Study Shows Surprising Range of Risk After Acute MI Among Patients With Diabetes

Confirming once again that not all diabetic patients are alike, a new study suggests there may be a way for physicians to identify those at high versus low risk of having adverse events after a heart attack. 

Another View. Study Shows Surprising Range of Risk After Acute MI Among Patients With Diabetes

“Our models could be used for an individualized approach to risk stratification at the time of discharge and identification of patients for whom closer follow-up and secondary prevention strategies can be targeted most aggressively,” write researchers led by Suzanne V. Arnold, MD, MHA (Saint Luke’s Mid America Heart Institute, Kansas City, MO).

The predictive models that Arnold and colleagues constructed for mortality and angina tell a surprising story of wide risk distribution within diabetic populations after acute MI.

The study used data from two large databases of MI patients—TRIUMPH and PREMIER—to tease out predictors of long-term mortality and angina. The TRIUMPH population consisted of 1,613 patients with diabetes (either known or diagnosed during the acute MI) discharged from 24 US hospitals, while the PREMIER database, which was used as the validation model, included 786 diabetic patients treated at 19 US hospitals.

Known and Unknown Predictors Emerge

Mortality rates at 1 and 5 years after discharge were 8.9% and 27.1%, respectively. The factors most strongly predictive of mortality after an acute MI were: higher serum creatinine, not working at the time of acute MI, older age, and lower hemoglobin on admission.

The mortality model showed good discrimination (c-index = 0.79; bootstrap validated 0.78) and excellent calibration. Patients in the lowest risk category had a 5-year mortality rate of 4% versus 80% for those in the highest category. Overall, about one-third of the entire study population had a 5-year risk of mortality that was under 10%, whereas those whose risk was above 50% accounted for just 16% of all patients.

The rate of angina at 1 year was 27%. The burden of angina present prior to acute MI was found to be the most important predictor of angina occurrence after acute MI, while other predictors included younger age, history of prior CABG, and chronic heart failure.

The angina model demonstrated moderate discrimination (c-index = 0.71; bootstrap validated 0.69) and excellent calibration. Patients in the lowest category of risk had a 1-year angina rate of 12% versus 59% for those in the highest risk category. As with mortality, distributions of risk for angina were wide and variable.

Predictors Unique, but Not Exclusive to Diabetes

According to the study authors, using the models at time of hospital discharge after acute MI may permit identification of patients who require closer follow-up and more intensive secondary prevention, “including careful discharge education and counseling provided for patients and their families.”

Interestingly, Arnold and colleagues say, while some predictors of mortality such as not working at time of acute MI and low hemoglobin are not found in traditional risk models, they also are not specific to underlying diabetes. Additionally, peak troponin, which has been previously identified as predictive of angina in acute MI patients, was not at all predictive in the diabetic population.

From a practical perspective, Arnold and colleagues say the risk models “would allow for better identification of patients’ risk so that follow-up and treatment strategies can be triaged and allocated accordingly, rather than just on the basis of having [diabetes].” They also may have economic implications, they add, given that angina drives repeat hospitalizations and healthcare costs.

Lower Risk Does Not Equal Less Treatment

Michael E. Farkouh, MD, MSc, of the Heart and Stroke Richard Lewar Centre of Excellence at the University of Toronto (Toronto, Canada), who was not involved in the study, told TCTMD that there may be a message for the global community “in the sense that where you have limited resources, risk stratifying could allow you to put those resources toward those at highest risk.”

But he expressed concern that the overall message of the study could be misinterpreted, leading to less clinical vigilance with those at lower risk.

“The message remains that all diabetics are high risk compared to nondiabetics,” Farkouh said. “There are patients who have a better outcome with regard to mortality and angina, but these are patients that traditionally are treated as high-risk patients and they should continue to be treated that way. Maybe what it’s telling us is that our current strategies are working.”

Farkouh added that demonstrating heterogeneity within the diabetic community with regard to outcomes is important, “but I’m not sure this paper tells us how we can translate that into a change in management. It certainly doesn’t convince me to be less aggressive with the low-risk people. Those people may still need to be treated aggressively to maintain that mortality advantage.”


  • Arnold SV, Spertus JA, Jones PG, et al. Predicting adverse outcomes after myocardial infarction among patients with diabetes mellitus. Circ Cardiovasc Qual Outcomes. 2016;Epub ahead of print. 


  • Arnold and Farkouh report no relevant conflicts of interest. 

Related Stories:


We Recommend