A new study in the Lancet reports that while cocaine use is much more likely to trigger a non-fatal heart attack than air pollution (i.e., coarse particulate matter or PM10), air pollution is the more deadly trigger because more people are exposed to it.
This is study is faulty because:
- The strength of association between exposure to air pollution and non-fatal heart attacks is essentially zero — i.e., a relative risk on the order of 1.01, as compared to 23.7 for cocaine; and
- The study employs attributable risk calculations to make the determination that more heart attacks are triggered by air pollution than cocaine. But attributable risk is an invalid methodology for calculating population risk.
This study is what I would expect if epidemiology were taught in junior high school; it is not something that belongs in the Lancet, which apparently has not learned too much from its MMR-autism fraud scandal.