Different linkage of depression to hypercortisolism early versus late after stroke. A 3-year longitudinal study.

BACKGROUND AND PURPOSE Using the dexamethasone suppression test, we studied the suppressibility of the cortisol axis and its clinical determinants at various time points after stroke. A major aim was to examine the dexamethasone test as a diagnostic tool for the diagnosis of major depression in stroke patients. METHODS The dexamethasone suppression test, major depression, functional ability, and disorientation were assessed in a cohort of 70 patients with acute stroke and after 3 months (n = 63) and 3 years (n = 43). RESULTS Early after stroke, 24% of the patients were nonsuppressors, with about the same proportion at 3 months (22%) and 3 years (21%). None of the controls (17 healthy elderly volunteers) were nonsuppressors. High cortisol levels early after stroke were significantly associated with functional impairment (r = 0.35; p = 0.003) and disorientation (r = 0.27; p = 0.03). Three years after stroke, high postdexamethasone cortisol levels were significantly associated with major depression (r = 0.57; p < 0.001). The sensitivity of the dexamethasone test was 70% and the specificity 97%. In a longitudinal analysis of the long-term survivors (n = 42), postdexamethasone cortisol values at 3 months predicted major depression at 3 years. CONCLUSIONS Hypercortisolism is associated with major depression late (3 years) but not early (0-3 months) after stroke. Patients with hypercortisolism 3 months after stroke are at risk of major depression later in the course and warrant careful follow-up from a psychiatric viewpoint.