Trajectory Matters for Heart Health Assessments Every Few Years

Tracking a person’s change in cardiovascular risk over time modestly improved prediction of incident cardiovascular disease (CVD) and survival beyond a single risk estimate alone in primary prevention, a large study showed.

Cardiovascular risk over nearly 20 years was better modeled when a repeat assessment at 5 years was added to earlier SCORE or ASCVD pooled cohort equations results:

• Harrell’s C index increased from 0.685 to 0.690 for SCORE; from 0.699 to 0.700 for ASCVD
• The Akaike information criterion fell from 17,255 to 17,200 for SCORE; from 14,739 to 14,729 for ASCVD
• The continuous net reclassification index was 0.353 for SCORE; 0.232 for ASCVD

“These findings were not sensitive to competing risks, missing data, or inclusion of specific CVD outcomes, and the results were replicated in individuals who improved their SCORE and ASCVD risk scores without medication,” reported Joni Lindbohm, MD, PhD, of University of Helsinki, Finland, and University College London, and colleagues.

They took study results and developed an online calculator that estimates a person’s CVD-free life-years as a function of two risk score measurements, the authors explained in the Lancet Digital Health.

The analysis was based on the Whitehall II longitudinal cohort of several thousand government employees in London.

Overall, the study supports the existing approach of tracking changes in cardiovascular risk over time based on known intervention effects.

“However, the approach has not been validated in intervention studies, does not aid in recommendation of a combination of lifestyle changes, and cannot be applied when no estimate of the effect of intervention exists as is the case when multiple lifestyle changes are combined. In these situations, measuring changes in risk scores could provide an alternative,” the authors said.

Their models showed that “commonly recommended lifestyle interventions might be insufficient to effectively prevent or delay the onset of cardiovascular events in most cases,” as “clinically meaningful improvement in the risk scores would require substantial changes in lifestyle, such as quitting smoking, or a decrease in systolic blood pressure or total cholesterol similar to that achieved by low-intensity antihypertensive or statin therapy.”

Whitehall II participants included in the present analysis (n=7,574) underwent clinical examinations in 5-year intervals from 1991 to 2016, and were followed up for incident CVD until 2019.

All had no history of CVD at baseline in 1991-1993, when they were ages 40-63. Just under 70% were men.

Over follow-up averaging 18.7 years, people tended to have SCORE and ASCVD risk scores rise over time.

Ultimately, approximately one in five individuals developed CVD — stroke, MI, coronary artery bypass grafting surgery, percutaneous coronary intervention, definite angina, heart failure, or peripheral artery disease — according to linked electronic health records.

Changing CVD risk scores were especially predictive of outcomes in younger people. For example, each 2-unit improvement in risk scores was associated with an additional 1.3 life-years free of CVD for SCORE and an additional 0.9 life-years for ASCVD — but only an additional 0.4 life-years free of CVD for SCORE and 0.3 life-years for ASCVD, respectively, at age 65.

Lindbohm’s group acknowledged that Whitehall II is an occupational cohort of relatively healthy individuals compared with the general population. “This means that the incidence of disease and prevalence of risk factors are likely to be an underestimation of those in the general population,” they stated.

Another caveat of the study was its reliance on participants self-reporting variables (e.g., use of antihypertensive medications) on questionnaires.