The study by Sumner and colleagues [1] confirms such metabolic consequences of gastric surgery as vitamin B₁₂ (cobalamin) and iron malabsorption. We agree with the editorial comment by Dr. Green [2] that achlorhydria and the intestinal blind loop could have caused bacterial overgrowth, resulting in malabsorption of vitamin B₁₂. In addition, some study patients could have been receiving histamine-2-receptor antagonists or been using proton pump inhibitors, which would aggravate malabsorption of vitamin B₁₂ [3]. However, Sumner and colleagues did not report which medication patients were receiving. A food (protein-bound) cobalamin-absorption test could have resolved some of the issues by diagnosing cobalamin deficiency, but this test is not readily available.

Sumner and colleagues do not mention whether their patients were inpatients or outpatients. This information would have been helpful because receiving a hospital diet for a few days can normalize serum folate levels and erythrocyte folate levels. Erythrocyte folate levels reflect the tissue store more accurately than do serum folate levels [4]. The authors do not provide their patients' history of alcohol intake; alcoholics can have normal serum folate levels but still have extreme folate deficiency. In some patients, cobalamin deficiency (as defined by Sumner and colleagues' second criterion under the definition of vitamin B₁₂ deficiency) could have been caused by deficiency in tissue. Although the homocysteine levels of these patients decreased after cobalamin was administered, some of the patients could also have received folate supplementation. Combining levels of serum homocysteine and methylmalonic acid has a high sensitivity for diagnosing cobalamin deficiency [5].

Although cobalamin and iron are frequently malabsorped in patients who have had gastric surgery, the conclusions of Sumner and colleagues (that is, a suggestion that vitamin B₁₂ replacement be used in patients who have had gastric surgery and who have vitamin B₁₂ levels less than 221 pmol/L) indicate that overdiagnosing cobalamin deficiency is still possible, especially in the absence of hematologic or neuropsychiatric manifestations.