Background: Continuous paid work supports cognitive health, but employment disruptions, especially among historically marginalized groups, may increase the risk of cognitive decline and Alzheimer's disease and related dementias (ADRD). Limited evidence exists on the relationship between caregiving-related job disruptions and cognitive outcomes, which may exacerbate disparities in ADRD risk. Objective :This study examines the association between caregiving-related job disruptions and late-life cognitive outcomes, with a focus on gender and racial/ethnic differences. Methods: Data from 7759 U.S. adults aged 55 and older in the 2018 Health and Retirement Study (HRS) and Life History Mail Survey (LHMS) were analyzed. Employment disruptions were classified as caregiving-related, other reasons, or continuous employment. Sequence analysis summarized employment history patterns, and multivariable linear regression models evaluated the association between the years of caregiving-related job disruptions between ages 18 and 50 and cognitive function scores at ages 55 and older, adjusting for demographic covariates. Results: Women experienced more caregiving-related job disruptions than men (0.57 versus 0.04 years, p < 0.001), and Hispanics had more disruptions than non-Hispanic Whites (0.84 versus 0.28 years, p = 0.01). Caregiving-related disruptions were significantly associated with lower cognitive scores among non-Hispanic White males, non-Hispanic Black males, and non-Hispanic Black females, but not among non-Hispanic White females or Hispanic males and females. Conclusions: Caregiving-related job disruptions may contribute to cognitive decline, particularly among non-Hispanic White males and non-Hispanic Black caregivers. Targeted policies and interventions addressing these disparities are essential to reduce ADRD risk across diverse caregiving populations.

Influential work in the gerontological literature in the 1990s has demonstrated that the discrepancy between objective and subjective evaluations of health increases with age. It is an open question though whether the direction and size of the age gradient in the discrepancies has changed over historical time. In this study, we examine cohort differences in age-related divergences in health evaluations using two markers of objective health (illnesses/functioning) and two markers of subjective health (overall/compared to others) obtained from cross-sectional samples of older adults in the Berlin Aging Studies in 1990-1993 (n = 516, aged 70-103) and 2018-2020 (n = 625, aged 68-94). We use both person-oriented and variable-oriented methodological frameworks that specifically facilitate study of aging as a process that propels individuals through a multidimensional health space: permutational multivariate analysis of variance (PERMANOVA) and structural equation modeling (SEM). The person-oriented PERMANOVA results revealed that divergence across markers of health increased with age, but did not evidence any cohort differences. Complementary results from variable-oriented SEMs corroborated that the four-dimensional variance-covariance structure did not differ across cohorts and that the age pattern was consistent across cohorts. Results suggest that objective and subjective indicators of health are increasingly discrepant after age 70, but reveal no evidence that age gradient of the divergences has changed across historical time.

The life course stress model suggests that stressors, such as loneliness, can accumulate and shape long-term cognitive trajectories, particularly in later life. While loneliness has been linked to cognitive decline, few studies have examined whether this relationship is bidirectional. This study investigated associations between loneliness and cognitive function over time. Data were drawn from four waves (2008–2022) of the Health and Retirement Study. The analytic sample included 12,144 adults aged 50 and older. A random-intercept cross-lagged panel model was used to separate within-person fluctuations from stable between-person differences, and a dual-process latent growth curve model examined long-term patterns in loneliness and cognitive function. Loneliness was measured using the 11-item UCLA Loneliness Scale, and cognition was assessed via immediate and delayed word recall tasks. Cross-lagged effects pathways between loneliness and cognitive function were not statistically significant, suggesting no within-person bidirectional relationship. However, the random intercepts for cognition and loneliness were negatively correlated (r = -0.22, p < .001), suggesting that individuals with higher loneliness tended to have lower cognitive function at baseline. Additionally, the random slopes for cognition and loneliness were also negatively correlated (r = -0.30, p < .001), indicating that individuals with greater declines in cognition also tended to experience increases in loneliness over time. These findings suggest that loneliness and cognitive function are interconnected, even in the absence of short-term directional effects. Understanding their long-term association can inform interventions aimed at preserving cognitive health and reducing loneliness in older adults.

Tips about how to remain mentally sharp abound in the media alongside reports about the personal and societal challenges of dementia. Multiple questions remain about the impact of these messages. We use 2022 panel data from the Health and Retirement Study (HRS) to examine concerns about cognitive health and proposals suggesting a mismatch between actual cognitive performance and subjective status. Eligible participants (N = 4098: Mage = 69.5; Range 50-101; 60% women) responded to five items in the Psychosocial and Wellbeing questionnaire about concerns regarding age-related challenges [staying mentally sharp, remaining in own home, health-care expenses, communication, and developing Alzheimer’s Disease (AD)] together with subjective and objective assessments of cognition, health, perceptions of aging, and demographic covariates. Staying mentally sharp was the top concern (71%) followed by staying in one’s own home (56%) and AD (51%). The items were intercorrelated (rs = 0.62 to 0.42). Logistic regressions predicting concerns revealed that women (ORs =1.25), people with more chronic health diagnoses (ORs =1.37), and depression (ORs =1.84) were more likely to be concerned. Objective memory performance was not associated with concerns, but people who rated their memory as poor were more likely to be concerned (ORs=1.95). Additional analyses, revealed that people who viewed their own aging more positively were less likely to be concerned (OR = 0.67). In a longitudinal HRS subsample (N = 225), 31% who were not concerned about AD in a 2010 became concerned in 2022. Future research should investigate if changes in cognition, physical health, depression, and media use contribute to concerns.

Continuous paid work supports cognitive health, but employment disruptions, especially among historically marginalized groups, may increase the risk of cognitive decline and Alzheimer’s disease and related dementias (ADRD). While prior research has explored employment trajectories and cognitive aging, limited evidence examines how caregiving-related employment disruptions affect cognitive outcomes, especially across race/ethnicity and gender. This study investigates the association between caregiving-related job disruptions and cognitive function among older adults, with a focus on disparities. Using data from 7,769 U.S. adults aged 55 and older in the 2018 Health and Retirement Study (HRS) and Life History Mail Survey (LHMS), we classified employment histories into caregiving-related disruptions, other disruptions, or continuous employment. Sequence analysis identified employment patterns, and multivariable linear regression models assessed the impact of caregiving-related job disruptions on cognitive function, adjusting for demographic and socioeconomic covariates. Results show that women experienced significantly more caregiving-related job disruptions than men (0.54 vs. 0.04 years, p < 0.001), and Hispanic individuals had more disruptions than non-Hispanic Whites (0.83 vs. 0.26 years, p = 0.010). Caregiving-related job disruptions were significantly associated with lower cognitive scores in both females (β=-0.07, p = 0.002) and males (β=-0.14, p = 0.03), with notable disparities among non-Hispanic White and non-Hispanic Black male caregivers. Findings suggest that caregiving-related employment disruptions may contribute to cognitive decline, particularly among non-Hispanic White males and non-Hispanic Black caregivers. Targeted policies and interventions addressing these disparities are essential to reduce ADRD risk across diverse caregiving populations.

BACKGROUND:Higher glycemic levels are associated with poor cognitive outcomes in older adulthood. Early-life educational experiences predict later-life cognition, but their role as a moderator of the association between hyperglycemia and cognition is unknown. Educational experiences in childhood promotes cognitive reserve, and greater cognitive reserve may reduce the deleterious effects of hyperglycemia on cognition in later-life. We hypothesized that higher glycemic level and poorer childhood educational experiences would be associated with worse memory outcomes in later-life, and that the negative association between glycemic level and memory would be stronger in individuals with worse educational experiences. METHODS:Participants comprised 8,142 adults from the Health and Retirement Study Life History Mail Survey (LHMS) with biomarker data (Table 1). Memory was assessed biennially (three timepoints) with immediate and delayed word recall. Glycemic level was indexed by hemoglobin A1c (HbA1c). Retrospectively reported childhood educational experiences included: learning problems, school context (predominantly white vs. non-white students), mathematics ability, reading/writing ability, and participation in extracurricular language/creative arts activities. Covariates were baseline age, sex/gender, race, ethnicity, birth cohort, parental education, wealth, and assistance with LHMS questionnaire completion. RESULTS:Latent growth curve models showed independent associations of HbA1c and all education variables with memory intercept, but not memory decline (Figure 1). Models separated by each education variable showed that the association between higher HbA1c and steeper memory decline was stronger in older adults who reported very low reading/writing (B = .19, SE=0.10, p = .04) and mathematics (B = -.22, SE=.11, p = .04) abilities (Figure 2). Patterns of association were stronger when HbA1c was restricted to non-normal (≥5.7%) and clinically elevated (≥6.5%) levels. Educational experiences did not moderate HbA1c associations with memory intercept. CONCLUSION:The current study provides preliminary evidence suggesting that early-life educational quality, particularly in mathematics and literacy, may shape hyperglycemia-related associations with memory aging. Poor educational experiences during formative years may hinder the development of cognitive reserve, exacerbating the adverse cognitive consequences of hyperglycemia. Future studies are needed to establish causal relationships, examine educational experiences beyond high school, and model memory trajectories over a longer follow-up period, to more fully characterize the contribution of early-life educational experiences to cognitive health in later-life.

INTRODUCTION:Multisystem biological risk, a marker of physiological dysregulation, accumulates over time in response to exposure and adaptation to chronic stress. Spousal concordance in health and health behaviors indicates potential for shared risk to accumulate within couples. This study examined spouses' shared multisystem biological risk as a modifier of individual risk and predictor of future functional limitations and mortality. METHODS:Biomarkers and physical measures from 3,856 heterosexual couples (77%White, 31% college degree) were collected from the 2008/2010 waves of the Health and Retirement Study to construct individual and shared frailty, cardiometabolic, and total risk indices and predict functional limitations and survival at 2016/2018. RESULTS:Multilevel Poisson and logistic regressions showed couples' cumulative shared frailty, cardiometabolic, and total risk to be associated with the number of functional limitations and survival status at follow-up. Spouses' shared cardiometabolic and total risks attenuated the effects of respective individual risks. Results were partially explained by partner selection and health experiences. Predicted probabilities were compared to gender-stratified models. The Receiver Operating Characteristic curve showed models of shared risk to have greater predictive power. CONCLUSIONS:The findings from this study indicate that the cost of adaptation manifests not only in individuals, but also through an additional pathway that is co-constructed and shared by spouses. These findings underscore the critical role of shared context between individuals and their spouses in the treatment process.

This symposium showcases new cross-disciplinary research into proposals about early-life and midlife contexts associated with later-life cognition and dementia. The early-career speakers utilize data from the Carolina African American Twins Study of Aging (CAATSA), Health and Retirement Study (HRS), and Panel Study of Income Dynamics (PSID) and apply advanced analysis methods. Presentations consider the influence of education, early adult and midlife neighborhood poverty, and pre-retirement work contexts in cognitive aging. Early-life education is well-known as a predictor of dementia. Johnson expands this research by revealing a unique role for education beyond zygosity and socio-environmental factors. Kalousová examines the relevance to late life cognition of early-to-late life residential contexts characterized by neighborhood poverty. Using sequence analysis, she identifies distinct patterns of exposure to neighborhood disadvantage over time. Cumulative exposures to areas with limited resources increased the likelihood of late-life cognitive impairment. Retirement from the workforce often contributes to increased stress, loneliness, and reduced cognitive engagement. Kayser, Xiang, and Smith apply cross-wave dynamic change models to HRS multi-wave data to disentangle bidirectional associations between changes in cognition and loneliness after retirement. They report that over the longer-term both cognitive decline and loneliness increase. Working longer and different retirement trajectories may mitigate cognitive decline. Oh used sequence analysis to identify six work-related trajectories prior to retirement. He finds that working longer and engaging in so-called “encore” careers reduce the probability of later-life dementia. The session concludes with an integrative discussion by Carr of the implications of these studies for future research and policy.

Higher glycemic levels are associated with poor cognitive outcomes in older adulthood. Early-life educational experiences predict later-life cognition, but their role as a moderator of the association between hyperglycemia and cognition is unknown. Educational experiences in childhood promotes cognitive reserve, and greater cognitive reserve may reduce the deleterious effects of hyperglycemia on cognition in later-life. We hypothesized that higher glycemic level and poorer childhood educational experiences would be associated with worse memory outcomes in later-life, and that the negative association between glycemic level and memory would be stronger in individuals with worse educational experiences. Participants comprised 8,142 adults from the Health and Retirement Study Life History Mail Survey (LHMS) with biomarker data (Table 1). Memory was assessed biennially (three timepoints) with immediate and delayed word recall. Glycemic level was indexed by hemoglobin A1c (HbA1c). Retrospectively reported childhood educational experiences included: learning problems, school context (predominantly white vs. non-white students), mathematics ability, reading/writing ability, and participation in extracurricular language/creative arts activities. Covariates were baseline age, sex/gender, race, ethnicity, birth cohort, parental education, wealth, and assistance with LHMS questionnaire completion. Latent growth curve models showed independent associations of HbA1c and all education variables with memory intercept, but not memory decline (Figure 1). Models separated by each education variable showed that the association between higher HbA1c and steeper memory decline was stronger in older adults who reported very low reading/writing ( B  = .19, SE=0.10, p  = .04) and mathematics ( B  = -.22, SE=.11, p  = .04) abilities (Figure 2). Patterns of association were stronger when HbA1c was restricted to non-normal (≥5.7%) and clinically elevated (≥6.5%) levels. Educational experiences did not moderate HbA1c associations with memory intercept. The current study provides preliminary evidence suggesting that early-life educational quality, particularly in mathematics and literacy, may shape hyperglycemia-related associations with memory aging. Poor educational experiences during formative years may hinder the development of cognitive reserve, exacerbating the adverse cognitive consequences of hyperglycemia. Future studies are needed to establish causal relationships, examine educational experiences beyond high school, and model memory trajectories over a longer follow-up period, to more fully characterize the contribution of early-life educational experiences to cognitive health in later-life.

Abstract Efforts to determine potentially modifiable factors associated with heterogeneity in cognitive status in old age point to the importance of participation in cognitive, physical, and social leisure activities and the relevance of historical changes in completion of post-secondary education. We use data from two birth cohorts in the US Health and Retirement [HRS: Baby Boomers (BB) born 1948-1959; Older cohorts (OC) born before 1948] to examine this proposal. Eligible participants (N=8503: Mage=71) provided post-secondary education information in the HRS Life History Mail Survey (LHMS, 2015-2017) and cognitive status and participation in leisure activities in the 2016 or 2018 HRS biennial waves. Cohorts differed in rates of college education (BB 53% vs OC 42%) and overall leisure activity participation (BB>OC, p<.001). A multiple regression analysis predicting cognitive status (normal vs impaired: Langa-Weir algorithm) revealed significant effects for college education (OR=O.47), leisure activities (OR=0.87), and cohort (ps <.001), after controls for age, gender, wealth, race, marital status, chronic illnesses, and depression. In follow-up cohort-stratified analyses, we grouped people by level of participation in specific activity types (top quartile vs rest). Interestingly, within the older cohort, the risk of cognitive impairment was significantly reduced by high participation in cognitive (OR=0.42) and physical (OR=0.71) leisure activities in addition to having college education (OR=0.47). Within the BB cohort, only people with high participation in cognitive leisure activities (OR=0.65) in addition to college education (OR=0.38) had a reduced risk. Future research should investigate why some leisure activities seem to confer benefits for cognitive resilience whereas others don’t.

Objectives: Structural racism creates contextual stressors that disproportionately affect Black, relative to White, older adults in the United States and may contribute to worse cognitive health. We examined the extent to which interpersonal, community, and societal stressors uniquely explain Black-White disparities in initial memory and memory change. Methods: The sample included 14,199 non-Latino Black and White older adults (M-age = 68.32, 19.8% Black) from the U.S. Health and Retirement Study who completed psychosocial questionnaires at baseline and a word list memory task every 2 years over an 8-year period. Interpersonal, community, and societal stressors were operationalized as self-reported everyday discrimination, neighborhood physical disorder, and subjective societal status, respectively. Latent growth curves modeled longitudinal memory performance. Stressors were modeled simultaneously and allowed to correlate. Covariates included age, sex, education, wealth, parental education, and Southern residence. Results: Compared to White participants, Black participants experienced more discrimination (beta = -0.004, standard error [SE] = 0.001, p < .001), more neighborhood physical disorder (beta = -0.009, SE = 0.002, p < .001), and lower perceived societal status (beta= -0.002, SE = 0.001, p = .001), each of which uniquely mediated the racial disparity in initial memory. Sensitivity analyses utilizing proxy-imputed memory scores revealed an additional racial disparity in memory change, wherein Black participants evidenced a faster decline than White participants. This disparity in memory change was only uniquely mediated by more everyday discrimination among Black participants. Discussion: Elements of structural racism may contribute to cognitive disparities via disproportionate stress experiences at multiple contextual levels among Black older adults. Future research should consider multilevel protective factors that buffer against negative impacts of racism on health.

Abstract Belonging is a fundamental human need that encourages fortitude throughout the life course. In the literature, belonging has been studied primarily from two perspectives: place attachment (Chawla, 1992, Scannell, 2021) and belonging within a social group (Allen, 2020). Less research, however, has been conducted about how belonging is experienced by the individual – self-perceptions of belonging. Using data from the Health and Retirement Study Life History Mail Survey (N=3,091, M Age =70.48 [50-98]) we investigated retrospective self-reports of belonging from three life stages, childhood, adolescence and midlife, as predictors of belonging and activity engagement in late life. First, we conducted linear regressions controlling for sociodemographic variables (gender, age, race, marital status). This revealed that self-reported feelings of belonging in childhood (β=.11, p<.001), adolescence (β=.13, p<.001), and midlife (β=.23, p<.001) were significantly predictive of heightened feelings of belonging in late life (Adj R2=.043,.051,.068). Second, we assessed self-reported belonging in childhood, adolescence, midlife and late life as predictors of late life engagement. These analysis revealed that belonging in childhood (β=.06, p<.001), midlife (β=.15, p<.001), and late life (β=.13, p<.001) were significantly predictive of heightened activity engagement in late life (Adj R2=.026,.037,.040). Overall this study demonstrates the importance of assessing how individuals think about their belonging retrospectively throughout their lifespan as a predictor of subsequent late-life belonging and activity engagement– both protective factors as a person ages.

Abstract Background Studies from Europe indicate an increase in social support exchanges among older adults, but whether this trend also applies to the US and the perception of these exchanges as positive or negative remains unclear. Furthermore, it’s uncertain if these changes also benefit those that face disadvantage, such as low income and racial minority groups, especially when considering intersecting factors. Methods This study employs cohort-sequential analysis of Health and Retirement Study (HRS) data collected from 2008 to 2022 in 8 waves to investigate cohort differences in positive and negative social support exchanges among older adults in the United States. It focuses on the influences of cohort, race, and income on social support dynamics, employing data from the psycho-social questionnaire. Results Despite a general increase in positive support exchanges, disparities in social support exchange persist based on race and socioeconomic status. Individuals with lower incomes and from ethnic minority backgrounds experience disproportionately more negative social support exchanges, indicating persistent inequalities. Discussion Recognizing the complexities of social support exchanges, particularly in relation to cohort, race, and socioeconomic status, is crucial for policymakers and healthcare professionals. Tailored interventions are essential to address disparities and promote equitable aging outcomes and overall well-being among older adults in the US.

Abstract Few studies examined the ways older workers self-identify their retirement status (named ‘retirement identity’) over time before exiting the workforce. Examining the trajectories of retirement identity transitions – between “not retired”, “partly retired”, and “completely retired” – clarifies the heterogeneous outcomes found in ‘static’ retirement identity. In this study, we distinguish and categorize these trajectories using sequence analysis, investigate the common sociodemographic and economic characteristics for each trajectory using multinomial logistic regression, and examine the associations between these trajectories and the physical and mental health conditions of older workers using latent growth modeling. Drawing on a sample of respondents from the Health and Retirement Study (HRS) from 1992 to 2020, we use the first nine waves of the initial HRS birth cohort (born 1931-1936, recruited in 1992, N=1,487) and the Early Baby Boomers (born 1948-1953, recruited in 2004, N=1,017). We determine and name four distinct categories of trajectories – early, late, traditional, and partial transitions – based on the timing and sequence of transitions. The membership of these trajectories is strongly associated with the birth cohort, educational attainment, labor income from career jobs, and household wealth. The latent growth modeling reveals that workers in late or partial transition trajectories tend to have better initial physical, mental, and cognitive health and experience slower deterioration of these than those in early or traditional transition trajectories. These results suggest a close relationship between health and self-identification of retirement status that may influence the work attachment and timing of workforce exit.

Abstract Butler (1963) proposed that engaging in a whole life review is especially beneficial for older adults because it promotes a sense of self-worth. Previous research highlights links between reminiscing and psychological well-being, however, life review narratives are rarely collected in the context of large, population-based aging studies. This study uses data from the HRS Life History Mail Survey (2017, 2019) to 1) examine factors that predict older adults (N=12,815; Mage= 68 years; range 50-101) willingness to share their life’s most important accomplishments (up to three); and 2) establish relations between content and complexity (content variety) of reported accomplishments and participants’ well-being. Participants who reported their accomplishments (59%), compared to those who didn’t, were more likely to be better educated, currently married, wealthier, have better memory and fewer depressive symptoms. Importantly, they also were more likely to perceive their life-as-a-whole as either completely or very satisfied (OR=1.315, CI=1.214-1.425, p<.001). Multinomial analysis of content and complexity of reported accomplishments revealed that people with higher life satisfaction were more likely to report relationship-related accomplishments (47%; family, friendship, or caring for others) compared to those who were most proud of their societal achievement (35%; education, work, military services), adjusted RRR=1.207, CI=1.061-1.372, p<.005. People, who wrote about personal qualities (e.g., their faith, health) did not differ from other two groups. The complexity of reported accomplishments was not associated with life satisfaction but was related to respondents’ education and memory. Our findings support the proposal that reviewing one’s life accomplishments enhances well-being in later life.

Background and Objectives:The growing interest in the impact of lifetime occupational exposures on later-life health underscores the need to expand and evaluate the quality of data resources. The present study took advantage of a retrospective life history survey fielded within the context of the Health and Retirement Study to assess the accuracy of retrospectively obtained information on job history. We evaluated hypotheses related to job history and respondent characteristics to understand more about factors associated with recall accuracy.Research Design and Methods:We used data from the Life History Mail Survey (LHMS), a self-administered survey conducted in 2015 and 2017. We compared the match rate of work status collected in the LHMS questionnaire with data collected concurrently during HRS core face-to-face or phone interviews from 1992 through 2016 with respect to jobs held at the time of the interview. We also conducted a limited set of comparisons of occupation and industry match.Results:The sample was 61.79% women, 82.12% White, and 8.57% Hispanic with a mean age of 74.70 years. The overall work status match rate was 83%. Jobs held longer ago were recalled with less accuracy. Jobs held for longer durations and that were full-time rather than part-time were recalled with greater accuracy. More complex job histories that involved a larger number of jobs were also associated with a lower match rate. Higher levels of conscientiousness and cognitive functioning were both associated with a higher match between the two sources of work status information. The occupation match rate was 69%, and the industry match rate was 77%.Discussion and Implications:A self-administered, paper-and-pencil questionnaire attempting to measure decades-long histories of autobiographically important dimensions of life can provide reasonably accurate historical employment information. Several factors are likely to influence the relative accuracy of recalled information.

Abstract Singlehood has long been identified as a potential risk factor for undermining well-being across adulthood, particularly in older ages. However, it is an open question what role other social resources may play. In the current study, we examine whether and how older adults experiencing marital transitions through divorce and spousal bereavement differ from their never married and continuously married peers in their well-being trajectories and in the roles that social support and strain play. We used six waves (2008 – 2018) of longitudinal data from 7,670 participants (Mage = 66 years, 56% women, 75% White) of the Health and Retirement Study (HRS). We also employed synthetic cohort matching to equate the groups on relevant background factors. First results of multi-level models revealed that bereavement trajectories were characterized by a steeper transition-related rise in depressive symptoms than the divorce trajectories, but also by a sharper drop in subsequent years. Both transition groups showed declines in depressive symptoms in the years after the event, yet those who had experienced divorce reached average levels of their continuously married peers after about three years, whereas the bereaved group continued to lag behind even after six years post-spousal loss. Overall, we found little evidence that the relevance of social resources for well-being trajectories differs across marital status groups. As one exception, we found that the predictive effects of strain from family and friends for elevated depressive symptoms was particularly pronounced among those who had experienced spousal loss.