Researchers spent 16 years evaluating 900,000 people who were cancer-free when the study began in 1982. They concluded that excess weight may account for 14 percent of all cancer deaths in men and 20 percent of those in women.
The study was big enough to back up a fat connection not only in cancers where it has been known for some time, but in eight where it hadn't been widely documented, lead researcher Eugenia Calle said.
Calle, whose study is in Thursday's New England Journal of Medicine, said she was surprised the link "really was the rule more than the exception."
A commentary said the study is 10 times greater than the largest previous research on the topic. Top researchers in both cancer and obesity said the research virtually proves they are linked.
"Because of the magnitude and strength of the study, it's irrefutable," said Dr. Donna Ryan, head of clinical research at the Pennington Biomedical Research Center in Baton Rouge. "It's absolutely convincing. And therefore it's frightening."
Dr. Robert Mayer of Harvard Medical School and the Dana Farber Cancer Institute in Boston said it's not certain whether one in five, six or seven cancers might be prevented or better treated if people lost weight.
"What's clear is that large studies of this sort — and this is the biggest and best to date — show very clearly this is a major health problem in this country," said Mayer, speaking for the American Society of Clinical Oncology.
The study by American Cancer Society relied on the body mass index using heights and weights reported by study participants. For instance, a 5-foot-11 person who weighs 175 pounds would have a BMI of 24.4, near the top of the normal range. A 5-foot-3, 175-pounder would be obese, with a BMI of 31.
For the study, a BMI of 18.5 to 24.9 was considered normal. Those who were overweight (25 to 29.9) or obese (30 or over were all compared to the normal group, and statistical analysis was used to adjust for smoking and other risk factors.
Earlier studies have found that excess weight contributes to cancers of the breast and uterus, colon and rectum, kidney, esophagus and gall bladder. This one also linked it to cancers of the cervix and ovary, multiple myeloma, non-Hodgkins lymphoma, pancreas, liver, and, in men, the stomach and prostate.
The researchers found no link between fat and brain, skin and bladder cancers.
There are two big reasons the overall link is stronger in women than in men, Calle said.
"More women are obese," she said. "And also, breast cancer plays a pretty big role here. That's obviously one of the most common cancers."
Too much body fat can influence cancer and cancer mortality a number of ways. It increases the amount of estrogen in the blood, increasing the risk of cancers of the female reproductive system. It increases the risk of acid reflux, which can cause cancer of the esophagus. It raises levels of insulin, prompting the body to create a hormone which causes cells to multiply.
Obesity also makes cancer harder to diagnose and treat. It's harder to see or feel lumps and bumps, and some patients don't fit into CAT scanners, Mayer said.
They also may avoid regular doctor's visits, "possibly because of their appearance or they just shy away from physicians," he said.
"The morbidly obese are harder to operate on, harder to plan radiation therapy for — often, they don't even fit into a radiation therapy machine," Mayer said. He said it also is hard to decide the right chemotherapy dose for the obese, because fat tissue sometimes absorbs the chemicals.
Both Ryan and Calle said attitudes must change about weight the way they did about smoking. They said communities, workplaces, schools and transportation all need to change to make it easier both to eat right and exercise. "We've developed a culture where you have to work really hard to eat right and exercise," Calle said. "We're kind of stacking the deck against ourselves.
"Until we accept that it is a bigger problem than one of individual discipline, we probably won't be too successful in turning it around."