Gender-related patterns of psychiatric disorder clustering among bariatric surgery candidates: A latent class analysis

Abstract Background Psychiatric disorders tend to distribute unevenly in women and men with severe obesity. The current research aimed to identify homogeneous clusters of concurrent psychiatric disorders among patients seeking bariatric surgery, by gender. Methods We recruited a consecutive sample of 393 candidates with obesity (311 women and 82 men) in a university-based bariatric center. Trained clinicians assessed psychiatric disorders through the Structured Clinical Interview for DSM-IV (SCID). Latent class analysis categorized pre-surgical patients into uniform clusters of co-occurring psychiatric disorders. Results For both genders, the 3-class psychopathological clustering was the best-fitting solution. Among women, the latent classes were: (1) “oligosymptomatic”, wherein 42% of patients showed low probability of psychiatric disorders; (2) “bipolar with comorbidities”, in 33%; and (3) “anxiety/depression”, in 25%. Among men, (1) “bipolar with comorbidities” was found in 47% of patients; (2) “oligosymptomatic”, in 40%; and (3) “anxiety/depression”, in 13%. For both genders, the probability of presenting eating disorders was higher in both “bipolar” and “anxiety/depression” classes. Substance use disorders was prominent among “bipolar” men. In comparison with “oligosymptomatic” class, the likelihood of higher BMI was observed among “bipolar” men and poorer work attainment among men with “anxiety/depression”. Limitation Participants was cross-sectionally drawn from a single bariatric center. Conclusions Pre-surgical men and women with severe obesity were distributed in three comorbidity profiles and revealed analogous psychopathological patterns. The class of “bipolar disorders” most likely presented comorbidity with eating and substance use disorder. This natural clustering of psychiatric disorders among bariatric patients suggests gender-related therapeutic approaches and surgical outcomes.