Atkins Diet Under Attack
It's all over the news -- variations on "Atkins Diet is Hard on Your Heart". So of course, I had to look at this new study in the April 2009 Journal of the American Dietetics Association in depth so we can try to understand its implications. The authors say that the purpose of the study was to compare three diets (the Atkins Diet, the South Beach Diet, and the very low fat Ornish Diet) on heart disease risk in the maintenance phase -- that is, when weight was not being lost.
The Bottom Line: If you're not into a lot of science details, I'll cut to the chase: this study is too small and preliminary to be of much use at this time. But it has other major problems as well.
It is certainly important to differentiate between the effects of weight loss and the effects of diet without weight loss, so I was interested. Note that although the authors say that they used the "maintenance phase" of a diet, they did not study maintenance following weight loss; in this context they merely meant a diet intended to maintain weight. (This has been misreported in the press, and could be important, as the body of someone post-weight loss is different than a body maintaining a normal or stable weight.) In the study, each subjects followed each of the diets for four weeks, with a four week "washout" period in between. (The "washout period" and the South Beach diet were both described as being similar to a "Mediterranean diet".)
Whenever I look at a new study involving low-carb diets, the first two questions I ask are, "How is low-carb defined?" and "What did people actually eat?" Time after time, I find that the answers to these questions are crucial in understanding the study. For example, in a recent study which purported to compare diets with various compositions (and was widely-reported as having a low-carb group), the authors specifically decided NOT to include a low-carb group. Unfortunately, in this current study the answer to these questions is "we aren't told." This is a particular problem with the South Beach and Atkins diets, where the amount of carbohydrate in the maintainence phase is tailored to the individual. How did they do this for a group that was not really in a "maintainence phase" at all, since they had never lost weight to begin with? We are not told.
So we have no idea what people actually ate, although we received a couple of hints. The authors were under the impression that the Atkins Diet "advocates a high saturated fat intake". This is not true. Atkins, where he made any recommendations about fat at all, advocated a mix of fats (more about Atkins' recommendations regarding fats). Then we are told that the diets differed on fat intake in that the Atkins group ate an average of "58" fat (I assume this was percentage of calories from fat, although this was not explained), the South Beach group "31" and the Ornish group "9". However, we are given no information about how the diets differed on the amounts or types of carbohydrates.
The Study Participants: 18 adults participated in this study. We are not told the gender mix nor the ages, although the recruitment strategy favors people in their mid-20's to early 30's. The subjects were healthy, "without metabolic disease" and not obese. I interpret this as meaning that the very people most likely to benefit from a low-carb diet -- those who are insulin resistant, have metabolic syndrome, obesity, prediabetes, or diabetes -- were specifically excluded from the study. This impression gains more validity when we see that the participants not only had normal lipid values (cholesterol, triglycerides, etc) but they had what would be considered "very good" values (e.g. average triglycerides below 100 and average HDL over 60) It seems likely, then, that these young, healthy people are likely to be those who's bodies process sugars well.
Results: Over the four weeks of the diet, the average of these 18 subjects tended to improve on most heart disease risk markers when on the "Ornish" diet, improve on some markers on the "South Beach" diet, and stay the same on the "Atkins" diet (I have to put these in quotes since, again, we don't know what people actually ate.) When on "Atkins", there was no significant change in any of the markers. The researchers conclude that "high saturated fat may adversely impact lipids and endothelial function during weight maintainence" (the chart they use to support this correlation is not impressive-looking, BTW -- see comment below by Dr. Richard Feinman for more on this).
My Thoughts on This Study: Three years ago, a study by Ronald Krauss was published comparing three diets, and using both a weight loss phase and a weight maintaining phase. The Krauss study had ten times the number of subjects as the current study, and the results were very different. Heart disease markers changed with low-carb diets along the familiar lines we've come to expect, with the most dramatic changes being improvements in triglyerides and HDL cholesterol. I have to wonder what made the outcome of this current study different. It might be the selection criteria of the subjects. It might be that there was a lot of individual variation that was washed out in the averages (with 18 subjects it would not have been hard to show the range of changes in the individual subjects, but the authors chose not to do this). It might be some aspect of the diet composition that is not described in the published paper. We simply don't know. I do find it interesting that the Krauss study was not even cited in this paper.
The results also run counter to the typical results of people who follow a low-carb way of eating for health purposes. Over the long term, people often report improvements that might not show up in the first few weeks or even months. I use myself as an example. I have maintained a moderate weight loss for about eight years. In each of those eight years, my LDL cholesterol has decreased, my HDL cholesterol has increased (those numbers are only 11 points apart now), and my triglycerides have remained lower than the people in this study. All this while eating a diet low in carbohydrate and high in fat. However, I would not have qualified to be a subject in this study.
Source:
Miller, Michael, Beach, Beach, Valerie, et al. "Comparative Effects of Three Popular Diets on Lipids, Endothelial Function, and C-Reactive Protein during Weight Maintenance." Journal of the American Dietetic Association. 109/4 (2009)
Photo Courtesy of Pricegrabber
Related Resources:
The figure showing the correlation does not indicate the slope of the line. You can draw a line by eye and you can see the slope is close to 0, in other words, although there was a correlation between saturated fat and vessel dilation, the correlation was very weak. A big change in saturated fat meant a very small change in vessel dilation. This is Atkins-bashing pure and simple.
The lack of citation and comment on Chris Gardner’s A-Z study and Jeff Volek’s work is a strong indication of author bias. Given the short duration of his study, the very small sample size, and the weak correlations, drawing conclusions about possible long term health risks tied to fat consumption in the maintenance phase of any weight control program is not good science. The final sample size was 18, yet they make generalizations to many people. The entire duration of the treatments was 4 weeks, yet they make statements about “long-term maintenance.” Dr. Miller and colleagues may have assigned Atkins, but that is not the same as following Atkins. Red Flag – no decrease in Triglycerides, the participants were obviously not following Atkins.
Jeff Volek, PhD, RD has published data on normal weight men and women following Atkins with little change in body weight for 4-6 weeks that showed dramatic decreases in TG and increases in HDL-C. He has also published extensive data in people who were near weight stability after 12 weeks of Atkins, and the Atkins diet markedly reduced triglycerides, increased HDL-C, decreased inflammatory markers, and decreased total saturated fatty acid content of the blood. He also has data in press showing the Atkins diet improved fasting and postprandial vascular function assessed using the same flow mediated dilation technique used in this study.
There is no one-size fits all weight-loss diet so of course results, depending upon the population participating in any given study will vary.
But the very group Atkins initially sought to help – patients with carbohydrate-related metabolic disorders who process food differently from those with normal carbohydrate metabolisms was not included in this study.
As you stated, it is those of us with diabetes, prediabetes, insulin resistance, etc. that have been shown in valid studies (in fact, in multiple studies) to respond well to true low-carb diets. Cardiovascular risk is decreased, weight loss is easier, diabetes is more readily controlled without medications, and in many cases fertility is often even restored to women who follow low-carb diets.
There are always going to be naysayers but fortunately, there will also be educated non-experts to set the record straight when the “experts” and press report either inaccurate or insignificant findings.
I am alive today because I began following Dr. Atkins plan in 1997. Prior to low-carbing I suffered from angina and had a very poor lipid profile. Now, I do not.
Thank you for debunking another slanted study.
Yet again another very flawed study to back up the bias of the authors. I’m so very tired of ‘news’ like this. This eating plan WORKS for so many of us, keeps us healthy if we are, improves our health if we are not, and the weight loss is a nice side effect.
Hugs,
G
Thanks for the indepth report on
this flawed study.
By the way…please comment on today’s (April 6) AARP online magazine featured Nateralova (spelling)the tennis champ is advocating people eat very very high
carb foods suggesting they will lose
weight. Perhaps if I played a couple
of games of tennis every day i could
eat the way she does! ..Or not!
They were checking to see if heart risk improved during maintenance. Maybe they didnt say that on Atkins it didnt need to improve. On Atkins my markers, trig and hdl vastly improved to a point the testers probably would have said were great. So how in the world or why in the world would I need to improve said markers?
Karen