Association of social isolation and loneliness with years of life lost across healthy lifestyle levels

Social isolation (SI) and loneliness are prevalent worldwide; however, their associations with life expectancy and whether these associations vary across healthy lifestyle levels remain unclear. This study aimed to investigate the associations of SI and loneliness with all-cause mortality and years of life lost (YLL), and to examine whether these associations varied across healthy lifestyle levels. Data from the UK Biobank were analyzed. SI and loneliness were assessed using established questionnaire-based measures, and healthy lifestyle was evaluated using a composite score based on five modifiable lifestyle behaviors. Flexible parametric Royston-Parmar survival models were used to estimate hazard ratios (HRs) for all-cause mortality and derive YLL estimates associated with SI and loneliness, both overall and within healthy lifestyle score (HLS) strata. Compared with participants aged 40 years with no social isolation and loneliness, those with only loneliness, only social isolation, and social isolation with loneliness had estimated YLLs of 0.81[95%CI:0.49–1.14], 1.02[95%CI:0.80–1.24], and 2.21[95%CI:1.83–2.58], respectively. Similar patterns were observed among participants aged 45, 60, and 65 years. At the age of 40 years, among participants with low HLS, the YLLs associated with only loneliness, only social isolation, and social isolation with loneliness were 1.05[95%CI:0.57–1.53], 1.48[95%CI:1.13–1.83], and 2.36[95%CI:1.83–2.88], while among participants with high HLS, the corresponding YLLs were 0.17[95%CI:-0.45-0.80], 0.47[95%CI:0.06–0.89], and 1.39[95%CI:0.66–2.11]. A similar pattern was also observed among participants aged 45, 60, and 65 years. SI and loneliness were each associated with YLL, and their co-occurrence was associated with greater YLL than either condition alone. YLL estimates associated with SI and loneliness were numerically smaller at higher HLS levels.