Dietary Carbohydrate Restriction in The Treatment of Diabetes and Metabolic Syndrome

Richard D. Feinman, PhD, Jeff S. Volek, PhD, RD and Eric C. Westman, MD 

Learning Objective:

After reading this article, the participant should be able to:

1. Understand in broad outline the benefits of carbohydrate restriction for diabetes and metabolic syndrome.

2. Have awareness of the literature studies on low-carbohydrate diets for the treatment of diabetes.

3. Discuss with patients the resources and information available for diets stressing carbohydrate restriction.

Dr. Feinman is Professor of Biochemistry at the State University of New York Downstate Medical Center, Brooklyn, NY 11203; Dr. Volek is Associate Professor of Kinesiology at the University of Connecticut, Storrs, CT  06269-1110; Dr. Westman is Professor of Medicine at Duke University School of Medicine, Durham, NC 27704 ;

The authors have disclosed that they have no significant relationships with or financial interests in any commercial organizations pertaining to this educational activity.


Word count text = 2004

Diabetes is fundamentally a disease of carbohydrate intolerance.  Reduction in dietary carbohydrate, alone or as an adjunct to pharmacology, is thus an intuitive approach to treatment.  The major therapy before the discovery of insulin, low carbohydrate diets are perceived as part of diabetes therapy by many physicians and laymen alike. Since the discovery of insulin, however, the standard diet has emphasized fat reduction and health agencies have specifically discouraged the use of low carbohydrate diets although the American Diabetes Association (ADA) (1) has recently given limited acceptance, at least for weight loss.

It has also been argued that carbohydrate restriction improves all the features of metabolic syndrome and, as such, provides an operational definition (2) as well as a treatment modality, an idea supported by prospective studies(3).

For the practitioner contemplating the use low-carbohydrate diets, there is a need to understand the basic rationale for a course of action different from standard recommendations.  Guidance on implementation is also required, especially for patients already on medication where one has to avoid the risk of hypoglycemia due to the combined effects of drug and diet.


DN, a 56 y/o Caucasian male presented with a 3 year history of type 2 diabetes, coronary artery disease, hyperlipidemia, hypertension, gastroesophageal reflux disease, sleep apnea, and depression. His weight was 131.1 kg, BMI 41.4 kg/m 2.  Blood glucose varied from 130 to 390 mg/dL and hemoglobin A1c was 8.2%.

Medications included Lantus insulin 46 units daily, Byetta 5 mcg twice daily, glipizide 10 mg twice daily, Nexium 40 mg daily, Toprol 100 mg daily, Enalapril 20 mg daily, Cymbalta 40 mg twice daily, Lyrica twice daily.

DN was started on a carbohydrate-restricted diet (< 20 grams/day) and followed at 1-2 weeks intervals to monitor adherence and medications. Because reductions in dietary carbohydrate allow return to more normal glycemic control and reduced insulin fluctuations, insulin and other glucose-lowering  drugs must be reduced in advance. Upon program initiation, DN’s Byetta was discontinued, insulin was reduced to 30 units but glipizide was continued.   After two weeks,  blood glucose was 94 to 172 with several readings below 100 mg/dL and  insulin was discontinued. By the fourth month DN had only been able to reduce his weight by 8 % to 120.6 kg (BMI 38.1 kg/m2), but blood glucose was 111 to 156, (HbA1c =5.7%) and he was able to discontinue all medication.


Several reports in the scientific literature, consistent with the case study, provide support for low-carbohydrate diets as an attractive alternative if not the preferred treatment in diabetic and pre-diabetic states (4) A recent small study has been followed to 44 months (5).

These reports were summarized in a multi-authored paper  that laid out the basic features of low-carbohydrate diets (6).  The major principles emphasized were:

1. Carbohydrate restriction improves glycemic control, the primary target of nutritional therapy and reduces insulin fluctuations.

2. Carbohydrate-restricted diets are at least as effective for weight loss as low-fat diets.

3. Substitution of fat for carbohydrate is generally beneficial for markers for and incidence of CVD. Whereas low-fat diets are generally more effective for reducing LDL, dietary carbohydrate is the strongly correlated with small dense LDL which are considered more atherogenic and carbohydrate restriction is the most effective method of lowering triglycerides and raising HDL.

4. Carbohydrate restriction improves the features of metabolic syndrome.

5. Beneficial effects of carbohydrate restriction do not require weight loss.

Points 2 and 5 are of particular importance in that the 2008 ADA Recommendations and Interventions (7) received some attention for admitting, for the first time, that “for weight loss, either low-carbohydrate or low-fat calorie-restricted diets may be effective in the short term (up to 1 year).”  The emphasis on weight loss rather than glycemic control seems odd to many but given how difficult it is to lose weight by any method, patients and physicians should be encouraged by the fact that benefits accrue to low-carbohydrate diets even if weight reduction is not attained .  Experiments in the literature have demonstrated dramatic improvements in triglycerides and HDL-C in patients with type 2 diabetes under conditions where reduction in body mass were minimal(8).


It is frequently suggested that lifestyle interventions are as successful as pharmacology in treating diabetes, but traditional lifestyle changes stressing high carbohydrate have the potential for increasing the need for drugs.  The 2008 ADA Recommendations contain the telling statement(9):

“Sucrose-containing foods can be substituted for other carbohydrates in the meal plan or, if added to the meal plan, covered with insulin or other glucose lowering medications. Care should be taken to avoid excess energy intake.”

One of the ironies in this recommendation is that insulin therapy is established to pre-dispose to weight gain(10).  Patient apprehension of the possibility of insulin-related weight gain has been cited as a barrier to insulin use(11).

Reduction in medication is sensibly taken as improvement in a medical condition.  The case study above is characteristic of the carbohydrate-restriction interventions and remains one of their strongest benefits(12). At the Duke Lifestyle Clinic, using diets of 20 grams carbohydrate maximum per day (VLCKD; Table 1) it has been possible to have patients taper off as much as 280 units of insulin in 3 weeks.  Because  there is  immediate improvement in glycemic control, patients already on glucose-lowering drugs must have medication reduced or eliminated before a change in diet. Blood pressure also improves, so patients taking antihypertensive medication require monitoring as well.


The concept of a metabolic syndrome (MetS; insulin resistance syndrome) has had great intellectual impact in medicine because it suggests that a broad and seemingly disparate set of conditions of overweight, hypertension, atherogenic dyslipidemia (high triglycerides and small dense-LDL-C and low HDL-C) have a common origin.  The underlying factor is likely insulin resistance, simultaneously a feature of the diabetic or pre-diabetic state.  Recent papers in this journal and others have raised the question as to whether the concept is really useful for the clinician (13). That is, would characterization of multiple markers as a syndrome lead to any different treatment than the sum of the treatments for each individual risk indicator?

The observation that traditional and emerging markers of MetS are precisely those that are targeted by carbohydrate restriction provides evidence for the appropriateness of the designation syndrome and suggests, as well, a unique method of treatment (14).  Thus, there are many ways to treat obesity but none are as effective as carbohydrate restriction at improving triglycerides.  Known since the fifties (15) this is probably the most robust response to any dietary intervention, Similarly, there are several pharmacologic approaches to raising HDL-C or improving hypertension but few target the other markers.  In addition, low-fat diets tend to lower HDL-C and, in any case, seem to require weight loss for beneficial effects whereas carbohydrate restriction does not.  Carbohydrate restriction is also of obvious value in improvement of glycemic and insulin responses which are frequently exacerbated by low-fat diets.

That the collection of markers is improved by a single type of intervention argues for the existence of a common (carbohydrate-sensitive) mechanism and suggests for the physician that treating any one marker by reducing carbohydrate has the potential to prevent the onset of others which may not be evident at the moment. This hypothesis was tested in a prospective study in which forty overweight subjects with atherogenic dyslipidemia were randomly assigned to dietary interventions restricted in fat (low fat diet, LFD) or carbohydrate (carbohydrate-restricted diet, CRD).  Subjects consuming the CRD had improved glycemic control and insulin sensitivity, greater reductions in weight and adiposity and improvements in several inflammatory markers(16).


L.G. is a 48 y/o Caucasian male with MetS. He was taking no medications and was started on a carbohydrate-restricted diet (<20 grams/day) and followed every 1-2 months to reinforce adherence for a 12 month period.  No medication was instituted.  Results were as follows:

Baseline 12 months
serum triglyceride (mg/dL) 473 218
blood glucose (mg/dL) 117 92
waist circumference(cm) 111 99
blood pressure (mm/Hg) 128/81 125/80
body weight(kg) 110.8 98.4
BMI(kg/m2) 31.3 27.8


Implementation of low carbohydrate diets is complicated by the lack of clear definitions.  Many studies show benefits proportional to the reduction in carbohydrate but there also appears to be a threshold effect reflected in the recommendation for very low carbohydrate diets, the so-called induction period of popular diets such as the Atkins Diet.  Table 1 lists definitions which would probably be accepted by most workers in low carbohydrate diets.

Carbohydrate-restricted diets do not generally specify what  the carbohydrate is to be replaced with leading critics to characterize them as high-fat or high-protein.  LaRosa was the first of several researchers to show that in practice, many dieters simply reduce calories by removing carbohydrates without replacement (17).  Protein in the diet generally tends to be relatively stable and the key question is the relative amounts of fat and carbohydrate.  According to the rationale for carbohydrate restriction, where  glucose and insulin are control elements, fat is expected to play a relatively passive role and specific recommendations to reduce fat are unnecessary. Undesirable effects of high fat are seen only under conditions where carbohydrate is moderate to high and, in any case, substitution of carbohydrate for fat is almost always deleterious (18).

A survey of an online support group, the Active-Low Carber Forums showed, perhaps surprisingly, that the major change for dieters was increase in the consumption of non-starchy vegetables (19). Low-carbohydrate diets may in some sense be characterized as high-vegetable diets.  Finally, although critics of low carbohydrate diets point to limited choices, there are now thousands of recipes and strategies on the Internet, the most comprehensive is probably the site at (


Exercise clearly enhances the proportion of fat loss, especially if the exercise involves resistance training. The combination of resistance training and diet has been investigated in two studies which both showed that men and women who consumed a low carbohydrate diet while performing resistance training 3 times per week had the greatest decreases in percent body fat (20).


Side effects and, conversely, the ability to realize therapeutic goals strongly affect adherence to any intervention but compliance with a diet is largely separate from the efficacy of the diet itself and depends on the motivation, external support and the overall features of the patient-physician interaction.  Low-carbohydrate diets attain at least as high a degree of compliance as low fat diets in experimental trials but such experiments may not offer the full potential for patient support.

The internet provides opportunity for support of people attempting to follow a low carbohydrate diet.  The Active-Low Carber Forums ( ) as of September, 2008 had more than 110, 000 members.  The forums and similar support groups specifically targeting people with diabetes offer support and suggestions for implementing dietary plans. The widespread use of glucometers suggests that people with diabetes my have a unique opportunity to participate in their own medical nutrition therapy.  Education and support from physicians will be most successful in patients who are able to discover for themselves which foods or combinations of food will allow good glycemic control.


In treating patients already receiving medication, physicians may need training in the use of low-carbohydrate diets but the literature and case studies are consistent with the intuitive value of such diets.

  1. Dr Feinman congrats for your blog. I consider you a major contributor in the nutritional treatment of the Metabolic Syndrome. I would like to ask a question that puzzles me. In terms of the general public and the medical community involved with this, especially dietitians and nutritionists, is there any awareness that carbohydrates are not an essential nutrient?

    • rdfeinman says:

      This is a good question. With respect to the medical community, I work in a medical center and there is a wide range of understanding of metabolism but a substantial appreciation by many physicians of the benefits of carbohydrate restriction. Most physicians are interested in helping the patient, are not particularly doctrinaire but many clinicians are overworked as it is and are unlikely to want to fight with the nutritional board of the American Diabetes Association who are part of and speak for the most narrow-minded, political segment of the organization. I suspect there is even more confusion among dietitians who are even more constrained by limitations on their licensing although many understand the physiology. And, of course, the general public has a still wider range of views and more confusion.

      On the specific question of there being no requirement for dietary carbohydrate, some take this as an objectionable recommendation for a zero carbohydrate diet which, if it were even possible (even red meat has glycogen), is not recommended by anybody. My current research activity involves collecting and tabulating patient responses to carbohydrate restriction in diabetes because the stories are so compelling. So why is there so much resistance? Why, don’t people understand the benefit of carbohydrate-restriction? Some have not the capacity to ask. Some simply don’t get it. A few are genuinely evil: consider political benefit beyond anything that happens to the patient. In the end, a large number of people do understand but it has not been possible, despite the blogs, despite the forums, to get people together. I had tried do this with Nutrition & Metabolism society and still am trying ( and our face book page). If you are volunteering to organize a meet-up or some kind of group in the UK, I’m at

  2. Thank you for the reply Dr Feinman. I suspect this antithesis of some people is partly to the “fad-diet” stereotype of restricting carbohydrates, and a slippery slope of the fact that carbohydrate is not essential. Another explanation is that restricting carbohydrates is too simple for some people to go under the label of “science based” , I find this in practice too. Whenever I inform clients that carbohydrates are not an essential nutrient the almost immediate response (or thought) is that they think I am suggesting they ban all carbohydrates forever.

    • rdfeinman says:

      My advice (and my own behavior) is to get all or most of your carbohydrate from vegetables. People do have to recognize, as explicitly said in the new Atkins book, that “a fruit is not a vegetable.” Along those lines, one of the popular experts that I like is Suzanne Sommers because she understands the problem (after all, what does Walter Willett know about having a weight problem): she lost a part in Starsky and Hutch because they told her she was “too chunky.” She recommends eating fruit in pieces in the course of the day (reduces insulin surges) and eating that way helps.

  3. Dr Tim says:

    Can you please make the article PDF accessible? Thank you.

  4. ken says:

    How is it I can eat 500-700 grams of carbs a day and yes the so called bad kind,and I lost 45lbs,gained 3 lbs of muscle and all blood work is fine ? and a heart strong enough for 2 people according to my cardiologist ? PS I’m 40yrs old and weigh about 170 lbs about 12% body fat

    • rdfeinman says:

      It’s because you’re a lucky guy. Dietary carbohydrate restriction is primarily therapeutic. Although many people stay on low-carbohydrate diets because they feel better, the dietary approach is primarily for people who don’t react to carbohydrate the way you do. You’re a lucky guy.

  5. So Ken you have 2000 to 2800 calories JUST FROM CARBS (!!!) everyday and still manage to stay at 12% bodyfat , at 40 years of age. Dr Feinman, I think you should do a case study here. If we assume that you get some protein and fat in your diet you could be eating 4000 calories easily. I can give some explanations to that: a) you don’t know how to count carbohydrates. You may count the weight of foods rich in carbs but not the actual energy/carbohydrate content. b) Your count is right and you do 3-5 hours of hard exercise every day which I personally think is highly unlikely. c) A combination of bad counting and 2-3 hours of exercise d) Your home and everything in it is made of DNP and e) You are lying, which doesn’t happen very often in the internet.

    • rdfeinman says:

      I take this as an offer to fund a case study, is that right? Until we are underway on that, we have to assume that Ken is roughly describing his experience. In early phases of low carb diets people describe rather remarkable intake of protein and fat that Ken would not believe. The principle “one size does not fit all” is almost always disingenuous but it is a good principle until you do get hard data from the case study. The check can be made out to the Research Foundation of the State University of New York and mailed to me.

    • Reziac says:

      I’d like to see this guy’s complete thyroid panel. I’d guess he has very good T3 conversion.

  6. Nevertheless, your work is excellent, and should receive more funding. Keep it up!


  7. I enjoy reading your posts. This one in particular I am spreading far and wide in the hope that people will learn how to ‘get well’ and loose weight and generally feel better!

    My question is – are you an advocate that if you ‘do this diet’ and get better, you can go back to living life, just like you did before? I certainly am not of this opinion. I do not believe for example Type 2 DM can be cured. I do however believe it can be put into remission – and stay that way, so long as you don’t regress to your old lifestyle….. YES lifestyle – not diet.

    • rdfeinman says:

      Thank you. I think that very few people can go back to carbs. In fact, from what I see and what I read — I am not a health care provider — there can be many variations and problems. What we know is less than what we don’t know and type 2 DM can be difficult but we do know the big principles. On the other hand, for many people a lifestyle change can be so beneficial as to put the question of a cure into the realm of semantics. In some sense, I definitely feel that my nearsightedness has been cured…if I wear my glasses.

  8. Peter Brown says:

    I found this post very enlightening. I am almost 54 years old. Four years ago I was diagnosed with Type 2 diabetes, high triglycerides, low HDL and moderate LDL but low overall cholesterol. I have exercised all my life, have always eaten well, have mostly eaten good wholemeal grains, rarely eat any junk food and eat lots of vegetables and some fruit. I don”t smoke and rarely drink. I weigh 72 kg and am 6 feet tall. My BMI is about 22. I have tried eating more fish, exercising more, eating low fat and eating no added sugar products. Nothing has helped. My readings have remained constant. Two years ago I became a vegetarian because I no longer enjoyed the taste of meat. My readings have remained constant. The one thing that has remained constant over the four years is that I have been eating a high carb diet composed of good whole grains. My usual breakfast is barley and oats with wheatgerm, psyllium fibre, linseed meal and oat bran. I do not take any medications as of yet. My doctor and diabetes educator just keep telling me I am unlucky and cant work out why I have these problems. I don’t think I am that unlucky. After reading on the internet the last two weeks I have started to eat a lower carb diet. Already I am noticing that when i take my glucose readings each day they are improving. The less carbs I am eating the less hungry I feel. I am eating predominantly good fats from nuts and avocados. I have stopped eating my favourite breakfast. I would like to thank you for this article and post. It confirms and reinforces the things I have been reading and provided me with some new information eg. the link between carbs and triglycerides (my doctor said they would improve when I ate fish but they didn’t most probably because my carbs stayed the same). I now have an APP on my Android phone which lets me know the carbohydrate, fat and protein profile of each food I eat. Once again thank you and keep up the great work. Pete from Perth, Australia.

  9. Dr Feinman, I am surprised I have not found this blog before, I did a lot of research in the past 3 years on the metabolic syndrome and diabetes. I first read your text in 2015 (Dietary carbohydrate restriction as the first approach in diabetes management: Critical review and evidence base), a few months after cutting all starchy foods from my diet…the subsequent change in my health, and the explanations from your text made me understand that I was probably a T2DM patient and did not know it. I thank you for your research and work, and I have quoted you numerous times, referring people to your text. Here in Quebec, T2DM is still badly understood..sadly, here also, the new approach is in a “different school of thought”…where carbohydrates restriction is not a prime…what a disaster.
    Anyway, my question for now, if you feel the interest in answering, is concerning the Hemoglobin A1c marker. Now, from your paper, you indicated that most markers, outside of HbA1c usually correct after a few weeks, maximum a month. What about HbA1c.
    The reason I ask is because I had not seen a doctor for 12 years, when suddenly, in 2012 I got an anaphylactoid shock. In the following, I tried to communicate with the profession some of the other symptoms I had, but nobody was listening, so I was left on my own. I was told the shock was probably allergies, was given an epipen, and that is it. I was obese at the time, BMI 31.
    I tried all kinds of diet to address my health condition (symptoms) afterwards. I eventually cut all flours from my diet, and then progressively, as I was learnig about food compositions, and diet, all starchy foods. The result is that I lost 65 pounds (BMI down to 22.1) in 3 months in a half with no effort at all. But in July 2015, I got an acute pericarditis (without complications), which got me to the hospital. At the emergency, I tried to explain to them that I believed I might have been T2DM, but was in some kind of remission…I made them laugh a little bit, but that is it. This past week, I went to get the blood work that was done in 2015, at the time of the pericarditis, and it showed that all markers were on line in a general blood work, safe for HbA1c which was at 6.7%. This is not a lot higher than the upper limit, but could it be an indication ( since it had been four months that I started changing my diet to very low carb) and had lost so much weight, could it be the sign that effectively I had been T2DM.
    Thank you.
    Denys Picard
    Canadian and US Citizen
    PS: I went on my diet on March 5th 2015, read your text on June 14 2015 and got my blood work on July 4th 2015. Further, parralel to this, I am stuck with neuropathies (known complications of T2DM, numbness leg and feet, carpel tunnel syndrome, cervical arthrosis (which has recently been correlated with T2DM in 2 studies), and just this week, diagnosed with very mild microbleeds (Also associated with Diabetes Type 1 and 2 (possible leukoaraiosis, with difficulty swallowing)).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s