LOS ANGELES » For years, people have tried going on extremely low-calorie diets, hoping to stave off illness and delay the effects of aging.

It’s an approach that is based on science showing that limiting proteins and sugars seems to impede processes in the body that lead to diabetes and even cancers. But it’s also a tough road that can sap dieters’ patience and strength, said University of Southern California aging researcher Valter Longo, who recently led a study searching for a better way to get the benefits of fasting without quite so much pain.

“We wanted to know, what if you let people eat normally, but then once every few weeks you fool the system into thinking it was starving?” he said.

The results of Longo’s experiment, published last week in the journal Cell Metabolism, showed that the fasting-mimicking plan did indeed work.

The research was conducted in three parts: in yeast, in mice and in people.

When Longo’s team alternated between offering yeast a nutrient medium and fasting them, they lived longer. Middle-aged mice that were fed a diet that mimicked fasting lived longer and had less fat, fewer cancers, less bone density loss and other positive effects.

To begin to understand the effects of a fasting-like diet on people, the team organized a pilot randomized clinical trial involving 38 healthy subjects ages 18 to 70. Half of the group were randomized into a control group that ate normally, returning for testing at the end of the three-month study.

The other half went through three five-day-long monthly cycles of the fasting-mimicking diet. Their food — all plant-based, with low carbohydrates, low protein and high levels of healthy fat — was delivered to them in a box and included powdered soups, nut bars and chips. It provided about 1,090 calories on the first day and about 725 calories on Days 2 through 5.

“We try to make it as close as possible to something that looks like normal food,” Longo said, adding that 95 percent of the dieters stuck with the plan — a success rate that surprised the researchers.

“I think people noticed a lot of results, and that motivated them to come back,” he said.

The test subjects who ate the special diet experienced drops in their fasting blood glucose levels and in factors associated with cancer and cardiovascular risk. In mice, the researchers saw increased numbers of stem cells, suggesting that starvation-like conditions killed off old, weaker cells and allowed younger, refreshed cells to emerge.

“Everything is getting a little younger and it goes back to working much better,” Longo said.

He and his colleagues are almost done with another study with 70 to 80 subjects and soon will sit down with officials at the Food and Drug Administration to see whether the diet might be appropriate for people with illnesses like cancer — whether it might be possible to reduce recurrences, for example, by prescribing a fasting-mimicking food plan.

Someday, he said, doctors might want to begin prescribing similar diets for patients with markers showing they are at risk for cancers and other illnesses — much as physicians today prescribe special diets and medications for patients with high cholesterol who are at risk for heart disease. (Longo has equity in the USC-backed medical food company L-Nutra; data analysis for the Cell Metabolism study was conducted by scientists who had no ties to the venture, he said.)

Longo, who eats sparingly and healthily to begin with but said he follows the stricter regimen once or twice a year, emphasized that people should not experiment with diets like this one without medical supervision.

“This is actually a strong intervention,” he warned.