Risque cardiovasculaire et dépression : interactions et associations dans la cohorte CONSTANCES (CARDIAC)

Background

Major depression is one of the leading causes of disability worldwide and is associated with an increased risk of cardiovascular (CV) mortality. The underlying mechanisms of this association remained poorly understood. Previous studies may have overlooked some potential confounding or mediating factors, such as hazardous health behaviors (e.g. at-risk diet, lack of physical activity, tobacco, cannabis and alcohol consumption) or psychotropic drugs, as well as the moderating role of socio-economic status (SES).

Objectives

The present project will take advantage of the CONSTANCES Cohort Study to examine:

1) The cross-sectional associations between depression and the following CV risk factors among individuals free of CV diseases: at-risk diet, obesity, dyslipidemia, diabetes, hypertension, resting heart rate, lack of physical activity, tobacco, cannabis and alcohol consumption.

2) The cross-sectional associations between depression and the presence of a metabolic syndrome among individuals free of CV diseases.

3) The cross-sectional associations between depression and the general CV risk among individuals free of CV disease.

4) The longitudinal associations between depression at baseline and the incidence of metabolic syndrome and CV diseases at follow-up.

5) The potential role of hazardous health behaviors (i.e. at-risk diet, lack of physical activity, tobacco, cannabis and alcohol consumption) or psychotropic drugs in mediating or confounding the association between depression and metabolic syndrome, general CV risk or incident CV diseases (see objectives 2, 3 and 4).

6) The potential moderating role of SES on the associations between depression and CV risk factors, metabolic syndrome, general CV risk or incident CV diseases (see objectives 1, 2, 3 and 4).

Methods

Depressive states will be measured at inclusion by the CES-D scale. The regular use of the following medications will be extracted from the SNIIRAM data from 4 years before baseline: lipid-lowering drugs, anti-hypertensive drugs, antidiabetic drugs, antidepressant drugs and antipsychotic drugs. CV risk factors will be collected through questionnaire and biometric and biological data. Based on a qualitative food frequency questionnaire, a principal component analysis will identify dietary patterns. Dyslipidemia, diabetes and hypertension will be inferred from questionnaires, medications and biological measures. Metabolic syndrome will be defined according to the 2005 revised NCEP ATP III criteria and general CV risk will be computed according to the calibrated 2008 Framingham score and the 10-year French CHD risk equation. The moderating role of 3 SES proxies will be examined: education, occupational status and income. Associations between depressive symptoms and CV risk factors will be examined among participants free of CV diseases at baseline with general linear model or binary logistic regression for continuous or binary variables. Associations between depressive symptoms and incident CV diseases will be examined with Cox’s regression models. Should a significant interaction be found between depressive symptoms and SES as regards a specific CV risk factor or disease, post hoc subgroup analyses will be performed to describe this interaction in details. The hypothesis of a potential confounding or mediating role of any factor will be tested through sequential adjustments. The percentages change of the depression’s OR after inclusion of additional variables to the models will be calculated as well as 95% CI using a bootstrap method.

Perspectives

A better understanding of the potential mediating, confounding or moderating factors of the association between depression and CV diseases is critical as it could ultimately inform targeted preventive interventions, with expected improvement in public health through policy making, as well as in personalized medicine, through a better stratification of the CV risk.