Disorders of Blood Sugar

Roberto Victor Illa, M.D.

Main Page

Getting off of insulin injections safely

Obesity & Weight Control

Referral Instructions

For Patients

For Physicians
and Researchers

Patient Stories

Dr. Illa's CV

Video Testimonials

Learn more about diabetes

To order a copy of the 4th edition of Dr Illa's book, click here or on the cover image, above.

Turning ADA on its Head:

A Conceptual Preview of the Illa Protocol

Roberto Victor Illa, M.D.
Stanley E. Lieberson, Ph.D.

October, 2007

Why is everyone following the American Diabetes Association protocol for the treatment of diabetes mellitus?

Published studies have shown it is a failed protocol (see References, at end). Why is it so ineffective, and what can be done to reverse this failure?


As is commonly known, diabetes is a disease in which the body's mechanism for maintaining optimal levels of glucose in the blood (blood sugar level) has failed, and blood glucose levels become high. Although there are other hormones involved, the body primarily uses alpha and beta cells, in the pancreas, to raise (alpha cells) and lower (beta cells) blood sugar levels as needed to maintain your sugars in the normal range.

Physicians have long known that hyperglycemia (high glucose levels) is harmful to many organs in the body, and chronic high blood sugar levels can harm circulation, vision, the brain, nerves and many other organs. This harm occurs not all at once, but slowly, over a long period of time, even years, bit by bit and piece by piece. Physicians have also long known that its opposite, hypoglycemia (low blood glucose levels), is also harmful to the body. However, unlike hyperglycemia, the harmful effects of hypoglycemia are immediate, and can be quite dramatic; effects include mental confusion, physical weakness, loss of consciousness, seizures, brain damage and death.

The American Diabetes Association (ADA) Protocol

Given that both hyperglycemia and hypoglycemia are harmful, and that hyperglycemia is a long-term threat while hypoglycemia is an immediate threat, one could be excused for thinking that the treatment of diabetes would stress the gradual decrease of the high blood sugar levels so as to avoid episodes of medication-induced hypoglycemia. We say “excused for thinking” this because in fact the ADA promotes just the opposite! Their protocol, also known as the “tight control” method, aggressively attacks high blood sugar levels. In their attempt to immediately bring down high levels, they risk providing too much medication and thereby inducing hypoglycemia, with its attendant immediate damage. This seems counter to what we know about short- and long-term dangers. (It also seems counter to the medical dictum “First, do no harm.”)

Turn the ADA protocol on its head!

So let us turn the ADA protocol on its head! Let us treat a patient with diabetes by intentionally (and quite purposefully, as you'll see) giving them a sub-therapeutic dose of medicine, a dose clearly too low to bring their blood sugar levels down to the optimal range. This isn't going to cure them, but it is not going to make them sick (hypoglycemic), either.

Now, let's see what this dose accomplishes. Let's give our diabetes patient a home glucose meter, and request that they measure their blood glucose level two or three times daily, before eating. And let us require them to report those readings after two or three days (or bring the meter back to the office for the physician to read it).

What will we see? We will see a brief graph of recent blood sugar readings. In particular, we will see what the effect of our sub-therapeutic medication dose was. (Treatment note: the patient also collects a few days readings before coming to the office for that first, sub-therapeutic dose, so we have before and after graphs.) Most likely, we will see that that our initial dose was too little to produce normal levels—no surprise there. But by carefully studying the graph, we can learn some useful and most interesting things about this patient's diabetes.

Let us look at some statistical aspects of these data. (Yes, we must get mathematical here—like it or not, biochemistry is described, to an important degree, by mathematics, no less than is physics or astronomy. If even a little math makes your eyes glaze over, just skim this section, or jump directly to the The Illa Protocol section, below.)

The data has a mean (that's the mathematical term for the commonly-used average value). This mean will, of course, be high, way above normal blood sugar levels. But what else is there to see? There is the standard deviation, which is simply a measure of the degree of variation of the points. A high standard deviation means that few points are near the mean, while a low standard deviation means that few points are distant from the mean. For example, in the figure above, you can see that the means of the two graphs are about the same, although the graph on the left has a greater “scatter” or standard deviation. The standard deviation of the blood sugar readings produced by your home glucose meter tells you something of value about your diabetes. It tells you that not only is your blood sugar level too high (the mean), but also that your body is reacting to it poorly, resulting in those wild swings a high standard deviation represents. We have found that these wild swings are of great importance in understanding the state of your body, just as the mean is. For example, some drugs, like insulin, tend to affect the mean (they bring the whole graph down, without greatly changing the standard deviation, the wild swings), while others, which target the alpha cells, act to reduce the swings, while not necessarily reducing the mean.

Congratulations! You have now learned that by simply looking at the graph of your blood sugar readings, you can learn about the relative health of your alpha cells and your beta cells, and thereby choose medications appropriate for your case.

Now let's use the “High S.D.” graph above and look again at the effect of “tight control” on your health. If a form of insulin is given to someone with a blood sugar graph resembling that of the High S.D., the effect will be to lower the whole graph. This, of course, will lower the mean, which is certainly a good thing. However, it will lower all of the points, including the low ones. Look at the lowest point of that graph. If you lower the mean but not the standard deviation, than you run the risk of lowering the graph so much that the lower points drop to the hypoglycemic level, thereby inducing those immediate adverse effects mentioned at the opening of this paper.

How to avoid this harmful effect? The ADA provides no guidelines, no specifics. Physicians are left with guessing how much insulin (for example) to use to maximally reduce the mean without risking dropping the lower points into the unhealthy region. Five units of insulin? Fifteen? Thirty? It is a gamble.

The Illa Protocol

With the Illa Protocol, however, gambling is removed. That sub-therapeutic initial dose, called a probe, and the blood sugar graphs produced by your home glucose meter, provide the information needed to set you on the course toward controlled blood sugar levels. This is because, in addition to the probe and the graphs, another principle of the Illa Protocol is to avoid hypoglycemia by making only very gradual, step-by-step changes to your diabetes medications. Although the probe fails to reduce your blood sugars to normal levels (as, in fact, it is designed to fail), the resulting graph provides the information (mathematically) to guide the physician to the next, gradual, increase in dosage. The Illa Protocol includes a number of tables with specific dosages of specific medications (all are FDA-approved—you probably already know the names of most or even all of them—there are no unstudied plant extracts here). Unlike the ADA protocol, the Illa Protocol specifies which medications, and in what exact doses and frequency, to use. The choice of which probe to give, and which tables of gradual steps to follow, is determined by a combination of your diabetes-related health (including that of your kidneys, since they clear insulin from your body) and your pre-probe graph, and is based on a—yes, that word again—mathematical analysis of your graph. The Illa Protocol uses a combination of the mean, standard deviation, and slope of the least-squares curve fit (the line most closely fitting the points in your graph) to determine your treatment. (Physicians probably never compute a standard deviations or least-squares fit once out of med school. Fortunately, we have created a software package to do these computations; the software package also tracks patients' blood sugar readings, produces graphs and even shows which tables of medications to use for a given patient and the step in the table—the specific dosage and frequency—appropriate for the patient, based on the most recent home meter readings.)

The result is a very specific treatment plan targeted to each individual, to their body's glucose-regulation functioning when they begin treatment under the Illa Protocol. The Illa Protocol guides the physician in making small, initially frequent, changes in the doses. (Note to patients: This does not require frequent visits to your physician! You can report your home glucose meter readings by phone, e-mail, fax, etc. Your body doesn't have to be in the physician's office to do this.) Another result of the Illa Protocol is that, if you follow the protocol, you will rarely, if ever, experience hypoglycemia.

In addition to the probe, the daily graphs and gradual, small changes in medications, there is another important principle in the Illa Protocol. The Protocol aims not only to safely bring your blood sugar levels under control, but also to improve your functioning by increasing your beta cell mass. (Recall that your beta cells, which produce insulin, are key in reducing high blood sugar levels, while your alpha cells do the opposite, they increase blood sugar levels when yours get too low.) If you have too few functioning beta cells, your body cannot regulate your glucose levels. A careful reading of the medical literature shows that a class of medications, called TZDs, can act to increase your beta cell mass. Unfortunately, not all physicians treating diabetic patients include TZDs in their treatment regemin. They may have read papers reporting that no effect was found with TZDs. Dr. Illa has realized that TZDs do not increase beta cell mass while the standard deviation (of your blood sugar readings) is large, but they do increase beta cell mass when the standard deviation is smaller, even if the mean is still high. Thus, those studies using uncontrolled (and high standard deviation) patients failed to find a positive effect for TZDs, and many physicians, seeing these reports, don't use this class of drugs. But when applied to patients with more controlled levels, TZDs have been shown to increase beta cell mass.

What this means is that (1) the Illa Protocol includes TZDs right from the start, and (2) when the Protocol brings your sugars into a more stable range, the TZDs begin to work, that is, you begin to increase your insulin-producing beta cells. This improvement will show up on your graphs, and the Illa Protocol takes that into account, and begins to gradually, again in a step-by-step fashion, decrease your medications (now going backwards down the steps of the tables). Once again, no sudden changes, no risk of hypoglycemia.

The Illa Protocol is applied to both Type I and Type II diabetics. It is helpful in both cases—patients will improve, and will not have hypoglycemic episodes—and in Type II cases it can lead to remission. Remission means controlled, normal blood glucose levels in the absence of medications. Dr. Illa has seen many of his patients experience periods of several months of remission (the longest so far: two years). When a patient relapses (as most or all will), their levels do not rise as high as before the Illa Protocol treatment was used, and a return to the treatment (probe, steps) quickly returns them to normal levels.

Diet and Weight Loss

Did you notice the absence of discussion about careful attention to strict dietary guidelines and regular exercise? Every physician should remind us of the benefits—and there are many—of attention to diet and exercise. But exercise is not a part of the Illa Protocol, and diet, well, the Protocol doesn't really include any specific diet requirements (other than to avoid alcohol), although you will be requested to be moderate and sensible in the ingestion of carbohydrates. But no strict requirements, no measuring, no counting, and no blaming yourself—or having the physician blame you—for another disappointing A1C reading. (In fact, because of the immediacy of the home glucose meter tracking required by the Illa Protocol, an A1C, which is a measure of average blood sugars over some three months, and tells you nothing about today or yesterday, is pointless.)

Does this make sense? Are you unhappy with your current, ADA treatment? Have you had even one hypoglycemic episode? (Hopefully not while driving!) Tired of measuring your cereal, your meat, the number of minutes spent walking, swimming or bicycling? Want to just measure your blood instead, and actually improve? Contact Dr. Illa (www.ChicoDiabetesDoctor.com, or phone 1 530 345-1157) and tell him about your diabetes and your treatment up to now. Even if you do not live in his area, he may still be able to help you, as only an initial visit and occasional follow-up visits are required in most cases, as long as you send him your daily blood sugar readings by e-mail, fax or phone.

If you are a physician and want to escape the ADA failure and the steady deterioration of your patients, ask Dr. Illa about the Illa Protocol Manual, available from the publishers, Amazon.com and other booksellers.

If you are an HMO administrator, worried about the ever-increasing costs of the diabetes pandemic, and not afraid to face down the lawyers in your organization (this fear is why the ADA holds sway, despite being appointed by nobody), talk to Dr. Illa, who can come to your facility and train your staff to use the Illa Protocol (they can avoid calculating standard deviations and slopes by use of the DiabetesControlTower software, which is a fully multi-user database application—you can use it whether you're a lone practitioner or a group of hundreds.)


Failure of ADA protocol::

1. Recent study demonstrating failure of ADA protocol: health.msn.com/centers/diabetes/articlepage.aspx?cp-documentid=100170830

2. "Although new classes of medications, and numerous combinations, have been demonstrated to lower glycemia, current-day management has failed to achieve and maintain the glycemic levels most likely to provide optimal health care status for people with diabetes." From Conclusions section of a consensus statement from the American Diabetes Association and the European Association for the Study of Diabetes, reported in Diabetes Care, 29:1963-1972, 2006 and posted at Diabetes Care's website.

TZDs and beta cells:

3. TZDs can regrow beta cells: Bell, DS. Beta-cell rejuvenation with thiazolidinediones. Am J Med. 2003 Dec 8; 115 Suppl 8A:20S-23S.