Comparison of depressive symptoms in type 2 diabetes using a two-stage survey design

Objective: To test the validity of the Patient Health Questionnaire-9 (PHQ-9) in adults with newly diagnosed Type 2 diabetes mellitus (T2DM) and compare the distribution of PHQ-9 items in those with and without Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) depression. Methods: A two-stage survey design was used in primary care centers (n = 96). In Stage 1, participants were administered the PHQ-9 and biopsychosocial measures. In Stage 2, PHQ-9 positives (score ≥10) and randomly selected PHQ-9 negatives (score <10) had a diagnostic interview, Schedule for Clinical Assessment in Neuropsychiatry 2.1. The sensitivity, specificity and receiver operating characteristics of the PHQ-9 were calculated. The means of PHQ-9 items were compared. Results: Stage 1: Prevalence of PHQ-9 positives (completing Stage 2) was 12.1% (n = 182). PHQ-9 positives were younger (mean [standard deviation] age = 52.8 [9.47] versus 56.2 [11.50] years, p < .001) and had higher body mass index (32.7 [7.08] versus 31.6 [6.25] kg/m2, p = .028) than PHQ-9 negatives (n = 1278). Stage 2: Prevalence of DSM-IV depression was 5.6% (n = 84). The optimal cutoff was ≥12 (sensitivity = 86.9%, specificity = 80.3%). PHQ-9 true positives scored significantly higher than PHQ-9 false positives on all items, excluding sleep (mean [standard deviation] score = 2.4 [0.98] versus 2.2 [1.06]), fatigue (2.5 [0.80] versus 2.3 [0.86]), and appetite (1.6 [1.23] versus 1.5 [1.26]). Conclusions: Over-identification of depression by PHQ-9 in T2DM may be driven by presence of symptoms in keeping with an organic origin.