The article by Angst and colleagues in this issue (1) proposes an additional disorder within the spectrum of bipolar disorders, involving people who have major depressive disorder and subthreshold hypomania symptoms. This condition differs from major depression in having more familial bipolar disorder, a younger age at onset, more severe symptoms, higher rates of comorbidity, and more depressive episodes. However, lifetime and 12-month histories of treatment as well as level of disability do not differ from those of major depression. Compared with patients who have bipolar I and II disorders, this new group manifests less severe manic symptoms, less disability, and less treatment (both lifetime and 12-month).

This epidemiological study, based on data from the National Comorbidity Survey Replication, was carefully done, with a large sample, using state-of-the-art sampling methods and data collection instruments. The authors findings support other surveys as well as clinical descriptions. From a Linnean standpoint, the validity of this new disorder (perhaps a "bipolar III" disorder) holds up well: it bears greater genetic loading, pathology, and morbid prognosis than does major depression, yet less morbidity and treatment than bipolar I and II disorders.

Clinicians might accept this new diagnosis for its ability to set a more accurate prognosis than previously possible. However, to truly welcome a new entity to their daily practice, clinicians may want more information, such as the following:

Cross-sectional epidemiological studies alone cannot provide this critical information. Controlled treatment studies, long-term prospective studies, and cost- effectiveness studies are needed. Although much work remains to be done, we owe a great debt to Angst and colleagues for bringing us this far.

Implications.

We are in a period when new technologies and more refined studies are identifying ever more new disorders. For example, a recent report suggests that major depression in the presence of "hyperthymic temperament" may comprise a distinct bipolar IV condition (2). Some psychiatrists are growing alarmed at this apparent propensity to create ever more diagnoses. Will each new edition of DSM produce a geometric increase in the number of psychiatric disorders? Will an exponential growth in diagnoses lead to greater confusion or to greater clarity in our clinical work? Will we become overwhelmed with a focus on subsyndromal "trees" while rendering ourselves incapable of comprehending the overall "forest"?

Several countervailing trends offer promise in balancing the dizzying array of new disorders, subdisorders, and sub-subdisorders. These trends (not in any order of implied significance) include the following:

We are entering an exciting period of change in psychiatric nosology. A tension exists between, on the one hand, the potential for a psychiatric Tower of Babel, in which categories proliferate so greatly as to undermine our ability to apply them reliably, and on the other hand, the potential of these new categories to inform and improve our care of people who are distraught, disabled, demoralized, or even destroyed by their conditions. We can—and I am confident that we will—fi nd sensible means of integrating these innovations into our daily work.