To the Editor.
—Reding et al1 have reported on dexamethasone suppression tests (DSTs) in stroke, and found that they correlate with depression but are not predictive of rehabilitation outcome. Unfortunately, their study has not specified whether and how patients' depressions were treated. If the depressed patients received treatment, the effect of treatment outcome on rehabilitation outcome would require assessment. If the patients with depression were not treated, ethical questions could be raised.Furthermore, two methodologic points need attention. First, the authors lumped dysthymic disorder and major depressive disorder together as "depression." In most of the psychiatric DST literature, it is major depression that is associated with abnormal DST results. A 49% prevalence of abnormal DST results in dysthymic disorders would be highly unusual. Thus, it would be useful to know how the patients were divided into major depressed and dysthymic groups. Second, the authors employed a 5-μg/dL cutoff for