lee KW. Reactive astrocytes express nitric oxide synthetase in the spinal cord of transgenic mice expressing a human Cu/Zn SOD mutation. NeuroReport. 1998; 9:1503–1506. 56. Ferrante RJ, Shinobu LA, Schulz JB, Matthews RT, Thomas CE, Kowall NW, Gurney ME, Beal MF. Increased 3-nitrotyrosine and oxidative damage in mice with a human copper/zinc superoxide dismutase mutation. Ann Neurol. 1997; 42:326–334. 57. Gurney ME. Transgenic animal models of familial amyotrophic lateral sclerosis. J Neurol. 1997; 244:S15–20. 58. Gurney ME, Cutting FB, Zhai P. Benefit of vitamin E, riluzole and gabapentin in a transgenic model of familial amyotrophic lateral sclerosis. Ann Neurol. 1996; 39:147–157. 59. Zemlan FP, Thienhaus OJ, Bosmann HB. Superoxide dismutase activity in Alzheimer’s disease: possible mechanism for paired helical filament formation. Brain Res. 1989; 476:160–162. 60. Hensley K, Carney JM, Mattson MP, Aksenova M, Harris M, Wu JF, Floyd RA, Betterfield DA. A model for beta-amyloid aggregation and neurotoxicity based on free radical generation by the peptide: relevance to Alzheimer disease. Proc Natl Acad Sci USA. 1994; 91:3270–3274. 61. Iannello RC, Crack PJ, de Haan JB, Kola I. Oxidative stress and neural dysfunction in Down syndrome. J Neural Transm. 1999; 57:257–267. 62. Behl C, Davis JB, Lesley R, Schubert D. Hydrogen peroxide mediates amyloid beta protein toxicity. Cell. 1994; 77:817–827. 63. Multhaup G, Ruppert T, Schlicksupp A, Hesse L, Beher D, Masters CL, Beyreuther K. Reactive oxygen species and Alheimer’s disease. Biochem Pharmacol. 1997; 54:533–559. 64. Yan SD, Yan SF, Chen X, Fu J, Chen M, Kuppusamy P, Smith MA, Perry G, Godman GC, Nawroth P, et al. Non-enzymatically glycated tau in Alzheimer’s disease induces neuronal oxidant stress resulting in cytokine gene expression and release of amyloid beta-peptide. Nat Med. 1995; 1:693–699. 65. Pappolla MA, Chyan YJ, Omar RA, Hsiao K, Perry G, Smith MA, Bozner P. Evidence of oxidative stress and in vivo neurotoxicity of beta-amyloid in a transgenic mouse model of Alzheimer’s disease: a chronic oxidative paradigm for testing antioxidant therapies in vivo. Am J Pathol. 1998; 152:871–877. 66. Furuta A, Price DL, Pardo CA, Troncoso JC, Xu ZS, Taniguchi N, Martin LJ. Localization of superoxide dismutases in Alzheimer’s Disease and Down’s syndrome neocortex and hippocampus. Am J Pathol. 1995; 146:357–367. 67. Hajimohammadreza I, Brammer M. Brain membrane fluidity and lipid peroxidation in Alzheimer’s disease. Neurosci Lett. 1990; 112:333–337.

OXIDATIVE AND FREE RADICAL MECHANISMS IN BRAIN AGEING

203

68. Kola I, Cristiano F, de Haan JB, Sumarsono S, Thomas R, Corrick C, Tymms M. Genes, Embryogenesis and Down syndrome. In: Moeloek F, Affandi B, Trounson AO, editors. Advances in human reproduction. Casterton: Parthanon Publishing, 1995; 309–320. 69. Kola I, Hertzog PJ. Animal models in the study of the biological function of genes on human chromosome 21 and their role in the pathophysiology of Down syndrome. Hum Mol Genet. 1997; 6:1713–1727. 70. Anneren KG, Epstein CJ. Lipid peroxidation and superoxide dismutase-1 and glutathione peroxidase activities in trisomy 16 fetal mice and human trisomy 21 fibroblasts. Pediatr Res. 1987; 21:88–92. 71. Brooksbank BWL, Balazs R. Superoxide dismutase, glutathione peroxidase and lipoperoxidation in Down’s Syndrome fetal brain. Dev Brain Res. 1984; 16: 37–44. 72. Sinet PM, Michelson AM, Bazin A, Lejeune J, Jerome H. Increase in glutathione peroxidase activity in erythrocytes from trisomy 21 subjects. Biochem Biophys Res Com. 1975; 67:910–915. 73. Percy ME, Dalton AJ, Markovic VD, Crapper McLachlan DR, Hummel JT, Rusk ACM, Andrews DF. Red cell superoxide dismutase, glutathione peroxidase and catalase in Down syndrome patients with and without manifestations of Alzheimer’s disease. Am J Med Genet. 1990; 35:469–467. 74. Bladier C, de Haan JB, Kola I. Antioxidant genes and reactive oxygen species in Down syndrome. In: Sen CK, Sies H, Baeuerle PA, editors. Antioxidant and redox regulation of genes. San Diego: Academic Press, 2000; 425–449. 75. Busciglio J, Yankner BA. Apoptosis and increased generation of reactive oxygen species in Down’s syndrome neurons in vitro. Nature. 1995; 378:776–779. 76. Odetti P, Angelini G, Dapino D, Zaccheo D, Garibaldi S, Dagna-Bricarelli F, Piombo G, Perry G, Smith M, Traverso N, Tabaton M. Early glyoxidation damage in brains from Down’s syndrome. Biochem Biophys Res Com. 1998; 243: 849–851. 77. Mann DMA, Esiri MM. The pattern of acquisition of plaques and tangles in the brains of patients under 50 years of age with Down’s Syndrome. J Neurol Sci. 1989; 89:169–179. 78. Neve RL, Finch EA, Dawes LR. Expression of the Alzheimer amyloid precursor gene transcripts in the human brain. Neuron 1988; 1:669–677. 79. Sohal RS, Allen RG. Oxidative stress as a causal factor in differentiation and ageing: a unifying hypothesis. Exp Geront. 1990; 25:499–522.

Chapter 12 THE ROLE OF NUTRITIONAL FACTORS IN COGNITIVE AGEING Janet Bryan

Introduction The association between nutrition and cognitive performance has become a topic of increasing scientific and public interest. 1 Nutrition may be an important, modifiable, life-style factor in age-related cognitive decline and there is a growing interest in the development of nutritional supplements as therapeutic agents aimed at enhancing or maintaining cognitive function or delaying cognitive decline.2 The role that food and nutrition has in the course of age-related cognitive change is yet to be clearly defined. However, early findings from epidemiological and experimental studies suggest that there may be a role for dietary components such as: folate and vitamins B-12 and B-6, antioxidants, omega 3 fatty acids, and herbal supplements like Ginkgo biloba. The assumption that food and its components can impact on cognitive performance and age-related cognitive change is based on the knowledge that the central nervous system (CNS) depends heavily for efficient functioning on a constant supply of almost all of the essential nutrients as well as glucose and oxygen.3 If brain function is optimised, one would expect cognitive function to be optimised as well. Understanding this link between brain function and cognitive function is important in the formulation of hypotheses about diet-cognition links and how we might test them. We need to base research in the area of diet and cognition on hypotheses about the mechanisms by which nutrients or other aspects of the diet might impact on the brain, and how this in turn affects cognitive performance. Nutrients may impact on the brain in

206

THE AGEING BRAIN

a number of ways such as: on neurotransmitter synthesis, on the structure of neurons, and on the vasculature of the brain. The effects of nutrients on the brain may be short-term and acute, or more longer-term due to dietary habits over an extended period of time. Importantly, if we understand the mechanisms by which nutrients affect the brain, we can more accurately predict the role that nutrients may have in cognitive ageing and which aspects of cognition are likely to be affected by nutritional factors. At a very basic level, we might expect nutritional factors to impact on fluid, rather than crystallised, abilities among older adults.4 Fluid ability is thought to reflect innate information processing that is determined by genetic and physiological factors such as the integrity of the CNS. Fluid ability is demonstrated in the capacity to process novel information, that is, to apply mental processes to situations that require no previous knowledge. Crystallised ability refers to the application of learned information and cultural experience acquired over the lifetime and therefore depends on our store of knowledge, education and cultural background. Due to its reliance on the integrity of the CNS, we might expect nutritional factors to have more of an impact on fluid, rather than crystallised, ability. Furthermore, there are some aspects of cognition that show marked agerelated decline, such as memory performance and cognitive resources, that may also be sensitive to nutritional factors. Cognitive resources refer to our mental capacity to perform cognitive tasks. The three identified cognitive resources are: speed of information processing (how fast we can think); working memory capacity (how much information we can simultaneously store and work on); and attentional capacity (how long we can concentrate for). Because these resources are so important for the efficient working of the cognitive system,5 any investigation of the links between nutrition and cognitive performance should include measures of cognitive resources.4 In summary, because cognition is multidimensional, it is crucial that pertinent tests of cognitive performance be selected in order to test hypotheses that specify the impact that nutritional factors might have on the cognitive system. Folate, and Vitamins B-12 and B-6 Recent research has focussed on the role of B vitamins, especially folate, B-12 and B-6, in cognitive ageing.6 Although most of the evidence for the link between B vitamins and cognitive performance is based on studies involving participants with clinical vitamin deficiencies, the effect of these vitamins might also be important for a broader population.3, 7 In particular, a mild to moderate folate deficiency is thought to be relatively common in the general population,8 while the incidence of clinical and subclinical levels of folate and B-12 deficiency has been found to be higher in the elderly.8–12 Stabler et al.13 estimate that low serum B-12 concentrations might be evident in 5–15% of the elderly population. They suggest multiple causes of B-12 deficiency in

NUTRITION AND COGNITIVE AGEING

207

older adults such as: pernicious anaemia, atrophic gastritis causing B-12 malabsorption, or previous gastric or intestinal surgery. Evidence is continuing to mount that higher intakes and serum concentrations of certain vitamins, particularly folate and B-12, might be beneficial for cognitive performance.14 Research to date indicates two inter-linked neurochemical mechanisms by which these B vitamins influence cognitive performance through their role in methylation in the CNS.15–17 The first mechanism (the hypomethylation hypothesis) posits that folate, with B-12 or B-6 as catalysing cofactors, may have a direct and possibly acute effect on the CNS via hypomethylation, which inhibits the synthesis of methionine and the formation of S-adenosylmethionine (SAMe). This in turn inhibits methylation reactions throughout the CNS involving proteins, membrane phospholipids, DNA, the metabolism of neurotransmitters such as the monoamines (e.g. dopamine, norepinephrine, serotonin), and melatonin, all of which are crucial to neurological and psychological status.16–19 The second mechanism (the homocysteine hypothesis) proposes that there is an indirect and possibly longer-term effect of folate, B-12 and B-6 on the functioning of the brain via the cerebrovasculature. Studies have demonstrated that high levels of homocysteine, largely attributable to low levels of folate, and B-12 or B-6, are associated with increased risk of vascular disease14,15,20–22 due to toxic effects on vascular tissue, or excessive production of excitotoxic sulphur amino acids, homocysteic acid and cystein sulphinic acid.11 Thus, these B vitamins may function to preserve the integrity of the CNS via their role in the prevention of vascular disease, which is crucial to cognitive function.15,16,23–25 Cross-sectional studies Most studies investigating the links between the B vitamins and cognitive performance among older adults have employed correlational designs. Goodwin et al.26 found significant associations between lower dietary intakes of folate and reduced abstract thinking and problem solving performance, but found no effects for B-12 or B-6. Ortega et al.10, 27 found that those with higher scores on the Mini Mental State Examination (MMSE) had higher blood levels and dietary intake of folate, but again found no effects for B-12 or B-6. Other studies have investigated the individual and combined relationships of folate and B-12 with cognition. Bell et al.28 found that individuals with less than median values of serum folate and B-12 had lower scores on the MMSE than those above the median value for either folate or B-12 or both. Wahlin et al.29 found that those with low serum folate or a combination of low folate and B-12 performed more poorly on tests of episodic memory than did those with normal folate and B-12 levels. Hassing et al.30 found that individuals with low serum folate levels performed more poorly on tests of episodic memory. Furthermore, there is some evidence to suggest that homocysteine levels mediate the relationship between the B vitamins and cognitive performance. Riggs et al.31 found that higher homocysteine, lower folate and B-6 blood levels were associated with poorer spatial copying performance, and that lower

208

THE AGEING BRAIN

B-6 was also associated with backward Digit Span performance. Interestingly, homocysteine levels were found to mediate the relationships between folate and B-6 and cognitive performance suggesting that these vitamins impact on cognition via homocysteine status. The results from these studies examining cross-sectional associations between B vitamins and cognitive performance suggest that low folate intake and/or status emerges as the most reliable associate of cognitive performance, either alone or in combination with low B-12. Many aspects of cognition appear to be related to B vitamin status, especially memory performance and measures of abstract reasoning. In addition, the relationship between the B vitamins and cognition may be mediated by homocysteine levels since homocysteine uniquely predicted cognitive performance after controlling for B vitamin status in the study by Riggs et al.31 Longitudinal studies The findings of cross-sectional studies have been supported by longitudinal studies. In a six-year follow up of a healthy subsample of the original Goodwin et al.26 study, La Rue et al.32 found a positive association between past intake of B-12 and B-6 and current cognitive status. Ebly et al.33 found that those in lowest serum folate quartiles were more likely to have cognitive loss and to be depressed, and that they were more likely to be demented, institutionalised and to have a higher mortality rate at two-year follow-up. Results from longitudinal studies allow for the examination of possible long-term effects of B vitamin intake and status on cognition at a later date, or the impact on cognitive change. The results of these studies suggest that prior intake of B vitamins is a predictor of cognitive performance at a later date. The findings of Ebly et al.33 suggest that low folate status may be a predictor of cognitive decline among older adults. Experimental studies Very few studies have manipulated B vitamin intake experimentally and assessed its affects on cognitive performance, and only three studies have used a placebo-controlled design. Tolonen et al.34 investigated B-6 status among older Finnish and Dutch adults aged between 64 to 96 years, and younger Dutch adults aged from 22 to 55 years. In addition, they gave daily oral doses of 2 mg of B-6 for one year to 20 Finnish older adults, while 24 matched participants received a placebo. Biochemical results clearly indicated that both Finnish and Dutch older adults had lower B-6 levels than the younger adults. Clock drawing performance was improved by supplementation relative to controls but there were no significant effects of supplementation on memory or Digit Span performance. Deijen et al.35 also investigated the effects of B-6 supplementation on cognitive performance and mood among 76 men aged between 70 and 79 years. They gave the men 20 mg of B-6 or placebo daily for three months. Significant positive effects of B-6 supplementation, compared with placebo, were found on measures of the amount of information retained in long-term memory,

NUTRITION AND COGNITIVE AGEING

209

but there were no effects for iconic or short-term memory. The authors concluded that B-6 supplementation might have a modest but significant effect in improving the storage of information, thereby reducing age-related memory loss. Fioravanti et al.36 used a double-blind, placebo-controlled study to assess the effects of folate supplementation (15 mg daily for 60 days, a dose well above the Recommended Daily Intake) on the memory performance of 30 community or aged-care dwelling participants, aged 70 to 90 years. Participants were selected for low folate levels ( or = 80 years of age) humans. Acta Neuropath. 1994; 88:212–221. 63. Ellis RJ, Olichney JM, Thal LJ, Mirra SS, Morris JC, Beekly D, Heyman A. Cerebral amyloid angiopathy in the brains of patients with Alzheimer’s disease: the CERAD experience part XV. Neurology. 1996; 46:1592–1596. 64. Hardy JA, Mann DMA, Wesler P, Winblad B. An integrative hypothesis concerning the pathogenesis and progression of Alzheimer’s disease. Neurobiol Aging. 1986; 7:489–502. 65. Terry RD, Katzman R. Lifespan and synapses: will there be a primary senile dementia? Neurobiol Aging. 2001; 22:347–248. 66. Linn RT, Wolf PA, Bachman DL, Knoefel JE, Cobb JL, Belanger AJ, Kaplan EF, D’Agostino RB. The preclinical phase of probable Alzheimer’s disease. A 13-year prospective study of the Framingham cohort. Arch Neurol. 1995; 52:485–490. 67. Morris JC, Storandt M, Miller JP, McKeel DW, Price JL, Rubin EH, Berg L. Mild cognitive impairment represents early-stage Alzheimer disease. Arch Neurol. 2001; 58:397–405. 68. Ebly EM, Hogan DB, Parhad IM. Cognitive imairment in the nondemented elderly. Results from the Canadian Study of Health and Aging. Arch Neurol. 1995; 52:612–619. 69. Brayne C, Nickson J, Johnson A. Cognitive function and dementia in six areas of England and Wales: the distribution of MMS and the prevalence of GMS organicity level in the MRC CFA Study. Psychol Med. 1998; 28:319–335. 70. Moynihan R, and Smith R. (2002) Too much medicine? Br Med J. 2002; 324: 859–860. 71. Shapin S, Martyn C. How to live forever: lessons of history. Br Med J. 2000; 321: 1580–1582. 72. Tallis RC. Brains and minds: a brief history of neuromythology. J Roy Coll Phys Lond. 2000; 34:563–567. 73. Novella JL, Jochum C, Jolly D, Morrone I, Ankri J, Bureau F, Blanchard F. Agreement between patients’ and proxies’ reports of quality of life in Alzheimer’s disease. Qual Life Res. 2001; 10:443–452. 74. Gallagher M, Gill M, Baxter MG, Bucci DJ. Semin Neurosci. 1994; 6:351–358. 75. Wainwright NWH, Surtees PG, Gilks WR. Diagnostic boundaries, reasoning and depressive disorder. I Development of a Probabilistic model for public health psychology. Psychol Med. 1997; 27:835–845. 76. 76. Surtees PG, Wainwright NWH, Gilks WR. Diagnostic boundaries, reasoning and depressive disorder. II. Diagnostic complexity and depression: time to allow for uncertainty. Psychol Med. 1997; 27:847–860.

Chapter 15 DETECTING ALZHEIMER’S DISEASE AT THE PRE-SYMPTOMATIC STAGES Gary W Small

Introduction As people liver longer, the risk for developing Alzheimer’s disease (AD) increases dramatically. In fact, the incidence appears to double every five years after age 60 years, suggesting that if people lived long enough, they would all develop the disease by a certain age. Although AD is the most common cause of late-life dementia, other causes, particularly vascular disease, do contribute to the occurrence of dementia. In fact, the burden of such vascular decline appears to contribute to a greater portion of dementia cases in the upper age groups. Whether it is pure AD, pure vascular dementia, or something along the continuum, these cases of dementia progress with time. The underlying lesions reach a threshold such that they lead to cognitive decline that interferes with daily life. With Alzheimer’s dementia, this slow insidious decline represents an accumulation of pathological features and declining neurotransmitter functions that begin well before the clinician can confirm a clinical diagnosis in practice. Adapted in large part from: Small GW. Structural and functional imaging of Alzheimer’s disease. In: Davis KL, Charney D, Coyle JT, Nemeroff C, editors. Neuropsychopharmacology: the fifth generation of progress. Philadelphia: Lippincott, Williams and Wilkins, 2002; 1231–1242.

260

THE AGEING BRAIN

Focus on Early Detection With the realization that the neuropathological changes of AD begin to accumulate perhaps decades before the disease is obvious clinically, studies have emphasized recruitment of subjects at time points years or decades before a physician begins to focus on early detection of AD at clinical stages. These studies might confirm a clinical diagnosis of probable AD.1 The ultimate goal is to develop tools to identify pre-symptomatic candidates for beginning preventive pharmacological treatments before extensive neuronal damage develops. Brain imaging has become an important tool for the development of surrogate markers that will effectively identify people with only mild cognitive losses who are likely to progress in their cognitive loss and eventually develop the full dementia syndrome of AD. As novel, disease-modifying agents emerge, these surrogate brain-imaging markers will be critical in determining drug efficacy and facilitating drug development in both animal models and human studies. Diagnostic Categories Clinical investigators have developed definitions for categorical pre-symptomatic stages that assist in clinical trials and communicating staging levels. Several diagnostic entities have been described in efforts to better characterize age-related cognitive decline. The mildest form of age-related memory decline is known as age-associated memory impairment (AAMI),2 characterized by self-perception of memory loss and a standardized memory test score > 1SD below the aged norms. In people 65 years of age or older, its estimated prevalence is 40%, afflicting approximately 16 million people in the United States.3 Only about 1% of such cases will develop dementia each year. The term AAMI has generated controversy since many people question the specific criteria and whether they define a stable or declining entity. A more severe form of memory loss is mild cognitive impairment (MCI), often defined by significant memory deficits without functional impairments. People with MCI show memory impairment that is > 1.5 SDs below aged norms on such memory tasks as delayed paragraph recall.4 Approximately 10% of people 65 years or older suffer from MCI, and nearly 15% develop AD each year.4,5 This condition also has generated controversy. Some experts consider MCI a risk state rather than a diagnostic entity. Despite such controversy, MCI appears to be a useful concept and may respond favorably to current symptomatic treatments. Brain imaging studies of pre-symptomatic AD focus on both these forms of age-related memory decline.

DETECTING ALZHEIMER’S DISEASE AT THE PRE-SYMPTOMATIC STAGES

261

Evidence for Pre-Symptomatic Changes Several areas of research have contributed to the idea that pre-symptomatic conditions do exist, including neuropathological, neuroimaging, and clinical investigations. Taken together, this work supports the notion that the dementing process leading to AD begins years before a clinical diagnosis of probable AD can be confirmed.6 Post-mortem studies of non-demented older people7 indicate that tangle density in healthy ageing correlates with age, but that some cases demonstrate widely distributed neuritic and diffuse plaques throughout neocortical and limbic structures. Other studies8 have found that neurofibrillary tangle density increases in some individuals, presumably those who will eventually develop AD very early in adult life, perhaps even by the fourth decade. The diffuse amyloid deposits in middle-aged non-demented subjects are consistent with an early or pre-symptomatic stage of AD and suggest that the pathological process progresses gradually, taking 20 to 30 years to proceed to the clinical manifestation of dementia.9 Other supportive evidence includes findings that linguistic ability in early life predicts cognitive decline in late life.10 High diffuse plaque density in non-demented older persons has been observed in the entorhinal cortex and inferior temporal gyrus, in association with acetylcholinesterase fibre density.11 Evidence from animal models also supports compromised hippocampal cholinergic transmission during ageing.12 Studies of glucose metabolic rates using positron emission tomography (PET)6,13,14 indicate lower regional brain metabolism in middleaged and older persons with a genetic risk (apolipoprotein E [ApoE ε4]) +ve status, lending further support for a prolonged pre-symptomatic AD stage. Structural Imaging Computerized tomography and magnetic resonance imaging The largest body of data comes from studies of structural imaging modalities in that clinical investigators have had the greatest access to these technologies. Studies of early detection logically follow from initial work demonstrating the differential diagnostic utility of a brain-imaging marker. For structural imaging, particularly magnetic resonance imaging (MRI), data have emerged on the use of regional atrophy patterns for the positive diagnosis of AD and other neurodegenerative disorders. Studies without neuropathological confirmation report the utility of medial temporal lobe atrophy, particularly hippocampal atrophy, on computerized tomography (CT) or MRI for the clinical diagnosis of AD.15 Some but not all quantitative MRI studies indicate that white matter hyperintensities correlate with neuropsychological functioning in both healthy elderly persons and demented patients.16,17 Other studies indicate loss of cerebral gray matter,18 hippocampal and parahippocampal atrophy,19 and lower left amygdala and entorhinal cortex volumes20 in patients with AD. In differentiating AD from older normal controls, the sensitivity of various

262

THE AGEING BRAIN

medial temporal atrophy measures ranges from 77% to 92%, with specificities ranging from 49% to 95%.21–23 In older MCI patients, hippocampal atrophy predicts subsequent conversion to AD.24 Of various analytic methods, computerized volumetric techniques are most accurate, but are currently labor-intensive and not widely available. A modified negative-angle axial view designed to cut parallel to the anterior-posterior plane of the hippocampus has been used to assess hippocampal volume using CT or MRI.15 Such hippocampal atrophy is a sensitive and specific predictor of future AD in patients with MCI. Baseline hippocampal ratings accurately predicted decliners with an overall accuracy of 91%. Neuropathological studies find that the sites of maximal neuronal loss for both AD and MCI are in the CA1, subiculum, and entorhinal cortex.15 Hippocampal atrophy also has been found to predict future cognitive decline in older persons without cognitive impairment followed for nearly four years. Visual assessments of medial temporal lobe atrophy on coronal MRI sections show significant correlations between estimated and stereologically measured volumes.25 Because the latter is much more labor-intensive, visual readings may be an alternative approach with greater efficiency. The hippocampus and the temporal horn of the lateral ventricles also may serve as antemortem AD markers in mildly impaired patients (mean MMSE score of 24).26 While hippocampal atrophy may distinguish AD from normal ageing, such atrophy may be non-specific, occurring in other dementing disorders.27 Magnetic resonance imaging hippocampal atrophy measures are not as sensitive as PET glucose metabolism measures, which begin decreasing before memory decline onset.28 The presence of MRI white matter hyperintensities does not improve diagnostic accuracy since they occur both in AD and healthy normal elderly.29,30 The entorhinal cortex (EC), a region involved in recent memory performance, is one of the earliest areas to accumulate NFTs.8 Histological boundaries of the EC from autopsy-confirmed AD patients and controls have been used to validate a method for measurement of EC size relying on gyral and sulcal landmarks visible on MRI.31 Such measures may be additional early AD detection markers. Several studies have addressed the interaction between regional atrophy and ApoE genotype. Increasing dose of ApoE ε4 allele was associated with smaller hippocampal, entorhinal cortical, and anterior temporal lobe volumes in already demented patients.32 A study of non-demented older persons found an association between ApoE ε4 dose and a larger left than right hippocampus.33 Combining medial temporal measures with other functional neuroimaging34 (34) or ApoE genotyping may improve the ability of any of these measures alone to predict cognitive decline.35 In vivo imaging of amyloid plaques and neurofibrillary tangles This is the most innovative structural imaging approach currently under development. Although not yet widely available, the technology offers con-

DETECTING ALZHEIMER’S DISEASE AT THE PRE-SYMPTOMATIC STAGES

263

siderable promise in studies of new drug development and eventually in differential diagnosis and early detection of dementia. The evidence for neuritic plaque (NP) and neurofibrillary tangle (NFT) accumulation years prior to clinical AD diagnosis suggests that in vivo methods that directly image these pathognomic lesions would be useful presymptomatic detection technologies. Current methods for measuring brain amyloid, such as histochemical stains, require tissue fixation on post-mortem or biopsy material. Available in vivo methods for measuring NPs or NFTs are indirect (e.g., cerebrospinal fluid measures).36 Studies that may lead to direct in vivo human A imaging include various radio-labelled probes using small organic and organometallic molecules capable of detecting differences in amyloid fibril structure or amyloid protein sequences.37 Investigators also have used chrysamine-G, a carboxylic acid analogue of Congo red; an amyloid-staining histologic dye; 38 serum amyloid P component, a normal plasma glycoprotein that binds to amyloid deposit fibrils;39 or monoclonal antibodies.40 Methodological difficulties that hinder progress with these techniques include poor blood–brain barrier crossing and limited specificity and sensitivity. In addition, most approaches do not measure both NPs and NFTs. In a recent study, Barrio et al. 41 used a hydrophobic radiofluorinated derivative of 1,1-dicyano-2-[6-(dimethylamino)naphthalen-2-yl]propene (FDDNP)42 with PET to measure the cerebral localization and load of NFTs and SPs in AD patients (N=7) and controls (N=3). The FDDNP was injected intravenously and found to readily cross the blood-brain barrier in proportion to blood flow, as expected from highly hydrophobic compounds with high membrane permeability. Greater accumulation and slower clearance of FDDNP was observed in brain regions with high concentrations of NPs and NFTs, particularly the hippocampus, amygdala, and entorhinal cortex. The FDDNP residence time in these regions showed significant correlations with immediate and delayed memory performance measures,43 and areas of low glucose metabolism correlated with high FDDNP activity retention. The probe showed visualization of NFTs, NPs and diffuse amyloid in AD brain specimens using in vitro fluorescence microscropy, which matched results using conventional stains (e.g. thioflavin S) in the same tissue specimens. Thus, FDDNP-PET imaging is a promising non-invasive approach to longitudinal evaluation of NP and NFT deposition in preclinical AD. Magnetic resonance spectroscopy This approach is another innovative technique that is only recently been studied in the context of dementia. Initial studies of MRS as a preclinical AD detection technique found significantly lower NAA concentrations in AD and AAMI subjects compared with controls.44 Mean inositol concentration was significantly higher in AD than in controls, whereas AAMI subjects had intermediate values. Another study focused on patients with Down’s syndrome because they invariably develop Alzheimer-type pathology by the time they reach their thirties or forties. Concentrations of myoinositol and

264

THE AGEING BRAIN

choline-containing compounds using 1H MRS were significantly higher in the occipital and parietal regions in 19 non-demented adults with Down’s syndrome and 17 age- and sex-matched healthy controls.45 Moreover, older Down’s syndrome subjects (42–62 years) had higher myo-inositol levels than younger subjects (28–39 years) suggesting that this approach may eventually be useful as a preclinical AD marker. Functional Imaging Positron emission tomography In recent years, clinicians have shown greater interest in the use of PET scanning for dementia diagnosis as access to PET centers has increased and mounting evidence of its diagnostic accuracy has come to light. Using fluorodeoxyglucose PET (FDG-PET), our group reported that parietal hypometabolism predicted future AD in people with questionable dementia46 and that even people with very mild age-related memory complaints have baseline PET patterns predicting cognitive decline after three years.47 These initial studies using PET for early AD detection emphasized family history of AD as a risk factor for future cognitive decline. A change in focus came with the discovery of the ApoE genetic risk for AD. The first report combining PET imaging and ApoE genetic risk in people with a family history of AD included 12 non-demented relatives with ApoE ε4, 19 relatives without ApoE ε4, and compared them to seven probable AD patients.14 “At-risk” subjects had mild memory complaints, normal cognitive performance, and at least two relatives with AD. Subjects with ApoE ε4 did not differ from those without ApoE ε4 in mean age (56.4 vs. 55.5 years) or in neuropsychological performance. Parietal metabolism was significantly lower and left-right parietal asymmetry higher in at-risk subjects with ApoE ε4 compared to those without ApoE ε4. Patients with dementia had significantly lower parietal metabolism than did at-risk subjects with ApoE ε4. The following year, Reiman et al.6 replicated these results and extended them to other brain regions. They found hypometabolism in temporal, prefrontal and posterior cingulate regions in a study of 11 non-demented ApoE ε4 homozygotes (4/4 genotype) and 22 ApoE ε3 homozygotes (3/3 genotype) of similar ages, i.e. in their mid-fifties, to those in our own initial study. They also applied an automated image analysis method, wherein metabolic reductions were standardized using three-dimensional stereotactic surface projections from FDG PET scans of AD patients compared with controls.48 The results from these two studies6,14 provided independent confirmation of an association between genetic risk and regional cerebral glucose hypometabolism. Our group recently confirmed these two initial reports in a study that included none of the subjects participating in our previous report on ApoE and PET.14 We studied 65 subjects in the 50 to 84 year age range (mean+SD

DETECTING ALZHEIMER’S DISEASE AT THE PRE-SYMPTOMATIC STAGES

265

= 67.3+9.4 years), with or without a family history of AD.49 Of the 65 subjects, 54 were non-demented (27 ApoE ε4 carriers and 27 were subjects without ApoE ε4), and 11 were demented and diagnosed with probable AD.49 The non-demented subjects were aware of a gradual onset of mild memory complaints (e.g. misplacing familiar objects, difficulty remembering names) but had memory performance scores within the norms for cognitively intact persons of the same age and educational level. The ApoE ε4 carriers had a small and non-significant but consistent reduction in cognitive performance. As predicted, baseline comparisons among the three subject groups indicated the lowest metabolic rates for the AD group, intermediate rates for the nondemented ApoE ε4 carriers, and highest rates for the non-demented group without ApoE ε4 in several cortical regions, including inferior parietal, lateral temporal, and posterior cingulate (Figure 1). Additional studies using FDG-PET have focused on older patients with Down’s syndrome who are at risk for AD.50 The investigators hypothesized that an audiovisual stimulation paradigm would serve as a stress test and reveal abnormalities in parietal and temporal cerebral glucose metabolism before dementia developed. At mental rest, younger and older patients with Down’s syndrome did not differ in glucose metabolic patterns. During audiovisual stimulation, however, the older patients showed significantly lower parietal and temporal metabolism. Families with familial AD linked to chromosome 14 or APP mutations have been studied with FDG-PET as well.51 In such families with early-onset AD, approximately half of relatives who live to the age at risk will develop AD. While pedigree members with AD show typical parietal and temporal hypometabolism, asymptomatic relatives at risk for AD show a similar but less severe hypometabolic pattern. Single photon emission computed tomography Johnson et al.52 used SPECT with a 99mTc-HMPAO to study longitudinal cerebral perfusion of patients with questionable AD (Clinical Dementia Rating [CDR] = 0.5)53 and controls. Regional decreases in perfusion in patients whose diagnosis converted to AD were most prominent in the hippocampalamygdaloid complex, the anterior and posterior cingulate, and the anterior thalamus. Including ApoE status did not influence results. A direct comparison of FDG-PET and HMPAO-SPECT in their ability to differentiate AD from vascular dementia indicated higher diagnostic accuracy for PET regardless of dementia severity.54 Using ROC curves, PET diagnostic accuracy was better than SPECT for MMSE > 20 (87.2% vs. 62.9%) and for MMSE < 20 (100% vs. 81.2%). Other studies have confirmed a lower sensitivity for even high-resolution SPECT compared with PET.55 Moreover, the parietal hypoperfusion observed using SPECT in AD patients has been observed in such other conditions as normal ageing, vascular dementia, post-hypoxic dementia, and sleep apnea.56

266

THE AGEING BRAIN

Figure 1. Examples of PET images (comparable parietal lobe levels) co-registered to each subject’s baseline MRI scan for an 81-year-old non-demented woman (ApoE 3/3 genotype; upper images), a 76-year-old non-demented woman (ApoE 3/4 genotype; middle images), and 79-year-old woman with AD (ApoE 3/4 genotype; lower images). The last column shows two-year follow-up scans for the non-demented women. Compared with the nondemented subject without ApoE ε4, the non-demented ApoE ε4 carrier had 18% (right) and 12% (left) lower inferior parietal cortical metabolism, while the demented woman’s parietal cortical metabolism was 20% (right) and 22% (left) lower, as well as more widespread metabolic dysfunction due to disease progression. Two-year follow-up scans showed minimal parietal cortical decline for the woman without ApoE ε4, but bilateral parietal cortical decline for the non-demented woman with ApoE ε4, who also met clinical criteria for mild AD at follow-up. MRI scans were within normal limits.49

Functional MRI Two recent studies have combined ApoE genotyping and fMRI in persons at risk for AD. Bookheimer at al.57 performed fMRI studies while 30 cognitively intact middle-aged and older persons (mean age 63 years) memorized and retrieved unrelated word pairs. The 16 ApoE ε4 carriers did not differ

DETECTING ALZHEIMER’S DISEASE AT THE PRE-SYMPTOMATIC STAGES

267

Figure 2. These statistical parametric maps of recall vs. control blocks for ApoE ε4 carriers and non-carriers were standardized into a common coordinate system. Both groups showed significant MRI signal intensity increases in frontal, temporal and parietal regions, and the ApoE ε4 group had greater extent and intensity of activation. The ApoE ε4 group showed additional activation in the left parahippocampal region, left dorsal prefrontal cortex, and other regions in the inferior and superior parietal lobes, and anterior cingulate.57

significantly from the 14 subjects without ApoE ε4 in age, prior educational achievement, or rates of AD family history. Brain activation patterns were determined during both learning and retrieval task periods and analyzed using between-group and within-subject approaches. Memory performance was reassessed on 12 subjects after two years of follow-up. The ApoE ε4 carriers had significantly greater magnitude and spatial extent of MRI signal intensity during memory performance in regions affected by AD, including bilateral hippocampal and left parietal and prefrontal regions (Figure 2). This pattern of activation was greater in the left hemisphere, consistent with the verbal nature of the task, and during the retrieval rather than the learning condition. Longitudinal data indicated that greater baseline brain activation correlated with verbal memory decline assessed two years later. The greater signal in

268

THE AGEING BRAIN

subjects with the ApoE ε4 genetic risk suggests that the brain may recruit additional neurons to compensate for subtle deficits. Moreover, the longitudinal data are encouraging that functional MRI may be a useful approach to prediction of future cognitive decline and early AD detection. By contrast, other kinds of memory tasks may produce different patterns of brain activation. In another study of persons at risk for AD, visual naming and letter fluency tasks were used to activate brain areas involved in object and face recognition during functional MRI scanning.58 Subjects in the highrisk group had at least one first-degree relative with AD and one ApoE ε4 allele. The low risk group was matched for age, education, and cognitive performance. The high-risk group showed reduced activation in the midand posterior inferotemporal regions bilaterally. Such decreased activation patterns could result from subclinical neuropathology in the inferotemporal region or in the inputs to that region. Longitudinal studies of glucose metabolism of persons at risk for dementia Two research groups — UCLA and University of Arizona — have reported their longitudinal FDG-PET follow-up data on non-demented persons at risk for AD. At UCLA, a total of 20 non-demented subjects (10 ApoE ε4 carriers and 10 without ApoE ε4) have received repeat PET and neuropsychological testing two years after baseline assessment (mean+SD for follow-up was 27.9+1.7 months).49 The 10 ApoE ε4 carriers available for longitudinal study were similar to the 10 non-carriers in mean+SD age (67.9+8.9 vs. 69.6+8.1 years) and educational achievement (14.4+1.8 vs. 16.4+2.8 years). Memory performance scores did not differ significantly according to genetic risk either at baseline or follow-up and the ApoE ε4 carriers and non-carriers did not differ significantly in cognitive change after two years. The ROI analysis of PET scans performed after two years showed significant glucose metabolic decline (4%) in the left posterior cingulate region in ApoE ε4 carriers. The SPM analysis showed significant metabolic decline in the inferior parietal and lateral temporal cortices with the greatest magnitude (5%) of metabolic decline in the temporal cortex (Figure 3). After correction for multiple comparisons, this decline remained significant for the ApoE ε4 group, wherein a decrease in metabolism was documented for every subject. Based upon these data from only 10 subjects, the estimated power of PET under the most conservative scenario is 0.9 to detect a 1-unit decline from baseline to follow-up using a one-tailed test. Such findings suggest that combining PET and AD genetic risk measures will allow investigators to use relatively small sample sizes when testing anti-dementia treatments in preclinical AD stages. The University of Arizona group also found that ApoE ε4 heterozygotes have significant two-year declines in regional brain activity, the largest of which is in temporal cortex, and that these reductions are significantly greater than those in ApoE ε4 non-carriers. Their findings suggest that as few as 22 cognitively normal, middle-aged ApoE ε4 heterozygotes

DETECTING ALZHEIMER’S DISEASE AT THE PRE-SYMPTOMATIC STAGES

269

Figure 3. Regions showing the greatest metabolic decline after two years of longitudinal follow-up in non-demented subjects with ApoE ε4 (SPM analysis) included the right lateral temporal and inferior parietal cortex (brain on the left side of figure). Voxels undergoing metabolic decline (p