The underwhelming effect of 5% weight loss on diabetes

There are a couple articles out (from Gina Kolata and Paul Campos) about an 11-year study that looked at the effect of a 5% weight loss on people with type 2 diabetes–specifically, its effect on heart attacks, strokes, and cardiovascular deaths. It ended two years early because there was no effect. There was also no effect on cholesterol, blood pressure, or blood sugar.

There was one difference between the control group and the weight loss group (as an acquaintance was quick to point out when I ‘liked’ the Campos article on FaceBook): the people on the diet-and-exercise program took fewer medications. It’s not clear how big the effect is or even if that was something the study was designed to measure vs. something the researchers decided to look at after the fact* to try to salvage something from the study (the details of the study have not been released yet). It’s a matter of fewer medications, rather than one group taking a lot and the other group taking none, so there’s a limit to how big the effect could be. I doubt that the improvement in quality of life from taking fewer meds outweighs the cost to quality of life of following the diet and exercise intervention. Then there’s the fact that the study ended early–probably a sign that they did not think the intervention was much benefit to patients. I’m mostly familiar with studies ending early when the treatment actually causes harm. If the effect on medication was a big one, I suspect they would have continued the study as scheduled.

And the diet was pretty strict. For people who weighed less than 250 lbs, it was 1200-1500 calories per day. For people who weighed more than 250 lbs, it was 1500-1800. The exercise program was less extreme: at least 175 minutes of moderate activity a week, 25 minutes more than what is currently recommended as a minimum for all Americans.

For reference, on the days when she runs 4 miles, MeMe Roth eats 1800 calories. When she doesn’t, she eats 1300 calories. (I’m using MeMe Roth as an example because there seems to be widespread agreement in both the general media and the fatosphere that her habits are extreme.) The US Department of Health and Human Services’ website says that vigorous exercise such as running counts double–you can replace 30 minutes of brisk walking with 15 minutes of running. Let’s say that MeMe Roth runs 12-minute miles**. 4 miles x 12 minutes x 2 (to convert to moderate exercise) = 96 minutes of moderate exercise. 175/96 = 1.8. So two of MeMe Roth’s runs equals a little more than the minimum of 175 minutes of moderate exercise. Let’s say MeMe Roth goes  for two runs a week, and gets slightly more than the minimum of 175 minutes, getting 192 minutes (96 x 2). So two days she eats 1800 calories, and the rest of the days she eats 1300 calories. Her total calories for the week are 10100, or an average of 1443, on the high end of the 1200-1500 calorie range for people <250 lbs in this study. So… for the amount of exercise they’re doing, the subjects are typically eating a little less than MeMe Roth does.

It’s not exactly comparable, though–my understanding is that Roth runs most days, so both her food intake and her exercise are higher than the subjects’. Longtime maintainer Debra SY also eats about 1800 calories and does a long bout of vigorous exercise almost every day. In Debra SY’s case, she’s maintaining a significant weight loss. In Roth’s case, most of her family has high BMIs and she believes that she is genetically predisposed to having a high BMI; based on the similarity of her calorie intake and exercise routine to Debra SY’s, I’m inclined to agree with MeMe Roth. (There’s a phrase I wasn’t expecting to write anytime soon.) After making this comparison, I think we’ve got yet more evidence that there’s a lot of endocrine stuff influencing weight beyond simple calories in/calories out***–because the study subjects were NOT maintaining radical weight loss. They were only maintaining a 5% loss. On a 200 lb person, that’s all of 10 lbs. (If weight gain is more a symptom than a cause of diabetes, that may be why such drastic measures were needed for subjects to maintain a 5% weight loss.) Even a principle investigator of the study implies that a 5% loss is not a lot compared to what most dieters hope to lose:

“We showed that meaningful weight loss — let’s put ‘meaningful’ in quotes — could be established and maintained,” Dr. Nathan said. “To me that means we did a good experiment. We had a fair test of this hypothesis.”

So: MeMe Roth-level caloric restriction for ten years just to maintain a 5% weight loss and take fewer medications than the control group? Not exactly a ringing endorsement of the power of dietary changes to change either appearance or health outcomes.

Oh, and of course in the comments of the Paul Campos article people are saying that the reason the intervention had no effect was because the 5% loss wasn’t significant enough. This was a pretty intensive intervention. If this intervention wasn’t enough, even though the experimental group had reduced their caloric intake to at least 377 calories less than the average American woman eats–and also less than she ate in 1971 (1542 calories per day), before the “obesity epidemic”–I just don’t think that a more extreme intervention is going to be feasible. It’s an even more dramatic reduction compared to the average American man–a reduction of more than 1000 calories per day. And somehow I doubt that these commenters were thinking about possible endocrine effects of exercise on set point.

*If you see an effect that the study was not designed to measure, its significance–the confidence with which we can say the result was a real effect and not due to chance–is different than if you planned to study it. That is because you’re not just looking at the few differences between groups your study was designed for, you’re looking at a bunch of things and cherry-picking out the ones where there’s actually a difference that you’re interested in.  The more different things you look at, the more likely that some of them will show a difference between the groups just by chance.

If the researchers decided after their study was over, to say, “Well, what about doctor’s visits?” “Nope, they’re the same.” “What about amputations?” “Same.” “What about vision problems?” “Same.” “What about blood sugar control?” “Same.” “What about number of medications?” “Oh, hey, those are different!” Well, in that case, we should be a lot less confident that we’re seeing a real effect caused by the diet and exercise intervention. (I don’t KNOW that this is what happened, but it is a possibility.)

**I really don’t know how long it takes MeMe Roth to run four miles. In another post, she says that she gets 1600-1800 calories and 1 hour of exercise a day. And she mentions not liking to run with food in her stomach in the next sentence, so it sounds like she mostly means running. If her typical run is four miles (which she’s indicated elsewhere) and it takes her an hour, then she’s a very slow runner. Maybe four miles is more of a minimum? Another source (see “Elle feature” near the bottom) indicates that she exercises “at least 45 minutes every single day, rain or shine, usually by running four miles and doing some strength work.” This indicates she runs faster than 12-minute miles. I’m starting to feel like a stalker, so I think I’ll just go with less-than-perfect accuracy so that this doesn’t get even weirder.

***and that this endocrine stuff is affected by exercise–and/or that only some people can maintain radical weight loss, even with these demanding measures. Not everyone is affected by exercise in the same way.

(If weight gain is as much or more a symptom than a cause of diabetes, this may explain why there was only a 5% weight loss despite such a low-calorie diet.)

This entry was posted in Uncategorized and tagged , , , , , . Bookmark the permalink.

7 Responses to The underwhelming effect of 5% weight loss on diabetes

  1. Meowser says:

    Oy, the comments on the Campos article. A lot of “they couldn’t possibly have lost that little weight with that degree of calorie restriction — LAWS OF THERMODYNAMICS!” and “they must have been lying about what they ate” and “they already had diabetes, so it was too late to save them, they should have lost the weight earlier.” People are just determined to hate the shit out of us.

  2. Hey, Meowser is here!
    Yes, if data contradicts their theories, and those theories are necessary to hate fat people, then clearly the data’s wrong.

  3. Michelle says:

    I’m really looking forward to taking a look at the full text of this when it is published.

  4. Pingback: Fat Agnostics and the New Consensus | closetpuritan

  5. Pingback: The Look AHEAD study: a rare long-term weight loss study | closetpuritan

  6. Pingback: Chocolate and Disillusionment | closetpuritan

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s