Unrecognized myocardial infarction: The association with cardiopulmonary symptoms and mortality is mediated via echocardiographic abnormalities of global dysfunction instead of regional dysfunction: The Olmsted County Heart Function Study

BackgroundThere are very few data describing the association of electrocardiogram-based unrecognized myocardial infarction (ECG-UMI) with nonanginal cardiopulmonary symptoms, echocardiographic abnormalities, and mortality in the community.MethodsWe studied 2042 Olmsted County residents, who were randomly selected and aged ≥45 years, by a survey questionnaire for symptoms, echocardiogram for structural abnormalities, and a 5-year follow-up for all-cause mortality. Unrecognized myocardial infarctions (n = 81) were diagnosed if ECG-based myocardial infarction (MI) criteria were met without the history of a documented recognized MI.ResultsIn UMI versus no MI controls, the prevalence (%) of dyspnea on exertion (49 vs 29), orthopnea (6 vs 4), palpitations (20 vs 15), and history of fluid overload (6 vs 1) was significantly higher (P < .05). The associations of exertional dyspnea and history of fluid overload with UMI were independent of age, sex, and pulmonary disease but had a significant reduction in their magnitude after adjusting for global dysfunction (diastolic or systolic dysfunction). All the 4 symptoms were associated with increased risk of mortality (hazard ratios ranging from 2.3 to 9.1, P < .0001), which was meaningfully attenuated by adjusting for ECG-UMI status. Global ventricular dysfunction had a more significant impact on this association than regional ventricular dysfunction (wall motion abnormalities).ConclusionsThe increased risk of mortality associated with symptoms is at least in part mediated via ECG-UMI. Structural abnormalities of global dysfunction play a greater role in mediating this risk than regional dysfunction, challenging the current clinical practice of calling an ECG-based MI false positive in symptomatic adults in the absence of wall motion abnormalities.