Fatality of coronary events : epidemiological studies of potential determinants

Abstract

Background: Although there has been a large number of improvements in prevention and care, coronary heart disease (CHD) was one of the main causes of death in 2010 around the world. The majority of CHD deaths are due to myocardial infarction (MI) and occur out-of-hospital. The research regarding determinants of MI fatality including out-of-hospital deaths is scant.

Aims: The overall aim of this project was to increase knowledge about the determinants of MI fatality in men and women from the Stockholm general population who suffer a first MI. Specific aims were to: 1) validate questionnaire data collected from close relatives of MI patients, 2) assess how known cardiometabolic risk factors are associated with MI fatality, 3) describe which comorbidities are the most common among fatal MI cases and to assess if they are associated to MI fatality, and 4) assess whether low-grade inflammation is associated to fatality of future coronary events.

Methods: Using material from the Stockholm Heart Epidemiology Program (SHEEP), a case-control study, the validity of questionnaire data provided by spouses/common-law spouses (proxy respondents) of non-fatal MI cases regarding 82 exposures was valuated. Using conditional logistic regression we calculated for each of the exposures, a “proxy bias”, based on information collected from 1) MI cases and controls [odds ratio A] and 2) proxies and the same set of controls [odds ratio B]. Disagreement was measured by calculating the ratio between odds ratio B and odds ratio A; 95% confidence intervals (CI) were calculated using resampling bootstrap with replacement. From the SHEEP, an inception cohort of first time MI cases was formed. Data were retrieved from questionnaires (filled in by a close relative if the case was fatal), physical examinations (for non-fatal cases), national registers and autopsy reports. Associations between selected cardiometabolic risk factors and MI fatality were assessed through calculations of risk ratios (RR) with 95% confidence intervals (CI) using binomial regression with log link. Presence of comorbidities among the fatal MI cases was mapped out and the number of previous hospitalizations was assessed. Associations between specific comorbidities, as well as number of previous hospitalizations, and MI fatality were assessed using the same modelling as for the cardiometabolic risk factors. A structured review of autopsy data was performed to identify additional indicators of comorbidities in fatal MI cases. Using material from the AMORIS cohort, sex specific associations between low-grade inflammation (using a score of five biomarkers: C-reactive protein, haptoglobin, white blood cell count, uric acid and albumin) and a fatal outcome in subjects who subsequently experienced a first coronary event were assessed. Odds ratios were calculated with 95% CI using logistic regression.

Results: For the vast majority of the exposures considered in the validation study, there was no significant disagreement between reports from MI patients and proxies. However, leisure time physical inactivity was overestimated by proxies compared to MI patients. Diabetes, but not hypertension and hyperlipidemia, was associated with MI fatality. Overweight, as compared to normal BMI, was inversely associated with MI fatality; the results for obesity went in the same direction. The results were adjusted for age, current smoking, parental history of premature cardiac death, number of previous hospitalizations, educational level, disposable income, and other cardiometabolic risk factors. An increased number of previous hospitalizations was associated with MI fatality after adjustments for sex, age and disposable income. Among comorbidities identified as prevalent in fatal cases, the following were, after adjustments (where possible), associated with 7-day fatality: heart failure, stroke, diabetes, alcoholism, psychiatric diseases, cancer, kidney diseases, epilepsy, rheumatoid arthritis and asthma. Indicators of comorbidities identified from autopsy data included silent MI, severe abdominal aortic atherosclerosis, liver steatosis, and underweight. An elevated inflammation score was after adjustments by age at the time of the coronary event, calendar year of the coronary event, diabetes, level of education, serum total cholesterol, serum triglycerides, and angina associated with increased fatality of future coronary events, both in men and in women.

Conclusions: MI patients and their spouses similarly reported data on a wide range of exposures including traditional cardiovascular risk factors, leisure time physical inactivity being an exception. Among the cardiometabolic factors under study, diabetes and presence of a low-grade inflammation, but not hypertension and hyperlipidemia was associated with an increased MI fatality. Overweight was associated to a decreased MI fatality. Repeated prior hospitalizations and/or heart failure, diabetes, stroke, psychiatric disease, alcoholism, cancer, renal diseases, epilepsy, rheumatoid arthritis and asthma were associated with increased MI fatality.