Skip to main content
NCBI home page
The Journals of Gerontology Series B: Psychological Sciences and Social Sciences logoLink to The Journals of Gerontology Series B: Psychological Sciences and Social Sciences
. 2018 Apr 16;73(Suppl 1):S73–S81. doi: 10.1093/geronb/gbx167

Secular Trends in Cognitive Performance in Older Black and White U.S. Adults, 1993–2012: Findings From the Chicago Health and Aging Project

Jennifer Weuve 1,✉, Kumar B Rajan 2, Lisa L Barnes 3,4, Robert S Wilson 3,4, Denis A Evans 2
PMCID: PMC6019012  PMID: 29669103

Abstract

Objective

To characterize secular trends from 1993 to 2012 in cognitive performance using a cohort of older black and white U.S. adults, and compare trends by race.

Method

Our data come from 8,906 participants of the Chicago Health and Aging Project (CHAP), a longitudinal, population-based cohort (age ≥ 67, 60% black). Participants underwent cognitive assessments in six 3-year study cycles from 1993 to 1996 through 2010 to 2012. We computed 3 measures of cognitive performance: global cognition, episodic memory, and perceptual speed.

Results

Mean performance in terms of global cognitive score followed a secular pattern of modest decline over the 6 study cycles. The trend was most pronounced for perceptual speed. Mean scores among black participants were consistently lower than those for whites; these disparities in mean performance narrowed over time, especially on perceptual speed, but appeared to widen at the last cycle. Global scores among the upper quartile of performers rose slightly, but scores among the lowest quartile of performers dropped precipitously.

Discussion

Between 1993 and 2012, secular trends in cognitive performance in this established cohort did not follow a clear pattern of improvement, contrasting with previous research. But patterns differed by cognitive domain, performance level, and race.

Keywords: Cognition, Epidemiology, Health disparities, Race

Growing evidence indicates that the incidence rate of dementia may be declining (Derby, Katz, Lipton, & Hall, 2017; Noble et al., 2017; Wu et al., 2017), possibly as a by-product of improved cognitive function later in life among later birth cohorts (Baxendale, 2010; Dodge, Zhu, Lee, Chang, & Ganguli, 2014; Gerstorf, Ram, Hoppmann, Willis, & Schaie, 2011; Rönnlund & Nilsson, 2009; Sacuiu et al., 2010). These data offer hope that impaired cognition and dementia will impose less of a burden on future populations than previously forecasted.

However, these secular declines only partially reveal trends in the spectrum of cognitive functioning, and, thus far, most data have come from cohorts of white older adults living in high-income countries. Thus, the full public health impact of the decline remains in question.

Dementia trends offer a limited window into how older adults’ cognitive performance at a given age has changed. For example, whereas trends in dementia prevalence reveal changes in the percentage of older adults whose cognitive performance falls below a clinical threshold, analogous trends in the distribution of age-specific cognitive scores reveal the wider gamut of functioning, which reflects the range of independence and dependence in the population and has direct bearing on the degree of supports needed for older adults. It is notable, for example, that reported trends in dementia prevalence have not pointed downward as consistently as trends in dementia incidence, possibly due to lengthening lifespans (Wu et al., 2017). Secular trends in specific cognitive functions may also provide clues as to factors that may have shaped time trends in overall performance and dementia incidence. For example, improvements in age-specific episodic memory could reflect population-scale delays in accruing neuropathology related to Alzheimer’s disease and other dementias (Wilson et al., 2013). Improvements in other functions, such as perceptual speed, may signal improvements in early-life phenomena, such as access to and uptake of education (Diamond, 2013).

Furthermore, most dementia and cognitive trend studies consist of persons of European ancestry in high-income countries, so it is not clear that these trends are universal. Black Americans are underrepresented in this work, with few characterizations of trends in black Americans’ dementia incidence and prevalence, and fewer still of trends in their cognitive performance in older adulthood. This paucity of data is striking for several reasons. First, compared with white Americans, blacks have lower cognitive performance (Early et al., 2013) and about twice the risk of developing dementia (Steenland, Goldstein, Levey, & Wharton, 2016; Weuve et al., 2018), making it critical to understand forces affecting cognitive risk in black Americans, especially as the population of older black adults is forecasted to comprise a larger segment of the older adult population in coming decades (Ortman, Velkoff, Hogan, & Census, 2014). Second, although a limited set of data has suggested a decline in dementia incidence among black Americans (Gao et al., 2016; Noble et al., 2017), it is unknown whether there are racial disparities in cognitive trends—for example, whether improving cognitive performance observed in older Europeans and white Americans (Baxendale, 2010; Dodge et al., 2014; Gerstorf et al., 2011; Rönnlund & Nilsson, 2009; Sacuiu et al., 2010) is also occurring in black Americans. Different trends in cognitive performance may point to modifiable factors, especially those operating at the population level, that have differentially shaped those trends. Education, a putative protective factor, is one example: gains in high school completion among blacks lag behind gains among whites by two decades (U.S. Census Bureau—Education and Social Stratification Branch, 2016).

To address these critical gaps, we analyzed data from the longitudinal, population-based Chicago Health and Aging Project (CHAP), a cohort comprising more than 10,000 older adults, 63% of whom self-identified as black and 37% as white, with ongoing enrollment and follow-up spanning nearly two decades. We used data from CHAP to estimate age- and sex-adjusted secular trends in performance according to three measures of cognitive function, and to compare these trends by race. With these analyses, which we standardized to the characteristics of the CHAP population, we sought to answer the question: given a population of a set composition in age, sex, and race, has the cognitive performance of that population changed over time? And does performance among older blacks and older whites follow the same secular trend?

Methods

Study Population

The Chicago Health and Aging Project (CHAP) began in 1993 with a census of individuals aged 65 years or older living in a geographically defined region on the south side of Chicago, Illinois (Bienias, Beckett, Bennett, Wilson, & Evans, 2003; Evans et al., 2003). Of those identified, 6,158 persons (79%) participated in a home interview. Beginning in the 2000–2002 cycle, additional people enrolled as they turned age 65 years, for a total of 10,802 participants through 2012. At each cycle, we restricted analyses to data from participants who were at least 67 years old, the oldest minimum age across the cycles. In total, 8,906 participants had the necessary data to be included in the present study. Participants were re-interviewed in 3-year cycles. Each data collection cycle consisted of in-home interviews of all participants. CHAP achieved follow-up assessments of about 89% of surviving participants. The Rush University Medical Center Institutional Review Board approved the study.

Assessment of Cognitive Function and Other Variables

Cognitive function

During their home interviews, all CHAP participants underwent a brief cognitive assessment comprising four tests of functions that typically decline in Alzheimer’s disease dementia. The Symbol Digit Modalities Test (Smith, 1982) measures perceptual speed, a component of executive function; the East Boston Memory Test (Albert et al., 1991) generates measures of both immediate and delayed episodic memory; and the Mini-Mental State Examination (Folstein, Folstein, & McHugh, 1975) measures several cognitive functions, including orientation, memory, language, and visual construction. For each of the four tests, we transformed the raw scores to z scores, using the baseline raw scores as the source of the standard deviation. Our analyses concerned three cognitive measures, all scaled to standard normal distribution to enhance comparisons across tests. The first, the global cognitive score, was created by first averaging the z scores from all four tests (composite z score) and then converting the resulting score to standard normal, using the baseline composite z score’s mean and standard deviation (Wilson et al., 2003, 2005). The second, the episodic memory score, was the average of the z scores from the two components of the East Boston Memory Test, which we further transformed to standard normal as done for the global score. The third measure, the perceptual speed score, was the z score from the Symbol Digit Modalities test.

Race

Participants self-reported their race using categories from the 1990 US Census.

Statistical Analyses

Imputation of cognitive scores

In cognitive studies of older adults, participants who are severely ill, perhaps near death, are less likely to undergo assessment. Those who do typically have poor cognitive scores. The missing assessments from such people may result in overestimation of the mean score. For participants who died during follow-up, we imputed their cognitive scores, using a previously developed method (Wilson et al., 2017), for all missing visit cycles after their last visit, through and including the cycle during which they died (<1% in 1993–1996 and 15–20% in the remaining cycles). Although this type of missingness did not vary systematically over time, it was more common among white participants than black participants from the 2001–2003 cycle onward. In our data, cognitive scores were more strongly associated with mortality than with age, so we used a mortality-based terminal decline approach to estimate missing scores, using previously described random effects change point models for composite cognitive test scores and individual tests of episodic memory and perceptual speed (Wilson et al., 2017). All of the analyses herein include directly measured test scores as well as the imputed scores. (Supplementary table 1).

Trends in mean cognitive performance

We adopted a nonparametric approach to characterize trends in average cognitive performance, overall, and by race, conducting separate analyses for each cognitive outcome. We fitted a generalized estimating equations regression model of the cognitive score in which the key independent covariates were indicator variables for each of the five study cycles following the 1993–1996 study cycle, which served as the referent period. Each model also included terms for age at assessment, sex, and, for the analyses of the total population, race. The models also included cohort indicators to account for original and successive cohort membership in the study population. To account for correlation between some participants’ repeated cognitive test measurements, we used a robust sandwich variance estimator based on a generalized estimating equations approach with an exchangeable working covariance matrix (Zeger & Liang, 1986). Using this approach, we estimated age- and sex-adjusted marginal mean cognitive test scores—global cognitive function, episodic memory, and perceptual speed—at the six data collection cycles. The sandwich covariance matrix was used to estimate the 95% confidence intervals around the mean values. The marginal mean for the cognitive tests were standardized to the characteristics of the entire study population rather than to time-specific characteristics, since the distribution of age, sex, and race showed variability over the years.

Trends in highest and lowest levels of cognitive performance

To explore shifts in the wider distribution of cognition in the population, we estimated the mean global cognitive scores of participants above the highest quartile (75th percentile) and below the lowest quartile (25th percentile) of performance. We estimated these means for each study cycle and race group, using the adjustments previously described.

Predicted dementia threshold

Performance on cognitive tests is integral to the diagnosis of dementia, including Alzheimer’s disease dementia (AD). We used the distribution of cognitive scores in the CHAP population to trace how well over time the population was performing relative to the AD threshold. In CHAP, a stratified Bernoulli sample of participants underwent detailed clinical diagnosis for dementia at each cycle, described previously in detail (Bienias, Beckett, et al., 2003; Bienias, Kott, Beck, & Evans, 2005; Bienias, Kott, & Evans, 2003; Evans et al., 2003; Weuve et al., 2018). Among those who were diagnosed with AD, black and white participants had similar global cognitive scores at the time they were diagnosed (Rajan et al., 2017). We gauged cognitive performance in the CHAP population relative to the level of performance at AD diagnosis: specifically, using a normal probability distribution function, we estimated the probability that each time-specific marginal mean global cognitive score was less than the mean global cognitive score at AD diagnosis. A larger probability corresponds to a larger proportion of the population scoring below the AD threshold.

Sensitivity and exploratory analyses

We conducted additional analyses that were not restricted by age and thereby included data from participants who were aged 65–66 at the time of testing (i.e., from all but the second cycle). Second, as a preliminary exploration into the influence of secular changes in education on our findings, we repeated our primary analyses by further adjusting for years of education.

All statistical analyses were created using R software (R Core Team, 2016).

Results

Table 1 summarizes the characteristics of CHAP participants at their baseline assessments; Supplementary Table 2 shows the age and education distributions of participants at each study cycle.

Table 1.

Characteristics of the CHAP Study Population at Enrollment

All participants (N = 8,906) Black participants (N = 5,384) White participants (N = 3,522)
Age, years, mean (SD) 75.0 (6.9) 73.6 (6.2) 77.0 (7.3)
Female, N (%) 5,501 (62%) 3,313 (62%) 2,188 (62%)
Education, years, mean (SD) 12.1 (3.6) 11.1 (3.5) 13.5 (3.2)
Global cognition, mean (SD) 0.084 (.866) −0.095 (.870) 0.358 (.780)
Episodic memory,a mean (SD) 0.110 (.946) −0.032 (.973) 0.327 (.859)
Perceptual speed,a mean (SD) 0.129 (.976) −0.163 (.914) 0.578 (.890)
MMSE,a mean (SD) 0.083 (.917) −0.066 (.979) 0.310 (.762)
Open in a new tab

aStandard normalized scores. For all scores, a higher score indicates better performance.

Table 2.

Age- and Sex-Adjusted Standardized Scores (and 95% Confidence Intervals) on Composite (Global) Cognition, Episodic Memory, and Perceptual Speed in Chicago Health and Aging Project, 1993 to 2012

Study cycle All participants Black participants White participants
Global cognition
 1993–1996 0.001 (−0.018, 0.020) −0.175 (−0.200, −0.150) 0.314 (0.286, 0.342)
 1997–2000 −0.120 (−0.142, −0.099) −0.291 (−0.317, −0.264) 0.183 (0.149, 0.216)
 2001–2003 −0.101 (−0.128, −0.073) −0.241 (−0.272, −0.209) 0.149 (0.110, 0.188)
 2004–2006 −0.094 (−0.127, −0.060) −0.236 (−0.273, −0.199) 0.160 (0.114, 0.205)
 2007–2009 −0.142 (−0.181, −0.103) −0.271 (−0.312, −0.230) 0.088 (0.036, 0.140)
 2010–2012 −0.179 (−0.222, −0.135) −0.327 (−0.372, −0.281) 0.084 (0.027, 0.141)
Episodic memory
 1993–1996 0.019 (−0.003, 0.041) −0.120 (−0.150, −0.090) 0.266 (0.233, 0.299)
 1997–2000 −0.091 (−0.115, −0.068) −0.228 (−0.258, −0.197) 0.151 (0.114, 0.187)
 2001–2003 −0.031 (−0.061, −0.001) −0.139 (−0.174, −0.104) 0.161 (0.118, 0.204)
 2004–2006 0.019 (−0.018, 0.056) −0.106 (−0.146, −0.065) 0.240 (0.191, 0.290)
 2007–2009 −0.001 (−0.043, 0.042) −0.112 (−0.157, −0.067) 0.198 (0.141, 0.254)
 2010–2012 −0.046 (−0.094, 0.002) −0.168 (−0.218, −0.117) 0.171 (0.109, 0.232)
Perceptual speed
 1993–1996 −0.007 (−0.027, 0.012) −0.311 (−0.336, −0.285) 0.532 (0.502, 0.562)
 1997–2000 −0.074 (−0.094, −0.053) −0.356 (−0.381, −0.331) 0.428 (0.396, 0.461)
 2001–2003 −0.050 (−0.076, −0.024) −0.279 (−0.308, −0.251) 0.357 (0.317, 0.396)
 2004–2006 −0.078 (−0.111, −0.046) −0.265 (−0.299, −0.231) 0.254 (0.206, 0.302)
 2007–2009 −0.138 (−0.174, −0.103) −0.300 (−0.336, −0.265) 0.150 (0.100, 0.200)
 2010–2012 −0.212 (−0.252, −0.172) −0.433 (−0.473, −0.392) 0.179 (0.125, 0.234)
Open in a new tab

Trends in Mean Cognitive Performance

The average age- and sex-standardized global cognitive score among both black and white participants fell over the six study cycles, by 0.15 standard deviation (SD) units among blacks and 0.23 SD units among whites (pdifference in race-specific trends < .0001; Figure 1). Over the last five cycles (i.e., excluding the first), mean global scores were consistent by comparison, ultimately declining 0.10 SD units among whites and 0.04 unit among blacks. Secular trends in mean episodic memory score, a component of the global score, were more stable after an initial decline from the 1993–1996 cycle, with performance among both black and white participants increasingly slightly (by 0.06 and 0.02 SD unit, respectively) over the last five cycles.

Figure 1.

Figure 1.

Open in a new tab

Age- and sex-adjusted mean cognitive scores (95% confidence interval), among black and white participants, 1993–2012.

Perceptual speed scores followed different secular trends. Age- and sex-standardized mean performance among black participants rose slightly before dropping over the 2004–2012 stretch by about 0.17 SD units. By contrast, white participants’ mean performance followed a more pronounced drop over the first five cycles, by 0.4 SD units, before leveling off.

Trends in the Upper and Lower Distribution of Cognitive Performance

Overall, the mean global score among the top quartile of performers at any given cycle rose followed a small increase over time. From the second to the last cycle, this mean rose by 0.05 SD units and 0.08 SD units among black and white participants, respectively (Figure 2). By contrast, the mean global scores in the lowest quartile fell dramatically over the six cycles, by 0.5 SD units and 0.8 SD units among black and white participants, respectively.

Figure 2.

Figure 2.

Open in a new tab

Age- and sex-adjusted mean global cognitive scores (95% confidence interval) in the upper and lower quartiles of performers at each cycle, among black and white participants.

Disparities in Performance

Mean cognitive performance as gauged by all three cognitive scores, especially perceptual speed, was consistently lower among black participants than among white participants (Figure 1; Supplementary Figure 1). On the global and episodic memory scores, disparities in mean performance narrowed slightly over time, for example, from 0.5 to 0.4 SD unit on the global score (Figure 3). The disparity in mean performance on perceptual speed narrowed more considerably, from 0.8 SD unit in the first cycle to 0.4 SD unit in the fifth cycle. From the fifth to the last cycle, mean performance disparities on all scores, especially perceptual speed, widened. Within the upper quartile of performers on the global score, the racial disparity in mean scores remained small (<0.1 SD unit), increasing slightly over the last two cycles. The performance disparity within the lower quartile shrank notably, from 0.3 to 0.02 SD units, largely because lowest-quartile performance among white participants dropped more rapidly.

Figure 3.

Figure 3.

Open in a new tab

Difference between the age- and sex-adjusted mean global cognitive scores (95% confidence interval) among black and white participants, 1993–2012.

Cognitive Performance Relative to the Dementia Threshold

The estimated probability that the marginal mean global score was less than the mean global score among persons diagnosed with AD mirrored the secular trends in global cognitive performance. Over the six study cycles, this dementia threshold probability increased by 8 and 6 percentage points, respectively, among black and white participants (Figure 4).

Figure 4.

Figure 4.

Open in a new tab

Estimated probability (95% confidence interval) that the mean global cognitive score dementia threshold was less than the score of persons at the time of Alzheimer’s disease diagnosis, by race, between 1993 and 2012.

Sensitivity and Exploratory Analyses

Including data from younger participants (aged 65–66) from the cycles that involved new enrollees showed a steeper drop in mean global scores from the first to the second cycle (Supplementary Table 3). Thereafter, mean scores among white participants remained stable, as opposed to declining slightly in our primary analyses. The trend among black participants was similar to that in our primary analyses.

In analyses of secular trends in mean global cognitive score, further adjustment for years of education changed findings very little (Supplementary Figure 2).

Discussion

In this established cohort, with data from nearly 9,000 black and white older residents of a large U.S. Midwestern city, age- and sex-standardized trends in cognitive performance and dementia prevalence did not follow a clear secular trend of improvement in the 19-year period spanning 1993–2012 and including six assessment cycles. Beneath the trends in mean global cognition, which followed a modest downward arc, were disparate or even countervailing secular trends according to cognitive domain—with steeper drops in perceptual speed than in episodic memory, especially among whites—and according to segment of the performance distribution—with decreases in performance among those in the lowest performers and slight improvements among the top performers. Moreover, although racial disparities in mean cognitive performance narrowed over the first five study cycles, a trend most notable for perceptual speed, there is some suggestion that they widened between the last two cycles.

These findings stand in contrast to trends observed in numerous other settings. Age-adjusted cognitive test performance improved over spans ranging from six to 30 years in study populations in Sweden (Finkel, Reynolds, McArdle, & Pedersen, 2007; Rönnlund & Nilsson, 2008, 2009; Sacuiu et al., 2010), Denmark (Christensen et al., 2013), Germany (Steiber, 2015), England (Baxendale, 2010; Skirbekk, Stonawski, Bonsang, & Staudinger, 2013), and the United States (Dodge et al., 2014; Gerstorf et al., 2011). These studies documented improvements in a range of functional domains, including learning and memory, executive function, visuospatial abilities, and perceptual speed. Reports of stable or diminishing performance have been less common and, in several instances, entailed a single domain out of two or more evaluated (Baxendale, 2010; Choi, Schoeni, Martin, & Langa, 2016; Engberg, Christensen, Andersen-Ranberg, & Jeune, 2008; Finkel et al., 2007; Gerstorf et al., 2011). The brief cognitive battery in CHAP has been validated against a more extensive battery and strongly predicts subsequent risk of dementia (Rajan, Wilson, Weuve, Barnes, & Evans, 2015). It taps domains similar to those tapped in previous studies, except for visuospatial ability. Studies that used tests were nearly identical to those used in CHAP reported mixed findings (Choi et al., 2016; Christensen et al., 2013; Engberg et al., 2008; Finkel et al., 2007; Steiber, 2015).

Although previous findings are generally consistent with beneficial effects on cognitive performance of more recent birth cohort, period, or both, they depict the experience of European and white U.S. older adults. Of the three U.S. studies, two were set in small regions that might not generalize to the whole population (Dodge et al., 2014; Gerstorf et al., 2011). Our study is among the first to chart trends in cognitive performance among U.S. black older adults and compare them with trends among older white adults. In the absence of comparable estimates in the literature, we look to reports on trends in the prevalence of dementia and cognitive impairment among black older adults. Results from three such studies lend credence to the possibility that the improving trend in older adult cognitive performance may not be universal. Among a cohort of older Black adults living in Indianapolis, dementia prevalence remained stable between 1992 and 2001 (Rocca et al., 2011). In data from the National Health and Aging Trends Study, dementia prevalence among black participants fell from 2011 to 2014 (from 19.3% to 14.6%) before increasing slightly in 2015 (15.4%) (Freedman, Kasper, Spillman, & Plassman, 2017). By comparison, during the same period, the prevalence among white participants fell progressively. Finally, in a study of the younger participants (aged 55–69 years) of the Health and Retirement Study, the prevalence of cognitive limitations in non-Hispanic black participants did not change between 1998 and 2014, whereas the prevalence among non-Hispanic white participants dropped by about 0.8% per year (Choi et al., 2016).

Our investigation draws several strengths in estimation and precision from features of CHAP. CHAP is one of the largest study cohorts of black and white older adults in the United States, with 5,384 black and 3,522 white participants. Participation in CHAP has been high, from enrollment throughout follow-up. We have adopted several measures—refined adjustment for age, adjustment for enrollment cohort, imputing scores—to minimize the influence of CHAP’s study design on these estimated trends. In particular, our imputation method allowed us to assign cognitive scores to participants who missed assessments prior to dying, providing a more comprehensive capture of cognitive performance in the population of older adults. The multiple assessment waves allowed us to articulate variation in performance over time—including periods of improvement, decline, and stability—along with variation over time in racial disparities in performance. A comparison of performance at two time points would not capture these phenomena. Moreover, the analysis of trends in the upper and lower quartiles of performance revealed discrepant trends among those with higher versus lower levels of cognition.

Limitations of this study also merit attention. CHAP represents a specific population of older adults living in a large Midwestern U.S. city, and trends in this population might be might be unique to this population. Nonetheless, within strata of race, the CHAP population is roughly comparable to the population of older U.S. adults in terms of educational attainment (U.S. Census Bureau, 2014), smoking status (Federal Interagency Forum on Aging-Related Statistics, 2012), and the prevalence of chronic conditions, such as diabetes and hypertension (Federal Interagency Forum on Aging-Related Statistics, 2012; Mozaffarian et al., 2015). Moreover, the trends in CHAP highlight the importance of identifying populations in which trends are less favorable. These “outliers” offer signals for targeting services and for identifying etiologic factors that have shifted the epidemic in the population. Second, we have limited this investigation to articulating cognitive trends over time, leaving explanations for those trends to future exploration. Over time, the CHAP population was composed of more participants with higher levels of education, suggesting that cognitive performance, especially perceptual speed (Diamond, 2013), might improve over time as a result. In exploratory analyses, trends in mean global score did not change remarkably when we adjusted for years of education. However, it is possible that education’s influence on these trends: is more pronounced in specific parts of the performance distribution, operates in a way that depends on reaching specific milestones (e.g., high school, college), or even varies by race (Weuve et al., 2018).

Conclusion

Using data from a large cohort of older black and white adults in the United States, secular trends over an 18-year span in cognitive performance varied by cognitive domain, performance level itself, and race. On the whole, these results stand in contrast to the upward trends observed in investigations of European and other U.S. white populations. The variable cognitive trends among CHAP’s black participants, in particular, echo findings on dementia prevalence from recent studies that have large numbers of black participants. Collectively, these observations indicate that monitoring cognitive and dementia trends in all segments of the population is fundamental to understanding the full burden of cognitive aging on society.

Supplementary Material

Supplementary data are available at The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences online.

Supplementary Data
Click here for additional data file. (56.4KB, docx)

Funding

This work was supported by the National Institute on Aging at the National Institutes of Health (R01AG11101, R01AG051635, R01AG057532, and P30AG012846).

Conflict of Interest

The authors declare that they have no conflicts of interest with the submitted work.

References

  1. Albert M., Smith L. A., Scherr P. A., Taylor J. O., Evans D. A., & Funkenstein H. H (1991). Use of brief cognitive tests to identify individuals in the community with clinically diagnosed Alzheimer’s disease. The International Journal of Neuroscience, 57, 167–178. [DOI] [PubMed] [Google Scholar]
  2. Baxendale S. (2010). The Flynn effect and memory function. Journal of Clinical and Experimental Neuropsychology, 32, 699–703. doi:10.1080/13803390903493515 [DOI] [PubMed] [Google Scholar]
  3. Bienias J. L., Beckett L. A., Bennett D. A., Wilson R. S., & Evans D. A (2003). Design of the Chicago Health and Aging Project (CHAP). Journal of Alzheimer’s Disease, 5, 349–355. [DOI] [PubMed] [Google Scholar]
  4. Bienias J. L., Kott P. S., Beck T. L., & Evans D. A (2005). Incorporating multiple observations into logistic regression models of incident disease. Proceedings of the Annual Meeting of the American Statistical Association—Section on Survey Research Methods [CD-ROM], 2767–2774. [Google Scholar]
  5. Bienias J. L., Kott P. S., & Evans D. A (2003). Application of the delete-a-group jackknife variance estimator to analyses of data from a complex longitudinal survey. Proceedings of the Annual Meeting of the American Statistical Association—Section on Survey Research Methods [CD-ROM], 539–544. [Google Scholar]
  6. Choi H., Schoeni R. F., Martin L. G., & Langa K. M (2016). Trends in the prevalence and disparity in cognitive limitations of Americans 55–69 years old. Journals of Gerontology: Social Sciences (this issue). [DOI] [PMC free article] [PubMed]
  7. Christensen K., Thinggaard M., Oksuzyan A., Steenstrup T., Andersen-Ranberg K., Jeune B.,…Vaupel J. W (2013). Physical and cognitive functioning of people older than 90 years: A comparison of two Danish cohorts born 10 years apart. Lancet (London, England), 382, 1507–1513. doi:10.1016/S0140-6736(13)60777-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Derby C. A., Katz M. J., Lipton R. B., & Hall C. B (2017). Trends in dementia incidence in a birth cohort analysis of the Einstein Aging Study. JAMA Neurology, 74, 1345–1351. doi:10.1001/jamaneurol.2017.1964 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Diamond A. (2013). Executive functions. Annual Review of Psychology, 64, 135–168. doi:10.1146/annurev-psych-113011- 143750 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dodge H. H., Zhu J., Lee C. W., Chang C. C., & Ganguli M (2014). Cohort effects in age-associated cognitive trajectories. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences, 69, 687–694. doi:10.1093/gerona/glt181 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Early D. R., Widaman K. F., Harvey D., Beckett L., Park L. Q., Farias S. T.,…Mungas D (2013). Demographic predictors of cognitive change in ethnically diverse older persons. Psychology and Aging, 28, 633–645. doi:10.1037/a0031645 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Engberg H., Christensen K., Andersen-Ranberg K., & Jeune B (2008). Cohort changes in cognitive function among Danish centenarians. A comparative study of 2 birth cohorts born in 1895 and 1905. Dementia and Geriatric Cognitive Disorders, 26, 153–160. doi:10.1159/000149819 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Evans D. A., Bennett D. A., Wilson R. S., Bienias J. L., Morris M. C., Scherr P. A.,…Schneider J (2003). Incidence of Alzheimer disease in a biracial urban community: relation to apolipoprotein E allele status. Archives of Neurology, 60, 185–189. doi:noc10242 [pii] [DOI] [PubMed] [Google Scholar]
  14. Federal Interagency Forum on Aging-Related Statistics.(2012). Older Americans 2012: Key Indicators of Well-Being. Washington, DC: U.S. Government Printing Office: Retrieved from: https://agingstats.gov/docs/PastReports/2012/OA2012.pdf [Google Scholar]
  15. Finkel D., Reynolds C. A., McArdle J. J., & Pedersen N. L (2007). Cohort differences in trajectories of cognitive aging. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 62, P286–P294. [DOI] [PubMed] [Google Scholar]
  16. Folstein M. F., Folstein S. E., & McHugh P. R (1975). “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189–198. [DOI] [PubMed] [Google Scholar]
  17. Freedman V. A., Kasper J. D., Spillman B. C., & Plassman B. L (2017). Short‐term changes in the prevalence and incidence of probable dementia: An analysis of the 2011–2015 National Health and Aging Trends Study. Journals of Gerontology: Social Sciences (this issue). [DOI] [PMC free article] [PubMed]
  18. Gao S., Ogunniyi A., Hall K. S., Baiyewu O., Unverzagt F. W., Lane K. A.,…Hendrie H. C (2016). Dementia incidence declined in African-Americans but not in Yoruba. Alzheimer’s & Dementia, 12, 244–251. doi:10.1016/j.jalz.2015.06.1894 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gerstorf D., Ram N., Hoppmann C., Willis S. L., & Schaie K. W (2011). Cohort differences in cognitive aging and terminal decline in the Seattle Longitudinal Study. Developmental Psychology, 47, 1026–1041. doi:10.1037/a0023426 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Mozaffarian D., Benjamin E. J., Go A. S., Arnett D. K., Blaha M. J., Cushman M.,…Turner M. B; American Heart Association Statistics Committee and Stroke Statistics Subcommittee (2015). Heart disease and stroke statistics—2015 update: A report from the American Heart Association. Circulation, 131, e29–e322. doi:10.1161/CIR.0000000000000152 [DOI] [PubMed] [Google Scholar]
  21. Noble J. M., Schupf N., Manly J. J., Andrews H., Tang M. X., & Mayeux R (2017). Secular trends in the incidence of dementia in a multi-ethnic community. Journal of Alzheimer’s Disease, 60, 1065–1075. doi:10.3233/JAD-170300 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Ortman J. M., Velkoff V. A., & Hogan H(2014). An aging nation: The older population in the United States, Current Population Reports, P95/01-1. Washington, DC: U.S. Census Bureau. [Google Scholar]
  23. R Core Team (2016). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. [Google Scholar]
  24. Rajan K. B., Wilson R. S., Barnes L. L., Aggarwal N. T., Weuve J., & Evans D. A (2017). A cognitive turning point in development of clinical Alzheimer’s disease dementia and mild cognitive impairment: A biracial population study. The Journals of Gerontology, Series A: Biological Sciences and Medical Sciences, 72, 424–430. doi:10.1093/gerona/glw246 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rajan K. B., Wilson R. S., Weuve J., Barnes L. L., & Evans D. A (2015). Cognitive impairment 18 years before clinical diagnosis of Alzheimer disease dementia. Neurology, 85, 898–904. doi:10.1212/WNL.0000000000001774 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Rocca W. A., Petersen R. C., Knopman D. S., Hebert L. E., Evans D. A., Hall K. S.,…White L. R (2011). Trends in the incidence and prevalence of Alzheimer’s disease, dementia, and cognitive impairment in the United States. Alzheimer’s & Dementia, 7, 80–93. doi:10.1016/j.jalz.2010.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Rönnlund M., & Nilsson L.-G (2008). The magnitude, generality, and determinants of Flynn effects on forms of declarative memory and visuospatial ability: Time-sequential analyses of data from a Swedish cohort study. Intelligence, 36, 192–209. doi:10.1016/j.intell.2007.05.002 [Google Scholar]
  28. Rönnlund M., & Nilsson L. G (2009). Flynn effects on sub-factors of episodic and semantic memory: Parallel gains over time and the same set of determining factors. Neuropsychologia, 47, 2174–2180. doi:10.1016/j.neuropsychologia.2008.11.007 [DOI] [PubMed] [Google Scholar]
  29. Sacuiu S., Gustafson D., Sjögren M., Guo X., Ostling S., Johansson B., & Skoog I (2010). Secular changes in cognitive predictors of dementia and mortality in 70-year-olds. Neurology, 75, 779–785. doi:10.1212/WNL.0b013e3181f0737c [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Skirbekk V., Stonawski M., Bonsang E., & Staudinger U. M (2013). The Flynn effect and population aging. Intelligence, 41, 169–177. doi:10.1016/j.intell.2013.02.001 [Google Scholar]
  31. Smith A. (1982). Symbol Digit Modalities Test Manual—Revised. Los Angeles: Western Psychological Services. [Google Scholar]
  32. Steenland K., Goldstein F. C., Levey A., & Wharton W (2016). A meta-analysis of alzheimer’s disease incidence and prevalence comparing African-Americans and Caucasians. Journal of Alzheimer’s Disease, 50, 71–76. doi:10.3233/JAD-150778 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Steiber N. (2015). Population aging at cross-roads: Diverging secular trends in average cognitive functioning and physical health in the older population of Germany. PLoS One, 10, e0136583. doi:10.1371/journal.pone.0136583 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. U.S. Census Bureau.(2014). P23-212, 65+ in the United States: 2010. Washington, DC: U.S. Government Printing Office. [Google Scholar]
  35. U.S. Census Bureau—Education and Social Stratification Branch.(2016). Table A-2. Percent of people 25 years and over who have completed high school or college, by race, Hispanic origin and sex: Selected years 1940 to 2015 Retrieved from https://www.census.gov/hhes/socdemo/education/data/cps/historical/
  36. Weuve J., Barnes L. L., Mendes de Leon C. F., Rajan K. B., Beck T., Aggarwal N. T., … Evans D. A (2018). Cognitive aging in black and white Americans: Cognition, cognitive decline, and incidence of Alzheimer disease dementia. Epidemiology, 29(1), 151–159. doi:10.1097/EDE.0000000000000747 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wilson R. S., Bennett D. A., Bienias J. L., Mendes de Leon C. F., Morris M. C., & Evans D. A (2003). Cognitive activity and cognitive decline in a biracial community population. Neurology, 61, 812–816. [DOI] [PubMed] [Google Scholar]
  38. Wilson R. S., Bennett D. A., Mendes de Leon C. F., Bienias J. L., Morris M. C., & Evans D. A (2005). Distress proneness and cognitive decline in a population of older persons. Psychoneuroendocrinology, 30, 11–17. doi:10.1016/j.psyneuen.2004.04.005 [DOI] [PubMed] [Google Scholar]
  39. Wilson R. S., Rajan K. B., Barnes L. L., Jansen W., Amofa P., Weuve J., & Evans D. A (2017). Terminal decline of episodic memory and perceptual speed in a biracial population. Neuropsychology, Development, and Cognition, Section B Aging, Neuropsychology and Cognition, 1–12. Epub ahead of print. doi:10.1080/13825585.2017.1306020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Wilson R. S., Yu L., Trojanowski J. Q., Chen E. Y., Boyle P. A., Bennett D. A., & Schneider J. A (2013). TDP-43 pathology, cognitive decline, and dementia in old age. JAMA Neurology, 70, 1418–1424. doi:10.1001/jamaneurol.2013.3961 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Wu Y. T., Beiser A. S., Breteler M. M. B., Fratiglioni L., Helmer C., Hendrie H. C.,…Brayne C (2017). The changing prevalence and incidence of dementia over time—Current evidence. Nature Reviews. Neurology, 13, 327–339. doi:10.1038/nrneurol.2017.63 [DOI] [PubMed] [Google Scholar]
  42. Zeger S. L., & Liang K. Y (1986). Longitudinal data analysis for discrete and continuous outcomes. Biometrics, 42, 121–130. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
Click here for additional data file. (56.4KB, docx)

Articles from The Journals of Gerontology Series B: Psychological Sciences and Social Sciences are provided here courtesy of Oxford University Press

RESOURCES