Abstract
Substance use disorders (SUDs) are common in those with bipolar disorder (BD) and are linked to a number of negative outcomes. The aim of the present study was to systematically review and perform a meta-analysis on correlates of SUDs in those with BD. A systematic review was performed on the online database PsychINFO, with specific keywords and inclusion criteria employed. A total of 1794 papers were produced and 49 were accepted. A meta-analysis was performed on all available data. Correlates of SUD reported by individual studies included various socio-demographic, clinical history, developmental, personality and family history variables. A series of meta-analyses found that lower current age and lower age ofsymptom onset, male gender, being unmarried, having BD type I, having a co-morbid anxiety disorder and a history of suicide attempts all predicted co-morbid SUD. Variables such as years spent in education, number of lifetime psychiatric hospitalisations and ethnic minority status did not reach significance. These results demonstrate a number of correlates of SUD in BD. These may enable the identification of high risk individuals so that interventions to prevent the development of SUD can be employed.