Aspirin Trial Mocked Astrological Subgroups

In 1988, the landmark ISIS-2 trial studied 17,187 suspected heart attack patients and found that aspirin reduced five-week vascular mortality by 23 percent. When journal editors requested subgroup analyses, the researchers divided patients by their 12 astrological signs to demonstrate how random chance could create misleading results. The analysis appeared to show that aspirin provided no benefit to Libras and Geminis, despite its conclusive overall lifesaving effect. The deliberately absurd comparison became a famous warning against trusting post hoc subgroup findings.

Previous Fact Next Fact