Identifying if Regression to the Mean is a Problem in a Study

Abstract

Regression to the mean is a statistical phenomenon where initial measurements of a variable in a non-random sample at the extreme ends of a distribution tend to be closer to the mean upon a second measurement. Unfortunately, failing to account for the effects of regression to the mean can lead to incorrect conclusions on the observed mean difference between the two repeated measurements in a non-random sample above or below the population mean of the measured variable. Study designs that are susceptible to misattributing regression to the mean as intervention effects are prevalent in nutrition and obesity research. This field often conducts secondary analyses of existing intervention data and/or evaluates intervention effects in those most at risk (i.e., those with observations at the extreme ends of a distribution). Thus, unsubstantiated conclusions as a result of ignoring the effects of regression to the mean in nutrition and obesity research also are common. Here, we outline best practices for identifying the presence of regression to the mean using a flow chart available as a web-based app. We also provide multiple methods to quantify the degree of regression to the mean so investigators can adjust analyses to include the regression to the mean effect, thereby isolating the true intervention effect. The identification of regression to the mean and implementation of proper statistical practices will help advance the field by improving scientific rigor and the accuracy of conclusions.