Over a 1-year period, we examined 42 consecutive patients with low serum cobalamin levels detected by primary screening test (S-protein binder, RIA). In 31 patients (74%) clinical cobalamin deficiency was confirmed, whereas the remaining 11 patients (26%) were characterized clinically as non-cobalamin deficient. The serum methylmalonic acid level was higher than 0.34 mumol l-1 (3 SD above the mean in normal controls) in 30 of the 31 clinically characterized cobalamin-deficient subjects, and below this level in 10 of the 11 non-deficient patients. We conclude that the serum methylmalonic acid assay provides an appropriate means of discrimination between cobalamin deficiency and non-cobalamin deficiency (efficiency = 0.95), and we recommend that the assay be adopted as the standard test for diagnosis of tissue cobalamin deficiency.

Methylmalonic acid (MMA) concentrations are increased in cobalamin (vitamin B12) deficiency, but the relative diagnostic usefulness of determination of MMA in serum vs urine has not yet been assessed. We obtained urine collections and matched serum samples from 28 healthy volunteers and from 20 consecutive patients admitted for clinical and hematological evaluation because of low cobalamin concentrations in serum. Increased concentrations of MMA in serum were found in 12 patients, in all of whom a clinical diagnosis of cobalamin deficiency was established. By contrast, cobalamin deficiency was excluded in seven of the eight remaining patients, who all had normal MMA concentrations. Here we report that linear relationships exist between MMA concentrations in serum (investigated range: 0.05-34.2 mumol/L) and MMA concentrations in urine (r = 0.74), concentrations relative to creatinine (r = 0.98), and MMA excretion rates (r = 0.97) (P less than 0.001 in each instance). Our data are consistent with glomerular filtration and passive reabsorption of MMA by the tubules. We demonstrate, for the first time, a negative correlation between concentrations of cobalamin and MMA in serum in clinical cobalamin deficiency (r = -0.69; P less than 0.01; n = 12); when the values for MMA were log transformed, the correlation with cobalamin was much better (r = -0.84; P less than 0.0005).

We demonstrated a within-patient inverse relationship between the concentrations of methylmalonic acid and cobalamin in plasma from these uremic patients. Despite the fact that only two of the patients developed subnormal plasma cobalamin values, we demonstrated a B12 depletion during the withdrawal period. Treatment with cyanocobalamin tablets once daily was found efficient, but the oral doses should possibly be increased.

No abstract

Over a 2-year period, we examined 48 patients with P-cobalamin levels in the difficult "grey zone" at the lower reference limit detected by a competitive protein binding assay using intrinsic factor as binder. In 21 of 30 patients (70%) with low P-cobalamins we could not establish the diagnosis of metabolic cobalamin deficiency, but 1 of 18 patients (6%) with low-normal P-cobalamin values was confirmed metabolically cobalamin-deficient. Half of these 30 patients with low P-cobalamins had neuropsychiatric disorders, but only one-third of the latter patients had metabolic cobalamin deficiency. Ten of the remaining 15 patients (67%) were characterized as non-deficient. In patients with low-normal P-cobalamin level, we found neuropsychiatric disorders in 5 of the 18 (28%), but none of these had metabolic cobalamin deficiency. We conclude that P-cobalamins below the reference interval combined with typical neuropsychiatric symptoms or findings are not diagnostic of cobalamin deficiency and that further analyses are necessary.