March 3, 2009

Extremely Bad Science: HEART2D "Proves" Fast Acting Insuiln is Worthless

Sometimes you see a study published that is so poorly conceived it leaves you wondering how lab chimpanzees were able to take over the department that ran the study and how they got into the journal office that published it.

The most recent study that falls into this category was published in this month's Diabetes Care.

Effects of Prandial Versus Fasting Glycemia on Cardiovascular Outcomes in Type 2 Diabetes: The HEART2D trial

It is very important that you pay attention to what is wrong with this study, because it is yet another study whose outcome is going to be used to justify denying you the care that could prevent you from developing complications.

The idea behind this study was to see if lowering post-meal blood sugars would be more effective in preventing heart attacks than using a strategy of lowering fasting blood sugar alone.

Okay. That sounds reasonable. Why am I trashing this study?

Here's why: The study took a group of people who had already had a heart attack. It assigned them to two groups. One received three shots a day of Humalog (Lispro). The other received either 2 shots a day of NPH or one shot of Lantus.

Do you see the problem with this study? Of course you do! They gave no basal insulin to the people who were given the post-meal insulin. But all these people started out with very high A1cs which tells you that their fasting blood sugar control was shot. So when those three shots of fasting insulin wear off, their blood sugar will go right back up.

That, apparently, is exactly what happened. At the time the study was stopped both groups of patients had nearly the same average A1cs: 7.7 ± 0.1 (Humalog) and 7.8 ± 0.1% (basal insulin.) The people using fast acting only would have had their blood sugars rise any time they were in the fasting state. The people using basal only had their blood sugars shoot up any time they were in the post-meal state.

No group was given insulin to control their blood sugars in both states. Brilliant, eh?

Endocrinologists do not prescribe post-meal insulin alone. They prescribe it in combination with basal insulin. That researchers in a major study involving 1,115 patients were so ignorant that they set up their experiment in a way that violated everything doctors know about blood sugar control is criminal. That the editors of Diabetes Care considered this study worthy of publication is even more so.

Criminal? That's a strong word. But it is probably not strong enough. Because you can be certain this study will be used by insurance companies as "Evidence Based Medicine" that justifies withholding expensive fast acting insulin from patients.

In fact, all this study really proved is that if you maintain your blood sugar at levels high enough to produce a 7.7% A1c and already have severe heart disease you have a one in three chance of having another heart attack.

This is not news. EPIC-Norfolk data already showed conclusively, drawing on a huge number of study subjects, that people with A1cs well over 7% have three times the risk of heart attack as those with an A1c of 5.5%-6% and between 4 and 5 times higher than those with an A1c of 4-5.4%.

So in addition to proving that using fast acting insulin without basal insulin is ineffective in people with very high blood sugars. this study also proves only that allowing patients who have already had a heart attack to maintain average blood sugars at levels known to quadruple cardiovascular risk in anyone will not change their risk.

But try telling that to your insurer when they tell you they are no longer paying for your fast acting insulin because "studies show it doesn't do anything.

13 comments:

Scott S said...

The scary part of this is the fact that the study, now with the G.W. Bush Administration's parting gift to the drug industry before he left office last year, will be used and given to salespeople to sell drugs. Most of those who buy into the sales pitch are family doctors who do not have the latest data on diabetes care anyway, but it will still harm many people thanks to sanctioned "off-label" prescription sales!

The Old Man and His Dog said...

How about this article in the NY Times "Harvard Medical School in Ethics Quandary ". Drug companies advancing their agenda through professors in schools! We don't stand a chance even with new Dr's

http://www.nytimes.com/2009/03/03/business/03medschool.html?_r=1&ref=business

Harold said...

I have a question about fast acting insulin. As I understand, it works in about 20 min when given subQ and a bit faster im. Regular insulin (humulin R) work in about 45 min and I believe in about 15 min im. If this is correct why not use regular im as it is so much cheeper than fast acting?

Anonymous said...

Seen in another light, this is just one more example that there are NO studies of those who attempt to be healthy and have good or normal blood glucose levels. RobLL

Anonymous said...

Sometimes you see a study published that is so poorly conceived it leaves you wondering how lab chimpanzees were able to take over the department that ran the study and how they got into the journal office that published it.

No it wasn't the chimpanzees, it was the baboons!

http://www.proteinpower.com/drmike/lipid-hypothesis/baboon-business/

Thanks you the likes of you, Michael Eades, Gary Taubes etc. for doing what the Peer Reviewers have failed to do in letting these "studies" out into the wild and ripping them apart.

"Statistics? Certainly Sir, which result are you paying for today?"

Jenny said...

Harold,

There is no reason not to use R. It has a longer activity curve than the analogs and if you use it a couple times of day the tail will actually give you a bit of a basal effect which the analogs won't do.

I used it very happily for several years, but the last 3 bottles I have bought have been completely ineffective. I don't know exactly why--it might be the pharmacies aren't taking good care of it--they don't sell much so it sits around for ever, or it might be that since I stopped responding to it after adding NPH for a few weeks something about NPH messed up my response to it.

I'm sad because I used it for quite a while and it is the only insulin that doesn't pack weight on me.

Bernard said...

Jenny thanks for bringing this idiocy to light. It constantly amazes me what passes as 'research'. I wish they'd used the funding for something useful. I'm lost for words at this terrible waste.

Anonymous said...

I use Humulin Regular and it's very effective for me (I am mildly diabetic and live in Switzerland), 2 or 3 units per meal (low carb) and don't need any basal (fasting readings: 4,4 - 5 mmol, pp usually 4,9 - 5,8 mmol). No hypoglycemia and no weight gain.

renegadediabetic said...

One thing to note is that these idiots are so fixated on heart disease that nothing else matters. If a treatment doesn't reduce the risk of heart disease, than it's worthless. The same reasoning was true for ACCORD and ADVANCE.

Who cares if you go blind or have to go on dialysis? As long as you don't have a heart attack, everything is fine. :)

Anonymous said...

I know a Type 1 who uses the logs to cover higher carb meals but prefers R for his usual low carb meals as it onsets and runs out slower and better covers lower GI things.

Talking of duff studies, have you seen this magnificent takedown of Ancel Keys?

http://high-fat-nutrition.blogspot.com/2009/02/cholesterol-presentation-between.html

This started off the technique of choosing your data according to your preconceived notions to the extent that totally crap papers are getting published simply because they confirm current dogma.

Anonymous said...

Your wrong. as a diabetes nurse i can assure you there are many type 2 patients who benefit from bolus only insulin regimes, especially those who have high carb meals and dont want to change. Many just need a touch of fast acting to get postprandial readings back in range. It is more widely believed now that it is a yo-yo of blood glucose readings that causes more harm rather than higher readings, hence trying to see if stopping post meal peaks reduces MI.
The point in general i willstress is that there is no 1 magic regime, everyone is different, so everyone needs a regime to suit them. I suggest you research things a little more thoroughly yourself before dissing a potentially lifesaving study.

Jenny said...

Anonymous 4:10's post shows a complete lack of comprehension of the point made in the post.

I love it when people post anonymously and insist they are Very Important People whose opinions trump any data that has been presented, too.

Don't you!

Unknown said...

What a great Review. Amen. Please if you have diabetes... keep taking meal time insulin and if it has been suggested for you, do it. It is in your best interest if your 2 hour after meal BG # are > 140 mg/dL! Also consider using pump therapy even if you have type II diabetes. There are pumps with tubing but also a pump without tubing called the OmniPod, if that has been part of your objection.