Chunk #49 — Reasons to be Concerned about the Published cGxE Literature — Problems with the Recipe: Statistical Concerns in cGxE Research — Power to detect and characterize different types of interactions
an ostensibly “significant” interaction is usually of a “cross-over” (i.e., disordinal) nature, especially when samples sizes are small (Sher & Steinley, 2013). Boardman et al. (2014) recently made a similar observation in reference to the emerging genome-wide gene-by-environment (GWGEI) approach (Cornelis et al., 2012; Mukherjee, Ahn, Gruber, & Chatterjee, 2012; Thomas, Lewinger, Murcray, & Gauderman, 2012) by demonstrating that when many interaction tests are performed, the most significant p-values will come from disordinal interactions even when such interactions are generated from random data (J.D. Boardman et al., 2014). Boardman et al. note these findings “conform to the differential susceptibility model but will not tell us anything meaningful about the way in which environments systematically moderate genetic factors…because they will likely be a statistical artifact” (p. 123).
