Log in

Post a comment - Jackdaws love my big sphinx of quartz [entries|archive|friends|userinfo]

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

Stuff [Jul. 6th, 2012|10:47 pm]
[Tags|, ]

I'm used to hearing misleading medical facts from my friends, from the newspapers, and every so often from leading medical journals. But I figured if there was one place I could rest assured would stay reasonable and objective, it would be cracked.com. I trusted cracked.com. I especially trusted David Wong, who's been my favorite contributor there ever since I read JDatE.

Unfortunately, Wong's latest column, Fat Is Officially Incurable, According To Science doesn't do him much credit. It takes some more-or-less correct-ish numbers, presents them in a misleading way, and then blows them totally out of proportion. It's almost as if it were some kind of comedy website, or something.

Wong starts like this: "Let's get this straight: The number of people who go from fat to thin, and stay there, statistically rounds down to zero. Every study says so. No study says otherwise. None."

He then proceeds to totally not cite any study that comes even close to saying that.

The studies he cites tend to support his next point: "You can lose and keep off some minor amount, 10 or 15 pounds, for the rest of your life -- it's hard, but it can be done. Rarer cases may keep off a little more."

The Secret Doctor Database Of Medical Information You Can't Access Because You're Not A Doctor - which my medical school has so far forgotten to remove my student access to - broadly supports his point. A lot of the diets studied - which range from Jenny Craig to Weight Watchers to less commercial medical programs - find that people who comply with them well can lose on average 5% to 10% of their weight (a study on Jenny Craig found 7.9%, one on Weight Watchers found 5.3%, one medical diet found 9.3%). In a moderately overweight person - let's say 200 lb - that corresponds to between between 10 and 20 lb.

Do people maintain this amount? The best that can be said is that sometimes they maintain some of it. Two studies that measured the effects of making people go to monthly counseling meetings after the diet, probably a much better case than you'd get in real life, found between 47% and 81% of people maintaining weight loss of at least 5% after 1.5 years.

On five year followup, results were a bit worse. People who lost significant amounts of weight during their diet on average only kept between 20% and 30% of the weight loss after five years. This isn't awful - one study on the high end of the range found some people who originally lost 17 kg on average who'd kept off 7 kg after four years - but it's nothing to be proud of either. The only really bright point is that people who are on diets based on exercise in addition to or instead of food restriction tend to do much better and pretty much keep the weight off as long as they keep exercising; they'd maintained fully half of their weight loss on average after 5 years.

So overall, Wong's point that "You can lose and keep off some minor amount, 10 or 15 pounds, for the rest of your life -- it's hard, but it can be done. Rarer cases may keep off a little more" seems broadly correct. Then he goes on to say: "And when I say "no one," I mean those cases are so obscenely rare that they don't even appear on the chart. They can't even find enough such people to include in the studies. It's like trying to study people who have survived falling out of planes."

Here's what actual science has to say on the matter of finding the people to include in studies:
In an effort to learn more about those individuals who have been successful at long-term weight loss, [Rena R.] Wing and [James O.] Hill established the National Weight Control Registry in 1994. This registry is a self-selected population of more than 4000 individuals who are age 18 or older and have lost at least 13.6 kg (30 lb) and kept it off at least 1 y. Registry members are recruited primarily through newspaper and magazine articles. Participants in the registry report having lost an average of 33 kg and have maintained the minimum weight loss (13.6 kg) for an average of 5.7 y.

So, put an ad in the newspapers asking for people who have lost lots of weight, and you can collect four thousand of them, which is more than enough to do studies if you really want. Needless to say, this fact didn't make it into Wong's article.

Wong then goes on to say: "Well, this person did the math, and as far as they could tell, two out of 1,000 Weight Watchers customers actually maintain large weight losses permanently. Two out of a thousand. That means if you are fat, you are 25 times more likely to survive getting shot in the head than to stop being fat."

Time to play the Trace The Original Article Game! The original article here turns out to be a study by Dr. Michael Lowe on Weight Watchers customers, Weight-loss maintenance in overweight individuals one to five years following successful completion of a commercial weight loss program. It concludes things like "At 5 y, 19.4% were within 5 lb of goal weight, 42.6% maintained a loss of 5% or more, 18.8% maintained a loss of 10% or more, and 70.3% were below initial weight." Figures that do not appear in the study include numbers such as "two" or "out of one thousand".

So where does the 2/1000 number come from? Well, according to the extremely-axe-having-to-grind blog that "did the math" for Wong, 3.9% of people were at or below goal weight after five years (note that "3.9%" is also one of the many numbers that does not appear in this article; I have no idea where the blog got it, but let's assume I missed it and it's correct.) Then the blogger notes that this study only addressed the top 6% of Weight Watchers users, so if we count everyone who didn't participate in the study as being in the study but failing to lose weight (!) that brings us down to 2/1000.

Note several flaws with this plan. First of all, I have no idea how Weight Watchers sets "goal weight", and I would be surprised if the blogger or David Wong does either. Second, they totally just looked in a business article to see how many people were in Weight Watchers, and counted everyone not in this study as failing! Third of all, I still don't know where that 3.9% number came from! Fourth of all, even if by some miracle all of that resulted in something resembling a statistic, it has no relation to David Wong's claim that sure, some people can keep off a little bit of weight, but no one can keep off a lot of weight. This Weight Watchers study wasn't even of people who were really overweight! The average participant had a BMI of between 25-30, ie not even technically "obese". Most people with a BMI 25-30 really could get back within normal weight by losing 15 pounds or so!

I'm not accusing the blogger at...uh, "Fat Fu" of doing anything unethical or sketchy. She admits right there in the post where she derives the 2/1000 number: "Now what I’ve just done. Is this any relationship to science? Picked out the most damning numbers in a slightly less successful study to paint a dismal portrait of Weight Watchers? Of course it’s not science. But why should I feel obligated to do that when weight loss researchers – including this one – aren’t engaged in anything remotely related to science either." That seems like a reasonable disclaimer, and goodness knows I've done the same sort of recreational statistics on occasion.

But Wong doesn't mention the disclaimer. He just cites the figure as support for his hypothesis, when it's an admittedly made up number made up in a context completely unrelated to his hypothesis. If we want to know whether random people who aren't too overweight on a Weight Watchers type program can lose weight and keep it off, we already have the answer - those unspectacular 10-15 pound losses we already mentioned from all the other studies. This is just addressing that same question, except replacing the whole "study" thing with made up numbers.

The rest of the article I have fewer qualms with. Yes, people who lose a lot of weight have some underappreciated metabolic difficulties in keeping it off. Yes, by far the most effective method for major weight loss is bariatric surgery, and yes, bariatric surgery has horrible side effects and shouldn't be anyone's first choice (or tenth choice) option.

So in summary, Wong is right that most diets result in relatively modest 10-15 pound weight losses that are hard but not impossible to maintain. He's wrong that no one ever loses more and doctors can't even find enough people who lose more to study them. He may or may not be wrong that it's very rare, but all of the "evidence" he cited to that effect turns out to be totally made up and to not provide the slightest shred of support for his position. And he's right that there are metabolic reasons why long term weight loss is really hard, and that bariatric surgery is a bad idea.

Actually, it's still like two or three times better than most people trying to address weight loss. But even though it's a comedy site and people taking medical advice from a comedy site deserve whatever they get (most likely cancer, I'd imagine), I still think it's just really really unethical to give unsupported medical information and dress it up with made-up numbers.

(this blog entry should not be taken to imply anything about society, overweight people, diets, carbohydrates, "paleo", acceptable BMIs, whether or not BMI is a stupid statistic which it is, the medical establishment, or your value as a human being. It's just asking people to not use made up medical facts. Please don't kill me.)

post comment:

No HTML allowed in subject


Notice! This user has turned on the option that logs your IP address when posting.