Electrocardiographic criteria for silent myocardial infarction: Impact of different definitions on detection rates and prognostic significance in the atherosclerosis risk in communities (ARIC) study: A comparison with evolving bundle branch blocks.

INTRODUCTION

It is unclear how differences in the electrocardiographic (ECG) definition of myocardial infarction (MI) impact detection rates and prognostic significance of silent MI (SMI).

METHODS

This analysis included 9188 participants (57.4 % women, 20 % black, age 62.6 ± 6.0 years) enrolled in the Atherosclerosis Risk in Communities study who had serial ECGs obtained between visit 1 (1987-1989) and visit 4 (1996-1998). Exclusions included known cardiovascular disease (CVD) prior to visit 1, ECG findings of MI or bundle branch block (BBB) at visit 1, or an adjudicated fatal and non-fatal MI events between visits 1 and 4. Using the Minnesota Code (MC) ECG Classification and in the absence of adjudicated MI, the following SMI definitions were derived: Standard MC MI [major Q-wave abnormality, or minor Q-wave abnormality plus major ST/T abnormality], only major Q-wave abnormality, standard significant serial Q-wave changes [Q1 to Q7], expanded MC serial Q-wave changes[Q1 to Q8], standard MC significant serial Q-wave changes or Standard MC significant serial ST/T changes, and evolving MC BBB. Cox proportional hazard models were used to examine the association of different definitions of SMI (compared to no new MI or evolving BBB) with fatal or non-fatal MI events ascertained after visit 4 until December 2016.

RESULTS

The prevalence of SMI ranged from 0.6 % to 7.0 % depending on the ECG criteria defining SMI. Presence of SMI was predictive of fatal/non-fatal MI regardless of the definition but with varying levels of association. Standard MC expanded serial Q-wave changes had the strongest adjusted relationship [Hazard Ratio: 2.53 (95 % Confidence Interval (CI): 1.60-4.01)] while evolving BBB had the weakest adjusted association [HR: 1.39 (95 % CI: 0.80-2.40)].

CONCLUSIONS

The prevalence and prognostic significance of SMI are impacted by the ECG criteria defining MI. A uniform approach(s) for detection of SMI in population studies which builds on the available standard definitions that fit different research scenarios is needed.