Effects of Dietary Composition on Energy Expenditure During Weight-Loss MaintenanceSome of the articles and studies I've discussed here have considered "adaptive thermogenesis," which can be described as the metabolic penalty we suffer when we lose weight. While reducing the size of one's body will reduce the amount of energy we need to operate and maintain that body, the act of dieting itself reduces our calorie burn (both at rest and total daily) more than would be predicted simply by the reduction in body mass. One recent study that I linked last month claimed that dietary composition didn't affect the degree of adaptive thermogenesis, and further concluded that the entire phenomenon didn't persist for two years.
This latest study has, on the first issue, reached a very different conclusion. Because it's both a unique finding and was better controlled than most nutritional studies, it's an interesting read. A lot of very knowledgeable people are still analyzing the methods and results to determine what firm conclusions should and shouldn't be drawn from the study; while that will probably be going on for a long while, it's worth mentioning here while the smarter folks debate the intricacies.
In a nutshell, the study took a group of 21 obese subjects (32 started, only 21 finished) of mixed gender and race, whose BMI averaged 34.4. For four weeks, their maintenance calorie level was determined by trial and error - tweaking the amount of energy they were fed until they were (supposedly) weight-stable. They were then fed a diet with 60% of those calories for 12 weeks (25% Pro, 30% Fat, 45% CHO). All of the subjects lost weight, averaging 13.6% total weight lost.
The subjects then went through another 4 weeks of weight-stability testing, determining their new maintenance calorie level through more trial and error. At that point, the subjects were given three different isocaloric diets in random order: Low fat (20P, 60C, 20F), low glycemic index (20P, 40C, 40F), and low carb (30P, 10C, 60F). Although it wasn't a true metabolic ward study, it was fairly tightly controlled, with energy expenditure being measured indirectly but in what can be called state of the art methods (doubly-labeled water).
At the end of the study, the authors concluded that differences in the diets resulted in different levels of both resting metabolic rate and total daily energy expenditure (resting metabolism + activity). More specifically, they concluded that on the low-fat diet, RMR decreased by 205 and total daily calorie burn decreased by 423. For the low-GI diet, RMR decreased by 166 and TDEE by 297. For the low-carb diet, RMR decreased by 138 and TDEE by 97. Through the use of an accelerometer, the researchers claimed that activity level was constant among the three diets. The study also claimed to find that the low-carb diet caused an increase in levels of cortisol (a stress hormone that has been implicated in weight gain) and C-reactive protein, a marker of chronic inflammation.
So what does this mean, practically speaking? I'm not sure. Some are claiming that we have finally established a metabolic advantage for low-carb diets - at the same level of activity and calories, the subjects seemed to somehow burn more calories throughout the day on a very low-carb diet). Others are claiming the exact opposite - because calories were controlled and constant during the crossover test stage, and because the subjects' weights didn't change, the supposed metabolic advantage is illusory or doesn't matter - the low-fat group didn't add weight back, and the low-carb group didn't lose more weight despite consuming the same number of calories but supposedly burning roughly 300 more per day.
Some of the limitations of an otherwise well-controlled study:
* Small sample size of 21
* Calorie intake was not controlled sufficiently to rule out cheating or other misreporting (I don't think this is very likely though)
* The amount of calories the subjects received during the crossover portion of the study was several hundred less than their supposed TDEE, as measured indirectly. Because the subjects stayed weight stable, this suggests that the indirect measurements (from which the conclusions were drawn) were inaccurate.
* Short duration of the test diets - just 4 weeks.
* The mismatch in protein between the low-carb diet and the low-fat/low-GI. While the authors claim this wouldn't account for the differences in RMR because that was tested in a fasting state (to eliminate the thermogenic effect of food), that may not be true, based on some of the reported RQ values (go
here for more detail on that, if you want it).
* The low-fat diet probably had insufficient levels of dietary fat, which can cause hormonal perturbations that might account for some of the observed differences. In this case, it's hard to untangle the effects from "too little fat" with the other factors (such as, arguably, "too much carbohydrate").
I could go on about this a lot longer, but a) I don't have more time right now and b) I don't want to make eyes glaze over any more than I have. Suffice to say, though, I don't think it can said definitively that this study does or does not establish any sort of metabolic advantage for a particular diet - it's hard to get around the fact that even if you accept the rest of the methodology and conclusions, the subjects who supposedly experienced this metabolic advantage didn't realize any changes in bodyweight, despite supposedly having the same activity level and same number of calories in. On the other hand, the entire study was designed and controlled much, much more rigorously than most, and the
possible flaws and limitations aren't necessarily actual flaws.
My Take on the Bottom LineBut let's accept the conclusions of the study entirely. Should everyone immediately switch to Atkins or a low-GI diet? In my opinion, no. Even accepting the study's conclusions, there was tremendous individual variability in the responses to the different diets. Some subjects supposedly had the least reduction in RMR/TDEE while on the low-fat diet. Others with the low-GI diet. Even if the author's conclusions are correct, we're still left with the following:
* All of the subjects lost weight through calorie reduction, and while consuming >200g CHO/day.
* All of the subjects stayed weight-stable while going through the 3 different test diets.
* The low-fat group (despite probably being below the sufficiency level for dietary fat) stayed weight stable while consuming >400g CHO/day.
* If there is a metabolic advantage for certain diets, a proposition that isn't at all impossible or even that far-fetched, there seems to be tremendous variation in the individual response.
On a side note, if anyone is using the cortisol/CRP levels as a justification for avoiding low-CHO diets, the best analysis I've seen so far suggests that those fears were overblown, and the increased levels in the low-CHO group were relatively minor.
Despite all of that, this is a very interesting contribution to the knowledge base. For those looking to maintain weight loss and not count calories, I think this study reinforces some existing conventional wisdom - experiment. See what works best for you. Don't drop below sufficiency levels on fats, protein, or micronutrients. While calories matter, the macronutrient composition of a diet may affect your long-term success at weight maintenance. Trial and error!