Body weight and reported versus measured weight loss as confounders of the dexamethasone suppression test

We examined the association between post-dexamethasone suppression test (DST) serum cortisol and body weight, self-report of weight loss during the episode, and measured weight loss during the first week of admission in a series of 245 depressed inpatients. Data on measured weight loss between two successive admissions was available in a group of 57 depressed inpatients. Reported weight loss during the episode and measured weight loss during the first week of admission were not related to DST nonsuppression. In contrast, DST nonsuppression was significantly more frequent in patients with measured weight loss between two successive admissions. This association was particularly strong in patients with below-average body weight and was practically nonexistent in patients with above-average body weight. Multivariate analysis indicates that a significant association between DST results and weight loss may be missed if self-report is substituted for direct measurement of weight loss and if potential confounders, such as total body weight, age, and sex, are ignored.