P39 Changes in loneliness and social isolation over time in adults aged over 50: the English Longitudinal Study of Ageing

Background Epidemiological studies commonly distinguish between transient and chronic loneliness, but little is known about how these experiences change over time and how they relate to the more objective experience of social isolation. We investigated how reports of loneliness and social isolation evolve over a period of twelve years, using data from the first six waves of the English Longitudinal Study of Ageing (ELSA). Methods ELSA is a cohort study which began in 2002 and follows a representative sample of men and women living in England who are aged over 50. Every 2 years, loneliness is measured using the 3-item UCLA Loneliness Scale and two direct questions, and social isolation is captured through questions relating to the availability, frequency and quality of relationships. First, we looked at trajectories of loneliness and social isolation over time, comparing the patterns produced by different measurement tools. We then explored associations between loneliness and social isolation using Pearson’s chi-squared test. Our analyses are based on participants for whom more than one wave of data were available (n = 12,481). Results A majority of people did not experience loneliness or social isolation, with prevalence varying according to the measurement tools used (from 2% using the UCLA 3-item scale to 14% using the CESD single-item question). Reports of loneliness and social isolation fluctuated significantly over time. Patterns of variation included moving in and out of loneliness and/or social isolation, either simultaneously or independently, and prolonged exposure to loneliness or isolation. Our presentation will include a discussion of factors (age, gender, health status and socio-economic status) associated with different social isolation and loneliness patterns. Conclusion Feeling lonely or being isolated are not static experiences. Our findings suggest that future epidemiological studies should consider loneliness and social isolation as time-varying factors, so as to explore the implications of different social relationship trajectories, including prolonged exposure to isolation, or changes in loneliness. Theoretically robust and practical approaches to the measurement of loneliness and isolation are essential, to enable evaluation of interventions that address social relationships.