Course materials should teach exposures as correlated histories
For the gene–environment interaction course materials, the environmental side should not collapse into one exposure measured once. A stronger teaching example would pair correlated agents—such as traffic-related air pollution, noise, heat, and neighborhood deprivation—with susceptible windows, exposure duration, and latency. The central test is whether an interaction persists after mixture-aware confounding control and alternative timing specifications, rather than whether a single product term reaches significance. Does the material distinguish biological interaction from statistical interaction and address exposure measurement error shared across mixture components?
