education · heart ·
Your Statin's Benefit Was Quoted As A Percentage. Here's The Other Percentage
Same trial, two true numbers. One sounds enormous and one sounds small, and which one you were told depends entirely on who was talking.
Here are two descriptions of the same trial result.
“Cut your risk by a third.” “Prevented one event for every 91 people treated for five years.”
Neither is a lie. They are the same data, and which one you heard tells you more about who was talking than about the drug.
This is the single most useful piece of numeracy a person can carry into a conversation about any medication, and statins are the clearest place to learn it.
The two numbers, in one paragraph
Suppose 2 people in 100 have a heart attack over five years without treatment, and 1 in 100 with it.
Relative risk reduction: 50%. The treated group had half the events.
Absolute risk reduction: 1 percentage point. Ninety-eight of those hundred people were never going to have a heart attack either way, and one was going to have one regardless.
Both are true. The first is the one that appears in headlines and drug advertising, because it does not change with the population and it is always the bigger-sounding figure.
What the actual statin data says
The Cholesterol Treatment Trialists collaboration pooled individual participant data from 27 randomised trials: 22 trials of statin versus control in 134,537 people, and five trials of more versus less statin (Mihaylova et al., The Lancet, 2012).
The headline relative figure: each 1.0 mmol/L reduction in LDL cholesterol reduced major vascular events by a rate ratio of 0.79. About a 21% proportional reduction, and it held largely irrespective of age, sex, baseline LDL or previous vascular disease.
Then the sentence that this whole article exists for. In people whose 5-year risk of a major vascular event was under 10%, each 1 mmol/L of LDL reduction produced an absolute reduction of about 11 major vascular events per 1,000 people over 5 years.
Eleven per thousand. Just over one percent, across five years.
In individuals with 5-year risk of major vascular events lower than 10%, each 1 mmol/L reduction in LDL cholesterol produced an absolute reduction in major vascular events of about 11 per 1000 over 5 years.
Mihaylova et al., The Lancet, 2012
And now the part we are obliged to tell you, because leaving it out would make us exactly what we criticise.
The authors who produced that number did not present it as evidence that statins are oversold. They used it to argue the opposite. Their next sentences say this benefit greatly exceeds any known hazards of statin therapy, that such people would not typically be considered suitable for statins under then-current guidelines, and that the guidelines might need to be reconsidered.
So the number is real, and the people who calculated it drew the expansionary conclusion from it. You can disagree with them. You cannot quote their figure while hiding what they concluded.
Why starting risk changes everything
The proportional benefit is roughly constant. The absolute benefit is not, because it depends on how much risk there was to reduce.
Compare the low-risk group above with people who already have coronary disease. In the Scandinavian Simvastatin Survival Study, 4,444 patients with angina or previous heart attack were randomised to simvastatin or placebo for a median 5.4 years (4S, The Lancet, 1994).
| Outcome | Placebo | Simvastatin | Absolute difference |
|---|---|---|---|
| Death | 256 (12%) | 182 (8%) | 4 points, about 1 in 25 treated |
| One or more major coronary events | 622 (28%) | 431 (19%) | 9 points, about 1 in 11 treated |
Same class of drug. Same kind of relative effect. A completely different proposition, because these people were going to have events.
That is the actual answer to “do statins work.” On average, yes, for major vascular events. Whether that is worth it for you depends on your absolute risk, your time horizon and the harms, and nobody can hand you that answer from a headline.
The side effect question, honestly
This is where our scepticism has to point in both directions, because the evidence here surprised us.
Fifty participants who had abandoned statins because of side effects were given twelve one-month bottles: four with atorvastatin, four with placebo, four empty. They rated symptoms daily. Sixty were randomised and 49 completed the year (Howard et al., Journal of the American College of Cardiology, 2021).
Mean symptom score was 8.0 in the no-tablet months. It rose to 16.3 in statin months. It also rose to 15.4 in placebo months, and the difference between statin and placebo was not significant (P = 0.388).
The symptoms were real. Taking a tablet, any tablet, roughly doubled them. The statin itself added almost nothing on top.
At six months after the trial, half the participants were back on statins.
This is not the same as calling anyone a liar, and we will not let it be used that way. These people felt what they felt. What the trial shows is that the sensation is poor evidence about its own cause, which is true of symptoms generally and is the reason blinded trials exist at all.
The harm side is not zero either. Across 13 trials and 91,140 people, statins raised new-onset diabetes with an odds ratio of 1.09, roughly one extra case for every 255 people treated for four years (Sattar et al., The Lancet, 2010). Small, real, and larger at higher doses.
What we are and are not saying
| Claim | Our line |
|---|---|
| We tell anyone to start or stop a prescription | Never |
| Relative risk reduction overstates personal benefit | Yes, and it is the number you are usually given |
| The absolute benefit is small in low-risk people | Yes. About 11 events per 1,000 over 5 years per mmol/L |
| The researchers behind that number think statins are oversold | No. They argued for wider use. We report that |
| The benefit is larger in people with existing heart disease | Yes. 4S found about 1 in 25 for death |
| Statins do not work | No. The proportional benefit is real and consistent |
| Muscle symptoms are imaginary | No. They are real, and mostly not caused by the drug |
| There are no harms | No. Diabetes risk is small but genuine |
Where the design shows up
The reason absolute risk is the honest number is that you are a person, not a population.
A relative risk reduction is a statement about a group. It tells you how the aggregate moved. It cannot tell you what happens to you, because most of the people in that group were never going to have the event, and the ones who were cannot be identified in advance.
That gap, between what is true of a thousand people and what will happen to one, is where almost all medical persuasion lives. It is not a trick. It is a genuinely hard problem. But the person being handed the prescription is entitled to see both numbers before deciding, and in most consultations only one of them gets said out loud.
“Buy truth, and do not sell it, get wisdom and instruction and understanding.” Proverbs 23:23, NASB
The verse carries the conviction. One hundred and thirty-four thousand participants carry both numbers.
What to do this week
Ask for the absolute number. The question is one sentence: out of 100 people like me, how many avoid an event over five years by taking this?
A good clinician will not be annoyed by that. It is the number they use to make the recommendation, and it is in every guideline risk calculator.
Then ask what the same calculator says if you change the inputs. Blood pressure, smoking, weight and activity all feed those risk scores, and moving them moves your absolute risk, which moves how much any drug can do for you. That is the lever this site is actually about.
And do not stop anything on the strength of an article, this one included. Coming off a cardiovascular medication has its own risks, and the way to do it is deliberately, with the person who prescribed it, for reasons you can both see in your own numbers. We covered what that process looks like in the rebound effect.
Read this next: how to grade a drug literature in general: two of medicine’s top editors say you can’t believe much of it. What happens when the pill count keeps rising: the prescribing cascade. And the lifestyle lever with the best trial evidence for reversing the underlying problem: the rebound effect and getting off medication.
Sources
- Mihaylova B, Emberson J, Blackwell L, et al. (Cholesterol Treatment Trialists’ Collaborators). “The effects of lowering LDL cholesterol with statin therapy in people at low risk of vascular disease: meta-analysis of individual data from 27 randomised trials.” The Lancet, 2012;380(9841):581-590. https://pubmed.ncbi.nlm.nih.gov/22607822/
- Scandinavian Simvastatin Survival Study Group. “Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S).” The Lancet, 1994;344(8934):1383-1389. https://pubmed.ncbi.nlm.nih.gov/7968073/
- Howard JP, Wood FA, Finegold JA, et al. “Side Effect Patterns in a Crossover Trial of Statin, Placebo, and No Treatment.” Journal of the American College of Cardiology, 2021;78(12):1210-1222. https://pubmed.ncbi.nlm.nih.gov/34531021/
- Sattar N, Preiss D, Murray HM, et al. “Statins and risk of incident diabetes: a collaborative meta-analysis of randomised statin trials.” The Lancet, 2010;375(9716):735-742. https://pubmed.ncbi.nlm.nih.gov/20167359/
Article metadata retrieved from PubMed. Photograph from Pexels. Scripture quoted from the New American Standard Bible.
Educational content from True Health Solutions. We do not diagnose or prescribe, and nothing here is a substitute for care from someone who knows your history. The body heals. We write about what supports that design.
Questions people ask
- What is the difference between relative and absolute risk reduction?
- Relative risk reduction is the proportional change between two groups. Absolute risk reduction is the difference in actual events. If 2 in 100 people have an event on placebo and 1 in 100 on a drug, that is a 50% relative reduction and a 1 percentage point absolute reduction. Both numbers describe the same result.
- How big is the absolute benefit of a statin?
- It depends entirely on your starting risk. In people whose 5-year risk of a major vascular event was under 10%, each 1 mmol/L of LDL reduction produced about 11 fewer major vascular events per 1,000 people over 5 years. In people who already had heart disease, the numbers are far larger.
- Does that mean statins do not work?
- No, and anyone using absolute risk to argue that is misusing it the same way the other side misuses relative risk. The proportional benefit is real and consistent. The question absolute risk answers is whether that benefit is worth it for a particular person, which is a different question.
- Are the muscle aches real?
- Sometimes, and less often than people assume. In a crossover trial where participants took statin, placebo and nothing in rotation, symptom scores were 16.3 on statin and 15.4 on placebo, with no significant difference. Most of the symptoms were real sensations that the tablet did not cause.
- Do statins cause diabetes?
- There is a small real increase. Across 13 trials in 91,140 people, the odds ratio for new diabetes was 1.09, which works out to roughly one extra case for every 255 people treated for four years. Higher intensity doses carry a larger increase.
