Relationship between atherosclerosis and late-life depression. The Rotterdam Study
Tiemeier H, van Dijck, Hofman A, et al.;
Commented by , 20 Apr 2004
Background
Depression in late life has been associated with vascular abnormalities. Individuals with brain infarcts along with subtle white-matter and deep gray-matter abnormalities have higher rates of depression, the basis of the "vascular depression hypothesis". The association between depression and coronary artery disease has also been consistently demonstrated. Direct measures of atherosclerosis, however, have not been evaluated in late-life depression
Purpose
Examine the association between atherosclerosis at different locations and depression.
Methods
The Rotterdam study, a cross-sectional population-based cohort study. 4019 subjects > or = 60 years, had non-invasive assessments of atherosclerosis in different locations: common carotid intima–media thickness, carotid artery plaques, ankle-brachial blood pressure index, and aortic atherosclerosis. An overall measure of extracoronary atherosclerosis was obtained in 3747 subjects. In a subgroup of 1986, coronary clacifications were used as a measure of coronary atherosclerosis.
Possible confounders included: age, sex, cognitive function, education, antidepressants, smoking, cholesterol, body mass index, blood pressure, diabetes, past myocardial infarction.
All subjects were screened for depression with the Center for Epidemiological Studies Depression Scale (CES-D). Individuals scoring > or = 16 were assessed for DSM-IV depressive disorders using the semi-structured Present State Examination. The relationship of atherosclerosis measures with depressive disorders and subthreshold depressive symptoms (CES-D > or = 16 alone) was evaluated by logistic regression, adjusting for potential confounders.
Results
A depressive disorder (major or minor depression, dysthymia) was found in 119 cases, and subthreshold depressive symptoms in 118. Individuals with depressive disorders were older (74 vs 72), more likely to be female (72% vs 57%) and more likely to have had a stroke (8% vs 3%).
There was no relationship between any measure of atherosclerosis and subthreshold depressive symptoms. Using the overall measure of extracoronary atherosclerosis, a dose-response relationship was found with depression, adjusted Odds Ratio (OR)=1.29 (1.01-1.68; 95% CI). On all individual extracoronary atherosclerosis measures, more severe atherosclerosis was associated with depressive disorders; however, only a trend level of significance was achieved. The strongest association was for aortic calcifications (p=0.06); severe aortic calcifications, adjusted OR= 2.0 (1.02-3.9; 95% CI).
There was a strong dose-dependent relationship between coronary calcifications and depressive disorders (p=0.004), but not subthreshold depressive symptoms. For severe coronary calcifications, OR = 3.89 (1.55-9.77; 95% CI).
Discussion
In summary, a relationship between extracoronary atherosclerosis and depressive disorders was only found when using the combined measure of atherosclerosis, and this association was modest. The more convincing association was between coronary atherosclerotic disease (CAD) and depression, a well-established association from prospective epidemiological research over the past twenty years.
The authors suggest their findings are consistent with the "vascular depression hypothesis" (ref. 1), which proposes cerebrovascular disease as the pathophysiologic basis of late-life depression. With the exception of the higher stroke rates observed in depressive disorders, there was in fact no evidence of cerebrovascular disease from their measures of atherosclerosis. Neuroimaging data would have valuable in exploring this important question.
As the authors acknowledge, the cross-sectional nature of the study limits causal inference. In addition, important confounders such as the impact of disability from severe atherosclerosis on depression, were not addressed.
The research adds further support to the well-established epidemiologic association between CAD and depression, however, little direct support for the vascular depression hypothesis can be drawn.
References
1. Alexopoulos A, Meyers BS, Young RC. Clinically defined vascular depression. American Journal of Psychiatry 1997;154:562-565