Abstract
Our aim was to evaluate the potential utility of mean corpuscular volume (MCV) and mean corpuscular hemoglobin (MCH) to detect hemochromatosis. We computed the accuracy of MCV and MCH cut-off points ≥ upper reference limits using data from 94 probands and 132 white controls. Our reference ranges are MCV 80.0-97.0 fL and MCH 26.0-32.0 pg. Sensitivity of MCV was 8.6-48.3% for men and 2.8-44.4% for women (cut-off points ≥105.0-≥97.0 fL, respectively). Sensitivity of MCH was 33.9-70.7% for men and 19.6-50.0% for women (cut-off points ≥34.0-≥32.0 pg, respectively). Using MCV and a hemochromatosis frequency typical of many western Caucasian populations (0.005), positive predictive values (PV+) were 2.1-100.0% in men and 4.2-100.0% in women. Using MCH, PV+ were 1.7-8.2% in men and 1.8-6.8% in women. We also calculated PV+ using the hemochromatosis frequency 0.015, which could occur in persons receiving medical care. Using MCV cut-off points ≥101.0 fL, PV+ were 8.9-100.0% in men and 100.0% in women with maximum sensitivities of 24.1% and 25.0%, respectively. Using MCH testing, PV+ was 21.5% in men (cut-off point ≥34.0 pg) and 18.2% in women (cut-off point ≥33.0 pg) with sensitivities of 33.9% and 37.0%, respectively. Using MCV or MCH, sensitivity and PV+ for the HFE genotype C282Y/C282Y were generally greater than for "nonclassical" HFE genotypes. All negative predictive values in our study were ≥98.5%. We conclude that supranormal values of MCV or MCH could be used to detect hemochromatosis in white persons of western European descent who are receiving routine medical care. Comparisons of MCV, MCH, and transferrin saturation testing and other implications of MCV and MCH testing for hemochromatosis in medical care are discussed.
Get full access to this article
View all access options for this article.