Scientists sometimes shift the scientific goalposts

It’s easy to believe that statins have dramatic life-saving properties. The reality is, however, that for the majority of people who take them, they don’t. In the biggest and best review published to date, statins were not found to reduce overall risk of death in individuals with no previous history of cardiovascular disease [1]. What this study shows is that for great majority of people who take statins, the chances of them saving their life are, essentially, nil (just so you know).

Of course, you wouldn’t expect everyone to take this finding lying down. A number of people responded to this study with letters to the journal in which it appeared, attempting to cast doubt on its findings. None of it amounted to much, but I thought I would focus on one response, which in my view demonstrates how some scientists and doctors attempt to shift the scientific goalposts to make their point and suit their ends.

The response came from Drs Gabriel Chodick and Varda Shalev [2]. The main thrust of their objections come in the form of three studies that were included in the review referred to above that they claim have ‘major limitations’. Here’s what they say about each of these studies:

“…their meta-analysis included 3 studies with major limitations: a significant decrement in low-density lipoprotein cholesterol levels over the study period in the placebo arm (Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial [ALLHAT]), old age at therapy initiation (Pravastatin in Elderly Individuals at Risk of Vascular Disease [PROSPER] Study), and incomplete information on low-density lipoprotein cholesterol levels over the follow-up period (Air Force/Texas Coronary Atherosclerosis Prevention Study [AFCAPS/TexCAPS]). All these studies showed negative results; their inclusion would have biased against finding a benefit to statin treatment.”

With regard to the first study, what Drs Chodick and Shalev seem to be saying is that the control group (the group treated with placebo rather than statin) saw natural reductions in cholesterol, so the benefits of taking a statin did not to show up. However, the impact that statins had on cholesterol levels relative to a control group is not important – the only important thing is the impact statins had on health (and, in particular, overall risk of death). This is also true for the last study highlighted by Drs Chodick and Shalev.

As regard the second study, it’s not clear why the advanced years of participants would be a barrier to determining the effectiveness of statins. Actually, the elderly are known to be at particularly high risk of cardiovascular disease, meaning that if anything, this population would, theoretically, be generally most likely to benefit from statin therapy.

In summary: none of Drs Chodick and Shalev’s objections hold any water at all. But they don’t stop there. Here’s the final paragraph from their letter.

“Also, randomized controlled trials are often characterized by limited follow-up periods. Therefore, all-cause mortality benefits may not be apparent in randomized controlled trials among a primary prevention population. It would be informative in this regard to take into account the results of large observational studies with longer follow-up periods to better capture the benefits of statins in primary prevention patients.”

What they’re saying here is that clinical trials don’t go on long enough to detect benefits. It’s better, in their mind, to revert to longer studies that are observational (also known as ‘epidemiological’) in nature. However, such studies look at associations between things, but can never be used to prove the benefits of statins. Only intervention studies can do this.

So, what the authors of this letter are effectively saying is that we should ignore the best evidence we have in favour of quite-useless epidemiological evidence.

One of the authors of this letter is, in fact, an epidemiologist, and really should know better. But then again, both of the authors work for a company that assists drug companies in, among other things, ‘reducing the time to market’ and the writing and submission of scientific articles for publication. See here for more details. It’s a clear conflict of interest, of course, and perhaps goes some way to explain why they make apparently spurious objections to existing evidence and appear to be calling for an approach that can never really get to the truth.

References:

1. Ray KK, et al. Statins and all-cause mortality in high-risk primary prevention: a meta-analysis of 11 randomized controlled trials involving 65 229 participants. Arch Intern Med. 2010;170(12):1024-1031

2. Chodick G, et al. Statins and all-cause mortality in high-risk primary prevention: a second look at the results. Arch Intern Med. 2010;170(22):2041-2

2 Responses to Scientists sometimes shift the scientific goalposts

  1. tal 15 August 2011 at 12:39 am #

    ” for many current scientific fields, claimed research findings may often be simply accurate measures of the prevailing bias “
    John Ioannidis

    Fits the likes of Drs Chodick and Shalev to a ‘T’

  2. DanC 15 August 2011 at 1:12 am #

    Wonderful post, Dr B. For further evidence of goalpost shifting look at what the AMA did to Linus Pauling in the 70s. How dare a Nobel chemist make an incursion into the sacrosanct world of medical trials with a cheap remedy (Vit C).

    What about the current rage, vitamin D? The benefits were discovered by “legitimate” medical researchers. Now, doctors are falling all over themselves to recommend ever higher mega-doses of the vitamin.

Leave a Reply