Impact of treatments or genotype on length of the small intestine (mean and standard deviation) panel A. Open symbols for DKO sets (Splenda, Metronidazole, vancomycin, and streptomycin) are males, dark symbols are females. For the non‐DKO mice, open symbols are Gpx1+/−Gpx2−/− males, dark symbols are Gpx1+/−Gpx2−/− females and half‐tone symbols are Gpx1−/−Gpx2+/− males. The values for non‐DKO mice resemble results from wild‐type mice. Statistics 1‐way ANOVA with Tukey's multiple comparisons test. Common letters indicate no significant difference, otherwise a > b, etc. Representative ileum histology (H&E) (panels B‐F). For histology, the original magnification was 10×. Non‐DKO mice (panel C) display histology nearly identical to wild‐type B6 with dense and orderly crypt/villus units (Esworthy et al., 2014). Most crypts have abundant Paneth (vertical white arrows) cells and goblet cells are visible (black arrowhead points to clusters). The control Splenda DKO mice (panel B) have disordered and distorted crypt/villus units that are less densely packed and lack Paneth and goblet cells. This particular image shows a probable crypt abscess (vertical open arrow). Panels D‐F show the generally beneficial impact of antibiotic treatment with some restoration of Paneth and goblet cells in better ordered crypt/villus units. These panels also show that some pathology remains, for example, less densely packed crypt/villus units in the metronidazole set (panel D)