Bipolar II disorder: reasons to recognize.

J Clin Psychiatry 2015;76(2):e222–e223 (doi:10.4088/JCP.14com09639). © Copyright 2015 Physicians Postgraduate Press, Inc. O 40 years ago, clinical observations generated proposals for a distinction among mood-disordered patients between those who experience mania in the course of their illness and those who experience only hypomania.1 Concerns that this distinction could not be made reliably delayed its inclusion in official nomenclature until DSM-IV appeared 25 years later. The availability of other operational definitions and, in particular, the Research Diagnostic Criteria,2 nevertheless fostered an accumulation of findings in support of bipolar II disorder as a valid category. The JCP article by Frankland et al3 is among the most recent of these and is one of the few that have compared symptom profiles of major depressive disorder (MDD) to those of bipolar I depression and bipolar II depression separately. Both of the bipolar depression groups differed from the MDD group in nonoverlapping ways. Thus, contrasts in depressive symptom composition can be added to the other lines of evidence that bipolar I and bipolar II disorders differ. The strongest support for the separation lies in family studies. At least 7 show that bipolar II disorder is more prevalent among the relatives of bipolar II probands than among those of bipolar I probands and, conversely, that bipolar I disorder is more prevalent among the relatives of bipolar I probands.4–10 A prospective component of that study also showed that patients who switch from MDD to bipolar II disorder are more likely to have bipolar II relatives, while those who switch to bipolar I disorder instead have a preponderance of bipolar I relatives.11 Although measures of diagnostic stability are rare, their results can provide robust evidence for validity. Only a few reports have described the likelihood of shifts from bipolar II to bipolar I disorder, but those that have found that this happens in only 5%10 and 7.5%12 of adult samples. A higher rate of 25% reported by the Course and Outcome of Bipolar Youth study13 suggests that diagnostic stability may be lower in child and adolescent populations. Follow-up studies have also revealed differences between the courses of bipolar I and bipolar II disorder, though with less consistency. Some have found that patients with bipolar II disorder experience more depressive morbidity, less time in elevated mood states, more comorbidity, a greater likelihood of suicide attempts, and more psychological impairment.14 A particularly interesting difference between groups is in the likelihood of seasonal patterns. Seasonality among bipolar patients appears to be either confined to,15 or much more prevalent in,16,17 bipolar II groups. Individuals with bipolar II disorder thus may be more photosensitive, and those with bipolar II depression might therefore derive more benefit from light therapy than those with bipolar I depression. Although bipolar depression, with or without seasonal patterns, is often more responsive to phototherapy than is MDD,18 there seems to be no equivalent comparison between bipolar I and bipolar II depression in responsiveness to light therapy. Only a few genomic studies have separated bipolar I disorder from bipolar II disorder, but there are indications that doing so may be very worthwhile. In one, sibling pairs concordant for bipolar II disorder characterized families with evidence of linkage to 18q but not families that did not show linkage.19 More recently, a genomewide association study analysis of a combined cohort revealed that genomewide significance for an association with the adrenomedullary gene on 11p was specific to bipolar II disorder.20 Support for the validity of bipolar II disorder has emerged despite the obvious difficulty of accurately diagnosing hypomania, a syndrome that is rarely current at the time of the diagnostic assessment and that shades into normal periods of excitement or elation. In addition, the boundary between mania and hypomania is largely based on duration, on the need for hospitalization, or on the degree of resulting impairment rather than on symptom number and type. The evidence stands, though, and clearly argues for efforts to characterize how the bipolar I/bipolar II dichotomy might bear usefully on treatment selection. Unfortunately, most large-scale treatment studies are industry sponsored and are designed to achieve an indication for a particular disorder, very rarely for its subtypes. So far, psychiatric drug effects more often generalize across disorders than show specificity to a subtype. Support for the bipolar I/bipolar II distinction seems stronger than that for most other subtypes in psychiatric nosology, though, and therefore may prove a more useful tool for tailoring treatment.