Does comorbid subthreshold anxiety predict treatment response in depression? Results from a naturalistic cohort study (the CRESCEND study)

Objective To investigate whether the anxious depression defined as depression with clinically significant anxiety but not comorbid anxiety disorder predicts poor outcomes of depression treatment in naturalistic clinical setting. Method From nationwide sample of 18 hospitals, 674 patients with moderate to severe depression who completed the DSM-IV-based Structured Clinical Interview (SCID) were recruited. Anxious depression was defined as not having comorbid anxiety disorder by SCID and having a Hamilton Rating Scale for Anxiety (HAM-A) total score ≥20. Participants were classified into three groups: anxious depression (N=259), non-anxious depression (N=351), or comorbid anxiety disorder (N=64). Rates of and time to remission and response and changes in scale scores were compared between these groups during 12 weeks treatment with antidepressant interventions freely determined by clinicians. Results No significant differences were observed in the Hamilton Rating Scale for Depression (HAM-D) remission rate and the time to achieve HAM-D remission between anxious and non-anxious depression after adjustment for variables is not equally distributed at baseline. There were also no significant differences in HAM-D and HAM-A response rate and time to responses between two groups. Patients with comorbid anxiety disorder showed less improvement on HAM-D and HAM-A score than did those with anxious depression despite similar baseline symptom severity. Limitation This study was observational, and the treatment modality was naturalistic. Conclusions Anxious depression did not predict worse outcome to antidepressants treatment. This finding might result from exclusion of comorbid anxiety disorder from anxious depression population and allowance of broad treatment modality.