Cerebral Hemodynamics and Incident Depression: The Rotterdam Study

Background: According to the vascular depression hypothesis, subclinical cerebrovascular disease can cause depression in older adults. To test this hypothesis, several cross-sectional studies have assessed structural brain parameters, but few have examined hemodynamic alterations in the brain. Methods: From the Rotterdam Study, we studied a cohort of 1494 participants (65+ years of age) free of depression, dementia, and stroke at baseline. In the middle cerebral artery blood flow velocities and vasomotor reactivity were measured with transcranial Doppler ultrasonography. All participants were repeatedly assessed for depressive symptoms with Centre for Epidemiological Studies-Depression scale (CES-D). Participants with depressive symptoms (CES-D >= 16) had a semi-structured interview, to classify the depression according to DSM-IV criteria. All analyses were adjusted for sociodemographic data, vascular risk factors, and incident stroke. Results: Lower peak-systolic, end-diastolic, and mean blood flow velocities at baseline were associated with higher CES-D scale scores at follow-up. Mean blood flow velocity predicted incident depressive symptoms (odds ratio [OR]:.74, 95% confidence interval [CI]:.60-.91, p = .004) and depressive disorders (OR:.83, 95% CI:.69-.98, p = .032), whereas decreased baseline vasomotor reactivity predicted incident depressive disorders only (OR:.66, 95% CI:.53-.83, p = .001). Conclusions: Lower blood flow velocity, indicating reduced cerebral metabolism, predicted depressive symptoms and depressive disorders. Reduced vasomotor reactivity, which might indicate cerebral microangiopathy, predicted depressive disorders only, in healthy older adults. These findings provide prospective evidence for vascular depression hypothesis.