Predicting the Development of Myocardial Infarction in Middle-Aged Adults with Type 2 Diabetes: A Risk Model Generated from a Nationwide Population-Based Cohort Study in Korea

Abstract

Background: Most of the widely used prediction models for cardiovascular disease are known to overestimate the risk of this disease in Asians. We aimed to generate a risk model for predicting myocardial infarction (MI) in middle-aged Korean subjects with type 2 diabetes.

Methods: A total of 1,272,992 subjects with type 2 diabetes aged 40 to 64 who received health examinations from 2009 to 2012 were recruited from the Korean National Health Insurance database. Seventy percent of the subjects (n=891,095) were sampled to develop the risk prediction model, and the remaining 30% (n=381,897) were used for internal validation. A Cox proportional hazards regression model and Cox coefficients were used to derive a risk scoring system. Twelve risk variables were selected, and a risk nomogram was created to estimate the 5-year risk of MI.

Results: During 7.1 years of follow-up, 24,809 cases of MI (1.9%) were observed. Age, sex, smoking status, regular exercise, body mass index, chronic kidney disease, duration of diabetes, number of anti-diabetic medications, fasting blood glucose, systolic blood pressure, total cholesterol, and atrial fibrillation were significant risk factors for the development of MI and were incorporated into the risk model. The concordance index for MI prediction was 0.682 (95% confidence interval [CI], 0.678 to 0.686) in the development cohort and 0.669 (95% CI, 0.663 to 0.675) in the validation cohort.

Conclusion: A novel risk engine was generated for predicting the development of MI among middle-aged Korean adults with type 2 diabetes. This model may provide useful information for identifying high-risk patients and improving quality of care.

Keywords: Diabetes mellitus, type 2; Myocardial infarction; Risk.

Fig. 1

A nomogram for the prediction of the 5-year probability of myocardial infarction. Each of the 12 variables was applied with scores from 0 to 100. Each variable corresponds to a specific point by drawing a line straight up to the score axis. The total score, which is the sum of the scores for each of the 12 variables at the bottom of the nomogram, ranges from 0 to 654.

Fig. 2

The 5-year incidence probability of myocardial infarction (MI) according to the total risk score.

Fig. 3

Incidence rate (per 1,000 person-years) based on the decile groups of total risk score in the development and validation cohorts. The numbers on the x-axis represent the range of the total risk score according to each decile group.