The effects of loneliness and social isolation on all-cause, injury, cancer, and CVD mortality in a cohort of middle-aged Finnish men. A prospective study

Abstract

Objectives

Loneliness and social isolation both increase mortality and are likely to affect health via several pathways. However, information on the potential pathways remains scarce. We investigated the associations between loneliness, social isolation, and mortality, and possible mechanisms underlying these connections.

Methods

The analyzed data comprised a prospective population-based cohort of Finnish men (42–61 years at baseline, n = 2588) who were followed up for an average of 23.2 years. Mortality data were obtained from the national population register in 2012. Cox proportional hazards analysis with adjustments for possible confounding factors was used to examine the associations between loneliness and social isolation at baseline and all-cause, injury, cancer, and cardiovascular disease (CVD) mortality. Mediation analysis was conducted to investigate the mechanisms underlying the associations of loneliness and social isolation with mortality.

Results

Loneliness predicted all-cause mortality, even after adjustments for all covariates. Loneliness predicted cancer mortality, except after adjustments for lifestyle variables or Human Population Laboratory (HPL) depression scores, and also predicted CVD mortality, except after adjustments for HPL depression scores. Social isolation predicted all-cause mortality and injury mortality. The effect of social isolation on all-cause mortality was mediated by loneliness and HPL depression scores.

Conclusions

Our findings suggest that both loneliness and social isolation increase the risk of all-cause mortality, while they have differing effects on different causes of death. Loneliness and depressive symptoms may mediate the effect of social isolation on increased mortality.

Keywords:

• All-cause mortality
• loneliness
• social isolation
• mediation
• depression
• longitudinal study

Acknowledgements

The corresponding author was supported by the Signe and Ane Gyllenberg Foundation while writing this manuscript.

Disclosure statement

The authors report no conflict of interest.