A meta-analysis published this week in the American Journal of Cardiovascular Drugs found that people who regularly climb stairs were 39% less likely to die of cardiovascular disease and 24% less likely to die of anything at all, compared with people who rarely climb them. Nine studies, 480,479 participants, a median of 14 years of follow-up. The University of East Anglia put out a release on 14 August and the story travelled fast.
Here is the part almost nobody mentioned. Those exact numbers — 39%, 24%, nine studies, 480,479 people, ages 35 to 84, 53% women — were presented by the same UEA team at ESC Preventive Cardiology in Athens on 26 April 2024, under the title “Evaluating the cardiovascular benefits of stair climbing.” The European Society of Cardiology issued a press release that day. It carried the same lead quote from Dr Sophie Paddock that ran in this week’s coverage: “If you have the choice of taking the stairs or the lift, go for the stairs as it will help your heart.”
Nothing in the evidence base changed between then and now. What changed is that the abstract cleared peer review and became a journal article. That is a real and useful milestone. It is not a new finding, and it is not new information about stairs.
The number the coverage dropped
The 2024 ESC release published the effect sizes with their confidence intervals. The all-cause mortality result was a relative risk of 0.76, 95% CI 0.62–0.94, p=0.01. That interval is the story. The headline says “24% less likely to die.” The data are consistent with anything from a 38% reduction down to a 6% reduction — and 6% is close enough to nothing that a single unmeasured confounder could erase it.
The cardiovascular-death result is firmer: RR 0.61, 95% CI 0.48–0.79, p=0.0002. Tighter interval, larger effect, and it does not cross the line. If you want one number from this paper, that is the one to keep.
Our take: Nobody randomised anyone onto a staircase. This is nine observational cohorts pooled together, and the obvious confounder is the entire mechanism — the people who habitually take the stairs are the people who can take the stairs. Cardiorespiratory fitness, joint health, body weight and absence of existing disease all determine stair-climbing before stair-climbing determines anything. The honest claim is that regular stair use is a good marker of a healthy body, and a plausible, free, low-friction way to add activity. The dishonest claim is that a staircase cuts your risk of cardiac death by 39%.
What the paper does and does not tell you
The authors’ own caveats are worth more than the headline. The analysis pooled studies “regardless of the number of flights of stairs and the speed of climbing” — so there is no dose. One cohort inside it pointed at roughly six flights a day, but the team says plainly that the ideal dose is not established and that future work should use wearables rather than self-report.
Self-report is the quiet weakness. Every input rests on people telling researchers how often they take the stairs — exactly the kind of question healthy people answer generously. Same structural problem we flagged in the orange-juice gene-expression study: soft measurement, hard write-up.
What to watch
- Wearable-measured replication. Accelerometer and altimeter data from cohorts like UK Biobank can count flights without asking anyone. Until someone runs that, the dose question stays open.
- Absolute risk, not relative. A 39% relative reduction on a small baseline risk is a small absolute reduction. No coverage this week put an absolute number on it. That number is the one that would tell you whether to care.
- Conference-to-journal lag as a news trap. Abstracts get press releases, then the journal version gets a second press release two years later. If a study feels familiar, check whether you already read it.
None of which is an argument against taking the stairs. They are free, already in the building, and carry no downside for anyone without a joint or mobility problem. Take them because it is easy — not because a number from April 2024 got a fresh coat of paint. Run any health headline through The Study Filter first.