Teaching examples for internal validity: published experiments later shown to be confounded in an interesting way

Question

There are some good examples of interesting/fun spurious correlations here (Examples for teaching: Correlation does not mean causation). Many/most of these are "funny" examples that aren't likely to affect real life decision making. In contrast, this thread (Real examples of Correlation confused with Causation) has some interesting examples of real-life misinterpretation of spurious correlations.

I'm wondering if anyone has cool/interesting favorites along the lines of the second thread, for a related-but-different phenomenon: published experiments (i.e., not correlational data) whose results were later discovered to be confounded in an interesting/unexpected/memorable way.

Answer 1

There is one example explained in Causality (2009) p. 358:

... a study involving injected vitamin E as a treatment for incubated children at risk for retrolental fibroplasia. The randomized experiment indicated efficacy for the injections, but it was soon discovered that the actual effective treatment was opening the pressurized, oxygen-saturated incubators several times per day to give the injections, thus lowering the barometric pressure and oxygen levels in the blood of the infants.

The problem here was that experimenters were not only doing $X$ (vitamin E injection), but doing $X$ and $Z$ (vitamin injection and opening the incubator), thus what they thought was the effect of one intervention ($X$) was in fact the effect of another ($Z$). This could have been detected with a placebo group in the study.