Longitudinal Predictive Validity of the DSM-5 Anxious Distress Specifier for Clinical Outcomes in a Large Cohort of Patients With Major Depressive Disorder