Skip to main content
. Author manuscript; available in PMC: 2014 Aug 19.
Published in final edited form as: JAMA. 2012 Sep 5;308(9):890–896. doi: 10.1001/2012.jama.11089

Table 2.

Prevalence of recognized and unrecognized myocardial infarction (MI) by CMR or ECG stratified by diabetes status. UMI by CMR were observed roughly twice as often as recognized MI. The prevalence of MI with the addition of ECG was significantly higher than the prevalence without ECG, but still significantly less than the increased prevalence with the addition of CMR (p<0.01 for both).

All 936 Participants Prevalence in 337
Participants with
Diabetes, No. (%,
CI)
Prevalence in 599
Participants without
Diabetes, No. (%,
CI)
No MI 688 (74%, 71–76%) 228 (68%, 63–73%) 460 (77%, 73–80%)
Clinically Recognized MI 91 (10%, 8–12%) 37 (11%, 8–14%) 54 (9%, 7–11%)
Unrecognized MI by ECG 46 (5%, 4–6%) 15 (4%, 2–7%) 31 (5%, 3–7%)
Unrecognized MI by CMR 157 (17%, 14–19%) 72 (21%, 17–26%) 85 (14%, 11–17%)
Cumulative MI by ECG 137 (15%, 12–17%) 52 (15%, 12–19%) 85 (14%, 11–17%)
Cumulative MI by CMR 248 (27%, 24–29%) 109 (32%, 27–37%) 139 (23%, 20–27%)

Abbreviations: CMR, cardiac magnetic resonance; ECG, electrocardiography; MI, myocardial infarction.