Objectives

Adverse childhood experiences (ACEs) are associated with higher risk of chronic disease, but little is known about the association with late life cognitive decline. We examined the longitudinal association between ACEs and late-life cognitive decline in the Study of Healthy Aging in African Americans (STAR).

Design

Linear mixed models with random intercepts and slope examined the association of individual and composite ACEs with cognitive change adjusting for years from baseline (timescale), baseline age, sex, parental education, childhood socioeconomic status and childhood social support. Participants reported whether they had experienced nine types of ACEs. Executive function and verbal episodic memory were measured up to three times over a 3-year period using the Spanish and English Neuropsychological Assessment Scales.

Settings

Kaiser Permanente Northern California members living in the Bay Area.

Participants

STAR is a cohort study of cognitive ageing launched in 2018 that has enrolled 764 black Americans ages ≥50 years (mean age=67.5; SD=8.5).

Results

Twenty-one per cent of participants reported no ACEs, 24% one ACE, 20% two ACEs, 17% three ACEs and 17% four or more ACEs. Compared with no ACEs, two ACEs (β=0.117; 95% CI 0.052 to 0.182), three ACEs (β=0.075; 95% CI 0.007 to 0.143) and four or more ACEs (β=0.089; 95% CI 0.002 to 0.158) were associated with less decline in executive function. There were no significant associations between number of ACEs and baseline or longitudinal verbal episodic memory or between individual ACEs and executive function or verbal episodic memory.

Conclusion

In this cohort of older black Americans, there was no association between ACEs and baseline cognition or cognitive change in verbal episodic memory; however, experiencing ≥ 2 ACEs was associated with less decline in executive function. These results may indicate that participants who survived to age 50+ and experienced ACEs may have cognitive resilience that warrants further investigation.

Abstract: Background: The oldest‐old are the fastest growing segment of the elderly population but very little is known about cognition in this age group; particularly in diverse populations. Our goal was to evaluate if domain specific cognitive change was different across different ethnoracial groups in those aged 90+. Method: LifeAfter90 (LA90) is an ongoing cohort of participants aged 90+ who are long‐term members of Kaiser Permanente Northern California participants. Participants (n = 984) were interviewed every 6 months for up to 3.5 years (1‐7 visits). Executive Function (EF) and Verbal Episodic Memory (VEM) were measured every six months using z‐standardized Spanish English Neuropsychological Assessment Scale. Racial/ethnic identify (Asian, Black, Latino, or White participants) was used in linear mixed models with random slopes and intercepts adjusting for baseline age, gender, education, interview mode, and practice effects. Result: Participants were 20% Latino, 23% Black, 24% Asian, 27% White, and 7% other individuals with a mean age of 92.4 (SD = 2.3) and a mean follow up time 1.1 years (Table 1). 39% of the cohort were men, 35% were college educated, and 29% were high school or less educated. Average annual change in EF was ‐0.06 (95% CI: ‐0.12, ‐0.00). Stratified models with Latino as the reference group showed White participants had significantly greater decline in EF (β = ‐0.13; 95%CI:‐0.20,‐0.06), followed by Asian participants (β = ‐0.09; 95%CI:‐0.17,‐0.01). EF scores among Black participants and participants who identified as other declined at a similar rate as Latino participants (β = ‐0.03; 95%CI:‐0.11,‐0.05; β = 0.00; 95%CI:‐0.10, 0.11; respectively) (Table 2: model 2). VEM had an annual change of ‐0.26 (95%CI: ‐0.40, ‐0.13), but there were no significant differences across ethnoracial groups in rate of decline. Conclusion: In this population of individuals aged 90+, decline in EF over the study period varied across ethnoracial group with White participants experiencing the fastest decline and Black, Latino, and other participants experiencing the slowest decline. Continued follow up will identify if there are differences in risk of cognitive impairment in this diverse population of oldest‐old. The results suggest the disparities in cognitive aging for those aged 90+ don’t mirror disparities seen in younger‐elderly ages.