Is ‘fit and fat’ a nonissue?

I wasn’t even going to bother mentioning this study — it was a data dredge looking for correlations using self-reported data from an unrelated clinical trial that was never designed to even examine or answer this question. Plus, the findings were so contrary to the body of evidence, I never imagined anyone would take it seriously.

Silly me.

The mainstream media, such as CBS News and WebMD, have been reporting that this study provides evidence that fat women are doomed to have heart disease, whether they exercise or not. “The findings underscore the importance of being fit and trim,” they’re telling us, and show that it’s impossible to be fat and fit. “Regardless of body weight, these data highlight the importance of counseling all women to participate in increasing amounts of regular physical activity and maintaining a healthy weight to reduce the risk of coronary heart disease,” they’re reporting. Worse, they say, only a third of the women supposedly met the Surgeon General’s guidelines for being physically active.

Not so fast...

Let’s read the study

This study just published in Archives of Internal Medicine, didn’t actually follow nearly 39,000 women for eleven years and find that fat and sedentary women had higher rates of heart disease, as many believe. Dr. Amy R. Weinstein, M.D., MPH and colleagues in Boston, selected data from questionnaires that had been gathered on women (average age 54 years) when they were enrolled in the randomized, controlled clinical trial of low-dose aspirin and vitamin E for the primary prevention of cardiovascular disease and cancer (the Women’s Health Study). They then plugged their data into a computer model, looking for correlations between physical activity and heart disease.

But the original randomized clinical trial hadn’t been designed to evaluate physical activity, so, not surprisingly, it hadn’t measured activity levels or followed them through the eleven year follow-up period. It had, however, found that BMI was not related to clinically confirmed cardiovascular disease, heart attacks, strokes, ischemic strokes or major cardiovascular events. [More about that in a minute.] So, how did Dr. Weinstein and colleagues manage to find a link? Only their computer model knows for sure, but we can begin to gain some insight by asking two questions: How was heart disease defined? How was activity defined?

Heart disease was “defined as a cardiovascular event including nonfatal myocardial infarction (MI = heart attack), coronary artery bypass graft, percutaneous transluminal coronary angioplasty, or coronary heart disease death.” They didn’t include the data on stokes, ischemic strokes, fatal MIs, or all-cause mortality from the original trial.

Physical activity was defined by asking the women to estimate the average time per week spent during the past year on 8 specific recreational activities (that included aerobic dance and exercise, swimming, tennis, squash, racquetball, bicycling, running, jogging, walking and hiking) and flights of stairs climbed per day. No activity was actually measured, it was all self-reported and narrowly defined as these specific exercise activities. As we’ve examined in a previous article, counting only leisure time activity also disregards work-related physical activity and other non-“exercise” physical activity. Recreational and sports activities often reflects socioeconomic status, but these authors didn’t factor in or adjust for social-economic status. The authors also said they didn’t include occupational physical activity because female health professionals would be expected to expend relatively little physical activity at work. [(!) They clearly have never spent a 12-hour shift as a nurse, running nonstop, constantly bending and lifting, and never even stopping to take a pee break! But I digress...]

They then reported the correlations, not even as relative risks, but as hazard (odds) ratios (described here]. These are a way to compare odds, rather than actual incidences, and can greatly exaggerate risks. The authors reported that increasing levels of activity correlated with significant reductions in risk for heart disease among obese women but that exercise didn’t completely eliminate the risks. They concluded: “Our study strongly supports current physical activity guidelines recommending 30 minutes of moderate activity daily to reduce chronic disease. This level of activity improved CHD risk regardless of baseline BMI. Furthermore, this study suggests that more than 30 minutes of physical activity daily further reduces the risk.”

But their findings didn’t support their case. There was no dose-response to exercise — more wasn’t better, it was worse. [Click on image to enlarge.]

While the hazard ratios their computer somehow derived (the “multivariate adjusted” column on the right that adjusted for age, smoking, hormone use, etc.) were higher among the obese women (BMI≥30) — there was no difference in cardiovascular disease among the ‘obese’ women doing no exercise and those doing the most. Plus, the most active ‘obese’ women, exercising 4 or more hours a week, could cut their risks nearly in half by exercising less. In fact, the lowest risks were those doing a mere 1 to 11/2 hours a week. This very same phenomenon was seen in the ‘normal’ weight women (BMI<25), who had increased risks exercising 30 minutes a day or more. [The Boston researchers were called on this back in 2001, when they released a nearly identical paper from this data. (See below.)]

In their report, they tried to explain away these findings by saying “the association between physical activity and CHD in normal-weight individuals remains unclear. Although the hazard ratios trended downward with increasing activity levels, our study was underpowered to detect small differences in risk between the normal-weight individuals.” But this is an uncredible argument, as there were 2.7 times as many ‘normal’ weight women as obese. So, they’re asking readers to ignore their evidence. Instead, they concluded that their data highlights the importance of counseling all women to do increasingly amounts of exercise and that fat women need “concurrent weight loss” and to maintain a “healthy weight” — even though their study didn’t show any benefit of increasing exercise, nor did it examine weight loss, either!

Reading the text reveals a couple of other paradoxes:

Obesity Paradox #13: They said they used only nonfatal heart attacks in their definition for heart disease because when they included the fatal heart attacks in their analysis of correlations with obesity, the “magnitudes of effect were weakened.” In other words, the obese women were more likely to survive heart attacks and it weakened their negative correlation between obesity and “cardiovascular disease,” as they defined it. This is the only clue the Women’s Health Study has let out that obesity offers a survival advantage. That’s an interesting thing about the Woman’s Health Study out of Brigham and Women's Hospital and Harvard Medical School — its methodology describes the great lengths they went to to gather mortality information, but not one of the dozens of published paper from this study has reported all-cause or any mortality data in relation to BMI or physical activity. Even when they reported that physical activity was unrelated to breast cancers among these post-menopausal women, they still didn’t reveal mortality data.

Obesity Paradox #14: Dr. Weinstein and colleagues also said it didn’t matter that they didn’t track BMIs throughout the eleven years of the study, because people’s weight status predictably remains constant even though weight gain with aging is normal for all. In other words, although virtually all women are dieting and trying to lose weight, they admit long-term weight loss is known to be ineffective. This is exactly what Dr. Katherine Flegal, Ph.D., and colleagues at the CDC’s National Center for Health Statistics reported in a 2005 issue of the American Journal of Public Health, in response to claims of an impending health crisis because the long-term effects of obesity aren’t known. CDC data reveals that few Americans actually lose weight and keep it off long-term, so the long-term effects of obesity are already known. [And, the body of evidence, has consistently shown that the natural weight gain women experience as they get older is associated with living longer.]

The original randomized clinical trial found no link

A point that everyone seems intent on ignoring is that the results of the original randomized, controlled clinical trial found no tenable correlation between BMIs and heart disease. Just as Dr. Ancel Keys, Ph.D., had shown more than 50 years ago. This perpetual focus on weight and heart disease is much ado about nothing.

The published Women’s Health Study clinical trial data provides a valuable opportunity to see how easily we can be mislead to believe risks are huge when they’re reported as relative risks or odds ratios, rather than telling us the actual numbers. Over the eleven years that the 39,876 older women were followed, 3.1% (1,251) died from all causes and only 20% of those (246) were due to cardiovascular causes. In other words, 0.6 % of the women died from heart disease during the 11 years.

Looking at actual incidences of strokes reported, for example, only 1.1% of ‘normal’ weight women suffered a stroke (221 of 19,849), compared to 1.4% of the ‘obese’ women (103/7,126). When reported as a relative risk, that’s a 30% higher risk associated with obesity. But it is an actual difference associated with obesity of 0.34% after 11 years or 0.03% a year difference — clinically meaningless and statistically untenable.

And 0.7% (140) of the ‘normal’ weight women had heart attacks compared to 1.27% (91) of the ‘obese’ women. Reported as a relative risk, it’s an 81% higher risk associated with obesity. But it’s an actual difference associated with obesity of 0.55% over 11 years or 0.05% a year difference — again, not clinically meaningful.

Credible evidence-based medical care does not take an untenable correlation and claim it proof of a causation and then flop it around into a treatment recommendation. The take-home lessons for us are to always ask ourselves what researchers are actually measuring and if they are asking the right questions. Next, always be suspicious of secondary analyses using data from a study that wasn’t designed to answer that question and is trying to splice meaning from untenable findings. Finally, don’t fall for scares based on relative risks or odds ratios, keep the actual data in perspective.

Might the benefits of exercise be being overblown?

The Women’s Health Study authors published a similar paper in the 2001 issue of the Journal of the American Medical Association, concluding that more exercise was associated with lower risks for heart disease (again, not revealing all-cause mortality). But, the actual findings showed the lowest risks with minimal exercise (an hour to an hour and a half each week of walking) and risks increased with more intense or long-duration activity. As Dr. Gerson T. Lesser, M.D., with the Department of Geriatrics and Adult Development at Mount Sinai School of Medicine in New York, wrote in JAMA:

The results of Lee et al. also show no trend to improved outcomes for progressively more vigorous activity... the group that expended 600-150 kcal/week actually had better outcomes than the group that expended 1500 kcal/week and greater. People older than 65 years who participated in social and productive activities, usually involving minimal physical activity, have been found to have significantly lower mortality over 13 years than nonparticipators, and the magnitude of mortality reduction was quite similar to the reductions of CHD noted by Lee et al. This near absence of dose-benefit relationship weakens the cause-effect hypothesis.

Self-evaluation of health has been found to be a significant predictor of mortality in longitudinal studies.

Dr. Lesser went on to explain how people who exercise differ in many ways from those who don’t and exercise may be a marker for any number of other factors. So, the correlations between exercise and better outcomes that have been widely reported in observational studies may be giving a false sense of causation. He is not alone in suggesting that the significance of “exercise” to longer, healthier life may be being overblown and obscuring the health benefits associated with other less physical activities.

The 13-year study he exampled was led by Dr. Thomas A. Glass, Ph.D., at Harvard University School of Public Health. The authors set out to examine the link between activity level and survival (not surrogate endpoints). They took a random sample of 2,761 men and women, aged 65 years and older, living in New Haven. Beginning in 1982, the researchers were seen face-to-face by investigators (blinded to the study hypothesis) every year for 13 years. The study authors then compared actual mortalities and the relationships to social activities (going to movies, restaurants and sporting events; attending church; playing cards, games and bingo; participating in social groups, and other social activities), fitness activities (active sports, walking, physical exercise of any type) and productive activities (everyday work such as preparing meals, shopping, community work, gardening, employment). They controlled for sociodemographics, age, marital status, education, race/ethnicity, and income.

What they found was that social and productive activities had about the same effect on mortality as exercise and fitness activities.

“Social and productive activities were observed to confer equivalent survival advantages compared with fitness activities,” they concluded. “This observation is important because it suggests that activities that entail little or no physical exertion may also be beneficial. A wider range of mechanisms, both physiological and psychosocial, may be involved in the association between activity and mortality than had been previously thought.” To rule out that some of the productive activities, such as gardening and shopping, could be contributing to cardiovascular fitness, they also analyzed their data excluding physically active social and productive activities.The beneficial correlations held.

Their discussion is worth reading, as this information rarely makes its way outside of the medical literature:

Previous studies of activity and mortality have assumed that physiological pathways mediate this association. This has led to substantial investment in exercise interventions in older people. The activities in which older people engage, however, result in a complex array of effects beyond improved fitness. Social activities may involve a broad range of goals, including leisure and enjoyment, reinforcement of social status and sense of worth, social engagement, and productivity...

Psychosocial pathways may be involved in the health benefits of activity. Recent research has shown that social contacts influence several biological factors. Substantial evidence indicates that social contacts may reduce the deleterious effects of psychological stress through enhancement of both cellular and humoral immune response. Secondly, social and productive activity involves the performance of meaningful social roles. These roles may have important consequences that go beyond the benefits of fitness alone. Meaningful social role performance promotes a sense of self efficacy that has been linked to several important health outcomes in later life. Social and productive activity may result in a sense of meaning and purpose in life, which has been linked to survival in at least one study. Finally, social and productive activity reinforces relationships as well as norms of reciprocity and mutuality. Social and productive activity may enhance social networks and social support, which have been linked to survival in several prospective studies.

Our findings corroborate those of a small number of studies that have found a link between survival and social activities that entail little or no physical exertion. This finding has important implications for public policy and clinical practice. Clinicians might recommend a broader range of activity options for older people. For patients with chronic conditions such as arthritis social activity may promote wellbeing more effectively than physical activity.

This wasn’t a clinical trial, of course, and they point out in a later issue that definite conclusions about cause and effect are unwarranted. However, “these findings build on a substantial body of studies that show the benefits of remaining active in later life,” they wrote. But it’s how “active” is being defined that needs closer consideration. The “exclusive emphasis on exercise and fitness activity may be overly narrow,” they said.

Although this outlook counters what has become almost a creed in our culture, it is what many doctors have been saying all along. Those were the conclusions of the renowned Dr. Keys, for example, who said in a University of Minnesota Update that he didn’t advocate exercise prescriptions. “I try to persuade people to have some pleasurable, safe exercise, preferably useful, when they are older,” he said. “Margaret and I get lots of pleasure from working in our yard,” he said.

As Dr. Dean Edell, M.D., author of Eat, Drink & Be Merry, reminded readers, people who are living their lives to the fullest are active and doing things, but it's a very rare centarian who has spent his/her life “exercising.” As he wrote:

You need less exercise than you think to accrue most of the health benefits. Exercise is not a magic cure for all that ails you and it can even have some very negative consequences... I place almost the highest premium on having fun.

© 2008 Sandy Szwarc