Tuesday, November 19, 2013

Study shows paying people to lose weight works. Or does it?


A new study looked at the effect of paying people to lose weight.

The authors randomized 100 people with BMIs ranging from 30 to 39.9 into four groups. Two groups received weight-loss education, one group with and one without payment. The other two groups received education plus behavior modification with again one group receiving financial incentives and the other not. To remain in the study, they were all supposed to have lost 4 pounds per month. Patients in the two financial incentive groups received $20 per month if they met their goal, and those not meeting the goal had to pay $20 per month which was pooled for a lottery among the participants at the end of the study.

A significantly larger percentage of those receiving remuneration completed the study. At the study's endpoint—12 months, the average weight loss for those in the paid groups was about 9 pounds compared to just over 2 pounds for the two unpaid groups. Using a two-way ANOVA, the incentives were estimated to have led to a weight loss of 6.5 pounds, which was statistically significant with a p value < 0.001.

The authors concluded, "Sustained weight loss may be achieved with financial incentives."

The paper was presented at the American College of Cardiology meeting last March and is available only in abstract form.

The study raises some questions.

The paid groups lost less than 1 lb per month. If the subjects were to have lost 4 lbs per month, why didn't they lose a minimum of 48 lbs, which would be 4 lbs x 12 months?

How durable was the weight loss? In other words, after the monetary incentive stopped, did the subjects regain the weight? It is well-known that many people regain weight after they go off their diets.

Does this study actually show that education and behavior modification are not very useful promoting weight loss? Then why should anyone bother?

It's one thing to do a study of 100 people, but if money truly is a good way to get people to lose weight, who is going to pay the millions of obese people in the US?

But here's the real question. How clinically important is a 6 to 9 lb weight loss for someone with a BMI of say 35?

If a man is 5'8" tall and weighs 230 lbs, he has a BMI of 35. If he loses 6 lbs, his BMI drops to 34. Does that decrease his risk for diabetes or hypertension? I think not.

This may be another example of a statistically significant result that is very likely not clinically important.

6 comments:

Anonymous said...

You right this study looks like a complete failure, the median weight loss value is not shown but I would be curious to know if any of the paid group actually reached the target to normally "stay" in the study. Hence the miserable attempt to save the study with meaningless statistics.

Skeptical Scalpel said...

Anon, you are probably correct. To be fair, it would have been nice to have read the entire paper. The reason I wrote this is that it was cited by some people on Twitter. I thought that comments were warranted.

DD said...

There are health benefits from losing even 5 to 10% of body weight. The method you describe in your post (paying people) is questionable but my comment reflects the need for providers to encourage modest weight loss due to its benefits.

It is also well established that for those with BMI exceeding 40 (morbidly obese) surgical treatment works best for weight reduction, improvement in comorbid conditions and overall health. Albeit surgery is not without risk, but, when performed at a bariatric surgery center that includes follow up, the complications are low.

Skeptical Scalpel said...

DD, I agree weight loss should be encouraged.

I'm not so sure that a 5% weight loss leads to health benefits if you are 300 lbs. That would get you all the way down to 285.

Surgery is probably better for the super obese.

Anonymous said...

Wonder if these companies have vending machines with junk food in them? I have a feeling that these companies could have saved themselves a bit of money just by ripping out the vending machines, and making the elevators run slower, so that it was more efficient to use the stairs.

Skeptical Scalpel said...

Most places, even most hospitals, have vending machines with junk food. That's a huge part of the problem.

Post a Comment

Note: Only a member of this blog may post a comment.