Cognitive reactivity versus dysfunctional cognitions and the prediction of relapse in recurrent major depressive disorder

Publication Type:

Journal Article


Journal of Clinical Psychiatry, Volume 76, p.e1306-e1312 (2015)

Other Number:



cognitive reactivity; dysfunctional cognitions; relapse; chronic depression; major depressive disorder


OBJECTIVE: Major depressive disorder (MDD) is a burdensome disease that has a high risk of relapse/recurrence. Cognitive reactivity appears to be a risk factor for relapse. It remains unclear, however, whether dysfunctional cognitions alone or the reactivity of such cognitions to mild states of sadness (ie, cognitive reactivity) is the crucial factor that increases relapse risk. We aimed to assess the long-term predictive value of cognitive reactivity versus dysfunctional cognitions and other risk factors for depressive relapse. METHOD: In a prospective cohort of outpatients (N = 116; studied between 2000-2005) who had experienced ≥ 2 previous major depressive episodes (MDEs) and were in remission (DSM-IV) at the start of follow-up, we measured cognitive reactivity, with the Leiden Index of Depression Sensitivity (LEIDS), and dysfunctional cognitions, with the Dysfunctional Attitudes Scale, simultaneously. Course of illness (with the primary outcome of MDE assessed by the Structured Clinical Interview for DSM-IV Axis I Disorders Patient Edition) and time to relapse were monitored prospectively for 3.5 years. RESULTS: Cognitive reactivity scores were associated with time to relapse over the 3.5-year follow-up and also when corrected for the number of previous MDEs and concurrent depressive symptoms (hazard ratio for 1 standard deviation [(HRSD); 20 points of the LEIDS, measuring cognitive reactivity] = 1.47; 95% CI, 1.04-2.09; P = .031). Rumination appeared to be a particularly strong predictor of relapse (HRSD = 1.60; 95% CI, 1.13-2.26; P = .007). Dysfunctional cognitions did not predict relapse over 3.5 years (HRSD = 1.00; 95% CI, 0.74-1.37; P = .93). Every 20-point increase on the cognitive reactivity scale resulted in a 10% to 15% increase in risk of relapse (corrected for previous MDEs and concurrent depressive symptoms). CONCLUSIONS: Cognitive reactivity-and particularly rumination-is a long-term predictor of relapse. Future research should address whether psychological interventions can improve cognitive reactivity scores and thereby prevent depressive relapses. TRIAL REGISTRATION: ISRCTN Identifier: 68246470.