A Tutorial on Interaction

TLDR

Interaction between exposures can be examined on additive or multiplicative risk scales and is related to statistical models such as linear, log‑linear, and logistic models. The tutorial introduces interaction between exposure effects and evaluates arguments for using additive versus multiplicative scales to assess interaction. The authors review methods for presenting interaction analyses, including mechanistic forms, robustness to confounding, continuous and qualitative outcomes, attribution techniques, case‑only estimators, and power and sample size calculations for additive and multiplicative interaction.

Abstract

Abstract In this tutorial, we provide a broad introduction to the topic of interaction between the effects of exposures. We discuss interaction on both additive and multiplicative scales using risks, and we discuss their relation to statistical models (e.g. linear, log-linear, and logistic models). We discuss and evaluate arguments that have been made for using additive or multiplicative scales to assess interaction. We further discuss approaches to presenting interaction analyses, different mechanistic forms of interaction, when interaction is robust to unmeasured confounding, interaction for continuous outcomes, qualitative or “crossover” interactions, methods for attributing effects to interactions, case-only estimators of interaction, and power and sample size calculations for additive and multiplicative interaction.