Background: Due to concerns about overlapping symptomatology between medical conditions and depression, the validity of the beck depression inventory (BDI-II) has been assessed in various medical populations. Although major depressive disorder (MDD) and primary insomnia (PI) share some daytime symptoms, the BDI-II has not been evaluated for use with insomnia patients.
Method: Participants (N=140) were screened for the presence of insomnia using the Duke structured clinical interview for sleep disorders (DSISD), and evaluated for diagnosis of MDD using the structured clinical interview for DSM-IV-TR (SCID). Participants' mean BDI-II item responses were compared across two groups [insomnia with or without MDD) using multivariate analysis of variance (MANOVA), and the accuracy rates of suggested clinical cutoffs for the BDI-II were evaluated using a receiver operating characteristic (ROC) curve analysis.
Results: The insomnia with depression group had significantly higher scores on several items; however, the groups did not differ on insomnia, fatigue, concentration problems, irritability, libido, increased appetite, and thoughts relating to suicide, self-criticism and punishment items. The ROC curve analysis revealed moderate accuracy for the BDI-II's identification of depression in those with insomnia. The suggested BDI cutoff of >or=17 had 81% sensitivity and 79% specificity. Use of the mild cutoff for depression (>or=14) had high sensitivity (91%) but poor specificity (66%).
Conclusion: Several items on the BDI-II might reflect sleep disturbance symptoms rather than depression per se. The recommended BDI-II cutoffs in this population have some support but a lower cutoff could result in an overclassification of depression in insomnia patients, a documented problem in the clinical literature. Understanding which items discriminate insomnia patients without depression may help address this nosological issue.