Social support and cognitive function in postmenopausal women

An active social lifestyle may protect against cognitive decline in older adults, but few studies have examined specific social support dimensions in relation to specific cognitive domains. This study examines associations between social support dimensions and cognitive function in older women.

The Early versus Late Intervention Trial with Estradiol (ELITE) was a randomized, double-blinded, placebo-controlled trial of oral 17β-estradiol versus placebo in postmenopausal women, designed primarily to evaluate the timing hypothesis on atherosclerosis progression and cognitive decline; the present analysis is a secondary, hypothesis-generating examination of the psychosocial substudy data (2009-2012). A total of 448 women completed psychosocial assessments every 6 months. Cognitive function was assessed at baseline, 2.5, and 5 years using composite scores for executive function, verbal memory, visual memory, and global cognition derived from 14 standardized tests. Perceived social support was assessed using the Medical Outcomes Study-Social Support Survey (MOS-SSS; 0-100 scale). Paired cognitive-psychosocial assessments were analyzed using linear mixed-effects models adjusted for age, marital status, blood pressure medication, BMI, education, income, race, and randomized treatment (hormone therapy or placebo).

A total of 298 women (mean age 60.3 y) provided 322 paired visits. Positive associations were observed between instrumental social support and visual memory (β(SE)=0.0044(0.0022), 95% CI: -0.00002 to 0.0088, P=0.051) and between positive social interaction and global cognition (β(SE)=0.0092(0.0046), 95% CI: 0.0001-0.0182, P=0.047). In age-stratified analyses, the positive social interaction-global cognition association was significant among women <65 years (P=0.042) but not women ≥65 years (P=0.34). No association survived false discovery rate (FDR) correction for multiple comparisons.

In this sample of healthy middle- to older-aged women, no association between specific dimensions of social support and cognitive domains was statistically significant after correction for multiple comparisons. These hypothesis-generating findings suggest any association is modest at best and warrant confirmation in studies designed specifically for this question.