The prevalence, odds and predictors of lifespan comorbid eating disorder among people with a primary diagnosis of bipolar disorders, and vice-versa: Systematic review and meta-analysis

Background: There are scarce and discrepant data about the prevalence and correlates of co-occurring eating disorders (EDs) among people with a primary diagnosis of bipolar disorder (BD), and vice-versa, compelling a systematic review and meta-analysis on the matter. Methods: MEDLINE/PsycINFO databases were systematically searched for original studies documenting BD⇌ED comorbidity across the lifespan, from inception up until April 20th, 2020. Random-effects meta-analysis and meta-regression analyses were conducted, accounting for multiple moderators. Results: Thirty-six studies involved 15,084 primary BD patients. Eleven studies encompassed 15,146 people with primary EDs. Binge eating disorder (BED) occurred in 12.5% (95%C.I.=9.4-16.6%, I2=93.48%) of BDs, while 9.1% (95%C.I.=3.3-22.6%) of BEDs endorsed BD. Bulimia Nervosa (BN) occurred in 7.4% (95%C.I.=6-10%) of people with BD, whereas 6.7% (95%C.I.=12-29.2%) of subjects with BN had a diagnosis of BD. Anorexia Nervosa (AN) occurred in 3.8% (95%C.I.=2-6%) of people with BDs; 2% (95%C.I.=1-2%) of BD patients had a diagnosis of AN. Overall, BD patients with EDs had higher odds of being female vs. non-ED controls. Several moderators yielded statistically significant differences both within- and between different types of BDs and EDs. Limitations: Scant longitudinal studies, especially across different EDs and pediatric samples. High heterogeneity despite subgroup comparisons. Limited discrimination of the quality of the evidence. Conclusions: The rates of BD⇌ED comorbidity vary across different diagnostic groups, more than they do according to the “direction” of BD⇌ED. Further primary studies should focus on the risks, chronology, clinical impact, and management of the onset of intertwined BD⇌ED across different ages, promoting a continuum approach.