Abstract
Objective:
To determine risk and protective factors for mild cognitive impairment (MCI) among persons 85 years and older.
Methods:
Participants in the population-based prospective Mayo Clinic Study of Aging were comprehensively evaluated at baseline and at 15 monthly intervals to determine incident MCI. At baseline, lifestyle factors in midlife and late life were assessed by self-reported questionnaire; vascular and comorbid conditions were abstracted from participants' medical records.
Results:
Of 256 participants who were cognitively normal at enrollment (median age 87.3 years, 62% women), 121 developed MCI at a median 4.1 years of follow-up. Predictors of MCI were APOE ε4 allele (hazard ratio [HR] 1.89; p = 0.008), current depressive symptoms (HR 1.78; p = 0.02), midlife onset of hypertension (HR 2.43; p = 0.005), increasing number of vascular diseases (HR 1.13; p = 0.02), and chronic conditions from the Charlson Comorbidity Index (HR 1.08; p = 0.006). Models were adjusted for sex and education, with age as the time variable. The risk of MCI was reduced for participants who reported engagement in artistic (HR 0.27; p = 0.03), craft (HR 0.55; p = 0.02), and social (HR 0.45; p = 0.005) activities in both midlife and late life, and in the use of a computer in late life (HR 0.47; p = 0.008).
Conclusions:
Chronic disease burden increases risk of MCI, whereas certain lifestyle factors reduce risk in persons 85 years and older. This implies that preventive strategies for MCI may need to begin in midlife and should persist throughout late life.
Individuals aged 85 years and older are the most rapidly growing group in the United States and worldwide.1 Studies of the oldest old are difficult to conduct and to interpret. Persons aged 90 years and older typically have sensory losses, difficulty providing valid and reliable information, high comorbidity, and a high prevalence of dementia2; most are typically women. Often, factors associated with risk of cognitive impairment at younger ages are no longer predictive, raising the possibility that multiple coexisting diseases might be more predictive than solitary diseases. Because many individuals aged 90 years and older already have early stages of mild cognitive impairment (MCI), studies are often cross-sectional and can only assess risk of dementia or Alzheimer disease (AD). Furthermore, potential interventions at these ages may have limited long-term benefit. The goal of this study was to identify risk and protective factors for incident MCI among cognitively normal persons aged 85–89 years at enrollment to the Mayo Clinic Study of Aging (MCSA).
METHODS
Study cohort at baseline.
Participants were randomly selected from among Olmsted County, Minnesota, residents for participation in the MCSA. Details of the study design and methodology have been published.3,4 Briefly, residents aged 70–89 years were identified using the medical records linkage system of the Rochester Epidemiology Project (REP).5 Eligible participants were invited to participate in person or by telephone. This study is limited to participants who were aged 85–89 years at enrollment (October 1, 2004, or March 1, 2008) and were cognitively normal at the baseline evaluation.
In-person evaluation.
The evaluation consisted of 3 components. A nurse or study coordinator interviewed the participant to assess memory and administered the Clinical Dementia Rating scale6 and the Functional Activities Questionnaire (FAQ)7 to an informant to assess participant functioning. A physician evaluation included the Short Test of Mental Status8 and a neurologic examination. A psychometrist performed neuropsychological testing using 9 tests to assess performance in memory, executive function, language, and visuospatial skills. The raw test scores were transformed into age-adjusted scores using normative data.9 Domain scores were computed by summing and scaling the age-adjusted test scores within domains to allow comparisons across domains.9
Diagnostic criteria.
MCI was diagnosed per published criteria—cognitive concern, impairment in 1 or more of the 4 cognitive domains, essentially normal functional activities, and absence of dementia3,4,10—and classified as amnestic (aMCI) or nonamnestic MCI (naMCI). Dementia was diagnosed according to DSM-IV criteria.11 Participants were considered cognitively normal if they performed within the normative range and did not meet MCI or dementia criteria.3,4,10
Exposures and covariates.
Demographic information, weight, height, and timed gait speed (m/s) were determined at the interview. A stroke history was obtained by the physician and validated in the medical record. Depressive symptoms in the previous month were assessed using the Neuropsychiatric Inventory Questionnaire (NPI-Q).12 Participants completed self-administered questionnaires on engagement in exercise and in cognitive activities in midlife (age 50 years) and late life (1 year prior to the evaluation). Medical comorbidities and date of onset of these conditions were abstracted from participant medical records using the REP medical records-linkage system.5,13 APOE genotyping was performed. Chronic disease burden was assessed from a weighted Charlson Comorbidity Index (CCI) score14 using ICD-9 codes (table e-1 on the Neurology® Web site at Neurology.org). This score takes into account disease severity, and was developed to assess impact of disease burden on health outcomes. Vascular disease burden was assessed as the total number of vascular diseases and related conditions: type 2 diabetes mellitus, hypertension, dyslipidemia, coronary artery disease, congestive heart failure, atrial fibrillation, peripheral vascular disease, stroke, and obesity.
Longitudinal follow-up.
Follow-up was performed at 15-month intervals. To avoid potential bias in making a diagnosis, clinical and cognitive findings from previous evaluations were not considered. Participants who declined the in-person evaluation at follow-up were invited to participate by a telephone interview that included the Telephone Interview of Cognitive Status–Modified,15 the Clinical Dementia Rating scale,6 and the NPI-Q.12
Standard protocol approvals, registrations, and patient consent.
This study was approved by the Institutional Review Boards of the Mayo Clinic and Olmsted Medical Center. Written informed consent was obtained from all participants.
Statistical analyses.
The date of MCI onset was assigned at the midpoint between the last assessment as cognitively normal and the first-ever assessment as MCI; participants who developed dementia without an MCI diagnosis were included. Persons who were lost to follow-up or died were censored at their last evaluation. We estimated follow-up time from baseline to onset of MCI, date of censoring, or last follow-up. We investigated bivariate associations of risk and protective factors with MCI using Cox proportional hazards models adjusted for sex and education, with age as a time variable. Exposures and covariates were APOE ε4 allele (any ε4 vs no ε4), type 2 diabetes, hypertension, dyslipidemia, body mass index, engagement in exercise and cognitively stimulating activities, high-sensitivity C-reactive protein, and smoking (current and former vs never). When possible, we characterized variables as present in midlife (≤65 years) or late life (>65 years); or in midlife only, late life only, or both. We performed stratified analyses by MCI subtypes and by sex. In multivariable models, we included variables that were significantly associated with MCI.
In sensitivity analyses, we estimated the annualized percent change (slope) in FAQ score, memory, and executive function z scores for each participant, and computed and compared the average slope for participants grouped by performance of activities in midlife or late life. Associations were considered significant at p values <0.05, using SAS version 9.3 (SAS Institute, Cary, NC).
RESULTS
Of the 301 participants who were cognitively normal at enrollment, 256 (85.0%; median age, 87.3 years, 62.1% [n = 159] women) had ≥1 follow-up evaluation (table 1). Women had a lower frequency of an APOE ε4 allele, were older, had a lower median CCI score, and had a lower vascular disease burden than men. Participants with no follow-up were younger (median age, 86.6 years; p = 0.03) but did not differ from those with follow-up regarding sex, education, APOE ε4 allele, diabetes, hypertension, gait speed, engagement in exercise, or cognitive activities.
Table 1.
Characteristics of study participants at baseline, overall and by sex
Predictors of incident MCI.
There were 121 incident MCI events over a median 4.1 (interquartile range 2.6, 6.1) years of follow-up. MCI risk was associated with an APOE ε4 allele, hypertension (midlife and late life), depressive symptoms, and increased with increasing CCI score and vascular disease burden (table 2). The risk was reduced in participants engaging in artistic, craft, and social activities in both midlife and late life, and computer use in late life. Estimates from the multivariable models were essentially unchanged (data are not presented). When stratified by midlife or late life, the risk of MCI was reduced in participants engaging in artistic, social, and group activities, and reading newspapers in midlife; and with artistic and craft activities, and computer use in late life (table e-2).
Table 2.
Risk and protective factors for mild cognitive impairment
Predictors of MCI subtypes.
In bivariate models, depressive symptoms were associated with an increased risk of aMCI; social activity in both midlife and late life was associated with reduced risk (table 3). The point estimates from the multivariable model were essentially unchanged, but the associations were marginally significant (data not presented). Significant predictors of naMCI were an APOE ε4 allele, diabetes (midlife and late life), hypertension (midlife and late life), increasing CCI score, and increasing vascular disease burden. The risk of naMCI was reduced with computer use in late life and with increasing gait speed. Estimates from the multivariable were similar; the association with education was significant (hazard ratio [HR] 0.74, p = 0.004). The associations of APOE genotype with naMCI persisted even with the ε3ε3 genotype as the reference category (table e-3).
Table 3.
Risk and protective factors for MCI stratified by MCI subtype
Predictors of MCI by sex.
In men, MCI was associated with depressive symptoms and higher CCI scores in bivariate models (table 4); estimates changed little in multivariate models (data not shown). In women, MCI risk factors were hypertension and diabetes in midlife and late life, and increasing vascular disease burden. The risk was reduced with engaging in crafts and artistic activities in midlife and late life, and computer use in late life in bivariate models, and changed little in multivariate models (data not shown).
Table 4.
Risk and protective factors for MCI stratified by sex
Sensitivity analyses.
Participants who never engaged in cognitively stimulating activities had greater average declines in FAQ score, memory, and executive function z scores compared to those who did; these differences did not reach statistical significance due to small sample sizes. For example, for computer use, decline in FAQ score was −0.790 for never vs −0.105 for computer use in late life only; decline in memory was −0.074 for never and −0.044 for late life only; decline in executive function was −0.170 for never and −0.108 for late life only. For craft activities, the decline in FAQ score was −0.782 for never, −0.691 for midlife only, and −0.439 for both midlife and late life; decline in memory was −0.071 for never, −0.063 for midlife only, and −0.055 for both midlife and late life; decline in attention was −0.159 for never, −0.192 for midlife only, and −0.125 for both midlife and late life. For engagement in social activities, the average decline in FAQ was −0.954 for never, −0.782 for midlife only, and −0.418 for midlife and late life; decline in memory was −0.120 for never, 0.041 for midlife only, and −0.082 for both midlife and late life; decline in attention was −0.176 for never, −0.177 for midlife only, and −0.138 for both midlife and late life.
DISCUSSION
In our cohort of 85- to 89-year-olds, the risk of MCI was elevated in participants with an APOE ε4 allele, hypertension onset in midlife, greater comorbidity and vascular disease burden, and depressive symptoms. By contrast, the risk was reduced with engagement in artistic, craft, and social activities in both midlife and late life, and with use of a computer in late life. Depressive symptoms and APOE ε4 were associated with aMCI, and vascular factors were associated with naMCI. Our findings suggest that strategies to reduce risk of MCI in the oldest old should include prevention and efficient management of vascular and other chronic diseases earlier in life. These nonpharmacologic interventions may have greatest benefit when initiated early and maintained. Furthermore, these efforts should begin in young adulthood or midlife, and should persist throughout late life.
A strength of our study was that it distinguished the impact of factors in midlife, late life, or both on MCI risk. For factors associated with reduced risk, this allowed us to disentangle the potential effects of reverse causation, where the outcome precedes and causes the exposure. Engaging in beneficial lifestyle activities in midlife only, or initiating them in late life, did not consistently confer benefit. Persons who performed certain activities only in midlife may have ceased to perform them in late life due to incipient cognitive impairment. Alternatively, those who performed activities in late life may have done so because they still could. Others may have discerned cognitive decline and initiated activities in late life as an effort to curb progression, suggesting reverse causation.
Our findings suggest that the burden of chronic conditions or vascular diseases may predict MCI in persons aged 85 years and older. Multiple chronic conditions may contribute to longstanding pathologic insults to the brain through cerebrovascular disease and related mechanisms, including endothelial dysfunction, inflammation, and oxidative damage, which lead to neuronal death, synaptic dysfunction, and cognitive impairment.16 Certain established risk factors for cognitive impairment (e.g., hyperlipidemia, smoking, C-reactive protein, obesity) may not have predicted MCI because of survival bias, or because the cumulative burden of disease may have greater impact on risk than a single disease at very old ages.
The association of APOE ε4 with aMCI is consistent with a neurodegenerative etiology. The association with naMCI, however, was unexpected, but may relate to the adverse vascular effects of APOE ε2 and ε4.17–19 APOE ε2 and ε4 alleles have been reported to increase atherogenic lipoproteins and accelerate atherogenesis.20 These effects are consistent with the hypothesized vascular etiology for naMCI, and consistent with our findings in table e-3. Thus, although the small numbers suggest a spurious association, the present findings may be real. However, they remain to be validated in a larger population-based cohort of oldest old.
The association of midlife hypertension with MCI underscores the need for aggressive prevention at younger ages. Targeted education of the general population regarding the association of vascular disease with MCI risk may promote lifestyle changes and treatment compliance. Effectively monitoring and managing persons with hypertension particularly in midlife may also prevent adverse cardiovascular outcomes that increase MCI risk.21,22
Interestingly, our findings suggest that nonpharmacologic preventive strategies may reduce naMCI risk in the oldest old. Higher education may reduce risk by increasing cognitive reserve, which in turn may delay clinical expression of symptoms or counteract vascular assaults on the brain.23,24 The reduced risk with computer use and with artistic or crafts activities suggest that these activities should be promoted throughout life. These activities may also increase cognitive reserve, maintain neuronal function, stimulate neural growth, and recruit alternate neural pathways to maintain cognitive function.25
By contrast, failure to observe protective factors for aMCI suggests that oldest old participants at risk for aMCI possibly have greater pathology resulting from both neurodegenerative and vascular effects that may be less amenable to nonpharmacologic interventions. Although we did not observe significant associations with exercise, the reduced HR for persons who exercised in late life suggests a potential benefit for MCI (0.55 for MCI, 0.66 for aMCI; table 2 footnote).
In sex-stratified analyses, MCI risk increased with increasing burden of chronic disease in men and with increasing burden of vascular disease in women. This difference raises the hypothesis that oldest old men may be sicker than women, and this multifactorial morbidity may contribute to MCI. The higher vascular disease burden in women, as observed with diabetes and hypertension, may increase MCI risk. A lower vascular disease burden in men may be due to the earlier onset of these conditions in men, leading to earlier mortality (survival bias). Finally, the protective effects of artistic and craft activities and computer use in women suggest opportunities for exploring comparable interventions in men.
Some of our findings are consistent with previous studies among persons 90 years and older. In particular, the beneficial effects of cognitively stimulating activities on cognition are consistent with previous findings.25,26 Absence of a protective effect of education on aMCI risk in the present study is consistent with the rapid cognitive decline observed in highly educated late-stage aMCI patients from a memory clinic.27 The association of increasing gait speed with decreased risk of naMCI is in keeping with the documented increase in the odds ratio of dementia with decreasing gait speed in persons in the 90+ study28 and with MCI in the MCSA cohort.29 In the MCSA, depressive symptoms were associated with an increased risk of aMCI.30 This is consistent with the putative role of aMCI as being a prodromal AD stage. Similarly, another study reported an association of late-life depressive symptoms with increased risk of AD, and suggested that recurrent depression was more likely to be etiologically associated with vascular dementia.31 Together, these findings suggest that depressive symptoms may be in the causal pathway or may be a marker for incipient aMCI or AD.
Certain of our findings, however, are inconsistent with prior studies among persons aged 90 years and older. Although the association of APOE ε4 with incident MCI has not been reported in the oldest old, the impact of the ε4 allele on dementia risk is thought to be minimal or absent in the oldest old.32 In the 90+ study, APOE ε4 carrier status (vs ε3ε3) was associated with dementia in cross-sectional but not in prospective analyses.33 The frequency of an APOE ε4 allele was only 8.1% in the incident cohort, compared to 18.8% in the present study. Thus, earlier mortality associated with APOE ε4 may preclude detection of significant associations with cognitive risk in studies among persons age 90 years and older. With the ε3ε3 genotype as the reference group, the association of the ε4 allele with MCI persisted (table e-3). Hypertension in midlife was associated with naMCI risk, but a similar association was not observed in a 90+ study cohort2 perhaps because of survival bias, cross-sectional design, different study endpoints, or to failure to take the age at onset of hypertension into account.
One potential limitation of our study is that cognitive activities and exercise were assessed only at baseline. Thus, we were unable to determine the effects of discontinuation of these activities during follow-up. Second, there may be residual reverse causation; declining cognition could have promoted engagement in cognitive activities in some participants.34 However, declining cognitive activities have been shown to predict cognitive impairment, but not the reverse.26 Given the old age of participants, there is a potential for recall bias. We previously assessed the reliability of the physical activity questionnaire in 87 persons who completed the questionnaire at 2 time periods. The internal consistency was moderate to good, a Cronbach α of 0.71, and the test-retest Spearman rank correlation coefficient was 0.50 for moderate exercise.35 Engagement in cognitive activities in midlife/late life may simply be a marker for healthy cognition, with a noncausal association with MCI. Finally, our findings are based on a cohort with a primarily Northern European ancestry.
Strengths of our study include the population-based, prospective design, the comprehensive evaluation of participants, and the assignment of cognitive status by consensus both at enrollment and follow-up. The ascertainment of chronic diseases and medical conditions from the medical records allowed us to reliably determine the presence and onset of risk factors in midlife and late life. The ascertainment of cognitively stimulating activities and exercise in a population-based cohort of cognitively normal 85- to 89-year-olds allowed us obviate the potential impact of reverse causation. Blinding the study evaluators to previous clinical assessments and cognitive diagnosis contributed to unbiased ascertainment of cognitive status during follow-up.
Supplementary Material
ACKNOWLEDGMENT
The authors thank the Mayo Clinic Study of Aging staff and participants for their involvement; Mary J. Dugdale, RN, Connie J. Fortner, RN, and Julie A. Gingras, RN, for the abstraction of medical record data; and Sondra L. Buehler for administrative assistance.
GLOSSARY
- AD
Alzheimer disease
- aMCI
amnestic mild cognitive impairment
- CCI
Charlson Comorbidity Index
- DSM-IV
Diagnostic and Statistical Manual of Mental Disorders, 4th edition
- FAQ
Functional Activities Questionnaire
- HR
hazard ratio
- ICD-9
International Classification of Diseases–9
- MCI
mild cognitive impairment
- MCSA
Mayo Clinic Study of Aging
- naMCI
nonamnestic mild cognitive impairment
- NPI-Q
Neuropsychiatric Inventory Questionnaire
Footnotes
Supplemental data at Neurology.org
AUTHOR CONTRIBUTIONS
R.O. Roberts had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: R.O. Roberts. Acquisition of data: R.O. Roberts, Dr. Machulda, Dr. Mielke, Dr. Knopman, Dr. Boeve, Dr. Geda, Dr. Petersen. Analysis and interpretation of data: R.O. Roberts, R.H. Cha. Drafting of the manuscript: R.O. Roberts. Critical revision of the manuscript for important intellectual content: Dr. Mielke, Dr. Machulda, Dr. Knopman, Dr. Geda, Dr. Petersen. Statistical analysis: R.H. Cha. Obtained funding: R.O. Roberts, Dr. Petersen, Dr. Knopman. Administrative, technical, or material support: R.O. Roberts, Dr. Petersen. Study supervision: R.O. Roberts.
STUDY FUNDING
Supported by the National Institute on Aging (U01 AG006786, P50 AG016574, K01 AG028573, and K01 MH68351), the Mayo Foundation for Medical Education and Research, and the Rochester Epidemiology Project (R01 AG034676).
DISCLOSURE
R. Roberts, R. Cha, M. Mielke, and Y. Geda report no disclosures relevant to the manuscript. B. Boeve has served as a consultant to GE Healthcare; receives publishing royalties from The Behavioral Neurology of Dementia (Cambridge University Press, 2009); and receives research support from Cephalon Inc., Allon Therapeutics Inc., the NIH/National Institute on Aging, the Alzheimer's Association, and the Mangurian Foundation. M. Machulda reports research support from the NIH/National Institute on Deafness and Other Communication Disorders. D. Knopman serves as Deputy Editor for Neurology® and on a Data Safety Monitoring Board for Eli Lilly and Co., and is an investigator in clinical trials sponsored by Janssen Pharmaceuticals Inc. R. Petersen serves on scientific advisory boards for Pfizer Inc., Janssen Alzheimer Immunotherapy, Merck Inc., Roche Inc., and Genentech Inc., and receives royalties for Mild Cognitive Impairment (Oxford University Press, 2003). Go to Neurology.org for full disclosures.
REFERENCES
- 1.Vincent GK, Velkoff VA. The Next Four Decades, The Older Population in the United States: 2010 to 2050, Current Population Reports. Washington, DC: US Census Bureau; 2010:P25–P1138. [Google Scholar]
- 2.Peltz CB, Corrada MM, Berlau DJ, Kawas CH. Cognitive impairment in nondemented oldest-old: prevalence and relationship to cardiovascular risk factors. Alzheimers Dement 2012;8:87–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Roberts RO, Geda YE, Knopman DS, et al. The Mayo Clinic Study of Aging: design and sampling, participation, baseline measures and sample characteristics. Neuroepidemiology 2008;30:58–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Petersen RC, Roberts RO, Knopman DS, et al. Prevalence of mild cognitive impairment is higher in men: the Mayo Clinic Study of Aging. Neurology 2010;75:889–897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.St Sauver JL, Grossardt BR, Yawn BP, et al. Data Resource Profile: the Rochester Epidemiology Project (REP) medical records-linkage system. Int J Epidemiol 2012;41:1614–1624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology 1993;43:2412–2414. [DOI] [PubMed] [Google Scholar]
- 7.Pfeffer RI, Kurosaki TT, Harrah CH, Jr, Chance JM, Filos S. Measurement of functional activities in older adults in the community. J Gerontol 1982;37:323–329. [DOI] [PubMed] [Google Scholar]
- 8.Kokmen E, Smith GE, Petersen RC, Tangalos E, Ivnik RC. The short test of mental status: correlations with standardized psychometric testing. Arch Neurol 1991;48:725–728. [DOI] [PubMed] [Google Scholar]
- 9.Ivnik RJ, Malec JF, Smith GE, et al. Mayo's Older Americans Normative Studies: WAIS-R, WMS-R and AVLT norms for ages 56 through 97. Clin Neuropsychol 1992;6:1–30. [Google Scholar]
- 10.Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med 2004;256:183–194. [DOI] [PubMed] [Google Scholar]
- 11.American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), 4th ed Washington, DC: American Psychiatric Association; 1994. [Google Scholar]
- 12.Kaufer DI, Cummings JL, Ketchel P, et al. Validation of the NPI-Q, a brief clinical form of the Neuropsychiatric Inventory. J Neuropsychiatry Clin Neurosci 2000;12:233–239. [DOI] [PubMed] [Google Scholar]
- 13.Melton LJ., III History of the Rochester Epidemiology Project. Mayo Clin Proc 1996;71:266–274. [DOI] [PubMed] [Google Scholar]
- 14.Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613–619. [DOI] [PubMed] [Google Scholar]
- 15.Welsh KA, Breitner JCS, Magruder-Habib KM. Detection of dementia in the elderly using telephone screening of cognitive status. Neuropsychiatry Neuropsychol Behav Neurol 1993;6:103–110. [Google Scholar]
- 16.Kelleher RJ, Soiza RL. Evidence of endothelial dysfunction in the development of Alzheimer's disease: is Alzheimer's a vascular disorder? Am J Cardiovasc Dis 2013;3:197–226. [PMC free article] [PubMed] [Google Scholar]
- 17.Baum L, Lam LC, Kwok T, et al. Apolipoprotein E epsilon4 allele is associated with vascular dementia. Dement Geriatr Cogn Disord 2006;22:301–305. [DOI] [PubMed] [Google Scholar]
- 18.Chuang YF, Hayden KM, Norton MC, et al. Association between APOE epsilon4 allele and vascular dementia: the Cache County study. Dement Geriatr Cogn Disord 2010;29:248–253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Liu X, Li L, Liu F, et al. ApoE gene polymorphism and vascular dementia in the Chinese population: a meta-analysis. J Neural Transm 2012;119:387–394. [DOI] [PubMed] [Google Scholar]
- 20.Mahley RW, Weisgraber KH, Huang Y. Apolipoprotein E: structure determines function, from atherosclerosis to Alzheimer's disease to AIDS. J Lipid Res 2009;50(suppl):S183–S188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Roberts RO, Geda YE, Knopman DS, et al. Cardiac disease associated with increased risk of nonamnestic cognitive impairment: stronger effect on women. JAMA Neurol 2013;70:374–382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Knopman DS, Roberts RO, Geda YE, et al. Association of prior stroke with cognitive function and cognitive impairment: a population-based study. Arch Neurol 2009;66:614–619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Vemuri P, Lesnick TG, Przybelski SA, et al. Association of lifetime intellectual enrichment with cognitive decline in the older population. JAMA Neurol 2014;71:1017–1024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rapp SR, Espeland MA, Manson JE, et al. Educational attainment, MRI changes, and cognitive function in older postmenopausal women from the Women's Health Initiative Memory Study. Int J Psychiatry Med 2013;46:121–143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Wilson RS, Boyle PA, Yu L, Barnes LL, Schneider JA, Bennett DA. Life-span cognitive activity, neuropathologic burden, and cognitive aging. Neurology 2013;81:314–321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wilson RS, Segawa E, Boyle PA, Bennett DA. Influence of late-life cognitive activity on cognitive health. Neurology 2012;78:1123–1129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ye BS, Seo SW, Cho H, et al. Effects of education on the progression of early- versus late-stage mild cognitive impairment. Int Psychogeriatr 2013;25:597–606. [DOI] [PubMed] [Google Scholar]
- 28.Bullain SS, Corrada MM, Shah BA, Mozaffar FH, Panzenboeck M, Kawas CH. Poor physical performance and dementia in the oldest old: the 90+ study. JAMA Neurol 2013;70:107–113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Mielke MM, Roberts RO, Savica R, et al. Assessing the temporal relationship between cognition and gait: slow gait predicts cognitive decline in the Mayo Clinic Study of Aging. J Gerontol A Biol Sci Med Sci 2013;68:929–937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Geda YE, Roberts RO, Mielke MM, et al. Baseline neuropsychiatric symptoms and the risk of incident mild cognitive impairment: a population-based study. Am J Psychiatry 2014;171:572–581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Barnes DE, Yaffe K, Byers AL, McCormick M, Schaefer C, Whitmer RA. Midlife vs late-life depressive symptoms and risk of dementia: differential effects for Alzheimer disease and vascular dementia. Arch Gen Psychiatry 2012;69:493–498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Juva K, Verkkoniemi A, Viramo P, et al. APOE epsilon4 does not predict mortality, cognitive decline, or dementia in the oldest old. Neurology 2000;54:412–415. [DOI] [PubMed] [Google Scholar]
- 33.Corrada MM, Paganini-Hill A, Berlau DJ, Kawas CH. Apolipoprotein E genotype, dementia, and mortality in the oldest old: the 90+ Study. Alzheimers Dement 2013;9:12–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ghisletta P, Bickel JF, Lovden M. Does activity engagement protect against cognitive decline in old age? Methodological and analytical considerations. J Gerontol B Psychol Sci Soc Sci 2006;61:P253–P261. [DOI] [PubMed] [Google Scholar]
- 35.Geda YE, Roberts RO, Knopman DS, et al. Physical exercise, aging, and mild cognitive impairment: a population-based study. Arch Neurol 2010;67:80–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.