Living alone, loneliness and mortality in hospitalized older adults: findings from the SAFES cohort

Population aging is a worldwide phenomenon. Social isolation and loneliness are common factors of social vulnerability among older people and are associated with morbidity and mortality. Loneliness is a subjective state and cannot be directly measured by objective criteria such as marital status or living alone, although the latter may have an indirect link to feeling lonely. The main aim of this study was therefore to investigate the association between loneliness, living alone and survival in a nationwide French cohort including adults aged 75 years or older who are admitted to hospital via the emergency department. Data were extracted from the SAFE cohort, a multicenter study of patients aged ≥75 hospitalized after emergency department visits (March 2001-January 2002). Data included socio-demographic, clinical variables, and living arrangements. Loneliness was assessed via CES-D. Survival at 1, 2, and 3 years was analyzed using Cox proportional hazards models. Bivariable and multivariable analyses were performed using Cox logistic regression models. A total of 1,306 patients were included in the SAFES cohort, with an average age of 85.0±5.9 years, and almost two-thirds women (65%). Among the persons who lived alone (N=467), 45% did not feel lonely. Among those who did not live alone, 34% did feel lonely. By multivariable analysis, after adjustment, we found that individuals who lived alone had longer survival than those who did not live alone. Our findings show that social isolation protects against mortality in this population. Loneliness, on the other hand, remains simply an indicators of vulnerability, but was not significantly associated with mortality. This is a clinical reality that should be identified as early as possible in order to implement appropriate strategies and management in a pluridisciplinary context.