Anaemia: cobalamin deficiency: methylmalonic acid was minimally incrementally useful.

Clinical bottom line (level 4)

  1. In the presence of known B12 deficiency, raised methylmalonic acid levels made cobalamin deficiency much more likely than folate deficiency (LR+21) .
  2. The commonest causes of folate deficiency were alcoholism and malnutrition.
  3. The commonest causes of cobalamin deficiency were pernicious anaemia and tropical sprue.
Savage et al: American Journal of Medicine 1994; 96: 239-246
Expires December 2003

The study

Setting: two university hospitals, USA

525 patients (aged ?, 52% female) cobalamin deficiency (406) or folate deficiency (119)

Independent ?blinded reference standard, applied in all patients from a non-consecutive inappropriate spectrum.
Reference standard:
  • Combination of
    • diagnostic marrow and/or blood smear- positive if megaloblastic bone marrow or hypersegmented neutrophils
    • response to cobalamin therapy- decrease in MCV 5 fl or more, increase in haemocrit 0.05 or more, resolution of thrombocytopaenia or leukopenia, improvement in neuropsychiatric symptoms
Diagnostic test:
  • serum methylmalonic acid- positive if more than 376 nmol/l
  • serum homocysteine- positive if more than 21.3 micromol/l

  • Folate deficiency diagnosed if:
    • diagnostic marrow or blood smear
    • serum folate less than 4 ng/ml
    • serum cobalamin more than 300 pg/ml
    • underlying disorder associated with folate deficiency
    • documented response to folic acid

The evidence


differential diagnosis number of patients prevalence
(95% CI)
cause of cobalamin deficiency- pernicious anaemia 310 76%
(72% to 81%)
tropical sprue 39 9.6%
(6.7% to 13%)
bowel resection 27 6.7%
(4.2% to 9.1%)
jejunal diverticula 6 1.5%
(0.3% to 2.7%)
vegetarianism 2 0.5%
(0.0% to 1.2%)
food cobalamin malabsorption 3 0.7%
(0.0% to 1.6%)
unknown cause 19 4.7%
(2.6% to 6.7%)
cause of folate deficiency-alcoholism 103 87%
(80% to 93%)
malnutrition 11 9.2%
(4.0% to 14%)
other 5 4.2%
(0.6% to 7.8%)


diagnostic test cobalamin deficiency folate deficiency LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
MMA >376 292 4 21
(8.2 to 56) 		99% 0.29
(0.25 to 0.34) 		50%
homocysteine >21.3 290 88 0.97
(0.85 to 1.1) 		77% 1.1
(0.78 to 1.5) 		79%
total 406 119


diagnostic test cobalamin deficiency folate deficiency LR
(95% CI)		 post-test probability
both positive 286 4 24
(9.0 to 62) 		99%
either positive 10 84 0.039
(0.021 to 0.073) 		10%
both negative 1 10 0.033
(0.0043 to 0.26) 		9%
total 297 98

Comments

  1. This was a case-control style study using an unusual and highly selected population. This skew may produce a 'spectrum bias' in the test results (i.e. the test characteristics are markedly different in this 'spectrum' of patients than those in whom it would usually be used). Results should be treated with caution.
  2. Pernicious anaemia was diagnosed using the Schilling test or serum antibody to intrinsic factor in 58% of patients with cobalamin deficiency and was considered likely in another 19%.

Citation

  1. Savage DG, Lindenbaum J, Stabler SP, et al: Sensitivity of serum methylmalonic acid and total homocysteine determination for diagnosing cobalamin and folate deficiencies. American Journal of Medicine 1994; 96: 239-246
Search Terms: reference in review article
Contributor: Chris Ball and Clare Wotton, June 2000
Reviewer: David L. Simel

Clinical Question.
Patient cobalamin or folate deficiency
Intervention or Exposure serum MMA and homocysteine
Outcome diagnosis of cobalamin deficiency