Blog | Tuesday, August 27, 2013

Does AHEAD set us back?

One of the more prominent items of medical news recently was the final analysis of the Look AHEAD trial, allegedly indicating that a lifestyle intervention and associated weight loss did not reduce heart disease risk in adults with diabetes. There is little we seem to love more than scientific findings that refute the current consensus and imply perpetual confusion, so media uptake of this story was predictably robust. Most of the coverage accentuated the negative, with headlines such as: "Disappointing Results for Weight Loss and Diabetes."

If the Look AHEAD trial truly has debunked the health benefits of weight control in diabetes, then it has potentially set us back rather severely. Fortunately, this ostensibly disappointing news is not, in fact, disappointing, or news.

The trial is not news because these same results were reported, and received widespread media attention, last fall. This matters, because double dipping gives the impression that this is the second trial in the past year to reach this same conclusion. A repetitive drumbeat is much more compelling than a single note, so this repetitive coverage makes the story seem bigger. For reasons not entirely clear to me, the story was reported widely in October 2012, when the study was stopped early, and again now when the data analysis was brought fully to the point of all "I"s dotted, and "T"s crossed. But this is the same study, and these are the same data. We are hearing the story twice, but it is just one story. So this isn't really today's news. It is yesterday's news.

Nor are the results truly disappointing, for a number of reasons. First and perhaps foremost among them is the fact that if you compare two groups that don't differ much in their treatments, you can't expect them to differ much in their outcomes.

Look AHEAD compared an intensive lifestyle intervention to fairly conventional diabetes education in a cohort of over 5,000 obese adults with type 2 diabetes. The data now making news relate to the 10-year follow-up, which proves to be important.

Initially, study participants in the intensive lifestyle group lost an average of roughly 8.5% of their body weight. But at 10 years, that had fallen to an average of 6%, indicating these folks had been gaining back some weight over recent years. In contrast, the control group lost only 1% of their body weight on average at the start, but by year 10 that was up to 3.5%, indicating these folks were gradually losing more weight. Several salient considerations jump out at us.

For the long-term outcomes making news now, the study was comparing a 6% weight loss to a 3.5% weight loss. That represents the effects of a weight loss difference of 2.5% of body weight. If two individuals started the study at 200 pounds each, this equates to looking for a difference in cardiac event rates because one now weighs 188 pounds, and the other now weighs 193 pounds. No one could reasonably expect to see much difference in outcomes with so small a difference in exposures, and frankly, that's just what the study showed: not much difference. In absolute terms, cardiac events did occur a bit less often in the intensive intervention group. The difference was just too small to qualify as statistically meaningful.

Second, given that the intensive treatment group was gaining back the weight they lost early while the control group was losing more weight slowly, the person weighing 193 pounds is now apt to be gradually losing weight, whereas the person weighing 188 pounds is now apt to be gradually gaining. There are physiologic effects of active weight loss and gain that influence cardiovascular risk independent of absolute weight, so this would further narrow the already tiny outcome gap for which we might have hoped. With almost no "exposure" difference, we simply could not reasonably expect much difference in outcome.

Third, as noted, the intervention group did have fewer cardiac events, just not enough fewer to matter in statistical terms. But what that means is: Even that very modest weight loss difference may have conferred a small benefit. It looks that way, we just can't be sure.

Fourth, the study endpoint was a composite of cardiac death, heart attack, stroke, or hospital admission for angina. Risk factors for these overlap substantially, but are not identical. A hybridized outcome measure like this will tend to bias the results of any study toward insignificance.

Fifth, and finally, the authors note that the rate of cardiac events was lower than expected in the entire study cohort. So much so, in fact, that they had to expand the definition of the study outcome so they would have enough events to measure. I trust we all know the adage: If it ain't broke, don't fix it. Closely related is this: If it has already been unbroken, fixing it more may not matter much (or be discernible). If the participants in this study were already being conscious about their health and experiencing a lower than expected rate of cardiac events, it would have been hard to show additional benefit on top of that. This, too, would bias results toward insignificance.

With those five declarations, the defense might rest. But, in fact, there is much more to this tale, as was beautifully expressed by Nanci Hellmich, writing for USA Today. Bucking the "Look AHEAD sets us back" trend that predominated, Ms. Hellmich entitled her article "Diabetes Study Finds New Weight Loss Benefits." That is exactly right, assuming one takes "new" with a grain of salt, for these very benefits were reported last fall as well.

Look AHEAD resulted in reduced rates of kidney disease, eye disease, and depression in the intervention group. There was also improved overall quality of life, fewer hospitalizations, enhanced mobility, and reduced medication use. And all of these benefits were seen despite the modest between-group differences noted above.

There is another point I emphasized when this story broke in the fall, worth reiterating now. The intensive intervention of Look AHEAD was based on the methods of the Diabetes Prevention Program. That study showed this lifestyle intervention, producing an average weight loss of 7%, could prevent the onset of diabetes in nearly 60% of high-risk individuals. Look AHEAD does nothing to invalidate those findings. It just may be that a lifestyle intervention works better when applied earlier. Once diabetes is established, perhaps the returns diminish somewhat, although Look AHEAD indicates there clearly are still returns.

A number of important messages issue from Look AHEAD. One is that very small intervention differences are apt to produce very small outcome differences. Another is that medicine of great value early may be of less value when administered late. Medication that can prevent asthma flairs is useless in treating an acute one. Ditto for migraine. This phenomenon is more common than not. If weight loss is medicine, the timing of its application may matter enormously. Yet even so, there was clear, and almost surprising, evidence of intervention benefit in Look AHEAD.

Yet another message is the importance of how, and even how often, medical news is reported. If headlines say "weight loss is of no benefit in diabetes," that is apt to be all readers take away. If the very same data could be reported as showing that "novel benefits are attached to weight loss in diabetics," then perhaps a bit too much of what we think we know comes down to how things get spun.

We may also be doing a disservice to the power of lifestyle interventions by focusing excessively on weight. While weight loss was a study goal, the real medicine was lifestyle. In the aggregate, the evidence is overwhelming that the same basic lifestyle prescription does prevent heart disease, and diabetes, along with every other major chronic disease.

Look AHEAD tells us that lifestyle medicine, though more powerful overall than all other medicine we've got, is not miraculous. It works as other medicine does, with effects depending on dose, and timing, and disease severity. Some lifestyle as medicine is always better than none, but for specific effects in specific populations, it is clearly possible to administer too little, too late, or both. This in turn simply indicates that lifestyle is real medicine, not a magical mix of unicorn horn and snake oil.

The power of lifestyle to prevent both diabetes and cardiovascular disease outright is irrefutably established. If we truly "get" AHEAD, however, understanding its details in context and look beyond the prevailing headlines, then it in turn should help us get ahead, not set us back. Even a little lifestyle medicine late still confers meaningful benefit.

But I dare hope we might look ahead to earlier, more consistent use of the largely latent power of prevention. Were we to do so, dramatically fewer of us would ever get diabetes. And this truth we hold to be self-evident, whether looked at ahead, behind, or sideways: the diabetes you never get will never increase your risk of heart disease.

David L. Katz, MD, FACP, MPH, FACPM, is an internationally renowned authority on nutrition, weight management, and the prevention of chronic disease, and an internationally recognized leader in integrative medicine and patient-centered care. He is a board certified specialist in both Internal Medicine, and Preventive Medicine/Public Health, and Associate Professor (adjunct) in Public Health Practice at the Yale University School of Medicine. He is the Director and founder (1998) of Yale University's Prevention Research Center; Director and founder of the Integrative Medicine Center at Griffin Hospital (2000) in Derby, Conn.; founder and president of the non-profit Turn the Tide Foundation; and formerly the Director of Medical Studies in Public Health at the Yale School of Medicine for eight years. This post originally appeared on his blog at The Huffington Post.