RESUMO
BACKGROUND: Evidence provides inconsistent findings on risk factors and health outcomes associated with loneliness. The aim of this work was to grade the evidence on risk factors and health outcomes associated with loneliness, using an umbrella review approach. METHODS: For each meta-analytic association, random-effects summary effect size, 95% confidence intervals (CIs), heterogeneity, evidence for small-study effect, excess significance bias and 95% prediction intervals were calculated, and used to grade significant evidence (p<0.05) from convincing to weak. For narrative systematic reviews, findings were reported descriptively. RESULTS: From 210 studies initially evaluated, 14 publications were included, reporting on 18 outcomes, 795 studies, and 746,706 participants. Highly suggestive evidence (class II) supported the association between loneliness and incident dementia (relative risk, RR=1.26; 95%CI: 1.14-1.40, I2 23.6%), prevalent paranoia (odds ratio, OR=3.36; 95%CI: 2.51-4.49, I2 92.8%) and prevalent psychotic symptoms (OR=2.33; 95%CI: 1.68-3.22, I2 56.5%). Pooled data supported the longitudinal association between loneliness and suicide attempts and depressive symptoms. In narrative systematic reviews, factors cross-sectionally associated with loneliness were age (in a U-shape way), female sex, quality of social contacts, low competence, socio-economic status and medical chronic conditions. LIMITATIONS: Low quality of the studies included; mainly cross-sectional evidence. CONCLUSIONS: This work is the first meta-evidence synthesis showing that highly suggestive and significant evidence supports the association between loneliness and adverse mental and physical health outcomes. More cohort studies are needed to disentangle the direction of the association between risk factors for loneliness and its related health outcomes.