Depressive symptoms, brain volumes and subclinical cerebrovascular disease in postmenopausal women: the Women's Health Initiative MRI Study

Abstract

Objective: Late-life depressive symptoms (DS) increase the risk of incident mild cognitive impairment and probable dementia in the elderly. Our objectives were to examine the relationship between elevated DS and regional brain volumes including frontal lobe subregions, hippocampus and amygdala, and to determine whether elevated DS were associated with increased subclinical cerebrovascular disease in postmenopausal women.

Methods: DS were assessed an average of 8years prior to structural brain MRI in 1372 women. The 8-item Burnam regression algorithm was used to define DS with a cut-point of 0.009. Adjusting for potential confounders, mean differences in total brain, frontal lobe subregions, hippocampus and amygdala volumes and total ischemic lesion volumes in the basal ganglia and the cerebral white and gray matter outside the basal ganglia were compared between women with and without DS.

Results: Depressed women had lower baseline global cognition and were more likely to have prior hormone therapy history. After full adjustment, DS at baseline were associated with smaller superior and middle frontal gyral volumes. Hippocampal and amygdala volumes, and ischemic lesion volumes were similar in depressed and non-depressed women.

Limitations: Depression was not assessed based on semi-structured interview, and MRI scans were obtained cross-sectionally rather than longitudinally. Longitudinal MRI assessments will be necessary to define the temporal relationships between DS and frontal lobe volumes.

Conclusions: Elevated DS were associated with lower volumes in certain frontal lobe subregions but not in the medial temporal lobe structures. Our findings support the role of frontal lobe structures in late-life DS among women.

Figure 1

Adjusted mean middle frontal gyrus (A) and superior frontal gyrus (B) volumes and 95% confidence interval after fully adjusting for potential confounding variables.* Footnote: ^*Confounding variables: Women’s Health Initiative clinical trial, arm, clinical site, age, race/ethnicity, education, body mass index, modified mini-mental state (3MS), smoking, prior use of hormone therapy, antidepressant therapy, cardiovascular disease history, diabetes, hypertension, intracranial volume, and time from randomization to magnetic resonance imaging (MRI) scan

Figure 1

Adjusted mean middle frontal gyrus (A) and superior frontal gyrus (B) volumes and 95% confidence interval after fully adjusting for potential confounding variables.* Footnote: ^*Confounding variables: Women’s Health Initiative clinical trial, arm, clinical site, age, race/ethnicity, education, body mass index, modified mini-mental state (3MS), smoking, prior use of hormone therapy, antidepressant therapy, cardiovascular disease history, diabetes, hypertension, intracranial volume, and time from randomization to magnetic resonance imaging (MRI) scan