Regression to the mean

Regression to the mean is a statistical phenomenon: when a value is measured and happens to be unusually high or low, a second measurement of the same thing tends to fall closer to the group average, because random fluctuation partly drove the first extreme reading. It can masquerade as a treatment effect. If people are enrolled in a study because their blood pressure, pain score, or lab value was extreme, many will improve on retesting regardless of any intervention. When reading a study, ask whether participants were selected for having extreme baseline values and whether a control group was included: without a comparison group, apparent improvement may be regression, not the treatment working. For example, a supplement given to people with the highest cholesterol readings may look effective simply because those readings drift down on retest. A randomized control group experiences the same regression, so a real effect is the difference between the arms, not the change within one arm.