May 17, 2007

Misunderstanding UKPDS - 7% is NOT a Good A1c.

The ADA board is full of people with A1c test results over 7% whose doctors tell them they are doing fine. They aren't. The doctors who tell them this are as irresponsible as if they told them not to worry about a "touch of cancer."

My guess is that uneducated doctors think an A1c near 7% is "fine" because they've only read the one line summary of the findings of the UKPDS. That "25 words or less" version is that people with Type 2 Diabetes who attained A1cs of 7% reduced the incidence of complications.

What this summary statement ignores, is that UKPDS also showed that, while the rate of complications in the population with A1cs of 7% was better than that in the population with A1cs of 9%, people with those 7% A1cs still developed microvascular complications at a very significant rate, and, even more importantly, they had not decreased their likelihood of dying from a heart attack or stroke.

Here's what the actual findings published in the British Medical Journal said:
Each 1% reduction in updated mean HbA1c was associated with reductions in risk of 21% for any end point related to diabetes . . . No threshold of risk was observed for any end point. [i.e. this stayed true as the A1cs continued to drop] . . .Any reduction in HbA1c is likely to reduce the risk of complications, with the lowest risk being in those with HbA1c values in the normal range (<6.0%)."

Does this sound to you like 7% is a good A1c? I don't think so.

Even more significant is that other studies show that to decrease the incidence of cardiovascular "incidents" i.e. heart attack and stroke, you need to lower A1c far below 7%.

The chart below, derived from the huge EPIC-Norfolk study makes it very clear that the risk of heart attack DEATH has already doubled at an A1c 6%. (from Medscape New Avenues for Complicated Patients with Type 2 Diabetes and Hypertension) The chart graphs the risk of death from various causes for various A1c levels. CHD: Coronary Heart Disease. CVD: Cardiovascular Disease (includes stroke).

Could any doctor who looked at this data complacently tell a patient that an A1c of 7.3% was fine?

Clearly, the only A1c that is truly "fine" is 5% or less. Many of us can't get there. I sure can't. But the 5%s are a whole lot better than the 6%s and with the 7 times higher risk of death at the 7% you'd have to be crazy to be complacent about an A1c that high.

If this is all making you nervous, don't despair. I know literally hundreds of people who have brought their A1cs into the 5% range, often starting from as high as 13%. The tools you use are cutting back on your carbohydrate intake until you can get under 120 mg/dl two hours after every meal and, if carb restriction isn't enough, well chosen meds, including the one drug that always works for people with diabetes: Insulin.

Normalizing blood sugar saves lives. It can be done. It must be done. And if your doctor won't help, find a new doctor. It isn't his eyes that will go blind, his feet that will get amputated, or his heart that will give out when you take his outdated and dangerous advice!


Anonymous said...

You're incorrect when you mention a 7 fold higher risk of death purely from a1c. What the study says is that the odds ratio is around 1.24 (i.e. a 24% higher risk) for a 1% increase in a1c. That is after controlling for other factors. If you don't control for age, wasit-hip ratio, BP, cholestrol etc., you get the 7 fold risk.

And for stroke, the Epic Norfolk study did find a threshold around 7%, below which stroke related events were almost the same at different a1c level.

None of this is to suggest that 7% is a good a1c. I too think that people should try to get below 6 (My last a1c was 5.2% without meds ). But I don't think one should be too alarmist either.

Bernard said...

I wonder how realistic it is for someone with Type 1 diabetes to get down to 6%. I think for folks with Type 2 that's more attainable.

For those of us blessed with Type 1, I think we might spend a lot of time riding in ambulances.

Though reducing carbs would certainly help. And perhaps Symlin would also reduce those post-prandial excursions.

Jenny said...


The UKPDS study ONLY involved people with Type 2 diabetes.

It was intended to duplicate DCCT which only involved people with Type 1. However, it found a much higher rate of retinopathy etc. occurring in people with Type 2 who had those 7% A1cs than had occurred in people with Type 1.

This is probably because people with Type 2 typically go 10 years with elevated blood sugars before they are diagnosed, so the damage is already well under way when they start controlling. People with Type 1 are diagnosed immediately.

Anonymous said...

So, I just got back from a Dr. visit showing my A1C at 7.1. This is down from 13 in just 45 days. To accomplish this I monitored 5 to 8 times a day and ate primarily protein, fats and cellulose... The only problem is that my LDLs went from 108 to 141. So I lose on that front.. Dr. increased the statin dosage, but I must say this past 2 months have been very very hard.. diet wise and excersice wise. P.S. I just now over 50 years old...

Jenny said...


LDL often goes up when you are losing weight. This is normal. After six months on a low carb diet most people see great improvements in their cholesterol profiles.

However, there is a lot of question as to whether LDL levels really mean much as far as heart disease. Gary Taubes' Good Calories, Bad Calories takes an exhaustive look at the data and reading it should set your mind at rest.

Anonymous said...

It's great to see people challenging the complacency of the ADA and mainstream doctors.

For Type 1 and 2's they only need to check out Dr. Richard Bernstein's book from the library and see if it works for them. While he seems to be controversial, he's certainly worth reading.

Pattidevans said...

I don't think it's unrealistic for T1s to get down to 6%. I have maintained an Hba1c under 6 for the past 7 years. This is not at the expense of multiple hypos, but of taking notice how my body works, of controlling my diet and learning how to take insulin accordingly. It's doable! There are other T1s in forums around the world who achieve the same. There are also thousands of T2s who can achieve it. A little learning and a little understanding achieve much.

Jenny, thanks for this article.

Ahmed said...

Although its true tht risk of complications at 7 is still present but you forgot to mention that according to WHO a level below 6.5 carries higher level of mortality from hypos so its not feasible to be under 6 please dont just say your opinion when you ddnt study the case well

Jenny said...


This is not an opinion and I have studied this subject exhaustively for many years. The only way that an A1c of 6.5% would reflect a risk from hypos is if you are using insulin incorrectly or taking an overdose of insulin secreting drugs.

If you lower your blood sugar by cutting back on carbohydrates and avoid the drugs that cause uncontrolled insulin secretion whether or not food is present (glimipiride, glipizide, etc.) you can get A1cs in the high 4 and 5% ranges with no hypos at all.
I know hundreds of people who do this and have done it myself for 14 years.