Social isolation and loneliness linked to poorer health across 18 countries

Researchers analysed harmonised data from five longitudinal cohorts—HRS, ELSA, SHARE, CHARLS and MHAS—in adults aged 50 years or older followed for 2–10 years during 2010–2020. The study included 94,122 participants across 18 countries. Adjusted country-specific estimates of baseline social isolation and loneliness in relation to repeated physical disease, mental disorder and self-rated health outcomes were pooled using random-effects meta-analysis. The authors also assessed cumulative and joint exposure, subgroup differences and potential mediation.

Social isolation was associated with higher probabilities of physical disease (pooled average marginal effect [AME] 0.007; 95% CI 0.003–0.011) and mental disorders (AME 0.026; 0.014–0.037), and poorer self-rated health (AME −0.048; −0.069 to −0.028). For loneliness, corresponding estimates were 0.011 (0.006–0.016), 0.243 (0.214–0.271) and −0.233 (−0.280 to −0.186). Repeated exposure and experiencing both isolation and loneliness were associated with poorer subsequent health. Associations were stronger in people with lower education, functional limitations or no partner.

The findings support considering social isolation and loneliness as distinct, potentially overlapping markers of health risk in later life. Because this was an observational analysis of existing cohorts, the associations do not establish causation or show that interventions to reduce social disconnection improve health.