Could Fasting Cure Diabetes? Evidence on Not Eating for Long Stretches Is Compelling—and Controversial

Weight gain may be driven not only by what we eat but also by our tendency to eat all day long. In the past few years, intermittent fasting has emerged as a popular trend in weight loss. A growing number of health professionals are also prescribing fasting to people with type 2 diabetes, which currently afflicts more than 29 million people in the U.S. Yet a recent study warns that going for long stretches without eating could cause the very damage it's supposed to prevent.

Type 2 diabetes is triggered in part by unhealthy eating, which renders the body resistant to insulin, a hormone produced by the pancreas. Without insulin, sugar from food can't enter our cells, leaving the blood with an excess amount of it. At first, the pancreas compensates by making more insulin, but eventually the demand wears out the digestive organ. Diabetics then become dependent on insulin injections to control their blood sugar.

Dr. Jason Fung, a kidney specialist, is convinced that fasting undoes that cycle: Not eating reduces blood sugar. As he points out, fasting is simply extending what we already do at night when we sleep. "It's supposed to be part of everyday life," says Fung, who co-founded the Toronto-based Intensive Dietary Management Program and wrote The Obesity Code and The Complete Guide to Fasting . Fasting can also send the body into ketosis, in which it burns fat rather than sugar. That helps with losing weight, which also helps slow diabetes.

Several recent studies support intermittent fasting. Research published June 5 in the scientific journal Cell Metabolism found that eating only between 8 a.m. to 2 p.m.—instead of the more common 8 a.m. to 8 p.m.—helped people with early signs of diabetes respond better to their body's natural insulin. The schedule also reduced blood pressure and appetite, two factors that worsen diabetes.

But some researchers are calling for caution, including Ana Bonassa, from the University of São Paulo, in Brazil. She and her colleagues presented a study in which rats subjected to intermittent fasting showed an increase in fat tissue, with damage to insulin-releasing cells in the pancreas. Those affects, Bonassa said, "could lead to diabetes and serious health issues."

Fung disputes the result. As with all clinical research, what happens with rats isn't necessarily applicable to humans. And, he points out, humans have gone for long periods without eating for most of human history. "If fasting gives us diabetes," he says, "then cavemen should have had a lot of it."