Marcuard and coworkers [1] have failed to show that omeprazole therapy causes the apparent changes in cyanocobalamin absorption. The authors used a modified Schilling test intended to evaluate absorption of protein-bound B₁₂, not free B₁₂ (cyanocobalamin). The absorption of free B₁₂ does not depend on acid. Further, they state in the abstract that "each volunteer served as his own control."

Whether such a form of control exists can be debated. Marcuard and associates [1] have actually compared the change from baseline between two groups of patients, each receiving the same interventions, including daily administration of omeprazole, but at a dose of 20 mg or 40 mg. No placebo group was identified, and therefore no opportunity existed for comparison of differences in mean results between treated patients and a true control group receiving the same contemporaneous interventions. Results of repeated normal and protein-bound cobalamin Schilling tests in untreated and treated patients are necessary to determine whether actual cyanocobalamin malabsorption occurred, or if cobalamin liberation from food protein failed. The latter is correctable by oral vitamin supplementation.

Other potential sources of bias include the unblinded administration of drug and of the various tests and analyses. Because the participants varied considerably in results of B₁₂ absorption tests that were done before treatment (with only minor absolute changes reported for several patients), the conclusions would be substantially affected by inapparent small biases caused by inappropriate study design. In conclusion, the effects of drugs can be determined only by double-blinded, randomized, controlled trials.