Abstract
Reported associations between depression and myocardial infarction in some studies might be explained by use of psychotropic drugs, residual confounding, and/or reverse causation (whereby heart disease precedes depression). We investigated these hypotheses in a large prospective study of UK women with no previous vascular disease. At baseline in median year 2001 (IQR 2001-2003), Million Women Study participants reported whether or not they were currently being treated for depression or anxiety, their self-rated health, and medication use during the previous 4 weeks. Follow-up was through linkage to national hospital admission and mortality databases. Cox regression yielded adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) for the first myocardial infarction event in those reporting treatment for depression or anxiety (subdivided by whether or not the treatment was with psychotropic drugs) v. not, and stratified by self-reported health and length of follow-up. During mean follow-up of 13.9 years of 690 335 women (mean age 59.8 years) with no prior heart disease, stroke, transient ischaemic attack, or cancer, 12 819 had a first hospital admission or death from myocardial infarction. The aHRs for those reporting treatment for depression or anxiety with, and without, regular use of psychotropic drugs were 0.96 (95% CI 0.89-1.03) and 0.99 (0.89-1.11), respectively. No associations were found separately in women who reported being in good/excellent or poor/fair health or by length of follow-up. The null findings in this large prospective study are consistent with depression not being an independent risk factor for myocardial infarction.
Highlights
Acute myocardial infarction and depression are important causes of morbidity and mortality globally
To enable comparisons to be made between any two groups, even if neither was the reference category, we estimated the Hazard ratios (HRs) for each exposure category with a group-specific 95% confidence interval derived from the variance of the log risk in that particular category (Easton, Peto, & Babiker, 1991; Plummer, 2004) We investigated the potential for proportional hazards violations by examining heterogeneity by the underlying time variable, but the exposure effect did not vary materially by attained age and did not vary materially when all of the covariates were allowed to vary by time
The associations were attenuated still further in the model that adjusted for all of the risk factors [HRs 1.11 and 1.13], and the reduction in the LR χ2 statistic indicates that about 87% of the association found in the minimally adjusted model was explained by deprivation, education, smoking status, alcohol consumption, BMI, physical activity, use of menopausal hormone therapy, and history of hypertension, diabetes, and parental heart disease
Summary
Acute myocardial infarction and depression are important causes of morbidity and mortality globally. Reported associations between depression and myocardial infarction in some studies might be explained by use of psychotropic drugs, residual confounding, and/or reverse causation (whereby heart disease precedes depression). We investigated these hypotheses in a large prospective study of UK women with no previous vascular disease. During mean follow-up of 13.9 years of 690 335 women (mean age 59.8 years) with no prior heart disease, stroke, transient ischaemic attack, or cancer, 12 819 had a first hospital admission or death from myocardial infarction. The null findings in this large prospective study are consistent with depression not being an independent risk factor for myocardial infarction
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
