End‐of‐Life Loneliness, Social Isolation, and Symptom Burden: A Nationally‐Representative Study

ABSTRACT Background Loneliness (subjective feeling of lacking connection) and social isolation (objective deficit in number of relationships or contact with others) are common at the end‐of‐life and can be detrimental to quality of life. Investigating the association between symptoms and end‐of‐life loneliness and social isolation could help inform targeted interventions. Methods We used nationally‐representative, cross‐sectional Health and Retirement Study data, including adults > 50 years old ( N = 2385) who died while enrolled. Respondents self‐reported on validated loneliness and social isolation measures within 1 year of death; 12 physical and psychological symptoms were determined via after‐death interviews with proxies. We use multivariable logistic regression to determine the adjusted probability of end‐of‐life loneliness or social isolation by each symptom, adjusting for socio‐demographic covariates. Results Respondents were on average 76 years old, 50% female, 82% identified as White, 10% Black, and 5% Latino. Loneliness was more common among the following symptoms ( p < 0.05): pain (30% vs. 20%), depression (35% vs. 18%), fatigue (29% vs. 22%), drowsiness (32% vs. 25%), and agitation (38% vs. 24%). Social isolation was more common in decedents who had difficulty breathing (22% vs. 14%, p = 0.03), drowsiness (30% vs. 17%, p = 0.006), and persistent cough (24% vs. 15%, p = 0.007). Conclusions In the last year of life, multiple psychosocial symptoms were associated with experiences of loneliness whereas physical symptoms were associated more with social isolation. While these relationships are complex, addressing the social sequela of physical and psychological symptoms may be an opportunity to improve overall well‐being at the end‐of‐life.