Michele Fornaro, Chiara Di Lorenzo, Francesco Monaco, Annarita Vignapiano, Luca Steardo
The review distinguishes comparatively credible prevalence findings from substantially less certain evidence on risk factors and outcomes. Its scope is limited to AN, BN, and BED and does not represent the full spectrum of feeding and eating disorders. The findings support multidisciplinary assessment while highlighting priorities for subtype-specific, longitudinal, and methodologically rigorous research.
OBJECTIVE: Anorexia nervosa (AN), bulimia nervosa (BN), and binge-eating disorder (BED) frequently co-occur with psychiatric and general medical conditions. The present umbrella review synthesized and graded the credibility of the corresponding meta-analytic evidence.
METHOD: We included systematic reviews with meta-analysis of observational studies reporting prevalence, risk factors, or outcomes associated with psychiatric/general medical multimorbidity in AN, BN, or BED, and the reverse associations. PubMed/MEDLINE, Scopus, Embase, PsycINFO, ProQuest, Google Scholar, and Web of Science were searched from inception to June 28, 2026, without language restrictions to identify journal publications and eligible grey-literature reviews. Meta-analytic estimates were recalculated when primary data were available and graded using established or adapted credibility criteria; methodological quality was assessed with AMSTAR-2.
RESULTS: Sixty-one meta-analyses yielded 213 estimates. Of 134 prevalence estimates, 91 (67.9%) met moderate-to-strong credibility criteria. Prominent estimates included any personality disorder in AN (49%; 95% CI = 42-56%) and BN (54%; 47-61%), non-suicidal self-injury in BN (37%; 31-43%), tobacco use disorder in BED (47.7%; 34.8-61.0%), and bipolar disorder in BED (9.1%; 3-23%). No re-pooled risk-factor or outcome association reached convincing or highly suggestive credibility.
CONCLUSIONS: The review distinguishes comparatively credible prevalence findings from substantially less certain evidence on risk factors and outcomes. Its scope is limited to AN, BN, and BED and does not represent the full spectrum of feeding and eating disorders. The findings support multidisciplinary assessment while highlighting priorities for subtype-specific, longitudinal, and methodologically rigorous research.