Relative Risk, Absolute Risk, and Number Needed to Treat: How to Read a Statin Trial

Why does the same statin result sound big and small at once?

A headline like "statins cut heart attacks by a quarter" is reporting a relative risk reduction, and a relative number tells you nothing about how often the event happened in the first place. To judge the real size of a benefit you convert that percentage into an absolute risk reduction, which is the plain arithmetic difference in event rates between the treated and untreated groups, and then into a number needed to treat, which is simply one divided by that difference. Do this and the same honest result can read as impressive or unremarkable depending only on which of the three numbers someone chose to put in the headline. None of them is a lie. They are answers to different questions, and the gap between them is where most confusion about statins lives.

This is an educational article, not medical advice. My aim is to hand you a method you can apply to any trial, using statins because they are the most argued-over example in preventive medicine.

The three numbers, and how one becomes the next

Start with the raw material every trial reports: the event rate in each arm. Suppose a study finds that 4 percent of untreated people have a heart attack over several years, against 3 percent of treated people. Those two rates are all you need.

The relative risk reduction compares the two rates as a ratio. Going from 4 percent to 3 percent is a drop of one quarter of the original risk, so the relative risk reduction is 25 percent. That is the number most likely to reach a press release, because it is the largest-sounding of the set and it stays large no matter how rare the event is.

The absolute risk reduction is the subtraction: 4 percent minus 3 percent is 1 percentage point. This is the number that actually reflects what happened to the whole group, because it keeps the size of the underlying risk in view. A 25 percent relative reduction on a 4 percent risk and a 25 percent relative reduction on a 0.4 percent risk produce very different absolute benefits, even though the relative figure is identical.

The number needed to treat turns that absolute difference into people. It is one divided by the absolute risk reduction, so 1 divided by 0.01 gives 100. You would treat 100 people for the study's duration to prevent one heart attack, and the other 99 would have had the same outcome either way. The American Academy of Family Physicians published a worked version of exactly this chain, and it makes the practical point plainly: a 12 percent relative reduction in all-cause mortality it discusses corresponds to an absolute risk reduction of about 0.6 percent and a number needed to treat of 167 over roughly four years, which is the same fact wearing two very different outfits.

What the statin evidence actually shows

The reason statins are the standard teaching case is that their relative reductions are genuinely real and their absolute reductions are genuinely modest, so the framing gap is wide. A 2022 systematic review and meta-analysis in JAMA Internal Medicine pooled 21 randomized trials, each running at least two years with more than 1,000 participants, and reported both kinds of number side by side.

For myocardial infarction, statin therapy was associated with a relative risk reduction of about 29 percent, which sounds decisive. The absolute risk reduction across those same trials was about 1.3 percent. That absolute figure implies roughly 77 people treated for about four and a half years on average to prevent one heart attack, the exact translation the authors themselves make.

For all-cause mortality, the relative risk reduction was about 9 percent and the absolute risk reduction about 0.8 percent. For stroke, roughly 14 percent relative against about 0.4 percent absolute. In every row the relative number is the eye-catching one and the absolute number is the sobering one, and both come from the same trials.

Why baseline risk changes everything

Here is the part that a single headline can never carry: the same relative risk reduction produces a bigger absolute benefit in people who start out at higher risk. Relative reduction is roughly portable across risk levels; absolute reduction is not. Split the meta-analysis by prevention setting and this shows up directly. In primary prevention, treating people without established heart disease, the absolute reduction in heart attacks was about 0.7 percent. In secondary prevention, treating people who already have vascular disease and therefore far higher baseline risk, it was about 2.2 percent. The relative reductions in the two settings were closer together than those absolute figures, which is precisely why you cannot infer your own likely benefit from a relative number alone.

The JUPITER trial makes the same lesson concrete from the other direction. It enrolled apparently healthy people with normal LDL cholesterol but elevated C-reactive protein, and rosuvastatin cut the primary composite endpoint by a large relative margin, published in the New England Journal of Medicine in 2008. A follow-on analysis in Circulation: Cardiovascular Quality and Outcomes translated that into a five-year number needed to treat of about 20 for the broad endpoint, favorable next to older primary-prevention estimates. The relative effect was striking; the number needed to treat is what tells you how many people shared in it.

A short checklist for any trial

When you meet a treatment claim, ask four things. First, what were the actual event rates in each group, not just the ratio between them. Second, is the headline a relative or an absolute figure, and if only the relative one is quoted, treat that as a signal to go find the other. Third, what was the baseline risk of the people studied, because that is what decides whether a real relative effect becomes a large or a tiny absolute one for someone like you. Fourth, over what time horizon, since a number needed to treat of 77 over five years is a different proposition from the same number over one year.

None of this argues for or against taking a statin. It argues for reading the number that answers your question rather than the number chosen to impress you, and for remembering that relative risk reduction, absolute risk reduction, and number needed to treat are three views of one result, not three different results.