Answers are grounded in this paper and cite their sources.
BACKGROUND: Loneliness in later life is a prevalent psychological experience that reflects both stable individual differences and changing social circumstances, yet little is known about the relative contribution of these components over time. The aims are: a) to partition the variance of loneliness into trait vs. occasion (state); and b) to accommodate covariates related to social contact (received and given help). METHOD: Data comes from three waves of the Survey of Health, Aging and Retirement in Europe. Loneliness was measured with UCLA-R Scale. RESULTS: The trait component accounted for approximately 60% of the latent variance, whereas the occasion component accounted for approximately 40%. The TSO model was extended to include covariates. Received help was positively associated with trait-like loneliness, whereas given help was negatively associated with it. Associations with occasion-specific components were generally weaker and mostly non-significant, with two statistically significant exceptions. CONCLUSIONS: Most of the loneliness seems to be trait-like. The time-varying loneliness was smaller in size, but relevant. Helping exchanges were more consistently associated with the stable component of loneliness than with occasion-specific variability.
No takes yet. Share an insight, caveat, or question.
Galiana et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: