Chronic loneliness and isolation phenotypes, incident functional impairment and mortality in England between 2004 and 2018

Social deficits are potential risk factors for premature mortality. Most research has focused on social deficits measured at single points in time. It remains unclear if the chronicity of loneliness affects its impact on adverse health outcomes. This study assessed the effects of chronic loneliness and isolation in predicting incident functional impairment and all-cause and cause-specific mortality. This longitudinal study used panel data from the English Longitudinal Study of Ageing, including 14 years of follow-up (waves 2-9, in 2004-2018). Social deficits over three waves (4 years) were measured using the UCLA loneliness scale and social isolation index, categorized as not present, fluctuating or chronic. We estimated the all-cause mortality risk with Cox proportional hazard modeling, and the Fine-Gray competing risk modeling was used to test the risk of functional impairment onset and cause-specific mortality. We analyzed 5,131 participants (mean age 67.6 years (s.d. 9.8)) in the mortality cohort (follow-up 9.8 years (IQR: 6.67-10.08)) and 4,279 participants (mean age 67.0 years (s.d. 9.6)) who were functional disability-free at baseline (follow-up 9.8 years (IQR: 7.17-10.17)). Compared with not being lonely/isolated, there was a higher risk of incident functional impairment among those with fluctuating loneliness (sub-hazard ratio (sHR) 1.30, 95% CI: 1.03-1.63) and chronic loneliness (sHR 1.58, 1.12-2.23), as well as chronic social isolation (sHR 1.41, 1.02-1.94). In survival analyses, compared with people who were not lonely/isolated, people experiencing fluctuating loneliness and social isolation had higher risks of all-cause mortality (loneliness HR 1.29, 1.13-1.48; isolation HR 1.15, 1.01-1.31). People with chronic isolation also had higher risks of all-cause mortality (HR 1.27, 1.05-1.55) and cancer-related mortality (sHR 1.69, 1.23-2.31). Over a 14 year follow-up, we found that chronic loneliness and isolation phenotypes were associated with aggravated risks of incident functional impairment and mortality. There was a potential dose-response relationship between chronicity of loneliness phenotypes and functional impairment onset and mortality. Preventing the onset of and transition to chronic loneliness and isolation in older age is a crucial target to support both the healthspan and the lifespan.