Sunday, February 25, 2018

Registered Dietitian Health Educators: how fat do you want to get?

I guess everyone has seen this:

Effect of Low-Fat vs Low-Carbohydrate Diet on 12-Month Weight Loss in Overweight Adults and the Association With Genotype Pattern or Insulin Secretion The DIETFITS Randomized Clinical Trial

How do you sum up the trial (apart from tedious)? These quotes do it up for me:

"The intervention involved 22 instructional sessions held over 12 months in diet-specific groups of approximately 17 participants per class. Sessions were held weekly for 8 weeks, then every 2 weeks for 2 months, then every 3 weeks until the sixth month, and monthly thereafter. Classes were led by 5 registered dietitian health educators who each taught 1 healthy low-fat class and 1 healthy low-carbohydrate class per cohort"

" emphasis on high-quality foods and beverages"

"...focus on whole foods that were minimally processed, nutrient dense, and prepared at home whenever possible."

This should be a good intervention.

Except decision making was then handed to the participants:

"Then individuals slowly added fats or carbohydrates back to their diets in increments of 5 to 15 g/d per week until they reached the lowest level of intake they believed could be maintained indefinitely"

End result of this is that 10% of the participants weighed more at the end of 12 months of closely supervised healthy eating by a Registered Dietitian Health Educator than they did at the start. Cracking intervention for these poor folks.

And in both the low fat and the low carb groups just under 5% (LF 4.3%, LC 3.6%) of participants developed metabolic syndrome, who had been relatively healthy before the start of the intervention. How can you manage this with "healthy" food and 22 meetings with a Registered Dietitian Health Educator? I'm impressed. If a Registered Dietitian Health Educator ever comes your way, RUN. Especially if they use the words "healthy" and "diet" in the same sentence.

Bottom line: Low carb diets only work when you limit the amount of carbs you eat. However "healthy" those carbs you add back in might be, depending on the opinion of a Registered Dietitian Health Educator, it's no longer a low carb diet. You'll get fat again.

Of course the same applies to low fat diets, especially if they are sugar restricted at the same time. Ultimately if you follow a low fat diet with as much added fat as you feel comfortable with, you're going to be disappointed with the results too. Adding back sugar will be even more disastrous. Sad but true.


Addendum: Gardner did essentially the same study in 2007 but made the mistake of publishing the weights alongside the carb intakes at each assessment interval. I wrote all over his graphs here. He didn't repeat the mistake. No one should imagine he's stupid. Or honest.


raphi said...

i call it the Moderate-fat vs Moderate-carb study

this extract from the study gets a lot of what's wrong into a single sentence ==> "Dietary intake of fats and carbohydrates was well differentiated between the 2 diet groups, as confirmed by diet assessment"

1) conflating statistical significance with biological significance
2) diet questionnaires are witchcraft not science

Peter said...

raphi, there is almost no end to the holes you can pick in this study. Gardner is not quite as criminal as folks like Barnard but he's not going to help anyone... Perhaps excepting his own ego.


spfldo said...

"ultimately if you follow a low fat diet with as much added fat as you feel comfortable with...". Personally, I eat a lot of fat, but almost zero carbs. I typically add a tablespoon of cream cheese with each of ten half slices of thick cut bacon, ham or sausage, every morning. With that, I also eat three or four eggs, and four to six ounces of cheese. Two or three times a day, I have a tab or two of butter and coconut oil. Beef, meatballs/meatloaf, roast chicken, sour cream...etc. for dinner. I fast 6/18 every day, and after initially losing 30lbs., haven't gained or lost a lb. In four years. Fat doesn't make you fat, in my experience.

bill said...

The Woo did a great analysis of the study:

Peter said...

spfido, you clearly need a Registered Dietitian Health Educator to teach you to eat essential carbohydrates and maybe brainwash you to believe you need at least 150g/d of carbs to feed your brain and waistline.

bill, of course.


cavenewt said...

Peter said

"Addendum: Gardner did essentially the same study in 2007 but made the mistake of publishing the weights alongside the carb intakes at each assessment interval. I wrote all over his graphs here. He didn't repeat the mistake."

After some searching, I finally located the supplementary material for the study.

The very last table is what's called "weight trajectory". I'll leave it to more knowledgeable people to interpret.

Peter said...

Thanks cavenewt,

Just a kilo ahead, on average, for the lowish-carb group. Not very exciting, sigh!


cavenewt said...

Peter: exactly. But as you have pointed out, it wasn't what we would consider a low-carb intervention. As Wooo pointed out, both arms were encouraged to lower their carbs! I would quibble with Wooo a little bit about significance of the very slight differences in weight loss/weight gain when comparing the two groups. And, there was no control group.

I think this study was trying to do too many things at once.

1. Comparing what they claim is low-carb vs low-fat

2. finding a sustainable diet for both groups (which ended up muddying the waters of #1 to a considerable extent)

3. using insulin sensitivity to split the group into the various dietary arms (which I have a gut feeling really screwed up the whole thing, but I can't articulate why)

4. Seeing how DNA affects all this (is this part of the analysis still ongoing? I've considered it mostly irrelevant for my interests.)

Puddleg said...

While the new Metsyn cases are what they are, there's also a significant improvement of TG/HDL in the low carb arm that's not happening to the low fat arm.

The insulin stratification by tertiles is odd, because it divides the study into 6 arms and reveals that despite there being over 100 in each arm, randomisation has not distributed weight very evenly in the high and low insulin bands.

And there is an interaction between insulin and diet in efigure 3b at 3-6 months, which is usually when the effect of adherence is strongest. Weight drops most, with a significant difference, in HLC Hi Ins vs HLF Hi Ins. Based on this you would in fact prescribe LCHF to the insulin-resistant - you'd just tell them to stick to it.

Medicine is prescriptive - it does not need to be permissive.

mem said...

You're still at it after all these years, Peter. And I am so glad.

Thank you! Your confirmation means much! NEVER stop blogging!!!

Peter said...

George, yes, on lipids the lowish carb arm does well at times.

mem, ta.


Anand Srivastava said...

Why is Gary Taubes conducting such studies?

Eric said...

Is he?

ctviggen said...

Did it seem like they were trying to do way too much with this study? They were trying to test low fat versus low carb, plus genotype, plus insulin secretion. Yikes! And, as already pointed out, they merged the LF versus LC to be too close to each other.

I mean, if you really want to compare genotype and insulin resistance, randomize them into an all-potato diet or a carnivory (all meat) diet! ;-) THAT would at least help point out the differences.

And I wouldn't mind seeing the results of that study. I see people who go on the potato diet or similar, with reported success and lowered blood sugar/HbA1c. Whereas when I eat potatoes, my blood sugar shoots through the roof. Why the differences?

I am surprised it has Ioannidis, who I (at least used to) consider one of the best researching minds out there.

cavenewt said...

Scratch my #3 above. I was reading the original methods, which got changed later on. Or I misunderstood, or something.

Unknown said...

Anyone interested in Stephans opinion on this study?

Seems he believes it's a "big deal":

cavenewt said...

@0949 JW

NuSI kicked in $3 million to the funding after the original study methods were already published. So I don't think it's fair to say it's a "NuSI-funded diet trial", which implies undue influence.

Aside from that, the article was more evenhanded than one might expect from Stephan.

From DIETFITS Supplement 2:

"3. Approached by the Nutrition Science Initiative (NuSI). In the fall of 2012, just after being funded by NIH, we were approached by the newly founded NuSI group with an extraordinary offer to provide an initial $3M in unrestricted funds to augment the study if we could also raise a matching $2M in unrestricted funds from donors within an 8-month time frame (January 1-August 31, 2013); i.e., a total augmentation of $5M to the initial $3.2 from NIH.

"The primary request from NuSI was to make the study more rigorous and more impactful. The specific factors agreed on were to make the study larger (more participants) and more extensive (more secondary and exploratory outcomes measures) without extending the timeline (faster pace of enrollment)."

cavenewt said...

@anand srivastava and @Eric

Taubes is still on the board of NuSI.

Unknown said...

I've had roughly the same experience. Lost 75 eating low carb. The less (carb) I eat, the better I do. I fast 15-20 hours Monday-Friday. Normally don't eat too many carbs (20-30g on most days). I've been zero carb for the last 3 weeks to see what would happen. I've done well. I'm 44 and hover between 9.5-10% bf. Train fasted, and actually have packed a lot of muscle onto my frame. I eat a lot of very fatty cuts of meat. Make mayo. Eat cheese, eggs, and cook everything I butter, lard, or bacon grease. As long as my carbs are low, the fat is of no real concern to me. Plus it makes everything taste delicious.

Eric said...

Aaron Carroll's take in the NYT:

His take: what a surprisingly well-designed study!

Peter said...

What does he think a crap design looks like???????????


JR said...

There is potentially an interesting sub group study within Dietfits: the size of fat cells.
McLaughlin found out a difference inbetween the lo-carb and lo-fat groups thru biopsy: despite the similar weight loss, the size of fat cells was consistently smaller in lo-carb group, and there was virtually no change within the lo-fat group. So, lo-carb lost fat mass, lo-fat did not so much. And the difference in insulin levels ...maybe the millions were not wasted after all.

There seems to be no papers as of yet, just this presentation, but this result from "official" and impactful research might turn into something interesting. Of course Peter has been posting on fat cell size and leakage and its meaning to T2D earlier on.