A systematic review of digital interventions for loneliness and social isolation

Abstract As loneliness and social isolation emerge as pressing public health concerns, identifying effective interventions is essential. Digital solutions offer flexible, scalable, and cost-effective approaches, yet their effectiveness remains uncertain. This systematic review and meta-analysis assess the impact of randomized controlled trials (RCTs) on digital interventions to reduce loneliness and social isolation. Following PRISMA 2020 guidelines, we searched seven databases and grey literature (2022-) and applied random-effects models to pool effect sizes by intervention type. A total of 40 RCTs involving 6,062 participants were included, with one-third focusing on younger individuals. Loneliness was assessed in 36 studies, while eight examined social isolation. Interventions were classified as psychological (k = 25), social (k = 4), activity-based (k = 4), robot-based (k = 4), and social media reduction (k = 5). Psychological interventions-especially those with group or social components-along with group-based activities and robotic pets, were effective in reducing loneliness. In contrast, social contact interventions, self-guided individual activities, and conversational robots showed limited impact. Social media reduction interventions suggested potential benefits, though results were not statistically significant. The evidence base exhibited moderate to high risk of bias, heterogeneity, and limited long-term follow-up. We provide specific recommendations for future interventions and research, including leveraging digital technologies for enhanced personalization, using digital tools for signposting non-digital interventions, systematically comparing digital and non-digital versions of the same intervention, and, most critically, examining the impact of increasingly popular AI-driven and humanlike social chatbots. Key messages • Psychological and group-based digital interventions - especially those incorporating social elements or robotic pets - are the most effective at reducing loneliness. • Some digital approaches, such as self-guided activities, conversational robots, and social contact interventions, show limited effectiveness, and results for social media reduction are inconclusive.