RE: The effects of consuming a high protein diet (4.4g/kg/d) on body composition in resistance-trained individuals
The newest study making the rounds (1) comes from the International Society for Sports Nutrition. The conclusion of the study states that "Consuming 5.5 times the RDA of protein has no effect on body composition in resistance-trained individuals who otherwise maintain the same training regimen. This is the first interventional study to demonstrate that consuming a hypercaloric high protein diet does not result in an increase in body fat". The language in the discussion only gets stronger: "The results of the current investigation do not support the notion that consuming protein in excess of purported needs results in a gain in fat mass. Certainly, this dispels the notion that ‘a calorie is just a calorie" . Anytime someone starts throwing around the 'a calorie isn't a calorie' mantra, I get a bit skeptical. Even if you think that manipulating macronutrient composition can lead to different effects on weight gain, this is usually explainable by changes in REE/TEE (2). To really prove calories in calories out isn't true for protein, you'd need well-designed experiments (adequate size, power, duration) with doubly labeled water to determine free-living energy expenditure, have food provided, with excess calories determined to be coming only from protein for each individual, collect feces to measure undigested calories excreted and have a pretty strong understanding of the genetics of all of this (discussed before - see here). Summary: it's close to impossible to do at this point in time (maybe an overfeeding study with identical twins in a metabolic ward?, though the doubly labeled water shortage would pose a problem).
This study, while an interesting endeavor, is farrr from definitive, and has some severe limitations that should prevent it from making loaded claims about calories in calories out.
Limitations:
1. Sample Size/Stats - forty mixed-sex, healthy resistance-trained individuals were recruited to partake in the study over 8 weeks. 10 dropped out, however, with 3 claiming to be unable to consume the amount of protein. This left us with a total of 10 controls, and 20 HP dieters and quite a small sample size, and very little power. The authors used a 2 way ANOVA to analyze their data, but a 2 way ANOVA has some assumptions in order to be appropriately used; 1 of these assumptions is that the 2 groups must have the sample sample size . Others include that there must be equal variances in the groups and that the data must be normally distributed. Houston, we've got some severe limitations to this data as a whole, and our ability to extrapolate anything from ANOVAs. To really make something of the data, you'd need to look at every participant individually.
2. Food Tracking - To be fair, it's great that the individuals took daily food logs. However, MyFitnessPal is far from scholarly and is often not accurate. If you look in Table 4, for the Dietary Intake, the macronutrients and kcals don't add up. Take the Pre and Post Changes in the Control group per day (top values). The Change in kcals for the Control group was 243kcals/day. The change in CHO/Pro/Fat were, respectively: 9, 11, 2 (this is on average, huge variation). Those macronutrient changes only account for 98kcals. 98kcals ≠ 243kcals.
3. The Controls weren't very good Controls:
-Control Group Exercised More - The Volume Load per day, found in table 3, training volume, increased for the control group. The high protein group increased slightly, about 1500. The control group increased by 4700.
-Control Group Dieted - Back to Table 4 again, the post kcal values, if we can trust them, are lower in the control group after the intervention. And not just a little bit lower, 243kcals lower on average. Taking into account the fact that the standard deviation beforehand is 639kcals, and 532 kcals after, some individuals may have been dieting a lot.
4. NEAT, TEF - the study tries to take a stab at the 'calories in, calories out' but didn't calculate Non-Exercise Activity Thermogenesis, which has been shown to increase resistance to fat-gain. The authors would also need to calculate the TEF of food from the high protein group, seeing as protein intake is largely associated with an increase in TEF. The study also doesn't report any REE/TEE values for the individuals/groups, though I believe this can be calculated from the BodPod data (correct me if I'm wrong)
5. Fecal Nitrogen Excretion and Other Biochem - I've seen Dr. Antonio mention that they're doing a followup to collect more data (Darwin bless whoever chooses to go through consuming this much protein per day). I'd love to see the authors calculate fecal nitrogen excretion- I'm doubtful that this large of a protein load would be entirely absorbed. I'll also be curious to see creatinine, albumin, CRP, etc etc.
6. Carbs make you fat - The authors try to outline in their paper that excess carbs will cause an increase in fat mass but protein won't. If you look at the (crappy) food intake data, the HP group increased their carbs by about 30g on average. Where'd those 120ish calories go? The carbohydrate argument seems to be alluding to the insulin-makes-you-fat hypothesis; however, individuals got their protein from whey protein shakes, which has quite a size-able effect on insulin, comparable to white bread (5,6). If calories from carbs make you fat because of insulin, shouldn't whey? I'm still just curious as to what's happening to all of those extra carbon skeletons from extra amino acids...
Something's funky about the conclusions of this study, and it's probably more than just the flatulence of people consuming 4.4g/kg/d of protein. The authors point out some of these things I've mentioned in their discussion, but still went so far as to say comparing group means with unequal variances casts doubt on calories in calories out. I would never look at this study and say, look, protein calories don't count! I would certainly never tell any future clients that protein calories don't count. From persons living in controlled settings, we know that calories count, higher protein diets (15-25%) preserve lean body mass, and that relatively higher protein diets (still within the AMDR ranges) increase satiety and reduce calorie intake (7,8,9). Even if you want to argue that low-carb diets have different effects on weight loss, adherence is absolute shit and they don't maintain weight loss as long as Mediterranean dieters (10,11).
1.http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4022420
2. http://jama.jamanetwork.com/article.aspx?articleid=1199154#COMMENT
3.http://www.ncbi.nlm.nih.gov/pubmed/9880251
4.http://www.ncbi.nlm.nih.gov/pubmed/15466943
5. http://ajcn.nutrition.org/content/82/1/69.full
6.http://www.nutritionandmetabolism.com/content/9/1/48
7.http://jama.jamanetwork.com/article.aspx?articleid=1199154
8. http://www.ncbi.nlm.nih.gov/pubmed/22215165
9. http://www.ncbi.nlm.nih.gov/pubmed/24675714
10. http://www.nejm.org/doi/full/10.1056/NEJMoa0708681
11. http://www.nejm.org/doi/full/10.1056/NEJMc1204792
This study, while an interesting endeavor, is farrr from definitive, and has some severe limitations that should prevent it from making loaded claims about calories in calories out.
Limitations:
1. Sample Size/Stats - forty mixed-sex, healthy resistance-trained individuals were recruited to partake in the study over 8 weeks. 10 dropped out, however, with 3 claiming to be unable to consume the amount of protein. This left us with a total of 10 controls, and 20 HP dieters and quite a small sample size, and very little power. The authors used a 2 way ANOVA to analyze their data, but a 2 way ANOVA has some assumptions in order to be appropriately used; 1 of these assumptions is that the 2 groups must have the sample sample size . Others include that there must be equal variances in the groups and that the data must be normally distributed. Houston, we've got some severe limitations to this data as a whole, and our ability to extrapolate anything from ANOVAs. To really make something of the data, you'd need to look at every participant individually.
2. Food Tracking - To be fair, it's great that the individuals took daily food logs. However, MyFitnessPal is far from scholarly and is often not accurate. If you look in Table 4, for the Dietary Intake, the macronutrients and kcals don't add up. Take the Pre and Post Changes in the Control group per day (top values). The Change in kcals for the Control group was 243kcals/day. The change in CHO/Pro/Fat were, respectively: 9, 11, 2 (this is on average, huge variation). Those macronutrient changes only account for 98kcals. 98kcals ≠ 243kcals.
3. The Controls weren't very good Controls:
-Control Group Exercised More - The Volume Load per day, found in table 3, training volume, increased for the control group. The high protein group increased slightly, about 1500. The control group increased by 4700.
-Control Group Dieted - Back to Table 4 again, the post kcal values, if we can trust them, are lower in the control group after the intervention. And not just a little bit lower, 243kcals lower on average. Taking into account the fact that the standard deviation beforehand is 639kcals, and 532 kcals after, some individuals may have been dieting a lot.
4. NEAT, TEF - the study tries to take a stab at the 'calories in, calories out' but didn't calculate Non-Exercise Activity Thermogenesis, which has been shown to increase resistance to fat-gain. The authors would also need to calculate the TEF of food from the high protein group, seeing as protein intake is largely associated with an increase in TEF. The study also doesn't report any REE/TEE values for the individuals/groups, though I believe this can be calculated from the BodPod data (correct me if I'm wrong)
5. Fecal Nitrogen Excretion and Other Biochem - I've seen Dr. Antonio mention that they're doing a followup to collect more data (Darwin bless whoever chooses to go through consuming this much protein per day). I'd love to see the authors calculate fecal nitrogen excretion- I'm doubtful that this large of a protein load would be entirely absorbed. I'll also be curious to see creatinine, albumin, CRP, etc etc.
6. Carbs make you fat - The authors try to outline in their paper that excess carbs will cause an increase in fat mass but protein won't. If you look at the (crappy) food intake data, the HP group increased their carbs by about 30g on average. Where'd those 120ish calories go? The carbohydrate argument seems to be alluding to the insulin-makes-you-fat hypothesis; however, individuals got their protein from whey protein shakes, which has quite a size-able effect on insulin, comparable to white bread (5,6). If calories from carbs make you fat because of insulin, shouldn't whey? I'm still just curious as to what's happening to all of those extra carbon skeletons from extra amino acids...
Something's funky about the conclusions of this study, and it's probably more than just the flatulence of people consuming 4.4g/kg/d of protein. The authors point out some of these things I've mentioned in their discussion, but still went so far as to say comparing group means with unequal variances casts doubt on calories in calories out. I would never look at this study and say, look, protein calories don't count! I would certainly never tell any future clients that protein calories don't count. From persons living in controlled settings, we know that calories count, higher protein diets (15-25%) preserve lean body mass, and that relatively higher protein diets (still within the AMDR ranges) increase satiety and reduce calorie intake (7,8,9). Even if you want to argue that low-carb diets have different effects on weight loss, adherence is absolute shit and they don't maintain weight loss as long as Mediterranean dieters (10,11).
1.http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4022420
2. http://jama.jamanetwork.com/article.aspx?articleid=1199154#COMMENT
3.http://www.ncbi.nlm.nih.gov/pubmed/9880251
4.http://www.ncbi.nlm.nih.gov/pubmed/15466943
5. http://ajcn.nutrition.org/content/82/1/69.full
6.http://www.nutritionandmetabolism.com/content/9/1/48
7.http://jama.jamanetwork.com/article.aspx?articleid=1199154
8. http://www.ncbi.nlm.nih.gov/pubmed/22215165
9. http://www.ncbi.nlm.nih.gov/pubmed/24675714
10. http://www.nejm.org/doi/full/10.1056/NEJMoa0708681
11. http://www.nejm.org/doi/full/10.1056/NEJMc1204792
Comments
Post a Comment