Clinical risk factors for bipolar disorders: A systematic review of prospective studies

Background Early phases and suspected precursor states of bipolar disorder are not well characterized. We evaluate the prevalence, duration, clinical features and predictive value of non-affective psychopathology as clinical risk factors for bipolar disorder in prospective studies. Methods We screened PubMed, CINAHL, PsycINFO, Embase, SCOPUS, and ISI-Web of Science databases from inception up to January 31, 2014, following PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and searched: bipolar disorder AND [antecedent* OR predict* OR prodrom* OR prospect* OR risk*] AND [diagnosis OR development]. We included only English language reports on prospective, longitudinal studies with two structured clinical assessments (intake and follow-up); no DSM intake diagnosis of bipolar-I or -II; diagnostic outcome was bipolar-I or -II. Details of study design, risk factors, and predictive value were tabulated. Results We found 16 published reports meeting selection criteria, with varying study design. Despite heterogeneity in methods, findings across studies were consistent. Clinical risk factors of bipolar disorder were early-onset panic attacks and disorder, separation anxiety and generalized anxiety disorders, conduct symptoms and disorder, ADHD, impulsivity and criminal behavior. Limitations Since risk factors identified in some prospective studies are predictive of other conditions besides bipolar disorder, these preliminary findings require replication, and their sensitivity, specificity and predictive value need to be assessed. Conclusions Clinical risk factors for bipolar disorder typically arise years prior to syndromal onset, include anxiety and behavioral disorders with unclear sensitivity and specificity. Prospectively identified clinical risk factors for bipolar disorder are consistent with retrospective and family-risk studies. Combining clinical risk factors with precursors and family-risk may improve early identification and timely and appropriate treatment of bipolar disorder. © 2014 Elsevier B.V.