Abstract P117: Episodes of Anger are Associated with a Higher Risk of Myocardial Infarction and Acute Coronary Syndrome: Systematic Review and Meta-Analysis
Introduction: Short-term psychological stress is associated with an immediate physiologic response. We assessed the hypothesis that brief episodes of anger may trigger the onset of acute myocardial infarction (MI) and other acute coronary syndromes (ACS). Methods: We performed a literature search of the CINAHL, Embase, PubMed and PsycInfo databases from January 1966 through June 2012 using the key words “anger”, “hostility”, “myocardial infarction” and “acute coronary syndrome” without restrictions. We also reviewed the reference lists of retrieved articles and spoke with authors of relevant work to identify unpublished data. We used an inverse-variance-weighted random effects model to perform a meta-analysis of all epidemiological studies that reported relative risks and 95% confidence intervals (CI) for the association between outbursts of anger and nonfatal MI or ACS. Results: The meta-analysis included four independent case-crossover studies of anger and risk of MI or ACS published between 1995 and 2007 and one unpublished report. Combined, there were 4546 MI cases (two studies) and 462 ACS cases (two studies), including 180 that experienced an episode of anger in the two hours prior to disease onset; two studies were restricted to first MI or ACS and two included first and recurrent events. All studies used the Onset Anger Scale to assess anger episodes, but the studies differed in the cutoff used to define exposure. The association between outbursts of anger and disease risk in the following two hours was 3.46 (95%CI 2.08-5.76; p<0.001). The results were similar for studies of first and recurrent events and were similar for studies of MI and ACS. Conclusions: There is a higher risk of MI and ACS shortly after outbursts of anger. Some of the between-study heterogeneity may be due to regional differences in the type of anger expression and the interpretation of the questions. Further research is necessary to evaluate whether the risk is lower among people who are regularly active.