A recently discovered hormone that regulates appetite helped obese people lose weight in the first study of its use as a drug in human beings, researchers reported yesterday.

Obese men and women who injected themselves every day with leptin for six months while dieting lost about 15 pounds, compared with a loss of only about three pounds in people taking placebo injections, the researchers said.

The treatment, however, was not without drawbacks. Nearly everyone found the shots irritated the skin, and more than one-third of people taking the highest, most effective dose of leptin chose to stop it after a month.

Nevertheless, the study, reported in today's Journal of the American Medical Association, demonstrated the potential of a novel approach to one of the country's biggest medical problems. About half of American adults are overweight or obese by current medical definitions.

"I think this is an opening shot of the molecular biology revolution in the treatment of obesity--even though we may not see it as a magic bullet in the immediate therapeutic sense," said Steven B. Heymsfield, a physician at St. Lukes-Roosevelt Hospital and Columbia University, who headed the study.

The discovery of leptin in December 1994 was greeted with optimism by many scientists because it acted principally on the brain, the organ that ultimately regulates appetite and body metabolism. Produced primarily by fat-storing cells, the hormone suppresses hunger and speeds up metabolic activity as its concentration rises in the bloodstream and brain.

People with a rare, inborn deficiency of leptin are extremely obese. Curiously, people with less severe weight problems tend to have an excess, not a lack, of leptin. Many physiologists theorize people in the second group don't respond adequately to leptin and that their bodies produce even more in an attempt to compensate. The new study, funded in part by Amgen, a Southern California biotechnology company that holds the patent on leptin, tested the idea that those people need even higher levels of leptin before they can get the hormone's appetite-curbing effect.

"Until this experiment, the pessimists were saying that 'leptin resistance' just couldn't be surmounted, and that we should be considering other targets," said Richard D. Palmiter, a biochemist at the University of Washington and the Howard Hughes Medical Institute who has studied the hormone's activity in mice. "People who are interested in the leptin receptor ought to be encouraged."

In the study, 73 volunteers, with an average weight of about 200 pounds, were randomly assigned to get daily injections of either placebo or leptin at one of four doses. (The highest dose was 30 times the concentration of the lowest.) Everyone was encouraged to exercise (by brisk walking) five days a week and was given a diet that provided 500 fewer calories than necessary to maintain weight.

After four weeks, people taking the placebo had lost about one pound, and people taking the highest dose of leptin, about four pounds. The volunteers were then given the chance to continue on the study or to stop. Sixty of the 70 who had gotten that far chose to continue.

Seven of the 10 who quit, however, were in the highest-dose leptin group. Members of that group ultimately reported having, on average, 42 episodes of redness, itchiness, inflammation or other reaction over the six-month study.

Virtually all of the weight loss occurred at the two highest doses. People taking the second-highest dose reported 33 episodes of skin reaction.

"There's a delivery problem," Heymsfield said. "The dose we can give is not really enough to get weight-normalizing effects."

Amgen officials have decided the formulation of leptin used in the study isn't good enough to warrant applying to the Food and Drug Administration for permission to market it to the public. The company is testing "a few" other forms of leptin, which it hopes will permit the use of higher, but less irritating, doses, said Andrea Rothschild, an Amgen spokeswoman.

Tests showed that virtually all the weight lost was fat. Volunteers taking the highest leptin dose ate less than people taking lower doses or placebo. The drug had no effect on the production or activity of insulin, which had been a major worry of some scientists.

Nearly three-quarters of the subjects taking leptin had antibodies to the hormone by the end of six months. Antibodies can either diminish or enhance the actions of the substances they attack. Although the antibodies appeared to have no effect on leptin's potency in this study, only longer-term study of the injected hormone will determine whether that will continue to be true, Heymsfield said.