Our investigation confirms the known metabolic consequences of gastric surgery using the most sensitive available tests—measurement of serum methylmalonic acid and total homocysteine levels—to determine vitamin B₁₂ deficiency. We did not determine the cause of vitamin B₁₂ deficiency but agree with Drs. Murthy and Visweswaraiah that achlorhydria, intestinal blind-loop syndromes, and histamine-2 antagonist therapy could have caused some of the malabsorption. Histamine-2 antagonists are frequently prescribed and available over the counter. We did not test for food protein-bound cobalamin absorption. A limitation of all vitamin B₁₂-absorption tests is that they only measure malabsorption at the time of the test and not whether vitamin B₁₂ deficiency is actually present. Further, the tests are not standardized.

Approximately 50% of the patients who had gastric surgery were inpatients and 50% were outpatients. Of the patients with vitamin B₁₂ deficiency, 12 were outpatients, 5 were inpatients, and 2 were residents of a Veterans Affairs nursing home.

Participants were not treated simultaneously with vitamin B₁₂ and folate. Because we used elevation of serum methylmalonic acid levels and low or normal vitamin B₁₂ levels as our major criteria for defining vitamin B₁₂ deficiency, we do not believe that intake or problems in measurement of either serum or erythrocyte folate levels would have any effect on data. Only one case met the second definition of vitamin B₁₂ deficiency (normal methylmalonic acid level and abnormal homocysteine level). Treatment with vitamin B₁₂ decreased this patient's total homocysteine level from 22.8 µmol/L to 10.9 µmol/L. In addition to problems in standardization of the erythrocyte folate test, the test cannot distinguish between folate or vitamin B₁₂ deficiency. Therefore, the test would not add any diagnostic information to that discovered through methylmalonic acid and homocysteine assays.

We agree with the concerns of Drs. Murthy and Visweswaraiah that tests for serum vitamin B₁₂ level are not specific and that if this level alone is relied on, vitamin B₁₂ deficiency could be overdiagnosed. In our study, only 2 of 22 controls with low vitamin B₁₂ levels had metabolic evidence of vitamin B₁₂ deficiency. Because patients who have had gastric surgery have an increased risk for vitamin B₁₂ malabsorption, the number who were deficient was much higher; therefore, fewer patients with normal vitamin B₁₂ levels would be treated on the basis of only the serum vitamin B₁₂ level. Further, overtreatment with vitamin B₁₂ would not be harmful, and it seems unwise to risk having a patient develop possibly irreversible demyelinating disease of the nervous system or anemia. Our investigation shows the advantage of using measurement of serum methylmalonic acid and homocysteine levels with the measurement of serum vitamin B₁₂ and folate levels to gain specificity of diagnosis and to maximize the benefits of treatment.