Association between loneliness and mild cognitive impairment in older adults: a meta-analysis of longitudinal studies

Objectives: Prior studies reporting the effects of loneliness on mild impairment cognitive (MCI) have generated inconsistent results. This meta-analysis aimed to investigate the longitudinal association between loneliness and risk of MCI among community-dwelling middle-aged and older adults. Method: Five electronic databases were searched from inception to 9 May 2023. Eligible studies examined the longitudinal association between loneliness and cognitive outcomes, including incident MCI, cognitive impairment, and cognitive decline. Odds ratios (OR) and 95% confidence intervals (CIs) were calculated using random-effects or fixed-effects meta-analysis. Sensitivity analysis and subgroup analysis were conducted. Publication bias was examined using Egger's and Begg tests. Results: Eight studies were included. Among the 45,032 participants, 10,570 were diagnosed with MCI/cognitive decline. Loneliness was positively associated with an increased risk of MCI (overall OR = 1.14; 95% CI = 1.05, 1.23), with moderate heterogeneity (I-2 = 44.2%). Sensitivity analysis have minimal influence on the aforementioned pooled effect. Subgroup analyses indicated stronger associations in studies which employed incident MCI as cognitive outcome (OR = 2.55, 95%CI = 1.31, 1.83), were conducted in non-Asia countries (OR = 1.52, 95%CI = 0.95, 1.20), and reported no depression adjustment (OR = 1.51, 95%CI = 1.04, 1.25). The association between loneliness and MCI was stronger among males compare to females. The Egger test and Begg test showed no evidence of significant publication bias (p =.493; p=.474). Conclusion: The findings indicated that loneliness was associated with an increased risk of MCI. Future longitudinal studies should evaluate potential cases of MCI through comprehensive clinical assessments by practitioners to draw robust findings on the association of loneliness with MCI.