February 3, 2010

The Lancet Study: Old Patients Victimized By Poor Treatment Die of Hypos from 70/30 Insulin Regimens

The latest outrage perpetrated on the diabetes community is a study published recently in the Lancet which is being widely reported in the media as proving that the 6.5% A1c is dangerous and that the "safest" A1c for people with diabetes is one of 7.5%.

The 7.5% A1c corresponds to an average blood sugar of 169 mg/dl (9.4 mmol/L). To get an average that high a person would typically be seeing blood sugars rising into the middle 200s after every meal, a level that the ADA has confirmed is associated with the development of retinopathy and eventually, blindness. That A1c also guarantees you will develop neuropathy. Two previous studies, ACCORD and ADVANCE both found that lowering A1c to 6.5 lowered neuropathy significantly.

The abstract of this study can be found here:

Survival as a function of HbA1c in people with type 2 diabetes: a retrospective cohort study. Dr Craig J Currie. The Lancet, Early Online Publication, 27 January 2010. doi:10.1016/S0140-6736(09)61969-3

Unfortunately, this study can only be understood by reading the full text which is NOT available for free. Someone sent me a copy so I was able to read it, and I will summarize the most important points that the full text makes for you. Save them off somewhere, because it is almost certain your doctor, who does not have access to the full text will NOT understand the conclusions of this study and will use it to deny you the treatment you need to keep your A1c in the low, safe zone that will prevent you from getting complications.

I am going to keep this brief to make it easy for you to grasp the concepts you will need to know to counter the arguments your doctor might make.

1. WHO WAS STUDIED: This Study analyzed records of about 50,000 British Patients Dating Back to 1986. In this study there were two and a half times as many patients with A1cs of near 11% as there were patients with A1cs near 7%. This reflects the abysmal treatment patients in the UK received during this period which was characterized by treatment that let them maintain very high A1cs for years before being offered insulin.

The average age of the patients with 11% A1cs was 59.7. The average age of those with the lowest A1cs was 67.4. This again reflects the fact that people in the UK had to spend years with very high blood sugars before being given insulin to lower blood sugars. The study methodology suggests very strongly that these people with 6.5% A1cs had low A1cs because they had been put on insulin, and that they had been put on insulin only after experiencing years of extremely high A1cs.

One reason for these very high blood sugars is that British patients, even today, are usually not given blood sugar meters until they have had uncontrolled diabetes for long enough to produce serious complications. They are never told that they can lower their blood sugars by cutting back on carbohydrates. Instead they are told to cut back on fat with the implication that fat is what raises their blood sugars.

Most of the patients in this study were not diagnosed with until their fasting blood sugar was 140 mg/dl (7.7 mmol/L) as this was the diagnostic standard until 1998. The Hoorn Study showed that fully half of all people diagnosed with this fasting blood sugar cut off already have significant neuropathy on the day of diagnosis. This means they probably had undiagnosed diabetes--post- meal blood sugars over 200 mg/dl (11 mmol/L) for as long as a decade before diagnosis.

The patients who'd been put on intensive therapy already had 2.5 times much severe diabetic kidney damage diagnosed by high creatinine levels as did the younger patients. Keep in mind that the younger patients also had the very high A1cs guaranteed to give them kidney problems as they grew older.

CONCLUSION: The people with low A1cs in this study were those who had spent many years with damagingly high blood sugars, both before and after diagnosis. Many had spent nearly a decade with A1cs near 11%. They were not put on insulin until they had already developed serious diabetic complications.

2. HOW WERE THEY TREATED: The patients given oral treatment were all given sulfonylurea drugs. Since this study goes back to the 1980s, this means they were given the first generation sulfonylurea drugs--the ones now known to promote heart attack. There is a black box warning on all sulfonylurea drugs in the US because of their proven association with increased heart attack death.

The full text of this study makes it clear that the patients with the 6.5% A1cs were those who had been put on insulin after many years of being out of control.

The study does not break out the kinds of insulin or the regimens used. After reading the diabetes discussion boards for years I can tell you what the typical UK insulin treatment was during the time of this study. It was a 70/30 insulin twice a day. 70/30 insulin is a combination of long acting and short acting insulin, usually in the UK involving NPH as the "long acting" insulin. The fast action portion cannot be matched to carbohydrate intake and NPH is notorious for its unpredictable activity curve and ability to provoke severe hypos.

The 70/30 insulin regimen is the one most likely to cause severe hypos, and from the discussion in this study, that appears to be exactly what it did. The group with the low A1cs had a very high rate of hypo. As stated in the study:
In this study, mortality was three times higher in patients in either the conventional or intensive treatment groups who had severe hypoglycaemia than in those who did not have severe hypoglycaemia.
One factor not discussed in this study is the damage caused by a high degree of blood sugar variation. It is now understood that large fluctuations in blood sugar--for example, a blood sugar that rises from 100 to 300 after a meal and then plunges down to 70 three hours later--are more damaging to tissues than blood sugars that stay stable.

When you give one or two fixed dose shots of 70/30 NPH based insulin to people who are eating high carbohydrate diets you guarantee just that kind of blood sugar variability. The 6.5% A1c will result from a blood sugar that goes high and then drops very low because it is an average.

The Kumamoto Study made it very clear that if A1cs are attained by keeping peaks low rather than going high and then low, far fewer complications result.

Finally, because the people in this study had been out of control for years, they were prone to autonomic neuropathy--damage to the nerves that control the systems the body uses to fight hypo. This made it more likely that if they started to have hypos, they would develope "Hypo unawareness" which causes the severe hypos that damage the brain and body. These are more likely in older Type 2s, which is the group put on the poorly designed insulin regimens used in the UK.

CONCLUSION: The low A1cs in this study were due not to good control but to the dangerous combination of blood sugar highs followed by severe hypos. Both these factors are known to increase mortality.


Explain that this study is irrelevant to patients who achieve a lower A1c by using a blood sugar meter to monitor their blood sugars and who keep their blood sugar in a narrow normal in a way that does not cause hypos.

Explain that this study is irrelevant to patients diagnosed before they have developed serious complications who avoid sulfonylurea drugs and use modern insulins less likely to provoke hypo.

Explain that doctors who only read one line summaries of studies may inadvertently harm patients. The authors of this study made clear its many limitations, but the reporting of this study even in the medical press is focusing only on its conclusions without citing the many cautions that the authors provided in the text.

Explain to your doctor that there is a large body of research that makes it clear that it is high post meal blood sugars, not A1c that cause neuropathy, retinopathy, and kidney damage, and that you intend to avoid developing any of these horrible complications by maintaining a safe, healthy blood sugar, as measured by blood sugar testing after meals, rather than relying on the A1c test to determine how you are doing.

You are fortunate in that you have read this information here. Sadly most people with diabetes will not, and it is very likely that their doctors will use this irrelevant result to deprive them of the treatment that could prolong their life and prevent them developing life-ruining complications. Insurers love this kind of study too as it justifies them in denying patients expensive care.



C said...

Thank you very much. I saw the study reported, and did wonder what you were going to make of it. I haven't had a chance to go through the text of the paper yet (but I will), but what you say seems to make a lot of sense.

Jack said...

I expect we're in for a wave of "evidence- based" studies which establish that benign
neglect of diabetes is the safest course -
For medical administrators and insurance


Susanne said...

I am just amazed at the level of ignorance regarding diabetes across the board. When my fasting BS tested at 138, my fancy expensive Dr. told me it was stress, not to worry. Thankfully, I found your Diabetes 101 book and website, and I was quickly able to determine this was NOT OK. With 2000 mg of Metformin and a very low carb diet, my fasting BS is now around 112. Thank GOD for the Internet, or I would still be eating 300 grams of carb a day and not be worrying.

Steve Parker, M.D. said...

Thanks for the valuable analysis, Jenny. I hadn't shelled out $31.50 for the full study yet. I may not now.

My prior reading on the study indicated the researchers focused on only two groups of diabetics: 1) those treated with metformin and sulfonylurea, and 2) those treated with a regimen including insulin.

The subjects in this study aren't really equivalent to current patients in the U.S.

Dr. Bernstein must be shaking his head in frustration and sadness.


Jenny said...


The two groups were a group of people taking met/sulfonylurea as an upgrade from one drug (probably a sulf, but this isn't stated) and a second group that had been on the oral regimen but had been "upgraded" to insulin from what is only described as "oral agents."

But since the data spanned patient records from 1986 to 2008, lumping them together as was done here is very misleading.

I hadn't even thought of the issue of how A1cs measured in 1986 compare to those measured today. I know that the lab my doctor used in 2001 used a different range than today's standard where 5.8% was equivalent to today's standardized 6.0%.

Nigel Kinbrum said...

Someone on a UK forum that I post on just quoted the above study as "proof" that I'm wrong about the optimum diet/medication for people with diabetes.

My reply linked to here. It's one step forwards, two steps back on this side of the pond.

renegadediabetic said...

Sometimes I think the deck is being stacked against us diabetics who want to maintain tight control. The irresponsibility in medical research is appalling. In addition to being poorly controlled for quite a while, these folks were no doubt on the "healthy" high carb diet and using what Dr. Bernstein calls "industrial strength" doses of insulin. Violating the "law of small numbers" results in wide BG swings, which is definitely bad for you.

Thanks for all you do in setting the record straight.

Big Nick said...

Thanks Janet for a very valuable analysis. I strongly believe that keeping blood glucose as near normal as possible will help stave off those horrendous cmplications. I was diagnosed in 2000 (A1c>10) and probably had T2 for 10 years before that. Fortunately I discovered Atkins & Bernstein in 2002 and so far I've avoided complications (A1c ~ 6.5 for 8 yeas). I find the notion that I will fare better with a higher A1c totally preposterous. There are many of us in the UK that have the knowledge to avoid the standard adviced meted out to T2s, but unfortunately the vast majority of people don't. And if you try to explain they think you are nut case!

Elaine said...

I started victoza 9 days ago. So far I love it. I love how it curbs my appitite and lowered my blood sugar. This is my first insulin type product that I have been on. I have only been on oral meds. I have lost 4 lbs in 9 days. That in itself I love. I also appreciate the fact that I only have to do one shot a day and no figuring out no doses on how much I need to take. Dont know if I can handle that.
The low readings is what is scarring me tho. So far my lows have been 66. Normally those come around 3:00 pm So I need to start figuring a afternoon snack. Problem is.... I am NOT hungary! I am not feeling any signs of a low, so that is scarry too. Before, I was able to recognize the signs, now I cant.
I welcome any suggestions.

Jenny said...


That is the kind of question you should be asking the doctor who prescribed this new, not-well understood drug, not strangers on a blog.