In Reply: We disagree with the suggestion of Drs Geller and Adams that our observed association between PPI use and hip fracture may be confounded by the pandemic of vitamin D deficiency. By definition, a confounder must not only be associated with the disease of interest, it must also be associated with the exposure under study, and it must not be affected by the exposure or the disease.1 Geller and Adams support their concern about confounding of the association between gastric acid suppression and hip fractures by citing evidence of vitamin D status as a risk factor for hip fractures. However, by not considering how PPI therapy might relate to vitamin D status, they ignored 2 of the 3 essential elements that characterize a confounder.
Yang Y, Lewis JD, Epstein S, Metz DC. Proton Pump Inhibitor Therapy and Hip Fracture Risk—Reply. JAMA. 2007;297(13):1429-1430. doi:10.1001/jama.297.13.1429-b