Statins and Big Pharma: The Argument Behind the Prescription

This is a history of an argument, and it should be read as one. It is not advice about any medicine, and nothing here is a reason for anyone to begin or stop taking anything; that is a conversation for a person and their doctor, over the actual facts of one body. What follows is the record of a debate that ran through the medical journals and the newspapers across roughly forty years — a debate about a class of pills prescribed to tens of millions of people, and about who paid for the evidence that put them there. The suspicion that gathered around statins was not stupid. Much of it grew from things that were genuinely true. The most interesting way to understand it is to build the sceptic’s case at its strongest, honestly, before asking where it holds and where it gives way.

The strongest version of the doubt

Start with the discomfort at the centre of it, because it is real. A statin is a drug most people take every day for the rest of their lives, often while feeling perfectly well, to lower a number most of them could not explain. It is prescribed at enormous scale. And a great deal of the evidence that established it was funded, run, and analysed by the companies that stood to sell it. Anyone who looks at that arrangement and feels a flicker of unease is responding to something that ought to produce unease.

The financial stakes were vast. Pfizer’s atorvastatin, sold as Lipitor, became the best-selling prescription drug in the history of the world, taking well over a hundred billion dollars across its patent life. When a product earns that much, the firm that owns it has every incentive to see its market expand — to see the threshold at which a doctor reaches for the prescription pad drop lower, and the pool of eligible patients grow wider. That is not a conspiracy theory. It is simply how the incentive points, and a careful person watches where incentives point.

Then there is the question of the data itself. For years, the raw patient-level results of the major statin trials were not open to independent researchers. They were held by the Clinical Trial Service Unit at Oxford and by the trial sponsors. Outsiders who wanted to check the safety findings for themselves — to count the muscle aches and the new diabetes diagnoses in the source records rather than in the summary tables — could not get at the numbers. In any other setting we would call that a reason for caution. If a party with a financial interest in a conclusion also controls the evidence for it and will not let you audit the evidence, scepticism is the correct response, and it would be wrong to dismiss it as a character flaw.

The steelman goes further, into the arithmetic. Much of the marketing of statins leaned on relative risk reduction, which flatters. Telling someone a pill cuts their heart-attack risk by “36 per cent” sounds transformative. But in primary prevention — treating people who have never had a heart attack or stroke, who are simply judged to be at some future risk — the absolute numbers are far smaller. If your five-year risk of an event is low to begin with, a large-sounding relative cut can translate into a very small change in your actual odds: a percentage point or two, sometimes less, meaning many people take the drug for years so that a few avoid an event. Present the same result as absolute risk and it looks different, and less overwhelming. A sceptic who insisted the public was being sold the flattering number over the honest one had a point that statisticians agreed with.

The trials the sceptics pointed to

The doubt was not free-floating. It fastened onto specific studies, and those studies deserve to be named accurately.

The foundational trials were genuinely striking, and the honest sceptic concedes them. The Scandinavian Simvastatin Survival Study — 4S — reported in The Lancet in 1994, and it was a landmark. It followed more than four thousand patients who already had heart disease and found that simvastatin reduced deaths substantially. The West of Scotland Coronary Prevention Study, WOSCOPS, published the following years, extended the finding toward people who had not yet had an event. These were large, long, and hard to wave away.

The arguments came later, over the newer trials and the wider claims built on them. JUPITER, published in 2008 and led by the Harvard cardiologist Paul Ridker, tested rosuvastatin in people whose cholesterol was not especially high but who had raised levels of an inflammatory marker, C-reactive protein. It reported a large relative benefit and was used to argue for treating a broader, healthier population. Critics noted that the trial had been stopped early — trials halted early for benefit can overstate the effect — and that Ridker held a patent interest in the C-reactive protein test the study helped promote. That was a real conflict, disclosed, and worth weighing.

ACCORD, reported around 2010, sat in the background of the intensity debate: a large trial in people with diabetes whose results complicated the assumption that pushing risk factors ever lower always paid off. And the flashpoint that turned a specialist disagreement into a public row came in 2013, when the American College of Cardiology and the American Heart Association issued a new risk calculator to decide who should be offered statins. Independent researchers, including the Harvard-based physician John Abramson, argued the calculator badly overestimated risk and would sweep millions of low-risk people into lifelong treatment. That mattered, because where you set the threshold decides how many healthy people take a drug they may barely benefit from.

The row that spilled into the newspapers

Then, in 2013 and 2014, the British Medical Journal published two articles that lit the fire in public. One was by Abramson, the other by the cardiologist Aseem Malhotra. Both cited a figure suggesting that side effects from statins occurred in around eighteen to twenty per cent of users — a startling rate, far higher than the trials reported. If true, it would mean the harms had been badly undercounted while the benefits were trumpeted.

The figure was wrong. It came from a misreading of an observational study, and both papers later issued corrections withdrawing it. But the argument did not end there, and this is where the steelman has to be careful and fair to both sides. Rory Collins, the Oxford epidemiologist who led the Cholesterol Treatment Trialists’ Collaboration — the group that pooled the trial data — reacted forcefully. He argued the erroneous side-effect figure was dangerous, that it would frighten people into stopping a drug that was saving lives, and he asked the BMJ to retract the articles. The journal’s editor, Fiona Godlee, convened an independent panel, which declined to retract; the correction of the specific number, the panel judged, was enough.

Underneath the personal heat was the substantive disagreement, and it is the honest core of the whole affair. Collins and the CTT held the pooled trial evidence and argued that in the randomised trials, side effects beyond a small excess of diabetes and rare muscle problems were genuinely uncommon — that the drugs were, on the trial record, well tolerated. The sceptics answered that the trials could not be fully checked because the raw data were not open, that real-world patients differed from trial volunteers, and that clinicians were seeing muscle complaints the summary tables did not capture. Two serious camps, one insisting the evidence was solid and the harms small, the other insisting the evidence could not be independently verified and the harms were being waved away. Both were describing something true. The evidence was strong. It also was held behind a wall.

Where the strong version stops holding

Having built the sceptic’s case at full strength, honesty requires walking to the place where even the strong version thins out — because the popular slogan that “statins are a Big Pharma scam” asks the individual truths to carry a weight they cannot bear together.

The first strain is the leap from conflict of interest to fabrication. Industry funding is a genuine reason to scrutinise a result harder; it is not, by itself, evidence the result is false. Compromised money and a correct finding can, and often do, coexist — the way to tell them apart is independent replication, and the cardiovascular benefit of statins in people who already have heart disease has been replicated across many trials, run in many countries, including ones with no commercial sponsor. When a finding survives that much independent repetition, “the company paid for it” stops being a sufficient answer. The conflict is real; the conclusion outran the conflict.

The second strain is the withheld-data argument, which is strong precisely where the scam framing is weak. It is entirely fair to say the raw statin data should have been open years earlier, and the campaign for open trial data — the AllTrials movement, launched in 2013 by Ben Goldacre and others — was in the right and largely won its case. But “we could not fully audit the data” supports one conclusion: reserve some judgement, demand access. It does not support the very different conclusion that the drugs therefore do not work. A locked archive is a reason to withhold certainty; it gives no licence to assert the opposite. The scam story quietly converts an absence of full verification into a positive claim of fraud, and that conversion is where it breaks.

The third strain is the side-effect question, and here a later trial did something rare in these disputes: it tested the disagreement directly. The SAMSON trial, reported in 2020, took people who had stopped statins because of side effects and gave them, in random order and blinded, months of the real drug, months of an identical dummy pill, and months of nothing. The striking finding was that most of the symptoms people suffered on the statin, they also suffered on the placebo. The effect was largely a nocebo — real, felt, distressing symptoms, genuinely experienced, but produced by the expectation of harm as much as by the chemistry. That does not mean nobody has a true drug reaction; some clearly do, and muscle problems and the small diabetes signal are real. It means the blanket claim that statins routinely poison their users did not survive being tested against a sugar pill.

Why the argument was never really about cholesterol

Notice what happens when the strong case is separated from the slogan. What survives is a set of legitimate, specific complaints: that a profit-driven industry funded and controlled too much of its own evidence, that benefits were sold in the flattering language of relative risk, that the primary-prevention threshold is a genuine judgement call where reasonable doctors still differ, and that the data stayed locked far too long. What does not survive is the sweeping conclusion the slogan wants — that the whole enterprise is a fraud and the pills do nothing. The reasonable parts are real. The unreasonable whole was assembled from them by a kind of emotional logic that is worth understanding rather than mocking.

Because the distrust did not come from nowhere, and it did not come only from statins. It arrived carrying the memory of times the suspicion was vindicated. The public had watched an industry get the science of nutrition badly, expensively wrong before — the decades in which dietary fat took the blame that sugar deserved, and the egg was cast as a villain on cholesterol evidence that later softened. It had learned, from the tobacco and sugar industries’ documented playbook of manufacturing doubt, that a company will fund the science that flatters its product and bury the science that does not. And it carried older, deeper wounds — the memory of thalidomide, a real pharmaceutical catastrophe in which a drug company’s assurances of safety were catastrophically false. A person who reaches for the statin bottle with that history in their head is not being paranoid. They are remembering.

That is the human core of it. The statin sceptic is often someone who has correctly learned that authority must be checked and that money bends evidence, and who has applied that hard-won lesson one step too far — from scrutinise the claim to reject the drug. The lesson is sound. The over-application is where the harm creeps in, because the same reasoning that rightly demands open data can, if it is not disciplined, talk a person who genuinely needs a proven medicine out of taking it.

The record does not resolve into a verdict to hang over anyone’s daily decision, and it should not be asked to. What it leaves is a sharper way of holding two things at once: an industry can behave in ways that fully earn our suspicion, and a medicine it sells can still, on independently repeated evidence, do roughly what it claims for the people it was tested in. Learning to keep both of those true at the same time — to distrust the seller without disbelieving the tested result — is harder than either slogan, the trusting one or the cynical one. It is also the only honest place the argument ever led.