However, there do not appear to be any serious adverse effects for those who do a water fast or a similar kind of fast where people consume a very small number of calories a day, said Krista Varady, professor of kinesiology and nutrition, who led the research, which is published in Nutrition Reviews.

"My overall conclusion is that I guess you could try it, but it just seems like a lot of work, and all those metabolic benefits disappear," Varady said. She stressed, however, that no one should undertake one of these fasts for more than five days without medical supervision.

Varady, an expert on intermittent fasting, said she wanted to study water fasting because she suddenly started getting contacted by journalists last fall who wanted to hear what she thought about it. She figured if she was going to comment, she should investigate the existing research.

The new paper is a literature review of eight studies on water fasting or Buchinger fasting, a medically supervised fast that is popular in Europe where people consume only a tiny amount of juice and soup a day. Varady's team looked at the results from each of those papers to see what story they cumulatively tell about the fasts' impact on weight loss, as well as a number of other metabolic factors.

Fasting did seem to spur noticeable short-term weight loss, the researchers found. People who fasted for five days lost about 4% to 6% of their weight; those who fasted for seven to 10 days lost about 2% to 10%, and those who fasted for 15 to 20 days lost 7% to 10%.

Only a few of the studies in the review tracked whether participants gained back the weight they had lost once the fast ended. In one of those, people gained back all they had lost in a five-day water fast within three months. In two other studies, only a small amount of the lost weight returned, but those studies encouraged participants to restrict their calorie intake after the fasts ended.

In contrast, it was clear that the metabolic benefits of the fasts disappeared soon after the fasts ended. Improvements to blood pressure, cholesterol and blood sugar levels were short-lived, returning to baseline levels quickly after participants started eating again.

Some of the studies included participants with Type 1 and Type 2 diabetes, who suffered no ill effects from the fasting, though they were monitored closely and had their insulin doses adjusted while fasting.

The most common side effects of these prolonged fasts were similar to those from intermittent fasting, Varady said, such as headaches, insomnia and hunger. There were no serious negative effects in the studies, such as metabolic acidosis or death.

She did note that the participants in these prolonged fasts lost about two-thirds of their weight in lean mass and one-third in fat mass. This is the opposite of what happens most of the time during weight loss, where more fat is lost than muscle. It makes sense that these extreme fasts would have this result, she said, because "your body needs a constant intake of protein. If it doesn't have that, then it draws from muscles."

Varady's research into intermittent fasting has looked at how well the regime works for weight loss, as well as at specific questions, such as whether intermittent fasting affects fertility -- she found it does not.

Varady said she would encourage someone hoping to lose weight to try intermittent fasting instead of a water fast, "because there's a lot more data to show it can help with weight management," she said.

The other UIC researchers in the study are Sofia Cienfuegos, Shuhao Lin, Vasiliki Pavlou and Kelsey Gabel, as well as Mark Ezpeleta, a former UIC student.