Association of social integration with cognitive status: results from the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) study

Abstract

AbstractBackgroundLack of strong social ties is associated with numerous morbidities including adverse late‐life cognitive outcomes. Prior work on social integration and cognition has been conducted in predominantly White cohorts. We evaluated the relationships between social integration and cognition in diverse older adults from major racial/ethnic groups in the United States using data from the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) study.MethodKHANDLE is a multi‐ethnic study of 1712 Kaiser Permanente Northern California members age 65+. Social integration was assessed using the Berkman‐Syme Social Network Index. A score ranging from 0 to 6 was calculated by assigning 1 point for each of the following domains: being married/partnered, volunteer activity, having monthly contact with children, relatives, and friends and daily contact with confidante (Figure 1). Cognition included three domains derived from the Spanish and English Neuropsychological Assessment Scales: verbal episodic memory (for White, Black and Latino participants), semantic memory and executive function. Covariate‐adjusted linear mixed‐effects models estimated a single association of social integration with cognition applicable to all three cognitive domains for the full sample and by race/ethnicity. Core models adjusted for age, gender, race, and education (years); expanded models further adjusted for variables that could either be determinants or consequences of social integration, including weekly alcohol use, average score for combined activities of daily living (ADL) and instrumental ADL, and income. We repeated these analyses using frequency of contact with confidante as the primary exposure.ResultHigher social integration score was associated with better cognition in core covariate‐adjusted models (β=0.065 (0.039, 0.091)); the association attenuated slightly when adjusting for expanded covariates (β=0.045 (0.019, 0.071)). In race‐stratified core models, higher social integration scores were associated with better cognition among Asian, Black, Latino, and White participants. In expanded models patterns were similar, but the association remained statistically significant only for Black participants (Table 1). Seeing a confidante daily was associated with higher cognition in the pooled model and for Asian participants (Table 2).ConclusionOur findings suggest that social integration and seeing a confidante daily were associated with better cognition across the four racial/ethnic groups.