Tuesday, July 28, 2009
More on Exercise without Diet doesn't produce Weight Loss and the Ethics of Research
Follow @mcphoo
Tweet
Despite the promising title, Eight weeks of resistance training can significantly alter body composition in children who are overweight or obese, the results don't offer what one might expect from it.
Most of us would say "body comp" means weight, girth measures, bmi changes. But no. It doesn't. In this case, it seems to mean fat kids who worked out got stronger, added lean mass, but didn't lose weight or overall fat. This is consistent with other studies of working out without diet to go with it:
First off, the main goal of the study was to explore resistance training rather than aerobic intervals training for obese kids. As far as i can tell, they are simply hypothesizing that some kids who are obese may prefer lifting to running type activities, and if that's the case, let's see what that does. That's not much of an hypothesis to test in a research environment is it?
Not surprisingly, since other research has shown this too, obese kids in an 8 week program make super strength and power gains too. Just like non-obese kids. And like adults of all ages who are new to lifting. Neural adaptations are taking place, and new tissue is being laid down.
Now the authors claim that their findings are great. They say
It's really great to see BMC going up too - that's something to keep for life.
But where does this take us? IF absolute fat doesn't go down, weight goes up, how does bf% go down? There's more new lean mass. as opposed to more (or less) fat. That's kinda fudging, isn't it?
These results (gaining lean mass; not losing weight overall) are consistent with both non-obese kids and adults. It tells us that muscle building mechanisms for the first 8 weeks of a program have an impact. That's good. What about the next 8 weeks and then the next 48 weeks?
When we work with adults who are overweight, we know that after 8-12 weeks, if their girth, weight, and fat - nothing on these measures seems to change - they are not feeling a whole lot of love and success or seeing it in their mirrors.
We know that the study reports here that eating habits didn't change throughout the study. They weren't logged too religiously, though, and unless an observer is making those logs, we know from other work that we ALL misreport food logs.
We know that folks may feel zippier from working out - and that's fabulous - but we also know with obese adults that without nutrition, all the jumping and pumping in the world will not shed the excess weight which is having the biggest negative impact on overall health.
What's the Point? In fairness, one might say, this study was *just* looking at effects of resistance training over 8weeks on obese kids. Is that good enough? These are real kids with real problems. Is this fair to them or the best we can do?
Consider this: the study doesn't explicitly state an hypothesis, eg: we postulate that fat kids who do resistance will have the same benefits as non-fat kids who start resistance.
Hypothesis Testing
Is that poor science not to have an explicit hypothesis? Generally speaking, in most fields, yes.
Because you have to defend why you hypothesize your position and show value of the study: why on earth would you think you'd need to see if fat kids respond to resistance training differently than skinny kids? When that's said outloud, kinda makes one go "hmm" no?
Without that rationale for the study being clear, what's the point of the work? It's rather gratuitous. The authors as said only suggest that some fat kids might not like aerobics or intervals so they need alternatives. Right! so the next point would be again to say, we have lots of results to show the benefits of resistance training for kids. What's the special thing you think you need to test in this population not covered by these other studies?
Well, these kids are obese.
Ok, so what? are you asserting that because of that, the effects of a resistance program may be different? if so why? what's the basis for that assertion and how will you test it? Are there special fat kid risk factors to test that one might think fat kids shouldn't do resistance? No? So what's the point?
The authors just show what we already know from a zillion ways past sunday: resistance training builds lean muscle. And even if absolute fat doesn't go down, because lean muscle goes up, the bf% ratio changes. And as we see in the charts, kids did gain weight - from the lean muscle.
That's why these kind of studies seem gratuitous to me. And heh, not every paper an academic writes is earth shattering. But something leaves me edgy here. Obesity is a real problem. This study is dealing with clinically overweight and obese kids (over 23-43% body fat in the study).
So we've confirmed that yet one more population benefits from resistance training. Was there any doubt, however, that that would be the case? Any hypothesis to test? No? then what's the study contributed, really? For 8 weeks kids got no nutritional counseling when the authors KNOW that obesity programs combine nutrition and exercise. "It is clear that, along with nutrition and lifestyle, exercise plays a significant role in overcoming obesity in children."
But if the authors had provided that counseling, that would have screwed up their results: they wouldn't know what was down to resistance vs what was down to diet.
Now ethically, we can say the children weren't harmed; in fact they are healthier than when they started. And still obese. And if they stick with their current training and their current activity they'll still be obese a year from now.
What's actually been proven here? hypothesis testing two. In the realm of statistics, one can simply set an alpha or confidence level - a percentage - by which if the results fall within that percentage, the results happened because of the intervention, not by chance. The way the authors set up this study, their signficance values don't claim that the body comp change is significant, but that the reason for the change is down to the intervention. That's right. Just that what change occurred is not because of chance but because of the progam. In other words they have an above 95% certainty that that body comp change is because of the training. Shocking.
Ok, that's not shocking but it means the title is: Eight weeks of resistance training can significantly alter body composition in children who are overweight or obese. All that can be claimed, surely, is that we know within the realm of probability, literally, that resistance training has an effect at changing body comp. To say a "significant" effect - again means kinda weasel words. In stats, significance just means there's a treatment effect. To normal human beings significant means "wow that was a big deal."
So the authors are being technically accurate, but less clear that perhaps saying "Resistance training does induce body comp changes in fat kids in 8 week protocol, really really"
So, bottom line, did we learn anything new from this study that we didn't know before? Are the results surprising in any way? Was a bold hypothesis demonstrated? Did researchers who know that nutrition and lifestyle along with exercise is a big part of dealing with obesity provide that information to their participants' families as part of the study or just say good bye to the participants at the end of the 8-weeks?
Tweet Follow @begin2dig
Most of us would say "body comp" means weight, girth measures, bmi changes. But no. It doesn't. In this case, it seems to mean fat kids who worked out got stronger, added lean mass, but didn't lose weight or overall fat. This is consistent with other studies of working out without diet to go with it:
McGuigan, MR, Tatasciore, M, Newton, RU, and Pettigrew, S. Eight weeks of resistance training can significantly alter body composition in children who are overweight or obese. J Strength Cond Res 23(1): 80-85, 2009-Update: as R.M. Koske rightly points out in the comments below, body comp is technically changes in fat/muscle/bone ratios. And it's not entirely fair to conflate a scientific definition of a concept used in a research journal with popular understanding. But i DO take issue with the term SIGNIFICANT body comp alteration. more on that below.
The purpose of this study was to investigate the effect of an 8-week resistance training program on children who were overweight or obese. Forty-eight children (n = 26 girls and 22 boys; mean age = 9.7 years) participated in an 8-week undulating periodized resistance training program for 3 d[middle dot]wk-1. Measures of body composition via dual-energy X-ray absorptiometry, anthropometry, strength, and power were made before and after the training intervention. There was a significant decrease in absolute percent body fat of 2.6% (p = 0.003) and a significant increase in lean body mass of 5.3% (p = 0.07). There were no significant changes in height, weight, body mass index, total fat mass, or bone mineral content. There were significant increases in 1-repetition maximum squat (74%), number of push-ups (85%), countermovement jump height (8%), static jump height (4%), and power (16%). These results demonstrate that the resistance training program implemented produces significant changes in body composition and strength and power measures, as well as being well tolerated by the participants. An undulating periodized program provides variation and significantly increases lean body mass, decreases percent body fat, and increases strength and power in children who are overweight and obese.
First off, the main goal of the study was to explore resistance training rather than aerobic intervals training for obese kids. As far as i can tell, they are simply hypothesizing that some kids who are obese may prefer lifting to running type activities, and if that's the case, let's see what that does. That's not much of an hypothesis to test in a research environment is it?
Not surprisingly, since other research has shown this too, obese kids in an 8 week program make super strength and power gains too. Just like non-obese kids. And like adults of all ages who are new to lifting. Neural adaptations are taking place, and new tissue is being laid down.
Now the authors claim that their findings are great. They say
An undulating periodized program provides variation and results in significant increases in lean body mass, decreased percent body fat, and increased strength and power.
It's really great to see BMC going up too - that's something to keep for life.
But where does this take us? IF absolute fat doesn't go down, weight goes up, how does bf% go down? There's more new lean mass. as opposed to more (or less) fat. That's kinda fudging, isn't it?
These results (gaining lean mass; not losing weight overall) are consistent with both non-obese kids and adults. It tells us that muscle building mechanisms for the first 8 weeks of a program have an impact. That's good. What about the next 8 weeks and then the next 48 weeks?
When we work with adults who are overweight, we know that after 8-12 weeks, if their girth, weight, and fat - nothing on these measures seems to change - they are not feeling a whole lot of love and success or seeing it in their mirrors.
We know that the study reports here that eating habits didn't change throughout the study. They weren't logged too religiously, though, and unless an observer is making those logs, we know from other work that we ALL misreport food logs.
We know that folks may feel zippier from working out - and that's fabulous - but we also know with obese adults that without nutrition, all the jumping and pumping in the world will not shed the excess weight which is having the biggest negative impact on overall health.
What's the Point? In fairness, one might say, this study was *just* looking at effects of resistance training over 8weeks on obese kids. Is that good enough? These are real kids with real problems. Is this fair to them or the best we can do?
Consider this: the study doesn't explicitly state an hypothesis, eg: we postulate that fat kids who do resistance will have the same benefits as non-fat kids who start resistance.
Hypothesis Testing
Is that poor science not to have an explicit hypothesis? Generally speaking, in most fields, yes.
Because you have to defend why you hypothesize your position and show value of the study: why on earth would you think you'd need to see if fat kids respond to resistance training differently than skinny kids? When that's said outloud, kinda makes one go "hmm" no?
Without that rationale for the study being clear, what's the point of the work? It's rather gratuitous. The authors as said only suggest that some fat kids might not like aerobics or intervals so they need alternatives. Right! so the next point would be again to say, we have lots of results to show the benefits of resistance training for kids. What's the special thing you think you need to test in this population not covered by these other studies?
Well, these kids are obese.
Ok, so what? are you asserting that because of that, the effects of a resistance program may be different? if so why? what's the basis for that assertion and how will you test it? Are there special fat kid risk factors to test that one might think fat kids shouldn't do resistance? No? So what's the point?
The authors just show what we already know from a zillion ways past sunday: resistance training builds lean muscle. And even if absolute fat doesn't go down, because lean muscle goes up, the bf% ratio changes. And as we see in the charts, kids did gain weight - from the lean muscle.
That's why these kind of studies seem gratuitous to me. And heh, not every paper an academic writes is earth shattering. But something leaves me edgy here. Obesity is a real problem. This study is dealing with clinically overweight and obese kids (over 23-43% body fat in the study).
So we've confirmed that yet one more population benefits from resistance training. Was there any doubt, however, that that would be the case? Any hypothesis to test? No? then what's the study contributed, really? For 8 weeks kids got no nutritional counseling when the authors KNOW that obesity programs combine nutrition and exercise. "It is clear that, along with nutrition and lifestyle, exercise plays a significant role in overcoming obesity in children."
But if the authors had provided that counseling, that would have screwed up their results: they wouldn't know what was down to resistance vs what was down to diet.
Now ethically, we can say the children weren't harmed; in fact they are healthier than when they started. And still obese. And if they stick with their current training and their current activity they'll still be obese a year from now.
What's actually been proven here? hypothesis testing two. In the realm of statistics, one can simply set an alpha or confidence level - a percentage - by which if the results fall within that percentage, the results happened because of the intervention, not by chance. The way the authors set up this study, their signficance values don't claim that the body comp change is significant, but that the reason for the change is down to the intervention. That's right. Just that what change occurred is not because of chance but because of the progam. In other words they have an above 95% certainty that that body comp change is because of the training. Shocking.
Ok, that's not shocking but it means the title is: Eight weeks of resistance training can significantly alter body composition in children who are overweight or obese. All that can be claimed, surely, is that we know within the realm of probability, literally, that resistance training has an effect at changing body comp. To say a "significant" effect - again means kinda weasel words. In stats, significance just means there's a treatment effect. To normal human beings significant means "wow that was a big deal."
So the authors are being technically accurate, but less clear that perhaps saying "Resistance training does induce body comp changes in fat kids in 8 week protocol, really really"
So, bottom line, did we learn anything new from this study that we didn't know before? Are the results surprising in any way? Was a bold hypothesis demonstrated? Did researchers who know that nutrition and lifestyle along with exercise is a big part of dealing with obesity provide that information to their participants' families as part of the study or just say good bye to the participants at the end of the 8-weeks?
Tweet Follow @begin2dig
PDFIf you enjoy these posts by all means you should
Subscribe to:
Post Comments (Atom)
COACHING with dr. m.c. 
2 comments:
Hm. I would not in the slightest say that I believe that body composition means means "weight, girth measures, bmi changes." In fact, I'd say the opposite. Body composition is percentage of fat mass to lean mass, something that neither BMI nor weight measure well at all. (That's part of why they're such poor measures of health.) Body composition changes may sometimes include weight, girth, and BMI changes, but not always.
I don't disagree at all with your assessment of the results and point of the study - it does seem pretty useless. But I really think that the title of the study is an accurate representation of the results and not at all misleading. I could be atypical in my reaction, though.
i hear what you're saying. they are accurate that they can claim body comp changes if bf% goes down.
but is that what a human being thinks when we talk about wanting body comp changes and go see a PT to get a program for that?
no we think of these other markers.
so yes i was conflating the technical with the popular. fair call.
thanks for dropping by
Post a Comment