Hereditary hemochromatosis is, by the reckoning of the liver societies, the most common identified genetic disorder in people of European descent — most frequent in populations of northern European, particularly Nordic or Celtic, ancestry, where it occurs at a prevalence of roughly 1 per 220 to 250 individuals [1]. Almost nobody has heard of it. Its treatment is the removal of blood, on a schedule, for years — a remedy that sounds like something a barber-surgeon would have offered, and which happens to work.
The combination is unusual enough to be worth explaining carefully: a common inherited condition, a genuinely effective and almost absurdly simple treatment, and a genotype that in most of the people who carry it never becomes a disease at all.
Start here: what hemochromatosis is
The body has no regulated route for getting rid of iron. Intake is controlled at the point of absorption in the gut, and once iron is in, it essentially stays. Hemochromatosis is an inherited failure of that control: the gut absorbs more dietary iron than the body needs, year after year, and the surplus is deposited in tissues.
The predisposition is to increased, inappropriate absorption of dietary iron, and the accumulation can lead to cirrhosis, hepatocellular carcinoma, diabetes and heart disease [1]. Joints are affected too, and not trivially: in a case-control study of 306 patients against 304 matched controls, osteoarthritis had been diagnosed in 50.5 percent of patients versus 28.9 percent of controls (adjusted odds ratio 2.5), with markedly higher odds of knee and hip replacement, and osteoporosis in 23.3 percent versus 4.6 percent (adjusted odds ratio 7.3) [2]. The classical picture also includes hypogonadism and skin pigmentation [3].
The genetics are straightforward. The principal defect is in the HFE gene: a missense mutation substituting tyrosine for cysteine at position 282 of the protein product, written C282Y. Roughly 85 to 90 percent of people with inherited iron overload are homozygous for it, with a small minority being compound heterozygotes carrying C282Y on one allele and H63D or S65C on the other; the remaining 10 to 15 percent have mutations in other iron-regulatory genes [1]. It is recessive: one copy is not enough.
Mechanistically, the wild-type HFE protein forms a stable complex with the transferrin receptor and lowers that receptor's affinity for transferrin; the C282Y mutation eliminates HFE's association with beta-2-microglobulin, prevents its cell-surface expression, and nearly completely abolishes the complex with the transferrin receptor [4]. HFE is thus a sensor in a regulatory circuit, and the mutation blinds it. A HuGE review gathers the population genetics and disease associations of the gene [5].
And then the complication that makes this disease interesting rather than merely unfortunate: many people with two copies never develop iron-overload disease at all. That is the subject of the centerpiece below, and it changes how the condition should be found and treated.
Pillar 1: measurement and diagnosis
The two iron numbers that matter
Two blood tests carry most of the diagnostic weight, and they measure different things.
Transferrin saturation is the fraction of the body's iron-transport protein that is currently carrying iron, calculated as serum iron divided by total iron-binding capacity. It rises early, before stores are large, which makes it the screening test. The conventional trigger is a value above 45 percent, with or without a raised ferritin, at which point testing for the hemochromatosis gene is recommended [6]. In the Busselton population study, 15 of the 16 C282Y homozygotes had a transferrin saturation of 45 percent or higher; the sixteenth was at 43 percent [7].
Ferritin reflects how much iron is actually stored. It is the number that says how much damage the accumulation has had time to do, and it is what treatment is titrated against. In the HEIRS screening study of 99,711 participants, among previously undiagnosed C282Y homozygotes, ferritin exceeded 300 micrograms per litre in 88 percent of men and 200 micrograms per litre in 57 percent of women [8].
The critical caveat is that ferritin is not specific. It is an acute-phase reactant, and it rises in inflammation and in metabolic liver disease as readily as in iron overload. A raised ferritin with a low or normal transferrin saturation should point away from hemochromatosis and toward other primary or secondary causes — the pattern named dysmetabolic hyperferritinemia [6]. Hyperferritinemia and inflammation travel together often enough that treating a ferritin number as an iron measurement is a recurring clinical error [9].
Genotyping, and the thing a genotype cannot tell you
Confirming C282Y homozygosity establishes the genetic susceptibility. It does not establish that a person has, or will get, the disease. The liver-society staging makes this explicit: stage 1 is the genetic disorder with no increase in iron stores — genetic susceptibility only; stage 2 is phenotypic iron overload without organ damage; stage 3 is iron overload with tissue and organ damage [1]. Only stage 3 is what a patient would call being ill.
Incomplete penetrance of this kind is a general problem in human genetics, not a quirk of this gene [10] [11], but hemochromatosis is one of its cleanest and most consequential examples.
Staging the organs
Where iron burden needs quantifying rather than inferring, imaging has largely displaced biopsy. Magnetic resonance can measure tissue iron directly; the T2-star technique was developed precisely to detect myocardial iron overload early, before cardiac function declines [12]. Liver biopsy retains a role in staging fibrosis in patients whose ferritin is high enough to raise the concern — in Busselton, biopsy was recommended for ferritin at or above 300 nanograms per millilitre, and among 11 homozygotes biopsied, 3 had hepatic fibrosis and 1 had cirrhosis [7].
The hormone underneath all of this
The regulatory logic became clear only after the gene did. Hepcidin, a peptide hormone made in the liver, is the principal regulator of systemic iron homeostasis: it controls plasma iron and tissue iron distribution by inhibiting intestinal iron absorption, iron recycling by macrophages, and iron mobilisation from hepatic stores, acting by binding the cellular iron exporter ferroportin and inducing its degradation [13] [14]. Hepcidin is the "stop absorbing" signal; ferroportin is the door it closes.
The disease is what happens when that signal fails. Mice lacking hepcidin expression develop severe tissue iron overload [15]; hepcidin transcription is regulated by anemia, hypoxia and inflammation [16], and interleukin-6 drives the hypoferremia of inflammation by inducing hepcidin [17] [18] — the same axis, pointed the other way. Bone morphogenetic protein signalling sets the hepcidin set-point in vivo [19], and hemojuvelin, a co-receptor in that pathway, is essential for dietary iron sensing, with its mutation producing severe overload [20].
That framework explains the whole family of iron-overload diseases at once. Mutations in HFE2 (hemojuvelin) cause chromosome-1q-linked juvenile hemochromatosis [21]; hepcidin is decreased in transferrin-receptor-2 hemochromatosis [22]; and digenic inheritance of mutations in HAMP (hepcidin itself) and HFE produces different disease types [23]. Genetic analysis of these disorders is what taught the field how iron homeostasis works [24] [25] [26] [27].
Centerpiece: a simple simulatable model of the penetrance gap
The question that decides how this disease should be found is quantitative: if the genotype is carried by roughly 1 person in 220 to 250 of northern European ancestry [1], why are patients so much rarer than that?
The model has two stages. The first is Hardy-Weinberg, which for an allele at frequency q in a randomly mating population gives the genotype frequencies as the expansion of (p + q) squared: carriers at 2q(1 − q) and homozygotes at q squared. This has no free parameters. The second stage multiplies the homozygote frequency by a penetrance P — the fraction of homozygotes who actually develop iron-overload disease — and P is both measured and far below one.
The published inputs are these. A survey of 5,956 chromosomes found a worldwide C282Y allele frequency of 1.9 percent, rising to 10 percent in Irish chromosomes, and found the allele absent from 1,042 African, 484 Asian and 644 Australasian chromosomes [28]. For penetrance, the HealthIron study followed 31,192 people of northern European descent for an average of 12 years and found documented iron-overload-related disease in 28.4 percent of male C282Y homozygotes (95 percent confidence interval 18.8 to 40.2) but only 1.2 percent of female homozygotes (0.03 to 6.5) [29]. A systematic review for the US Preventive Services Task Force put the range differently and more cautiously: data were insufficient for a precise estimate, but suggested up to 38 to 50 percent of homozygotes may develop iron overload, with up to 10 to 33 percent eventually developing hemochromatosis-associated morbidity [30].
The Hardy-Weinberg spine is checked rather than assumed. The Busselton study reported, independently, that 424 of 3,011 white adults (14.1 percent) were heterozygous and 16 (0.5 percent) were homozygous [7]. Feeding only the heterozygote frequency into 2q(1 − q) gives an allele frequency of 7.6 percent, and squaring that predicts 1 homozygote in 172 — a number the model never saw, against an observed 1 in 188. Two other independent estimates bracket it: HEIRS found 0.44 percent, or 1 in 227, among non-Hispanic whites [8], and the liver-society guideline quotes 1 per 220 to 250 for northern European populations [1]. All four agree within a factor of 1.5, which is what a randomly mating population is supposed to look like.
Applying the measured penetrances to that population, under an illustrative 50/50 sex split, gives roughly 0.083 percent of the population as homozygous men with iron-overload disease and 0.0035 percent as homozygous women with it — about 1 person in 1,160 altogether.
That is the teaching point, and it is computed rather than asserted. The genotype is about 6.8 times more common than the disease, and carriers are about 164 times more common. Two consequences follow directly.
First, genotype-first screening of the general population would identify a large number of people who will never become ill, and the evidence base does not support doing it: the Task Force review concluded that research remains insufficient to confidently project the impact of, or estimate the benefit from, widespread or high-risk genetic screening [30]. Screening starts with transferrin saturation — a measure of what iron is actually doing — and reaches for the genotype only afterwards [6].
Second, the enormous male-female asymmetry in the figure is not a modelling artifact. It reflects the measured 28.4 percent versus 1.2 percent penetrance [29], and the conventional explanation — menstruation and pregnancy as lifelong physiological phlebotomy — is the same mechanism as the treatment described below.
Two honest limits. The penetrance estimates disagree between sources, which is why the figure carries a band rather than a line: the Busselton investigators found that only half their homozygotes had clinical features and a quarter had ferritin that stayed normal over four years [7], while the Task Force range reaches 33 percent [30]. And Hardy-Weinberg assumes random mating and no selection, which is an approximation in populations with the founder structure this allele's northern-European distribution implies [28].
Pillar 2: treatment
Therapeutic phlebotomy
The treatment is to take blood out. Because roughly a quarter of the body's iron is in circulating red cells and there is no excretory route, removing blood forces the body to draw on its iron stores to rebuild the haemoglobin, and the stores come down.
The regimen is specific. Therapeutic phlebotomy should be started in men with a serum ferritin of 300 micrograms per litre or more and in women at 200 or more, regardless of whether they have symptoms; it consists of removing one unit — 450 to 500 millilitres — of blood weekly until the ferritin falls to 10 to 20 micrograms per litre, after which the level is maintained at 50 or below by periodic removal [3].
What it achieves is worth stating plainly, because it is the strongest claim in this review. Applied before iron overload becomes severe, phlebotomy prevents the complications: hepatic cirrhosis, primary liver cancer, diabetes mellitus, hypogonadotrophic hypogonadism, joint disease and cardiomyopathy. In people who already have established iron-overload disease, weakness, fatigue, raised hepatic enzymes, right-upper-quadrant pain and hyperpigmentation are often substantially alleviated [3]. A natural-history study of 277 C282Y homozygotes followed a mean of 7.3 years reported actuarial survival of 95, 93 and 66 percent at 5, 10 and 20 years, with cirrhosis and diabetes the major determinants of long-term survival, and concluded that long-term survival is excellent in homozygotes diagnosed and treated before cirrhosis and diabetes develop [31].
The asymmetry there is the whole clinical message. Treatment started early prevents essentially everything; treatment started after cirrhosis does not undo it. Timing, not the treatment, is the hard part.
Diet, and what to leave alone
Dietary management is a supporting measure, not a treatment. It means avoiding medicinal iron and mineral supplements, avoiding excess vitamin C — which increases iron absorption — and avoiding uncooked seafood, this last because iron-loaded patients are unusually susceptible to Vibrio infection. Together these reduce the rate at which iron re-accumulates and help reduce complications of liver disease and diabetes [3]. Ordinary oral iron in food does not cause overload except in the genetically predisposed or in people with ineffective erythropoiesis [1].
Chelation, and when it is needed
Where phlebotomy cannot be used — significant anaemia, poor venous access, cardiac disease that makes volume removal unsafe — iron must instead be bound and excreted. Chelation is the mainstay of iron control in the transfusional overload of the thalassemias [32], and the same T2-star imaging that detects myocardial iron is used to guide it [12]. Chelation is more burdensome and more toxic than taking a unit of blood, which is precisely why phlebotomy's simplicity matters.
Screening the family
Because the condition is recessive and common, the siblings of an affected person have a one-in-four prior probability of being homozygous — far higher than any population screen would yield. Case-finding in relatives is the setting where the genotype-first approach is justified, and the prevalence of C282Y homozygosity is indeed higher in family members of probands and in other high-risk groups defined by signs, symptoms and phenotypic screening [30]. Surveillance of people already known to be iron-loaded is a distinct and better-founded activity than population screening [33].
Pillar 3: what is unresolved
Who actually needs treating
The penetrance question is not settled, and it is the one that matters most. The measured proportions range from the 28.4 percent of men in HealthIron [29] to the wide bounds the Task Force review would commit to [30], with a population study finding half its homozygotes had clinical features and a quarter never even developed a raised ferritin [7]. Whatever modifies penetrance — other genetic variants, sex, alcohol, blood loss, metabolic liver disease — is only partly identified, and reduced penetrance in general remains poorly understood at a molecular level [10] [11]. Until it is, a substantial number of people are treated for a risk they may not have had.
Hepcidin-targeted therapy
If the disease is a hepcidin deficiency in effect, the pharmacological answer is to supply the missing signal rather than to remove blood. The mechanistic groundwork is in place: hepcidin's action on ferroportin is understood [13] [14], the bone-morphogenetic-protein pathway that sets its level is mapped [19] [20], and animal models with defined lesions exist [15]. Turning that into a drug that outperforms a weekly blood draw — cheap, effective, and with a side effect that helps the blood supply — is a high bar.
The screening debate
It is genuinely unresolved rather than merely unfunded. The arguments for screening are that the condition is common, detectable years before damage, and treatable with something trivial. The arguments against are that penetrance is low and imprecisely known, that available research is observational and inconsistently reported, and that no study has shown earlier phlebotomy in screen-detected people reduces morbidity or mortality compared with treatment after ordinary clinical diagnosis — one of the three questions the Task Force review set out to answer and could not [30].
The cell-biology tail
How iron actually kills cells is being rewritten. Iron-dependent cell death has been characterised in murine models of hemochromatosis [34], labile plasma iron is redox-active and chelatable [35], and the intracellular regulatory machinery is being dissected in vivo [36]. Whether any of this changes management is not yet clear.
Dig deeper in lmmol
Hemochromatosis is one of two great inherited disorders of iron and heme that lmmol covers. The other is sickle cell disease, which is the mirror image in an instructive way: a mutation in the oxygen-carrying protein itself rather than in the sensor that regulates how much iron the body takes in, and one whose transfusion-based treatments can cause the very iron overload that hemochromatosis produces spontaneously [32]. Both diseases end up in the liver, where the differential diagnosis of a raised ferritin runs straight into fatty liver disease — metabolic liver disease raises ferritin without iron overload, and telling the two apart is the point of measuring transferrin saturation [6]. Iron deposition in the pancreas is why untreated overload causes type 2 diabetes and glycemic control problems [1], deposition in the myocardium connects to heart failure [12], and the arthropathy and bone loss link to osteoporosis [2]. For the contrast between a recessive disease with low penetrance and a dominant one with essentially complete penetrance, see Huntington's disease; for a recessive disease whose genotype does predict phenotype closely, see cystic fibrosis. The full collection is at health.