Tuesday, November 07, 2006

Can Diabetics Have Normal Blood Sugars with Diet Alone?

Yesterday Dr. Mike Eades, in his post Saturated Fat info at ASBP, mentioned the presentation given by Dr. Eric Westman - Low-GI, Low-Carbohydrate Diets in Diabetes - at the American Society of Bariatric Physicians meeting last week. I was unable to attend the meeting, but Dr. Westman provided me a copy of the presentation since it's a topic I'd like to discuss here today.

Early in his presentation, Dr. Westman highlights the American Diabetes Association (ADA) goals of Medical Nutrition Therapy:
  • Attain and maintain optimal metabolic outcomes including:
  • Blood glucose levels in the normal range or as close to normal as safely possible to prevent or reduce the risk for complications of diabetes
  • A lipid and lipoprotein profile that reduces the risk for macrovascular disease
  • Blood pressure levels that reduce the risk for vascular disease
  • Prevent and treat the chronic complications of diabetes
  • Improve health through healthy food choices and physical activity
  • Address individual nutritional needs taking into consideration personal and cultural preferences and lifestyle

This provided a nice segway into the comparison of dietary trial outcomes where carbohydrate included was measured by glycemic index and/or load or total carbohydrate in the diet.

Beginning with evidence dating back to 1982, he highlighted that carbohydrate directly leads to an increase in post-prandial glucose and insulin response; included recent data showing glycemic index is a major factor, but with more fat included the response is reduced in meals; and then followed-up with data from 2003 that clearly found glucose and insulin response was most favorable when subjects consumed a low-carbohydrate diet.

One of the most interesting slides was the one that went way back - to the 1920's - to review how diabetes was treated with diet alone before insulin! Citing three different publications, he showed that the dietary recommendation back then was clearly low-carbohydrate; with the recommendation for someone weighing 60kg (132-pounds) being 1795-calories with 10g carbohydrate (40 calories), 75g protein (300 calories), 150g of fat (1350 calories) and 15g of alcohol (105 calories). Imagine that, a low-carb diet was the treatment way back before we had any pharmaceuticals to offer.

Fast forward to recently published data and he presents a dozen studies that convincingly make the case that it isn't just the quality of the carbohydrate (GI-GL), but the reduction in total carbohydrate in the diet that offers the most significant improvements for those with diabetes.

Two case studies, in fact, are compelling because they not only provide data showing significant improvements, they show subjects HbA1c reduced to below 6% - into the normal range! [While below 6% is "normal," later this month I'll include an article about what levels are optimal to reduce the risk of cardiovascular disease and why]

The first highlighted in the presentation was published in December 2003 in journal Metabolic Syndrome and Related Disorders, The Effects of a Low-Carbohydrate Regimen on Glycemic Control and Serum Lipids in Diabetes Mellitus. The case study involved a "chart review...of 30 patients who self-reported the consumption of 30 g of carbohydrate daily, followed a strict insulin regimen, monitored blood glucose levels at least four times daily, and had follow-up clinical visits or phone calls with their physician. For both type I and type II diabetics, there were significant improvements in glycemic control and mean fasting lipid profiles at follow-up."

How significant?

HbA1c levels dropped, over an average of 21.4 months, from 7.9 to 5.7 - much lower than the ADA target of 7 and the International Diabetes Federation (IDF) target of 6.5.

Additionally, significant improvements were reported for:

  • weight (average 5kg weight loss)
  • LDL (decreased from 155.4 to 129.7)
  • Triglycerides (decreased from 106.8 to 73.6)
  • HDL (increased from 50.4 to 73.6) and
  • Total Cholesterol-to-HDL ratio (decreased from 4.99 to 3.42).

Non-significant changes included a reduction in total cholesterol from 229 to 222, and a reduction in the use of insulin from 32 units to 25 units on average.

As the researchers noted, "A carbohydrate-restricted regimen improved glycemic control and lipid profiles in selected motivated patients."

In the he second case study, Clinical Experience of a Carbohydrate-Restricted Diet: Effect on Diabetes Mellitus, published in the September 2003 issue of the same journal, researchers undertook a a similar review of patient charts over a period of 18-months, and again, significant changes were found for triglycerides and HbA1c.

In this study, HbA1c was reduced from 10 to 5.9 over the 18-months of follow-up - again within the normal range and well below the targets set by the ADA and the IDF.

Most compelling were some of the unpublished data from Dr. Mary Vernon's practice that was presented to show the dramatic effect reducing carbohydrates in the diet has on HbA1c - the improvement is not only almost immediate, it is striking how quickly HbA1c falls below the ADA and IDF targets!

  • Male, 50, 26-months low-carb, HbA1c 7.0 ---> 5.3
  • Female, 58, 8-months low-carb, HbA1c 6.4 ---> 5.5
  • Female, 49, 12-months low-carb, HbA1c 6.0 ---> 5.1
  • Female, 39, 3-months low-carb, HbA1c 16.8 ---> 5.3
  • Male, 44, 4-months low-carb, HbA1c 8.7 ---> 4.8
  • Female, 69, 5-months low-carb, HbA1c 8.1 ---> 5.4
  • Female, 33, 15-months low-carb, HbA1c 10.9 ---> 4.8
  • Male, 50, 26-months low-carb, HbA1c 9.0 ---> 5.3
  • Female, 36, 18-months low-carb, HbA1c 9.2 ---> 5.5

Now to be fair and include all the relevant data, it must be said that the dietary approach does not reduce HbA1c below the targets for everyone. Of the thirteen profiles highlighted, four patients did have significant improvements, but did not see an improvement to below 6.0 and it is noteworthy that these four individuals had the highest HbA1c levels of the subjects included:

  • Male, 59, 16-months low-carb, HbA1c 12.0 ---> 7.4
  • Female, 49, 12-months low-carb, HbA1c 12.5 ---> 7.5
  • Male, 56, 2-months low-carb, HbA1c 12.0 ---> 6.8
  • Female, 35, 3-months low-carb, HbA1c 11.3 ---> 6.3

In comparing low-carbohydrate trials to conventional studies, the data is powerful, especially for reduction of HbA1c:

  • DCCT trail, conventional approach, HbA1c to 8.9
  • DCCT trial, intensive treatment, HbA1c to 7.1
  • Hays, 100g carbohydrate daily, HbA1c to 6.9
  • Bernstein, 30g carbohydrate daily, HbA1c to 5.7
  • Vernon, 20g carbohydrate daily, HbA1c to 5.9
  • Yancy, 42g carbohydrate daily, HbA1c to 6.3 (most significant to 5.8 in those who discontinued medication and attained glycemic control by diet alone)

At the end of the day, Dr. Westman nailed it - a low glycemic index diet may be politically correct, but it's the low-carbohydrate diet that offers the best glycemic control and improvements in data from studies of subjects with diabetes.

As he noted, low-carbohydrate diets are low glycemic index diets. The real difference is that a low glycemic index diet may still include up to 50% of calories from carbohydrate whereas a low-carbohydrate diet specifically limits the consumption of total carbohydrates in a day with careful attention to not only the total carbohydrate, but the quality of the source of carbohydrate.

As the case studies presented clearly show, attaining glycemic control is possible when the diet is modified to restrict carbohydrate intake. Significant improvements are found in not only the HbA1c levels, but also triglycerides, HDL, LDL and weight.

It's November and it's National Diabetes Month - isn't it time we started to ask why the ADA refuses to even offer patients the option to try a low-carbohydrate diet to control their blood sugars?


  1. This is exactly the approach recommended by Dr. Richard Bernstein in The Diabetes Solution. In fact, he suggests that diabetics minimize carbohydrate intake to 30g of carbohydrate a day or less, even if they are using insulin to offset the effects of those carbohydrates. Why? Because there is a sizeable margin of error permitted in reported carbohydrates. If you plan to take in 60g of carbohydrates at a single meal (not atypical for the ADA recommendations), and you're off by 20% (the allowable error), that's an error of 12g of carbohydrates, which can represent 60 mg/dL of blood sugar difference in a type 1 diabetic. If you're shooting for 90 mg/dL (close to normal), you've now got a resulting range of 30 mg/dL (exceedingly hypoglycemic and lifethreatening) to 150 mg/dL (unacceptably high no matter what the ADA thinks is reasonable; this level of blood sugar has been shown to be damaging.) Oops. If a type 1 plans for 6g instead, then he or she might see a range of +- 6 mg/dL, or 84 to 96 mg/dL. Perfectly acceptable.

    Now, I'm a type 2 diabetic, and I believe you were referring more to type 2s, but the principle's really the same; it's just that the accidental high now means that I'm burning out my remaining pancreatic function trying to cope with it, which is a fine way to effectively be type 1 down the road. (Admittedly the hypoglycemic episode is less likely for me.) On the other hand, in the case of low carbohydrate intake, the error again would be small and the danger of overworking the remaining beta cells consequently less. Indeed, I might actually see some improvement in the absence of inadvertant spikes...

    Bernstein calls this the Law of Small Numbers. His book should be required reading for diabetics, honestly. Substantial portions of it can be viewed on the book's website. I have no affiliation, etc, except for being someone whose diabetes control has increased remarkably thanks to Bernstein and the Eades'.

  2. Excellent post....now, how likely is it that we're going to see Dr Westman's results in the national media??? Slim to none is my guess!

  3. http://www.annals.org/cgi/reprint/142/6/403.pdf

    Effect of a Low-Carbohydrate Diet on Appetite, Blood Glucose Levels,
    and Insulin Resistance in Obese Patients with Type 2 Diabetes

    Ann Intern Med. 2005;142:403-411.

    In this study, co-sponsored by the ADA;

    “During the low-carbohydrate diet, 24-hour glucose profiles improved dramatically and mean hemoglobin A1c decreased from 7.3% to 6.8 % (P < 0.01) in only 2 weeks.”

    It is tragic that so many millions of people are condemned to relentless progressive degenerative injury by a system that ignores it’s own data rather than confront the failures of it’s dogma.

  4. the emperor has no clothes!!