The severity of iron overload among HFE p.C282Y homozygotes varies due to heritable and acquired factors but remains incompletely explained. Using exome sequencing, we previously demonstrated that the GNPAT p.D519G variant is associated with extreme iron loading among HFE p.C282Y homozygous males who did not report excessive alcohol consumption or other known risk factors for severe iron loading (Hepatology 2015;62:429-439). Demonstrating an effect of GNPAT p.D519G across a wider range of iron phenotypes has yielded mixed results in several study cohorts.

The current investigation was undertaken to examine the question whether the association previously observed between iron stores and p.D519G in p.C282Y homozygous men also exists in p.C282Y homozygous women. We used an extreme phenotypes design comprising 11 high- and 57 low-expressing HFE p.C282Y homozygous females. Criteria for high expressers were age >18 and either (a) mobilized body iron >10 g by quantitative phlebotomy or (b) hepatic iron concentration >236 µmol/g dry weight (reference range 0-36 µmol/g). Criteria for low expressers included age >50 y and serum ferritin <200 µg/L or <3.0 g iron removed by phlebotomy to achieve serum ferritin <50 µg/L. Linear regression was applied with GNPAT p.D519G allele count as the independent variable and either of two continuous phenotype measurements as the dependent variable. We tested for associations between GNPAT p.D519G allele count and (a) serum ferritin level at presentation and (b) total iron removed by phlebotomy to achieve iron depletion among women for whom complete genotype and alcohol consumption data were available (N=68). We adjusted for age, alcohol consumption reports, and body mass index at presentation.

Serum ferritin at presentation was associated with alternate allele count for rs11558492 (corresponding to GNPAT p.D519G). We observed an increase in serum ferritin of 430 µg/L for each increase in the number of variant alleles in GNPAT p.D519G (p=0.025, 95% CI=[19.9, 840]). Total iron removed was also associated with genotype. We observed an increase of 1.9 g iron removed for each increase in the number of variant alleles in GNPAT p.D519G (p=0.019, 95% CI=[0.34, 3.4]).

In women with HFE p.C282Y homozygosity, GNPAT p.D519G is significantly associated with serum ferritin phenotypes at presentation and with iron removed by phlebotomy in an extreme phenotypes model. Some women with GNPAT p.D519G do not have high iron phenotypes at presentation. This suggests that some women with HFE p.C282Y homozygosity and GNPAT p.D519G also have genetic or environmental factors that decrease iron absorption.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.