Exploring the effect of loneliness on all-cause mortality: Are there differences between older adults and younger and middle-aged adults?

Objective: This study aims to investigate the association between loneliness and all-cause mortality over a six-year follow-up period using the overall sample and by age groups (18-59 years and 60+ years). Method. Data from a longitudinal, prospective study of a nationally-representative sample of the Spanish non-institutionalized adult population were analysed (n = 4467). Mortality was ascertained via linkage to the National Death Index or obtained during the household visits. The UCLA Loneliness Scale was used to measure loneliness. Sex, age, education, physical activity, tobacco consumption, body mass index, disability, depression, living situation, and social participation were also considered as covariates. Multivariable Cox proportional hazard models were carried out. Results. A higher level of loneliness was not associated with mortality risk in fully covariate-adjusted models over the entire population (HR = 1.02; 95% CI = 0.94, 1.12). The interaction term between loneliness and age groups was significant, indicating that the rate for survival of loneliness varied by age (HR = 1.29; 95% CI = 1.02, 1.63 for young- and middle-aged individuals; HR = 0.96; 95% CI = 0.89, 1.04 for older adults). Conclusions. The development of interventions aimed at tackling loneliness among young- and middle-aged adults might contribute to a mortality risk reduction. Future research is warranted to test whether our results can be replicated.