Recent epidemiological studies suggest that the prevalence of bipolar disorder might be misdiagnosed initially as unipolar depression due to the difficulty to detect episodes of hypomania. The Hypomania Checklist (HCL-32), validated in Spanish, is a self-report questionnaire with 32 hypomania items designed to screen for hypomanic episodes.

To examine the prevalence of hypomania in patients with unipolar depression. Corroborate the efficacy of the HCL-32 to detect symptoms of hypomania.

The presence of hypomanic symptoms was assessed by the HCL-32 in a sample of 128 subjects diagnosed with bipolar I disorder (n = 30), bipolar II disorder (n = 1), unipolar depression (n = 57), and anxiety disorder (n = 15) according to DSM-IV-TR criteria. A control group of healthy subjects was selected (n = 25).

The discriminative capacity was analyzed by the ROC curve. The AUC was 0.65 which did not indicate a good capacity. The sensitivity (S), specificity (E) and prevalence (P) of hypomania in unipolar patients for the following cut-off points were :14: S = 81.6%,95%CI(69.8, 93.5); E = 30.1%,95%CI(19.7,40.6); P = 74.1%; 15: S = 77.6%,95%CI(64.9,90.3); E = 37.4%,95%CI(26.3,48.4); P = 67.2%; 16: S = 59.2%,95%CI(44.4,73.9); E = 55.4%,95%CI(44.1,74.0); P = 51.7%; 17: S = 55.1%,95%CI(40.2,70.1); E = 57.8%,95%CI(46.6,69.1); P = 48.3%.

The HCL-32 has a high sensitivity but a low specificity as screening instrument. This might explain the high proportion of hypomania found in this study. The difference with previous studies is that our sample was heterogeneous, unstable and serious. This suggests that the HCL-32 is not valid for any psychiatric sample. Future research should develop more specific instruments with better external validity.