Saturday, December 20, 2014

The A-Z Weight Loss Study - Atkins Wins?

Lately, a 2007 study has been circulating Facebook and the blogosphere, mostly being cited as support for a low-carb diet.  Brainy Kris Gunnars reviewed it on his blog as a high-quality study showing that high-fat diets outperform vegetarian/vegan ones in terms of weight loss and cardiovascular risk markers. 

After reading that, something twigged in my brain. I realised that we are talking about this study:

Comparison of the Atkins, Zone, Ornish, and LEARN Diets for Change in Weight and Related Risk Factors Among Overweight Premenopausal Women.

I read it a while back and put it on my pile of studies that don't say much. I'm going to quickly critique the study now, as an example of how sometimes, what we think we see, is not what is there. You have to look beyond what you read on the internet or on Facebook. Don't trust anyone, especially bloggers ;).

The first, and most important issue, is compliance.

Leaving aside the fact that dietary compliance was assessed by telephone food recall, the A-Z study appears to have a serious problem with compliance to the diets, except Atkins. There are a couple of things that stand out to me.

The Zone Diet is supposed to be 40:30:30 (percentage of carbs, protein, fat) and yet, by the end of the study, the Zone dieters were achieving just 20%, and their protein intake had been dropping at each recall.  

The Ornish diet is supposed to be very low fat (less than 10% of calories from fat) and yet by at the 12 month recall, the Ornish dieters were eating approximately 29.8% of their calories as fat. Surprisingly, these low-fat, vegetarian dieters somehow managed less than 20g of fibre a day. That's not even the RDI, which in NZ is 25g a day. This suggests to me that they were perhaps eating a few too many processed carbs instead of veges.  Just a guess there. Therefore, although the Atkins group were Atkins-ing, the Ornish group weren't Ornish-ing and the Zoners were not in the Zone. The LEARN group didn't meet their dietary targets (55-60% calories from carbs, less than 10% calories from saturated fat) either.

So really, we are not comparing the results of the Atkins, LEARN, Zone and Ornish diets at all. This is major. Other studies like the POUNDSLOST trial have shown that dietary compliance matters more than nutritional breakdown when it comes to weight loss.  I think the A-Z authors knew they had a problem there. A year later, some members of the original study group re-analysed the data for compliance and found this: 

Twelve-month weight change in the most vs least adherent tertiles, respectively, was −8.3±5.6 vs −1.9±5.8 kg, P=0.0006 (Atkins); −3.7±6.3 vs −0.4±6.8 kg, P=0.12 (Zone) and −6.5±6.8 vs −1.7±7.9 kg, P=0.06 (Ornish).

That messes with the notion that only the Atkins group lost lots of weight, although the most compliant of them did lose the most. It seems that the most compliant lost more than twice the previously reported amount on the Zone and nearly three times the previously reported amount on the Ornish diet. The LEARN diet was not included in the adherence analysis. The authors of the review concluded this (bolding is mine):

Regardless of assigned diet groups, 12-month weight change was greater in the most adherent compared to the least adherent tertiles. These results suggest that strategies to increase adherence may deserve more emphasis than the specific macronutrient composition of the weight loss diet itself in supporting successful weight loss.

Oh dear. That's interesting. I wonder why that follow up study isn't bounding around the interwebs? Probably because it's a predictable and boring finding.  Nobody likes to cheer for the mainstream.

Let's have a closer look at the actual results from the original study.

Firstly, there is something important to realise from the get-go.  Apart from the difference between Atkins and Zone Diet (which as mentioned above, was not really), the weight lost between groups was not statistically significant at 12 months. That means that given the group sizes, we can't be sure that the apparent difference in weights is not just chance.

There is also something weird going on with bodyfat levels. It's a bit of an elephant in the room, if you ask me.  Have a look at this:

The readers of Fit to Blog are pretty smart and I'm sure you can see what's going on.  In the last six months of the study, all groups except Ornish are regaining fat and increasing their BMI. The LEARN group regained the most bodyfat, with Atkins second.

I would dearly love to see a follow up study to see where these individuals (all women) are now. That would be truly interesting, yes? I wonder how they are doing seven years later? 

Just a little note about the other significant findings, which were blood pressure, triglycerides and HDL. At 12 months, the Atkins group did show significant lowering of triglycerides compared to the Zone group, significant lowering of blood pressure compared to all groups and significant increases in HDL ('good' cholesterol) compared to Ornish. You can see this in the table above. The results with little a's and b's beside them are significant results, but numbers that share a superscript are not significantly different to each other (making sense??). I was a little 'meh' about these findings though, because the group as a whole had pretty good levels of these things at baseline, and of course, we still have the compliance issue.

In summary, it is really difficult to see how the Atkins foundation (it's even on their Wiki page), and other low-carb fans, can take this study and use it to smack the internet with the healthfulness of their diet. I'll need to see something a whole lot more convincing than this before erecting an altar to Atkins in my office.  



  1. thanks for this, Sara. Quite a few specialists at our hospital got into Atkins and lost weight some years ago but now have pretty much regained or, yikes, are now even bigger.

    1. I've seen that happen so often, with many different diets. A while back, a group of my friends paid $1000 for the Cohen diet. All but one have regained all the weight or more. There are random successes though. I've got a friend that quietly did Jenny Craig and never regained. There are success stories all over the place, people that did a reality check and tidied up their lifestyle. Sometimes I think chance of success is directly correlated with how big a deal a person makes of it all. Haha.

  2. Hi Sara!

    Very interesting how people have serious problems to comply with strict diet recommendations. Nevertheless I think the problem with staying lower than 50g of carbohydrates a day in the LCHF case must come from still eating wheat and taking too much sugar. I don't think it should be difficult at all to stay at lower than 50g (certainly not at lower than 150g) with rice and potatoes Jaminets' style (even with pulses). I don't think HCVLF is a natural diet for humans (we have gallbladders and need liposoluble vitamins) but it could be an option to try as I have commented elsewhere. I think that using a glucometer should be standard when one wants to know how healthy is one's diet for oneself.

    It would have been very informative if the A-Z reanalysis had included body fat statistics.


    1. Hey Andre, how goes it? Just looking at the fibre of all the diet groups, I'd guess that there weren't many beans and starchy veges going down at all! The participants were paid (not extravagantly, but the amount increased at each checkpoint), and I wonder if there were too many in the study for the money, not taking it seriously.

      The glucose meter I have used is Accutrend, quite a nifty device. I haven't used it for a while, but did note that when I stopped low-carbing, the postprandial readings were awful for a while then became normal. I'm concerned that long term low-carbing can actually reduce the body's ability to metabolise carbs and so you can get this rebound hyperglycemia. That's not to say low-carbing doesn't have a therapeutic role, but .. well, that's another whole blog post really....

      I agree with you on the fat soluble vits and that they should have given the BF% changes in the second study. You'd expect the higher protein groups to have maintained more lean mass, right? They went to the bother of getting DEXA scans! And then.. didn't use that data in the second analysis. :-/

    2. Hi again! Doing mostly great around here!

      Certainly I doubt the random trial is of much use for this subject just because of what you have said about "forced" compliance. I think that knowing what others may accomplish wishfully with a given diet is informative enough to try it or not.

      Of course going too low carb is going to impair glucose tolerance in the short term. Actually I am not there and I don't even count (when I do) vegetable carbohydrates outside of fruit and starches as advised by the Jaminets. I think I am mostly below 50 but some days I approach 150g/day. I have measured my blood glucose even after we eat octopus with lots of boiled potatoes and lots of olive oil and it stays well below 140 at 1 hour and 120mg/dl at 2 hours.

      Yes, more protein should give rise to higher lean mass as observed in other studies. I would bet that the data favored the HFLC group.

      Happy new year!

  3. Anonymous1:52 am

    Hi Sara,
    you say "the Atkins group were Atkins-ing", but they were not. 35% of energy from carbs is not a low-carb diet. That is for sure the reason why the participants gained weight again: you eat carbs again, you gain weight again.
    Adherence was a problem in this study but I can't see how adherence to a bad diet is a solution for obesity. I disagree with that opinion.

    I don't follow the Atkins diet but I eat low-carb. My clinical results are published in my site, and I am healthier than ever. My weight has been stable for the last 12 months, without any kind of effort.

    Best regards

    1. Yes, I thought about that when I wrote it. But, after induction, I think the Atkins diet doesn't specify an amount of carbs, but just instructs you to increase carbs by 5g a week until you are comfortable with your weight loss or maintenance and then reduce if you want to start the weightloss again. They may well have been doing that (thus 'following the diet'). They did 2-3 months induction, then 7 months. In looking again at the study though, their 'target' was 50g/day, and they overshot that by about 80g/day, if their calorie estimate is correct (and it seems really low, IMO). So, technically, not very low carb, but still low-ish, and possibly still following the Atkins guidelines (?).

      The idea that adherence is crucial is not my opinion, it's from studies, however these are not long-term studies the usual is 12-24 months follow up. I think the idea is that commercial diet programs (of any sort) are designed so that you 'accidentally' eat fewer calories. So if you stick with the guidelines, you'll be in calorie deficit. Permanent solution for obesity? I don't think any 'diet' is.

      Congrats on the health improvements. I had a look. Nice to see the liver enzymes come right! It can be done, that's for sure!

    2. Anonymous1:24 pm

      thanks. I really had a health problem (obese with NAFLD, GERD and too high ferritin levels). My health improvement is the reason why I now have 100% adherence to this way of eating.

      Time will tell if low-carb is the permanent solution for me. I sincerely hope it is.
      Best regards

    3. What I'd really like to see is a study really looking at carb source in various contexts. It's a constant annoyance to me that, for example, in a study where people do better on a low carb diet, or replace fats with carbs and do badly, they never specify where those carbs came from. It could be veges or beans or it could be grains or refined grains. GERD is tricky. You've done very well.


I love to hear from you! Tell me what's in your brain, your heart or your dinner plate :D.