The Advanced Mediterranean Diet

Diabetes Solution is not a "diet book" in the sense of a weight-loss program.  It's a diet as in "the habitual food and drink of a person," in this case a person with diabetes who hopes to control the diabetes and prevent complications through proper eating.  Few people without diabetes will be interested in this review.  Per Amazon.com's rating system, I give this book five stars ("I love it").

Dr. Bernstein gives away thousands of dollars' worth of medical advice in this masterpiece, Diabetes Solution.  It's a summation of his entire medical career and a gift to the diabetes community.

The book starts off with some incredible testimonials: reversal of diabetic nerve damage, eye damage, and erectile dysfunction.  They're a bit off-putting to a skeptic like me, like an infomercial.  Dr. Bernstein is either lying about these or he's not; I believe him.  His strongest testimonial is his own.  He's been a type 1 diabetic most of his life, having acquired the disease at a time when most type 1's never saw 55 candles on a birthday cake.  He's in his mid-70s now and still working vigorously.

I only found one obvious error and assume it's a misprint. He writes that 95% of people born today in the U.S. will eventually develop diabetes.  The U.S. Centers for Disease Control, on the other hand, predicts that one in three born in 2000 will be diagnosed.

Dr. Bernstein delivers lots of facts that I can neither confirm nor refute.  He's a full-time diabetologist; I am not.

The central problem in type 1 diabetes is that, due to a lack of insulin,  ingested carbohydrates lead to spikes (elevations) in blood sugar.  The sugar elevations themselves are toxic.  The usual insulin injections are not good imitators of a healthy pancreas gland. So Dr. Bernstein is an advocate of low-carb eating (about 30 g daily compared to the usual American 250-300 g).  He says the available insulins CAN handle the glucose produced by a high-protein meal.

Dr. B reminds us that insulin is the main fat-building hormone, which is one reason diabetics gain weight when they start insulin, and why type 2 diabetics with insulin resistance (and high blood insulin levels) are overweight and have trouble losing weight.  You can have resistance to insulin's blood sugar lowering action yet no resistance to its fat-building (fat-storing) action.  Insulin also stimulates hunger, so insulin-resistant diabetics are often hungry.

"Carbohydrate counting" is a popular method for determining a dose of injected insulin.  Dr. B says the gram counts on most foods are only a rough estimate—far too rough.  He minimizes the error by minimizing the input (ingested carbs).  From his days as an engineer, he notes "small inputs, small mistakes."

Dr. B also cites problems with the absorption of injected insulin.  Absorption is variable: the larger the dose, the greater the variability.  So don't eat a lot of carbs that require a large insulin dose.  For adult type 1 diabetics, his recommended rapid-acting insulins doses are usually three to five units.  If a dose larger than seven units is needed, split it into different sites.

He recommends diabetics aim for normal glucoses (90 mg/dl or less) almost all the time, and hemoglobin A1c of 5% or less.  This is extremely tight control, tighter than any expert panel recommends.  He says this is the best way to avoid the serious complications of diabetes. 

Here's a smattering of "facts" in the book that had an impact on me, a physician practicing internal medicine for over two decades.  I want to remember them, incorporate into my practice, or research further to confirm:

• Hemoglobin A1c of 5% equals an average blood sugar of 100 mg/dl (5.56 mmol/l).  For each one % higher, average glucose is 40 mg/dl (2.22  mmol/l) higher.
• He's against any drugs that overstimulate (“burn out”) the remaining pancreas function in type 2 diabetics: sulfonylureas, meglitinides, “phenylalanine derivatives”.  Pancreas-provoking agents cause hypoglycemia and destroy beta cell function.
• The insulin sensitizers are metformin and thiazolidinediones.  He likes these.
• Blood sugar normalization in type 2 diabetes and early-stage type 1 can help restore beta cell function.
• He often speaks of preserving beta cell function.
• He believes in "insulin-mimetic agents" like alpha lipoic acid (especially R-ALA, and take biotin with either form) and evening primrose oil.  These  are no substitute for insulin injections but allow for lower insulin doses.  ALA and evening primrose oil don't promote fat storage like insulin does.
• He says many cardiologists take ALA for its antioxidant properties [news to me]
• He says rosiglitazone works within two hours [news to me] but later admits it may take 12 weeks to see maximal benefit
• One of his goals is to preserve beta cell function if at all possible
• He prefers rosiglitazone over pioglitazone due to fewer drug interactions
• "Americans are fat largely because of sugar, starches, and other high-carbohydrate foods."
• He's convinced that people who crave carbohydrates have inherited the problem, which also predisposes to insulin resistance and type 2 diabetes.  Low-carb diets decrease the cravings for many, in his experience.
• In small amounts, alcohol is relatively harmless: dry wine, beer, spirits.  Very few doctors have the courage to say this.
• If you're in a restaurant, you can use urine sugar test strips and saliva to test for presence of sugar or flour in food
• A rule of thumb: one gram of carbohydrate will raise blood sugar about 5 mg/dl (0.28  mmol/l) or less for most diabetic adults weighing 140 lb (64  kg) and about 2.5 mg/dl (0.139 mmol/l) in a 280-pounder (127  kg).  This must refer to type 1 diabetics or a type 2 with little residual pancreas beta cell function; variable degrees of insulin resistance and beta cell reserve in many type 2s would make this formula unreliable.
• Be wary of maltodextrin in Splenda: it does raise blood sugar.
• Much new to me in his section on artificial sweeteners.  Be wary of them.
• He avoids all grains, breads, crackers, barley, oats, rice, and pasta.
• Most diet sodas are OK.
• Coffees with 1-2 tsp milk is OK.  Cream is OK.
• He eats NO fruit and recommends against it.
• He avoids beets, corn, potatoes, and beans. A slice of tomato in one cup of salad is OK.  A small amount of onion is OK.
• String beans and snow peas are OK.
• Cooked vegetables tend to raise blood sugar more rapidly than raw.
• Use "Equal" aspartame tabs as a sweetener.  Don't use "powdered" Splenda.
• Avoid nuts: too easy to overeat.
• For desert: sugar-free Jell-O Brand Gelatin.
• Yogurt?  Plain, whole milk, unsweetened.  Flavor with cinnamon, Da Vinci syrups, baking flavor extracts, stevia or Equal.
• Avoid balsamic vinegar.
• Need fiber?  Bran crackers or soybean products.
• "Ideally, your blood sugar should be the same after eating as it was before."  85 mg/dl (4.72  mmol/l) is his usual goal.  If blood sugar rises by more than 10 mg/dl (0.56 mmol/l) after a meal, either the meal has to be changed or medication changed.
• Protein is a source of glucose: keep protein amounts at meals constant from day to day, especially if taking glucose-lowering drugs.
• The lowest-carb meal of the day should be breakafast.  Why?  Dawn phenomenon.
• He promotes strenuous, prolonged exercise, especially weight training (extensive discussion and instruction in book).
• Many diabetics on insulin need dose adjustments in 1/2 and 1/4 unit increments [news to me: if I ordered 4 and 1/4 units of insulin at the hospital, the nurses would laugh].
• Typical rapid-acting insulin doses for his adult type 1 patients are 3-5 units.  The "industrial doses" of insulin seen or recommended by many physicians reflect diets too high in carbohydrate.
• He says Lantus only acts for nine hours (nighttime injection) or 18 hours (AM injection).
• He doesn't like mixed insulins (e.g., 70/30).
• Humalog and Novolog are more potent than regular insulin, so the dose is about 2/3 of the regular insulin dose
• "Only a few of the 20 available [home glucose monitoring] machines are suitably accurate for our purposes."  "None are suitably accurate or precise above 200 mg/dl [11.11 mmol/l]."
• Vitamin C in doses over 250 mg interferes with fingertip glucose monitors.  Later he says doses over 500 mg cause falsely low readings.
• He prefers regular insulin (45 minutes before meal) over Novolog and Humalog, because of its five-hour duration of action.
• Insulin users need to check glucose levels hourly while driving.
• His personal basal insulin is 3 units Lantus twice daily.
• He urges use of glucose (e.g., Dextrotabs) to correct hypoglycemia.
• He says hypoglycemia is rare on his regimen.
• He has an entire chapter on gastroparesis.

His recommended eating program in a nutshell:

• Some similarities to the Atkins diet, which he never mentions.
• No simple sugars or "fast-acting" carbs like bread and potatoes, because even type 2s have impaired or nonexistent phase 1 insulin response.
• Limit carbs to an amount that will work with your injected insulin or your remaining phase 2 insulin response, if any.
• "Stop eating when you no longer feel hungry, not when you're stuffed."
• Follow a predetermined meal plan (each meal: same grams of carb and ounces of protein)
• Six g (or less) of carbs at breakfast, 12 g (or less) at lunch and evening meal.  So his patients count carb grams and protein ounces.
• Supplements are not required IF glucoses are controlled and eating a variety of veggies.  Otherwise you may need B-complex or multivitamin/multimineral supplement.
• Recipes are provided.

His has four basic drug treatment plans, tailored to the individual.  They are outlined in the book.  Dr. B provides detailed notes on what he does with his personal patients.

Overall impressions:

• Too complicated for most, and they won't give up higher carb consumption.  It requires a high degree of committment and discipline.  In fact, I’ve never had a patient tell me they were on the Bernstein program.
• If I had type 1 diabetes, I might well follow his plan or the Diabetic Mediterranean Diet, NOT the high-carb diet recommended by the ADA and many dietitians.
• And if I had type 2 diabetes?  Diabetic Mediterranean Diet first, Diabetes Solution as second choice.
• If one can get his hemoglobins A1c down to 5% with other methods, would that be just as good?  Dr. B would argue that all other methods have blood sugar swings that are too wide.
• Many new ideas and techniques here, at least to me.
• He pretty much reveals his entire program here, which is priceless.
• I'm not sure this plan will work unless the patient's treating physician is on-board.
• His personal testimony and breadth of knowledge are very persuasive.

-Steve Parker, M.D.

Disclosure:  I was given nothing of value by Dr. Bernstein or his publisher in return for this review.