Background
The Hypomania Checklist (HCL-32) was developed to identify subthreshold bipolarity in patients with major depression. An HCL-32 version with fewer items has been suggested.

Methods
Principal component analysis (PCA) without rotation was used to identify active/elevated mood versus risk-taking/irritable behaviour in the HCL-32. Using the Bech–Rafaelsen Mania Scale as index of clinical validity a shorter version was developed. Item response theory analysis was used to evaluate whether the total score of the HCL-32 was sufficient to measure subthreshold bipolarity. The short 13-item Mood Disorder Questionnaire (MDQ) was used for comparison.

Results
In accordance with the SCID-II criteria, we included 59 bipolar I and 63 unipolar (depressed) outpatients who had recently been discharged from inpatient treatment. In the HCL-32, PCA identified the two contrasting factors: active/elevated mood versus risk-taking/irritable behaviour. The clinical validation analysis focussed on 20 HCL items as the most acceptable (HCL-20). Item response analysis accepted that the total scores of the HCL-32/HCL-20 were a sufficient statistic, as was the total score of the MDQ. Among the unipolar (depressed) patients not responding to their antidepressive medication, subtreshold bipolarity was identified in 55% of patients using the HCL-20, 36% using the HCL-32, but only 18% using the MDQ.

Limitations
Only outpatients recently discharged from inpatient treatment were studied. A further limitation is that 9.5% of the unipolar patients had only suffered from one episode, which, however had led to hospitalisation.

Conclusion
The HCL-20 was found to identify subthreshold bipolarity in up to 55% of inpatients with major depressive disorder not responding to antidepressive medication.