A new study of Swedish twins published in JAMA Internal Medicine reveals that being heavy may not come with the health costs we typically assume of it, such as heart attack or an early death, though researchers did find a link between higher body mass index and Type 2 diabetes. The findings call into question the wisdom behind recommending weight loss as a preventative treatment in some cases.

In groups of twins that were studied, around 5 percent reported a heart attack during 12 years of follow-up. Among twins with lower BMI, about 16 percent died compared to 14 percent of twins with higher BMI. But heavier twins were roughly twice as likely to develop diabetes over the study’s length.

"What the study does show is that there's a strong association between obesity and diabetes, which leads us to conclude that weight reduction interventions can be more effective against diabetes than when it comes to reducing the risk of heart attack and mortality," said lead author Dr. Peter Nordström, a chief physician in the Department of Community Medicine and Rehabilitation at Sweden’s Umeå University, in a statement.

Researchers examined the medical records of over 4,000 identical twin pairs previously enrolled in a long-running population study. The middle-aged pairs each had one twin with a higher body mass index (BMI) than the other and were tracked for an average of 12 years as they entered old age. To their surprise, the researchers found no noticeable difference in the rate of heart attacks or premature death between the group of twins with higher BMI and the group with lower BMI, even when they exclusively looked at obese individuals or twin pairs where the weight difference was dramatic. Supporting earlier research, however, they did find a sustained link between Type 2 diabetes risk and higher BMI.

Because the authors looked exclusively at identical twins and found no difference in heart attack or mortality risk between the two groups, the authors suggest that genetic factors may better explain why earlier research had shown a connection between these conditions and higher weight. In contrast, the findings only reaffirm how closely connected obesity is to the development of Type 2 diabetes.

Though most research has shown a positive link between poorer health and higher weight, other studies have muddled the picture.

For one, scientists have long known about the so-called obesity paradox, where obese people with chronic conditions like heart disease or diabetes are better protected against dying than their thinner counterparts. Some research has also shown that people who are merely overweight live slightly longer than normal weight people. And an extensive randomized clinical trial of obese people with Type 2 diabetes found that while people who were on a diet and exercise plan lost slightly more weight than people who were given the standard diabetes treatment after eight years time, they weren’t any less likely to die early or develop cardiovascular disease.

When the authors of the current study looked at all 8,000 participants at once, they found that people with a relatively high or low BMI had a greater risk of dying early, and obese individuals were 37 percent more likely to develop heart attack or die early than were normal weight people, even after accounting for other factors like age, smoking history, and level of physical activity. Because of the earlier findings though, the authors concluded that may be due more to genetic factors than obesity per se. It might also be the case that being overweight or obese isn’t necessarily damaging to our heart health, but only up until a certain point.

Of course, even if weight loss doesn’t necessarily improve your chances of staving off a heart attack or dying too young, lifestyle changes like regular exercise or a healthier diet certainly will, regardless of how much or little weight you end up losing.

Source: Nordström P, Pedersen N, Gustafson Y, et al. Risks of Myocardial Infarction, Death, and Diabetes in Identical Twin Pairs With Different Body Mass Indexes. JAMA Internal Medicine. 2016.