Background: Plasma homocysteine has been reported to be useful in the evaluation of patients with suspected vitamin B12 or folate deficiency. In November 1998, Canada began its mandatory fortification of all flour, and some corn and rice products, with folic acid. We evaluated the status of folate and vitamin B12 in Ontario since this fortification program began, and also studied the role of plasma homocysteine in the assessment of vitamin B12 deficiency since that time.
Methods: A retrospective cross-sectional study design was performed using a community database of all Ontario samples analyzed by MDS Laboratories, a major provider of diagnostic laboratory services in Canada. All consecutive single-patient fasting samples for plasma homocysteine collected between January 1 and September 30, 1999 were included, as well as corresponding red cell folate and serum B12 concentrations. Data for serum folate were included when available. Descriptive statistics included the arithmetic and geometric means for each measure, as well as the lower and upper centile values. After excluding cases with a concomitant serum creatinine > 120 micromol/L or red cell folate < 215 nmol/L, we established the test properties of a plasma homocyteine level of 15 micromol/L or greater for the diagnosis of "low" (< 120 pmol/L) or "indeterminate" (i.e., between 120 and 150 pmol/L) serum vitamin B12 concentrations.
Results: The mean age of all subjects was 58.4 years (95% CI 57.4 to 59.4). Plasma homocysteine samples were obtained from 403 males (56.7%) and 308 females. The geometric mean homocysteine concentration for the entire population was 8.3 micromol/L, and was significantly higher among males (9.3 micromol/L) than females (8.3 micromol/L) (unpaired t-test: 2-p < 0.0001). The geometric mean serum folate concentration was significantly higher in females (35.8 nmol/L) than males (33.6 nmol/L) (2-p < 0.0001), as were the mean red cell folate levels (females 966.8 nmol/L, males 949.3 nmol/L; 2-p < 0.0001). Serum vitamin B12 concentrations were available for 692 subjects, with a geometric mean of 322.0 pmol/L. Again, mean vitamin B12 was higher in females (332.5 pmol/L) than males (314.3 pmol/L) (2-p < 0.0001). The fifth centile for vitamin B12 was 134.6 pmol/L. A plasma homocysteine concentration > 15 micromol/L did not discriminate between cobalamin concentrations below versus above 120 pmol/L (positive and negative predictive values 7.4% and 97.2%, respectively), nor did it discriminate "indeterminate" B12 levels between 120 and 150 pmol/L (positive and negative predictive values 6.3% and 94.0%, respectively).
Conclusion: In a large select group of Ontarians, serum and red cell folate concentrations appear to be higher than expected, possibly due to a recent national folate fortification programme; cobalamin levels are no higher than expected. Given our inability to detect mild B12 deficiency using such indicators as plasma homocysteine, and considering the substantial growth in the elderly segment of the Canadian population, occult cobalamin deficiency could become a common disorder. Accordingly, we recommend either consideration of the addition of vitamin B12 to the current folate fortification programme, and/or the development of better methods for the detection of cobalamin deficiency.