Note that the abstract of underlying paper[1] is much more carefully drawing conclusions from the study. It mainly claims that ther exists an association betwee

Note that the abstract of underlying paper[1] is much more carefully drawing conclusions from the study. It mainly claims that ther exists an association between "climbing five flights of stairs daily" and an "over a 20% lower risk of ASCVD", but not of a direct causation (as the linked article does). Therefore the title of the article: "Walking more than five flights of stairs a day can cut risk of heart disease by 20%, study says" is misrepresenting the findings. When the study says that there is an association, the causation might also be the other way round, ie. a higher disposition of heart disease leads to people not climbing a lot of stairs. The last sentence of the abstract points into this direction: "Participants who stopped stair climbing between baseline and resurvey had a higher risk [32%] of ASCVD compared with those who never climbed stairs." When stopping stair climbing is worse correlated with ASCVD risk than not even starting stair climbing, one might suspect that there are other hidden causes that govern the correlation, such as people getting old or sick, which is their real cause for their increased ASCVD risk, and thus move into the basement or use the elevator more often, which in turn decreases their climbing stairs a lot. If this is the case, than there is no proof that climbing 50 steps a day can cut your risk of heart disease at all. We just have a correlation, no causation. To really find out whether there exists a causation, an intervention study would be necessary: we would need a representative sample of people of the particular group we are interested in (in this case people who do only climb very few steps a day). One half of the sample is not changing anything; the other half from now on consistently climbs 50 steps a day for a certain period of time. Afterwards we compare the two. (Ideally we would like to perform a double-blind experiment[2], but this seems impossible in this case.) [1] https://www.atherosclerosis-journal.com/article/S0021-9150(23)05221-8/fulltext https://www.atherosclerosis-journal.com/article/S0021-9150(2... [2] https://en.wikipedia.org/wiki/Blinded_experiment https://en.wikipedia.org/wiki/Blinded_experiment