Depression as a proxy of negative affect? A critical examination of the use of the CES-D in type 2 diabetes

Objectives. Recent research suggests depression as measured by self-endorsed symptoms lists is associated with poor health outcomes in chronic illness. Yet, it is probable that these lists of symptoms reflect other concepts such as general distress or negative affect when used as dimensions. Methods. To test for this hypothesis, we explored associations of the Centre of Epidemiological Studies-Depression, Radloff ([CES-D], 1977) with disease severity in diabetes and how trait negative affect from the Profile of Mood States ([POMS]; Usala & Hertzog, 1989; adaptation by Cohen et al., 1995) impact these associations in a sample of 502 people with type 2 diabetes. Results. We found that the CES-D included two dimensions of negative and positive experience. Each CES-D component was independently linked to disease severity. However, controlling for trait negative affect suppressed the correlation between the CES-D negative experience component and disease severity. Item-level analyses revealed that the negative experience component of the CES-D bore an emotional tone of sadness but not anger. Conclusions. When using the CES-D, distinguishing positive and negative components is necessary. Self-reported depression symptoms from the CES-D have no incremental validity over negative affect. (C) 2009 Elsevier Masson SAS. All rights reserved.