Nutrition news changes more quickly than the Buffalo restaurant scene.

No wonder there is confusion.

Amid the deluge of health information about the link between diet and disease, it is difficult to separate fact from fiction, well-grounded theory from wishful thinking.

Is all fat bad? What about cholesterol? Does that cup of coffee pose a danger?

As a pioneer in epidemiological research, the study of how disease affects groups of people, Saxon Graham has spent his career at Roswell Park Cancer Institute and the University at Buffalo pondering those questions. Many of his studies have contributed to the confusion, too.

"Science is a work in progress," he said. "Conclusions come from evidence that accumulates over many years, not from the latest finding."

It was Graham, a professor emeritus of social and preventive medicine at UB, who in the late 1950s used sophisticated interviewing methods on large groups of people to find the potential health benefits of vitamin A and in the early 1970s the protection a diet high in raw vegetables offered against some cancers.

The findings seem obvious today. But at the time, they were not.

To be sure, food was something to be eaten, not studied. Most scientists did not accept research that was not based on animal or human tissue studies.

"Thirty years ago, the experts -- the mouse men -- argued that you couldn't study diet and disease the way I was doing it," Graham said. "We helped people realize that human behavior, not just experiments with rodents or test tubes, can lead to higher or lower risk for disease."

Colleagues and others noted Graham's contributions last week when he received an honorary doctorate of science as part of sesquicentennial celebrations at the UB School of Medicine and Biomedical Sciences.

"He's a giant in his field," said Dr. John Naughton, dean of the Medical School and UB's vice president of clinical affairs. "Epidemiology, as the name suggests, used to deal with epidemics of infectious diseases. Saxon led the way in using it to study diet and disease."

Graham worked at Roswell Park in the late 1950s, when he noticed relationships between ethnicity and economic status and disease -- for instance, that Polish people seemed to develop some cancers more often than other groups.

He began to ask himself whether the differences could have resulted from different diets.

But his work went largely unnoticed until the late 1970s, when Sir Richard Doll, a world-renowned epidemiologist at Oxford University, wrote about it. When Doll talked, people listened.

After four decades with a ringside seat in the scientific arena, Graham has disappointing news for anyone hoping a special list of foods or supplements will keep them thin and healthy -- there is no magic bullet.

More research is needed, but current scientific evidence does offer sound dietary advice:

Eating foods from plants is good for you, although it is not known yet exactly what ingredients protect against disease.

Diets high in fat increase risk for heart disease and a host of cancers, including colon and prostate.

Carotenoids, the colorful pigments that make squash yellow and spinach green, in food form, not in supplements, may reduce the risk of some cancers.

High-fiber diets -- plant more so than grain fiber -- decrease the risk of cancer.

Obesity can lead to many illnesses, including heart disease, diabetes, hypertension and cancer.

Scientists do not know enough to say anything confidently about the potential harmful or beneficial influences of such foods as coffee, tea, fish or alcohol.

"I can show you 15 papers that show booze causes cancer and 15 papers that show it doesn't," said Graham. "We probably won't know what the answer is until the evidence adds up slowly over many

Graham diverges from some of his colleagues on the subject of red meat.

"We've overstated the risks," he said. "There is a puritanical streak among epidemiologists. We tell people not to smoke, not to drink, not to eat red meat. But I don't see the data showing red meat is that bad.

"Even if there is a small risk, life is full of risks. All I know is, I had a steak last night, and it tasted pretty good."

As boring as it sounds, Graham says the secret to health is simple: avoid cigarettes, exercise regularly and eat a balanced diet.

If Graham has a pet peeve these days, it centers around the common misconceptions about cancer.

There's the notion of having cancer in the family. Said Graham: "It's a myth."

To be sure, inherited genetic defects can predispose people to specific cancers. But cancer is a diverse assortment of diseases in many sites throughout the body. There's no such thing as cancer, in general, running in a family.

Graham doesn't know whether or when cancer will be cured. But he contends the nation's strategy for fighting the disease doesn't make sense.

A significant portion of the cancer cases in the United States stem from cigarette smoking and poor diets. But the war on cancer focuses nearly all of its resources on high-tech cures, rather than on prevention and education.

"The research community is oriented toward clinical trials," he said. "That's important. But we haven't placed enough attention on getting people to stop smoking and eat healthy."