Prognostic Significance of Silent Myocardial Infarction in Newly-Diagnosed Type 2 Diabetes: UKPDS 79

Study Questions:

What is the prevalence of silent myocardial infarction (SMI) in people with newly-diagnosed type 2 diabetes (T2D), and its relationship to future myocardial infarction (MI) and all-cause mortality?


A total of 5,102 people with newly-diagnosed diabetes participating in the 30-year UKPDS (United Kingdom Prospective Diabetes Study) were included in the study. Patients were seen quarterly in UKPDS clinics until 1997. The 3,277 surviving patients entered a 10-year post-trial study in a usual care setting. Exclusion criteria included severe vascular disease, definite or possible coronary heart disease at baseline, stroke within the previous year, and incomplete data. Cox proportional hazards regression was used to examine outcomes by SMI status.


The prevalence of SMI in the full cohort of 4,827 increased progressively from 17.5% (845 patients) at baseline to 30.1% (680 of 2,259 patients) at 12 years. The presence of Q waves was typically transient, with 53% of those with SMI at baseline also having electrocardiograms (ECGs) of an SMI at year 3. Median follow-up was 17 years. Of 1,967 patients with complete baseline data (37% of total UKPDS cohort), 326 (16.6%) had ECG evidence of SMI at enrollment. Those with SMI were more likely to be older, female, sedentary, and nonsmokers compared with those without SMI. Their mean blood pressure was greater despite more intensive antihypertensive treatment, they were more likely to be taking aspirin and lipid-lowering therapy, and they had a greater prevalence of microangiopathy. Evidence of an SMI at the time of diagnosis of T2D was associated with a 49% increased rate of subsequent fatal MI and a 26% increased rate of all-cause mortality after adjustment for conventional cardiovascular risk factors. Fully-adjusted hazard ratios (95% confidence interval) for those with, versus those without, SMI in multivariate models that included UKPDS Risk Engine variables were 1.58 (1.22–2.05) for first fatal MI, and 1.31 (1.10–1.56) for all-cause mortality. Hazard ratios for first fatal or nonfatal MI, and for first nonfatal MI, were nonsignificant. The Net Reclassification Index showed no improvement when SMI was added to these models, and the Integrated Discrimination Index showed that SMI marginally improved prediction of fatal MI and all-cause mortality.


Approximately one in six UKPDS patients with newly-diagnosed T2D had evidence of SMI, which was independently associated with an increased risk of fatal MI and all-cause mortality. However, identification of SMI does not add substantively to current UKPDS Risk Engine predictive variables.


Silent MI does not have the incremental predictive power of the UKPDS Risk Engine variables including age at diagnosis, gender, ethnicity, smoking, glycated hemoglobin, systolic blood pressure, and total:high-density lipoprotein cholesterol ratio. Nevertheless, the finding should prompt assessment of the adequacy of management of modifiable cardiovascular risk factors in T2D.

Clinical Topics: Dyslipidemia, Vascular Medicine, Atherosclerotic Disease (CAD/PAD), Lipid Metabolism, Nonstatins

Keywords: Coronary Artery Disease, Great Britain, Myocardial Infarction, Follow-Up Studies, Lipoproteins, Risk Factors, Electrocardiography, Prognosis, Cholesterol, Incidence, Cardiology, Cardiovascular Diseases, Diabetic Angiopathies

< Back to Listings