The Short Answer: Relative Risk vs. Absolute Risk
A headline claiming something “cuts your risk by 50%” is reporting relative risk — how much bigger or smaller one group’s risk is compared to another’s. It says nothing about absolute risk, the actual chance of the event happening at all. The same result can be reported as dramatic or negligible depending on which number a source chooses, and both can be accurate at once.
An Evidence Ladder: From Definition to Real Numbers
Step 1: What Absolute Risk Means
Absolute risk is the real, standalone chance of an event in one group over a set time period. If 1 out of every 10 people with a given exposure develop a condition, the absolute risk in that group is 10%, or 1 in 10. It is a plain count, not a comparison to anything else.
Step 2: What Relative Risk Means
Relative risk compares one group’s absolute risk to another’s by dividing one by the other. In a frequently cited illustration, if 17% of smokers develop lung cancer compared with 1% of nonsmokers, the relative risk is 17 divided by 1, or 17 — smokers in that comparison were 17 times as likely to develop the disease. A relative risk of 1 means no difference between groups; above 1 means higher risk with exposure; below 1 means lower risk.
Step 3: The Same Data, Two Framings
In a hypothetical trial, 20% of an untreated group had a poor outcome versus 12% of a treated group. The absolute risk reduction is the plain gap: 20% minus 12%, or 8 percentage points — about 8 people helped out of every 100 treated. The relative risk is 12% divided by 20%, or 0.6, usually reported as a “40% relative risk reduction.” Both descriptions come from the identical dataset; the relative figure sounds larger, the absolute figure tells you how many people are actually affected. From that same absolute figure, you can also calculate a number needed to treat — roughly 100 divided by 8, or about 13 people treated for every one additional person who benefits.
Step 4: When a Big Ratio Sits on a Tiny Base
Relative risk can sound alarming even when absolute risk barely moves. The general annual risk of a seizure is roughly 0.057%. A specific medication that raises this to 0.16% produces a relative risk near 2.8 — reported as “2.8 times more likely.” The actual absolute increase, though, is about 0.1 percentage points, or roughly 1 additional case per 1,000 people per year. Both framings are accurate; they answer different questions.
Step 5: Exposure Can Also Lower Risk
The same math applies when an exposure is protective. In one example, 40% of people who exercised moderately were overweight or obese, versus 70% of those who did not. The relative risk is 40% divided by 70%, or 0.57 — meaning exercisers were about 43% less likely to be overweight or obese in that comparison (1 minus 0.57). A separate medication trial found 25% of exposed volunteers developed an upset stomach versus 2.4% of unexposed volunteers, a relative risk of about 10.4. In both cases, the relative number alone still hides the underlying group sizes and absolute rates.
What the Math Confirms — and What It Doesn’t Tell You
- Confirmed: relative risk and absolute risk are calculated differently and can move independently — a large relative change can sit on a tiny absolute change, or the reverse.
- Confirmed: absolute risk reduction converts directly into a number needed to treat, giving a concrete count of how many people must be exposed or treated for one additional outcome to change.
- Not something a relative-risk number alone tells you: your own baseline risk. The ratio is only as meaningful as the baseline group it was calculated against, and a different baseline can turn the same relative risk into a very different absolute effect.
- Not settled by this framework alone: whether a specific study’s population, dose, and follow-up period actually match your own situation — that’s a separate evaluation from the risk arithmetic itself. See our guide on confidence intervals and how precise a study estimate really is for the next layer of that question.
A Practical Checklist for Reading Risk Numbers
- Look for both numbers. If a source reports only a percentage change (a relative figure), ask what the actual, real-world rates were in each group being compared.
- Find the baseline. A relative risk means little without knowing the starting risk it was multiplied against.
- Check the time frame. Absolute risk only makes sense attached to a specific period — a year, five years, or a full study follow-up.
- Subtract the two absolute risks yourself if only one framing is given. That difference is the absolute risk reduction or increase, a number you can act on more directly than a ratio.
- Watch for one-sided framing: benefits are often reported as relative risk reduction (sounds larger), while harms are sometimes reported as relative risk increase (also sounds larger). Converting both to absolute terms puts them on equal footing.
As a quick decision rule: if a source gives you only a relative-risk percentage and no baseline rate, treat the claim as incomplete until you can find or calculate the absolute numbers behind it.
For background on how researchers classify the strength of a body of evidence before any of these numbers should be trusted, see Evaluating Evidence Quality. If the number you’re reading concerns a side effect rather than a benefit, our guide to how clinical studies measure and report safety data covers how those same risk figures are typically presented.
Frequently Asked Questions
Is absolute risk or relative risk more useful for a personal health decision?
For an individual decision, absolute risk figures — and the related number needed to treat — are generally more directly useful, because they describe how many real people out of a given group would actually be affected. Relative risk is useful for comparing the size of an effect across studies, but it depends heavily on the baseline risk of the group it was measured in.
Can two studies report very different-sounding numbers from the same underlying data?
Yes. A single result can be legitimately reported as, for example, a “40% relative risk reduction” or an “8 percentage point absolute risk reduction,” and both descriptions can come from the exact same dataset. Neither framing is incorrect on its own; the difference is which question each one answers.
What does “number needed to treat” mean?
Number needed to treat is a way of expressing an absolute risk reduction as a count of people. It is calculated by dividing 100 by the absolute risk reduction percentage, and it tells you roughly how many people would need to receive an intervention for one additional person to benefit.
Medical Disclaimer
This article is for general educational purposes only and does not constitute medical advice. It does not diagnose any condition or recommend starting, stopping, or changing any treatment. Discuss how any specific study’s findings apply to your own health with a qualified healthcare provider. If you are experiencing a medical emergency, contact your local emergency services immediately. This guide was prepared by the ClinicalStudyConnect.com Research Desk.