The widespread use of statins is the end-point to the postulate, starting early last century, that high serum cholesterol causes heart disease. While there are no different types of cholesterol in this context, there are different lipoproteins that carry them around the body. Cholesterol itself will not dissolve in blood (fat and water, eh), so it is carried around in a protein and phospholid encapsulated ‘boat’, a lipoprotein. There is a distribution of sizes for the lipoproteins, so we get the well known LDL (low-density lipoprotein – larger package) and HDL (high-density lipoprotein – smaller package) amongst others (Chylomicrons, VLDL, IDL). But the cargo is the same – the cholesterol in each lipoprotein is not a different thing. In this postulate, elevated LDL (and perhaps decreased HDL) causes increased risk of coronary disease. Of note, all of the evidence linking high LDL to atherosclerosis is epidemiological; as far as I know, there’s no RTC or concrete causative pathway. However, if we take the correlation as causation and ‘ground truth’, the rush was on in the 1950’s to identify drugs to lower serum cholesterol.

As an aside, there is little correlation between dietary cholesterol and serum cholesterol levels. That’s because your body produces at least 80% of the cholesterol in your system. Cholesterol is an absolute necessity for life – you will die without it. It’s critical for cell membrane construction and maintenance, hormone production, bile fluid, etc. It’s too important to be left to exogenous sources, so we make our own, largely in the liver. So even if you want to lower cholesterol, there’s no particular reason to stop eating eggs – your body is smart, knows what it needs, and will make more of it.

A lipoprotein. The different ‘types’. Their function.

 

Given that we make our own cholesterol, if you want to reduce it pharmacologically you need to modify the production chain. The first drug to reduce serum cholesterol was Triparinol, released in 1959. It was withdrawn in 1962 as it was causing cataracts and accelerated heart disease (!). After the fact, there was some evidence (and court case to boot) that the dangers of the drug were known from the pre-approval trials and were simply not reported or were ignored – not sure where I’ve heard this before, but it seems familiar… Triparinol was not a statin – its method of action is to inhibit the final step in cholesterol creation.  But if one only inhibits the final stage of a multi-part process, you get a build-up of intermediate products and that  build up was probably the root cause of the adverse affects. Complex systems, who’d of thunk?

On the other hand, the statin class of drugs inhibits the rate controlling enzyme (HMG-CoA reductase) for cholesterol generation, hence they slow down the entire chain of processes that create cholesterol, not just truncate a single step. The first statin (mevastatin) was isolated from a fungus in Japan (what is it with the Japanese and fungi?). In the late 70’s Sankyo (Japan) and Merck (US) were in a race to bring a statin to market. No too long after the start of trials in dogs, rats, and monkeys, Sankyo halted development due to significant adverse events in the their trials. There was little success in actually lowering cholesterol in rats, and over 50% of the dogs in the trials developed liver cancer. While there has been criticism that the dosing was extremely high (poison is in the dose), within months, Merck pulled out of development of their drug as well (lovastatin, also isolated from a fungus and building on the work of the Japanese in the 70’s – following a trip by the Merck CEO to Japan in the late 70’s). However, Merck did continue research on lovastatin and the statin class of drug, and, 5-6 years later (1987), got FDA approval for lovastatin, marketed as “Mevacor” (AKA Altocor, Altoprev), the first commercial statin drug.

Atorvastatin (marketed as Lipitor) was patented in 1986 by Pfizer and approved for use in 1996.

Cerivastatin (marketed as Baycol, Lipobay) was developed and marketed by Bayer A.G. as a competitor for Lipitor. It was withdrawn in 2001 due to a high incidence of rhabdomyolysis – rapid break down of skeletal muscle which is potentially fatal as it can lead to kidney failure. The rate of rhabdomyolysis with Baycol was 50-80 times higher than with other statins.

That’s right, “than other statins” meaning all statins have this side effect. Of course with any medication, side effects will happen (except vaccines – they are known to be 100% safe). With statins, all the various products list a common set of side effects – not surprising as they all have the same method of action. Side effects include increased diabetes risk, high serum liver enzymes (liver damage), muscle pain, reduced cognitive function, the aforementioned rhabdomyolysis, weakness, blurred vision, “insert fast talking voice at the end of all pharma commercials”. While “side effects are rare” (except in cases where they aren’t…), as many as 30% of users experience muscle pain; Generally, 20% of patients discontinue statin use due to ‘intolerable’ side effects.

One of the original lipitor ads. All statin adds follow the same M.O.

Of course, if the effectiveness of statins in preventing heart disease is high, side effects may be tolerable. What is the effectiveness? Most people have probably heard the claim that regular statin use reduces risk of heart attacks by 30%.  The fine print is actually relatively big in this lipitor add, but it’s often not. They are talking about relative risk, not absolute risk. In this specific case, reduction in events from 3% (control) to 2% (treatment). So it’s the tried and true tradition of reporting relative risk instead of absolute risk to make your intervention appear to be more effective than it is. In general, for all statins, one only sees a 1% reduction in heart attack risk, so one needs to treat 100 people to avoid 1 heart attack. In fact, the statin advertisements were one of the driving forces behind the ‘strong recommendation’ from the CDS and FDA that absolute risk should be reported for all drug interventions; advice that seems to have been forgotten recently for some reason…

On it’s face, 1% is not a great benefit – one might still choose it, but it’s a bit different calculus than what people think of when they hear 36% reduction. However, that number doesn’t even account for a variety of other factors. One factor is that essentially all of that ‘benefit’ comes from a reduction in secondary events – i.e. there seems to be essentially no benefit in preventing a first heart attack, all the benefit appears to be in the cohort that has already had a cardiovascular event. A second factor is that we are looking at only one metric for a positive outcome, heart attacks. But what we are really interested in is overall health; not having a heart attack is small comfort if you instead die from some other adverse impact from the statin. There was a similar bias in a recent set of trials for some vaccine or something. The only end point was whether the vaccine group was protected from contracting the virus – and even that number was gamed by the study design! One can see the folly of that approach with a simple, though extreme, thought experiment. Say one develops a cancer treatment and goes to trial in an RCT with the only end point metric being the remission of cancer in the control versus treatment. At the end of the trial, we find that 10% of the control group died of cancer and 0% of the treatment group died of cancer. Success! Except that 50% of the treatment group died from massive organ failure. But by the study end-point metric, the trial was a success.

There have been some meta analyses that try to tease this out and it tends to be a wash for statins: “The Effect of Statins on Average Survival in Randomized Trials, an analysis of the end point postponement”, Kristensen et. al. 2015, BMJ Open and “Evaluation of Time to Benefit of Statins for the Primary Prevention of Cardiovascular Events in Adults Aged 50 to 75 Years: A Meta-Analysis”, Yourman et. al. 2021, JAMA Intern Med.  (fights urge to put links in footnotes…)

Part of the reason why statins may not provide an overall mortality benefit, let alone when weighed against quality of life, is that their primary affect – namely lowering of LDL cholesterol might  actually be  harmful. If LDL is not implicated in increased risk of heart disease and lowering it has other deliterious effects, overall health will be reduced on a statin. Indeed there is some evidence that the very small benefit of statins in reducing heart attacks is not due to their effects on cholesterol, but rather on anti-inflammatory properties, especially vascular inflammation.

In any case, that was my main thought putting together this article – cholesterol has gotten a bad rap and we’re off chasing our tails trying to reduce it by dietary interventions and pharmacologically and doing great harm, especially from the dietary side (statin use seems to be relatively harmless – I wouldn’t put too much of my money on that proposition though – if potentially unpleasant) in the process. But what was intended to be a short section on statins and why their cholesterol lowering effects may not be a good thing, became very long. I’ll stop now and leave a discussion about why the dominance of the diet-heart hypothesis in the medical orthodoxy might have had and continues to have long term negative impacts on human health and well being to another time!