Statins Cause Diabetes

[ofonorow]

Alan Gaby, MD, has written an excellent editorial on page 86 of the June 2010 TOWNSEND LETTER FOR DOCTORS AND PATIENTS. His observation, probably from the same study we have discussed on the side-effects of statins, is that statins cause diabetes, and that because most diabetics eventually develop heart disease, statins will eventually lead to heart disease. As we have discussed, he points out that most (if not all) statin studies are terminated before this negative effect of statins can be scientifically recorded.

A recently published meta-analysis of 13 randomized controlled trials that included a total of 91,140 participants found that treatment with cholesterol-lowering statin drugs signficantly increased the risk of developing diabetes....

The increase in diabetes incidence of this new study was relatively small (9% compared with placebo), and 255 patients had to be treated with a statin drug for four years to cause one new case of diabetes. However, the benefits of statin drugs are relatively small. In one study, for example, 100 patients had to be treated with a statin drug for an average of 3 years in order to prevent one heart attack. Thus, for every heart attack prevented by a statin, approximately 30 new cases of diabetes will occur. Considering that one of the major long-term complications of diabetes is heart disease, and considering that most statin trials have lasted only several years, the results of the new meta analysis suggest that the cardioprotective effect of statin drugs will begin to wane with time. That possibility is not encouraging because many patients prescribed statins are relatively young and are expected to stay on the treatment indefinitely.

He goes on to theorize that the effect is probably due to less vitamin D caused by lower cholesterol levels.

[godsilove]

The increase in diabetes incidence of this new study was relatively small (9% compared with placebo), and 255 patients had to be treated with a statin drug for four years to cause one new case of diabetes. However, the benefits of statin drugs are relatively small. In one study, for example, 100 patients had to be treated with a statin drug for an average of 3 years in order to prevent one heart attack. Thus, for every heart attack prevented by a statin, approximately 30 new cases of diabetes will occur.

I don't see how he calculated that 30 new cases of diabetes will occur for each heart attack prevented. In fact, even going by the nnts that he provides (255 to result in a new case of diabetes vs 100 to prevent a heart attack), the benefit clearly outweighs the risk.

The abstractof the paper which reported these findings even notes: "Statin therapy is associated with a slightly increased risk of development of diabetes, but the risk is low both in absolute terms and when compared with the reduction in coronary events."

[ofonorow]

Hmmm. I agree, something seems wrong and I sent an email to Dr. Gaby.

However, the larger point is that if these drugs cause diabetes, then one would expect that over time, they will cause heart disease. And the criticism stands, they terminate early. i.e., there are few or no long-term studies.

[Johnwen]

He's taking the 9% and using 100 people as an example. 9% per year Times 3years.
9% of 100 = 9 9X3=27 approx 30. will get Diabetes Thank you statins!
During that 3 year period 1 of the 100 will have not had a heart attack because they used a statin drug!

What a miracle drug! 30 people will have to buy insulin from Big Pharma. MORE MONEY$$$
It also only deprieve's the cardioligist of 1 patient but hay they bought the Statins. MORE MONEY$$

[godsilove]

He's taking the 9% and using 100 people as an example. 9% per year Times 3years.
9% of 100 = 9 9X3=27 approx 30. will get Diabetes Thank you statins!
During that 3 year period 1 of the 100 will have not had a heart attack because they used a statin drug!

The calculations are incorrect.

The 9% represents a relative risk, not an absolute risk - and it's based on a four year period. So if 10 out of 100 patients might normally be newly diagnosed with diabetes over a four year period, with statin use this number would go up to 11 patients.

Again, you only need to take a look at the nnt's that the author is using to see that he's made a gross error in his calculations. If the nnt to prevent one heart attack with statins over a three year period is 100, you would have to have a number needed to harm (for newly diagnosed diabetes over a three-year period) of roughly 3 to arrive at the figure that he cites. That is, there would be an extra case of diabetes for every three patients treated with a statin - and this clearly isn't the case.

I assume the Townsend Letter is not peer-reviewed, otherwise such an error would have been caught before it went to print. In any case, I hope Dr. Gaby has the integrity and humility to issue a corrigendum.

[Johnwen]

The increase in diabetes incidence of this new study was relatively small (9% compared with placebo), In one study, for example, 100 patients had to be treated with a statin drug for an average of 3 years in order to prevent one heart attack. Thus, for every heart attack prevented by a statin, approximately 30 new cases of diabetes will occur.

If you read for what it says!
Not what you think it said!
You might understand whats being said.

In one study, for example

Does this say it was this study or another study he was talking about.????

[godsilove]

The increase in diabetes incidence of this new study was relatively small (9% compared with placebo), In one study, for example, 100 patients had to be treated with a statin drug for an average of 3 years in order to prevent one heart attack. Thus, for every heart attack prevented by a statin, approximately 30 new cases of diabetes will occur.

If you read for what it says!
Not what you think it said!
You might understand whats being said.

In one study, for example

Does this say it was this study or another study he was talking about.????

Your understanding of relative risk is flawed.

Let's consider the numbers from the actual study. The meta-analysis involved around 90,000 participants of whom 2226 on statins developed diabetes and 2052 on placebo developed diabetes.
How do these numbers conform with what you think the study says? They don't.

[Johnwen]

Your understanding of relative risk is flawed.

Let's consider the numbers from the actual study

Which study the one the articles about or the one he's talking about???

[godsilove]

Your understanding of relative risk is flawed.

Let's consider the numbers from the actual study

Which study the one the articles about or the one he's talking about???

The 9% figure is from the meta-analysis. So if you're going to use this number you need to to use it correctly. It does not mean that 9% of patients who receive statins will get new-onset diabetes, which is what your calculations seem to be assuming. So you can't multiply 9% by 100 and say that 9/100 patients treated with a statin over a period of one year will develop diabetes.

The nnt of 100 to prevent a heart attack is taken from a different study. I still have no idea of how he arrived at the conclusion that for every heart attack prevented by a statin, 30 patients would develop diabetes. Even if you're assuming that this is based on figures from a single study and not the metaanalysis, the number of patients that would have to be treated with statins to result in an extra case of diabetes would be roughly 3. If this were the case, this would have thrown off red flags long ago. Compare this to the meta-analysis of several studies which found that this number is actually around 255, with little heterogeneity between studies. Obviously, something is off in Dr. Gaby's calculations.

He also fails to give any perspective on the number needed to treat of 100 to prevent one heart attack. This figure is likely taken from a primary prevention trial, i.e in patients who show no signs of heart disease. If you look at secondary prevention trials, the nnts are much lower.

Owen, please let us know if you get a response from Dr. Gaby. To paraphrase Mark Twain, he is certainly entitled to his own opinions, but not his own facts.

[Johnwen]

[godsilove]

[Johnwen]

[godsilove]

[Johnwen]

Because the people who were involved with the study keep saying the "Rate of Incidence" for diabetes during the study. They don't say that they calculated anything! In fact one of the Doctors in the studies which is on that you-tube post Say's,"Perhaps we need to revaluate how we adminster statins to those who don't have elevated levels of choestrol!" Appearantly their seeing a risk that arose from actual incidence during the study. Your saying they calculated the risk they say," They experienced the risk."

Their also saying 9% as is Dr Gaby saying 9% so for all intents and puposes 9% increase is the rise that can be expected. So in futher studies they can use 9% as a baseline and see if other types of statins are above or below the expected level of 9%.

[godsilove]

Because the people who were involved with the study keep saying the "Rate of Incidence" for diabetes during the study. They don't say that they calculated anything! In fact one of the Doctors in the studies which is on that you-tube post Say's,"Perhaps we need to revaluate how we adminster statins to those who don't have elevated levels of choestrol!" Appearantly their seeing a risk that arose from actual incidence during the study. Your saying they calculated the risk they say," They experienced the risk."

Their also saying 9% as is Dr Gaby saying 9% so for all intents and puposes 9% increase is the rise that can be expected. So in futher studies they can use 9% as a baseline and see if other types of statins are above or below the expected level of 9%.

I don't recall either of those doctors in the youtube videos saying that 9% is the rate of incidence. They describe it as the increased risk, i.e. the relative risk of newly diagnosed diabetes. Look up the terms "relative risk", "incidence" and "prevalence" and you just might understand why you and Dr. Gaby are misusing the 9% figure to reach a conclusion that does not follow from the actual data.

Simple logic ought to tell you that even if the nnt to prevent one heart attack is 100, and the nnh to result in one new case of diabetes is 255 - it is absurd to conclude that 30 patients will develop diabetes for every heart attack prevented.