What follows is a poorly written article that is a good example of the kind of bad healthcare journalism that I frequently rail against on this blog. To be clear, my thoughts on statins are nuanced (as I will explain during the course of this post) and the central point of the article – that statins are generally safe and well tolerated, and that they have been proven effective in cutting the risk for cardiovascular disease is, on a certain level, correct. My beef (pun intended) with the article is not in and of itself that conclusion. We live in an age of internet driven conspiracy theories, and some people have taken to blaming statins for every imaginable problem. An article that pushes back on that narrative by presenting the more mainstream view on statins is perfectly fair game.
Rather, what galls me about the article is that it makes wildly exaggerated claims about the benefits of statins, while ignoring the nuance around the criticisms of their side effects.
Take, for example, the opening paragraph of the article, which states that “For every 10,000 people who take these cholesterol-lowering drugs, 1,000 will avoid major cardiovascular events.” That’s a confusingly worded sentence, but if you read my last post on risk reduction and number needed to treat (NNT), it implies an NNT of 10 (because saying 1,000 out of 10,000 takers of statins avoid a heart attack is the equivalent of saying that one in ten people who take these medications will avoid a heart attack, i.e. for every ten patients who take these drugs, one will have a major tangible benefit).
This is a stunning claim. I have never heard any doctor or scientist claim that statins have an effect size even close to this large. Where does the author get this outlandish statistic? I have no idea, because he doesn’t cite any source in the article. I even went so far as to email him and ask, but have received no response (I’ll assume my email went into his spam folder, rather than that he is ignoring my question.)
In point of fact, depending on which study you are looking at, and which group of patients you are talking about, statins are generally thought to have an NNT of about 50 in people who have already had a heart attack, and an NNT of around 300 in those who have have high cholesterol but have not yet had a heart attack. Or, to more accurately re-word it in the author’s clumsy formulation: for every 10,000 people who take these cholesterol-lowering drugs, somewhere between 20 and 200 will avoid major cardiovascular events, which obviously is wonderful if you are one of those people, but is not even close to the 1,000 who he claims will be helped.
So it is fair to state that statins have benefits. But implying that the NNT for them is 10 is as divorced from reality as stating that the way to cure stomach ulcers is via exorcism of a demon.
Which brings me to the first point about statins that I wish to make: it is simultaneously true that they reduce the risk for heart attack and stroke, and that they have a fairly small benefit and are unlikely to help the vast majority of people who take them. Both things are true, and therefore a doctor may reasonably recommend – and a patient may reasonably decide – that this benefit is or is not worth it to them based on their personal circumstances and preferences.
If statins aren’t all that helpful for many people, what about their downsides? After all, every drug has side effects. If they didn’t, then even the possibility of only a very small benefit would be worth it.
Here the author is a bit on firmer ground when he reviews a recent study published in the prestigious journal Lancet which demonstrated that many of the side effects that statins are accused of causing turn out to not be common when evaluated by scientific studies.
However, he builds straw man arguments throughout the article. For example, he notes that statins are accused by many of causing weight gain, and that the recent Lancet review shows there to be no statistically significant association between statin use and weight gain. But that neglects the argument that is made by proponents of this theory, which rests on three facts:
Statins are clinically shown to reduce GLP1 levels. This is unambiguously true and has been demonstrated in a well done randomized controlled trial. If the term GLP1 sounds vaguely familiar to you, it is a hormone that helps, amongst other things, to reduce hunger after eating and give you that “full feeling.” The current class of popular weight loss medications like Ozempic, Wegovy, and Mounjaro all work by giving patients a peptide that mimics GLP1, thus reducing hunger. Therefore, on a biochemical level, statins function as an opposite to something like Ozepmic, so it is very plausible that statins might lead to increased hunger and weight gain.
There is a trend in statin users toward an increased risk for type 2 diabetes. Why this is so is controversial, with many statin advocates pointing out – reasonably – that patients who are already on their way to developing diabetes are much more likely to be put on a statin by their doctors than patients who do not have blood sugar issues. Nonetheless, it would be consistent with a reduction in GLP1 levels for people to gain weight and see a rise in blood sugar levels, so again, this is a piece of evidence that would make sense if statins did in fact cause weight gain.
Finally, while studies do not show a STATISTICALLY SIGNIFICANT association between statin use and weight gain, they do show a TREND toward weight gain. The term “statistical significance” is important, because it implies that an association between two variables is so strong that we cannot assume it to be merely coincidental. However, the absence of a statistically significant relationship between two variables does not prove that there is no relationship between them. It only tells us that the relationship is not strong enough to be absolutely certain about. To use an analogy, if nobody witnesses Bob Smith committing a murder, it does not prove that Bob Smith didn’t commit the murder. It just reduces our ability to confidently state that he is the murderer. Statistical significance, in this analogy, is the presence of a witness. Its absence is not a complete guarantee of innocence.
Taking the three facts above, if statins lower GLP1 levels (which they do), and associate with an increased risk for type 2 diabetes (which they do) and there is a trend toward weight gain even if it does not reach statistical significance (which there is), it is in fact not crazy to be concerned that statins may cause weight gain.
The term “may” in the above sentence is important. Somebody who tells you with complete certainty that statins cause weight gain is getting ahead of the science, and you can very reasonably make the case that statins are unlikely to cause massive weight gain. However, someone who states that they see some lines of evidence to suggest weight gain might be a concern is being perfectly reasonable and expressing an opinion that is well supported by the known facts.
Which brings me to my second point about statins: they do have side effects and can sometimes, like all medications, cause harm. It’s not crazy or conspiratorial to be concerned about these possible side effects. This can be true even while acknowledging that hundreds of millions of people throughout the world take these medications, and the vast majority of them do just fine.
My take on statins then is that they have their place. They can lower the risk of heart attacks and strokes, and are well tolerated by the majority of people who take them. I do prescribe them when I feel it is appropriate, and would no more take them completely off the table for my patients than I would any other tool.
On the flip side, the overall benefit of a statin is not that large in most patients, and there are some valid reasons to have concerns about their side effects. Many patients can reasonably forego them.
The correct position to take regarding statins then is to carefully weigh their pros and cons in the context of each patient’s specific set of circumstances, and to factor in patient preferences, before deciding on when and how to use them. To me, neither the worst conspiracy theorists, nor the overly bullish author of this article, get the story correct.
