CC75C : Abstracts

Papers published and in press, with abstracts.

CC75C study design and overview

An epidemiological study of the dementias in Cambridge: from clinical progression to neuropathology.
Brayne C, Huppert FA, Xuereb JH, Gertz H-J, Chi L-Y, McGee MA, Paykel ES, Harrington C, Mukaetova-Ladinska E, O'Sullivan A, Dening T, Freer C and Wischik CM.
In: Iqbal K, Winblad B, Nichimura T, Takeda M and Wisniewski HM (Eds) Alzheimer's disease: biology, diagnosis and therapeutics. John Wiley & Sons Ltd., 1997, pp 11-20.

In the 1980s a large poppulation survey of people aged 75 years and over was undertaken by O'Connor and colleagues in Cambridge city, UK. This provided an estimation of the prevalence of dementia and formed part of the EURODEM group of dementia prevalence studies. The non-demented population was followed a little over two years later and the incidence of dementia, clinically diagnosed subtypes and cognitive decline as measured by MMSE were reported. diagnostic stability from the incidence interview to follow - up interview, in those diagnosed as clinically demented (including minimal), was good. Cognitive decline in those interviewed on three occasions showed a tendency for decline across the entire population, more marked with advanced age, and was greater in those receiving diagnoses of any severity of dementia. Decline was greater in those with less education and lower social class. The apolipoprotein E ɛ4 allele showed the anticipated association with both decline and clinical diagnosis in a subsample. The first 50 brains from the linked neuropathological study have been assessed. The findings were examined in relation to the final diagnosis, and the cognitive decline, based on all the survey material. Good agreement was found in the AD category, but not in the vascular categories. Neuropathological lesions map loosely to the clinical state of individuals, but not closely. Further larger studies of a similar nature with careful characterization of potential confounders and effect modifiers are necessary to clarify the tru relationship of putative genetic and pathophysiological indices to patterns of decline during life.

The Cambridge Project for Later Life: Design and preliminary results.
Brayne C, Huppert F, Paykel E and Gill C.
Neuroepidemiology, 1992; 11(Suppl 1): 71-75.

The Cambridge Project for Later Life is the follow-up at 2.4 years of the Hughes Hall Project for Later Life, a prevalence study of dementia in Cambridge city in which 40% of the population aged 75 and over were screened for dementia. In the follow-up, 1,173 people were screened a second time, and using Mini-Mental Scale Examination scores were selected for a more intensive interview with CAMDEX. This was followed by detailed neuropsychological testing and magnetic resonance imaging in a smaller proportion of individuals.

Dementia and dementia sub-types

The prevalence of dementia as measured by the Cambridge Mental Disorders of the Elderly Examination.
O'Connor DW, Pollitt PA, Hyde JB, Fellows JL, Miller ND, Brook CPB, Reiss BB and Roth M.
Acta Psychiatrica Scandinavica, 1989; 79: 190-198.

General practice patients aged 75 years and over were screened for cognitive impairment using the Mini-Mental State Examination. Those scoring 23 or below and a sample of those scoring 24 or 25 were assessed using the Cambridge Mental Disorders of the Elderly Examination (CAMDEX), a structured interview schedule specifically designed to detect mild dementia. The CAMDEX includes a mental state examination, a psychiatric history, detailed cognitive testing and an informant interview. The prevalence of dementia in 2311 patients was found to be 10.5%, about half that found in most earlier studies. Possible reasons for this low rate are discussed.

The validity of informant histories in a community study of dementia.
O'Connor DW, Pollitt PA, Brook CP and Reiss BB.
International Journal of Geriatric Psychiatry, 1989; 4: 203-208.

The CAMDEX is a structured interview schedule intended for use in community studies of dementia. It includes an informant interview in addition to a mental status examination and cognitive testing. In a Cambridge dementia prevalence study, the information given by informants, most of whom were relatives, was highly consistent with elderly subjects' scores on cognitive testing and with the observations made by research psychiatrists. Their reports were in serious disagreement with other parts of the assessment in only a small number of cases. Neither social class nor the relationship between informant and subject appeared to have any bearing on the consistency of informants' reports.

Mild dementia: Perceptions and problems.
Pollitt PA, O'Connor DW and Anderson I.
Ageing and Society, 1989; 9: 261-275.

The focus of this paper is on the beginnings of dementia - on the grey area where normal and abnormal ageing seem to overlap, but where a diagnosis can be established. We look at a group of elderly people diagnosed as suffering from mild dementia and at the relatives most closely involved with them and whom we had assumed to be their carers. Our principal interest is in the relatives' perception of the deterioration in intellectual function, and in their awareness of and response to problems associated with it. Contrary to expectation, these relatives did not see themselves as carers, or the elderly person as demented. Spouses often saw their partner as no more disabled than themselves; and more generally, the relationship between them often showed a high degree of reciprocity. Sons and daughters were usually aware of changes in their parents' behaviour but tended to explain them in terms of normal ageing. Improvement in the process of early identification and the creation of more appropriate services are generally seen as desirable. Questions are raised about the usefullness and justification for intervention in a situation which is not yet recognised by those involved as requiring it.

Memory complaints and impairment in normal, depressed and demented elderly persons identified in a community survey.
O'Connor DW, Pollitt PA, Brook CPB and Reiss BB.
Archives General Psychiatry, 1990; 47: 224-227.

Normal, depressed, and demented elderly persons who were identified by means of a community survey were asked to assess their memories and to complete a battery of memory tests. Depressed elderly persons reported memory problems more often than normal subjects and reported indecisiveness, impaired concentration, and mental slowing more often than demented subjects. However, memory complaints and memory performance correlated poorly in the normal and depressed groups. Normal and demented elderly persons who reported memory problems achieved higher scores on a series of questions about depression than those who denied memory problems.

Problems reported by relatives in a community study of dementia.
O'Connor DW, Pollitt PA and Reiss BB.
British Journal of Psychiatry, 1990; 156: 835-841.

The supporters of 120 demented elderly people completed rating scales of the frequency and severity of the problems they faced, the amount of strain they experienced and their psychological well-being. Problems and strain increased with the degree of dementia. Problems relating to physical dependency, forgetfulness and inertia were relatively common, while disturbed behaviours were relatively uncommon. Physical dependency and disturbed behaviour were less well tolerated than forgetfulness and inertia, but all problem categories were positively associated with measures of strain. Wives reported more problems and strain than husbands, and co-resident children were under greater strain than children who lived independently.

Co-existing depression and dementia in a community survey of the elderly.
O'Connor DW, Pollitt PA and Roth M.
International Psychogeriatrics, 1990; 2(1)

We report here on the coexistence of dementia and depression in a community population aged 75 years and older. Complete information about mood and cognition was available for 286 cognitively intact subjects selected for assessment because of their low scores on the Mini-Mental State, and for 158 mildly and moderately demented subjects. Severely demented subjects, who were incapable of providing information, were excluded. Five percent (8/158) of demented subjects also fulfilled criteria for major depressive disorder Diagnostic and Statistical Manual of Mental Disorders, third edition (DSM-III) compared with 9% (27/286) of cognitively intact subjects. No substantial differences existed in the symptoms reported by demented depressives and nondemented depressives, but subjects who suffered from both disorders were so markedly apathetic that their depression might easily have been overlooked had specific enquiries not been made. Depression was particularly associated with dementia secondary to multi-infarct and Parkinson's disease. When reviewed one year later, 2 of the 18 surviving depressed, nondemented subjects showed evidence of dementia. Both presented unusual diagnostic difficulties, however, and no evidence emerged that large numbers of elderly people will be misclassified in community surveys that include a mental state examination, cognitive testing, and an informant interview.

The contribution of CAMDEX to the Diagnosis of Mild Dementia in Community Surveys.
O'Connor DW.
Psychiatric Journal of University of Ottawa, 1990, 15(4): 216-20.

Distinguishing between cognitively intact and mildly demented elderly people poses particular difficulties. Now that modern criteria, such as those contained in the DSM-III-R, specify that cognitive deficits be sufficient to interfere with the pursuit of daily tasks, investigators must consider the relative effects of personality, physical illness and handicap, sensory deficits and functional psychiatric disorders since all of these factors contribute to loss of independence. Simple tests and questionnaires are inadequate to the task and are subject to demographic bias but structured assessments conducted by skilled clinicians using operational criteria have much to offer, notwithstanding the increased costs and time required. This paper discusses some of the issues involved and presents evidence concerning the use of CAMDEX in a recent British survey.

A follow-up study of dementia diagnosed in the community using the Cambridge Mental Disorders of the Elderly Examination.
O'Connor DW, Pollitt PA, Hyde JB, Fellowes JL, Miller ND and Roth M.
Acta Psychiatrica Scandinavica, 1990; 81: 78-82.

Elderly Cambridge residents diagnosed as demented using the Cambridge Mental Disorders of the Elderly Examination (CAMDEX) were reviewed approximately 12 months later. Diagnoses were confirmed in 133 of 137 surviving cases (97%). Subjects said to have minimal dementia (cognitive impairment insufficient to warrant a diagnosis of dementia proper) had a varied outcome. Only 6 out of 29 survivors showed progressive intellectual deterioration and 13 were reclassified as normal. Subjects passed as normal in the first year of the study were reviewed using the Mini-Mental State Examination. We cannot be certain how many were actually dementing, but our findings suggest that only a small number of false negative diagnoses were made in the first year of the study.

Continued clinical validation of dementia diagnosed in the community using the Cambridge Mental Disorders of the Elderly Examination.
O'Connor DW, Pollitt PA, Jones BJ, Hyde JB, Fellowes JL and Miller ND
Acta Psychiatrica Scandinavica, 1991; 83: 41-45.

We describe the outcome 2 years later of elderly Cambridge residents who had been diagnosed as demented using the Cambridge Mental Disorders of the Elderly Examination. Mortality rates were high but diagnoses were confirmed for all of the 43 moderate and severe cases who survived for this period, and for 51 of the 56 subjects who had been rated initially as mildly demented; 28 of the 56 mildly demented subjects progressed to more severe levels of impairment. As many as 12 of the 24 original minimally demented cases showed evidence of intellectual deterioration, which lends weight to the validity of this experimental category.

The progression of mild idiopathic dementia in a community population.
O'Connor DW, Pollitt PA, Hyde JB, Fellowes JL, Miller ND and Roth M.
Journal of the American Geriatric Society, 1991, 39, 246-251.

Thirty-one subjects aged 75 years and over who were identified as suffering from mild, idiopathic dementia in a large community survey were reviewed at annual intervals for 2 years. Diagnoses and severity ratings were based on defined criteria following a mental state examination, a medical and psychiatric history, detailed cognitive testing, and an interview with relatives or other key informants. Fourteen subjects became more severely demented within 2 years. The initial cognitive test battery failed to reveal any differences between respondents whose dementia advanced and those whose condition remained unchanged, but, in the former group, subjects' symptoms had been present for longer, and a greater proportion had been recognised as demented, or possibly demented, by their general practitioners. We suggest that subjects whose dementia progressed had actually been more severely disabled at the time of identification.

The influence of education and social class on the diagnosis of dementia in a community population.
O'Connor DW, Pollitt PA and Treasure FP.
Psychological Medicine, 1991; 21: 219-224.

We have reported previously that poorly educated elderly people and those of low social class were at much increased risk of scoring below the customary cut-point on the Mini-Mental State Examination, a widely-used, brief cognitive screening test. As part of the same study, subjects who scored 23 or less on the MMSE out of a maximum of 30 points, and a sample of those who scored 24 or 25 points, were assessed by psychiatrists using a structured, diagnostic interview. Assuming that persons who scored 26 points or above were cognitively intact (our data suggest that 2% or less were not), neither educational attainment nor social class had any influence on the likelihood that subjects would be diagnosed as demented. Our data suggest that social and psychological factors contribute substantially to cognitive test scores and serve to emphasize the importance of detailed assessment procedures in epidemiological surveys of dementia.

The Prevalence of Dementia in Europe: a collaborative study of 1980-l990 findings.
Rocca WA, Brayne C, Breteler MMB, Clarke M, Cooper B, Copeland JRM, Dartigues JF, Hofman A, Hagnell O, Heeren TJ, Engedal K, Jonker C, Lindesay J, Lobo A, Mann AH, Molsa PK, Morgan K, O'Connor DW, Sulkava R, Kay DW, Amaducci L, and Da Silva Droux A.
International Journal of Epidemiology, 1991, 20: 736-748.

To obtain age- and gender-specific estimates of the prevalence of dementia in Europe and to study differences in prevalence across countries, we pooled and re-analysed original data of prevalence studies of dementia carried out in some European countries between 1980 and 1990. The study followed these steps: census of existing datasets, collection of data in a standardized format, selection of datasets suitable for comparison, comparison of age and gender patterns. From the 23 datasets of European surveys considered, 12 were selected for comparison. Only population-based studies in which dementia was defined by DSM-III or equivalent criteria and in which all subjects were examined personally were included. Studies in which institutionalized subjects were not investigated were excluded. Age- and gender-specific prevalences were compared within and across studies and overall prevalences were computed. Although prevalence estimates differed across studies, the general age- and gender-distribution was similar for all studies. The overall European prevalences for the five-year age groups from 60 to 94 years, were 1.0, 1.4, 4.1, 5.7, 13.0, 21.6 and 32.2%, respectively. In subjects under 75 years the prevalence of dementia was slightly higher in men than in women; in those aged 75 years or over the prevalence was higher in women. The prevalence figures nearly doubled with every five years of increase in age.

Incidence of dementia in a population older than 75 years in the United Kingdom.
Paykel ES, Brayne C, Huppert FA, Gill C, Barkley C, Gehlhaar E, Beardsall L, Girling DM, Pollitt P and O'Connor D.
Archives of General Psychiatry 1994; 51: 325-332.

BACKGROUND: Incidence studies have been relatively neglected in psychiatric epidemiology. They are particularly important for dementia, since prevalence rates are affected by length of survival, which itself falls with increasing age and presence of dementia. METHODS: Two-wave community study of 1195 elderly subjects aged older than 75 years, restudied 2.4 years after a community prevalence study. A two-stage method was used, comprising the Mini-Mental State Examination followed in a stratified sample by the Cambridge Examination for Mental Disorders of the Elderly (CAM-DEX) interview. Incidence rates were based on person-years at risk. RESULTS: Annual incidence rates for dementia were 2.3% for subjects initially aged 75 to 79 years, 4.6% for ages 80 to 84 years, and 8.5% for ages 85 to 89 years, approximately doubling every 5 years. Rates did not differ significantly by sex, educational level, or social class. Twice as many additional individuals received a diagnosis of minimal dementia not reaching case threshold. CONCLUSIONS: The findings show high rates of new onset dementia, increasing markedly with age, and suggest rapid acceleration of one or more processes that is common in advanced age.

The incidence of clinically diagnosed subtypes of dementia in an elderly population. Cambridge Project for Later Life.
Brayne C, Gill C, Huppert FA, Barkley C, Gehlhaar E, Girling D, O'Connor D and Paykel E.
British Journal of Psychiatry, 1995; 167: 255-262.

BACKGROUND. In developed countries, most dementia appears to be due to Alzheimer's disease and vascular dementia. We report rates for incidence of subtypes of dementia based on clinical diagnosis. METHOD. This study was a 2.4-year (s.d. 2.6 months) follow-up of a cohort aged 75 years and over, seen initially in a prevalence study of dementia. A screening interview in 1173 survivors was followed in a subsample of 461 respondents by a diagnostic interview 1.8 months after screening (s.d. 1.5 months). This comprised a standardised interview with respondent and informant, with venepuncture where possible. Clinical diagnoses of subtypes were made by specified criteria. RESULTS. The incidence of Alzheimer's disease of mild and greater severity was 2.7/1000 person-years at risk (1.6-4.4); in men 1.5 (0.8-2.7) and in women 3.3 (1.8-5.9). The incidence of vascular dementia was 1.2/100 person-years at risk (0.7-1.9); in men 1.1 (0.4-2.8) and in women 1.2 (0.7-2.0). Alzheimer's disease, but not vascular dementia, showed a marked increase with age, particularly in women. Rates for minimal dementia of different subtypes showed similar age and sex effects, but were much higher for Alzheimer's disease than vascular dementia. CONCLUSIONS. The striking rise in incidence rates of dementia in the very old appear to be due to Alzheimer's disease, while rates for vascular dementia remain relatively constant. These trends are particularly marked for minimal dementia, but emphasise the importance of Alzheimer's disease in the community as a cause of cognitive decline of all degrees.

Cognitive decline in an elderly population; a two-wave study of change.
Brayne C, Gill C, Paykel ES, Huppert FA and O'Connor DW.
Psychological Medicine, 1995; 25: 673-683.

A sample of 1111 survivors from a population aged 75 years and over completed the Mini-Mental State Examination (MMSE) twice, separated by 28 months on average. There was a mean decline of 1.3 points in MMSE score. With increasing age, mean drop in score also increased. The proportion at each age identified as newly cognitively impaired according to any standard cut-point on MMSE rose markedly. Mean decline was greater in women than men even after adjustment for age. Cognitive change on the MMSE was approximately unimodally and normally distributed. This distribution was a marked contrast to the distribution of MMSE scores themselves, which was skewed due to truncation of scores at the maximum. The decline was not due to the inclusion of individuals with physical impairment. These findings indicate that cognitive decline, like dementia, becomes increasingly common with advancing age, and suggest that dementia may be regarded as one extreme of the continuum of cognitive decline.

Vascular risk factors and incident dementia: results from a cohort study of the very old.
Brayne C, Gill C, Huppert FA, Barkley C, Gehlhaar E, Girling DM, O'Connor DW and Paykel ES.
Dementia, 1998; 9: 175-180.

The contribution of vascular pathology to the manifestation of dementia and the importance of vascular risk to measures of cognitive function is being increasingly recognized. In particular, confirmation of this risk points towards approaches for prevention in large sections of the population. Information on determinants of incident dementia is increasing, but still relatively few studies of risk have been based on incident cases of dementia in very elderly populations. In this study based on incident cases of dementia in a population aged 75 and over, vascular risks were obtained from informants of the respondents with incident dementia. When compared with controls the factors associated with incident dementia were history of heart attack (odds ratio 2.9), transient ischaemic attacks (4.8), cerebrovascular accidents (3.4), family history of first-degree relatives with dementia (4.0), and occupational exposure to vibrating instruments (1.4). If only Alzheimer's disease, clinically diagnosed, was included, diabetes (1.4) and a history of dementia in first-degree relatives (6.6) emerged. Thus, vascular risk continues to be of importance in the oldest age groups.

Incidence of dementia and cognitive decline in over-75s in Cambridge: Overview of cohort study.
Paykel ES, Huppert FA and Brayne C.
Social Psychiatry and Psychiatric Epidemiology, 1998; 33: 387-392.

This paper summarises the methods and some of the findings of a large cohort study of dementia and cognitive decline in subjects aged over 75 years in Cambridge, particularly regarding the incidence wave. From a sample of 1968 subjects previously studied in a prevalence study in 1985-1987, survivors were restudied at 2.4 years, in a two-stage design employing the Mini; Mental State Examination (MMSE) and the Cambridge Examination for Mental Disorders of the Elderly (CAMDEX). High incidence rates of dementia were found, which rose steeply with age, particularly for Alzheimer's disease. New minimal dementia and milder cognitive impairment were also common. Cognitive decline on the MMSE showed a near normal, non-bimodal distribution. The sample has since been restudied at intervals for a total of up to 9 years to document longitudinal cognitive change. Brains have been obtained for post mortem neuropathological and molecular biological study, particularly of the early sequential changes associated with cognitive decline and dementia.

The EURODEM collaborative re-analysis of case-control studies of Alzheimer's disease: implications for public health.
Brayne C.
International Journal of Epidemiology, 2002; 20(Suppl 2): 568-571.

In the EURODEM pooling and re-analysis of case-control studies of Alzheimer's disease it has been possible to examine putative risk factors with increased power to detect associations. The fundamental problems of case and control selection persist, such as use of prevalent cases, selection through contact with specific services, difficulties of control choice. Risk factors such as family history and head trauma are shown again, although the biases introduced in collection of exposure data could still account for these findings. Other associations which are shown, such as smoking may be accounted for by factors related to survival and use of prevalent cases. The direct public health implications of these findings are limited. Intervention based on many of the associations found in this re-analysis would have relatively low impact on overall rates of Alzheimer's disease because of the small proportion of the population exposed. The total public health impact of any such intervention would be also limited according to the contribution which Alzheimer's disease makes to overall rates of dementia. Improvement of cardiovascular indices may improve the cerebrovascular status of the population, possibly reducing the incidence of vascular dementia. Other broad strategies to maintain health and function would seem prudent, but specific recommendations to reduce the incidence of Alzheimer's disease, or to slow progression of the disorder cannot be recommended on the basis of these re-analyses. It is clear that more research is needed to understand the risks of different pathologies related to Alzheimer's disease as well as dementia and cognitive change generally in the population.(ABSTRACT TRUNCATED AT 250 WORDS)

Neuropsychology

The influence of education, social class and sex on Mini-Mental State scores.
O'Connor DW, Pollitt PA, Treasure FP, Brook CP and Reiss BB.
Psychological Medicine, 1989; 19: 771-776.

The Mini-Mental State Examination was administered to 1865 general-practice patients aged 75 years and over. Even when demented cases were removed from analysis, respondents with relatively little education, together with those in social classes III-manual and below, were significantly more likely to score below the cut-off point used in North American community surveys to denote 'cognitive impairment'. Education and social class influenced scores on all sections within the MMSE with the exception of registration. Sex influenced scores on tests of calculation and spelling backwards but had no effect on total scores. These findings emphasize the importance of investigating low scorers in more detail before making a diagnosis of dementia.

The reliability and validity of the Mini-Mental State in a British community survey.
O'Connor DW, Pollitt PA, Hyde JB, Fellows JL, Miller ND, Brook CP and Reiss BB.
Journal of Psychiatric Research, 1989; 23(1): 87-96.

The Mini-Mental State (MMSE) was administered to 2302 general practice patients aged 75 yr and over. Those scoring 23 or less and a sample of those scoring 24 or 25 were selected for further investigation using the Cambridge Mental Disorders of the Elderly Examination. Inter-observer reliability was high with a mean kappa value of 0.97. Eighty-six per cent of respondents judged to have organic mental disorders scored 23 or less on the MMSE and 92% of those judged to be cognitively intact scored 24 or more. However, only 55% of respondents who scored 23 or less were demented or delirious while a number of relatively well educated, mildly demented subjects scored 24 and 25. The customary cut-off point of 23/24 may need to be revised in future community studies. MMSE scores cannot be used to make even tentative psychiatric diagnoses; more detailed investigation of low scorers is essential.

A comparison of clinical, psychometric and behavioural memory tests: findings from a community study of the early detection of dementia.
Beardsall L and Huppert FA.
International Journal of Geriatric Psychiatry 1991; 6: 295- 306.

Diagnostic criteria for dementia require assessment of the ability to learn new information. Ideally, this assessment should be brief and sensitive to early impairment. We present comparative data on a variety of memory tests administered to a population sample of elderly community residents who are being studied longitudinally. Of the three classes of memory tests, clinical, psychometric and everyday tests, discriminant function analysis indicated that, overall, the best discrimination between demented and non-demented individuals was achieved by some everyday tests: recalling a news item, retracing a route, and recalling a name after learning a name-photograph association. However, of the 11 tests examined in total, two were identified which together discriminated almost as efficiently as all 11 measures combined. These were recalling a name and recalling six objects. By comparison, the principal memory item in the Mini Mental State Examination, the recall of three words, produced poor discrimination. The two simple measures which provided good discrimination in this population sample may therefore be recommended to assess the ability to learn new information, where the aim of the assessment is the diagnosis of dementia.

Prospective memory impairment as an early indicator of dementia.
Huppert FA and Beardsall L.
Journal of Clinical and Experimental Neuropsychology 1993; 15(5): 805-821.

Theoretical accounts of the cognitive processes involved in prospective memory imply that performance on such tasks will be more vulnerable than retrospective memory tests to the early stages of dementia. This hypothesis is examined in elderly subjects from a general population sample which includes demented subjects. We report the findings on three tests of prospective memory from the Rivermead Behavioural Memory Test (Wilson et al., 1985) and compare it with measures of retrospective memory for recently acquired information. In contrast to retrospective memory where subjects with minimal dementia perform at a level intermediate between normals and more demented subjects, the subjects with minimal dementia perform as poorly as more demented subjects on the prospective memory tests. These findings provide evidence that prospective memory tasks are particularly sensitive to the early stages of dementia. Covariance analysis and the pattern of intercorrelations found between prospective and retrospective memory lead to alternative hypotheses about the cognitive processes involved in prospective memory and the way in which they break down.

Improvement in NART word reading in demented and normal older persons using the Cambridge Contextual Reading Test.
Beardsall L and Huppert FA.
Journal of Clinical and Experimental Neuropsychology 1994; 16(2): 232-242.

A test that assesses ability to pronounce correctly a list of irregular words (NART) has become widely used to estimate IQ particularly in elderly and demented subjects. This estimate relies on the assumption that words that are not pronounced correctly were not previously in the subject's vocabulary. This assumption was questioned by the results of a community study showing that older adults commonly mispronounced even simple NART words that were almost certainly in their vocabulary. The present study investigated the extent to which putting NART words into sentences facilitated correct pronunciation. Both normal and demented subjects substantially improved their performance when the words were seen in context. The improvement was most marked for demented subjects and for poor or average readers as compared with skilled readers. Since irregular words can only be pronounced correctly if they are in the individual's vocabulary, it appears that the NART underestimates reading ability and, hence, underestimates premorbid IQ in certain groups. We conclude that the Cambridge Contextual Reading Test (CCRT) provides a more accurate estimate of reading ability and IQ in elderly and demented subjects.

CAMCOG - A concise neuropsychological test to assist dementia diagnosis: Socio-demographic determinants in an elderly population sample.
Huppert FA, Brayne C, Gill C, Paykel ES and Beardsall L.
British Journal of Clinical Psychology 1995; 34: 529-541.

The CAMCOG, which forms part of the CAMDEX interview (Roth et al., 1986, 1988), is a brief neuropsychological battery designed to assess the range of cognitive functions required for a diagnosis of dementia, and to detect mild degrees of cognitive impairment. It was administered to a population sample of 418 elderly people (aged 77 and above) in their place of residence. The data show that in contrast to the Mini-Mental State Examination, total CAMCOG scores are well distributed and there is no ceiling effect. Examination of the association between CAMCOG scores and socio-demographic variables (age, sex, education and social class) shows that each exerts a significant, and independent, effect upon performance. CAMCOG also includes a number of subscale which assess individual areas of cognitive function. Of the eight major subscales (orientation, language, memory, attention, praxis, calculation, abstract thinking, perception), age was significantly related to all but attention; sex with attention, praxis, calculation and perception; education with language and abstract thinking; and social class with language and perception. In all these analyses, the results were adjusted for the effects of the other socio-demographic variables using analysis of variance. However, education and social class are highly correlated variables and when the association with education is examined without adjusting for social class, attention and praxis are also found to be significantly related to education. Caution must therefore be taken when using the CAMCOG (or any other cognitive test) as a screening test for dementia, using a single, predetermined cutpoint. In general, the combination of brevity and breadth of the CAMCOG, along with its distributional properties, makes it an attractive neuropsychological test for use in the community or the clinic.

Psychometric properties of the CAMCOG and its efficacy in the diagnosis of dementia.
Huppert FA, Jorm AF, Brayne C, Girling DM, Barkley C, Beardsall L and Paykel ES.
Aging, Neuropsychology and Cognition, 1996; 3(3): 201-214.

The Cambridge Cognitive Examination (CAMCOG) is a concise neuropsychological test for the assessment of cognitive impairment in elderly people. It was designed specifically to assist in the diagnosis of dementia in an early stage. CAMCOG assesses a broad range of cognitive functions, as is required for the diagnosis of dementia, and it minimizes floor and ceiling effects by covering a range of item difficulty. We report here some psychometric properties of the CAMCOG in two stage of population survey of people over the age of 75 years. Total score on the CAMCOG was found to have excellent internal reliability and test-retest reliability. The reliability of the individual subscales, which corresponded to different cognitive abilities, was also acceptable. Principal components analysis on the individual CAMCOG items revealed two easily interpretable factors, corresponding to general intelligence and memory. CAMCOG scores were very effective in differentiating between demented and nondemented individuals. The CAMCOG total score, as well as each subscale score, differed significantly between nondemented individuals and those with the diagnosis of mild dementia or minimal dementia. Regression analysis demonstrated that CAMCOG score alone was a strong predictor of dementia diagnosis. CAMCOG total scores showed high levels of sensitivity and specificity in differentiating between nondemented individuals and those with a diagnosis of mild dementia. The cut-point which produced the highest levels of both sensitivity and specificity was 80/81, with values of 93% and 87% respectively.

Short NART, CCRT and Spot-the-Word: Comparisons in older and demented persons.
Beardsall L and Huppert FA.
British Journal of Clinical Psychology 1997; 36(4): 619-622.

This study compares the efficacy of three measures of premorbid intelligence: the National Adult Reading Test (NART), the Cambridge Contextual Reading Test (CCRT), and the Spot-the-Word Test. The results in a population sample of elderly and demented participants show that test efficacy varies between groups. In a demented group and a normal group of average readers, the CCRT leads to a higher estimate of premorbid word reading ability than the NART. Spot-the-Word results in good performance by normal groups and participants with minimal dementia, but performance is grossly impaired in subjects with mild/moderate dementia. We conclude that each test may be appropriate for specific groups.

Decline across different domains of cognitive funtion in normal aging: results of a longitudinal population-based study using CAMCOG.
Cullum S, Huppert FA, McGee M, Dening T, Ahmed A, Paykel ES and Brayne C.
International Journal of Geriatric Psychiatry, 2000; 15(9): 853-62.

Dementia is an important cause of disability in the elderly. There is evidence that cognitive impairment in dementia is on a continuum with cognitive impairment in the non-demented elderly. In order to investigate this possibility, we need detailed knowledge about the population distribution of cognitive function and change in cognitive function. The aim of this study is to describe the change in different domains of cognitive function over 4 years in a population based sample of non-demented elderly people, and to investigate the effect of sociodemographic variables and baseline cognitive function on change in each of the cognitive domains. Respondents from two group gerneral practice lists (n=503) were interviewed using the Cambridge Cognitive Examination (CAMCOG) at the incidence wave of the Cambridge City Over - 75 Cohort Study and after a mean time period of 3.9 years. One hundred and thirty five of 212 mon-demented subjects seen at follow-up completed the CAMCOG at both interviews. The annual rate of change in total CAMCOG score was - 1.6 points per year (p<0.001). There was statistically significant decline in all of the CAMCOG subscales. Greater decline in the Memory subscale was associated with less education (p=0.03). Greater decline in the Attention/Calculation subscale was associated with manual social class (p=0.05). Greater decline in the perception subscale was associated with older age (p = 0.03). Decline in specific cognitive domains may indicate a reversible phase of cognitive impairment and deserves further investigation.

Cognitive functions in UK community-dwelling African Caribbean and white elders: a pilot study.
Richards M, Brayne C, Dening TR, Abas M, Carter J, Price M, Jones C and Levy R.
International Journal of Geriatric Psychiatry, 2000; 15: 621-630

In recent years there has been interest in risk of cognitive impairment and dementia in populations of African origin. Little is known about this risk in older African Caribbean residents in the UK. One difficulty is lack of consensus over an adequate cognitive test battery for this community. Forty-five African Caribbean and 45 age and gender matched white community residents were recruited by household enumeration of an inner London electoral ward. These participants were administered the MMSE during a screening interview. Where possible, this was followed up by tests from the CERAD and CAMCOG neuropsychological batteries, a medical examination, and a structured interview with an informant. Based on these data, a psychiatrist blind to ethnicity independently rated 86 of these participants (41 of the African Caribbeans, all 45 of the whites) as cognitively normal, cognitively impaired, or demented. Of 41 African Caribbeans, 18 (44%) were rated as cognitively normal, 9 (22%) were rated as cognitively impaired, and 14 (34%) were rated as demented. Of the 45 whites, 39 (87%) were rated as cognitively normal, 4 (9%) were rated as cognitively impaired, and 2 (4%) were rated as demented. African Caribbeans scored significantly lower than whites in most cognitive test scores, which was not accounted for by their lower educational and occupational attainment, or their higher frequency of cardiovascular risk factors. African Caribbean elders in the UK appear to be at high risk of cognitive impairment and dementia. However, the influence of potential confounding factors such as socio-economic position and ill-health, and the effect of cultural test bias, cannot be ruled out. Copyright 2000 John Wiley & Sons, Ltd.

Population norms for the MMSE in the very old: estimates based on longitudinal data.
Dufouil C, Clayton D, Brayne C, Chi LY, Dening TR, Paykel ES, O'Connor DW, Ahmed A, McGee MA and Huppert FA.
Neurology, 2000; 55: 1609-1613

OBJECTIVE: To report the percentile distribution of Mini-Mental State Examination (MMSE) scores in older people by age, sex, and education level, estimated from longitudinal data, after correcting for loss due to dropout.

METHODS: The Cambridge City over 75 Cohort is a population-based study of a cohort of 2106 subjects age 75 years and older at study entry followed up over 9 years. At each of the four waves, cognitive function was assessed using MMSE. Based on these data, the relationship between age and MMSE score was modeled. Percentile distributions by age, sex, and education level were provided using inverse probability weighting to correct for dropouts.

RESULTS: Performance on MMSE was related to age in men and women. In women, at age 75, MMSE score ranged from 21 (10th percentile) to 29 (90th percentile). At age 95, the range was 10 (10th percentile) to 27 (90th percentile). The upper end of MMSE distribution was slightly modified with age, whereas the lower end of the distribution was very sensitive to age effect. A similar pattern was observed in both sexes.

CONCLUSION: These findings provide norms for MMSE scores in subjects age 75 years and older from longitudinal population-based data. Such norms can be used as reference values to determine where an individual’s score lies in relation to his or her age, sex, and education level.

A comparison of parametric models for the investigation of the shape of cognitive change in the older population.
Graciela Muniz, Fiona Matthews, Carol Brayne
BMC Neurology 2008, 8:16

Background
Cognitive decline is a major threat to well being in later life. Change scores and regression based models have often been used for its investigation. Most methods used to describe cognitive decline assume individuals lose their cognitive abilities at a constant rate with time. The investigation of the parametric curve that best describes the process has been prevented by restrictions imposed by study design limitations and methodological considerations. We propose a comparison of parametric shapes that could be considered to describe the process of cognitive decline in late life. Attrition plays a key role in the generation of missing observations in longitudinal studies of older persons. As ignoring missing observations will produce biased results and previous studies point to the important effect of the last observed cognitive score on the probability of dropout, we propose modelling both mechanisms jointly to account for these two considerations in the model likelihood.

Methods
Data from four interview waves of a population based longitudinal study of the older population, the Cambridge City over 75 CohortStudy were used. Within a selection model process, latent growth models combined with a logistic regression model for the missing data mechanism were fitted. To illustrate advantages of the model proposed, a sensitivity analysis of the missing data assumptions was conducted.

Results
Results showed that a quadratic curve describes cognitive decline best. Significant heterogeneity between individuals about mean curve parameters was identified. At all interviews, MMSE scores before dropout were significantly lower than those who remained in the study. Individuals with good functional ability were found to be less likely to dropout, as were women and younger persons in later stages of the study.

Conclusions
The combination of a latent growth model with a model for the missing data has permitted to make use of all available data and quantify the effect of significant predictors of dropout on the dropout and observational processes. Cognitive decline over time in older persons is often modelled as a linear process, though we have presented other parametric curves that may be considered.

Neurobiology

Quantitative analysis of tau protein in paired helical filament preparations: implications for the role of tau protein phosphorylation in PHF assembly in Alzheimer's disease.
Wischik CM, Lai RYK, Gertz HJ, Edwards CR, Xuereb JH, Paykel ES, Brayne C, Huppert FA, Mukaetova-Ladinska EB, Roth M and Harrington CR.
Neurobiology of Aging 1995; 16(3): 409-431.

In Alzheimer's disease, there is a major redistribution of the tau protein pool from soluble to PHF-bound forms. PHF-bound tau can be distinguished from normal tau by acid reversible occlusion of a generic tau epitope in the tandem repeat region and characteristic sedimentation in the if-II protocol developed in this laboratory. We show that 85% of tau bound in the PHF-like configuration can be recovered in the if-II PHF-fraction. Less than 1% of this material was phosphorylated at the mAb AT8 site in aged clinical controls or in cases with minimal or mild dementia. Of tau phosphorylated at the mAb AT8 site, only 12% was found to co-sediment with PHFs. These low levels could not be explained by postmortem dephosphorylation. As more than 95% of PHF-tau is not phosphorylated, even at early stages of pathology, it is misleading to use the terms "PHF-tau" and "phosphorylated tau" as though they were synonymous, particularly as this implies a pathogenetic role which phosphorylation need not have.

Examination of phosphorylated tau protein as a PHF-Precursor at early stage Alzheimer's disease.
Lai RYK, Gertz HJ, Wischik CM, Xuereb JH, Mukaetova-Ladinska EB, Harrington CR, Edwards PC, Mena R, Paykel ES, Brayne C, Huppert FA and Roth M.
Neurobiology of Aging 1995; 16(3): 433-445.

Hyperphosphorylated tau protein which can be isolated on the basis of insolubility in 1% sarkosyl (A68-tau fraction) is thought to represent a precursor pool for PHF assembly, associated histologically with neuritic pathology, which feeds into a more resistant tangle-associated PHF pool via cross-linking and proteolysis. We examined these predictions at the earliest detectable stages of neurofibrillary pathology. We report that there is no evidence that neuritic pathology represents an early pathologic stage, no evidence of an association between neuritic pathology and phosphorylated tau, no evidence of selective accumulation of phosphorylated tau at early stages of pathology, and no evidence for a precursor/product relationship between phosphorylated tau and PHFs during progression of pathology. We conclude that altered phosphorylation is a secondary process affecting 5% of PHFs and does not explain PHF assembly in Alzheimer's disease.

The relationship between clinical dementia and neuropathological staging (Braak) in a very elderly community sample.
Gertz H-J, Xuereb J, Huppert F, Brayne C, Kruger H, McGee MA, Paykel ES, Harrington C, Mukaetova- Ladinska E, O'Connor DW and Wischik CM.
European Archives of Psychiatry and Clinical Neuroscience, 1996; 246(3): 132-136.

The neuropathological staging model proposed by Braak and Braak (1991) implies that the evolution of neurofibrillary pathology follows a predictable sequence and can be ordered in a regular regional hierarchy. A total of 42 cases of an elderly population sample, which had been prospectively clinically assessed, were examined. Clinical diagnosis was made according to the CAMDEX criteria, and the sample reported here did not include cases were vascular dementia according to the criteria proposed by Chui et al. (1991). The neuropathological staging procedure was applied as originally proposed by Braak and Braak (1991). In addition, in all cortical laminae and regions which are essential for the staging model neurofibrillary tangles were quantified. Demented cases had significantly more areas involved and more advanced neuropathological stages. Cases with stages 1-3 tended to be non-demented, and cases with stages 4-6 tended to be demented. However, there was a considerable degree of overlap and no clear-cut threshold could be established. This brings into question the diagnostic value of the staging model.

Examination of the validity of the hierarchical model of neuropathological staging in normal ageing and Alzheimer's Disease.
Gertz H-J, Xuereb J, Huppert FA, Brayne C, McGee MA, Paykel ES, Harrington C, Mukaetova-Ladinska E, Arendt T and Wischik CM.
Acta Neuropathologica, 1998; 95: 154-158.

The neuropathological staging model of Alzheimer's disease proposed by Braak and Braak [Acta Neuropathol (1991) 82 : 259] requires that the evolution of neurofibrillary pathology follows a predictable pattern that can be ordered in a regular regional hierarchy. We have operationalized the neuropathological staging system to permit testing of its validity. Forty-two cases were derived from an epidemiological study of cognitive function in an elderly population for which post-mortem brain tissue was collected. Cases with neuropathological diagnoses other than Alzheimer's disease and normal aging were excluded. Neurofibrillary tangle counts were determined in all cortical laminae and regions used for staging. There was a significant correlation between the overall extent of neurofibrillary pathology and the number of regions affected. There were frequent order violations in the proposed hierarchy: 19 instances (45%) involving entorhinal and transentorhinal cortices, and 16 instances (38%) involving CA1 of hippocampus and entorhinal cortex. Only 6 out of 42 cases conformed in all regions to the expected hierarchy. Nevertheless, 90% of the cases had 2 order violations or less, supporting the approximate validity of the hierarchy.

a-Synuclein inclusions in Alzheimer's and Lewy body diseases.
Mukaetova-Ladinska E.B, Hurt J., Jakes R., Xuereb J., Honer W.G. and Wischik C.M.
J Neuropathol Exp Neurology, 2000; 59: 408-417

α-Synuclein has assumed particular neuropathological interest in the light both of its identification as a non-β-amyloid plaque constituent in Alzheimer disease (AD), and the recent association between dominant inheritance of Parkinson disease (PD) and 2 missense mutations at positions 30 and 53 of the synuclein protein. We report a systematic study of α-synuclein, tau, and ubiquitin immunoreactivity in representative neurodegenerative disorders of late life. The α-synuclein association with Lewy bodies is variable, peripheral, and is not stable with respect to proteases or acid treatment, whereas there is no association with Pick bodies. Stable patterns of immunoreactivity included neurites and a novel inclusion body. Although there is an overlap between the presence of Lewy bodies and stable α-synuclein immunoreactivity, this is seen only in the presence of concomitant neuropathological features of AD. The novel α-synuclein inclusion body identified in pyramidal cells of the medial temporal lobe in particular was found in AD and in the Lewy body variant of AD, and was associated neither with ubiquitin nor tau protein. The inclusion is therefore neither a Lewy body nor a PHF-core body, but may be confused with the Lewy body, particularly in the Lewy body variant of AD. Abnormal processing of α-synuclein leading to its deposition in the form of proteolytically stable deposits is a particular feature of the intermediate stages of AD.

Neuropathological findings in the very old: Results from the first 101 brains of a population based longitudinal study of dementing disorders.
Xuereb J, Brayne C, Dufouil C, Gertz H, Wischik C, Harrington C, Mukaetova-Ladinska E, McGee MA, O'Sullivan A, O'Connor D, Paykel ES and Huppert FA.
Ann NY Acad Sci (Suppl), 2000; 903: 490-496.

We report a unique longitudinal epidemiological study of cognitive decline in the elderly population of the city of Cambridge, UK. A population sample of people aged 75 and over was surveyed between 1984-1996 (n = 2,616) and followed 2.4, 6, and 9 years later. CAMDEX diagnostic criteria were used for clinical assessment, and the neuropathological protocol (in 101 cases) was based on the CERAD method, with additional features to allow Braak staging of neurofibrillary pathology. The main findings are of the heterogeneity of lesions to be found in very old populations, and the existence of considerable overlap in the pathologies found in the demented and nondemented. It seems that white matter (ischemic) pallor an amyloid angiopathy, as well as neuritic plaques, neurofibrillary tangles and Lewy body formation are all lesions that increase the likelihood of dementia.

Staging of cytoskeletal and b-amyloid changes in human isocortex reveals biphasic synaptic protein response during progression of Alzheimer's Disease.
Mukaetova-Ladinska EB, Garcia-Siera F, Hurt J, Gertz H-J, Xuereb JH, Hills R, Brayne C, Huppert FA, Paykel ES, McGee M, Jakes R, Honer W, Harrington CR and Wischik CM.
Am J Pathol, 2000; 157: 623-636.

We have examined the relationships between dementia, loss of synaptic proteins, changes in the cytoskeleton, and deposition of ß-amyloid plaques in the neocortex in a clinicopathologically staged epidemiological cohort using a combination of biochemical and morphometric techniques. We report that loss of synaptic proteins is a late-stage phenomenon, occurring only at Braak stages 5 and 6, or at moderate to severe clinical grades of dementia. Loss of synaptic proteins was seen only after the emergence of the full spectrum of tau and ß-amyloid pathology in the neocortex at stage 4, but not in the presence of ß-amyloid plaques alone. Contrary to previous studies, we report increases in the levels of synaptophysin, syntaxin, and SNAP-25 at stage 3 and of α-synuclein and MAP2 at stage 4. Minimal and mild clinical grades of dementia were associated with either unchanged or elevated levels of synaptic proteins in the neocortex. Progressive aggregation of paired helical filament (PHF)-tau protein could be detected biochemically from stage 2 onwards, and this was earliest change relative to the normal aging background defined by Braak stage 1 that we were able to detect in the neocortex. These results are consistent with the possibility that failure of axonal transport associated with early aggregation of tau protein elicits a transient adaptive synaptic response to partial de-afferentation that may be mediated by trophic factors. This early abnormality in cytoskeletal function may contribute directly to the earliest clinically detectable stages of dementia.

The response of elderly community residents to request brain donation: An interim report
Lynn Beardsall 1, Claire Barkley 1, Angela O'Sullivan 2
Int J of Geriatric Psychiatry, 2004; 7(3): 199-202.

Normal neurohistology of the elderly brain is largely unstudied and unknown. Normal brains are essential as controls for postmortem investigations in longitudinal, prospective studies which aim to document and clarify the natural history of Alzheimer's disease. Ethically, consent should be sought from these subjects and their relatives at the beginning of such studies. This article reports the responses of normal and demented elderly community residents to request for brain donation. Eighty-five per cent of the normal control group and 83% of incident cases of dementia signed declarations of intent for postmortem. During the first year of brain collection, 24 of 33 donated brains have been successfully collected at postmortem. The responses and attitudes of the subjects and their relatives and the value of obtaining such consent are discussed.

Population-based neuropathological studies of dementia: design, methods and areas of investigation – a systematic review
Zaccai J, Ince P and Brayne C
BMC Neurology 2006, 6:2

Abstract

Background
Prospective population-based neuropathological studies have a special place in dementia research which is under emphasised.

Methods
A systematic review of the methods of population-based neuropathological studies of dementia was carried out. These studies were assessed in relation to their representativeness of underlying populations and the clinical, neuropsychological and neuropathological approaches adopted.

Results
Six studies were found to be true population-based neuropathological studies of dementia in the older people: the Hisayama study (Japan); Vantaa 85+ study (Finland); CC75C study (Cambridge, UK); CFAS (multicentre, UK); Cache County study (Utah, USA); HAAS (Hawaï, USA). These differ in the core characteristics of their populations. The studies used standardised neuropathological methods which facilitate analyses on: clinicopathological associations and confirmation of diagnosis, assessing the validity of hierarchical models of neuropathological lesion burden; investigating the associations between neuropathological burden and risk factors including genetic factors. Examples of findings are given although there is too little overlap in the areas investigated amongst these studies to form the basis of a systematic review of the results.

Conclusion
Clinicopathological studies based on true population samples can provide unique insights in dementia. Individually they are limited in power and scope; together they represent a powerful source to translate findings from laboratory to populations.

Loss of synaptophysin and synaptosomal-associated protein 25-kDa (SNAP-25) in elderly Down syndrome individuals
E. C. Downes, J. Robson, E. Grailly, Z. Abdel-All, J. Xuereb, C. Brayne, A. Holland, W. G. Honer and E. B. Mukaetova-Ladinska
Neuropathology and Applied Neurobiology (2008), 34, 12–22

Aims: This study quantified the density of the synaptic proteins synaptophysin and synaptosomal-associated protein 25-kDa (SNAP-25) in brains from elderly Down’s syndrome individuals. Methods: Six areas (frontal, occipital, parietal and temporal lobes, cerebellum and hippocampus) of post mortem brains from elderly Down’s syndrome (DS) individuals (with reported functional memory problems and pathologically established Alzheimer’s disease) and elderly controls were studied.
Results: Collectively in the six brain areas studied, there were significantly lower levels of both synaptophysin and SNAP-25 immunostaining in the DS group compared with controls. The elderly control group displayed a significant decrease in the densities of synaptophysin and SNAP-25 as a function of age at death (AAD; P 0.001), whereas the DS group only showed a significant decrease as a function of AAD for synaptophysin. Assessing synaptic density as a function of Braak stage (BST) revealed a significant decrease in synaptophysin density for both groups. SNAP-25 was only significantly decreased as a function of BST in the DS group. Synaptic protein density was also shown to vary according to gender. Thus, both DS and control female subjects had a higher synaptic density of SNAP-25 than their male counterparts. In contrast, male DS and control individuals had a significantly greater density of synaptophysin than females.
Conclusions: These results indicate that elderly DS individuals have lower synaptic densities of both analysed proteins than cognitively intact elderly individuals. Although AAD and BST show varying significance to decreases in protein density for both DS and
control groups, results suggest that gender differences also play a role in the type of synaptic protein lost in elderly DS individuals.

Genetics

The tRNA Gln 4336 mitochondrial DNA variant is not a high penetrance mutation which predisposes to dementia before age 75 years.
Tysoe C, Robinson D, Brayne C, Dening T, Paykel ES, Huppert FA and Rubinsztein DC.
Journal of Medical Genetics, 1996; 33: 1002-1006.

The genetic factors that predispose to Alzheimer's disease (AD) are heterogeneous. Two recent reports have suggested that a mitochondrial DNA mutation within the tRNAGln gene, located at position 4336, may be a risk factor for AD, as it was found in 10/256 (3.9%) cases with AD confirmed by necropsy. Although low prevalences of this mutation were detected in non-demented subjects in both of these studies, the controls were not carefully matched with the AD cases. We have investigated the frequency of this mutation in two community based elderly cohorts in Cambridgeshire, who have participated in longitudinal studies of cognitive function. The 4336 mitochondrial mutation was detected in 8/ 443 people examined. These people were found to be non-demented at ages 74, 81, 84, 86, 89, 90, 91, and 102 years, in contrast to the previously described cases whose onset of dementia occurred between 60 and 76 years (mean 68). Accordingly, we believe that this mitochondrial variant is not a high penetrance mutation which predisposes to dementia before the age of 76 years.

Apolipoprotein E genotype and the prediction of cognitive decline and dementia: a prospective study of the very elderly.
Brayne C, Harrington C, Wischik C, Huppert FA, Chi LY, Xuereb J, O'Connor D and Paykel ES.
Dementia, 1996; 7: 169-174.

An increased apolipoprotein E (ApoE) type ɛ4 allele frequency is associated with both sporadic and familial late-onset Alzheimer's disease (AD). The age of onset of disease in patients homozygous for the ɛ4 allele appears to be decreased by approximately 15 years compared with E2/3 individuals. In order to assess the influences of this allele on both dementia and cognitive decline in the elderly wer have determined the ApoE genotype of 150 individuals over the age of 75 years who have taken part in a logitudinal study. Homozygosity for the ɛ4 allele was rare. Of the 2 homozygotes, 1 was severely demented but the other did not receive a clinical diagnosis of dementia. The latter individual did demonstrate marked cognitive decline over a 28-month period. There was a consistent association between the presence of an ɛ4 allele and both the clinical diagnosis of dementia and cognitive decline. These findings confirm a genetic heterogeneity in late-onset sporadic AD and prompt caution in the use of ApoE genotype to predict an elderly individual's susceptibility to either dementia or cognitive decline.

Analysis of the Apo E/Apo CI, angiotensin converting enzyme and methylenetetrahydrofolate reductase loci as candidates affecting human longevity.
Galinsky D, Tysoe C, Brayne C, Easton D, Huppert F, Dening T, Paykel ES and Rubinsztein DC.
Atherosclerosis, 1997; 129: 177-183.

Genetic factors are likely to affect human survival, since twin studies have shown greater concordance for age of death in monozygotic compared to dizygotic twins. Coronary artery disease is an important contributor to premature mortality in the UK. Accordingly, we have chosen genes associated with cardiovascular risk, apo E/apo C-I, angiotensin converting enzyme (ACE) and methylenetetrahydrofolate reductase (MTHFR), as candidates which may affect longevity/survival into old age. An association study was performed by comparing allele and genotype frequencies at polymorphic loci associated with these genes in 182 women and 100 men aged 84 years and older with 100 boys and 100 girls younger than 17 years. MTHFR allele and genotype frequencies were similar in the elderly and young populations. Apo C-I allele and genotype frequencies were significantly different in the elderly women compared to the younger sample (P < 0.05). No difference was observed in the elderly men. At the neighbouring apo E gene, we only observed a difference between genotypes in the elderly women and the young sample; however, this did not retain significance when the genotype frequencies of the young sample were adjusted to values expected from the allele frequencies on the basis of Hardy-Weinberg equilibrium and compared to observed genotypes in elderly men and women. In contrast to previous studies, apo E2 was not overrepresented in the elderly men or women. Thus, the proposition that apo E2, E3 and E4 protein isoforms are themselves functionally associated with increasing risks for early death, may be too simplistic. The I/I ACE was depleted in the elderly males but not the elderly females. Furthermore, significant differences were observed between ACE genotypes in elderly men and elderly women. These data suggest that the penetrance of loci which influence survival may vary according to sex. The depletion of the ACE I/I genotype in elderly men is generally consistent with a previous study which found decreased frequencies of the I allele in French centenarians compared to younger controls. However, these results are apparently paradoxical, since others have suggested that the I allele is associated with increased cardiovascular risk. Clarification of the overall effect of a genotype on survival will be vital if therapies are to be considered which target specific genetic variants.

Analysis of Alpha-1 Antichymotrypsin, Presenilin-1, Angiotensin Converting-Enzyme and Methylenetetrahydrofolate Reductase Loci as candidates for dementia.
Tysoe C, Galinsky D, Robinson D, Brayne CE, Easton DF, Huppert FA, Dening T, Paykel ES and Rubinsztein DC.
American Journal of Medical Genetics (Neuropsychiatric Genetics), 1997; 74, 207-212.

The genetic factors which predispose individuals to dementia in old age have not been fully defined. Although the apolipoprotein E4 allele accounts for a proportion of the genetic risk for late-onset Alzheimer disease (AD), it is neither necessary nor sufficient to cause this disease. Recent suggestions that other loci are involved in dementia risk have been supported by findings of associations of genotypes at the alpha-1 antichymotrypsin (ACT) and presenilin-1 (PS-1) loci with AD. We investigated these loci in two community-based aged Cambridgeshire populations: the rural Ely population (cohort 1) comprised 60 pairs of demented and nondemented elderly individuals, with a mean age of 84.2 years; and the Cambridge city population (cohort 2) comprised 81 pairs all over age 84, with a mean age of 87.3 years. Since vascular risk factors are likely to impact on dementia risk, we also examined the angiotensin-converting enzyme (ACE) and methylenetetrahydrofolate reductase (MTHFR) genes as candidates. ACE, ACT, PS-1, and MTHFR genotype and allele frequencies were not significantly different in cases and matched controls. These data support the doubts which have been raised about the involvement of the PS-1 and ACT polymorphisms in late-onset dementia.

ApoE and apo CI loci are associated with dementia in younger but not in older late onset cases.
Tysoe C, Galinsky D, Robinson D, Brayne CE, Huppert FA, Dening T, Paykel ES, Easton DF and Rubinsztein DC.
Dementia and Geriatric Cognitive Disorders, 1998; 9: 191-198.

Numerous groups have confirmed that apolipoprotein E allelic variation accounts for a proportion of the genetic risk for late-onset Alzheimer's disease (AD). However, there is a paucity of data on the impact of this locus on the overall risk of dementia (as opposed to AD) in the elderly. Most studies have ascertained specifically AD cases from hospital clinics or brain banks and many demented cases have vascular dementia or mixed AD and vascular pathology. We have examined the closely linked apo E and apo CI loci in demented cases and non-demented controls from two community-based aged Cambridgeshire populations: the rural Ely population (cohort 1) comprised 60 pairs of demented and non-demented elderly individuals, with a mean age of 84.2 years (SD = 6.11); the Cambridge city population (cohort 2) comprised 81 pairs all aged over 84 with a mean age of 87.7 years (SD = 2.9). The younger Ely cohort showed significant allelic associations with dementia at the apo E and apo CI loci, which were not replicated in the older Cambridge cohort. These data suggest the possibility of age-dependent penetrance for different candidate genes in late-onset dementia. We propose a number of explanations to account for the stronger associations we observed between dementia and apo CI, compared to the neighbouring apo E locus. Our data are compatible with the possibility that specific alleles or genotypes may confer different risks for overall dementia, compared to AD.

Alzheimer Disease is not associated with polymorphisms in the angiotensinogen and renin genes.
Taylor A, Ezquerra M, Bagri G, Yip A, Goumidi L, Cottel D, Easton D, Grimley Evans J, Xuereb J, Cairns NJ, Amouyel P, Chartier-Harlin M-C, Brayne C and Rubinsztein DC.
American Journal of Medical Genetics 2001; 105: 761-4.

Hypertension has been implicated as a risk factor for Alzheimer disease (AD) and dementia in epidemiological studies of humans. It is thus possible that there are common genetic determinants for hypertension and AD. Epidemiological, clinical, and experimental data suggest that the renin-angiotensin-aldosterone system is a critical regulator of blood pressure. The presence of an MboI site in an RFLP in the renin gene and the Thr at the Met/Thr polymorphism at codon 235 (M235T) of the angiotensinogen gene have been reported to be associated with hypertension. These variants were studied in autopsy-confirmed AD cases and matched controls from the U.K. While no association was detected with the renin polymorphism, a weak deleterious effect was observed in cases homozygous for the angiotensinogen Thr allele. However, this association was not observed in a French cohort of clinically diagnosed AD cases and controls, suggesting that the initial observation was a type I error. Thus, these polymorphisms are unlikely to be associated with AD risk. Copyright 2001 Wiley-Liss, Inc.

No evidence for an association between Saitohin Q7R polymorphism and Alzheimer's disease.
Cook L, Brayne CE, Easton D, Grimley Evans J, Xuereb J, Cairns NJ and Rubinsztein DC.
Annals of Neurology 2002; 52(5): 690-691.

Introduction: Conrad and colleagues recently described a novel gene called Saitohin located between exons 9 and 10 of the human tau gene.[1] Saitohin encodes a 128-amino acid protein with a similar tissue expression pattern to tau. A single nucleotide polymorphism was discovered in the gene that results in an amino acid change (Q7R). In a study of white Alzheimer's disease (AD) cases (n = 51) and controls (n = 30), the RR genotype and the R allele were found to be significantly overrepresented in the AD cases (odds ratio [OR], 11.92 for genotype; OR, 3.11 for allele).[1] These results, if correct, suggest that the Saitohin R allele may confer risks of a similar magnitude to apolipoprotein E (ApoE) 4.[2] The effect of Saitohin genotypes on AD risk was independent of ApoE genotype status.

Candidate Gene Association Studies of Genes Involved in Neuronal Cholinergic Transmission in Alzheimer's Disease suggests Choline Acetyltransferase as a candidate deserving further study.
Cook LJ, Ho LW, Wang L, Terrenoire E, Brayne C, Grimley Evans J, Xuereb J, Cairns NJ, Turic D, Hollingworth P, Moore PJ, Jehu L, Archer N, Walter S, Foy C, Edmondson A, Powell J, Lovestone S, Williams J and Rubinsztein DC.
Am J Med Genet (Neuropsychiatric Genetics) (In press)

Consistent deficits in the cholinergic system are evident in the brains of Alzheimer's Disease (AD) patients, including reductions in the activities of acetylcholine, acetylcholinesterase (AChE), and choline acetyltransferase (ChAT), increased butyrylcholinesterase (BChE) activity, and a selective loss of nicotinic acetylcholine receptors (nAChRs). Accordingly, we have analyzed polymorphisms in the genes encoding AChE, ChAT, BChE, and several of the subunit genes from neuronal nAChRs, for genetic associations with late-onset AD. A significant association for disease was detected for a non-coding polymorphism in ChAT (allele chi(1) (2) = 12.84, P = 0.0003; genotype chi(2) (2) = 11.89, P = 0.0026). Although replication analysis did not confirm the significance of this finding when the replication samples were considered alone (allele chi(1) (2) = 1.02, P = 0.32; genotype chi(2) (2) = 1.101, P = 0.58) the trends were in the correct direction and a significant association remained when the two sample sets were pooled (allele chi(1) (2) = 12.37, P = 0.0004; genotype chi(2) (2) = 11.61, P = 0.003). Previous studies have reported significant disease associations for both the K-variant of BChE and the coding ChAT rs3810950 polymorphism with AD. Replication analyses of these two loci failed to detect any significant association for disease in our case-control samples. Copyright 2004 Wiley-Liss, Inc.

Longitudinal perspective

Cognitive Impairment in elderly people: population based estimate of the future in England, Scotland and Wales.
Melzer D, Ely M and Brayne C.
British Medical Journal, 1997; 315(23): 462.

Introduction: Financing long term institutional care for elderly people is a contentious political issue. Cognitive impairment is the main reason for such care, and to inform the policy debate we constructed a population based set of estimates of the future number of elderly people and the characteristics of cognitive impairment among them on the basis of a national survey of disability in adults by the Office of Population Censuses and Surveys. We report future trends based on stable age specific rates of cognitive impairment and disability.

Changes in health, disability and contact with services in a very elderly cohort: A six year follow-up study.
Dening TR, Chi L-Y, Brayne C, Huppert FA, Paykel ES and O'Connor DW.
Age and Ageing, 1998; 27: 23-33.

OBJECTIVES: To study the relationships between global self-rated health, reported physical symptoms and depressive symptoms and the receipt of community services by very elderly people, and to examine changes in these variables over time. DESIGN: Three-wave study with follow-up at 2.4 and 6 years after first interview. Structured interview, incorporating cognitive examination (Mini-Mental State Examination) and enquiring specifically about overall self-rated health, physical symptoms and depressive symptoms. SETTING: Community setting in city of Cambridge, UK. PARTICIPANTS: 2609 were initially recruited: all patients aged 75 years and over from lists of six general practices (and one in three from a seventh practice). At 2.4 years, 1173 individuals re-examined and at 6 years 628 individuals. MEASUREMENTS: General health self-rated in comparison to others of similar age and individual physical and depressive symptoms self-rated as present or absent. Symptoms were added to produce physical health and depressive symptom scores. Data presented from cross-sectional analysis of 6-year sample; also examined longitudinal data from all three waves of study for ageing and cohort effects. Finally the effect of health variables on the receipt of services was examined. Statistics used included chi(2) and non-parametric statistics for continuous data, also odds ratios for likelihood of receiving services. RESULTS: At 6 years, 70% rated their overall health as good or very good. Overall self-rated health showed both ageing and cohort effects, improving with increasing age and especially with more recent cohort. Reported physical symptoms increased with ageing. Depression scores also increased with ageing but the relationship between depressive symptoms and ageing was less clear-cut. Receipt of services was associated with poor self-rated health and reported physical symptoms as well as with ageing. Higher depression scores at 2.4 years were associated with increased service receipt at 6 years, indicating a lag between the symptoms and the service response. Individuals in the more recent cohort were less likely to receive services, but those who did so received more frequent contact. CONCLUSIONS: Although very elderly people have a high prevalence of reported physical symptoms, they often rate their overall health as good. There was a stronger relationship between ageing and physical symptoms than with depressive symptoms. Symptoms of both kinds influenced the likelihood of receiving services, although there was a lag between depressive symptoms and service response. Cohort effects on service receipt may reflect changes in public service policy.

Cognitive decline in the old old is underestimated by current epidemiological studies: MMSE change in the Cambridge City over 75 cohort over nine years.
Brayne C, Spiegelhalter DJ, Dufouil C, Huppert FA, Chi L-Y, Dening TR, Paykel ES, O'Connor DW and Ahmed A.
Journal of the American Geriatrics Society, 1999; 47(11): 1283-8.

Summary

Objective: To measure cognitive change using a brief measure over nine years and adjust for attrition in the sample.
Sample: The Cambridge City Cohort (CC75C), a complete sample of the 75 years and over age group from five group general practices in Cambridge city, with a systematic third of a further practice, all followed on four occasions.
Setting: Cambridge city, UK, respondents' places of residence.
Procefures: Brief interview administered by trained interviewer containing a shor cognitive scale, the Mini-mental state examination, at baseline, 2.4 years, 6 years and 9 years.
Results: Decline in MMSE occurred across the population, greater in the oldest age groups. Attrition at later stages of the follow - up was associated with greater decline at earlier stages. Adjusting the results for loss to the sample leads to higher estimates of decline, with the older age groups declining faster from lower levels.
Conclusions: Cognitive decline in the oldest age groups is considerablyy underestimated by traditional follow-up studies. If populatioin samples are to reflect the health and social needs of this frail group accurately, they must adjust findings for effect of attrition.

Longitudinal studies of ageing: a key role in the evidence base for improving health and quality of life in older adults.
Huppert Felicia A., Brayne Carol, Jagger Carol and Metz David.
Age and Ageing, 2000; 29: 485-486.

Knowledge about the factors that influence the quality of our lives in old age is of major importance as population ageing becomes a reality world-wide. Implementation of policies to promote healthy ageing will need to be based on the best available evidence. A variety of seminal studies are addressing possible influences, including genetics, early-life factors and environmental risks and lifestyle in mid and late life. Common to many of these studies is their longitudinal design. Longitudinal studies are key to understanding patterns of ageing and the services which an ageing population is likely to require. Only a longitudinal design can yield information about the dynamics of change as people age, and can thereby suggest how outcomes might be linked to putative influences.

Population norms for the MSE in the very old - Estimates based on longitudinal data.
Dufouil C, Clayton D, Brayne C, Chi L-Y, Dening TR, Paykel ES, O'Connor DW, Ahmed A, McGee MA and Huppert FA.
Neurology, 2000; 55: 1609-1613.

OBJECTIVE: To report the percentile distribution of Mini-Mental State Examination (MMSE) scores in older people by age, sex, and education level, estimated from longitudinal data, after correcting for loss due to dropout.

A systematic literature review of attrition between waves in longitudinal studies in the elderly shows a consistent pattern of dropout between differing studies.
Chatfield M, Brayne C and Matthews F
Journal of Clinical Epidemiology 2005, 58(1) 13-19

Abstract

Objectives
Longitudinal studies of the elderly are complicated by the loss of individuals between waves due to death or other dropout mechanisms. Factors that affect dropout may well be similar from one study to another. This article systematically reviews all large population-based studies of the elderly (published 1966–2002) that report on differences in individual characteristics between people who remain and people who dropout at follow-up.

Study design and setting
A systematic review of articles that investigate attrition after baseline interview.

Results
Twelve studies were found that investigated dropout other than death using unadjusted, multivariable methods or both. The unadjusted analyses showed many significant factors related to attrition. Multivariable analyses showed two main independent factors were related to increased attrition: increasing age and cognitive impairment. People who were very ill or frail had higher dropout rates, and people in worse health were less likely to be recontactable.

Conclusions
Multivariable methods of analyzing attrition in longitudinal studies show consistent patterns of dropout between differing studies, with a small number of key relationships. These findings will assist researchers when planning studies of older people, and provide insight into the possible biases in longitudinal studies introduced by differential dropout.

Clinical practice and services for the oldest old

Practice observed: Do general practitioners miss dementia in elderly patients?
O'Connor D, Pollitt PA, Brook CP, Reiss BB and Roth M.
British Medical Journal, 1988; 297: 1107-1110.

General practitioners and community nurses were asked to rate the likelihood of dementia for each of their elderly patients. Cases of dementia were identified by research psychiatrists using the Cambridge mental disorders of the elderly examination (CAMDEX), a new structured diagnostic interview. General practitioners correctly identified dementia as at least a possibility in 121 of the 208 cases found. Nevertheless, they mistakenly rated as demented several patients suffering from functional psychiatric disorders, in particular depression. Community nurses correctly identified dementia as at least a possibility in 64 of the 74 demented patients known to them, but they incorrectly suspected dementia in a greater proportion of instances. Both general practitioners and families appeared to have low expectations of what general practice has to offer demented elderly people. General practitioners should take the initiative in diagnosing dementia in very elderly patients who show signs of the condition. In some cases it may be secondary to treatable disorders, and in others all that may be required are understanding, support, and advice to families.

A community survey of mental and physical infirmity in nonagenarians.
O'Connor DW, Pollitt PA, Brook CPB and Reiss BB.
Age and Ageing, 1989; 18: 411-414.

One hundred and thirty-two people aged 90 years and over were identified in a study of the prevalence of dementia. One third lived in institutions, a third were demented and half were unable to prepare a simple meal or do light housework. None was entirely free of mental or physical disability but many functioned satisfactorily at a simple level with the help of family members and the domiciliary services.

The distribution of services to demented elderly people living in the community.
O'Connor DW, Pollitt PA, Brook CPB and Reiss BB.
International Journal of Geriatric Psychiatry, 1989; 4: 339-344.

One hundred and sixty demented, community-resident elderly people were identified by means of a community survey. Subjects who lived alone received home helps and meals-on-wheels more often than those who lived with supporters but there were no differences between the two groups with respect to district nursing or day care. Services increased in line with the degree of mental infirmity and nearly all moderately and severely demented people who lived alone were known to at least one service. We suggest that the home help and meals-on-wheels services, as presently constituted, may have limited applicability to dependent old people who live with able-bodied supporters. If services are to provide effective assistance, they may need to adapt their provision to the actual needs of elderly people and their relatives.

Does early intervention reduce the number of elderly people with dementia admitted to institutions for long term care?
O'Connor DW, Pollitt PA, Brook CPB, Reiss BB and Roth M.
British Medical Journal, 1991; 302: 871-875.

OBJECTIVE--To test whether early diagnosis and practical help reduce the number of elderly people with dementia admitted to institutions. DESIGN--Controlled trial of effect of help from a multidisciplinary team on admission rates of people with dementia. SETTING--Seven general practices in Cambridge. SUBJECTS--2889 subjects aged 75 and over, of whom 159 were identified as having dementia with a two stage community survey. Eighty six subjects were referred for extra help if they or their supporters wished. The other 73 subjects had access to the usual services and served as controls. INTERVENTION--Subjects and families in the action group were offered a wide range of help, including financial benefits, physical aids, home helps, respite admissions, practical advice, and psychiatric assessments. MAIN OUTCOME MEASURE--Permanent admission to long term care within two years after diagnosis. RESULTS--Early intervention did not affect admission rates in subjects who lived with supporters. By contrast, nine of the 14 (64%) subjects with moderate or severe dementia living alone were admitted in the action group in the study's second year compared with only one of 13 (8%) controls (p = 0.004). CONCLUSIONS--Some people with moderate or severe dementia who lived alone and were at serious risk may have been identified earlier by the resource team. Without the team these people would not have become known to the responsible authorities until families, neighbours, and wardens became unable to cope. The study was conducted during the team's formative period, however, and greater experience might have allowed some subjects to remain at home for longer.

Depressive symptoms in the very elderly - their prevalence and significance.
Girling DM, Huppert FA, Brayne C, Paykel ES, Gill C and Mathewson D.
International Journal of Geriatric Psychiatry, 1995; 10: 497-504.

A community sample of 1173 very elderly people (aged over 77) was interviewed by trained lay interviewers using a structured interview, including questions relating to emotional and physical health and social circumstances and the Mini-Mental State Examination. Depressive symptoms such as loss of energy and feelings of tension and irritability were found frequently. High scores for depressive symptoms were associated with female sex, poor subjective physical health, loneliness and use of statutory services. Depressive symptom score was not found to be associated with age. Although only 6% of the sample described feeling depressed most of the time, 21% of respondents admitted that they sometimes felt that life was not worth living.

The prevalence of depression in a cohort of the very elderly.
Girling DM, Barkley C, Paykel ES, Gehlhaar E, Brayne C, Gill C, Mathewson D and Huppert FA.
Journal of Affective Disorders, 1995; 34: 319-329.

In a community study of 1173 very elderly (> or = 77 years) subjects, a screening interview was followed by a CAMDEX diagnostic interview in a subsample of 461. The estimated prevalence of DSM-III-R major depressive disorder in the community sampled was 2.4% (95% CI 0.9%, 4.0%). Using CAMDEX criteria, the prevalence of depressive illness was 3.0% (95% CI 0.7%, 5.3%). 10% of those who had a diagnostic interview were rated as having depressive symptoms of mild or moderate severity. Of these, approximately 1/3 met diagnostic criteria for major depressive disorder. The significance of these findings and the possible need for wider criteria for depression in the elderly are discussed.

Predictors of hospital contact by very elderly people: a pilot study from a cohort of people aged 75 years and over.
Chi L-Y, Brayne C, Todd CJ, O'Connor DW, Pollitt PA.
Age and Ageing, 1995 Sep; 24(5): 382-8

In 1991 the United Kingdom population was 57.8 million of whom 9.1 million (15.7%) were aged 65 years or older [1]. Forecasts indicate that the population aged over 65 will increase to 14.0 millions in 2031 (22.5% of the population). The largest proportional increase will be among those surviving into very old age. This group is believed to make the highest demand on health services [2].

Several demographic factors have been identified as associated with admission to long-term care. These are being very old, being female, living alone, and not having a spouse. These factors all tend to be associated with limited informal care networks which can be a major determinant of failure to maintain independence [3]. In 1988, 20% of men and 16% of women aged 75 and over reported an inpatient hospital stay in the previous year [4]. Outpatient hospital service use increased in Great Britain between 1972 and the mid-1980s, but since then there has been little change [5]. There are a number of explanations for this [6], and in order to assess the impact of the expected demographic changes we must fully understand the determinants of service use by elderly people. Self-rated health has repeatedly been shown to be strongly and independently related to survival [7, 8]. As such it could be strongly predictive of hospital contact.

Dementia incidence and prevalence increases with age [9, 10]. Recent estimates suggest that about half a million people suffer from dementia in the UK, and it is estimated that this will increase by 20% during the early years of the next century [11]. Caring for mentally disordered elderly people is one of the most difficult situations facing families and professionals [12, 13].

Estimating the numbers and characteristics of elderly people with cognitive disability in local populations.
Ely M, Melzer D, Opit L and Brayne C.
Research Policy and Planning, 1996/1997; 14: 13-18.

This paper describes a model which produces quantitative estimates of the numbers and characteristics of elderly people with cognitive disability at a level comparable to moderate or severe dementia or confusion. It produces robust estimates for local populations based on the numbers of people in the age groups: 65-74, 75-84 and 85 +. It is available as a simple software package designed for use by those involved in planning services for the elderly. It can provide a basis for evidence-based health care or social care decisiions, or for informed policy debate.

Local Population Differences and the Needs of People with Cognitive Impairment.
Melzer D, Ely M and Brayne C.
International Journal of Geriatric Psychiatry, 1997; 12: 883-887.

Introduction. Variations in local population age structure have attracted less attention than national population ageing. As moderate and severe cognitive impairment is a major cause of need for long-term care, population-based estimates of the numbers and characteristics of this group were calculated, to explore the effects of local differences.

Method. The UK Office of Population Census and Surveys (OPCS) study of disability in adults (N > 14000) was reanalysed. A group with moderate or severe cognitive impairments was identified and age-specific estimators of sociodemographic characteristics, household types, disabilities and service use were combined with population estimates for district health authorities in England and Wales.

Results. The proportion of the 65 plus population who are 85 plus varies from 8% to 15% across districts, equivalent to national population projections for 1986 and 2031 respectively. The estimated prevalence of the study group varies from 53 to 70 per 1000 population aged 65 plus, with 34-48% of cases aged 85 plus. Curiously, the proportion with severe disabilities varies little across districts. If national norms applied, local rates of institutionalization would vary from 18 to 27 per 1000 aged 65 plus.

Conclusion. Local differences in population age structure are large compared to national changes over decades. Local differences have substantial effects on overall prevalence and on the proportion of the cognitively impaired who would be institutionalized if national patterns applied. Service design should be influenced by these complex variations, with estimates modified by local surveys.

Cognitive impairment: a challenge for community care. A comparison of the domiciliary service receipt of cognitively impaired and equally dependent physically impaired elderly women.
Ely M, Brayne C, Huppert FA, O'Connor DW and Pollitt PA.
Age and Ageing, 1997; 26: 301-308.

Objectives: to compare the domiciliary service receipt of cognitively impaired and equally dependent physically impaired elderly women prior to the passing of the UK Community Care Act. Methods: secondary analysis of a population survey conducted in 1986 in the city of Cambridge. The analysis used data on 1585 women aged 75 and over living in the community. The effect of type of impairment on the receipt of domiciliary services (meals-on-wheels, home help and community nursing) is measured using a multivariate model which allows for adjustment for dependency level and other potential confounding factors. Results: the odds of an elderly woman getting help from any of the domiciliary services whilst not being significantly affected by cognitive impairment (odds ratio 0.7, 95% CI 0.5-1.2) are increased by physical impairment (odds ratio 1.8, 95% CI 1.2-2.5). Similar results were found for the home help service. The differences were exaggerated in the case of the community nursing service, whilst receipt of meals-on-wheels was similar for women of with all types of impairment. Conclusions: in the late 1980s, cognitively impaired elderly women received less help from the domiciliary services than equally dependent physically frail women who lived in similar household circumstances. The development of specialist services appropriate to the needs of cognitively impaired elderly people presents a challenge to community care policy, especially since this group are at high risk of institutionalization.

Suicidal thinking in community residents over eighty.
Rao R, Dening T, Brayne C and Huppert FA.
International Journal of Geriatric Psychiatry, 1997; 12: 337-343.

OBJECTIVE: Main objective: to study the relationship between suicidal thinking and both cognitive impairment and depression. DESIGN: Random sample selected for interview, all of whom were a cohort in a pre-existing epidemiological study of dementia. SETTING: Community residents. PATIENTS AND OTHER PARTICIPANTS: Participants aged over 81. Study excluded the following: moved out of area/died/ too frail/severe communication difficulties/refused interview, refusal by GP/family/carers, 300 names selected at random from database. 170 eligible participants approached: 31 refused, 125 interviewed, 125 informants approached for interview; 118 interviewed. MAIN OUTCOME MEASURES: CAMDEX, 15-item Geriatric Depression Scale (GDS), and Scale for Suicidal Ideation (SSI) (including informant versions of latter 2 scales). RESULTS: 9 people showed suicidal thinking, all women: 6 had clinical evidence of cardiovascular/cerebrovascular disease. Those with suicidal thinking showed higher CAMDEX depression scores, weaker strength of the wish to go on living, higher rates of expressing wish to die and higher rates of depressive illness and mixed DAT/multi-infarct dementia as primary psychiatric diagnoses. No significant associations between suicidal thinking and GDS scores, Alzheimer-type dementia alone, awareness of memory difficulties or severity of dementia. CONCLUSIONS: Results show association between suicidal thinking and both depression and mixed DAT/multi-infarct dementia, but do not support an association between suicidal thinking and awareness of memory problems/severity of dementia. Given the methodological limitations, the significance of the results should be viewed with caution. Further exploration of the role of cerebrovascular disease in depressive disorder is suggested.

Attitudes towards death - a community study of octo- and nonagenerians.
Rao R, Dening T, Brayne C and Huppert F.
International Psychogeriatrics, 1997; 9(2): 213-221.

Attitudes to death were explored in 125 community residents aged 81 and older. Those people who thought about dying had less frequent contact with their interview informant; those who thought about death more frequently showed less severe cognitive impairment, a greater severity of depressive symptoms, and were more likely to be unmarried and were more commonly reported to be depressed by their interview informant. Worries about dying show no association with sociodemographic or clinical variables. The commonest category of response from spontaneous comments was concerning the circumstances of dying. The results are discussed in light of other research findings, and emphasize the influence of low mood and social factors on death attitudes. This may have implications for closer examination of such attitudes in the assessment of depression and suicidal risk as well as in the care of the dying.

Aging and lymphocyte subpopulations: whole-blood analysis of immune markers in a large population sample of healthy elderly individuals.
Huppert FA, Solomou W, O'Connor S, Morgan K, Sussams P and Brayne C.
Exp.Gerontol, 1998; 33 (6): 593-600

Immune status was determined in a representative sample of elderly people by measuring lymphocyte subsets in whole-blood samples as part of an epidemiological study of the population aged 65 and over. Venepuncture was undertaken in more than 500 individuals who took part in an extensive interview that focused on the lifestyle and psychosocial determinants of healthy aging. The results show that median levels of all lymphocyte subsets tend to decline as the age of the sample increases. In the total sample there were significant age effects (p < 0.05) on total lymphocytes, CD3, CD4, and CD19 (B cells); age differences did not reach significance for CD8 and CD57. There were also significant sex differences (p < 0.05) on CD3, CD4, and CD19, and in all cases women had higher values than men. When we selected a particularly healthy subsample who did not report any illness and took no medication, the findings were unchanged. We conclude that the peripheral expression of lymphocytes appears little affected by aging-related illnesses in the general population, but is affected by aging itself. The study provides reference values for the lymphocyte measures, which can be regarded as having greater validity than the values usually cited.

Very old drivers: findings from a population cohort of people aged 84 and over.
Brayne C, Ahmed A, Dening TR, Chi L-Y, McGee M and Huppert FA.
International Journal of Epidemiology, 2000; 29: 704-707

BACKGROUND: Increases in longevity will involve a significant increase among the number of drivers in the very old, who are at greater risk of being involved in road accidents. Data are thus needed from studies of older populations to characterize those still driving, the reasons for giving up and to help formulate appropriate policies for dealing with the problems faced and created by an increase in older drivers. METHODS: A driving questionnaire was administered to surviving members of a cohort comprising a representative sample of individuals aged >/=84, the Cambridge City over 75 Cohort. Out of 546 survivors 404 completed the driving questionnaire at the 9-year follow-up. In addition, subjects were assessed, at baseline and at each follow-up, for cognitive performance using the Mini-Mental State Examination (MMSE) and for physical impairment using the Instrumental of Activities in Daily Living (IADL) scale. RESULTS: Of the sample, 37% had driven in the past, and 8.4% were still driving, the majority regularly. The drivers tended to be younger (mean age 86.6 years), men (71%) and to be married (67.7%). Although physical disability and cognitive impairment are common in this age group, current drivers had few physical limitations on their daily activities and were not impaired on MMSE. None of the current drivers had visual impairment and 22.6% had hearing loss. Of those who had given up driving, 48.5% had given up at the age of >/=80. The commonest reasons for giving up driving were health problems (28.6%), and loss of confidence (17.9%). One-third reported giving up driving on advice. CONCLUSION: A process of self-selection takes place among older drivers. People over the age of 84 who are still driving have generally high levels of physical fitness and mental functioning, although some have some sensory loss. Given the likely increase in the number of older drivers over the next decades, safety will be improved most by strategies aimed at the entire driving population with older drivers in mind, rather than relying on costly screening programmes to identify the relatively small numbers of impaired older people who continue to drive.