When Numbers Rise Together

A city health office once noticed that barangays with more fast-food outlets also had higher rates of high blood pressure. A councilor quickly proposed limiting new fast-food permits, arguing that the restaurants were making people sick. The proposal was popular, but a statistician on the staff urged caution. The two figures rose together, she explained, but that did not prove that one caused the other. Barangays with many fast-food outlets also tended to be the most densely populated and the most commercial; their residents worked longer hours, slept less, and walked less because they commuted by vehicle. Any of these factors could raise blood pressure on its own, and any of them could also attract restaurants. A third factor that drives two others in this way is called a confounding variable. The statistician did not claim that fast food was harmless. Her point was narrower: before restricting businesses, the city should gather evidence that could separate the effect of diet from the effects of stress, sleep, and inactivity, for example by comparing residents of similar barangays who differ mainly in what they eat. Policies built on mistaken causes, she warned, can burden the public without improving anyone's health.