The association of late‐life volunteering with cognitive function and cognitive decline in the KHANDLE and STAR cohorts
Yi Lor, Marianne Chanti‐Ketterl, Emily Hokett, Evan Fletcher, Zvinka Zlatar, Paola Gilsanz, Elizabeth Rose Mayeda, M. Maria Glymour, Lisa L. Barnes, Rachel A. Whitmer- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
Prior studies have shown that volunteering in late life is associated with better physical health, social activity, and cognition, but the bulk of evidence is limited to non‐Latino White or Asian populations. The purpose of this analysis was to determine whether volunteering is associated with differences in domain‐specific cognition or cognitive change in a diverse cohort.
Method
Kaiser Healthy Aging and Diverse Life Experiences Study (KHANDLE) and Study of Healthy Aging in African Americans (STAR) are harmonized cohort studies of long‐term Kaiser Permanente members ages + and 50+ respectively, consisting of 48% Black, 20% White, 17% Asian, and 14% Latinx individuals. Participants self‐reported performing volunteer work for the 12 months prior to baseline for religious, educational, health‐related, or charity organizations. Executive function (EF) and verbal episodic memory (VEM) were measured across three waves using the Spanish and English Neuropsychological Assessment Scales (SENAS) and scores were z‐standardized to the baseline. Linear mixed‐models with random intercepts and slopes evaluated the association of volunteering with baseline cognition and cognitive change over 3 visits (mean follow‐up = 1.2 years) with years since baseline as the timescale and adjusting for baseline age, sex, education, income, practice effects, and interview mode (phone vs in‐person)
Result
Of the 2476 participants with average baseline age of 73.7 (SD = 8.2) years, 1167 (42.3%) had volunteered in the 12 months prior to baseline (Table 1). Those who volunteered had on average higher baseline EF (β = 0.184; CI = 0.110, 0.257) and VEM (β = 0.114; CI = 0.040, 0.189) compared to those who did not report volunteering (Table 2, Model 2). There were no statistically significant associations of volunteering and EF or VEM decline, though the direction of the associations suggested that those who volunteered had slower cognitive decline than those who did not volunteer.
Conclusion
In this diverse cohort, volunteering was associated with significantly higher baseline cognition, and suggested to be associated with less cognitive change over an average of 1.2 years, though not statistically significant. With additional follow up, we will examine whether volunteering is protective against cognitive impairment, and how physical and mental health may mediate this relationship.