(Greenville News) -- Four days a week, Dale Wallace dons a sweatsuit and climbs on the elliptical machine at the gym.
Her leg muscles burn as she pumps, so she focuses on the TV and jokes with a pal to push through the pain.
The 59-year-old Easley woman has lost about 20 pounds in the last three months, and she wants to lose 30 more. It's impossible, she said, to ignore the constant messaging that being overweight is bad.
"All your life you're programmed to think, 'I need to be skinnier. I need to lose weight,' " Wallace told GreenvilleOnline.com. "Especially women."
So when she learned about new research that holds that overweight and moderately obese people may live longer than their normal-weight counterparts, she was dumbstruck.
"I am shocked by that," she said. "You don't ever hear that."
Published in the Jan. 2 issue of the Journal of the American Medical Association, the meta-analysis examined 97 studies on weight and mortality covering nearly 3 million adults in the United States, Canada, Europe, Australia, China, Japan, Brazil, Israel, India and Mexico.
It concluded, as expected, that people with a body mass index above 35 - or the most obese - had a 29 percent higher mortality rate than people of normal weight, or those with a BMI of 18.5 to 25.
But it also found that people who are overweight, or those with a BMI of 25 to 30, and the moderately obese, or a BMI of 30 to 35, had a 6 percent and 5 percent lower risk of death, respectively, than people of normal weight.
And while the research may offer comfort to some, it set off a firestorm of controversy.
Dr. Walter Willett, chair of the Department of Nutrition at the Harvard School of Public Health, called the study "a pile of rubbish."
Colorado law professor Paul Campos, author of "The Obesity Myth," meanwhile, called it "a very careful meta-analysis of an enormous amount of data."
Lead author Dr. Katherine M. Flegal of the U.S. Centers for Disease Control and Prevention's National Center for Health Statistics stands behind the research.
"We looked at almost 100 studies all over the world and they were all in peer-reviewed journals," she told GreenvilleOnline.com. "They were not a bunch of junk."
A third of American adults are overweight and another third are obese, with the highest rates in the South, according to the U.S. Centers for Disease Control and Prevention.
And excess weight has been linked to diabetes, cardiovascular disease and other health conditions.
Dr. Steve Newman, medical director for Bon Secours St. Francis Health System's Medical Weight Loss program, said the research is intriguing.
"It does have some clinical implications. But we have to be careful not to jump to conclusions," he said. "What we don't want to do is allow this article to justify an unhealthy lifestyle because obesity is still a major contributor to morbidity and mortality."
The researchers said their findings "are consistent with observations of lower mortality among overweight and moderately obese patients" - that is, many clinical studies show that slightly heavier people have better survival rates when they are sick.
And one possible explanation, they say, is that heavier people may see their doctors earlier than thinner people and therefore are treated earlier. Another is that some increased body fat could be beneficial in fighting off disease.
In an accompanying JAMA editorial, Drs. Steven B. Heymsfield and William T. Cefalu of the Pennington Biomedical Research Center in Baton Rouge, La., pondered whether being overweight as defined by BMI can "have a protective association with mortality."
While it makes sense for people 65 and older to have a little fat reserve in case they get sick, according to Dr. Sandy Hardee, medical director of Greenville Hospital System's LoseWell program, that's not true for the general population.
"I believe this study sends the wrong message to patients," she said.
Measure for measure
Newman said health-care providers should consider the implications of the study, including whether obesity should be defined as a higher BMI.
"It's not saying being obese is OK," he said. "It's saying we may need to redefine what we say is obese in terms of health and mortality."
Willett said the research provides no reason to rethink the definition of being overweight.
But one problem Newman sees with the study is using BMI - which is calculated as weight divided by height squared - in the first place.
It's not the most accurate way to measure body fat, he said. For example, a top athlete could have a higher BMI than a couch potato because muscle weighs more than fat, he said.
"I'm not saying BMI is not useful. In general, BMI is fairly accurate," he said. "But you always want to put it in context. You use it as a tool and look at the person behind the number."
Hardee holds a similar view.
"My main concern with this analysis is that it looks at just one factor. BMI is one indicator, but it's not a perfect parameter," she said. "It's just one piece of data when there need to be additional markers."
Other things that should be considered include waist circumference, waist-to-hip ratio and abdominal fat, along with blood pressure and heart rate, she said.
Flegal said she studied BMI because it has a standard international definition. And while she agrees it's not perfect, she said it's "a pretty good" measure of fatness.
"It's no worse than others, like waist circumference. They are related to each other very closely and all correlated to some extent with body fat," she said.
"Interpretation is more what we ought to be talking about ... because the words are loaded. Overweight especially. People don't have good understanding of what BMI we're talking about."
But Willett, who dismissed Flegal's work as rubbish, said the study is flawed and misleading.
"In the last several years, two other major analyses ... have been conducted on the relation of body weight to mortality," Willett, who could not be reached for comment, wrote on the school's website.
"These involved the collaborative efforts of over 150 scientists and were far superior to the recent JAMA paper because they combined the original data from over 70 studies," he continued.
"After eliminating distortions due to the effects of serious disease on body weight, these studies showed clearly that both overweight and all grades of obesity are associated with increased mortality."
But Campos, who argues that the weight-loss industry has fomented an "obesity panic message" to generate profits, said Willett's reaction is indefensible.
"I don't think that it's an appropriate reaction for an academic to call a study - especially one published in the most prominent medical journal in the U.S. whose authors are all distinguished senior scientists at top government institutions - a pile of rubbish," he told GreenvilleOnline.com.
Furthermore, Campos said, Flegal's research "merely points out there is no correlation between increased mortality and weight until you get to a quite high weight level, and the vast majority of people classified as weighing too much are not at those levels. It's also quite cautious about drawing any strong conclusions."
Willett's main criticism was that lean people who were ill weren't adequately separated from active, healthy lean people.
"The most serious problem in the Flegal paper is that their normal-weight group included a mix of lean and active people, heavy smokers, patients with cancer or other conditions that cause weight loss, and frail elderly people who had lost weight due to rapidly declining health," he wrote.
"Because the overweight and obese groups were compared to this mix of healthy and ill persons who have a very high risk of death, this led to the false conclusions that being overweight is beneficial, and that grade 1 obesity carries no extra risk."
But Campos said that if Willett would exclude the thin sick people from the study, he should also exclude the sick overweight people.
"There's no coherent methodological critique that's being made here," he said.
Flegal said most of the studies did take the thin sick people into account and concluded it didn't make any difference to the numbers.
"It's hard to argue that every (study) has something deeply wrong with it," she said.
But Hardee, who's been getting questions from patients about the study, said she's skeptical of its findings because there are too many unknowns about the participants, including their illnesses and quality of life.
"It's just hard to wrap your head around that," she said.
Nonetheless, both Hardee and Newman say one of the most perplexing aspects of the research is that it had nearly 3 million participants across multiple countries.
"Certainly large numbers are valuable to a study because you're not going to have one or two outliers that will affect the overall outcome," said Hardee.
"It went across all boundaries - socio-economic as well as cultural," added Newman.
They also agreed that the study did show a higher level of obesity is associated with increased mortality.
Flegal said it's logical that if the study confirmed increased mortality at the highest obesity levels, which was expected, it would also confirm them at the lower levels if they existed.
And while the researchers said the findings "may help to inform decision making in the clinical setting," Newman said he will continue to push patients to be as healthy as they can be.
Hardee said she will too.
"The study will not change my approach to patients who are normal weight, overweight or obese," she said. "I look at the whole patient - weight, height, blood pressure, heart rate, lipid panel, waist circumference, diabetic status, smoker or nonsmoker, etc."
Though Wallace believes people should be judged by their character and heart as opposed to their weight, she said she will continue with her regimen.
"We have always been told it is best to have our BMI to (less than) 30 ... or this could lead to severe diseases down the road," she said. "My main goal was to feel better. And I'm more active. I have more energy. I sleep better. And I feel better."