Nonparametric Estimation of the Potential Impact Fraction and the Population Attributable Fraction With Individual‐Level and Aggregated Data

The estimation of the potential impact fraction, including the population attributable fraction, with continuous exposure data frequently relies on strong distributional assumptions. However, these assumptions are often violated if the underlying exposure distribution is unknown. In this article, we discuss the impact of distributional assumptions in the estimation of the population impact fraction, showing that distributional violations lead to biased estimates. We propose nonparametric methods to estimate the potential impact fraction for aggregated data, where only the exposure mean and standard deviation are available, or individual data, where the full exposure distribution can be estimated from a sample of the target population. The finite sample performance of the proposed methods is demonstrated through simulation studies. We illustrate our methodology with a study of the impact of eliminating sugar-sweetened beverage consumption on the incidence of type 2 diabetes in Mexico. We also developed the R package pifpaf to implement these methods.