Abstract / Summary
Abstract Background Substance use disorders (SUDs) frequently co-occur with bipolar disorder (BD) in clinical settings and are associated with poorer outcomes, yet hospital-based data from sub-Saharan Africa are scarce. We determined the prevalence of SUDs and identified associated factors among adults with BD in Uganda. Methods This cross-sectional study consecutively enrolled 404 adults (≥ 18 years) with BD confirmed using the Mini-International Neuropsychiatric Interview (MINI) from the inpatient and outpatient units of Butabika National Referral Mental Hospital. SUDs were defined as moderate- or high-risk scores on the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST). Sociodemographic, clinical, psychosocial, and environmental data were collected by structured questionnaire. Factors independently associated with any SUD were identified using multivariable binary logistic regression. Results The prevalence of any SUD was 22.5% (95% confidence interval [CI] 18.7–26.8), with alcohol use disorder most common (17.3%), followed by cannabis (9.7%), tobacco (9.4%), and khat (2.7%). Easy access to substances (adjusted odds ratio [aOR] 25.55, 95% CI 12.84–50.84), family history of SUD (aOR 11.41, 95% CI 4.94–26.31), and high suicidality risk (aOR 3.15, 95% CI 1.40–7.10) were independently associated with higher odds of any SUD, whereas female sex (aOR 0.31, 95% CI 0.18–0.52) and a social support system (aOR 0.20, 95% CI 0.08–0.52) were protective. Illness-course indicators were not independently associated with SUD. Conclusions Nearly one-quarter of adults with BD had a co-occurring SUD, predominantly alcohol. Because illness severity did not predict SUD, routine ASSIST screening is warranted for all patients, with concurrent suicidality assessment and particular attention to those reporting easy access or a family history of SUD. The findings support integrating brief, low-resource interventions and substance-availability measures into BD care in Uganda.