(The New York Times News Service) -- The obesity warnings are everywhere. The Centers for Disease Control and Prevention says that with 35 percent of Americans overweight and 27 percent obese, obesity "has risen at an epidemic rate during the past 20 years."

In December, Dr. David Satcher, who was surgeon general, said obesity would soon succeed tobacco as the leading cause of preventable deaths in America.

In California, there are calls for taxes on soft drinks. Across the country, there are demands that schools banish soda and candy vending machines from within their walls. In Pennsylvania, one school district sent letters to parents telling them their children were too fat.

But underlying the cries of alarm are some assumptions about obesity and health that, some leading researchers say, have yet to be supported by facts. No one is claiming that it is good to be obese or that fatness does not matter to health. The problem, they say, is that when it comes to obesity, nothing is as straightforward as it might appear.

The obesity researchers say they are not trying to be contrarians. They simply are asking a fundamental question that has gotten lost in the growing cries of alarm over the fattening of America. Is obesity, they ask, a symptom or a disease?

Some strongly suspect it is a symptom. And losing weight, they say, may be suppressing the symptom but doing little or nothing for the underlying illness, just as taking aspirin for a fever may do nothing for the sickness that had fever as a symptom. Moreover, obesity experts add, not every person with the symptom of obesity necessarily has a disease that can increase their chances of an early death.

"People should be very, very careful in thinking about obesity and health," said Rudolph Leibel, the head of the division of molecular genetics at Columbia University's College of Physicians and Surgeons.

Leibel said his own years of experience studying fat people have convinced him that many have none of the common risk factors that go along with the condition - high blood sugar levels, high blood pressure and high cholesterol levels.

In his studies, Leibel said, he purposely recruits people who are obese but have no other health problems. That way, he explained, when he studies the effects of weight loss he reduces the likelihood that his data will be confounded by other disorders.

"You can find people like that - they're out there," Leibel said. "People would say, they must be dyslipidemic, they must be diabetic," he said, referring to abnormal blood lipid and blood sugar levels that are supposed to be endemic in the obese. But, he said, "they are not."

And it is not just humans who can be inured to the effects of excess body fat. "There are mouse strains that are infamous for this effect," Leibel said. When scientists give one strain, the c57 black 6j mice, an obesity gene, the animals grow hugely fat but never develop diabetes. But if they give the same gene to another mouse strain, called kayliss, the mice not only grow fat but are dead of diabetes in five months.

Both strains of mice develop insulin resistance, the so-called pre-diabetes state that often afflicts fat people, Leibel said. But the c57black6j mice compensate - their pancreases produce more insulin. The kayliss mice try, but fail, to make enough insulin to protect themselves.

Many obese people do not even have insulin resistance, said Ethan Sims, an obesity expert who is an emeritus professor at the University of Vermont. In a recent paper published in the journal Metabolism, he described a study at the University of Vermont's medical school of 43 sedentary, postmenopausal, obese women. The women had participated in obesity research at the university and were not recruited because of their insulin sensitivity or resistance. But, he said, 17 of them turned out to have completely normal insulin responses.

Reubin Andres, the chief of the metabolism section of the intramural section at the National Institute on Aging, said there was a crucial question about such people that he now wants to address: What happens to them over a long period of time? Do they eventually develop diseases, like diabetes or heart disease, that are more common among the obese? What is their mortality rate?

The National Institute on Aging is analyzing its data to get some answers, Andres said. Since 1958, it has been following thousands of people, gathering data on their health, including body weights and blood sugar levels. "It is important to know what this population is like and what happens to them," Andres said.

At the National Institute of Diabetes and Digestive and Kidney Diseases, a different question is being asked: Do fat people who already have diabetes and who lose weight become healthier? In particular, do they have a lower risk of heart attacks and strokes, does their diabetes improve, and do they live longer?

Although there is a widespread belief that weight loss will improve health, a number of large studies that looked at the health of the formerly fat have raised questions about whether that is true.

Andres explains that weight loss can improve blood sugar levels, blood pressure and cholesterol levels.

"On a population basis, you can expect all of those bad things to improve," he said. "It all makes sense. If you lose weight and all these things improve, it has to be good for you." But, he added, "the only problem is that when you look at mortality rates, they don't look good."

Jules Hirsch, an obesity researcher at Rockefeller University, concurred and provided evidence from studies that followed thousands of people for years, keeping track of who lost weight, who kept it off, who became ill and who died. Repeatedly, investigators reported that fat people who lost weight and kept it off had more diabetes and more heart disease and a higher death rate than people whose weight never changed.

"It all does tend to indicate that weight loss is not associated with lower mortality but is actually associated with higher mortality," said Katherine Flegal, an epidemiologist at the National Center for Health Statistics. But, she and others noted, the limitation of the studies is that they cannot distinguish between people who lost weight because they went on diets and those who lost weight because they became ill.

That, said investigators at the National Institute of Diabetes and Digestive and Kidney Diseases, is why the institute is recruiting 5,000 obese people with diabetes for its study and asking how weight loss will affect their health.

"Many studies have demonstrated that short-term weight loss has beneficial effects on risk factors such as high blood pressure and cholesterol," the institute writes in its description of the study. "However, observational studies have raised concerns about negative effects of weight loss and weight cycling. Some of these studies suggest increased, not decreased, mortality; however, most cannot distinguish voluntary from involuntary weight loss."

Hirsch said that, in the meantime, he wished the message could get out that truly fat people really are different from people of normal weight. "There is some sort of extraordinary genetic and environmental mix that has programmed people to be set for greater fat storage," he said. "Of course they overeat. But the significant issue is that they have another illness and the symptom of that is the overeating."

"A reduced fat person is not a normal person," Hirsch said. "If you take two women who both weigh 130 pounds, but one used to weigh 200 pounds and one always weighed 130, they are not the same."

Copyright 2002 The New York Times News Service. All rights reserved.