Calories in vs. calories out, caloric quality etc.

[BeauJangles]

Originally posted by: brikis98

Originally posted by: Kipper
Now that we're on the same page (*phew*), I feel obligated to point out that there are two sides to this coin. Our modern sources of meat (beef, pork, poultry, lamb) are ALSO the product of hundreds of generations of selective breeding for desirable characteristics - fat and muscle makeup. They, too, bear very little resemblance to their ancestors, and their 'modern' diets, in turn, can hardly be said to resemble that of their wild-ancestors - because they are vegetarian and eat...plants (which as you've mentioned, have changed). This all creates a somewhat problematic historical/anthropological case for applying the Paleolithic diet in a modern supermarket. That, and the scientifically questionable idea of "carbohydrate avoidance" combine to make a bit of a shaky argument.

The idea of a "diet built into your genes" is a bit sexy, but I think it really ends there. There is such variability in human diets (ranging from the mostly protein-fat diets of the Inuits) to a more diverse Mediterranean palate, to the heavily plant-based Buddhist diets that the Paleolithic notion doesn't really seem to fit. Many of these populations show impressive lifespans coupled with a virtual absence of "Western" (read: American) diseases, so I think we can pretty much all agree that whatever the problem is, it probably lies with the "meat with a side of meat and meat for dessert" American lifestyle.

I agree completely, which is why I personally don't follow the paleo diet. There are many logical flaws with the reasoning behind it and it's incredibly tricky to accomplish in the modern world. Having said, that, just about any diet that gets us away from the Western Diet and gets people to stop eating processed crap is probably a good thing. And even if the modern economy means you can't be 100% truly paleo, even going 70% of the way there is bound to be a huge improvement.

My own dietary practices sound similar to yours, in that I try to eat a diet of whole foods similar to what Michael Pollan describes in "In Defense of Food". I avoid processed crap, including highly processed grains, cereals, pastas, candy, bread, etc. However, if people were eating it a few hundred years ago - whether it's whole grains, milk, legumes, etc - it's fine by me. I wrote a bit about it in my diet experiment thread, although I haven't updated that in a while.

:thumbsup: I'm all about taking a logical approach to dieting. I eliminated soda from my diet a long time ago. I've really limited my intake of true crap, which is usually now a brownie or two once or twice a week. Other than that, I try to strike a balance between healthy and enjoyable. I find myself eating a lot of salads, chicken, and eggs. My girlfriend is vegetarian, so she's introduced me to some good non-meat-based meals.

My budget is pretty tight right now so, while I do get solid portions of fruits and veggies, I can't afford to be as healthy as I'd like. Given the choice, I'd be buying organic meat these days, but it's just too pricey. I'd probably also increase my fruits and veggies intake and really try and cut out the remaining pasta and other stuff I eat now.

*shrug* a buddy of mine at our gym went strict paleo a few months ago. His results (in terms of performance) have been mixed. While his metcon times have improved, his lifts have either stalled out or started to progress much more slowly. Now, of course, that doesn't mean paleo had anything to do with it, but he thinks it might be linked.

 

[brikis98]

Originally posted by: interchange
actual research study

This study is horrendous. There was a good critique of it on performancemenu, and to put it simply, the conclusions had nothing to do with the data they got, and the news agencies that reported it perverted the outcomes even further. For example, check out CNN's article about the study. A few things to note:

* The study very conspicuously omitted some of the most popular diets and macronutrient breakdowns, namely low carb and ketogenic. The lowest percentage of carbs in any of the 4 groups was 35%, which is not "low carb" (just as 35% fat wouldn't be "low fat") and certainly not ketogenic. Seems a bit odd with all the fuss over low carb and Atkins diets, doesn't it? Moreover, despite that, did you notice how CNN's article (and probably many others) has "low carb" in the title?

* The study introduced several key variables that utterly invalidate the conclusion: they asked ALL the participants to change the type of food they ate! In particular, they recommended ALL groups to "include 8% or less of saturated fat, at least 20 g of dietary fiber per day, and 150 mg or less of cholesterol per 1000 kcal. Carbohydrate-rich foods with a low glycemic index were recommended in each diet." Also, all groups were advised to do "90 minutes of moderate exercise per week." So, really, the weight loss observed from the study could've been from any of the factors above and had nothing to do with the calories at all. How did such a study and its conclusion ever get past a single peer review?

* And here's the kicker (straight from the performancemenu discussion): the subjects followed the suggested macronutrient breakdowns so poorly, that in reality, all 4 groups ate roughly the same thing:

The goals were (C, P, F):
65/15/25
55/25/20
45/15/40
35/25/40

The food actually eaten was:
53/20/27
51/21/28
48/20/33
43/21/35

So, protein was exactly the same, carbs only ranged from 43 to 53, and fat from 27 to 35. It's not surprising that the results are similar in the different groups.

 

[Kipper]

Originally posted by: brikis98

This study is horrendous. There was a good critique of it on performancemenu, and to put it simply, the conclusions had nothing to do with the data they got, and the news agencies that reported it perverted the outcomes even further. For example, check out CNN's article about the study. A few things to note:

* The study very conspicuously omitted some of the most popular diets and macronutrient breakdowns, namely low carb and ketogenic. The lowest percentage of carbs in any of the 4 groups was 35%, which is not "low carb" (just as 35% fat wouldn't be "low fat") and certainly not ketogenic. Seems a bit odd with all the fuss over low carb and Atkins diets, doesn't it? Moreover, despite that, did you notice how CNN's article (and probably many others) has "low carb" in the title?

I just *had* to post. Apologies in advance. My two cents:

I think NEJM deserves the benefit of the doubt here. It is one of the most prestigious biomedical journals out there - for what it's worth, they generally don't publish crap. The problem is that the naming convention chosen by the researchers was taken at face value by the media (one of the problems of having untrained reporters reading scientific research) and they ran with it. On a related note, I've found that the New York Times generally does a good job of summarizing studies.

The diets are in fact what they say. The Accepted Macronutrient Distribution Ranges (AMDR) for fat and carbohydrate are 25-35%, and 45-65%, respectively. So from this point of view, a diet at 20% would be "low fat" and a carbohydrate diet between 35/45% would be on the lower end, and vice versa. The researchers weren't out to test ketogenic or Atkins diets, so complaining about that is a bit problematic. Also, this is part of a very good study design, because it blinds the participant to the diet they are receiving and prevents behaviors that would confound results. They mention that in their conclusion. For instance, if you're in the study and you receive information which looks awfully like the Atkins diet without the name, you go out and buy a book on Atkins dieting, and wham, he study is compromised because you've figured out what group you are in. Also, the design allows comparison of two groups against each other with a different intake of the macronutrient in question. But food studies are really a messy science, especially when you're dealing with people self-reporting their own intake.

* The study introduced several key variables that utterly invalidate the conclusion: they asked ALL the participants to change the type of food they ate! In particular, they recommended ALL groups to "include 8% or less of saturated fat, at least 20 g of dietary fiber per day, and 150 mg or less of cholesterol per 1000 kcal. Carbohydrate-rich foods with a low glycemic index were recommended in each diet." Also, all groups were advised to do "90 minutes of moderate exercise per week." So, really, the weight loss observed from the study could've been from any of the factors above and had nothing to do with the calories at all. How did such a study and its conclusion ever get past a single peer review?

I don't really see the problem with these "healthy lifestyle/eating" guidelines in the study, or how they invalidate the conclusion (that weight loss can be achieved with any sort of P/C/F diet recommendation causing a deficit). Particularly because the groups are all being treated equally, and were given the same general recommendations, albeit with plans tailored to their particular diet group. Individual adherence aside, the other 'problem' is, ethics. You CAN ethically tell people to continue doing whatever they keep on doing if you are simply observing, but in an intervention study (where researchers group people and change something in peoples' lives), you cannot give people recommendations to eat unhealthy foods in varying quantities to "see what happens." The ethics review boards which govern studies involving human subjects won't allow it. A bad analogy, but it's as if you were trying to measure the effects of certain types of cigarettes on people, so you randomly assigned them to to groups, some which would smoke more, and some which would smoke less.

I'm a bit skeptical that 90 minutes/week exercise would produce the same sort of weight loss, especially because a large proportion of these participants were obese and obese people have trouble engaging in most strenuous sport. That said, exercise recommendations are usually standard practice in weight-loss studies, because people are probably GOING to exercise. Instead of leaving that variable open, the researchers close it as best they can by giving specific recommendations.

* And here's the kicker (straight from the performancemenu discussion): the subjects followed the suggested macronutrient breakdowns so poorly, that in reality, all 4 groups ate roughly the same thing:

The goals were (C, P, F):
65/15/25
55/25/20
45/15/40
35/25/40

The food actually eaten was:
53/20/27
51/21/28
48/20/33
43/21/35

So, protein was exactly the same, carbs only ranged from 43 to 53, and fat from 27 to 35. It's not surprising that the results are similar in the different groups.

True, the researchers didn't end up studying what they set out to study. Neither is it the prettiest study with the best data. But they did have some interesting findings. The thing I find most interesting is that although the participants were given recommendations that varied in macronutrient composition, they failed to hit their macronutrient targets (making it impossible for researchers to actually find the effects of any particular macronutrient, if any - "We did not confirm previous findings that low-carbohydrate or high-protein diets caused increased weight loss at 6 months...") and instead ALL ended up eating relatively the same thing. However, it didn't seem to make a difference in the end because all groups (the 80% or so who completed the study) managed to maintain caloric reductions and lose weight in spite of the varying recommendations. It would have been nice to get month-by-month data on adherence, but they don't have that - only measured twice, six months in and at the conclusion of the study (a weakness). My suspicion is that adherence lasts for around 2-3 months to the specific meal plan recommendations, and then people get bored of what they're eating and change it up a bit. They revert to habitual macronutrient proportions (as suggested in the study), but maybe make enough substitutions that they are able to sustain a caloric loss.

The take-home message? Do whatever works for you, what you can adhere to, whatever you can stick to in order to produce a caloric deficit. A bit "duh," but someone's got do "duh" studies, just in case what we assume is "duh" happens to be wrong...

 

[brikis98]

Originally posted by: Kipper
The diets are in fact what they say. The Accepted Macronutrient Distribution Ranges (AMDR) for fat and carbohydrate are 25-35%, and 45-65%, respectively. So from this point of view, a diet at 20% would be "low fat" and a carbohydrate diet between 35/45% would be on the lower end, and vice versa.

The problem is that it's incredibly deceptive to the public, especially the way it was reported. Simply having less carbs than in the AMDR does not make it a "low carb" diet, as CNN would have you believe. In fact, I don't know of ANY diet that would call itself "low carb" that has 45% carbs. It's absurd and the researchers should have known that.

Originally posted by: Kipper
The researchers weren't out to test ketogenic or Atkins diets, so complaining about that is a bit problematic. Also, this is part of a very good study design, because it blinds the participant to the diet they are receiving and prevents behaviors that would confound results. They mention that in their conclusion. For instance, if you're in the study and you receive information which looks awfully like the Atkins diet without the name, you go out and buy a book on Atkins dieting, and wham, he study is compromised because you've figured out what group you are in. Also, the design allows comparison of two groups against each other with a different intake of the macronutrient in question.

I completely understand this limitation, but lets be reasonable: if the researchers are going to make a bold conclusion that the ONLY thing that matters is calorie consumption and that macronutrient breakdown is unimportant, there is no reasonable way you can leave out ketogenic diets. You can't go around saying things like "the key really is that it's calories. It's not the content of fat or carbohydrates, it's just calories," if you left out one of the most controversial and popular macronutrient breakdowns.

Originally posted by: Kipper
But food studies are really a messy science, especially when you're dealing with people self-reporting their own intake.

They are indeed, which is all the more reason that these scientists (and the people reporting on them) need to be MUCH more careful about the conclusions they publish. People take this stuff seriously and when they see a study that says "only calories matter", they will think that they can be healthy while eating utter crap. There is an enormous difference between 2000 calories of junk food and 2000 calories of fruits, veggies and lean meats, but this study - despite enormous gaps in its reasoning and lots of unaccounted variables - makes it seem like either choice is equivalent.

Originally posted by: Kipper
I don't really see the problem with these "healthy lifestyle/eating" guidelines in the study, or how they invalidate the conclusion (that weight loss can be achieved with any sort of P/C/F diet recommendation causing a deficit).

Are you serious? All groups were told to eat less saturated fat, less cholesterol, more fiber, eat foods with a low glycemic index and exercise more. It is quite possible that even if the participants in the study didn't make any other changes at all in their lifestyles, that these factors alone could've accounted for the weight loss. In other words, they introduced a bunch of variables, all known to improve health measurements (including body weight), and without a control group, we CANNOT decide which of the following conclusions is true:

1. The weight loss was due to calorie restriction
2. The weight loss was due to eating more fiber
3. The weight loss was due to eating foods with a low glycemic index
4. The weight loss was due to exercising more
5. The weight loss was due to an interaction between calorie restriction & low glycemic index

There are a dozen other possibilities as well and yet, the study arbitrarily picks number 1. This is TERRIBLE science.

Originally posted by: Kipper
Individual adherence aside, the other 'problem' is, ethics. You CAN ethically tell people to continue doing whatever they keep on doing if you are simply observing, but in an intervention study (where researchers group people and change something in peoples' lives), you cannot give people recommendations to eat unhealthy foods in varying quantities to "see what happens." The ethics review boards which govern studies involving human subjects won't allow it.

I understand the issue of ethics, but I must AGAIN comment that if a study is going to make such a definitive conclusion as "calories are all that matters", you cannot have so many unaccounted for variables. I'm also not sure it's actually unethical to tell people to eat various macronutrient breakdowns with further instructions. If you can teach people how to measure their fat/carb/protein intake (such as sites like fitday), I don't think it's unethical to not mention things like fiber or saturated fat.

Originally posted by: Kipper
A bad analogy, but it's as if you were trying to measure the effects of certain types of cigarettes on people, so you randomly assigned them to to groups, some which would smoke more, and some which would smoke less.

Here's a slightly better analogy of what this study actually did: imagine we were doing an experiment to determine which of two new high tech sneakers, A or B, helps reduce running injury rates more. However, to be "ethical" and make sure the athletes didn't hurt themselves, we had both groups learn the POSE running method. At the end of the study, we found that both groups had reduced injury rates by 20%. Is it reasonable to conclude that both sneakers are equally effective at reducing injury rates? Of course not! The POSE technique - just like more fiber, more exercise, low glycemic foods in the original study - is known for producing the exact same kind of effects that we're studying. Maybe without POSE, sneaker A would've reduced injury rates 10%, but sneaker B only 5%. But perhaps POSE is so effective that it totally overshadowed our results. There is no way to tell and it's bad science to arbitrarily pick one. It's this kind of awful science that is incredibly prevalent in diet research that is a huge part of why we can't stop this obesity epidemic.

Originally posted by: Kipper
I'm a bit skeptical that 90 minutes/week exercise would produce the same sort of weight loss, especially because a large proportion of these participants were obese and obese people have trouble engaging in most strenuous sport.

Actually, in my experience, the obese and untrained are ALWAYS the ones that see the biggest benefit when they start to exercise. "Beginner gains" are often rapid and noticeable and you just cannot discount exercise as an unaccounted for variable.

Originally posted by: Kipper
That said, exercise recommendations are usually standard practice in weight-loss studies, because people are probably GOING to exercise. Instead of leaving that variable open, the researchers close it as best they can by giving specific recommendations.

Now see, here's where I'm skeptical: it's not too likely that these obese people would've started (and especially stuck with) a bunch of extra exercise unless specifically instructed to. By making it a requirement, the researchers virtually guarantee that exercise is an uncounted for variable. Otherwise, it would've been more likely that these obese people would've otherwise maintained their previous lifestyle which may or may not have included exercise, but certainly wasn't making them any thinner.

Originally posted by: Kipper
True, the researchers didn't end up studying what they set out to study. Neither is it the prettiest study with the best data. But they did have some interesting findings. The thing I find most interesting is that although the participants were given recommendations that varied in macronutrient composition, they failed to hit their macronutrient targets (making it impossible for researchers to actually find the effects of any particular macronutrient, if any - "We did not confirm previous findings that low-carbohydrate or high-protein diets caused increased weight loss at 6 months...") and instead ALL ended up eating relatively the same thing. However, it didn't seem to make a difference in the end because all groups (the 80% or so who completed the study) managed to maintain caloric reductions and lose weight in spite of the varying recommendations.

If the conclusion had said something like "it's hard to maintain a diet that emphasizes a particular macronutrient" or "most people regress to a 50/20/30 c/p/f breakdown, but as long as they have a calorie deficit they still lose weight" I would not complain. But instead, their conclusion was as follows:

"Reduced-calorie diets result in clinically meaningful weight loss regardless of which macronutrients they emphasize."

Even if we ignore the extra variables (fiber, GI, exercise), even this conclusion is EXTREMELY misleading. It can be interpreted in two ways:

(1) Different macronutrient breakdowns have no impact on weight loss as long as calories are the same. [CANNOT BE CONCLUDED FROM EVIDENCE]
(2) Weight can be lost through a calorie deficit on a 50/20/30 breakdown. Most people can't maintain diets with a particular macronutrient emphasized and regress to a ~50/20/30 breakdown. Under these conditions, any initial macronutrient breakdown choice is irrelevant and all that matters is a caloric deficit. [CAN BE CONCLUDED FROM EVIDENCE IF UNACCOUNTED FOR VARIABLES ARE IGNORED]

Unfortunately, just about everyone interprets it using option #1, especially the news agencies that report on this to the public. This is tragic.

Originally posted by: Kipper
The take-home message? Do whatever works for you, what you can adhere to, whatever you can stick to in order to produce a caloric deficit. A bit "duh," but someone's got do "duh" studies, just in case what we assume is "duh" happens to be wrong...

And this is exactly why we haven't been able to beat obesity. Most people CAN'T stick to a caloric deficit when eating the same foods they are used to. Perhaps eating a high carb diet increases hunger levels or screws up insulin levels. Perhaps a high fat diet is too calorie dense. Whatever it is, changing the type of food eaten is probably more important for most Americans than the quantity. It's not that quantity doesn't matter, but trying to change that first is likely to fail. On the other hand, changing the type of food you eaten very often leads to a spontaneous decrease in the quantity eaten as well.

 

[Kipper]

Originally posted by: brikis98

The problem is that it's incredibly deceptive to the public, especially the way it was reported. Simply having less carbs than in the AMDR does not make it a "low carb" diet, as CNN would have you believe. In fact, I don't know of ANY diet that would call itself "low carb" that has 45% carbs. It's absurd and the researchers should have known that.

The researchers are NOT writing for the general public. They are writing to a biomedical audience. NEJM is not on the magazine rack next to "USWeekly" and other trashy mags. To be sure, scientists need to consider the effects of publishing research, but you cannot fault these researchers for people who don't know how to read studies interpreting them incorrectly. That's ludicrous. The problem is CNN, not NEJM.

I completely understand this limitation, but lets be reasonable: if the researchers are going to make a bold conclusion that the ONLY thing that matters is calorie consumption and that macronutrient breakdown is unimportant, there is no reasonable way you can leave out ketogenic diets. You can't go around saying things like "the key really is that it's calories. It's not the content of fat or carbohydrates, it's just calories," if you left out one of the most controversial and popular macronutrient breakdowns.

Okay, problem. They didn't say that macronutrient breakdown was unimportant. They said that RECOMMENDATIONS for specific macronutrient intakes appeared to be meaningless. They started out trying to determine if specific macronutrients could be responsible for weight loss but when there was low adherence they had to draw the best conclusions that they could. The conclusion was that from the evidence, all other factors being equal, it had to be the calories that produced the weight loss. You can point out that they "left out" certain "popular" macronutrient ratios, but you're missing the point. Part of the purpose of this study was to address the shortcomings of previous studies, which included "small samples, underrepresentation of men, limited generalizability, a lack of blinded ascertainment of the outcome..." They had to blind the subjects and researchers, and it's fairly hard to do that if you are putting someone on a ketogenic diet. The non-inclusion of certain "popular/fad" diets is a consequence of the original study design. It weakens the study, to be sure, but it does not necessarily make the study "bad," because blinding is a strength and is tougher to pull off in diet studies, where you can't give people a placebo.

They are indeed, which is all the more reason that these scientists (and the people reporting on them) need to be MUCH more careful about the conclusions they publish. People take this stuff seriously and when they see a study that says "only calories matter", they will think that they can be healthy while eating utter crap. There is an enormous difference between 2000 calories of junk food and 2000 calories of fruits, veggies and lean meats, but this study - despite enormous gaps in its reasoning and lots of unaccounted variables - makes it seem like either choice is equivalent.

I'm agreeing with you substantively that the nutrient density of calories does matter, but back to the study. If your objection is that "people will generalize," I'm going to point again that the problem is with the people who interpret the research, not the researchers themselves, because their intended audience IS NOT THE GENERAL PUBLIC. Moreover, I don't quite see how you get the idea that their ultimate conclusion is that "only calories matter." Their findings suggest that behavioral factors are more important (although they caution against generalizing in this finding) than macronutrient metabolism, something I would tenatively agree with. There is also suggestion that people revert to old macronutrient ratios in the long term when participating in weight loss studies (note that this does not mean that they went back to old eating habits). The ultimate conclusion is that "diets that are successful in causing weight loss can emphasize a range of fat, protein, and carbohydrate compositions that have beneficial effects on risk factors for cardiovascular disease and diabetes."

Are you serious? All groups were told to eat less saturated fat, less cholesterol, more fiber, eat foods with a low glycemic index and exercise more. It is quite possible that even if the participants in the study didn't make any other changes at all in their lifestyles, that these factors alone could've accounted for the weight loss. In other words, they introduced a bunch of variables, all known to improve health measurements (including body weight), and without a control group, we CANNOT decide which of the following conclusions is true:

1. The weight loss was due to calorie restriction
2. The weight loss was due to eating more fiber
3. The weight loss was due to eating foods with a low glycemic index
4. The weight loss was due to exercising more
5. The weight loss was due to an interaction between calorie restriction & low glycemic index

There are a dozen other possibilities as well and yet, the study arbitrarily picks number 1. This is TERRIBLE science.

You can call it "terrible science," but the bottom line is that diet studies are not as "clean" as pharmaceutical studies or studies done on rats. You want a "clean" study? Do it on monkeys or rats, not on people. There are too many ethical and individual considerations that go into a human study to make it so clean (the "cleanest" you will get are pharmaceutical studies). Your complaint, as far as I see it, doesn't really have a point.

Yes, it is problematic to introduce "other" variables, but everything you've mentioned is a hallmark of a healther diet. You really cannot discuss macronutrients without discussing dietary fiber or saturated fat (glycemic index otherwise, but most "healthy" foods tend to be lower GI anyway). How, do I ask you, are you supposed to account for the effect of fiber intake in someone's diet? How exactly do you mitigate the ethical quandary of telling someone NOT to eat fibrous foods? You can't. You also have to assume, as a researcher, that the people that sign up for these studies are motivated (it was part of the selection criteria) ergo, they will be more attuned to nutritional literature. If you want to minimize external influence, you have to introduce these "new variables," as you suggest, but at least they are within your control, and you minimize the effects of external influence.

I understand the issue of ethics, but I must AGAIN comment that if a study is going to make such a definitive conclusion as "calories are all that matters", you cannot have so many unaccounted for variables. I'm also not sure it's actually unethical to tell people to eat various macronutrient breakdowns with further instructions. If you can teach people how to measure their fat/carb/protein intake (such as sites like fitday), I don't think it's unethical to not mention things like fiber or saturated fat.

I think it's a bit of a stretch to say that the conclusion of this study is "calories are all that matters." There is that implication, but I'm not seeing it written that explicitly (I may be wrong, feel free to correct me here). There is suggestion that a behavioral component was involved, which is obvious. Although IRBs will vary, many of them err on the side of caution and will balk at the slightest hint of unethical research behavior. In this case, it COULD be legitimately argued that withholding information on saturated fat diets would be unethical in the context of an interventional weight loss study, especially because researchers are measuring markers of cardiovascular disease. This is a research study and not a nutrition class, so there is a higher ethical standard. In a basic nutrition class, you may not want to teach about saturated fat or fiber beyond the cursory basics, but in a two-year study it would be pretty hard to keep saturated fat and fiber out of the discussion.

Here's a slightly better analogy of what this study actually did: imagine we were doing an experiment to determine which of two new high tech sneakers, A or B, helps reduce running injury rates more. However, to be "ethical" and make sure the athletes didn't hurt themselves, we had both groups learn the POSE running method. At the end of the study, we found that both groups had reduced injury rates by 20%. Is it reasonable to conclude that both sneakers are equally effective at reducing injury rates? Of course not! The POSE technique - just like more fiber, more exercise, low glycemic foods in the original study - is known for producing the exact same kind of effects that we're studying. Maybe without POSE, sneaker A would've reduced injury rates 10%, but sneaker B only 5%. But perhaps POSE is so effective that it totally overshadowed our results. There is no way to tell and it's bad science to arbitrarily pick one. It's this kind of awful science that is incredibly prevalent in diet research that is a huge part of why we can't stop this obesity epidemic.

My cigarette smoking analogy was intended to demonstrate the ethical problem with such a cigarette study, not as an analogy to the current study itself, but I do understand your objection that other variables may have been responsible (see above). Fair enough. But how much? It's unlikely that they all played a role together, but then the variables you mentioned above (low SCFA, fiber, GI, etc.) are all hallmarks of a healthy diet.

I personally think the problem with "stopping" the obesity epidemic is because it's so hard to change human behavior and less with diet research, which really doesn't tell us anything we already don't know - moderation and basic foods.

Actually, in my experience, the obese and untrained are ALWAYS the ones that see the biggest benefit when they start to exercise. "Beginner gains" are often rapid and noticeable and you just cannot discount exercise as an unaccounted for variable.

Do you mean with resistance exercise or with cardiovascular exercise? "Moderate" exercise can be as simple as a brisk walk, which many wouldn't consider exercise (myself included). I think it's fairly safe to say given the weight loss these people experienced that these people didn't turn into Olympic athletes over the course of two years. Although some exercise is usually better than none, and the benefits may be large, I'm not so sure that the role of exercise is as large you as you think it may be. As mentioned previously, the obese are VERY untrained. It's also magnified by the problem that they have a large amount of weight and it takes a lot of energy to move that weight around. They may also suffer from joint pain and other associated mobility problems. I don't see how ninety minutes of moderate exercise a WEEK (that's a bit more than 15 mins a DAY) would have made such a huge difference in the big picture.

Now see, here's where I'm skeptical: it's not too likely that these obese people would've started (and especially stuck with) a bunch of extra exercise unless specifically instructed to. By making it a requirement, the researchers virtually guarantee that exercise is an uncounted for variable. Otherwise, it would've been more likely that these obese people would've otherwise maintained their previous lifestyle which may or may not have included exercise, but certainly wasn't making them any thinner.

I don't follow. You are dealing with (presumably) motivated people. It is a reasonable assumption that motivated people will seek healthful behaviors (or they wouldn't have signed up for the study). Ergo, it is a reasonable assumption that they will try to exercise. KNOWING this, to minimize the effects you give them a goal, not a requirement of 90 min/week. True there was very little followup. True, this makes it a bit unaccounted for, but I still think you are overplaying the role that it would play in weight loss.

If the conclusion had said something like "it's hard to maintain a diet that emphasizes a particular macronutrient" or "most people regress to a 50/20/30 c/p/f breakdown, but as long as they have a calorie deficit they still lose weight" I would not complain. But instead, their conclusion was as follows:

"Reduced-calorie diets result in clinically meaningful weight loss regardless of which macronutrients they emphasize."

Even if we ignore the extra variables (fiber, GI, exercise), even this conclusion is EXTREMELY misleading. It can be interpreted in two ways:

(1) Different macronutrient breakdowns have no impact on weight loss as long as calories are the same. [CANNOT BE CONCLUDED FROM EVIDENCE]
(2) Weight can be lost through a calorie deficit on a 50/20/30 breakdown. Most people can't maintain diets with a particular macronutrient emphasized and regress to a ~50/20/30 breakdown. Under these conditions, any initial macronutrient breakdown choice is irrelevant and all that matters is a caloric deficit. [CAN BE CONCLUDED FROM EVIDENCE IF UNACCOUNTED FOR VARIABLES ARE IGNORED]

Unfortunately, just about everyone interprets it using option #1, especially the news agencies that report on this to the public. This is tragic.

The sentence you've quoted above is a FINDING, not a conclusion. The conclusion was that macronutrient proportions do not appear to matter when making dietary recommendations that will cause weight loss. Big difference. There is also suggestion that behavioral factors play a role, as their findings seem to reflect those of other studies (essentially that accountability and social support are vital tools in weight loss). Once again, you may be right that people are going to cherry-pick from the literature, but I don't see how that is the researcher's problem. Nor does it make the study bad. That would be like saying Hegel has bad philosophical arguments because they are hard to understand.

And this is exactly why we haven't been able to beat obesity. Most people CAN'T stick to a caloric deficit when eating the same foods they are used to. Perhaps eating a high carb diet increases hunger levels or screws up insulin levels. Perhaps a high fat diet is too calorie dense. Whatever it is, changing the type of food eaten is probably more important for most Americans than the quantity. It's not that quantity doesn't matter, but trying to change that first is likely to fail. On the other hand, changing the type of food you eaten very often leads to a spontaneous decrease in the quantity eaten as well.

I would agree with you, but the findings seem to point to the fact that people will revert to old macronutrient ratios (but not necessarily the same foods). Obviously, we don't have intake data to corroborate this assumption, but that has to be the fact because otherwise caloric intake would not have changed so much over the course of the study.

 

[brikis98]

Originally posted by: Kipper
The researchers are NOT writing for the general public. They are writing to a biomedical audience.

If you are doing a research of how varying macronutrient ratios impacts weight loss, then 45% carbs cannot be your "low carb" diet no matter who the intended audience is. For example, if you read the introductory paragraph to the study, they cite 8 studies that "showed that low-carbohydrate, high-protein diets resulted in more weight loss over the course of 3 to 6 months than conventional high-carbohydrate, low-fat diets". Well, if you look through these studies, the vast majority of them limit carbs to <30g per day, which will obviously be far less than 45%. Through out the rest of the study, they compare their results to previous low carb studies, which is just NOT appropriate.

Originally posted by: Kipper
Okay, problem. They didn't say that macronutrient breakdown was unimportant. They said that RECOMMENDATIONS for specific macronutrient intakes appeared to be meaningless.

Let me copy and paste a few lines from the study for you:

(1) "Reduced-calorie diets result in clinically meaningful weight loss regardless of which macronutrients they emphasize."
(2) "When nonnutritional influences are minimized, as they were in our study, the specific macronutrient content is of minor importance, as was suggested many years ago"
(3) "In conclusion, diets that are successful in causing weight loss can emphasize a range of fat, protein, and carbohydrate compositions that have beneficial effects on risk factors for cardiovascular disease and diabetes"

Do you see the word "recommendation" in these sentences? Do you see a mention of the fact that the subjects did not stick to the emphasized macronutrient content? I don't. And that's why every one of those statements is very ambiguous in meaning. They could be (and clearly have been) interpreted as implying that "macronutrients composition makes no difference to weight loss" when in reality, the only thing you could really conclude from this study is that "people can't stick to their assigned macronutrient composition, and when they end up eating the same exact stuff while in a caloric deficit, shockingly, they lose the same amount of weight." In fact, I'd say statement (2) above may suggest that the researchers felt that "calories are all that matters" was a foregone conclusion and have interpreted the results in a way to support this.

Originally posted by: Kipper
They started out trying to determine if specific macronutrients could be responsible for weight loss but when there was low adherence they had to draw the best conclusions that they could. The conclusion was that from the evidence, all other factors being equal, it had to be the calories that produced the weight loss.

So, what you're saying is that this "groundbreaking" study determined that, if everyone eats the same type of food, then reducing the quantity leads to weight loss? Wow, I'm glad millions of dollars were spent on this research. Good thing they didn't make any claims about the type of food being important.....

Originally posted by: Kipper
Moreover, I don't quite see how you get the idea that their ultimate conclusion is that "only calories matter."

I listed some quotes above that suggest that, but here's another one, directly from Dr. Frank Sacks, one of the authors: "The key really is that it's calories. It's not the content of fat or carbohydrates, it's just calories."

Originally posted by: Kipper
You can call it "terrible science," but the bottom line is that diet studies are not as "clean" as pharmaceutical studies or studies done on rats. You want a "clean" study? Do it on monkeys or rats, not on people. There are too many ethical and individual considerations that go into a human study to make it so clean (the "cleanest" you will get are pharmaceutical studies). Your complaint, as far as I see it, doesn't really have a point.

So... you admit that the study is flawed and then say that there's no point in me complaining about that? I believe that many of the variables they didn't control for could've been handled much better, but you'd prefer to just "shrug" and say diet studies aren't "clean"? It's the non-chalant acceptance of such crap science that has allowed diet research to become the mess that it is.

Originally posted by: Kipper
Yes, it is problematic to introduce "other" variables, but everything you've mentioned is a hallmark of a healther diet.

That depends on your definition of a "healther diet." For example, many diets (e.g. Atkins) don't concern themselves with dietary cholesterol and lots of research shows that it probably has no real effect on cardiovascular disease. The same thing goes for saturated fat. Now, perhaps these are "healthy" measures according to the AHA, but I'd say support for their guidelines is far from universal.

Originally posted by: Kipper
You really cannot discuss macronutrients without discussing dietary fiber or saturated fat (glycemic index otherwise, but most "healthy" foods tend to be lower GI anyway). How, do I ask you, are you supposed to account for the effect of fiber intake in someone's diet? How exactly do you mitigate the ethical quandary of telling someone NOT to eat fibrous foods? You can't.

Huh? I never said to recommend avoiding fibrous foods or suggest people eat more saturated fat. I'm saying that ALL you tell people is to eat a particular macronutrient breakdown and let them figure out how to get there. I'd say this is an incredibly important part of the diet. For example, eating low carbohydrate may get people to eat more fruits & veggies to satisfy the carb needs, as they are less carb dense than pasta, bread, cereal, etc. Or maybe it won't. Trying to eat less fat may push people away from red meat to lean meats. Or maybe it won't. But I believe that those kinds of decisions are a key aspect of how effective the diet is. Forcing everyone to eat the same type of food not only introduces a whole new variable, but may also distort the normal way people go about achieving one macronutrient breakdown or another. In fact, perhaps this is why the participants couldn't stick to their assigned macronutrient breakdowns, so the mere act of introducing these new dietary variables destroyed any chance of proper adherence to the recommended diets.

Originally posted by: Kipper
My cigarette smoking analogy was intended to demonstrate the ethical problem with such a cigarette study, not as an analogy to the current study itself, but I do understand your objection that other variables may have been responsible (see above). Fair enough. But how much?

I don't know how much and neither do the researchers. It could be a 5% difference or it could be 100%. It's the fact that we don't know that makes the study's conclusions questionable.

Originally posted by: Kipper
Do you mean with resistance exercise or with cardiovascular exercise? "Moderate" exercise can be as simple as a brisk walk, which many wouldn't consider exercise (myself included).

In my experience, the more untrained the person is, the faster their gains early on, regardless of the type of exercise or intensity. It's only in the later stages that the intensity needs to be continuously increased to keep producing results. For example, early on, you can make some pretty big gains in endurance/stamina by just doing a brisk walk a few times a week. As time goes on, you'll need to walk either faster (or jog), longer or more often to see results. Same thing for weight training: in the early stages, almost anything works. Go to the gym, do a little bench, some curls, some sit-ups and your muscle mass increases and fat decreases. After a few months, to keep making gains, you need a more structured program with proper choices for weight progression, reps, sets, etc.

Originally posted by: Kipper
I think it's fairly safe to say given the weight loss these people experienced that these people didn't turn into Olympic athletes over the course of two years. Although some exercise is usually better than none, and the benefits may be large, I'm not so sure that the role of exercise is as large you as you think it may be. As mentioned previously, the obese are VERY untrained. It's also magnified by the problem that they have a large amount of weight and it takes a lot of energy to move that weight around. They may also suffer from joint pain and other associated mobility problems. I don't see how ninety minutes of moderate exercise a WEEK (that's a bit more than 15 mins a DAY) would have made such a huge difference in the big picture.

Since the researchers were harping about calories being the key issue, 90 minutes of exercise a week can make a very substantial difference in the number of calories burned. This is especially true for the very obese, who would actually burn a lot of energy even doing "moderate" exercise. But once again, it's the fact that we don't know how much of a difference it made that casts doubt over the study's conclusions.

Originally posted by: Kipper
I don't follow. You are dealing with (presumably) motivated people. It is a reasonable assumption that motivated people will seek healthful behaviors (or they wouldn't have signed up for the study). Ergo, it is a reasonable assumption that they will try to exercise. KNOWING this, to minimize the effects you give them a goal, not a requirement of 90 min/week. True there was very little followup. True, this makes it a bit unaccounted for, but I still think you are overplaying the role that it would play in weight loss.

I'm just speculating here, but it is my opinion that far more of the participants ended up exercising because the study told them to than would have if exercise was not mentioned at all. Since exercise adds an unmeasured variable to this study, this yet again reduces the certainty of the study's conclusion.

Originally posted by: Kipper
The sentence you've quoted above is a FINDING, not a conclusion.

No it's not. It's listed in the "Conclusions" section of the abstract.

 

[Kipper]

Originally posted by: brikis98
If you are doing a research of how varying macronutrient ratios impacts weight loss, then 45% carbs cannot be your "low carb" diet no matter who the intended audience is. For example, if you read the introductory paragraph to the study, they cite 8 studies that "showed that low-carbohydrate, high-protein diets resulted in more weight loss over the course of 3 to 6 months than conventional high-carbohydrate, low-fat diets". Well, if you look through these studies, the vast majority of them limit carbs to <30g per day, which will obviously be far less than 45%. Through out the rest of the study, they compare their results to previous low carb studies, which is just NOT appropriate.

Alright, my last post regarding this study.

You seem to be harping on and on about the definition of low-carb when it really has no bearing on the findings. Whether they misapply the definition of low-carbohydrate in the introduction has little bearing on their research, and they do not cite other low-carb studies while discussing their own findings/data. The one other point where they mention low-carb diets in the conclusion, they mention that they in fact did not confirm those findings - obviously, since the adherence was so poor. I don't see how this difference of definitions is so dangerous, especially since they are not confirming the findings of other low-carbohydrate studies (if they did that would surely be a problem). If in fact you are correct in stating that they've grossly misrepresented the carbohydrate in other studies (and I haven't looked) then that is certainly a failing on the researchers' part. However, whether or not the definition of "low carb" is 20% or 40% has little bearing on the study's findings.

Let me copy and paste a few lines from the study for you:

(1) "Reduced-calorie diets result in clinically meaningful weight loss regardless of which macronutrients they [b[]emphasize.[/b]"
(2) "When nonnutritional influences are minimized, as they were in our study, the specific macronutrient content is of minor importance, as was suggested many years ago"
(3) "In conclusion, diets that are successful in causing weight loss can emphasize a range of fat, protein, and carbohydrate compositions that have beneficial effects on risk factors for cardiovascular disease and diabetes"

Do you see the word "recommendation" in these sentences? Do you see a mention of the fact that the subjects did not stick to the emphasized macronutrient content? I don't. And that's why every one of those statements is very ambiguous in meaning. They could be (and clearly have been) interpreted as implying that "macronutrients composition makes no difference to weight loss" when in reality, the only thing you could really conclude from this study is that "people can't stick to their assigned macronutrient composition, and when they end up eating the same exact stuff while in a caloric deficit, shockingly, they lose the same amount of weight." In fact, I'd say statement (2) above may suggest that the researchers felt that "calories are all that matters" was a foregone conclusion and have interpreted the results in a way to support this.

Okay, no need to get all condescending. The term is "emphasize," which I am interpreting as "recommend" - because that's what they did. Researchers "emphasized" certain diets to participants, each with a macronutrient focus - without the clients' knowledge, obviously, since both sides were blinded. I thought that much was obvious, but you're going to interpret it differently...and that's fine.

Your first passage is taken directly out of the first paragraph of the conclusions section, and is a legitimate conclusion drawn from what the study did. Each diet emphasized macronutrients differently resulting in a caloric deficit, they each lost a clinically significant amount of weight (meaning not due to natural weight fluctuation, obviously). I don't see what is so troublesome about that. Obviously, they can't make any sort of generalizations about the macronutrients because as you've suggested - people didn't adhere.

The second passage also seems legitimate. They minimized non-nutritional influences (popular culture, media, etc.) by giving specific prescriptions and providing their OWN nutrition education/advice/"emphasis"/etc. Controlling this external variable as best they could was part of the rationale for doing the study, although you seem to have problems with that, too. Unfortunately, interpreted alone and out of context, the statement doesn't make much sense. The researchers summarize the findings of other studies, mentioning how various studies emphasizing either low carb, high-carb, vegetarian, etc., ALL seem to produce weight loss, therefore suggesting that the ratios don't matter. Their conclusion is that if you minimize non-nutritional influences (i.e. non-biological), any diet can be effective. What I take this to mean is that there is no reason to focus on any particular nutrient if you educate the patient correctly and with "enthusiasm and persistence."

Third statement is like the first. The researchers emphasized (by recommending) certain ratios, but everybody lost weight. The point isn't whether or not they adhered, it is that they lost weight in spite of the diverging recommendations (er, emphasis).

So, what you're saying is that this "groundbreaking" study determined that, if everyone eats the same type of food, then reducing the quantity leads to weight loss? Wow, I'm glad millions of dollars were spent on this research. Good thing they didn't make any claims about the type of food being important.....

Whoa, I never said anything about groundbreaking. This all started by me saying that some of the findings were interesting (i.e. that they deserve further exploration). As I've said, some of what the study says is a bit "duh," but when it comes to research someone needs to present evidence that 2+2=4.

So... you admit that the study is flawed and then say that there's no point in me complaining about that? I believe that many of the variables they didn't control for could've been handled much better, but you'd prefer to just "shrug" and say diet studies aren't "clean"? It's the non-chalant acceptance of such crap science that has allowed diet research to become the mess that it is.

Of course the study is flawed. All studies are flawed in some way. What I'm saying about diet studies is that they are inherently messy - ESPECIALLY when dealing with people, because you cannot control every single aspect of a person's biology, you have no way of controlling their genetics, you don't even know that they ate exactly what they report eating - the variables beyond the researcher's control are through the roof. Ergo, they are inherently messy. I think that given the conditions, these researchers did a decent job of controlling and minimizing these variables. Recall that one of their aims was to try to limit the influence of "cultural norms, scientific novelty, and media attention" (among others). Educating all groups equally using your own standards and information is a decent, effective way to do that, but the study is still going to be messy regardless - because it involves people. There is no getting around that - at least, not without some serious ethical violations.

That depends on your definition of a "healther diet." For example, many diets (e.g. Atkins) don't concern themselves with dietary cholesterol and lots of research shows that it probably has no real effect on cardiovascular disease. The same thing goes for saturated fat. Now, perhaps these are "healthy" measures according to the AHA, but I'd say support for their guidelines is far from universal.

Well, here we are arguing over semantics again. If you want to discuss the finer points of a healthier diet, that's fine by me but it occurs to me that minimizing saturated fat intake and eating a fibrous diet are hallmarks of healthful diets. When I mean healthful, those are diets associated with longevity and reduced incidence of diseases. Now if you want a more objective measure, the recommendations put out by the USDA's FNB are pretty authoritative and constitute the basis for US and Canadian recommendations. The WHOs recommendations are pretty similar, and they serve as the basis for dozens of more nations' recommendations. Support for those standards is widespread, but nothing is ever set in stone, as I'm sure you realize.

Huh? I never said to recommend avoiding fibrous foods or suggest people eat more saturated fat. I'm saying that ALL you tell people is to eat a particular macronutrient breakdown and let them figure out how to get there. I'd say this is an incredibly important part of the diet. For example, eating low carbohydrate may get people to eat more fruits & veggies to satisfy the carb needs, as they are less carb dense than pasta, bread, cereal, etc. Or maybe it won't. Trying to eat less fat may push people away from red meat to lean meats. Or maybe it won't. But I believe that those kinds of decisions are a key aspect of how effective the diet is. Forcing everyone to eat the same type of food not only introduces a whole new variable, but may also distort the normal way people go about achieving one macronutrient breakdown or another. In fact, perhaps this is why the participants couldn't stick to their assigned macronutrient breakdowns, so the mere act of introducing these new dietary variables destroyed any chance of proper adherence to the recommended diets.

The problem with just telling people a macronutrient ratio is that it would have informed them (either researchers or subjects as to what group they were in, undermining the "blindness" of the study - which was one of the reasons for doing it in the first place. I'm sure there are studies that do exactly that. Consumer behavior is a pretty tricky science and not even corporations with their massive marketing budgets have it down to a science (although it seems packing a food with fat and sugar is sure-fire way to sell it like gangbusters). The researchers gave sample menus based on the desired macronutrient ratios, and it didn't seem to work. Maybe this did cause the lack of adherence, as you suggested, but informing them was out of the question because it would have violated the entire study design.

I don't know how much and neither do the researchers. It could be a 5% difference or it could be 100%. It's the fact that we don't know that makes the study's conclusions questionable.

Yes, but exercise recommendations are part of the standard recommendations they usually give people in these weight loss studies. Simply because they do not "account for the exercise" is not a reason to discredit the entire study, which is certainly what you appear to be doing in the last few posts. It is a weakness, to be sure, but not a particularly novel one. There are few cheap, indirect, accurate ways to actually estimate the amount of energy expended in exercise that there's really little point in measuring it. You accept that it's something you can't control as well as you'd like, you do your best, and you move on. The attempt by the researchers to limit the participants to 90 minutes/week seems to be a cursory, although well-intentioned attempt to do so.

In my experience, the more untrained the person is, the faster their gains early on, regardless of the type of exercise or intensity. It's only in the later stages that the intensity needs to be continuously increased to keep producing results. For example, early on, you can make some pretty big gains in endurance/stamina by just doing a brisk walk a few times a week. As time goes on, you'll need to walk either faster (or jog), longer or more often to see results. Same thing for weight training: in the early stages, almost anything works. Go to the gym, do a little bench, some curls, some sit-ups and your muscle mass increases and fat decreases. After a few months, to keep making gains, you need a more structured program with proper choices for weight progression, reps, sets, etc.

I'd agree with you completely. The question is, how much did they exercise? Given the rapid progression most people experience, my guess is that they didn't that much. On average, I would say weight loss was not incredibly high given the time frame. Slightly more than half lost less than 5% of their body weight. Perhaps the rest went crazy at the gym and exercised more than ninety minutes. We don't know. Is that a weakness? Certainly. Does it "cast doubt" over the conclusions? I'm not inclined to think so. Fifteen minutes is not a lot of time for producing that much of a expenditure in calories, especially not in a population where exercise induced asthma is not uncommon, they are generally untrained and/or inactive, etc.

Overall, is it the best of studies? No. It would have been interesting to see what happened if the subjects had actually adhered to the diets. However, I think your reaction to the study is a bit exaggerated. There are a number of interesting findings which I think are worthwhile to take away from the study and consider - until, of course, the next study comes along and disproves them.

 

[brikis98]

Originally posted by: Kipper
You seem to be harping on and on about the definition of low-carb when it really has no bearing on the findings. Whether they misapply the definition of low-carbohydrate in the introduction has little bearing on their research, and they do not cite other low-carb studies while discussing their own findings/data. The one other point where they mention low-carb diets in the conclusion, they mention that they in fact did not confirm those findings - obviously, since the adherence was so poor. I don't see how this difference of definitions is so dangerous, especially since they are not confirming the findings of other low-carbohydrate studies (if they did that would surely be a problem). If in fact you are correct in stating that they've grossly misrepresented the carbohydrate in other studies (and I haven't looked) then that is certainly a failing on the researchers' part. However, whether or not the definition of "low carb" is 20% or 40% has little bearing on the study's findings.

As I've already said, they compare themselves to "previous trials" of low carb studies in both the intro and the conclusion, but since their low carb is vastly different than that of other studies, this is inappropriate. And just from the perspective of reason, when they were designing the study, they should've seen that the lack of a truly low carb or ketogenic diet is a BIG weakness in the study. They are the most controversial and well known macronutrient breakdowns for dieting today and probably the most important ones to research. They cite over a dozen of them in the intro (including many with the word "Atkins" in the title) but then don't include them in the study. Maybe they did have a good reason for it, but if so, they should've clearly explained it. Omitting this explanation is highly suspect.

Originally posted by: Kipper
Okay, no need to get all condescending. The term is "emphasize," which I am interpreting as "recommend" - because that's what they did. Researchers "emphasized" certain diets to participants, each with a macronutrient focus - without the clients' knowledge, obviously, since both sides were blinded. I thought that much was obvious, but you're going to interpret it differently...and that's fine.

As I said, all those sentences are ambiguous and can be interpreted in multiple ways. Given that numerous news agencies reported the findings with the wrong interpretation, this is at the very least poor writing ability on the part of the researchers. At the worst, it's bad science.

Originally posted by: Kipper
Of course the study is flawed. All studies are flawed in some way. What I'm saying about diet studies is that they are inherently messy - ESPECIALLY when dealing with people, because you cannot control every single aspect of a person's biology, you have no way of controlling their genetics, you don't even know that they ate exactly what they report eating - the variables beyond the researcher's control are through the roof. Ergo, they are inherently messy. I think that given the conditions, these researchers did a decent job of controlling and minimizing these variables. Recall that one of their aims was to try to limit the influence of "cultural norms, scientific novelty, and media attention" (among others). Educating all groups equally using your own standards and information is a decent, effective way to do that, but the study is still going to be messy regardless - because it involves people. There is no getting around that - at least, not without some serious ethical violations.

Yes, diet studies on humans are messy. I personally believe this one is much messier than others, and much messier than it needed to be.

Originally posted by: Kipper
Well, here we are arguing over semantics again. If you want to discuss the finer points of a healthier diet, that's fine by me but it occurs to me that minimizing saturated fat intake and eating a fibrous diet are hallmarks of healthful diets. When I mean healthful, those are diets associated with longevity and reduced incidence of diseases. Now if you want a more objective measure, the recommendations put out by the USDA's FNB are pretty authoritative and constitute the basis for US and Canadian recommendations. The WHOs recommendations are pretty similar, and they serve as the basis for dozens of more nations' recommendations. Support for those standards is widespread, but nothing is ever set in stone, as I'm sure you realize.

My point with this was two fold: (1) what is considered "healthy" is changing all the time, so saying that less saturated fat, less dietary cholesterol, etc are hallmarks of every "healthy" diet is very inaccurate and (2) introducing these factors into the study could have easily invalidated the results and without a control group, we have no way of knowing.

Originally posted by: Kipper
Simply because they do not "account for the exercise" is not a reason to discredit the entire study, which is certainly what you appear to be doing in the last few posts. It is a weakness, to be sure, but not a particularly novel one.

If this was the only weakness in the study, then sure, we could probably ignore it. But it's not. When all of the weakness are put together, I think a reasonable person has to discredit most of the study's findings. As I said, the only conclusions they can soundly make are that "sticking to various macronutrient breakdowns is hard" and that "calorie deficits lead to weight loss." Everything else is speculation.

Originally posted by: Kipper
Overall, is it the best of studies? No. It would have been interesting to see what happened if the subjects had actually adhered to the diets. However, I think your reaction to the study is a bit exaggerated. There are a number of interesting findings which I think are worthwhile to take away from the study and consider - until, of course, the next study comes along and disproves them.

My reaction is exaggerated because the average person's reaction - in no small part due to the way news agencies present the study - is going to be "macronutrient breakdowns don't matter." This is not something you can conclude from the study and many people will (and already have) interpret this in the worst possible way (I can eat crap and be healthy!). I'm upset at the researchers for the weaknesses in the study and the way the results were presented, but obviously, it's not their fault that news agencies distort this data even further. I therefore feel compelled to present a strong case against this study to, at the least, make people read it more carefully to see what the findings really were.

 

[brikis98]

Originally posted by: Kipper

Here's a slightly better analogy of what this study actually did: imagine we were doing an experiment to determine which of two new high tech sneakers, A or B, helps reduce running injury rates more. However, to be "ethical" and make sure the athletes didn't hurt themselves, we had both groups learn the POSE running method. At the end of the study, we found that both groups had reduced injury rates by 20%. Is it reasonable to conclude that both sneakers are equally effective at reducing injury rates? Of course not! The POSE technique - just like more fiber, more exercise, low glycemic foods in the original study - is known for producing the exact same kind of effects that we're studying. Maybe without POSE, sneaker A would've reduced injury rates 10%, but sneaker B only 5%. But perhaps POSE is so effective that it totally overshadowed our results. There is no way to tell and it's bad science to arbitrarily pick one. It's this kind of awful science that is incredibly prevalent in diet research that is a huge part of why we can't stop this obesity epidemic.

My cigarette smoking analogy was intended to demonstrate the ethical problem with such a cigarette study, not as an analogy to the current study itself, but I do understand your objection that other variables may have been responsible (see above). Fair enough. But how much? It's unlikely that they all played a role together, but then the variables you mentioned above (low SCFA, fiber, GI, etc.) are all hallmarks of a healthy diet.

I meant to post this in an earlier reply but forgot: here is an extended version of the analogy above. Hopefully, it'll illustrate more clearly the weaknesses I see in this diet study.

Introduction
Some studies show that Asics, which have lots of cushioning, are better. Other studies show that Nikes, which have little cushioning, are better. Still other studies show that Vibram FiveFingers, which have no cushioning at all, are the best of all. We wish to do an experiment to see how the amount of cushioning a sneaker has affects running injury rates.

Methods
(1) Get a group of runners and get a baseline injury rate measurement
(2) Split the runners into two groups: one group will run in Asics, the other in Nikes. [We will not explain why Vibram FiveFingers aren't tested or why there is no "control" group.]
(3) The logos will be removed from the shoes, so neither the researchers nor the runners will know what they are wearing.
(4) The participants will be instructed to run 3 times a week and after a year, we'll measure injury rates again.
(5) To ensure our study is ethical, we'll take measures to make sure we follow the typical recommendations of a "healthy exercise program." All participants will be told to stretch before and after running - wouldn't want any pulled hamstrings! We'll also have them do a strength and conditioning program to ensure their heart & muscles can take it. Oh, and of course we'll teach them proper running technique. After all, every healthy exercise program includes these obvious measures! [No mention will be made that these very measures could reduce injury rates and that we are not controlling for this factor at all.]

Results
After one year, we measure the injury rates and find they are 20% lower. However, we notice that after just a few months, both groups actually stopped wearing their assigned sneakers and instead switched back to their old New Balance shoes with a medium cushion sole. Conclusion: wearing Nikes and Asics result in clinically meaningful reduction in injury rates, regardless of how much cushioning the sneakers have. [We should say that participants were unable to stick with using Nikes or Asics, but instead will use this ambiguous sentence and hope you get the implication.]