Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Jul 1.
Published in final edited form as: Exp Aging Res. 2022 Aug 5;49(4):334–346. doi: 10.1080/0361073X.2022.2106717

Cognitive Reserve and Cognitive Functioning Among Oldest-Old Adults: Findings from the Georgia Centenarian Study

Peter Martin 1, Yasuyuki Gondo 2, Gina Lee 1, John L Woodard 3, L Steven Miller 4, Leonard W Poon 4
PMCID: PMC9899291  NIHMSID: NIHMS1830098  PMID: 35929967

Abstract

Living a long life does not guarantee the maintenance of optimal cognitive functioning; however, similar to older adults in general, cognitive reserve may also protect oldest-old adults from cognitive decline. The purpose of this study was to assess cognitive reserve among centenarians and octogenarians and to evaluate a process model of cognitive reserve. A total of 321 centenarians and octogenarians from the Georgia Centenarian Study were included in this study. Cognitive reserve components included level of education, occupational responsibility, current social engagement, past engaged lifestyle, and activity. Cognitive functioning was measured with the Mini-Mental Status Examination. Structural equation modeling was computed, and the overall model fit well, χ2 (df=3) = 5.02, p = .17; CFI = .99, RMSEA = .05. Education is directly and indirectly related to cognitive functioning through occupational responsibility and past engaged lifestyle. Current social engagement is related to cognitive functioning directly and indirectly through current activities. The four direct predictors (i.e., education, current social engagement, current activity, and past engaged lifestyle) explained 35 percent of the variance in cognitive functioning. The results provide important information for cognitive reserve theories with implications for interventions that build cognitive reserve.

Cumulative resources over the life span are important contributors to well-being in later life (Martin & Martin, 2002). As the population of older adults increases, continued attention should be given to cognitive well-being. What accounts for individual differences in cognitive functioning in later life? How do older adults prevent cognitive decline and maintain healthy cognitive functioning? Cognitive reserve may hold the key to answering these questions.

Considering psychological resources regarding cognitive function and decline, a number of studies have revealed that cognitive reserve, such as life experiences, social engagement, educational and occupational exposure, and leisure activities are associated with slower rates of cognitive decline in late-life (Stern, 2009; Wang, 2002; Wang et al., 2012). Cognitive reserve refers to “the adaptability (i.e., efficiency, capacity, flexibility of cognitive processes) that helps to explain differential susceptibility of cognitive abilities or day-to-day function to brain aging, pathology, or insult” (Stern et al., 2020, p. 1306). Cognitive reserve is a theoretical construct that explains an individual’s ability to delay or avoid the onset of cognitive impairment (Stern, 2002). Such reserve is thought to accumulate over the life span with stimulating experiences, and this reserve may decrease the risk of dementia (Blondell et al., 2014; Harrison et al., 2015).

Recent literature suggests that education, occupational complexity, leisure and physical activities, as well as mentally stimulating activities supply cognitive reserve, acting as protective factors for reducing the likelihood of developing dementia (Cadar et al., 2016; Harrison et al., 2015; Meng & D’Arcy, 2012; Tucker-Drob et al.,2009; Stern et al., 2020). Marioni et al. (2012) used the term cognitive lifestyle, which includes education, midlife occupation, and late-life social engagement, in their study to investigate the combined and independent effects of these factors on cognitive transitions. Using such measures, Marioni et al. found evidence for the association between cognitive lifestyle and compression of cognitive morbidity. More education and more complex mid-life occupation were related to reduced duration of moderate-to-severe impairment before death, as well as decreased rate of transitioning from no impairment to mild impairment. Late-life social engagement was also related to the decreased rate of transitioning from mild to moderate to severe impairment, and lower mortality risk for cognitively non-impairment (Marioni et al.).

Educational attainment is the most commonly used cognitive reserve proxy. Numerous studies have used level of education as an independent variable for predicting better cognitive function in late life. There is some evidence that although educational attainment is associated with cognitive performance, it is not associated with the rate of age-associated cognitive decline in late-life (Muniz-Terrera et al., 2008; Tucker-Drob et al., 2009; Wilson et al., 2009). Wilson et al. (2009) examined the associations between educational attainment and cognition as well as the rate of change in cognitive functioning among 6,000 older residents from a community in Chicago over roughly 14 years using mixed effect methods. Their finding suggests that high educational attainment was associated with high cognitive functioning at baseline. However, this association was probably not a linear relationship. Rate of cognitive decline was not associated with level of educational attainment.

Although there is much evidence that educational attainment is associated with cognitive functioning, it is not clear what mechanisms explain the association. Therefore, studies on cognitive reserve have gone beyond education in recent years and have used multi-component assessments to capture cognitive reserve more comprehensively (Lavrencic et al., 2018). Occupational complexity, social participation, and engagement in activities may explain the association between education and cognitive functioning. For example, Smart et al. (2014) reported that occupational complexity with work and with people was associated with better cognitive performance at age 70, after controlling for childhood cognitive skills, education, and social deprivation. Similarly, Richards et al. (2019) recently reported that educational attainment and midlife occupational complexity exerted strong influences on mental status in later life. Studies on cognitive reserve, however, have not investigated the interrelationship of cognitive reserve components and how they build reserves over the life span with effects in very late life.

A Japanese study indicated that greater social group participation was associated with the prevention of cognitive decline for women, after controlling for age, family, and health-related variables (Tomioka et al., 2018). James et al. (2011) also indicated that cognitive decline was reduced among individuals who were socially active when compared to individuals who were infrequently socially active.

Cognitive reserve has not widely been studied among the oldest-old (often defined as 85 years and older) population. Lavrencic et al. (2018) examined a link between cognitive reserve and cognitive function among the oldest-old population. They reported that higher cognitive reserve (i.e., education, occupation-based social class, marital status, engagement in mental activities, social participation, and physical activity level) was associated with a reduced prevalence of dementia and better cognitive performance among oldest-old adults. Another study linked more education with lower dementia prevalence (Corrada et al., 2008) and less cognitive impairment (Goveas et al., 2016; Yaffe et al., 2011) among oldest-old women. However, as Lavrencic et al. (2018) pointed out, some studies failed to find a relationship between education and dementia incidence, and these conflicting results may be due to an overreliance on education as a cognitive reserve index. More work is needed to assess additional cognitive reserve components among oldest-old adults. Kliegel et al. (2004), for example, indicated that intellectual activities before age 80 and educational level were associated with better cognitive functioning in a sample of centenarians.

When including very old adults with different levels of cognitive impairment, it is impossible to assess the life-span accumulation of cognitive reserve directly. Therefore, researchers have to rely on proxy information. Previous studies have shown that assessing cognitive reserve as a combination of proxy measures provides a more accurate indication of an individual’s reserve level compared to only using educational level as a single measure (Lavrencic et al., 2018). Harrison et al. (2015) conducted a review of the literature and indicated that education, occupation, and mentally stimulating activities are protective against the risk of dementia, particularly when combining different indicators in the analyses.

Most of the research on cognitive reserve uses an overall summary index. The important components that make up this overall index have not been studied in their relationship and with regard to outcomes. The current study aimed to investigate the significant direct and indirect pathways of cognitive reserve variables as they relate to cognitive functioning among centenarians. As an index, higher cognitive reserve was associated with better cognitive function (Lavrencic et al., 2018). However, the cognitive reserve index does not allow to test the mechanisms by which education continues to exert an influence on cognitive function in very late life. We therefore propose direct and indirect pathways from education via other cognitive reserve components to cognitive function.

Figure 1 displays the cognitive reserve process model of our study. The model specifies education as an early life influence that may predict how involved individuals become in their occupation with different levels of responsibility. Education and occupational responsibility then determine how engaged individuals are over the life span with the effect on relatively high or low levels of activity and cognitive function in late life.

Figure 1.

Figure 1

Open in a new tab

Cognitive Reserve Process Model

We hypothesized a number of potential pathways from education and occupational responsibility indicating that more highly educated individuals and those with more job responsibility would be more engaged in social and lifelong activities in later life with additional benefits in cognitive reserve.

Method

Participants

Centenarians and octogenarians (N = 321) from the Georgia Centenarians Study (GCS) were included in the study. In the original design of the study, nonagenarians were not included to draw a stronger contrast between centenarians and a younger age group. Proxy data provided the information about social engagement, occupational responsibility, activity, and past engaged lifestyle. Proxy information was included to reach a more representative sample of participants, because cognitively impaired individuals cannot always provide reliable information.

Centenarian proxies were selected following a general procedure: if spouses were alive, they were first selected as proxies; second, adult children were contacted to provide proxy reports. For participants who had more than one living child, proxies were nominated by the participants. The third possible group of proxies included other family members, and the final group included community members. Most of the proxies were family members (88.75%), but friends (5.1%) and neighbors (4.0%) were also selected in a few cases. Because proxies could have been in advanced age, all proxies were tested for cognitive functioning (i.e., MMSE), and cognitively impaired proxies were excluded from the study.

There were 239 centenarians and near-centenarians, and their mean age was 100.2 years (SD = 1.91, Range, 98–108). There were 82 octogenarians with a mean age of 84.4 (SD = 2.87, Range, 80–90). About 22% of the participants were male (n = 71), and 78% were female (n = 250, Table 1). The majority of the participants were White/Caucasian (76.3%). There were 76 Black/African Americans, or 23.7% of the total sample. The majority of participants (82.4%) were widowed, separated or divorced. The mean score of the Mini-Mental State Examination (MMSE, Folstein, Folstein, & McHugh, 1975) for all participants was 19.02.

Table 1.

Descriptive Results

F Percent M SD Range
Age 96.13 7.21 80.5–108.6
Gender
   Male 71 22.1
   Female 250 77.9
Ethnicity
   White/Caucasian 245 76.3
   Black/African American 76 23.7
Marital Status
   Never Married 12 3.8
   Widowed/Separated/Divorced 262 82.4
   Married/Partnered 44 13.8
Occupational Responsibility
   Low 23 9.5
   Medium 85 35.0
   High 135 55.6
Education (in years) 11.17 3.93 0–20
Past Engaged Lifestyle 4.20 1.90 0–8
Current Social Engagement 1.86 0.92 0–3
Activity 0.71 2.56 (−4)-4
MMSE 19.02 9.28 0–30
Cognitive Reserve Index (CRI) 0.61 0.18 0.13–0.96
Open in a new tab

Ethical approval for this study was obtained from the Institutional Review Board at Iowa State University, XX-026.

Measures

Cognitive reserve measures were chosen to closely mirror a similar index developed by Lavrencic et al. (2018). These variables include education, marital status, current social engagement, occupational responsibility, activity, and past engaged lifestyle. In addition, a measure of global cognitive functioning was included. All variables (except for the MMSE) were collected from proxy informants so that low functioning centenarians could be included.

Cognitive Functioning

Cognitive functioning was assessed with the Mini-Mental State Examination (MMSE, Folstein et al., 1975), which measures five aspects of cognitive functioning with 30 items: orientation, registration, attention and calculation, recall, and language. MMSE scores range from 0 to 30. Higher scores indicate higher cognitive functioning. The reliability for MMSE in this study was α = .88.

Education

Total years of education of participants was used as a measurement for educational attainment. Higher scores indicate more years of education.

Marital Status

Current marital status of participants was measured on six levels (1 = currently married, 2 = living with a partner, 3 = separated, 4 = divorced, 5 = widowed, 6 = never been married). The scale was recoded so that there were three levels of marital status: 0 = never married, 1 = no current partner, and 3 = currently having a partner.

Occupational Responsibility

Proxy data provided the degree of occupational responsibility (Hultsch, Hertzog, Small & Dixon, 1999). Proxies responded to the question, “What was the degree of responsibility he/she had in his/her major lifetime occupation?” Answers included three levels (1 = low, 2 = medium, 3 = high). Higher scores indicate higher levels of perceived occupational responsibility.

Activity

“Reduced activity” includes the sum of four subscale items drawn from the Multidimensional Fatigue Scale (Smets, Garssen, Boke, & Haes, 1995). Proxies were asked to respond to the following prompt: “The next questions are about the level of energy your relative has on any given day.” The four items relate to daily activity: “S/he feels very active,” “S/he thinks s/he does very little in a day.” “S/he thinks s/he does a lot in a day,” and “S/he gets little done.” (−1 = disagree, 0 = neutral, 1 = agree). Reduced activity items were reverse coded, so that higher scores indicate higher activity. The total score ranged from −4 to +4. Cronbach’s alpha for this scale was α = .86.

Current Social Engagement

For current social engagement, proxies indicated the number of times in the past week they had spent some time with someone (Fillenbaum, 1988). The item was coded as 0 = Not at all, 1 = Once per week, 2 = 2–6 times per week, and 3 = Once a day or more.

Past Engaged Lifestyle

Past engaged lifestyle was defined by a series of cognitive engagement tasks (Hultsch et al., 1999; Martin et al., 2009). We asked proxies whether the centenarians or octogenarians had ever been engaged in tasks, including learning a foreign language, going back to school for more education, doing volunteer work, traveling to a foreign country, preparing his/her own taxes, giving a public talk or lecture, and balancing his/her checkbook (taking care of finances). Proxies answered “yes” or “no” to these questions. The items were summarized to a combined score. The reliability of past engaged lifestyle was α = .68.

Cognitive Reserve

The cognitive reserve index was coded the same way Lavrencic et al. (2018) coded this variable. All variables were recoded between 0 (low reserve) and 1 (high reserve). The cognitive reserve index contained the following variables. Education was recoded into three groups (0–9 grade = 0; 10–11 = 0.5 and 12+ years = 1). Occupational responsibility was recoded into three groups (low = 0; medium = 0.5, and high = 1).

Physical activity was measured by reverse scores of the reduced activity fatigue subscale (2 to 4 = 0; −1 to 1 = 0.5, and −4 to −2 = 1). Marital status was defined as never married = 0, widowed/separated/divorced = 0.5, and married or partnered = 1. Current social engagement was defined with a variable from the OARS social resources (Fillenbaum, 1988, “How much time do you spend with someone,” 0 = not at all, 0.33 = once a week; 0.66 = 2–6 times/week; 1 = once a day or more). Engaged life style (e.g., volunteering, balancing a checkbook, public speaking, learning a foreign language) was used as the measure of mental activities throughout life (1 to 2 activities = .25; 3 to 4 activities = .50, 5 to 6 activities = .75 and 7 to 8 activities = 1). A summary scores was then computed and divided by the number of items, so that the total score could range from 0–1.

Analyses

First, descriptive analyses were computed to assess means and standard deviations for all variables. Second, a cognitive reserve summary score was computed, and mean group differences were analyzed for high and low functioning participants.

Finally, the cognitive reserve process model was analyzed using Mplus (Muthén, & Muthén, 1998–2017). We evaluated a structural model of cognitive reserve, hypothesizing that education predicted occupational responsibility, which in turn predicted past engaged lifestyle and current social engagement. Past engaged lifestyle and current social engagement in turn were hypothesized to predict activity levels, which are (together with other direct effects) hypothesized to predict cognitive function. Indirect effects were computed within Mplus. Because some measures were based on rather crude ordinal scales, we used robust estimators (MLR) for the SEM analyses.

Results

Table 1 summarizes the descriptive data of our variables. On average, participants had eleven years of education, and most participants indicated that they had at least a “medium” level of responsibility in their job. Participants were engaged in several lifestyle activities, and the average number visits was several times a week. In addition, participants were somewhat below average on activities.

To replicate Lavrencic et al.’s (2018) findings, we first computed the distribution of cognitive reserve index (CRI). The scores ranged from .13 (low reserve) to .96 (high reserve). The range of scores is somewhat larger than the one reported by Lavrencic et al. We conducted independent t-tests for two groups of cognitive functioning, and the group with high MMSE scores (defined as a score >= 24) also scored higher on the CRI (M = .70, SD = .14), t (179) = 7.13, p < .05 when compared to the group with low MMSE scores (M = .54, SD = .17).

Table 2 summarizes bivariate correlations among the cognitive reserve variables and the cognitive functioning scores. Results indicated moderate to strong correlations between cognitive functioning and all reserve variables, except for occupational responsibility. Past engaged lifestyle showed high associations with education and occupational responsibility, whereas current social engagement and activity were only moderately associated with other cognitive reserve variables. Furthermore, the association between cognitive functioning and the cognitive reserve index was highly significant, r(181) = .56, p < .001.

Table 2.

Bivariate Correlations Among Cognitive Reserve Variable Components

1 2 3 4 5 6 7 8
1. Marital Status (married) 1.0
2. Education .10 1.0
3. Occupational Responsibility .01 .35** 1.0
4. Past Engaged Lifestyle .15* .62** .42** 1.0
5. Current Social Engagement −.14* .12* .15* .21** 1.0
6. Activity .15* .09 .10 .15* .14* 1.0
7. Cognitive Reserve (CRI) .27** .71** .54** .67** .48** .55** 1.0
8. MMSE .22** .43** .16 .48** .24** .36** .56** 1.0
Open in a new tab
*

p < .05.

**

p < .01.

Figure 1 summarizes the structural model specifying pathways from education to cognitive functioning. The model fit well with the data, χ 2 (df=3) = 5.02, p = .17; CFI = .99, RMSEA = .05. The results of the model are displayed in Figure 2.

Figure 2.

Figure 2

Open in a new tab

Cognitive Reserve Process Model with Significant Pathways

Note. χ 2 (df=3) = 5.02, p = .17; CFI = .99, RMSEA = .046.

*p < .05. **p < .01. ***p < .001.

Cognitive functioning was significantly associated with education, current social engagement, activity, and past engaged lifestyle. Activity was associated with current social engagement, and past engaged lifestyle was associated with occupation and education. Finally, occupation was significantly associated with education. The four direct predictors (i.e., education, current social engagement, activity, and past engaged lifestyle) explained 35 percent of the variance in cognitive functioning.

A number of indirect (mediating) associations were assessed in Mplus (Table 3). Results reveal that the association of education with cognitive functioning was mediated through occupational responsibility and past engaged lifestyle, as well as through engaged lifestyle on cognition. A third mediation was obtained for education, past engaged lifestyle, and activity to cognition. The indirect association of current social engagement with cognition via activity was also significant.

Table 3.

Significant Indirect Pathways to Cognitive Functioning

Pathway Regression Coefficient t
Education→Occupation→Past Engaged Lifestyle→Cognition .02 2.83**
Education→Past Engaged Lifestyle→Cognition .15 3.31**
Education→Past Engaged Lifestyle→Activity→Cognition .02 2.11*
Current Social Engagement→Activity→Cognition .04 2.00*
Open in a new tab

Note. Regression coefficients are standardized regression coefficients.

*

p < .05.

**

p < .01.

***

p < .001.

Discussion

The purpose of our study was to investigate the significant direct and indirect pathways of cognitive reserve variables associated with cognitive functioning among oldest-old adults. The study is one of the first to “decompose” the cognitive reserve variable. Rather than using cognitive reserve as only one summary variable, this research developed a process model indicating how the variables of cognitive reserve, as life span accumulation, interrelate to predict different cognitive outcome levels in late life. The findings reveal that in addition to the level of education, more activity, current social engagement, and past engaged lifestyle directly predicted better cognitive functioning among oldest-old adults. The findings partially confirms our hypothesis because occupational responsibility was not significantly associated with cognitive functioning of oldest-old adults.

The fit of the cognitive reserve model confirms that our model fits well with the data, and the pathways we predicted from education to cognitive functioning through occupational responsibility, current social engagement, past engaged lifestyle and activity were significant. Our results suggest that educational attainment alone, and in combination with other cognitive reserve variables, are directly and indirectly related to cognitive function in very late life. The total effect of all indirect pathways was about the same size as the direct effect of education on cognitive function. This overall finding is consistent with other studies indicating that activity engagement directly and indirectly affect cognitive tests (Cadar et al., 2016; Harrison et al., 2015; Meng & D’Arcy, 2012; Tucker-Drob et al., 2009; Zhang et al., 2019). Opdebeeck et al., (2015) also reported that higher educational level, as well as cognitively stimulating activities were related to better performance on cognitive tests. However, compared to other studies, occupational responsibility was not directly associated with cognitive function in this sample. Perhaps occupational responsibility was not a distinguishing factor for this cohort of centenarians. Furthermore, the single item measure of occupational responsibility may not have been an appropriately scaled measure. Other more detailed occupational complexity measures may have yielded different results.

The major hypothesis of the study was that education in combination with other variables build up cognitive reserve over the life span. We hypothesized that education would affect occupational responsibility, which in turn would influence current social engagement, past engaged lifestyle, and activities that all would be components of cognitive reserve.

We noted a significant pathway from education via occupational responsibility, past engaged lifestyle to cognitive reserve. Although some of these reserve components were collected by proxies retrospectively, the results support a life-span model that includes important components building up over time. In other words, the participants who had earned more years of education had a higher likelihood of working with more responsibility, which perhaps promoted an engaged lifestyle. An engaged lifestyle then was positively associated with cognitive functioning in later adulthood. Additional significant pathways started at education with a direct effect on past engaged lifestyle to better cognition in late life. Finally, education also affected global cognitive functioning via past engaged lifestyle and activity.

The pathway from education and occupational responsibility through activity on cognitive functioning was not significant. The literature suggests that educational and occupational attainment and activity add to cognitive reserve (Cadar et al., 2016; Wang, 2002; Wang, Xu, & Pei, 2012; Stern, 2009), but it is unclear whether occupational responsibility particularly predicts higher levels of activity in very late life.

Current social engagement was directly associated with cognitive function, as well as a mediated effect through activity. This finding is consistent with earlier research that higher levels of current social engagement in later life promotes activity and is associated with better cognitive function (Krueger et al., 2009.

The results of this study suggest that the construct of cognitive reserve should be decomposed in its specific components, such as education, occupational responsibility, engagement, and activity. Furthermore, cognitive reserve components need to be placed in a life-span timeline, with education as an early and lasting influence, and lifelong engagement as distal variables combined with current social engagement and activity as proximal variables. This would be consistent with models that emphasize lifelong developmental adaptation (Martin & Martin, 2002).

There are limitations to the current study. The sample of our study involves a relatively small number of octogenarians and centenarians from a limited geographic area in the United States. Therefore, our findings may not generalize to the general population of older adults. Furthermore, our sample consisted mainly of women. All measures, except for the cognitive outcome measure, were obtained from proxy informants. Most of these proxies were the participants’ adult children who most likely knew about the occupational role their parents played and their education and lifestyle; nonetheless, there may have been a bias or inaccuracy when evaluating proxy variables or current cognitive reserve components.

The measures we included in the study, such as occupational responsibility, may not be the best measures in assessing the cognitive reserve model, considering that the degree of responsibility of early 20th century cohorts may not predict social engagement or activity level. The activity measure was also very general – focusing on daily general activities, rather than specific physical or mental activities. Past engaged lifestyle was used as a lifespan measure; proxies could have included lifestyle activities during any time of the life span. There may be quite a difference between someone learning a foreign language in early life versus someone who took learning a language in midlife. Similarly, there may be differences between individuals who travelled to another country as a child versus someone who travelled extensively throughout life.

Using the MMSE as the only measure of cognitive function in late life is also limiting. Future studies should include more comprehensive measures of cognitive function. The model should also be evaluated with different potential moderators, such as comparisons by gender or for different ethnic groups. Finally, the causal relationships between past engaged lifestyle and cognitive functioning is unclear because bidirectional effects are plausible and the cross-sectional design of the study does not allow causal inferences. Cognitive functioning may be the result of an engaged lifestyle, but higher levels of cognitive functioning may also lead to higher levels of engaged lifestyle. As others have pointed out, a decline in cognitive functioning may limit engagement in activities (Hultsch et al., 1999).

In spite of these limitations, our study revealed some noteworthy findings about cognitive reserve that may protect from cognitive decline or impairment. The results from our study support previous findings on cognitive reserve but also go beyond studies that only use cognitive reserve as a single summary construct and evaluating different cognitive reserve components in their temporal, life span order. Educational attainment, degree of occupational responsibility, activity, social engagement, and engaged lifestyle are important in order to stay cognitively healthy in very late life.

ACKNOWLEDGMENTS

The Georgia Centenarian Study was funded by 1P01AG17553 from the National Institute on Aging, a collaboration among The University of Georgia, Tulane University Health Sciences Center, Boston University, University of Kentucky, Emory University, Duke University, Wayne State University, Iowa State University, Temple University, and University of Michigan. Additional authors include S. M. Jazwinski, R. C. Green, M. MacDonald, M. Gearing, W. R. Markesbery (deceased), J. L. Woodard, M. A. Johnson, J. S. Tenover, I. C. Siegler, W. L. Rodgers, D. B. Hausman, C. Rott, A. Davey, and J. Arnold. Authors acknowledge the valuable recruitment and data acquisition effort by M. Burgess, K. Grier, E. Jackson, E. McCarthy, K. Shaw, L. Strong, and S. Reynolds, data acquisition team manager; S. Anderson, E. Cassidy, M. Janke, and J. Savla, data management; M. Poon, project fiscal management. The first author also acknowledges the support by the Fulbright Commission and by United States Department of Agriculture, Hatch Project Grant, IOW04116 for his work on this project.

References

  1. Blondell SJ, Hammersley-Mather R, & Veerman JL (2014). Does physical activity prevent cognitive decline and dementia? A systematic review and meta-analysis of longitudinal studies. BMC Public Health, 14, 510. doi: 10.1186/1471-2458-14-510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Cadar D, Stephan BCM, Jagger C, Johansson B, Hofer SM, Piccinin AM, & Muniz-Terrera G (2016). The role of cognitive reserve on terminal decline: a cross-cohort analysis from two European studies: OCTO-Twin, Sweden, and Newcastle 85+, UK: A cross-cohort analysis on education and terminal decline. International Journal of Geriatric Psychiatry, 31(6), 601–610. 10.1002/gps.4366 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Corrada MM, Brookmeyer R, Berlau D, Paganini-Hill A, & Kawas CH (2008). Prevalence of dementia after age 90: results from the 90+ study. Neurology, 71, 337–343. doi: 10.1212/01.wnl.0000310773.65918.cd [DOI] [PubMed] [Google Scholar]
  4. Fillenbaum GG (1988). Multidimensional functional assessment of older adults: The Duke Older Americans Resources and Services Procedures Hillsdale: Lawrence Erlbaum Associates. [Google Scholar]
  5. Folstein MF, Folstein SE and McHugh PR (1975), “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician, Journal of Psychiatric Research, 12(3), 189–198. [DOI] [PubMed] [Google Scholar]
  6. Goveas JS, Rapp SR, Hogan PE, Driscoll I, Tindle HA, Carson Smith J… Espeland MA (2016). Predictors of optimal cognitive aging in 80+ Women: The Women’s Health Initiative Memory Study. Journals of Gerontology, Series A: Biological Sciences and Medical Sciences, 71(suppl 1):S62–S71. doi: 10.1093/gerona/glv055 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Harrison SL, Sajjad A, Bramer WM, Ikram MA, Tiemeier H, & Stephan BCM (2015). Exploring strategies to operationalize cognitive reserve: A systematic review of reviews. Journal of Clinical and Experimental Neuropsychology, 37(3), 253–264. 10.1080/13803395.2014.1002759 [DOI] [PubMed] [Google Scholar]
  8. Hultsch DF, Hertzog C, Small BJ, & Dixon RA (1999). Use it or lose it: Engaged lifestyle as a buffer of cognitive aging? Psychology and Aging, 14, 245–263. [DOI] [PubMed] [Google Scholar]
  9. James B, Wilson R, Barnes L, & Bennett D (2011). Late-life social activity and cognitive decline in old age. Journal of the International Neuropsychological Society, 17(6), 998–1005. doi: 10.1017/S1355617711000531 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Kliegel M, Zimprich D, & Rott C (2004). Life-long intellectual activities mediate the predictive effect of early education on cognitive impairment in centenarians: a retrospective study. Aging and Mental Health, 8, 430–437. doi: 10.1080/13607860410001725072 [DOI] [PubMed] [Google Scholar]
  11. Krueger KR, Wilson RS, Kamenetsky JM, Barnes LL, Bienias JL, & Bennett DA (2009). Social engagement and cognitive function in old age. Experimental Aging Research, 35(1), 45–60. doi: 10.1080/03610730802545028 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Lavrencic LM, Richardson C, Harrison SL, Muniz-Terrera G, Keage HAD, Brittain K, … Stephan BCM (2018). Is There a Link Between Cognitive Reserve and Cognitive Function in the Oldest-Old? The Journals of Gerontology: Series A, 73(4), 499–505. 10.1093/gerona/glx140 [DOI] [PubMed] [Google Scholar]
  13. Marioni RE, Valenzuela MJ, van den Hout A, Brayne C, Matthews FE, & MRC Cognitive Function and Ageing Study. (2012). Active Cognitive Lifestyle Is Associated with Positive Cognitive Health Transitions and Compression of Morbidity from Age Sixty-Five. PLoS ONE, 7(12), e50940. 10.1371/journal.pone.0050940 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Martin P, Baenziger J, MacDonald M, Siegler I, & Poon LW (2009). Engaged lifestyle, personality, and mental status among centenarians. Journal of Adult Development, 16, 199–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Martin P, & Martin M (2002). Proximal and distal influences on development: The model of developmental adaptation. Developmental Review, 22, 78–96. [Google Scholar]
  16. Meng X, & D’Arcy C (2012). Education and Dementia in the Context of the Cognitive Reserve Hypothesis: A Systematic Review with Meta-Analyses and Qualitative Analyses. PLoS ONE, 7(6), e38268. 10.1371/journal.pone.0038268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Muniz-Terrera G, Matthews F, Dening T, Huppert FA, Brayne C, & CC75C Group. (2008). Education and trajectories of cognitive decline over 9 years in very old people: methods and risk analysis. Age and Ageing, 38(3), 277–282. 10.1093/ageing/afp004 [DOI] [PubMed] [Google Scholar]
  18. Muthén LK, & Muthén BO (1998–2017). Mplus user’s guide: Statistical analysis with latent variables (8th ed.). Los Angeles, CA: Muthén & Muthén. [Google Scholar]
  19. Opdebeeck C, Martyr A, & Clare L (2015). Cognitive reserve and cognitive function in healthy older people: a meta-analysis. Aging, Neuropsychology, and Cognition, 23(1), 40–60. doi: 10.1080/13825585.2015.1041450 [DOI] [PubMed] [Google Scholar]
  20. Richards M, James SN, Sizer A, Sharma N, Rawle M, Davis DH, & Kuh D (2019). Identifying the lifetime cognitive and socioeconomic antecedents of cognitive state: seven decades of follow-up in a British birth cohort study. BMJ open, 9(4), e024404. doi: 10.1136/bmjopen-2018-024404 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Smart EL, Gow AJ, & Deary IJ (2014). Occupational complexity and lifetime cognitive abilities. Neurology, 83, 2285–2291. doi: 10.1212/WNL.0000000000001075 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Smets EMA, Garssen B, Bonke BD, & De Haes JCJM (1995). The Multidimensional Fatigue Inventory (MFI) psychometric qualities of an instrument to assess fatigue. Journal of psychosomatic research, 39(3), 315–325. 10.1016/0022-3999(94)00125-O [DOI] [PubMed] [Google Scholar]
  23. Stern Y (2002). What is cognitive reserve? Theory and research application of the reserve concept. Journal of the International Neuropsychological Society, 8, 448–460. doi: 10.1017/S1355617702813248 [DOI] [PubMed] [Google Scholar]
  24. Stern Y (2009). Cognitive reserve. Neuropsychologia, 47(10), 2015–2028. 10.1016/j.neuropsychologia.2009.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Stern Y, Arenaza-Urquijo EM, Bartrés-Faz, Belleville S, Cantilon M, Chetelat G, Ewers M, Franzmeier N, Kempermann G, Kremen WS, Okonkwo O, Scarmeas N, Soldan A, Udeh-Momoh C, Valenzuela M, Vemuri P, Vuoksimaa E, and the Reserve, Resilience and Protective Factors PIA Empirical Definitions and Conceptual Frameworks Workgroup (2020). Whitepaper: Defining and investigating cognitive reserve, brain reserve, and brain maintenance. Alzheimer’s & Dementia, 16(9), 1305–1311. 10.1016/j.jalz.2018.07.219 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Tomioka K, Kurumatani N, & Hosoi H (2018). Social participation and cognitive decline among community-dwelling older adults: A community-based longitudinal study. Journal of Gerontology, Series B: Psychological Sciences and Social Sciences, 73, 799–806. doi: 10.1093/geronb/gbw059. [DOI] [PubMed] [Google Scholar]
  27. Tucker-Drob EM, Johnson KE, & Jones RN (2009). The cognitive reserve hypothesis: A longitudinal examination of age-associated declines in reasoning and processing speed. Developmental Psychology, 45(2), 431–446. 10.1037/a0014012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Wang H-X, Karp A, Winblad B, & Fratiglioni L (2002). Late-life engagement in social and leisure activities is associated with a decreased risk of dementia: a longitudinal study from the Kungsholmen project. American Journal of Epidemiology, 155, 1081–1087. [DOI] [PubMed] [Google Scholar]
  29. Wang H-X, Xu W, & Pei J-J (2012). Leisure activities, cognition and dementia. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 1822(3), 482–491. 10.1016/j.bbadis.2011.09.002 [DOI] [PubMed] [Google Scholar]
  30. Wilson RS, Hebert LE, Scherr PA, Barnes LL, Mendes de Leon CF, & Evans DA (2009). Educational attainment and cognitive decline in old age. Neurology, 72(5), 460–465. 6. doi: 10.1212/01.wnl.0000341782.71418.6c [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Yaffe K, Middleton LE, Lui LY, Spira AP, Stone K, Racine C, Ensrud KE, Kramer JH. (2011). Mild cognitive impairment, dementia, and their subtypes in oldest old women. Archives of Neurology, 68, 631–636. doi: 10.1001/archneurol.2011.82 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Zhang W, Tang F, Chen Y, Silverstein M, Liu S, & Dong X (2019). Education, activity engagement, and cognitive function in US Chinese older adults. Journal of the American Geriatrics Society, 67(S3), S525–S531. [DOI] [PubMed] [Google Scholar]

RESOURCES