The duration of undiagnosed bipolar disorder: Effect on outcomes and treatment response

Introduction There are commonly long delays between the onset of bipolar disorder (BP), seeking of treatment and acquiring a bipolar disorder diagnosis. Whether a longer duration of undiagnosed bipolar disorder (DUBP) leads to an inferior treatment response is unclear in the literature. Method We conducted two studies with independent samples of BP patients who had received a first-time diagnosis of BP - first investigating whether DUBP was related to clinical and social outcomes at the time of assessment (n=173) and, second, whether response to mood stabiliser medication was affected by DUBP when assessed three months following assessment and intervention (n=64). Results Participants' mean DUBP was 18-20 years (from the onset of mood episodes). After controlling for age, a longer DUBP was associated with employment difficulties, whereas a shorter DUBP was associated with a history of engaging in self-harm behaviours, as well as a reduced likelihood of experiencing social costs as consequence of the mood disorder. The majority of study variables were statistically unrelated to DUBP. In a multivariate analysis, age was the only predictor variable to make a significant contribution to the DUBP (33%). Across the 3-month intervention period, participants improved significantly on all but one outcome measure. The participants' likelihood to improve, become worse or experience minimal/no change over the study period was not significantly related to the DUBP. Limitations Self-reporting poses a risk to measurement precision. Being a naturalistic observation, no specific dose of medication was prescribed. The small sample of BP I patients provided insufficient statistical power to undertake meaningful separate analyses of the BP I and BP II participants. Conclusion Early detection and intervention remains important for helping to reduce morbidity and risks associated with untreated BP. However, the variation in DUBP was mostly a function of age and did not substantially affect clinical status at assessment, or lead to an inferior response to mood stabilising medication. © 2014 Elsevier B.V