HFEgenotypes, haemochromatosis diagnosis and clinical outcomes to age 80: a prospective cohort study in UK Biobank

Objectives HFE haemochromatosis genetic variants have an uncertain clinical penetrance, especially to older ages and in undiagnosed groups. We estimated p.C282Y and p.H63D variant cumulative incidence of multiple clinical outcomes in a large community cohort. Design Prospective cohort study. Setting 22 assessment centres across England, Scotland, and Wales in the UK Biobank (2006-2010). Participants 451,270 participants genetically similar to the 1000-Genomes European reference population, with a mean 13.3-year follow-up through hospital inpatient, cancer registries and death certificate data. Main outcome measures Cox proportional hazard ratios of incident clinical outcomes and mortality in those with HFE p.C282Y-p.H63D mutations compared to those with no variants, stratified by sex and adjusted for age, assessment centre and genetic stratification. Cumulative incidences were estimated from age 40 to 80 years. Results 12.1% of p.C282Y+/+ males had baseline (mean age 57) haemochromatosis diagnoses, with age 80 cumulative incidence of 56.4%. 33.1% died vs. 25.4% without HFE variants (Hazard Ratio [HR] 1.29, 95% CI: 1.12-1.48, p=4.7*10 -4 ); 27.9% vs 17.1% had joint replacements, 20.3% vs 8.3% had liver disease, and there was excess delirium, dementia, and Parkinson’s disease, but not depression. Associations, including excess mortality, were similar in the group undiagnosed with haemochromatosis. 3.4% of p.C282Y+/+ females had baseline haemochromatosis diagnoses, with cumulative age 80 incidence of 40.5%. There was excess incident liver disease (8.9% vs 6.8%; HR 1.62, 95% CI: 1.27-2.05, p=7.8*10 -5 ), joint replacements and delirium, with similar results in the undiagnosed. p.C282Y/p.H63D and p.H63D+/+ men or women had no statistically significant excess fatigue or depression at baseline and no excess incident outcomes. Conclusions Male and female p.C282Y homozygotes experienced greater excess morbidity than previously documented, including those undiagnosed with haemochromatosis in the community. As haemochromatosis diagnosis rates were low at baseline despite treatment being considered effective, trials of screening to identify people with p.C282Y homozygosity early appear justified. Strengths and limitations of this study We analyzed largescale data on community volunteers from the UK Biobank, one of the world’s largest HFE genotyped cohorts. We have analyzed incident disease outcomes during an extended follow-up period of mean 13.3 years. We have provided the first clinical outcome data to age 80 years in those with haemochromatosis genotypes, including those undiagnosed with haemochromatosis at baseline, expanding the life-course evidence on HFE penetrance. UK Biobank participants were somewhat healthier than the general population, but HFE allele frequencies were similar to previous UK studies. Incident outcomes were from hospital inpatient and cancer registry follow-up, so did not rely on potentially biased patient self-reporting, but community diagnosed conditions may be underestimated.