February 2, 2009

Why Doctors are Telling Type 2s Not to Lower Blood Sugar--And Why They Are Wrong

Doctors who get all their diabetes "education" from newsletters that summarize recent research in oversimplified sound bites have misinterpreted two recent studies in a way that is starting to hurt people.

I know this because I am starting to get email from people with Type 2 whose doctors have warned them that it is useless and possibly dangerous to lower blood sugars and even, occasionally told them to raise their blood sugars.

These are Type 2s who are not using insulin. Occasionally a Type 1 who achieves a 5% A1c may be flirting with danger because they are achieving that A1c by spending a lot of time hypo. But this is not the case here. These people are controlling with diet.

My guess is that this latest rash of terrible medical advice is the result of doctors adding a garbled version of the findings of a new study to their earlier misinterpretation of the highly questionable ACCORD study --a study I discussed in detail HERE, HERE, and HERE

The new study was published in December in the prestigious New England Journal of Medicine and was a five year long study of 1791 military veteran performed at VA hospitals.

It concluded that for this group of veterans, lowering blood sugar "had no significant effect on the rates of major cardiovascular events, death, or microvascular complications." Because the group who lowered their blood sugar here--defined as achieving an A1c of 6.9% had a high rate of hypos, the conclusion seems to be that lowering blood sugar even to the ADA 7.0% target is a dangerous waste of time for people with Type 2.

Here's the link to the study:

Glucose Control and Vascular Complications in Veterans with Type 2 Diabetes William Duckworth, et al.

This study needs to be looked at very carefully, because its results contradict the results of every other large study that has ever been done with people with Type 2 diabetes, including UKPDS, UKPDS-followup, ACCORD and ADVANCE, all of which found significant improvements in microvascular complications as A1c dropped below 7.0%.

Though the amount of cardiovascular improvement varies from study to study in all these earlier studies, most found some improvement, and only one study, ACCORD, found a tiny rise in heart attack incidence in the group with lower blood sugar, a finding that was not confirmed by ADVANCE, a larger, longer study that used the same blood sugar target but did not lower blood sugars using the same promiscuous mixing of side-effect-rich diabetes drugs.

Most importantly, ALL these studies found that lowering A1c lowered the rate of microvascular complications--neuropathy and its resultant amputation, retinopathy, and kidney failure.

I do not have full text access to this latest New England Journal of Medicine article, but here is what the abstract reports:
Median glycated hemoglobin levels were 8.4% in the standard-therapy group and 6.9% in the intensive-therapy group. The primary outcome occurred in 264 patients in the standard-therapy group and 235 patients in the intensive-therapy group. There was no significant difference between the two groups in any component of the primary outcome or in the rate of death from any cause. No differences between the two groups were observed for microvascular complications. The rates of adverse events, predominantly hypoglycemia, were 17.6% in the standard-therapy group and 24.1% in the intensive-therapy group.

I do not question that they found what they say they found. In this group of veterans, whose average age was 60, there was no difference at all between those who had a 8.4% A1c and those with the 6.9% A1c except that one in four of the latter group experienced significant hypos.

But I do demand a bit more explanation about WHY this was the case, and I am disgusted with the editors for not demanding that this be included in the abstract of the study.

As it wasn't, I can only guess what the explanation might be, but these guesses are worth consideration.

A couple points stand out here. The abstract states that 40% of these veterans had already had a heart attack. That suggests to me that they were in much poorer shape than the average person with diabetes. This probably has a lot to do with their outcome.

But I am well aware that the level of care at many veterans hospitals has become very poor, with patients having to wait weeks and months for clinic appointments, and with doctor shortages meaning that these hospitals are often staffed by graduates of foreign medical schools whose training in diabetes may be very poor who do not speak English very well.

Because of this, people with good health insurance who live in more affluent neighborhoods with community hospitals and a choice of doctors avoid them. So the VA hospitals are used primarily by poor and minority veterans. This may be why their health by the time they are 60 years old is so bad: It is what you would expect of people who have had poor access to health care, healthy food, education, safe jobs, clean air, water and safe neighborhoods where you can take a walk in the evening --things that are taken for granted by more affluent Americans.

So right away, I wonder if what we are seeing in this study is not that lowering blood sugar is ineffective in Type 2 diabetes, but that in people who live in polluted neighborhoods, who work dangerous jobs that expose them to poisonous chemicals, and end up living with undiagnosed by highly abnormal blood sugar for 30 years while received almost no medical care, to the point where almost half of them have had heart attacks at a relatively young age, lowering blood sugar is not enough to undo the damage already done.

This is a very different conclusion.

We know that Black people living in inner cities in the U.S. have a far higher rate of amputation than more affluent people. We also know that amputation is completely avoidable when people have access to doctors who are educated in how to treat diabetes and when patients are taught how to use a blood sugar meter, adjust their carbohydrate intake down to a tolerable level, and inject insulin using a basal/bolus regimen.

Black people in inner cities don't get that kind of care. Just as so many Black people had to wait of 8 to 12 hours to vote because their precincts could not afford to buy enough voting machines, while affluent suburbanites could zip into the polls, vote and go home, access to health care is very different in the U.S., depending on where you live and how wealthy you are.

But whatever the explanation for the findings of this study. Too much data has accumulated to take its results at face value. Clearly, something was very wrong here and the researchers owed it to the medical community to explain why their subjects' outcomes were so completely at odds with that experienced by every other population of people with Type 2 ever studied.

If your doctor draws any other conclusion from this study, find a new doctor. We have almost 20 years of data now that support the finding that lowering blood sugars makes a dramatic difference in whether or not people experience microvascular complications and that the lower the blood sugar the lower the incidence of microvascular complications. Even ACCORD found this to be true!

The only remaining debate is whether lowering blood sugars can have any impact on established heart disease and the answer to that question has not yet been settled.

Any doctor who tells you otherwise and urges you to worsen your control is a danger to your health. Don't debate such a doctor. Fire him. It's that simple.

NOTE: Check out the comments, you'll learn more about what was in the full text of this article. When I get more time I'll update this information too, as several people have sent me the full text version.


Brenda Bell said...

I don't have access to the full-text either, but I can certainly hazard an educated guess to the reasons for the poor outcomes. First guess would be spikes and dips. Remember that for those of us without CGMs -- which is probably more than 90% of T2s -- we never learn the true magnitudes of spikes, dips, and area-under-curve. Consider that most oral antidiabetic regimens available in the US have an activity curve closer to that of basal insulin than bolus insulin -- meaning that if close attention is not paid to diet, there will be large spikes and dips. Add to that, VA-treated patients not on insulin are generally told to test not more than once a day, maybe twice. (This is similar to the instructions given Medicaid/Medicare patients, based on Medicaid/Medicare coverage of strips.) The result of this is a good A1c disguising a lot of diabetes-related damage going on behind the scenes.

Another issue to consider is that some of the previous trials that came up with similar (similarly-flawed) conclusions used pharmaceuticals that have since been shown to pose cardiovascular risk to the very class of patients on whom they were used for the purposes of intensive BG control.

bsc said...

I have access to the full text, and as you might expect, not only was the study designed to only look at certain things, but the actual study has been misinterpreted widely. These veterans were already in bad shape, having diabetes an average of 11 years, an HbA1c going into the study of 9.5%, a BMI of 31% on average and as noted 40% already had cardiovascular events. What is most dissapointing in the study is that while the results show virtually no improvement in microvascular outcomes, but in many cases, the original rate of microvascular problems was not reported. The P values all indcated nothing can be drawn from the results, but one might suspect that is because the baseline rates of microvascular problems was so high. Given a population with ten years of out of control blood sugars and probably high microvascular complications, dropping their blood sugars down to a marginal HbA1c 6.9% was unlikely to be enough to affect the already major damage.

Scott S said...

This is crazy, it is well-established that glycemic control is imperative, so any doctor who recommends that patients with type 2 do not focus on glycemic control is obviously not paying attention to the UK Perspective Study (as well as the type 1 DCCT) which have proven unequivocally that glycemic control does prevent complications. Soundbytes are not the same as CME (continuing medical education) credits!!

Anonymous said...

Something else to consider. With type 2 the VA believes you only have to check your blood sugar every other day. They will give you a free meter and 50 test strips to last 90 days. Makes it very difficult to get good control.

renegadediabetic said...

Stupid, Stupid, Stupid!!!!!

This is basing conclusions on one flawed study while ignoring the preponderance of the evidence from other studies. Sadly, after reading Taubes, I've come to expect this kind of nonsense & non-science from the medical establishment.

Anonymous said...

We're already seeing this in parts of the UK, often coming from nurses and presumably the source of the information is the PCTs (our equivalent of HMOs) Added to the refusal of test strips thanks to Farmer


(some PCTs are now printing leaflets advising against testing which further reduces their test strip budgets)

saving money becomes more important than saving diabetics as we have all brought it on ourselves through greed and sloth, so any paper such as this will be widely disseminated.

I would put money on the Heart Healthy diet being behind the decline and the high A1cs in the first place, and of course the cardiovascular disease. You might almost think they are trying to kill everyone who can't live on Healthy Whole Grains

The Old Man and His Dog said...


I just want to thank you for all the information you provide. It has helped me tremendously. I just had my recent blood work done. A year ago my a1c was 9.0 and my fasting BS was 199. My most recent a1c is 6.1 and fasting BS is 87.! I'm headed for the 5% club!


Thank you!!

Jenny said...


That is WONDERFUL news! I'm so glad you are doing so well. Heck, I've never tested with an 87 fasting blood sugar ever. I'm happy to get into the 90s!

Anonymous said...

I am disappointed that you comment on the NEJM article and say you have not read it! You have only read the abstract. I would think that you can get the whole article from a library and read it in its entirety before commenting on it.