To the Editor: A simpler explanation of the data in the articles by Alessandro Rotondo, M.D., et al. (1) and Dean F. MacKinnon et al. (2) was not disqualified. It is that panic disorder is part of a larger anxiety picture that includes generalized anxiety disorder or posttraumatic stress disorder (PTSD). The symptoms of these disorders can mimic bipolar disorder and thereby produce the appearance of an association between bipolar disorder and panic disorder. These reports did not exclude influences from these disorders.

DSM-III-R and DSM-IV criteria for a manic episode overlap with those of generalized anxiety disorder. Criterion A, a 1-week period of irritable mood, is met by the irritability of anxiety. Criterion B can be met by any four of these common anxiety symptoms: expressions of entitlement, habitual insomnia, pressure to ventilate, subjective thought racing, distractibility from anxiety, agitation, and indulgences to compensate for feelings of deprivation. The remaining criteria concern exclusions and severity. Meeting the criteria for a manic episode in this way is a technicality rather than evidence of bona fide mania. There is similar overlap in the criteria for anxiety disorders and a hypomanic episode. The risk of false positive identification of mania would be lower if relatively specific signs were required, e.g., observable euphoria or derailment; however, the analyses were not restricted to patients with such signs. One of us (C.M.S.) has seen dozens of patients who had been diagnosed as having bipolar disorder but whose symptoms were explained by PTSD or generalized anxiety disorder.

By this reasoning, the comorbidity of panic disorder with generalized anxiety disorder (3, 4) should explain at least some of the apparent association between panic disorder and bipolar disorder. A link from panic disorder to other anxiety disorders is simpler and more ordinary than a link to bipolar disorder; by Occam’s razor, it takes precedence unless disproved.