Interventions to address feelings of loneliness: A systematic umbrella review
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Rationale: Loneliness is being increasingly recognised as a public health concern with serious adverse consequences. Given its prevalence, it is important to understand what kind of strategies are effective for reducing loneliness and for whom.Objective: This umbrella review provides a comprehensive overview of the effectiveness of loneliness interventions to date, summarising evidence from 79 systematic reviews and meta-analyses. Method: A systematic search for systematic reviews and meta-analyses on loneliness interventions was conducted using PsychINFO and MEDLINE in November 2024. Only studies focusing specifically on loneliness were included. No exclusion criteria on target group, intervention type or study design were applied to get the broadest possible evidence base. Results: The evidence on the effectiveness of loneliness interventions is mixed, with some intervention types working better than others. In line with the existing literature, we find that psychological interventions show the strongest and most consistent effects in alleviating loneliness across population groups. Providing social support or promoting social connection also show some positive impact, though their effects are generally weaker than those of psychological approaches. Interestingly, multicomponent interventions yield consistently positive results across studies but have mostly been conducted with samples of older adults. A number of limitations of the existing literature were noted, including some methodological weaknesses as well as low geographical and demographic heterogeneity. Policy implications: These findings enhance our understanding of which interventions to alleviate loneliness have the strongest potential and for whom, thus are of importance to policy-makers and practitioners focusing on this issue.