Social isolation in mid-life: associations with psychological distress, life satisfaction and self-rated health in two successive British birth cohorts

Abstract

Background and Objectives: Social isolation can be quantified using indicators across a range of social contexts e.g., household composition, friends and family, employment and community. Little is known about the way in which different forms of isolation differentially impact health, whether they produce a dose-response effect, and whether different forms of isolation interact in an additive or multiplicative way to impact psychological distress, life satisfaction and general health. The current study focuses on social isolation in mid-life, a life stage often neglected by the field. Research Design and Method: Data (N=32,391) were analysed from two successive British birth cohort studies: 1970 British Cohort Study (BCS) N=16,585 and the 1958 National Child Development Study (NCDS) N=15,806 focusing on participants social isolation during mid-life (ages 42-46). Linear multivariable regression models were run to investigate the independent, dose-response and interactive associations of social isolation indicators with psychological distress, life satisfaction and self-rated general health. Subsequently, all models were stratified to examine sex differences, and a cohort interaction was tested to assess cohort effects. An additional latent class analysis investigated whether different forms of isolation cluster in mid-life. Results: Independent associations varied by form of isolation and specific health outcomes, for instance, being out of education and employment was associated with all outcomes, living alone was only associated with lower life satisfaction and a lack of frequent contact with friends and relatives, no labour market participation and limited community engagement were associated with lower life satisfaction and self-rated general health. There was a dose-response relationship with higher social isolation scores associated with greater psychological distress, lower life satisfaction, and poorer self-reported general health. The effects of different combinations of social isolation on health appears to be additive, with no consistent sex and cohort differences observed. Conclusions: Results support a multi-domain, multi-context approach to studying social isolation and provides justification for research that investigates the separate and combined effects of different forms of social isolation; and the relevance of investigating social isolation in relation to the complete state of physical, mental and social wellbeing. Overall, isolation in its various forms was found to be detrimental for health, in mid-life, and over time in Great Britain. In particular, social isolation was associated most consistently with lower levels of life satisfaction. Efforts to reduce social isolation and its negative health impacts must recognise the complexity of experience across contexts and for different populations.