WHY IS THE standard treatment for obesity -- diet and exercise -- so rarely effective? The doctors who prescribe it have a simple answer: They blame the patient.

Lots of patients do sneak snacks or skip knee bends. But even for the most highly motivated, best disciplined patients, the statistics are dismal: less than 5 percent manage to lose any significant amount of weight, and less than 10 percent of these successful patients -- one in 200 who start a traditional weight-loss program -- are able to keep their weight down for a significant time.

Imagine this kind of record in the treatment of any other disease. Cancer, for instance: If a cancer therapy had this low a success rate, we would discard it and presume that the few cures had occured by spontaneous remission.

Yet the traditional diet-and-exercise therapy persists. Doctors, dietitians, self-help diet clubs, best-selling books -- all rely on some variation of calorie counting and exercise. If you haven't lost weight, it must be that you haven't followed the rules.

But obesity is not that simple. For many patients, it may be that the treatment doesn't work because the theory is wrong. That doubt has finally brought capable scientists around to examining the question. As a result, we are just now approaching an understanding of what makes fat people fat.

First, consider people who maintain a normal weight. They usually do so without any particular conscious effort. Although they are often careless about their eating habits and although their eating and activities vary enormously, their weight remains stable.

An increase of 10 calories of food a day theoretically should increase a person's weight about a pound a year; 40 pounds by an adult's middle years. Most people weight is much more stable than that. this stability is much greater than most of us could achieve simply by conscious effort. With even herculean attention to-detail it is not possible to be that accurate in your calorie counting.

It seems inescapable that some kind of nonconscious regulators are functioning in normal-weight humans. These regulators assembly information about caloric intake and activities and send the body various signals to control calorie balance. These signals, controlling hunger and satiety, for example maintain yourr weight at a fixed level. All other mammals have these regulators. It would be a biological curiosity, and a very destructive one, if this kind of system did not function in humans.

Many fat people have normally functioning regulators set at abnormal weight.

Various medical or psychological events (pregnancy, puberty and divorce are good examples) appear occasionally and unpredictably to shift the regulator to an abnormal setting. Most people who are fat tend to keep their weight stable if abnormally high, and to return to about that same weight after loss. Nothing about losing weight appears to change the regulators.

Recently a new type of regulator has been explored and the disease should become easier to understand. Again, consider normal people. Many studies show that normal volunteers, when overfed even by huge quantities, do not gain what they should by caloric calculations. In one study some lean adults overfed by over a million calories during a 200-day experiment did not gain weight. The excess calories somehow seem to be dissipated.

The new studies identify how lean animals waste the unneeded calories: They burn them, increasing the skin's surface temperature, and radiate the heat away. There appear to be at least two mechanisms for the creation of heat. One is anatomically localized in a special form of fat, known as brown adipose tissue, located just beneath the skin. A second, which involves the transfer of sodium and potassium into and out of cells, appears to function in all cells. Both systems function abnormally in obese animals and humans. Obese animals do not convert thier excess calories into heat, while lean animals can do so readily.

These remarkable discoveries explain the paradox of obese rats who cannot survive in a laboratory cold room while lean rats, with no insulation, do fine. The fat rat cannot generate enough heat; the lean rats have an excellent system for creating heat. For many fat people, the problem may not be insufficient calorie burning because of no exercise, but because of abnormal heat creation.

Calorie needs for humans may vary far more than we have assumed. Since heat creating may be abnormal in many fat humans their caloric requirements may be much lower than we have ever calculated, perhaps too low to allow them the luxury of a normal diet. They may be right that "everything I eat turns to fat," and it probably is true that at least some fat patients really don't overeat. Surely some obesity is caused by overeating. But for many fat patients a spartan and religiously cautions intake may be more than enough. This will be biologically protective if we have a famine, but in 20th century America it is a biological disaster. We no longer die of starvation. We now die of degenerative diseases and these diseases are almost always made worse by obesity.

A series of new studies also demonstrate that fatness is associated with abnormalities in various parts of the bocy's chemistry. Hormone function is different in obesity. Some hormones prevent the body from losing weight; they function to sustain fatness. Small endorphin molecules, which function as one of the brain's neurochemical information transmitters, are probably abnormal in some types of obesity. Substances which prevent the endorphins from functioning in their usual way correct obesity in certain types of fat rodents.

Many people, including physicians and patients, are convinced that obesity is an emotional disorder. Obese patients usually identify situations in which their food intake gets out of control: "I eat as a reward" or "I eat when I am depressed, lonely, bored, etc."

Psychological factors have always been assumed to be characteristic of obesity, but many of the assumed differences have never been verified. Is there a personality pattern or emotional disorder associated with obesity? Are obese people fat and jovial? Isn't it an addiction just like cigarettes, alcohol or even drugs? The answers here are probably "no." Personality factors seem to play a very small role in the generation of obesity. Many studies indicate that fat people are much like everyone else. Personality traits which do exist in obesity appear only to be those which form in rection to a chronic disease. The major role of stress and emotional factors is not to cause obesity but to make it more difficult to control.

Despite this diversity, some regulator abnormalities probably exist in most forms of obesity. The normal regulators collect an enormous amount of information. They assess food intake, and compare it to our activity, exercise and rest. They gauge the social situations and stress of our lives. They send hunger and satiety signals that control our eating behavior. They secrete hormones and neurochemicals to affect our metabolism. They regulate our system of generating and wasting heat in response to excess calories. Considering the enormous variation in our patterns of food and activity, they do all this with remarkable precision. In one way or another, these regulators are probably malfunctioning in fat humans.

All of this doesn't mean that dieting and exercise should be scrapped. "Eat less and move more" is still good advice. Newer variations on this theme, particularly comprehensive programs using aggressive very low calorie diets (modified fasting) together with training in behavior modification, are achieving control in as many as 50 percent of those who start. The old therapy can be improved and it can succeed more often, but future therapy will depend on a more direct attack on the faulty regulators.

Take hope. Rational medications are likely to be developed soon. Perhaps we will even be using brown adipose tissue transplants to enable obese patients to burn their calories. Consider the irony of fat transplants to cure obesity.