Social connections and mortality in Chilean older adults: analysis of 20 years of follow-up

OBJECTIVE: To estimate the effect of social connections aspects on mortality among the Chilean older population. MATERIALS AND METHODS: This cohort study analyzed a sample of 2 265 community-dwelling adults aged 60+ from Santiago, Chile, with a median follow-up of 15 years. Social connection clusters were derived via cluster analysis across five dimensions: household size, marital status, group participation, recreational social activities, and perceived social support. All-cause mortality (2000-2019) was the outcome. Cox regression models, Kaplan-Meier curves, and interaction analyses were conducted, adjusting for sex, age, education, income, multimorbidity, and depression. RESULTS: Among older adults without multimorbidity at baseline, being in the 'without a partner and social participation' and 'no perception of support and no social participation' clusters increased mortality risk compared to the reference cluster ('with social participation, partner, and support'). CONCLUSION: These findings suggest that older individuals in better health (without multimorbidity) benefit the most in terms of longevity from being socially integrated.