In the ongoing nutritional war between adherents of low-fat and low-carb diets, the skeptical cardiologist has generally weighed in on the side of lower carbs for weight loss and cardiovascular health.
I’ve questioned the vilification of saturated fat and emphasized the dangers of added sugar. I’ve even dabbled in nutritional ketosis.
The science in nutrition is gradually advancing and the DIETFITS study recently published in JAMA is a welcome addition.
DIETFITS is a really well done study which provides important insights into three huge questions about optimal diet:
- Should we choose a low-fat or a low-carb diet for weight loss and cardiovascular health?
- Do baseline insulin dynamics predict who will respond to low-fat versus low-carb diet?
- Can we predict who will respond to low-fat versus low-carb by genetic testing?
The Details Of DIETFITS
Stanford investigators recruited 609 San Francisco area individuals between the ages of 18 to 50 years with BMI of 28 to 40 and randomized them to a “healthy” low-fat diet or a “healthy” low-carb diet.
During the first 8 weeks of the study, low-fat participants were instructed to reduce fat consumption to <20 gm/ day while the low carb participants were instructed to reduce digestible carbohydrate to <20 gms/day.
Then individuals were allowed to add back fats or carbs 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.” Importantly no explicit instructions for energy restriction were given.
The “healthy” instructions for both groups were as follows
- maximize vegetable intake
- minimize intake of added sugar, refined flours and trans-fats
- focus on whole foods that are minimally processed, nutrient dense and prepared at home whenever possible
Dietfits Outcomes-Diet And Weight
- Total energy intake was reduced by 500-600 kcal/d for both groups
- The low-fat vs the low-carb intake at 12 months was 48% versus 30% for carbs, 29 vs 43% for fat and 21 vs 23% for protein.
- Mean 12 months weight change was -5.3 kg for low-fat vs 6-6.0 kg for low-carb which was not significantly different
- There was no difference between groups in body fat percentage or waist circumference
- Both diets improved lipid profiles and lowered blood pressure, insulin and glucose levels
- LDL (bad cholesterol) declined more in the low-fat group whereas HDL (good cholesterol) increased more and triglycerides declined more in the low-carb group.
Thus both diets were successful for weight loss and both improved risk markers for cardiovascular disease after a year.
DIETFITS- Can Genes and Insulin resistance Predict Best Diet?
Surprisingly, the study found no significant diet-genotype interaction and no diet-insulin secretion interaction with weight loss.
This means that they could not predict (as many believed based on earlier studies) who will benefit from a low carb diet based on either currently available genetic testing or a generally accepted measure of insulin resistance.
As the authors point out, these findings “highlight the importance of conducting large, appropriately powered trials such as DIETFITS for validating early exploratory analyses.”
As you can imagine this study has led to quite an uproar and backlash from dedicated combatants in the macronutrient wars.
A reasoned summary and response from Andreas Eenfeldt, a low carb proponent can be found on his excellent low carb/keto Diet Doctor site here.
If I’m allowed to speculate, the reason that we did not see any major additional benefit from low carb in this study is that the groups ended up so similar when it came to bad carbs. The low-fat group ended up eating fewer carbs too (!) and significantly less sugar, while the low-carb group ended with a somewhat weak low-carb diet, reporting 130 grams of carbs per day.
Eenfeldt emphasizes that low-fat diets never “win” these macronutrient dietary skirmishes:
On the whole, this study adds to the 57 earlier studies (RCTs) comparing low carb and low fat for weight loss.
From a standing of 29 wins for low carb, zero for low fat and 28 draws, we now have 29 wins for low carb and 29 draws. The wins for low fat stay at zero.
Larry Husten at Cardiobrief.org in his analysis of the study quotes a number of experts including Gary Taubes, the low carb pioneering journalist
Taubes speculates “that the weight loss may have been similar not because any diet works if you stick with it and cut calories (one possible interpretation) but because of what these diets had in common — avoid sugar, refined grains, processed foods. Whether the low-carb arm would have done even better had Gardner kept their carbohydrates low is something this study can’t say. (And Ornish [low-fat diet proponent] would probably say the same thing about fat consumption.)”
The low-fat or vegan disciples seem to have had a muted response to this study. I can’t find anything from John McDougal , Dean Ornish, Caldwell Esselstyn or Joel Fuhrman.
Readers feel free to leave comments which link to relevant analysis from the low-fat proponents.
Dietfits-Perspective Of The Participants
Julia Volluz at Vox wrote a fascinating piece recently which involved interviewing some of the participants in this study.
She points out that although the average DIETFITS participant lost over 10 pounds, “Some people lost more than 60 pounds, and others gained more than 20 during the year.”
She obtained permission from the lead author, Christopher Gardner and interviewed “Dawn, Denis, Elizabeth*, and Todd — two low-fat dieters and two low-carb dieters — about their experiences of succeeding or faltering in trying to slim down”
I highly recommend reading the entire article for details but Volluz concludes
And that leads us to one of the burning mysteries of diets: how to explain why some people fail where others succeed — or the extreme variation in responses. Right now, science doesn’t have compelling answers, but the unifying theme from the four study participants should be instructive: The particulars of their diets — how many carbs or how much fat they were eating — were almost afterthoughts. Instead, it was their jobs, life circumstances, and where they lived that nudged them toward better health or crashing.
DIETFITS-Importance of “Healthy” Diet
Most likely the success of both of these diets is due to the instruction that both groups received on following a “healthy” diet. This guidance is remarkably similar to what I advocate and is something that combatants in the diet wars ranging from paleo to vegan can agree on.
The JAMA paper only provides the description I listed above but Volluz adds that participants were instructed to:
… focus on whole, real foods that were mostly prepared at home when possible, and specifically included as many vegetables as possible, every day … choose lean grass-fed and pasture-raised animal foods as well as sustainable fish ... eliminate, as much as possible, processed food products, including those with added sugars, refined white flour products, or trans-fats … prepare as much of their own food as possible. …
Indeed, if you want to see a very detailed description of the instructional process for participants check out the very detailed description of the methods here.
Yours in Health,
N.B. I was searching for a reasoned response to this study from the low fat camp and to my surprise came across this fascinating video featuring the lead author of the study, Christopher Gardner, on (no fat/vegan) John McDougal’s YouTube site. Gardner is clearly on the side of sustainable, local , ethical food consumption but to his credit, his research , publications and comments on DIETFITS don’t reveal this.
7 thoughts on “Low-Fat Versus Low-Carb Diet: DIETFITS Show Both Can Work If They Are “Healthy””
I found Dr Mcdougall dietitian Jeff Novick’s short response on their forum I will copy here as it appears a bit down the page>
“Here you go…
As I guessed… 🙂
In the end, it will be the calories (and adherence) that mattered most.
Of course, as we know, how you get those calories (limiting total food vs adjusting calorie density of the food) is the real issue for longterm compliance.
A few more quick points
Point 1: Neither diet matched the true diets advocated by the proponents of either type of diet (low-fat or low-carb)
In regard to low fat, most all popular low fat diets from Pritikin, Ornish, McDougall, Esselstyn, Barnard, aim for about </= 15% at most. In one of McDougall’s studies, they patients maintained 15% for the full year. The starting diet clearly represented the SAD diet, with about 35% fat, 45% carb and 17% Pro. But the low fat diet at its lowest was 24% fat, was 26% by 6 months and was 29% by a month.
Same with low carb. Most advocates want Carb much lower that the study had, usually less than 20% if not lower. They started at 23%, then went to 27 by 6 months and 30% by 1 year.
Point 2: Regression to the mean
The study highlights what happens in real world application which is that diets are hard to follow and as time passes they “regress to their former mean” and results fade. Over time, we see a clear direction of this regression with less and less success. If they kept their regression pace up for 18 months to 2 years they would both be very close to the starting diet. Or again they would have just quit from lack of success and gone back to their original diet.
Point 3: Diet Quality vs Calorie Density
The authors, and several of those who were interviewed for articles, discuss emphasizing diet quality over quantity or counting calories. However, "quality" is a generic widely misused term like "natural, organic, and whole" that doesn't have any real meaning and in the end won't help. What we need is to focus on understanding calorie density, satiety, intact fiber/cal, avoiding liquid calories, etc. The author of the study said to focus on
– More whole foods
– More vegetables specifically
– Less added sugar
– Less refined grains
Unfortunately, as I have described here, whole foods has no meaning and is another misused term. What we want is intact and/or minimally processed foods. More vegetables is good as is less sugar and less refined grains. But is leaves off other calorie dense foods including the most calorie dense refined food of all, oil. So, if we were to improve these we would say…
– Increase intact and minimally processed foods
– Make 50% of your plate non-starchy vegetables
– Minimize added sugar (and salt, and oil)
– Minimize refined grains.
Let's compare that to my 5 Pillars of Healthy Eating
1) Plant-Centered – Center your plate and your diet predominately around plant foods (fruits, vegetables, starchy vegetables, roots/tubers, intact whole grains, and legumes (beans, peas & lentils).
2) Minimally Processed – Enjoy foods as close to "as grown in nature" with minimal processing that does not detract from the nutritional value &/or add in any harmful components.
3) Calorie Dilute – Follow the principles of calorie density choosing foods that are calorie adequate, satiating and nutrient sufficient.
4) Low S-O-S – Avoid/minimize the use of added Salts/sodium, Oils/Fats and Sugars/sweeteners
5) Variety – Consume a variety of foods in each of the recommended food groups
Someday the Diet Wars will hopefully end, once everyone figures out that clean fresh food IS the answer, and that genetically we all have slight differences that enable each individual to fare well on grains, or dairy or different types of animal proteins in addition to fresh produce. It’s never going to be a right and wrong, clean fresh produce across the board will always do well for everyone, with what works for each individual in addition to that clean fresh produce being what it is. Someday we will stop spending money on all these arguments and plow it into what matters- clean healthy foods for eveyone. Your body will tell you clearly what does and doesn’t work for YOU. Learn to listen to it, and respect it, and we can put all the over the counter meds and big pharma in the rubbish. Now THAT would be a great leap forward for ALL of us!
I skimmed the actual study linked in your article. Noted that participants had to agree to being randomly assigned to low fat or low carb. To me, this suggests most were relatively indifferent as to which they had to cut back on to lose weight. This suggests one reason outcomes were similar. I’d guess, in general, the most successful dieters cut back on those ‘bad’ foods they are most willing to – fats for some folks, carbs for others, and a mix for others.
Yes. This was a randomized study. That aspect is part of the strength of the study.
Those who responded had to be willing to accept the random assignment to either low fat or low carb, thus, accepting of either.
Given equal outcomes , I view the success of the diet as due to the common instruction on “healthy” eating in general.
We have been on Paleo – it worked well, for a while – and was easy to maintain – for diabetes. But the winner for us is McDougal’s ‘Starch Solution’. We eat starch. Every meal, starch is the main course with everything else as the ‘side’. I started beginning of December 2017, husband started January 2018. I’ve lost 9kg and sugar is normal. Husband has lost 10 kg and sugar is normal. For diabetics who couldn’t even get that on Paleo it is amazing. It works for us. I posted the same anecdote on Dr Kendrick’s blog and got shot down by anti vegan fundamentalists. So I ask commentators to leave me alone, on this blog. The only adjustment we do to the Starch Solution, is to eat fish once a week and (sometimes) two eggs in an omelette. I am not promoting anything – I’m just saying what is working for US. Thanks for your interesting blog.
I very much appreciate your site and look forward to each post. Invariably they lead me to other articles and research and add to my understanding of the complex issue of what foods are best for good health.
Excellent! Thank you for this summary. The over-simplified question of which is better is so contentious. Your thoughts are very helpful and unifying.