Reply

Abstract

Editors note: The above letter was referred to the authors of the original paper, and their reply follows. In reply: We appreciate Dr. Norman's feedback on our article.1However, after reviewing the available evidence,2-6 we found that there is no‘gold standard' test at this time that will identify true vitamin B12tissue deficiency. The most commonly used standard is serum methylmalonic acid and/or homocysteine elevation greater than 2 standard deviations, as was utilized in our study1 and as has been utilized in several other clinical studies.7, 10There are limitations to these measurements, as pointed out by Dr. Norman. However, we find no evidence of consensus on a gold standard for measuring tissue vitamin B12deficiency. If anything, the leaning is toward serum assessments.” A low (22%) positive predictive value of below normal serum vitamin B12level to identify vitamin B12deficiency was reported by Matchar2 when the diagnosis of B12deficiency was made by clinical criteria. This argues that a low serum B12level, is, by itself, not sufficient to make the diagnosis of vitamin B12deficiency. Dr. Norman has reported B12deficiency, defined by urinary methylmalonic acid (MMA) >5 /μ.g/mg creatinine, using a spot urine specimen.3However, these values of MMA were not correlated with clinical manifestations of true B12 deficiency. In the same study, five of 35 individuals (14.2%) had elevated urinary MMA levels despite normal serum B12levels. This evidence contradicts the claim that urinary MMA/creatinine ratio is the gold standard for true vitamin B12tissue deficiency. One can anticipate several limitations to the use of a spot urinary MMA/creatinine ratio in determining vitamin B12deficiency. Chanarin4, 5 showed evidence 30 years ago that all B12deficient patients had low serum B12levels and most, but not all, had elevated urinary MMA levels by 24-hour urine collection, but not by a spot urine sample. Rasmussen12 reported that a protein meal caused a significant stepwise increase in the urinary excretion of MMA. The urinary concentrations of MMA varied as measured on multiple consecutive 3-hour urine collections, apparently because of variations in diuresis during the day. These findings suggest that a complete 18- to 24-hour urine collection is necessary for valid estimation of urinary excretion of MMA.13In clinical practice, however, 24-h urine collection is difficult. As pointed out by Rasumssen,13 there are several points in favor of serum assays of MMA for an indication of vitamin B12deficiency. These include: 1 there is no need for an additional assay for creatinine; 2 the dietary influences in serum in normal persons are far less marked than in urine12; and 3 measurements can be done on stored serum samples drawn at the time of B12level determination. Matchar2 comments that there is no gold standard for coalmine deficiency and recommends the use of clinical criteria to classify patients as coalmine-deficient or nondefi-cient in the absence of a single perfect diagnostic test. We concur, and we suggest that further double-blind, prospective randomized studies are necessary to determine the gold standard test for identifying true B12tissue deficiency.