A controlled study of MRI signal hyperintensities in older depressed patients with and without hypertension

OBJECTIVES: To compare the frequency/severity of signal hyperintensities-li kely markers of cerebrovascular disease-in the subcortical gray and deep wh ite matter on magnetic resonance imaging (MRI) scans of brains of hypertens ive and normotensive older depressed and nondepressed comparison subjects. DESIGN: Between-groups comparison of cross-sectional MRI data employing ana lyses of covariance controlling for the effects of age, gender, and height. SETTING. A comprehensive inpatient-outpatient geriatric psychiatry service in a university hospital. PARTICIPANTS: Nondemented older depressed (n = 81) and nondepressed compari son (n = 70) subjects divided into four groups (hypertensive depressed (n = 40), hypertensive normals (n = 21), normotensive depressed (n 41), normote nsive normals (n = 49)). MEASUREMENTS: Signal hyperintensities, were rated on T-2 weighted MRI scans blind to patient diagnoses employing two standardized hyperintensity ratin g systems (Fazekas, Boyko). RESULTS: Hypertensive depressives had significantly more-severe hyperintens ity ratings in both subcortical gray and deep white matter than did normote nsive depressives and controls (P < .05) and significantly more-severe hype rintensity ratings only in subcortical gray matter (P < .05), than did hype rtensive controls. Hypertensive controls had significantly more-severe rati ngs in deep white matter than either normotensive group (P < .05). CONCLUSIONS: Findings suggest a relationship between deep white matter hype rintensities and hypertension (regardless of depressive state), and a parti cular role of subcortical gray matter hyperintensities (possibly interactin g with more-severe deep white matter lesions) in older depressed hypertensi ves, as compared with older depressed normotensives of similar ages and sev erity of depression. These data support possible heterogeneous pathogenic c ontributions in late-life depression subgroups, one of which appears to be influenced by cerebrovascular disease.