Northwestern Medicine Develops 30-Year Heart Failure Risk Prediction Model

Sadiya Khan, MD

A recent study from Northwestern University suggests investigators may have developed a model possible of providing patients and clinicians with an accurate prediction of an individual’s lifetime risk of heart failure.

Using data from nearly 25,000 patients from 5 population-based cohorts, investigators were able to develop and validate a long-term incident heart failure risk equation that led to the creation of a sex- and race-specific equation for prediction of long-term risk of heart failure with high discrimination.

“These new models offer the opportunity for clinicians and patients to begin discussions at the individual level for opportunities to start prevention earlier in the life course,” said Sadiya Khan, MD, assistant professor of cardiology and epidemiology at Northwestern University Feinberg School of Medicine and a Northwestern Medicine physician, in a statement. “This moves the field forward by offering a precision approach to prevention and moving beyond risk associated with having or not having hypertension or diabetes.”

The study sample used by investigators included pooled individual-level data from the ARIC Study, CARDIA Study, Framingham Heart Study, Framingham Offspring Cohort, and the MESA study. Using adult patients aged 20-59 years without a history of cardiovascular disease at baseline, investigators identified 24,838 participants with 599,551 person-years of follow-up for inclusion in the current study.

This cohort was 55% female, 25% Black, and had a mean age of 43±12 years. Investigators noted the risk factor profile of the cohort was generally representative of the US population. The rate of incident heart failure among the entire study cohort was 4.0 per 1000 person-years, with the highest rates in Black men at 5.1 per 1000 person-years and the lowest rates in White women at 3.5 per 1000 person-years.

Sex- and race-specific equations for 30-year risk of heart failure included age, systolic blood pressure, BMI, total cholesterol, HDL-C, and smoking, diabetes, and hypertension treatment status. These equations were validated for competing risk of non-heart failure death. Investigators pointed out model discrimination and calibration were assessed using 10-fold cross-validation. Additionally, the model was applied to varying risk factor patterns for systematic evaluation.

Upon analysis, discrimination of the 30-year risk model for heart failure was excellent in all groups, with Harrell’s c statistics of 0.82 (95% CI, 0.80-0.83) and 0.84 (95% CI, 0.82-0.85) in White and Black men and 0.84 (95% CI, 0.82-0.85) and 0.85 (95% CI, 0.83-0.87) in White and Black women, respectively. Investigators noted this is consistent with previously published risk prediction equations, citing the 2013 ACC/AHA Pooled Cohort Equations for 10-year risk prediction of atherosclerotic cardiovascular disease.

Using the model, the estimated risk of developing heart failure in an average 40-year-old non-smoker with an untreated systolic blood pressure of 140 mm Hg, which is high, and body mass index of 30 kg/m² was 22.8% in a Black man, 13.7% in a White man, 13.0% in a Black woman, and 12.1% in a White woman.

“Once someone develops symptoms of heart failure, the window for prevention has closed, which is a missed opportunity, given that the risk of dying in the five years after diagnosis is 50%, similar to a cancer diagnosis,” Khan added.

This study, “Development and Validation of A Long-Term Incident Heart Failure Risk Model,” was published in Circulation Research.