The main issue, initially raised in a commentary published in The Lancet in November, is with a formula developed and published with the guidelines to assess the risk of having a heart attack or stroke.
The commentary in The Lancet suggested that the new formula predicted that too many people would have a heart attack or stroke. This concern was raised when the chance of having a heart attack or stroke predicted in the new guidelines was evaluated in several “validation studies” that were not used in developing the formula.
In one example, the formula was tested against data from the Reasons for Geographic And Racial Differences in Stroke (REGARDS) study. REGARDS is based at the School of Public Health at the University of Alabama at Birmingham.
REGARDS investigators have responded to the commentary with a letter published in the AHA journal Circulation, saying it is premature to draw firm conclusions about potential overestimation of risk using the new risk formula.
“The new formula was developed in several large studies from the 1980s and 1990s,” explained Paul Muntner, Ph.D., professor in the UAB Department of Epidemiology. “When scientists tried to validate it using data from more contemporary studies, including REGARDS, they found it considerably overestimated the risk of heart attack and stroke.”
“As epidemiologists, we are very familiar with the studies used in developing and validating the formula,” Muntner explained. “We identified four major differences in the studies that were used to develop the formula versus those used to validate it. These differences between studies may be the reason for the overestimation of risk, rather than the equation itself.”
Muntner and his colleagues describe these four differences:
- Several of the studies used to develop the formula documented strokes and heart attacks in the study by both keeping in contact with the participants in their study and also screening all records in every hospital in the region where the study was done to find additional heart attacks or strokes that had occurred. In contrast, the studies used to validate the formula documented events only by staying in contact with people in the study. “For REGARDS, people in the study are from across the nation, making it impossible to review every hospital record for every hospital in the nation,” said George Howard, one of the authors of the response from UAB. It is possible that up to 25 percent of stroke and heart attack events may be missed without screening all hospital records, which is costly to implement.
- Statins are known to reduce the risk for having heart attacks and strokes by as much as 20 percent to 30 percent. The use of statins has tripled among U.S. adults in the past 12 years. The validation studies were conducted in an era of widespread statin use, but the analysis did not account for people initiating statins.
- The use of operations to open arteries that feed the heart or brain, especially coronary stenting to keep heart arteries open, has increased over the past 25 years in the United States. These procedures were not evaluated in assessing differences in risk for heart attacks or strokes between the studies.
- The length of time people were followed to see if they had heart attacks or strokes in the validation studies was short. The longer people are followed, the older they get; and since they are older, the risk of having a heart attack or stroke is higher. The scientists in the REGARDS study are planning to re-do the analysis with a longer follow-up, so that the estimates will be more comparable.
Muntner and colleagues believe these four considerations make it premature to draw firm conclusions about potential overestimation of risk using the new risk formula.
“As with all science, knowledge is gained incrementally and will continue to evolve as the risk formula is studied further,” Muntner said. “We look forward to using the contemporary data available from the nationwide REGARDS study to inform the ongoing discussion surrounding the new formula with the scientific rigor that the public deserves.”