Temporal trends in prevalence of diabetes mellitus in a population-based cohort of incident myocardial infarction and impact of diabetes on survival

OBJECTIVE: To determine the temporal trends in prevalence of confirmed diabetes mellitus (DM), time from the date DM criteria were met to myocardial Infarction (MI), and impact of DM on survival. SUBJECTS AND METHODS: A retrospective cohort design was used to identify residents of Olmsted County, Minnesota, with incident MI from 1979 to 1998. The MI cases were characterized according to prevalent DM. Cases with and without DM were followed up for vital status until January 1, 2003. RESULTS: Of 2171 MI cases, 364 (17%) met criteria for prevalent DM. In the age- and sex-adjusted logistic regression models, the odds of prevalent DM increased 3% with each increasing year between 1979 and 1998 (95% confidence interval [CI], 194,5%; P=.007). Survival for MI cases with DM was unchanged between 1979-1983 and 1994-1998 (P=.74). For all MI cases, age-, sex-, and DM-adjusted risk of death decreased 3% from 1979 to 1998 (95%, CI, 10/6-5%) per year for 28-day survival (P=.02) and 2% (95% CI, 1%-3%) per year for 5-year survival (P=.02). There was a significant adverse effect of DM on 5-year survival after MI (age-, sex-, and calendar year-adjusted hazard ratio, 1.70; 95% CI, 1.38-2.09; P < 001). The adverse effect of DM persisted after adjusting for other cardiovascular disease risk factors, MI severity, and reperfusion therapy (hazard ratio, 1.66; 95% CI, 1.34-2.05; P < 001) and was unchanged over time (interaction between DM and calendar year, P=.63). CONCLUSION: These data indicate that the prevalence of DM among patients with MI is increasing and that its adverse impact on survival after MI remains unchanged.