Diet study: American Diabetes Association vs. Low Carb Ketogenic

A randomized pilot trial of a moderate carbohydrate diet compared to a very low carbohydrate diet in overweight or obese individuals with type 2 diabetes mellitus or prediabetes (Saslow et al., 2014)

Disclaimer: this study was not ground-breaking; it was confirmation of a phenomenon that is starting to become well-known, and soon to be the status quo. That is, advising an obese diabetic patient to reduce their carb intake consistently produces better results than advising them to follow a low fat, calorie restricted diet.

The two diets:

Moderate carbohydrate diet: 45-50% carbs; 45 grams per meal + three 15 gram snacks = 165 grams per day; low fat, calorie restricted (500 Calorie deficit).  Otherwise known as a “low fat diet (LFD).”

In their words: “Active Comparator: American Diabetes Association Diet.  Participants in the American Diabetes Association (ADA) diet group will receive standard ADA advice. The diet includes high-fiber foods (such as vegetables, fruits, whole grains, and legumes), low-fat dairy products, fresh fish, and foods low in saturated fat.

Very low carbohydrate diet: Ketogenic; <50 grams of carb per day, no calorie restriction, just a goal of blood ketones 0.5 – 3 mM.

In their words: “Experimental: Low Carbohydrate Diet.  Participants will be instructed to follow a low carbohydrate diet: carbohydrate intake 10-50 grams a day not including fiber. Foods permitted include: meats, poultry, fish, eggs, cheese, cream, some nuts and seeds, green leafy vegetables, and most other non-starchy vegetables. Because most individuals self-limit caloric intake, no calorie restriction will be recommended.

Both groups were advised to maintain their usual protein intake.


Food intake.  Those assigned to LFD went above and beyond the call of duty and reduced Calories by 792.  And as expected, those assigned to keto spontaneously reduced intake, by 697 Calories (for the record, that’s a little more than usual).  Both groups came pretty close to meeting the other recommendations – carbs on LFD were 139 grams (40% of calories); and 58 grams for the ketogenic dieters.  Protein intake declined in the LFD group, which presents a potential confounder as the final differences were pretty big: 106 vs. 68 grams per day (25 vs. 20% of calories).

Weight loss: Despite eating fewer calories and undergoing a greater reduction in calorie intake, those assigned to LFD lost less weight than those assigned to the ketogenic diet (5.7 vs. 12.1 lbs).

food intake data

This wasn’t a metabolic ward study, so dietary information was self-reported via 24-hour dietary recalls.  However, most of the ketogenic dieters achieved blood ketones greater than 0.3 mM, which is difficult to do if they were under-reporting carbohydrates (which was their only main instruction). 

Body composition: It wasn’t assessed in this study, but more protein and calories in the ketogenic diet strongly suggests better preservation of lean mass… in general, less dietary protein is required to maintain nitrogen balance as total calorie intake increases; the ketogenic diet in this study was higher in both protein and calories.  

Other notable changes: CRP, a marker of inflammation, followed body weight.  That is, it declined more in the ketogenic dieters.  Fasting insulin and glucose also improved more in this group.


just a reminder: the MCCR diet (aka “low fat diet”) is recommended by the American Diabetes Association for the dietary treatment of diabetes.  And it’s losing.

HbA1c figures


The most important finding of this study, and the one people should be talking about (in my opinion): 7 out of 11 keto dieters (64%) were able to reduce their anti-diabetic medications, whereas only 2 out of 13 low fat dieters did (15%).


What’s most compelling about this finding is: despite reducing their anti-diabetic medications (which should impair glucose homeostasis), markers of blood glucose control still improved in ketogenic dieters.

This isn’t an isolated finding:

Nutrition Disinformation, Part I. (Mediterranean Diet Fail)

How to define a healthy diet. Period. (numerous examples)

Nutrition Disinformation 2.0.  (Look AHEAD Study Fail)

Nutrition Disinformation III (William Yancy gets it right!)

The common thread in each of those posts: carbohydrate restriction is numero uno for reducing the need for prescription medications in obese and diabetic patient populations.



Comprehensive measures of well-being, affect, hunger, and appetite were also reported, I suspect, because the study was partly funded by the National Institute of Mental Health.  And interestingly, almost everything got better, even in the keto group.  Perhaps this is due to the slow rate of weight loss (~4 lbs/month), reduced hypoglycemic events, or fewer drug-related side effects… then again, interpreting mental health questionnaires isn’t my strongest suit.


Study weakness: small sample size. This is not a statistical weakness, as they had enough power to detect significant differences between the groups; rather, having only a dozen subjects per group doesn’t leave very much room for heterogeneity, individual variability, etc.

Food intake data:

everybody lies

It looks like everyone under-reported, to a degree.  That said, for what it’s worth, the 24-hour dietary recall isn’t the worst way to assess food intake.  The textbook advantage of this method is that it’s usually very accurate – all of the diet over the past 24 hours is reported, even the unhealthy parts (this is unique to the 24-hour recall; people tell you about the Twinkie they ate yesterday, not the one they ate last week).  And it’s far more comprehensive than: what did you eat yesterday, dude?

The textbook disadvantage is that it’s only 1 day, which may not fully reflect the usual routine.  So it needs to be repeated, which I don’t think happened in this study… but the increase in blood ketones and reduction in triacylglycerols pretty much confirms that the ketogenic dieters were following instructions.  And as discussed below, the changes in body weight are also in line for what we would expect if the low fat dieters were following instructions.

Duration: 3 months isn’t too short.  This is average, and may even be on the longer side.  No, it’s not 5 years-to-life, which is what detractors will say; but in reality, 3 months is pretty good for a diet study.  And sure, weight re-gain is going to happen, it nearly always does… the researchers plan on doing 6 and 12-month follow-ups.

Critics of most diet and exercise studies routinely cite small sample sizes, short duration, and sub-par assessment tools.  But these things are expensive.  If the questions you want answered require big numbers, long duration, and extremely accurate tools, either ask better questions or be smarter.  

Besides the results about medication usage, another interesting finding was the pseudo-spontaneous reduction in calorie intake in the low carb group.  It worked in an ‘ad libitum’ setting, which suggests it can work in real life.  I say “pseudo-spontaneous” because they actually, mindfully, reduced carbohydrate intake, but these calories weren’t fully compensated for, leading to an energy deficit.  The LFD group actually restricted more calories than they were asked, but they lost less weight.  So they were either more severely under-reporting food intake or energy expenditure declined to a greater extent.  Either way, it doesn’t bode well.

And lastly, as mentioned above, these finding are not surprising… the authors of this study may have unwittingly “stacked the deck,” in a sense.  They specifically included obese, insulin resistant subjects, and Chris Gardner has clearly shown that this characteristic strongly defines who will lose more weight on low carb diets.  That is, if the researchers recruited both insulin resistant and [could find enough] insulin sensitive obese participants, and [oddly] assigned the resistant to low carb and the sensitive to low fat, both groups may have been more equally successful.  just sayin’


calories proper


Become a Patron!



Be Sociable, Share!
  • “We thus believe the dietary intake data are likely to reflect in part what participants understood we wanted to hear.”

  • “A non-calorie-restricted low-carbohydrate diet is effective as an alternative therapy for patients with type 2 diabetes.”

    • Jack Kruse

      Who has been vilified for saying this for 4 yrs now? LOL CALORIES DONT MATTER. Electrons do.

  • charles grashow

    What about the studies showing a high carbohydrate macrobiotic diet that’s successful in taking T2D off insulin?

    • Have they tested that diet against a low carb diet?

      I found this study, but there was no comparison group:

      • charles grashow
        • thanks for the links. It seems like a relatively healthy, micronutrient-dense, whole foods-based diet which is probably a *significant* improvement on their previous diet.

      • Galina L.

        They involved specialists in the preparation of a microbiotic food and to teach people to properly do the same because the glysemic responce on carbs was so important , from the links Charles provided – “Carbohydrate intake is the first determinant of the post-prandial glycemic response; however, this response displays great individual variability. This variability is thought to be related to several factors, including the type of carbohydrate or starch (amylose versus amylopectin), the food preparation methods (cooking procedures, heating); the fasting time, the pre-prandial glucose level, the distribution of macronutrients in the diet and the individual doses of insulin and resistance levels [10]. Moreover, some studies suggest that the current standard definition of macronutrients fails to capture important information [11].”

        From the logical perspective, low-carbohydrate diet provides good control of blood glucose and doesn’t require specialists to be involved. May be in the society where food shortages are the fact of everyday life like it is in Cuba, macrobiotic diet has a special importance, but in US it is more practical to skip on carbs.

  • Pingback: Diet study: American Diabetes Association vs. Low Carb Ketogenic()

  • Wenchypoo

    Everybody lies, and it’s NEVER lupus! 🙂

  • Wenchypoo

    Yes, carb removal from a diabetic’s diet should be #1, but (tongue in cheek here) “we don’t want to rock the political boat and mess up the subsidy scheme we have going with grain farmers, now do we?”

    Government rigs commodity markets, pays farmers for votes, farmers vote and make political contributions, and we get the metabolic shaft.

  • 165g/day carbs is still relatively low carb – just forced that way by restriction of overall calories, a miserable existence for sure.

    I just love the Aussie diabetes council’s advice for how to eat (and maintain a disease state):

    • ” a miserable existence for sure”

      Many of those mental health indicators suggested it may not have been so bad! …or maybe just ‘better than baseline?’ I don’t know!

      • johnnyv

        It was a diet though not maintenance which pushes up the relative ratio of fat due to stored fat usage.
        Yes a high carb baseline I recall being rather bad comparatively, well I looked bad with translated to feeling bad 🙂

    • I dont understand. it’s like these folks are insane…how can you be so blind when there’s so much evidence (very compelling) that low fat makes things worse? if they keep doing this, my only consideration is that governments want to kill us on purpose…

      • I’m reminded of this from a few years ago:

        “Someone at some point in government has done their sums: Man A lives on the govt approved diet, he works all his life and drops dead at 60 of a heart attack (or spends the last 10 years off work due to diabetes related illnesses). Max cost to govt: 10 years healthcare & benefits. Man B lives on low carb diet, works until he’s 65 and then claims pension until he dies at 100. Cost to govt: 35 years of pension payments, plus care home fees. Govt answer: Let them eat cake. (Make sure cake is v. cheap)”

        • yup! quite comprehensive and SAD

      • Noname

        Things usually make sense. People are usually not bad. It may be that there are some folks who do not do well with fat. By recommending they eat LFHC they wanted to save them. But the rest of the population by going LFHC may have been the road to metabolic disorder, syndrome and eventually diabetes. Its a narrative that makes me feel all good inside, no idea if a word of it is true. No doubt in my mind some folks shouldn’t eat saturated fat. We have the DNA machines, we no longer have precondition issues in medicine, just need no preconditions in insurance and I don’t see why we just don’t start decoding people, diets, etc. This is analytically tractable.

  • Just another chip in the carbohydrate house of cards. Thanks for putting the study on the radar!

  • Joe

    No one disputes the fact that putting diabetics on a low-carb diet is going to improve insulin sensitivity and help them get off meds…The real question is will a low carb diet be good for them long-term? Without carbs, calories must come from animal protein and fat. The former in amounts greater than 2-3 ounces a day has been shown to, net, leach calcium from bones, and the latter degrades the liver. Not a good strategy long-term. Don’t be swayed by the low-carb craziness. The problem is that not all carbohydrates are the same. Whole grains are extremely effective at reversing type 2 diabetes, as are sweet potatoes, potatoes, beans, etc. Refined carbohydrates aggravate diabetes and contribute to heart disease…

    • Michael Trumper

      Can you share your sources. I haven’t seen any credible research that supports this finding, but I am open to reading new sources.

      • “Not a good strategy long-term.”

        there really aren’t any good sources for or against this claim, depending on how long is “long-term.”

    • Bill! The retards/vegans who think epidemiology is science have found your blog…

    • “The problem is that not all carbohydrates are the same.”

      I agree. Same goes for proteins (albeit to a lesser extent) and fats!

    • Thomas Hemming Larsen

      Even if your main source of calories is carbs you’ll still have to eat some protein and I don’t think 2-3 ounces is more than what most people on a high carb diet consume. Besides, I don’t think that people on a high fat diet need to consume more protein. I know some people are using it as an excuse to consume a lot of bacon but I think that idea is flawed. Anecdotally, I would say that I eat less protein on a high fat diet compared to when I was on a high carb diet.

    • Rakel

      I’ve been eating a low carb diet (around 50g a day, sometimes a little more but usually under 100g) for the last 3-4 months. I was diagnosed with diabetes and starting showing a non-alcoholic fatty liver about a year ago. The carbs I’m eating mainly come from vegetables. I do occasionally eat sweet potatoes. I eat meat, dairy and eggs, very little fruit. Some fast food/eating out, but I also make a lot of food from scratch. I do use some sugar substitutes such as stevia and erythirol.

      I just had new labwork done last week and I have reversed my diabetes (a1c and blood sugars now in prediabetic range), all my lipid values have improved (or are completely in normal ranges), and all my liver functions are normal. There’s no doubt in my mind that this diet is healthy for me. I tried eating more carbs and limiting sugar and fat, and only had very modest improvements. In the last 4 months I’ve improved more than the 8 months before that.

      If I kept doing what I was doing before I started lchf, I’m sure over time it would have improved more, but ultimately this way of eating is better for ME. But I also am very obese and lived with insulin resistance and PCOS all of my adult life. I am certainly more sensitive to carbs and sugars than most people. In fact I can eat the exact same thing as a thin person, they don’t gain and I gain like crazy.

  • Jody Rew

    With consideration to the article, I think everyone should check out Paleo Diet and give it some thought because it also incorporates a low carb diet. I’ve written a book about it and you can check it out in here for more in depth understanding.

    • Hi Jody, thanks for the link… we’re no stranger to the Paleo diet 😉

  • Pingback: Low-fat Diets Not Effective: How Paleo, Vegan, Vegetarian & Atkins Can Win At Weight Loss With New Low-Carb Data | Fito News()

  • Pingback: Low-fat Diets Not Effective: How Paleo, Vegan, Vegetarian & Atkins Can Win At Weight Loss With New Low-Carb Data - News Fetch()

  • Pingback: Low-fat Diets Not Effective: How Paleo, Vegan, Vegetarian & Atkins Can Win At … | HealthCookware()

  • Pingback: Low-fat Diets Not Effective: How Paleo, Vegan, Vegetarian & Atkins Can Win At … | FixingApples()

  • Pingback: Low-fat Diets Not Effective: How Paleo, Vegan, Vegetarian & Atkins Can Win At …()

  • Pingback: What The Health: A Wolf’s Eye Review – Natasha's Kirkwood Healthy Living Blog()

  • Pingback: What The Health: A Wolf’s Eye Review – Gym Buddies()

  • Pingback: What The Health: A Wolf’s Eye Review | My Well Care Blog()