Topic: "10 Body Myths Debunked" - counting calories?

[Alex Simmons]

Well, sure--but their driving on the wrong side of the street.

am I too late for the pedantic correction portion of the thread?

Well it depends on what the correction is.  It could be an error in word choice, or the correct word choice but in an incomplete sentence.

[ConspicuousCarl]

As you previously explained, you have based your conclusion on the fact that the 12-month difference in weight loss between the Atkins and Zone diets was only 0.4 kg greater than the diet effect quoted in the paper that the study had 80% power to detect—2.7 kg.  But both Meg and I have tried to explain to you that this comparison is irrelevant.  The 2.7 kg figure was merely used for planning purposes, and has no role in the analysis.

So there is another reason for them to declare that there is NOT a significant difference between Ornish and Atkins, a difference of about 2.6.  Go ahead and point me to the correct reason for that.  They sure seem to be taking that into account.

Although adherence to the 4 sets of dietary guidelines varied within each treatment group and waned over time . . . we believe that the adherence levels obtained are a fair representation of studying the diets and variations in macronutrient intake under realistic conditions and, therefore, increase the external validity of the findings.

So, it is neither true that the study's findings were barely significant nor that the study was compromised by problems with compliance or its measurement.

The big zinger, from which you pretend to have spared me, is to quote the authors ACKNOWLEDGING the possibility of the very thing I said could be the case?  But I am "wrong" because the authors say they are not especially concerned with the difference between biological and compliance causes?

So let's pretend that I believe your claim to being condescending.  This whole time, you are pretending to argue that the study design and results are not fairly assessed as being open to a possibility that the weight loss was caused by compliance, even though you are sitting on the useless point that the authors actually think that is true but don't care.

Whatever.

And you seem to have forgotten that this study was offered up as supporting a specific claim.  Saying that the people behind the study are not concerned with exactly which effect was measured does NOT change the fact that the study is weak evidence for the subject addressed here.

[ConspicuousCarl]

It is also worth noting that, even if they were OK with measuring weight loss variation from compliance variation, the design of the study wasn't really all that great even for that.  As I said, a dozen random calls to ask about the previous 24 hour food intake in order to measure their calorie intake for the entire year is open to all kinds of problems, even if you aren't trying to get them to eat a specific amount.

[jt512]

As you previously explained, you have based your conclusion on the fact that the 12-month difference in weight loss between the Atkins and Zone diets was only 0.4 kg greater than the diet effect quoted in the paper that the study had 80% power to detect—2.7 kg.  But both Meg and I have tried to explain to you that this comparison is irrelevant.  The 2.7 kg figure was merely used for planning purposes, and has no role in the analysis.

So there is another reason for them to declare that there is NOT a significant difference between Ornish and Atkins, a difference of about 2.6.  Go ahead and point me to the correct reason for that.  They sure seem to be taking that into account.

First of all, the difference in 12-mo weight loss between Atkins and Ornish was 2.1 kg.  Secondly, they did not "declare" that there was no statistically significant difference between these diets.  Reporting a non-statistically significant result is not a declaration; it's the lack of a declaration.  When two research findings are not statistically significantly different, as a rule, you can't make any claim about what the true difference is.  In particular, a non-statistically significant difference does not imply that there is no difference in fact.  That is, failure to reject the null hypothesis is not, as a rule, support for the null hypothesis.

Third, the authors' stating that they found no significant 12-mo weight loss difference between Atkins and Ornish is not based on a comparison of the observed difference with 2.7 kg.  The authors explain how they made determinations of statistical significance in the Statistical Analysis section of the paper: "Differences among diets for 12-month changes from baseline were tested by ANOVA.  For statistically significant ANOVAs, all pairwise comparisons among the 4 diets were tested using the Tukey studentized range adjustment. . . . All statistical tests were 2-tailed using a significance level of .05."  Simply put, they compared the diet results, not with 2.7 kg, but with each other, calculated p-values for each pair of diets, and declared the pair of diets significantly different when the calculated p-value was less than .05.  I hope that this is finally clear.

Although adherence to the 4 sets of dietary guidelines varied within each treatment group and waned over time . . . we believe that the adherence levels obtained are a fair representation of studying the diets and variations in macronutrient intake under realistic conditions and, therefore, increase the external validity of the findings.

So, it is neither true that the study's findings were barely significant nor that the study was compromised by problems with compliance or its measurement.

The big zinger, from which you pretend to have spared me, is to quote the authors ACKNOWLEDGING the possibility of the very thing I said could be the case?  But I am "wrong" because the authors say they are not especially concerned with the difference between biological and compliance causes?

First of all, I did more than quote from the paper.  I explained why the study design and the author's comments imply that they purposefully left compliance up to subjects, and therefore that your criticism about compliance was mistaken.  The aim of the study was to determine if these diets lead to differences in weight loss, as the diets are actually implemented by dieters in the real world under real-world, rather than laboratory, conditions.  The fact that you ignored my explanation does not mean that my entire argument (the so-called "zinger") consisted of one quote from the paper.  What I said you were wrong about is that actual compliance, or its measurement, was a weakness of this study.

So let's pretend that I believe your claim to being condescending.  This whole time, you are pretending to argue that the study design and results are not fairly assessed as being open to a possibility that the weight loss was caused by compliance, even though you are sitting on the useless point that the authors actually think that is true but don't care.

I would like to pretend that I can make sense out of that paragraph, but I'm not that good an actor.

And you seem to have forgotten that this study was offered up as supporting a specific claim.  Saying that the people behind the study are not concerned with exactly which effect was measured does NOT change the fact that the study is weak evidence for the subject addressed here.

Looking back, the study was offered in support of two claims: (1) that "counting calories is ineffective for weight loss" and (2) that "weight loss in the group with the most weight lost was directly proportional to their carbohydrate intake."

With regard to the first claim, although I don't agree with the claim, I do agree that the study supports it.  Subjects on the Atkins diet, which did not count calories, lost more weight than subjects on any of the other diets, including the Zone and LEARN diets, which prescribed caloric restriction.

The second claim doesn't make sense as worded, but it is true that the group with the greatest weight loss, Atkins, had the lowest carbohydrate intake.  Carbohydrate intake for the Atkins group was significantly lower than every other diet group at every post-baseline assessment.  The method of assessing dietary intake, which you mis-characterized as "shaky," was, in fact, adequate to assess carbohydrate intake at the group level. 

So, no, I do not agree that the study is weak evidence for these claims.

Jay

[jt512]

It is also worth noting that, even if they were OK with measuring weight loss variation from compliance variation, the design of the study wasn't really all that great even for that.  As I said, a dozen random calls to ask about the previous 24 hour food intake in order to measure their calorie intake for the entire year is open to all kinds of problems, even if you aren't trying to get them to eat a specific amount.

And, as I have said, you're wrong about that, too.  Dietary intake was assessed by 3 random 24-hr recalls per subject per assessment period (baseline, 3, 6, and 12 mo).  If we assume an average of 70 subjects per diet group, then there are 210 person–days of dietary data per diet group per period.  That is an adequate sample size to make reliable comparisons between diet groups for major nutrients.  Look at the p-values in the last column of Table 2.  The fact that many p-values are highly significant is prima facie evidence that the sample size is adequate. 

To put it another way, the investigators are not treating the data as a sample of "a dozen" days per subject to measure the individual subject's intake for the whole year, because they're not trying to characterize the intake of individual subjects and the relevant period isn't a year.  Rather, they're treating the data as a sample of 210 days per diet group per period, because they're trying to characterize the diet of each diet group as a whole for each assessment period individually.

Jay

[fatblond]

This just in:

http://www.ajcn.org/content/early/2012/01/17/ajcn.111.026328

Basically they found that the 'conventional wisdom' is right, it's the calories (stupid).

Um.. Actually all they said was that those four diets lost more fat than lean mass, whatever those diets were (can't view the study).

If one actually viewed many of the published NIH studies that are claimed to be measuring "lo-carb" diets, often they are not "lo-carb" but simply reduced carb with very little control for the type of carb.   I doubt anybody that is a self avowed lo-carber that has lost weight would say they get 35% of their diet from Carbs. 

A typical induction phase for Atkins for reference allows maybe 20-25g of carbs, with a maximum of about 100g to remain ketotic during maintenence.  (Some can go higher, and exercise allows more, ala TKD, CKD, etc.)  Roughly 80, 100, and 400 kcal total.  Assuming an average 2000kcal diet, no exercise, then 35% of calories would be 700 kcal or about 175g of carbs on the LOW end.  That's a lot of carbs.

These diet trials were not studying lo-carb, the type of carb, atkins, whatever.  They were determining if if caloric composition amongst a few diets mattered.  It didn't.

How one can conclude that calories matter, and that "conventional wisdom" is right from mistating the results is beyond me.

[DRmeg378]

Most processes in the body are continuous, I would think. If there is an effect of the composition of diet one would expect to observe it even in less extreme cases. Unless you have a theory why not, and it is adequately grounded on data.

[jt512]

Most processes in the body are continuous, I would think. If there is an effect of the composition of diet one would expect to observe it even in less extreme cases. Unless you have a theory why not, and it is adequately grounded on data.

Although physiologic processes may be a continuous function of some dietary variable, as a rule they are sigmoidal over a broad range of the variable.  There are upper and lower thresholds beyond which changes in the dietary variable have little effect.  Between the thresholds there is a range of the dietary variable in which the response is relatively linear; however, in some cases, the slope is large enough that the processes can be well approximated as dichotomous.  This is the case for the relation between essential nutrient intakes and their respective deficiency diseases, and for the case mentioned a couple posts up for ketosis and carbohydrate intake.  It could also be true (and there is at least some evidence for it) for appetite and protein intake.  It may be that the relationship looks like this:

If the depicted relationship is true, then over a large range of protein intake, there is negligible effect on appetite, but once a threshold intake is crossed, there is a steep decline in appetite.  At high enough intake, another threshold is reached, and there is no further appetite suppression (so even the highest protein diet won't stop you from eating altogether).  Since high-protein diets are also usually low-carbohydrate, the inverse relation would be observed if the x-axis were carbohydrate intake.  It could also be that low-carbohydrate contributes to appetite suppression independently from high-protein intake.

At any rate, the main point is that it is sigmoidal relationships that are the norm for biologiclal effects of dietary variables, not linear ones, when the dietary variable is considered over a wide range.

Jay

[lonely moa]

I thought the diet comparisons a few posts back included the ketogenic Atkins diet, which is NOT a high protien diet but a high FAT diet.  Appetite in normal humans is regulated in large part by leptin signals from the fat tissue with regard to the energy stored, isn't it, not the amount of protien eaten?

[DRmeg378]

Thanks jay!

[WeeDavie]

I thought the diet comparisons a few posts back included the ketogenic Atkins diet, which is NOT a high protien diet but a high FAT diet.  Appetite in normal humans is regulated in large part by leptin signals from the fat tissue with regard to the energy stored, isn't it, not the amount of protien eaten?

Leptin is the signal generated by Fat that signals to the brain (hypothalamic energy balacne circuit) the amount of Fat in adipose tissue. The energy balance circuit then manipulates all energy related process (including but not exclusily apeptite) to maintain the level of Fat at the correct level.

However, in obese individuals with high amounts of Fat, leptin levels are high as expected, but (crucially) the energy balance does not respond to the high level of leptin by induing weight loss. Two issue have been identitied, a failure to transport leptin across the blood brain barrier and 'leptin resistance' reduced response to leptin in leptin responsive neurons within the energy balance circuit.

The energy balance circuit does however notice when leptin levels drop in obese people due to loss of fat during dieting and responds by activation counter-regulation mechanisms to increase fat/adipose deposits. It basically acts as if the body is starving despite being full of energy.

[lonely moa]

I'm doing an n=1 observational lately.  Three months ago (due to a wee accident) I was cut off from from my weekly routine of 6-8 hrs bicycle training, 4-6 hours hill walking, 2 or 3 weight training sessions, whatever else came up and work on the farm.  I haven't changed my diet in composition (I may be eating a bit more volume, though) and I haven't lost or gained any weight.  I haven't had a look at my body composition yet (9% bodyfat at the start).  I guess I have another 3 months to go without working out... hate to think what my VO2 max will shrink to, but I doubt my weight will change much. 

I did get an HDMI cable today so the sedentary lifestyle will continue.  CS has reccommended a canadian site with great (and free) videos.

[Alex Simmons]

hate to think what my VO2 max will shrink to, but I doubt my weight will change much. 

I wouldn't be too concerned about your VO2max.   It's only partially impacted by training/detraining - and is largely genetically determined.

VO2max isn't a particularly good predictor of performance amongst athletes of relatively similar ability. 

The percentage of VO2max you can sustain at maximal sustainable aerobic levels (think max effort for ~ 1 hour) is a far more trainable and important component of fitness.  That will have declined somewhat during a period of detraining.  But if you have some years of training behind you, then it can be largely reclaimed in a matter of 6-12 weeks of quality work. 

How quickly depends on how detrained you are, chronic exercise history and genetic predisposition to training response.

Pithy Power Proverb - "Choose your parents wisely".

[lonely moa]

hate to think what my VO2 max will shrink to, but I doubt my weight will change much. 

I wouldn't be too concerned about your VO2max.   It's only partially impacted by training/detraining - and is largely genetically determined.

VO2max isn't a particularly good predictor of performance amongst athletes of relatively similar ability. 

The percentage of VO2max you can sustain at maximal sustainable aerobic levels (think max effort for ~ 1 hour) is a far more trainable and important component of fitness.  That will have declined somewhat during a period of detraining.  But if you have some years of training behind you, then it can be largely reclaimed in a matter of 6-12 weeks of quality work. 

How quickly depends on how detrained you are, chronic exercise history and genetic predisposition to training response.

Pithy Power Proverb - "Choose your parents wisely".

Just about to turn 60... like to see it (vo2)  match my age for a bit longer.  I really miss the hours on the bike... I do an 85/15 sort of week.  I miss the HIIT bit almost as much as the cruise.  I am a bit surprised that no spare tyre has appeared but what's left of the right shoulder shakes like a bowl of jelly now... yuk.

[341gerbig]

Good god people, What do I put in my mouth?


Prepare for useless personal anecdote people!!!!!


I was eating 2300 calories in 45% cals from carbs, 30% from protein, and  25% from fat, and i was caught at a weight loss plateau for months.

I switched my macros to 25% cals from Carbs, 50% from protein, and from 25% fat, and i went back to my pre-plateau weight-loss of 0.5 - 1 pounds lost  a week.

This is under the exact same exercise program.