The Feeding Doctor on Pinterest

the “evidence” you might see that child and adolescent weight management doesn’t increase disordered eating or EDs…

Posted by on Feb 14, 2011 in Blog Posts | 26 comments

Everything has to be “evidence-based” these days. While in theory that is a good thing, when the “evidence” itself is poor (which it is more often than I am comfortable with*) then I would argue we are worse off. If a question is “answered,” and published in a reputable journal,  many will take it for face value.  But did it really answer a question, provide clarity to difficult issues, or did it just muddy the waters? When a study is used to champion a cause, or cited by others who like the conclusion with little discussion of the merits of the study, it is lazy and possibly harmful.

The following article might be a case in point.  “The Influence of a Behavioral Weight Management Program on Disordered Eating Attitudes and Behaviors in Children with Overweight.” ( Follansbee-Junger, JADA  2010 1653-9) The conclusion of the study says, “These findings do not provide evidence that behavioral interventions lead to an increase in unhealthy eating attitudes and behaviors.” (It does go on to say further study is warranted as most article do these days.)

I’m throwing this out there, because this is a contentious issue. An issues that needs good science. If first we must do no harm, then this is an important question. Do current standard practices in weight management for children and teens cause an increase in disordered eating and eating disorders? (This is a separate question from does the current standard practice work, which many, including myself,  argue it does not, but I’m too tight for time to pull a bunch of references…) I argue that this study has brought us no closer to knowing the answer, but that many will take the conclusion and run with it…

I worry now that every study on children and adolescents and weight, every public health initiative  forged to fight the “obesity epidemic” will cite this study as “evidence” that weight loss programs for kids do not cause a rise in disordered eating or eating disorders. It will be cited as proof that refutes and argues concerns of harm, particularly from the eating disorder community that decries the rise in eating disorders, and in younger and younger children.

Here is my brief analysis of this flawed study, and why their wily conclusion, of “do not provide evidence of harm” should not be conflated with “doesn’t cause harm.”

1) this was a NON-RANDOM sample, that means families were recruited for a weight loss behavioral program for children and teens (I would have liked the authors to have a bit more of a discussion of their study limitations as many studies do…)
2) there were 68 initial families, 50 by the end of the study
3) the control group was “tainted”  in that it was not a true control. It was made up of the self-selected folks on the waiting list who were presumably worried about their children and their weight
4) they used what they called a “validated and reliable” child feeding questionnaire, while the author of the questionnaire herself said, “Further work with the instrument is needed to establish its reliability and validity.” (Child Feeding Questionnaire, Birch and Fisher, Appetite 2001)
5) they did not attempt to either validate (are you measuring what you say you will measure) or confirm reliability (consistency of the tool) though that is commonly done. What is odd is that they did test for internal consistency (a validity measure) for one of their tools, the Schwartz Peer Victimization Scale, so we can presume that they could have tested their Child Feeding Questionnaire tool…

They did find that “across all conditions, higher levels of body dissatisfaction, peer victimization, parent restrictive feeding practices, and concern for child weight at baseline predicted higher levels of disordered eating attitudes at follow-up,” which is consistent with past findings on the effects of worry over weight and restriction. Oh, and peer victimization, or bullying is NOT benign and is pervasive. (And they did check for validity of the peer victimization tool, so we could argue this is a more convincing finding…)

The “worry over weight”  points to why the control group might be “tainted,” since it is a group of parents and youngsters self-selected and already concerned, with no direction on a waiting list, which might lead to higher concern and restriction…

So, alas, in my opinion, the study is of such poor design that any results are not generalizable or relevant. I just fear that in the current era of “evidence” that folks who want to say that weight loss interventions don’t cause harm will use this as their “proof.”

What do you think? Do you work in the ED community? Keep an eye out for this paper… Remember the name Follansbee, it might just start popping up in those bibliographies…

*One of the worst things I encountered while practicing medicine in the office setting was not having the time to really dig into the data and studies. So many medications were “black-boxed”(taken off the market or used only with dire safety warnings.) Meds that only a few years earlier I was told were standard of care based on the “evidence.” From Tequin, Avandia, universal hormone replacement therapy, Bextra… I remember having hesitancy with Avandia, and not wanting to use it so quickly, and the company I worked for had a quality control group that said if I didn’t use them with diabetics, my “report card” would show I was not in compliance with “best practice” and I would have consequences… I guess I’m just a little skeptical of the “evidence” these days. Are you?

Share and Enjoy:
  • Print
  • Digg
  • Reddit
  • StumbleUpon
  • Tumblr
  • del.icio.us
  • Facebook
  • LinkedIn
  • Twitter

26 Comments

Join the conversation and post a comment.

  1. unscrambled

    So, we’re in the middle of an OMGBESITY epidemic, and the study that’s going to be generalized to every kid (I agree with you, it’ll get cited all over the place) is 50 families? Pretty sweet.

    Katja, I’m with you as a (budding) physician researcher. What passes for “evidence” is slim indeed, particularly if we’re dealing with things that Everyone Knows Are True.

    Keep it up, you and a few others give this student some hope that not everyone in my future profession is a schmuck. :)

    • katja

      of course, I’ve sort of transitioned out of your future profession in a strict sense :)
      Hang in there, I know many amazing, dedicated people in the field of medicine. You will be an important voice out there!

  2. Holly

    A question that occurred to me in reading about this study (on your blog only; I haven’t looked at the original report)relates to its specific focus on children “with overweight.” Is there evidence, elsewhere, which shows that a consistent clinical and/or inter-family focus on weight loss, calorie restriction, or body shape is associated with an increase in disordered eating/ED behavior in children of “normal” weight? I admit to not being so familiar with the literature that I could cite any particular study offhand, but my gut tells me that there is probably reasonable evidence that such a focus is contraindicated and has a known potential to cause some degree of psychic harm.

    I wonder, then, about the chain of logic which leads researchers such as these to focus their efforts specifically on children of above-average weight. To partially answer my own question, I realize that one reason is that children of “normal” weight are probably far less likely to be referred to a medical weight loss program, and thus harder to sample. That said, it seems like a more useful study design would look at data from a longitudinal study (easier said than done, I realize) and assess correlations between weight change over time, doctor/family attitudes, and ED symptoms.

    Otherwise, and assuming my initial claim (that we know a focus on weight is potentially harmful for “normal-weight” children) is correct, there doesn’t seem to be a good evidentiary basis for the a priori assumption that “overweight” children are likely to respond differently (i.e. more positively) to a negative focus on weight than their thinner counterparts. In fact, it seems to presume that there is something fundamentally different about the mental makeup and responses of fat and thin children.

    I apologize if my logical chain is missing a few links; it’s been a very long day.

    • Elizabeth

      I doubt that you could get approval to do this kind of study on “normal” weight children, because there would be no potential clinical upside for the kids in doing “behavior intervention” on them.

  3. karen

    I think evidence like statistics can be molded to say what you want to say. I’ve known (and possibly been) people who develop ED as a result of dieting. They haven’t all been teens. I’ve also seen middle aged women who after a little success at weight loss become addicted and can’t stop dieting or obsessing about food and exercise and that “last 5 pounds”… even after they have lost another 20! I know it’s not scientific, but I think there is a greater link between dieting and ED than smoking and cancer. Maybe they just haven’t found a good way to measure it yet.

    • katja

      thank you. This is a common clinical observation, that dieting begins, for some spins into a full-blown ED, for others becomes a way of life, cycling on and off diets for years, restricting, binging etc. The diet is common in BED, at least initially… As they say, genetics loads the gun, society (diet, etc) pulls the trigger. I know there are many experts who can cite the data better than I can on the links between dieting and EDs… I also know there are people who really, really want to push diets, “lifestyle modification” (if it’s eating try to get you to eat fewer calories or to control your weight, it’s a diet) following dietary guidelines, etc onto people and want to claim that there is no harm…

  4. The WellRounded Mama

    Evidence-based medicine is all well and good, but we have to remember…..garbage in, garbage out. If the study isn’t a good one, then conclusions from it or including it in a meta-analysis distorts things.

    I won’t reiterate what you said, but I had very similar concerns to yours.

    • katja

      i agree. Meta-analyses are only as good as the studies they exclude, or include. (I’ll do another breakdown soon of a recent NJEM review on BMI and badness where the studies they excluded were as interesting as what they included…)
      I’ve read some about the shenanigans of the drug companies in terms of research, what gets published, what doesn’t, who writes “studies” (badscience had a bit on this) and I just have trouble trusting their “evidence.” It makes me happy I’m not wrestling with decisions about statins, and all the other meds on a daily basis:)

  5. Datura

    I’m sorry, but since when is “overweight” something you have? Children with overweight? What language are they speaking? Have they somehow developed alternative rules of grammar?

    • katja

      this is more tricky that it would seem. I actually feel their pain on this one. (Though they could have said weight acceleration, but that is not the standard definition of “overweight.”) The terms are really tricky. I don’t use “overweight” or “obese” in terms of the BMI cutoffs (which is how most people in public health/medicine use them.) There are increasing rates of children who are at the higher end of the Bell Curve in terms of weight, so how to describe that is tough. For the individual child, I talk about “unstable” weight, or weight acceleration, but these terms truly are tricky. What do you think we should say? I struggle with this all the time. How to describe what is going on, without relying on meaningless terms, or using terms that people “think” they know all about…

  6. Twistie

    Like Kelly, I find myself really wondering about those 18 dropouts from the study. What happened to them? Why did they leave the study?

    It also does bother me, like you, that the entire sample was taken from a pool of people who were already staunchly pro-weight intervention and that they worked hard to validate one study tool, but not the other, especially when the one they didn’t bother to validate is the one whose author feels it needs more validation before being relied on too heavily.

    In short, while I don’t have training in the creation and proper conduct of medical studies, I can see holes in the organization and methodology that I could drive a Buick through.

    And this is a good reminder to be just as rigorous in looking at the evidence I like, as well as that which disagrees with my personal bent.

    • katja

      thank you for the reminder! I belong to a journal club, and we tend to find many, many faults with the articles we read, though I do think we are rigirous about all studies. I tend to like any study more that has a frank discussion about it’s own design flaws, or puts it’s findings into perspective. I think it is natural to want to find “proof” that agree with our own viewpoints. I was willing to read this study, and do think the question asked is important. That’s why it’s so frustrating when the quality or design is so poor. I happen to believe weight loss programs will increase disordered eating, but a great study that said otherwise might have me looking further.
      I have to say, a handful of years ago, as I discovered and came to this whole body of literature, I was convinced that the standard approach was the way to go. I had a lot of dissonance around the whole “fat and fit” (HAES, I hadn’t even heard the word) approach when I started reading this stuff. I had to read alot to convince myself. I always tell people, I can’t convince you, but at least be open to where you feel uncomfortable, to what challenges your beliefs, to the part where you keep coming back and saying, “but!…” Read there, learn there and try to be open and see what happens…

  7. KellyK

    Isn’t a drop-out rate of 18 people (26%) also a red flag? It seems high to me.

    • katja

      I think it’s pretty high. I’d have to look how many were from the control group vs intervention. Of note, most weight loss studies have extremely HIGH drop-out rates. 20,30,40% (just like any diet) and often in the analysis those numbers are “controlled” for, or they do statistical analyses to see where they would fall if they stayed in the study. I usually don’t see much discussion about how much the intervention must suck, which is why people drop out of weight loss studies/diets. Also interesting, HAES (Health at Every Size) studies tend to have a MUCH lower dropout rate.

      • KellyK

        usually don’t see much discussion about how much the intervention must suck, which is why people drop out of weight loss studies/diets.

        That’s something I think merits more attention too. If people are dropping out because the intervention is unsustainable in their lives (or, heck, because their kid is starting to develop worrisome behaviors related to eating), that puts a whole different spin on the results.

        • Emily

          It’s actually interesting to look at the meta-analyses that have been done on diets (all of which show that they work for something like 2% of the population, while at least 30% gain more weight than when they started), because every exclusion/drop-out/etc. criterion makes the diets more likely to succeed.
          It would be one thing if, say, we saw that clinical trials showed that dieting works for weight loss, but wider distribution was not able to replicate that. That sort of thing is very common – you get a super-motivated, “clean” bunch of people, and the treatment works better. The fact that dieting doesn’t even work under the most ideal circumstances really makes you wonder why people keep recommending it.

          • katja

            I totally agree. A few years ago I was in a meeting with a large insurance company about HAES and feeding info, their wellness/quality guy said they just finished a study on VLCL diets (Very low calorie liquid) and shock- found they didn’t work! I find it cruel and unusual, boardering, if not malpractice to recommend VLCL diets which are proven over and over again to be harmful. Yes, indeed, why does the medical establishment still recommend what doesn’t work? My flip answer is money, my second answer is many just don’t know any better and truly do care for the health of the patient but are uninformed…

        • katja

          they didn’t expound on the dropouts… Would be interesting…

          • Emily

            Well, as a medical researcher myself, I think it’s more complicated than money. (Wanted to make the same response to the Campos article as well.) Most researchers have transferable skills and if they really wanted to make money would be working for pharma or medical consulting or something. That said, their jobs usually do require grant funding, and they have to chase what’s fundable. If the NIH wants to fund obesity research, that’s what people are going to study.

            Most of the people doing this kind of research do have good intentions, I think. (There’s just not enough glory, power, or money in epidemiology to do it for less good reasons.) They genuinely believe that obesity is a major health problem, and that they should tackle it. But they’re as vulnerable to societal prejudices as anyone else, and it’s easier to tackle a problem you find ugly. I’d also add that BMI is nice and easy to measure, and it’s easy to find, say, a relationship between BMI and heart disease using readily available data – it’s much harder to measure fitness and nutrition in detail, do the analysis, and realize that most of that association goes away once you’ve taken them into account.

          • katja

            I guess when I say money, I mean systems-wide, rather than accusing the researchers themselves. It seems the grant money is there for certain hot topics. How do you propose we see more research on eating competence or feeding dynamics. The same old research on weight behavior management stuff keeps getting funded (who are the people on the committees determining what gets funded, what companies do they consult for or are invested in? I remember some gossip about the NEJM Berrington de Gonzalez, BMI and mortality study that folks involved also had connections with lap-band surgery companies and this article came out and immediately was followed by an application for FDA approval for the device starting at BMI of 30, vs 35 or 40) I guess I am just suspicious, but I do imagine that most folks doing this research really do want to help. What I wonder if how can we keep on the blinders? In practice, I KNEW standard of care didn’t work, but I always blamed the patient somehow… How do we cast a critical eye on “accepted fact” or dogma about weight, BMI, health etc…

          • katja

            oh, and thanks for writing in, I’ll try to do these kind of break-downs of research about once a month, and could use a critical eye :)

          • Emily

            I think the question for most of us is how do we get ANYTHING funded, let alone something innovative! That said, I do think a lot of the research that’s nationally funded (as opposed to internally or company-funded – much more sketchy) has moved towards environmental interventions – improving access to healthy food and physical activity. I just worry that the outcome is always BMI – why can’t we worry about blood pressure and cholesterol and let weight be what it is?

            People just don’t look at this stuff with a critical eye. I attended a lecture lately where they were presenting a weight loss/physical activity intervention in teachers. The guy presenting said, in the course of the presentation, “Calories in, calories out.” and “The teachers pick up their kids at the end of the day, go through the drive-in at McDonald’s, and then go home and eat supper.” Well, if they were really eating an extra McDonald’s meal a day and there’s nothing more to it than calorie balance, they would gain 5 lbs a month. But because it fits his mental image of fat people stuffing themselves, he didn’t even see the contradiction in what he was saying.

            A lot of researchers who go into this area are the sort who really don’t understand how someone could be fat – surely everyone would be skinny and running marathons like they themselves do, if they were less lazy/uneducated/misguided. Maybe we need affirmative action for obese researchers!

  8. Regina T

    As a super fat person, and a mother of two girls, I find I have to bite my tongue to keep from questioning my picky 10 yr old’s food choices. She has ADD, thus making her more senitive to things than others may be, and this has always made her a finicky eater. I use to blame myself before the ADD diagnosis—because maybe I didn’t offer enough choices (though I did), or didn’t try hard enough (though I did) to ensure that she made all the right food choices. She is a protein eater and veggie/fruit avoider, but fresh fruit and veggies are always offered. Silently, I worry that she will always make the less nutritious food choices, but in my heart I know that isn’t true. I am, though, bound and determined to keep the same negative and fat fearmongering that I endured growing up out of my house so that I don’t pass that self hate onto my girls. My youngest is in the 98th percentile in height and weight and her pediatrician is the absolute BEST when it comes to downplaying that chart. She is so protective of her patient’s psyche as equally as their health. I’m thankful to have such an ally. My kid gets bombarded enough with tv programming and even books that highlight bodies that are different in a negative way….she just won’t ever hear it from her parents, nor her doctor.

  9. Nicole

    Thanks for this. I can say (only anecdotally, of course) that the interventions my parents tried with me, on the advice of pediatricians, set me up for lifelong feeding and weight issues. Whether they crossed into the realm of eating disorders, I can’t say. I do know that not one–seeing a dietitian, taking part in special fat-girl exercise schemes, Weight Watchers–ever did a thing to make me less fat. In fact, the more I focused on weight, the less involved I became in martial arts, swimming, biking–all the physical activities I did regularly and loved as a child and young teen. I poured all my energy into “aerobics” so that I could “burn calories”. I would have been a lot better off staying with the activities I loved.

    • katja

      Thank YOU for your comment. Perfect. It’s part of what makes me crazy about so many of these interventions, games around “how many calories” in a can of soda, or how long you have to ride a bike to work off a slice of pizza. All rot. All makes us doubt eating in a tuned in way…

  10. Ines Anchondo

    Good post, Katja. I often say experience-evidence based practice is based. I think this is part of what Malcolm Gladwell was getting at in his book Blink.