Skip Navigation


Age and Ageing Advance Access originally published online on January 13, 2006
Age and Ageing 2006 35(2):154-160; doi:10.1093/ageing/afj030
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow supplementary data
Right arrow All Versions of this Article:
35/2/154    most recent
afj030v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Matthews, F. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Matthews, F. E.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the British Geriatrics Society. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Risk factors for incident dementia in England and Wales: The Medical Research Council Cognitive Function and Ageing Study. A population-based nested case–control study

MRC Cognitive Function and Ageing Study, Agustin G. Yip1, Carol Brayne1, Fiona E. Matthews2

1 University of Cambridge, Department of Public Health and Primary Care, Cambridge, UK
2 MRC Biostatistics Unit, Institute of Public Health, Cambridge, UK

Address correspondence to: F. E. Matthews. Email: fiona.matthews{at}mrc-bsu.cam.ac.uk


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 Key points
 Role of funding source
 Conflict of interest
 References
 
Objective: to investigate a number of prospectively collected factors (sociodemographic, medical and behavioural) and their association with incident dementia in a population-based cohort.

Design: nested case–control analysis (at 2 and 6 years) of a population-based cohort study.

Setting: individuals aged 65 years and above from five centres in England and Wales: two rural (Cambridgeshire and Gwynedd) and three urban (Nottingham, Newcastle and Oxford).

Participants: a total of 4,075 individuals from a detailed assessment group, with risk measured at baseline.

Main outcome measure: incident dementia at 2 and 6 years.

Methods: logistic regression was used to calculate crude odds ratios (ORs) for various risk factors and ORs adjusted for age, sex, education and social class.

Results: age (90+ versus 65–69 years OR = 25.6, 95% confidence interval (CI) = 11.6–56.9) and sex (women versus men OR = 1.6, 95% CI = 1.1–2.4) were directly associated with dementia, with a trend by years of education (Ptrend = 0.02) but not social class. Poor self-perceived health (versus good) increased the risk for incident dementia (OR = 3.9, 95% CI = 2.2–6.9). Alcohol and smoking (never, past and current) were neither strongly protective nor predictive. Stroke was strongly related to incident dementia (OR = 2.1, 95% CI = 1.1–4.2), as was Parkinson’s disease (OR = 3.5, 95% CI = 1.3–9.3), and exposure to general anaesthesia (GA) was inversely associated with dementia development (OR = 0.6, 95% CI = 0.4–0.9, with a trend with increasing GA exposure; P = 0.003).

Conclusion: in this large multicentre and long-term population-based study, some well-known risk factors for dementia, of vascular and Alzheimer’s type, are confirmed but not others. The association between self-perceived health—a robust predictor of later health outcomes—and incident dementia, independently of other potential risks, warrants further study.

Keywords: incident dementia, dementia, risk factors, population-based study, CFAS, elderly


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 Key points
 Role of funding source
 Conflict of interest
 References
 
Dementia is the most common disorder affecting the brain in older people, estimated to affect 550,000 individuals over 64 years in England and Wales, with up to 165,000 new cases every year [1, 2]. Increasing numbers are expected, and the costs of dementia care can only escalate. Risk estimates from previous epidemiological studies have not provided consistent results, although a consensus is emerging that education and dietary and vascular factors are likely to be important. How important they are to the UK’s ageing population is less clear and can only be estimated in representative samples [3, 4].

The Medical Research Council (MRC) Cognitive Function and Ageing Study (CFAS) is the largest UK-based multicentre study into functional and cognitive health of the elderly, which has provided robust estimates for prevalence and incidence. We report results of an analysis of relevant risk factors for incident dementia based on sociodemographic, medical, behavioural and family history data collected from study participants at prevalence screen.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 Key points
 Role of funding source
 Conflict of interest
 References
 
CFAS is a population-based study of dementia and cognition in six sites in the UK. It includes individuals aged 65 years and above from general practice-based populations including institutions, and it began in 1991.

The centres used here are set in two rural (Cambridgeshire and Gwynedd) and three urban (Nottingham, Newcastle and Oxford) locations. A detailed description of the design and the instruments has been published elsewhere [1, 2].

A two-phase sampling design was used (see Appendix 2: Supplementary data are available at Age and Ageing online: www.ageing.oxfordjournals.org). A total of 13,004 participants had a screening interview. Of these, 2640 had a detailed assessment, and these included individuals with all cognitive abilities, including the cognitively normal. A follow-up assessment after 2 years ascertained incident cases amongst the fully assessed group.

Those initially screened but not further assessed were screened again at 2 years, and a further sample was assessed in detail (7,175 screened and 1,443 assessed). A third wave (6 years) screened and assessed 1,489 of 2,016 non-demented survivors of both groups. Response rates in each of the sites ranged from 71 to 79%.

Screening interview (Wave 1)
The screening interview contained the following: sociodemographic, behavioural and self-reported medical variables, including years of full-time education, main occupation during working life [5], self-perceived health, vascular disease [angina [6], intermittent claudication [6], myocardial infarction (MI), any cardiovascular disease (CVD), stroke, transient ischaemic attack (TIA) and diabetes mellitus], neuropsychiatric disease or trauma [epilepsy, head injury, exposure to general anaesthesia (GA), depression, psychiatric disorder, Parkinson’s disease (PD), headache, meningitis/encephalitis and shingles], respiratory disorders (chronic bronchitis, asthma and any pulmonary disease), miscellaneous medical conditions (peptic ulcer disease, pernicious anaemia, thyroid disease and arthritis), smoking and drinking habits, Mini-Mental State Examination [7], questions about organic-type mental symptoms from the Automated Geriatric Examination for Computer Assisted Taxonomy (AGECAT) [8] and family history of a range of diseases including Alzheimer’s disease (AD) (specific wording of questions available on the website http://www-cfas.medschl.cam.ac.uk).

Subject assessment interviews (Waves 2 and 3)
Dementia diagnosis was derived from the Geriatric Mental State at assessment interview using the AGECAT algorithm. This produces a score of between 0 and 5 in an organicity scale, for which 3 and above is equivalent to dementia as diagnosed by DSM-III-R [9, 10]. Proxy schedules were used for diagnosis when access to the subject was denied. Respondents were interviewed by trained interviewers from professions allied to medicine.

Definition of cases and controls
The standard methodology for defining cases and controls within nested studies was used and is explained here briefly [11]. Two case–control groups (1 and 2) were identified: one for Wave 2 interviews and another for Wave 3 interviews (see Appendix 2: Supplementary data are available at Age and Ageing online: www.ageing.oxfordjournals.org). Group 1 was defined from the study population of individuals not demented at baseline and interviewed at Wave 2 (n = 1,346), together with those who were non-demented at the incidence assessment interview (n = 1,240). Group 2 was defined from those who were not demented at baseline or Wave 2 and who were interviewed at Wave 3 (n = 1,489). Dementia was defined as scoring 3–5 on the AGECAT diagnostic algorithm; controls scored <3. Cases for Group 1 were defined as individuals who obtained a dementia diagnosis during the interval between the baseline and Wave 2 assessments (n = 133); their controls were those assessed at Wave 2 who did not have dementia diagnoses (n = 2,453). Cases for Group 2 were individuals not demented at Wave 2 but who developed dementia between Wave 2 and Wave 3 assessments (n = 142); their controls were individuals who never had a dementia diagnosis (n = 1,347). Thirty individuals with missing educational status at baseline were excluded from the analysis.

Analysis
To create accurate confidence intervals (CIs) from two-phase studies, it is necessary to backweight cases and controls to the population using inverse probability weighting. The weights are defined as the inverse probabilities of being selected for assessment at the second wave (according to age, sampling strategy and cognitive status at first and second wave interview and path to second wave assessment). Logistic regression was used to calculate crude odds ratios (ORs), and ORs were adjusted for age, sex, education and social class for the various risk factors. Stata 7.0 [12] was used with the option pweight to specify sampling weights.

MRC CFAS has Multi-centre Research Ethics Committee’s approval and ethical approval from the relevant local research ethics committees. All participants gave informed consent. Version 6.3 of the data is used.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 Key points
 Role of funding source
 Conflict of interest
 References
 
Two hundred and seventy-seven cases of dementia were identified at Waves 2 and 3, and 3,800 controls were identified for analysis (9,752 person years at risk equivalent). Their sociodemographic characteristics are presented in Table 1. Full results on dropout and attrition are given elsewhere [13], but individuals who were older or with cognitive impairment at baseline were more likely to die or refuse the next interview wave. Age was an important independent risk factor for dementia. Estimates show increasing risk with advancing age, ranging from an OR of 1.8 in people aged 70–74 years to 25.6 in those aged 90+ when compared with people aged 65–69 years. Women were at marginally increased risk of incident dementia compared with men (OR = 1.6, 95% CI = 1.1–2.4). Increasing years in full-time education was associated with a decrease in incident dementia (Ptrend = 0.02), though social class was not associated with dementia. There was no difference in risk across rural and urban sites (Table 2).


View this table:
[in this window]
[in a new window]
  		Table 1.. Distributions and incident dementia risk estimates for age, sex and education—Medical Research Council Cognitive Function and Ageing Study

 

View this table:
[in this window]
[in a new window]
  		Table 2.. Self-perceived health, alcohol and smoking history and incident dementia risk—Medical Research Council Cognitive Function and Ageing Study

 

Self-perceived health was predictive of incident dementia (Table 2). Compared to the response ‘good’, those who replied ‘excellent’ were at similar risk (OR = 1.0, 95% CI = 0.6–1.7), whereas those who replied ‘fair’ or ‘poor’ were at increased risk (OR = 1.5, 95% CI = 1.0–2.3 and OR = 3.9, 95% CI = 2.2–6.9, respectively, Ptrend<0.001).

Risk for never drinkers was not different from risk for ever drinkers (OR = 0.8, 95% CI = 0.5–1.3). Similarly, smoking and dementia were only weakly associated: past versus never smokers (OR = 0.7, 95% CI = 0.5–1.0) and current versus never smokers (OR = 0.9, 95% CI = 0.5–1.5) (Table 2).

Angina pectoris and intermittent claudication, self-reported history of MI, hypertension and TIA were not associated with dementia. Individuals with a self-reported history of stroke were at increased risk (OR = 2.1, 95% CI = 1.1-4.2). Only PD (OR = 3.5, 95% CI = 1.3–9.3), of the neurological and psychiatric disorders examined (i.e. depression, epilepsy, headaches, serious head injury, PD and meningitis/encephalitis), and shingles (OR = 0.6, 95% CI = 0.4–0.9), of all other medical histories (i.e. arthritis, asthma, chronic bronchitis, diabetes mellitus, shingles, thyroid disease, peptic ulcer disease and pernicious anaemia), were associated with dementia. The CIs were often wide for the rare disorders (Table 3). Exposure to GA was inversely associated with dementia; this effect was consistent as a trend with increasing exposure to anaesthesia (never GA: OR = 1.7, 95% CI = 1.1–2.7; one GA: OR = 1.4, 95% CI = 0.9–2.0; two–five GA: referent 1.0; more than five GA: OR = 0.7, 95% CI = 0.3–1.5, Ptrend = 0.003).


View this table:
[in this window]
[in a new window]
  		Table 3.. Medical history and incident dementia risk—Medical Research Council Cognitive Function and Ageing Study

 

Multivariable analysis (data not shown) indicated that increased incidence of dementia was independently associated, in women, with increasing age, poor self-reported health, smoking history, not having had GA, reporting PD, shingles and stroke with minimal impact on effect sizes.

There were no increased risks associated with reporting a first-degree family history of a range of disorders (see Table in Appendix 1: Supplementary data are available at Age and Ageing online).


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
Key points
 Role of funding source
 Conflict of interest
 References
 
In this population-based analysis of 275 incident cases over 7 years (9,752 person years equivalent), increasing age, female sex, poor self-perceived health, history of stroke and PD were associated with increased risk of dementia. Shingles, being an ex-smoker and exposure to GA were independently associated with decreased risk of dementia. In this cohort, previously reported factors such as social class and drinking history or other medical and family history were not found to be associated with dementia.

Whilst the diagnostic measure used in CFAS is not clinician based, it is clinically based in its content and systematic in its application. The Geriatric Mental State was designed to examine and record the mental state of elderly subjects (in a manner equivalent to DSM-III-R criteria [10]). AGECAT [8], the algorithmic diagnosis applied to the interview content, is the program containing the diagnostic algorithm. This incorporates the cognitive, functional and behavioural impairments that characterise dementia and has been shown to be robust [14, 15]. No clinical subtyping of cases was included in the analysis, and the ratios are for dementia rather than subtype. The neuropathological work of this study has shown the common occurrence of both vascular pathology and Alzheimer’s type pathology on demented as well as non-demented individuals post-mortem [16].

As with other nested case–control studies, the results are subject to potential bias from missing data arising from loss to follow-up. Around 40% of the individuals had been lost to the study by the second wave of assessment interviews (mostly due to death or refusal), and another 40% before Wave 3. Most nested case–control studies do not report their detailed loss to follow-up, and therefore, it is difficult to directly assess the relative effects of such drop out across studies. In MRC CFAS, risk for drop out changes according to the length of the study and with factors that are strongly associated such as age, cognition, functioning and self-perceived health [13]. These factors would make the estimates presented here more conservative. For incidence case ascertainment, the bias inherent in the follow-up process has been further checked by using two time points. Estimates that change direction between waves were likely to have been influenced by bias and are less likely to be robust. This is shown by the sex effect within our study, which gets stronger with longer follow-up, consistent with an increase in mortality in men.

Health variables
Exposure was measured using interview questions. The questions were designed and pre-tested to be as unambiguous as possible, there was limited information available beyond never-ever exposure and the physical illness section was self-report rather than clinically confirmed. Stroke is known to be reasonably accurately reported and shows a clear risk pattern. This effect size is similar to that reported from the Framingham Study [17]. The increased risk with PD is consistent with that in the reported literature [18] and shows that this finding can be detected, even with self-reported data. The length of follow-up is important as a short follow-up would be biased by such incipient cases. These results are of clinical relevance since individuals with stroke and PD are likely to be under the care of medical specialities that can detect, manage support and individuals with incident dementia. The lack of strong contribution of Lewy body pathology to dementia expression [16] in the neuropathology study could be a sample size issue or possibly that dementia syndrome in PD is associated with other unmeasured aspects of pathology of PD. The protective effect of GA is of interest and may reflect less access to health care for those who are cognitively impaired but not demented [19]. Other vascular risk factors are measured by self-report, and whilst Rose angina questionnaire is validated, there are no biological measures.

Sociodemographic variables
The study sample is well defined, population based and can be seen as representative of England and Wales, given the lack of variation in prevalence and incidence of dementia across the five sites [1, 2]. The prospective design features include a well-defined sampling frame, large initial sample, targeted sampling for assessment (including cognitively normal individuals), careful standardisation of screening and assessment instruments, repeated follow-up of assessed individuals and an analytical strategy that adjusts for the complex sampling design.

The study, which is robust for the oldest age groups, confirms the consistent rise in incidence dementia in the older age groups with no reduction in risk. The lower risk for men may be due to the known mortality of men at earlier stages of cognitive morbidity [20]. If men begin to show the same survival as women, then that pattern of dementia may change [21]. The lack of a strong education effect within the UK population could be due to the standardisation of the diagnostic method or due to the relative homogeneity of the education of the UK population; the small effect seen is not independent of the other risk factors. It does highlight that changes in levels of secondary education are unlikely to have major knock-on reduction in dementia at later ages. The potential effect of tertiary education is too rare in our population to test this as a protective factor.

Alcohol and smoking
The lack of either risk or protection effect of alcohol adds to the current debate in this area, since there are such conflicting findings in the literature, with some advocating its positive effects [22] and others its potential damaging effects [23]. Although the measure of exposure was crude, the proportion of heavy duties in the population is tiny and we would expect to see a beneficial effect even if there were a J-shaped curve. The measure of exposure avoids the misclassification of ex-drinkers who have given up due to illness. Similarly, ex-smokers were at lower risk than both current and never smokers. This finding is consistent with the published literature supporting the assumption that the main effect being observed is a survival one.

Self-perceived health is a well-recognised predictor of mortality and functional status [20], but its relationship to incident dementia has not been well explored [24]. This emerged in an analysis as an unexpected predictor of dementia even after rigorous control of other variables including other health status measures [25]. Global measures are likely to play an increased role in research in older populations. They are key to quality of life and may prove to provide prediction of outcomes to identify individuals at risk beyond specific individual risk factors.

In conclusion, this large prospective multicentre study of dementia among older people in England and Wales confirmed age, sex, history of stroke and PD as key risk factors but did not support factors such as education, smoking, drinking and other vascular disease. The lack of positive associations within a population-based study suggests caution in the interpretation of single-risk factor studies from specific cohorts. These results from such larger-population representative studies may provide clues to why some trials have disappointing results. New cohort studies could identify biological measures from mid-life, but prospects for easily addressed risk factors with major effects are unlikely. The association between self-perceived health and incident dementia risk was unexpected and warrants further investigation.


   Key points
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
Role of funding source
 Conflict of interest
 References
 

  • Large UK-based multicentre study examining longitudinal risk of dementia in representative populations.
  • Strong increased risk with age, moderate with stroke and Parkinson’s disease, minimal for female sex and higher levels of education.
  • Good/excellent self-reported health is associated with long-term protective effect, but smoking and alcohol consumption are not. History of GA was associated with reduced risk.


   Role of funding source
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 Role of funding source
Conflict of interest
 References
 
The funding bodies have had no influence on the paper or decision to publish.


   Conflict of interest
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 Role of funding source
 Conflict of interest
References
 
None declared.


   Acknowledgements
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 Role of funding source
 Conflict of interest
 References
 
MRC CFAS has been funded by programme grants from the Medical Research Council and Department of Health. We thank all the respondents, their families and their primary care teams, from across the country, for their continued participation in CFAS.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 Role of funding source
 Conflict of interest
 References
 

  1. MRC CFAS. Cognitive function and dementia in six areas of England and Wales: the distribution of MMSE and prevalence of GMS organicity level in the MRC CFA Study. The Medical Research Council Cognitive Function and Ageing Study (MRC CFAS). Psychol Med 1998; 28: 319–35.[CrossRef][Web of Science][Medline]
  2. Matthews FE, Brayne C, Medical Research Council Cognitive Function and Ageing Study Investigators. The incidence of dementia in England and Wales: findings from the five identical sites of the MRC CFA Study. PLoS Med 2005; 2: e193.
  3. Launer LJ, Andersen K, Dewey ME et al. Rates and risk factors for dementia and Alzheimer’s disease: results from EURODEM pooled analyses. EURODEM Incidence Research Group and Work Groups. European Studies of Dementia. Neurology 1999; 52: 78–84.[Abstract/Free Full Text]
  4. Lindsay J, Laurin D, Verreault R et al. Risk factors for Alzheimer’s disease: a prospective analysis from the Canadian Study of Health and Aging. Am J Epidemiol 2002; 156: 445–53.[Abstract/Free Full Text]
  5. Elias P, Halstead K, Prady K. Computer Assisted Standard Occupational Coding. London: HMSO, 1993.
  6. Rose GA. The diagnosis of ischaemic heart pain and intermittent claudication in field surveys. Bull World Health Organ 1962; 27: 645–58.[Web of Science][Medline]
  7. Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–98.[CrossRef][Web of Science][Medline]
  8. Copeland JR, Dewey ME, Griffiths-Jones HM. A computerized psychiatric diagnostic system and case nomenclature for elderly subjects: GMS and AGECAT. Psychol Med 1986; 16: 89–99.[Web of Science][Medline]
  9. Copeland JR, Kelleher MJ, Kellett JM et al. A semi-structured clinical interview for the assessment of diagnosis and mental state in the elderly: the Geriatric Mental State Schedule. I. Development and reliability. Psychol Med 1976; 6: 439–49.[Web of Science][Medline]
  10. American Psychiatric Association. Diagnostic, Statistical Manual, Edition III, revised edition. Washington, DC: American Psychiatric Association, 1987.
  11. Breslow NE, Day NE. Statistical Methods in Cancer Research: Volume II. The Design and Analysis of Cohort Studies. Lyon, France: International Agency for Research on Cancer, 1987.
  12. Statacorp. Stata Statistical Software: Release 7.0. College Station, TX, 2001.
  13. Matthews FE, Chatfield M, Freeman C, McCracken C, Brayne C, MRC CFAS. Attrition and bias in the MRC cognitive function and ageing study: an epidemiological investigation. BMC Public Health 2004; 4: 12.
  14. Collighan G, Macdonald A, Herzberg J, Philpot M, Lindesay J. An evaluation of the multidisciplinary approach to psychiatric diagnosis in elderly people. BMJ 1993; 306: 821–4.[Abstract/Free Full Text]
  15. Ames D, Flynn E. Diagnosis of psychiatric disorder in elderly general and geriatric hospital patients – AGECAT and DSM-III-R compared. Int J Geriatr Psychiatry 1994; 9: 627–33.[CrossRef]
  16. Neuropathology Group. Medical Research Council Cognitive Function and Aging Study. Pathological correlates of late-onset dementia in a multicentre, community-based population in England and Wales. Neuropathology Group of the Medical Research Council Cognitive Function and Ageing Study (MRC CFAS). Lancet 2001; 357: 169–75.[CrossRef][Web of Science][Medline]
  17. Ivan CS, Seshadri S, Beiser A et al. Dementia after stroke: the Framingham Study. Stroke 2004; 35: 1264–8.[Abstract/Free Full Text]
  18. Aarsland D, Zaccai JH, Brayne C. A systematic review of prevalence and incidence studies of dementia in Parkinson’s Disease. Mov Disord 2005; 20: 1255–63.[CrossRef][Web of Science][Medline]
  19. Chi LY, Brayne C, Todd CJ, O’Connor DW, Pollitt PA. Predictors of hospital contact by very elderly people: a pilot study from a cohort of people aged 75 years and over. Age Ageing 1995; 24: 382–8.[Abstract/Free Full Text]
  20. Neale R, Brayne C, Johnson AL. Cognition and survival: an exploration in a large multicentre study of the population aged 65 years and over. Int J Epidemiol 2001; 30: 1383–8.[Abstract/Free Full Text]
  21. Hebert LE, Scherr PA, McCann JJ, Beckett LA, Evans DA. Is the risk of developing Alzheimer’s disease greater for women than for men? Am J Epidemiol 2001; 153: 132–6.[Abstract/Free Full Text]
  22. Letenneur L, Larrieu S, Barberger-Gateau P. Alcohol and tobacco consumption as risk factors of dementia: a review of epidemiological studies. Biomed Pharmacother 2004; 58: 95–9.[CrossRef][Medline]
  23. Anttila T, Helkala EL, Viitanen M et al. Alcohol drinking in middle age and subsequent risk of mild cognitive impairment and dementia in old age: a prospective population based study. BMJ 2004; 329: 539.
  24. Brown JW, Liang J, Krause N, Akiyama H, Sugisawa H, Fukaya T. Transitions in living arrangements among elders in Japan: does health make a difference? J Gerontol B Psychol Sci Soc Sci 2002; 57: S209–S220.
  25. Idler EL, Russell LB, Davis D. Survival, functional limitations, and self-rated health in the NHANES I Epidemiologic Follow-up Study, 1992. First National Health and Nutrition Examination Survey. Am J Epidemiol 2000; 152: 874–83.
Received April 5, 2005; accepted in revised form November 21, 2005.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 AJPHHome page
V. C. Crooks, J. Lubben, D. B. Petitti, D. Little, and V. Chiu
Social Network, Cognitive Function, and Dementia Incidence Among Elderly Women
Am J Public Health, July 1, 2008; 98(7): 1221 - 1227.
[Abstract] [Full Text] [PDF]

Home page
 Int J EpidemiolHome page
C. Brayne, C. McCracken, and F. E Matthews
Cohort Profile: The Medical Research Council Cognitive Function and Ageing Study (CFAS)
Int. J. Epidemiol., October 1, 2006; 35(5): 1140 - 1145.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow supplementary data
Right arrow All Versions of this Article:
35/2/154    most recent
afj030v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Matthews, F. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Matthews, F. E.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?