Repeated cycles of weight loss and gain may be linked to higher risk for stroke, heart attack, and death in people with pre–existing coronary artery disease, according to a study published online in the New England Journal of Medicine.

Led by researchers at NYU Langone Medical Center, the study was the first to measure the effect of “weight cycling” on health outcomes in people with pre–existing heart disease. People with the largest weight changes were found to experience 136 percent more strokes, 117 percent more heart attacks, and 124 percent more deaths than those with the smallest shifts in weight.

Those in the high–fluctuation group had weight changes as large as 3.9 kilograms (or roughly 8.6 pounds), while weight varied by around 0.9 kilograms (just under 2 pounds) in the group with the smallest shifts in weight.

“Our findings suggest that we need to be concerned about weight fluctuation in this group that is already at high risk due to coronary disease,” says lead study author Sripal Bangalore, MD, director of the cardiovascular outcomes group in the Cardiovascular Clinical Research Center at NYU Langone.

“Even though this analysis was not designed to find out the causes of increased risk with body weight fluctuations, we need to examine how we can help Americans keep weight off, rather than having it go up and down,” says Dr. Bangalore, also an associate professor of medicine in the Leon H. Charney Division of Cardiology.

The research team reviewed data on 9,509 men and women with coronary artery disease who participated in the Treating to New Targets trial, which originally concluded in 2005. Study participants were between the ages of 35 and 75. All had coronary artery disease, high cholesterol levels, and some history of heart problems. Half were being treated with cholesterol–lowering drugs in intensive doses to see if this resulted in fewer deaths. All were monitored for a median of 4.7 years.

The analysis linked shifts in body weight to statistically significant differences in outcomes only in people who were overweight or obese at the beginning of the study, but not for people who started with normal weight. Body weight changes were also strongly linked to an increase in newly diagnosed diabetes, and associations persisted regardless of a person’s average body weight and traditional risk factors for heart disease.

The researchers caution that their re–analysis does not show a cause–and–effect relationship between weight cycling and poor outcomes, but only an association. The authors also recognize that they were unable to tell if people lost weight intentionally, unintentionally, or due to illness, or if any eventual heart problems resulted directly from the weight loss, change in weight, or illness.

Dr. Bangalore hopes that the current findings will lead to further study of weight fluctuation in people with coronary artery disease and to the development of related practice guidelines once all the evidence is in.