Haemochromatosis Arthropathy

First, the stop sign. If you are breathless, your heart is racing or skipping, or you feel faint, stop reading and get medical help today. If you are taking iron, speak to your doctor before your next dose.

Make a fist. Look at which knuckles will not close.

Ordinary hand arthritis goes for the fingertip joints and the base of the thumb. This condition goes for the big knuckles of the index and middle fingers, which is backwards. If those two are the worst ones in your hand, or if you have an arthritic ankle you never injured, ask your doctor for two blood tests together: ferritin and transferrin saturation.

Ten seconds to check. People with this condition waited an average of nine years for someone to order those tests.

The Verdict

Your body hoards iron for decades, and it attacks the wrong knuckles. One blood test finds it.

Think of iron like water in a house with no drain. A normal body turns the tap down once the tank is full. This one cannot, so iron keeps arriving and gets stored in the walls. The liver, heart and pancreas are rooms you can dry out by draining the tank, which is exactly what the treatment does. The joints are the room where the damp already reached the plaster, and that is why draining the tank protects everything else and still does not repair that one wall.

What this actually is: an inherited fault that makes you absorb too much iron, and the joints are usually the first thing to complain.
What most people get wrong: taking the iron back out protects your liver, heart and pancreas, and it does not fix the joints, so nobody should promise you that it will.
Start here: ask for ferritin and transferrin saturation, two blood tests, done together.

Best for

Adults with joint pain in an unusual pattern, especially the index and middle knuckles or a never-injured ankle, starting younger than you would expect. Anyone with a family history of iron overload, liver disease or early joint replacement.

Skip if

You have any heart symptoms, or you are already being treated with iron. Both of those need a doctor first, not a self-check. This page is not a substitute for being examined.

Want the full evidence? Keep scrolling

Treatment

What Works

Read the grading honestly. The strongest thing on this list is a blood test, and there is no top-tier physical therapy treatment here because none has ever been tested.

Get ferritin and transferrin saturation measured STRONG

The single highest-value action available. Joint pain came an average of 9.0 years before the haemochromatosis diagnosis in a group of 199 patients, and the musculoskeletal clinic is where those people spend most of that decade. Finding it early and treating the iron prevents liver scarring, liver cancer, diabetes and heart muscle disease.

Expected timeline: a result within days. The benefit to your organs is lifelong. The benefit to the joints is none, and that is not a typo.

Stop any iron-containing supplement STRONG

Iron is not safe in this condition. That includes multivitamins with iron in them, which people rarely think of as medication. Raise it with your doctor rather than stopping a prescription unilaterally.

Standard arthritis management for the affected joints MODERATE, BORROWED

Activity and load management, joint-specific strengthening, range-of-movement work, supportive footwear or insoles where the ankle is involved, and pain relief through your doctor.

Labelled honestly: the evidence behind this is strong for ordinary arthritis and has never been tested in this condition. Nothing suggests it is harmful. Nothing demonstrates it works here.

Being told the truth about what to expect MODERATE

Specifically: removing the iron protects the organs and will not fix the joints, and new joints can become sore while your iron levels are perfectly controlled. That is the disease behaving normally, not the treatment failing and not you doing something wrong. Counselling counts as treatment here because getting it wrong causes real harm.

Hand therapy for grip and daily tasks EMERGING

Sensible and conventional. Recommended in a 2018 review that cites no trial for it. Presented as what it is.

Exercise Prescription

Read this before the table. No exercise programme has ever been tested for this condition, anywhere. What follows is the conventional starting point used for arthritic joints in general, not a prescription extracted from a study, and the sets and repetitions are a cautious opening position rather than a proven dose. If you have any known heart involvement, or you get breathless or dizzy, do none of this until your doctor has cleared you.

Exercise	How to do it	Sets × reps	How often	Pain guide
Gentle fist closing	Open the hand wide, then slowly curl the fingers into as full a fist as you can manage. Do not force it	2 × 10	Daily	Stretch and effort fine. Sharp pain means back off
Putty or ball squeeze	Squeeze a soft therapy putty or stress ball, hold 3 seconds, release slowly	2 × 10	Daily	Aching afterwards fine. Swelling next morning means too much
Ankle alphabet	Sitting, lift the foot and slowly trace the letters A to J with the big toe	1 set	Daily	Movement, not strain
Calf raises, two feet	Hold a worktop, rise onto the toes slowly, lower slowly	2 × 10	Every other day	Mild ache fine. Stop if the ankle joint itself sharpens
Sit to stand	From a dining chair, stand and sit slowly, without hands if you can	2 × 8	Every other day	Thigh effort is the goal. Lingering joint pain means less depth

Progression. Weeks 1 to 2, get the movements comfortable and change nothing. Weeks 3 to 4, if nothing is flaring, add a few repetitions or a little more range, one change at a time. Week 5 onward, keep going with whatever is tolerable. This is a long-term condition, so the aim is a routine you will still be doing in a year, not a course you finish.

What Doesn't Work

Expecting the iron treatment to improve your joints. It does not, and new joints can become affected while your iron levels are perfectly controlled. This is the most commonly held wrong belief about the condition, including among clinicians, because it follows so naturally from the idea of removing the cause.
Ultrasound scans hunting for active joint inflammation. The finding turns up in 83 out of 100 people who have this condition and no joint problem at all, at much the same rate as ordinary hand arthritis, and it does not track how anyone actually feels.
Genetic screening of everyone with early arthritis. 940 people under 70 with severe hip arthritis were screened and no excess was found. The trigger is the unusual pattern, not simply having arthritis young.
Blaming your job or your training. There was no link between this arthritis and a history of heavy physical work. It persists as an explanation because the joints look like wear and tear and people want a reason.

Safety first

Red Flags

When to get checked, and when to get checked today.

Dark cinematic anatomical study relating to urgent referral signs

Breathlessness, a racing or irregular heartbeat, or feeling faint

Iron can build up in the heart muscle. Get a same-day medical assessment. Do not do an exercise test and do not push through a training session to see how it goes.

You are taking iron tablets, or a multivitamin that contains iron

Iron is not safe for people with this condition. Speak to your doctor before your next dose. Do not simply stop a prescribed medicine on the strength of a web page, because tiredness can also mean the opposite problem, and only a blood test tells the two apart. Get the conversation booked quickly.

Aching index and middle knuckles, or an arthritic ankle you never injured, under about 55

This is the pattern that suggests the condition. Ask your doctor for ferritin and transferrin saturation, two blood tests done together.

Joint pain plus lasting tiredness, plus skin looking permanently tanned, new diabetes, or loss of sex drive

Suggests iron is affecting several organs, not just the joints. See a doctor promptly.

A joint that turns suddenly hot, red and very painful, especially with a fever

Needs same-day assessment. A crystal-related flare and a joint infection can look identical, and only fluid drawn from the joint separates them.

Already diagnosed, and no recent liver check

Joint disease here is a marker that the liver may be affected too. In one group of 112 patients, the absence of arthritis ruled out advanced liver scarring 95 times out of 100.

Who to see: your family doctor first, for the two blood tests. That covers most people. Liver specialist once it is confirmed. Rheumatology if the joint picture is genuinely unclear. Emergency or same-day care for any heart symptom.

Activity

Return to Training

These are clinical checkpoints, not evidence-based thresholds. No return-to-activity rule has ever been validated in this condition, so treat the list as a structured way to think rather than a test to pass. The first box is the one that is not negotiable, and unlike the others it does not resolve when the joints settle.

Heart checked, and no breathlessness, palpitations or dizziness on exertion. This comes first
Iron blood tests requested, and the diagnosis under way or already settled
Iron-containing supplements reviewed with whoever prescribes for you
Joint pain during your target activity is mild, and settles within 24 hours
No new joint has become sore in the last 4 weeks
You can grip your working loads without the knuckles ending the set
You can say in your own words that removing the iron protects your organs and will not fix your joints

There is no evidence that training makes this arthritis worse and none that resting helps it, so the general answer is keep going and work around the specific joints. A joint that is clearly worse at six weeks than at the start, despite sensible management, is a reason for a medical review rather than a reason to try harder.

Trust anchor

How Confident Should You Be

Overall: MODERATE and it varies a great deal by claim, so here it is broken out rather than averaged.

The index-and-middle-knuckle patternHIGH

Joint pain arriving years before the diagnosisMOD-HIGH

Ankles being affected out of all proportionMOD-HIGH

Removing the iron not fixing the jointMODERATE

Iron level deciding whether the joint disease startsMODERATE

How much more likely joint replacement isLOW on the size, high on the direction

Ultrasound inflammation as something to treatEVIDENCE AGAINST

Any exercise dose, progression, hands-on treatment or recovery timelineNO EVIDENCE

What would change my mind about iron removal not fixing the joint

This is the most important claim on the page and it rests on long-term expert observation rather than a trial. Follow 200 or more newly diagnosed people, assess the same joints at diagnosis, at the end of the intensive blood-removal phase, and again at two and five years, and count how many new joints become affected. If the joints improved as the iron came down, this claim is wrong and the whole page changes. If new joints kept appearing at the same rate as before treatment, it moves to high confidence. Nobody has run it.

What would change my mind about the new 2025 scoring system

It was built in 154 confirmed cases against 120 deliberately chosen look-alikes, in people already known to carry the gene fault and to be loading iron. Its authors say plainly that it is for selecting patients into research studies and is waiting to be tested elsewhere. Apply it to an ordinary, unselected clinic and report how it performs there, and it either becomes a usable clinical tool or it does not. Until then it stays a research instrument, and at roughly 7 in 10 real cases caught, a low score is not a reason to stop looking.

Next step

Most joint pain is exactly what it looks like. Occasionally it is the first sign of something a blood test would have caught years earlier. The Verdict works through one of these every week, evidence first.

Join The Verdict, free

The Full Picture

The Full Picture — Anatomy, Diagnosis & Evidence

What's Actually Going On

Dark cinematic anatomical study of joint and tissue structure

A hormone called hepcidin normally tells your gut to stop absorbing iron once you have enough. In this inherited condition it does not work properly, so iron keeps arriving and accumulates over decades in the liver, pancreas, heart, skin, and joints.

In the joint it behaves like fast-forwarded osteoarthritis in the wrong places: exuberant bony spurs, onset younger than expected, and involvement of joints that ordinary osteoarthritis tends to leave alone. A 2026 review describes it as being defined by bone and cartilage destruction rather than by persistent inflammation of the joint lining.

There are two parts to it, and they do not behave the same way. Iron level appears to decide whether the joint disease starts: a ferritin above 1,000 at the time of diagnosis carried an odds ratio of 14.0 for the typical arthritis, though on only 10 events, so the direction is credible and the size is not. Iron level appears to have very little to do with whether it then progresses, because it keeps advancing after the iron is removed, new joints get recruited during maintenance treatment, and textbook cases occur in people whose iron was never high. The specialist who set out that puzzle in 2018 ended by asking whether the gene fault damages joints through some route that has nothing to do with iron at all. Nobody has answered him.

Running alongside both is a crystal problem. Calcium pyrophosphate deposits, visible on X-ray as chondrocalcinosis, appeared in a pooled 29% of knees, and the two conditions share biological pathways. That is also why a florid or young-onset case of crystal arthritis should prompt a check for iron overload.

How to Identify It

Dark cinematic anatomical study relating to clinical assessment

The distribution is the finding, not the severity.

Ferritin and transferrin saturation, together

The test that actually changes the patient's life, and it is not a physical therapy test. In the inherited form, a transferrin saturation above 50% with ferritin above 300 in men and women past the menopause, or above 45% with ferritin above 200 in women, is enough to make the diagnosis alongside the gene result.

European liver guideline, 2022.

The 2025 scoring system (8 items, threshold 5 out of 11)

The first formal criteria this condition has ever had, covering age at onset, findings at the knuckles, fingertip joints and ankles, and previous hip or ankle surgery. Two warnings matter more than the numbers. It is a research tool for selecting study patients, not a diagnosis, and its authors say so. And it catches only about 7 in 10 genuine cases, so a low score is not a reason to stop looking.

Catches roughly 71 in 100 real cases; correctly clears about 93 in 100 look-alikes, measured in a hand-picked comparison group. Kiely and colleagues, 2026.

Hand and ankle X-rays

Joint space narrowing, small cysts under the joint surface, and bony spurs concentrated at the index and middle knuckles. Calcium deposits at the knee. Bone hardening at the hip.

Pooled across 36 studies, 2025. The intervals are very wide, so read the ranking of features rather than the percentages.

Ultrasound — specifically not recommended here

Inflammation of the joint lining showed up in 96 of every 100 affected patients, and in 83 of every 100 patients with this condition who had no joint problem at all, at much the same rate as ordinary hand arthritis. None of the ultrasound scores lined up with pain or with function.

87 patients across three groups, 2017.

Which Look-Alike Is It

Dark cinematic anatomical study contrasting joint patterns

This is the section that does the real work, because the condition is not rare so much as misfiled.

Ordinary hand osteoarthritis targets the fingertip joints and the base of the thumb and leaves the big knuckles alone. When the index and middle knuckles are the worst joints in the hand, the distribution itself is the clue.

Rheumatoid arthritis also goes for the big knuckles, but with soft boggy swelling, long morning stiffness, bone erosion rather than bony overgrowth, and positive blood markers. In this condition bony enlargement showed up in about two thirds of patients and obvious soft swelling in only one in seven.

Crystal arthritis overlaps rather than merely resembling. The two coexist, and the referral is the same in either direction.

Ankle arthritis after an old injury is the common explanation, so take the history properly. It is the absence of an injury that should make you look harder, because ankle joint replacement was roughly 9 times more likely in this condition.

The Debate

Does it behave like ordinary arthritis?

The genetics reference (revised 2024): management of the joint disease does not differ from managing these conditions in anyone else.

Two rheumatology reviews (2018): it is a joint disease in its own right, differing in distribution, age of onset, the character of the bony overgrowth and the speed of progression.

Both survive once you separate what to do from what to expect. There is no specific treatment, so care defaults to ordinary arthritis care. The natural history is clearly not ordinary, so the counselling should not be either.

Will removing the iron help the joints?

The intuition, and it follows from the liver guideline: early treatment prevents cirrhosis, liver cancer, diabetes, heart disease and arthritis.

The long-term observation: the arthritis does not respond, new joints get affected during maintenance, and classic cases occur with no iron overload at all.

The guideline's claim is about preventing it in people who do not have it yet. Once the joint disease exists, nothing shows that removing iron reverses it. This is the single most important counselling point on the page.

Should young people with bad hip arthritis be screened?

2012, 27,848 people genotyped: carriers of the double gene fault had nearly 6 times the odds of needing both hips replaced.

2018, 940 people under 70 with end-stage hip arthritis screened: no excess of the gene fault or of iron overload at all.

Neither study is wrong, and the reconciliation is arithmetic. One starts from the gene and asks what happens to those people. The other starts from the arthritis and asks how many carry the gene. The fault is rare enough that a genuinely large individual risk is still invisible in an unselected clinic. So: do not screen arthritis populations, and do test the individual whose pattern is odd.

Honest Limitations

The evidence describes the disease and never tests a treatment

What the research shows: how common it is, which joints, what happens over time, how much surgery it leads to, and now how to classify it.

The real-world gap: not one study has randomly assigned anyone with this arthritis to any physical therapy, exercise programme, hands-on treatment or machine. Searching the medical literature for this arthritis together with physiotherapy returns three records in total, one of which is a reference chapter.

What we did about it: everything below the top tier on this page is labelled as borrowed from ordinary arthritis care. A clinician can finish this literature confident about spotting the condition and with nothing proven to offer for it. That is the honest position, and stating it beats inventing a protocol.

How likely carriers are to actually get ill is misremembered in both directions

What the research shows: over 12 years, iron-related disease developed in about 28 of every 100 men with the double gene fault and about 1 of every 100 women.

The real-world gap: carrying the fault is common in people of northern European descent. Getting ill from it is not. Overstate it and you frighten families into unnecessary testing. Understate it and you miss the roughly 28 in 100 men with the fault who reach a joint replacement by 80.

What we did about it: the figure is given with the male and female split attached every time it appears, never as a single number.

The pooled percentages are not usable as percentages

What the research shows: a 2025 review pooled 36 studies out of 2,473 screened.

The real-world gap: its own authors write that heterogeneity and the small number of usable studies mean the data's meaningfulness is limited. Two of its headline ranges run from 9% to 90% and from 3% to 99%.

What we did about it: the ranking of features is used throughout this page and the percentages are not quoted as if they were rates.

Sources

Kiely PD, Finzel S, Farisogullari B, et al. 2026. EULAR 2025 classification criteria for haemochromatosis arthropathy. Annals of the Rheumatic Diseases. PMID 41193334. Derivation cohort of 154 cases and 120 look-alikes; 8-item model; correctly clears 93.3% of look-alikes and catches 71.4% of cases. Authors state external validation is awaited.
Engelhardt S, Buroh S, Schwarzer G, et al. 2025. Clinical and imaging features of haemochromatosis arthropathy. EULAR Rheumatology Open. PMID 42368602. Systematic review and meta-analysis; 2,473 screened, 36 included. Authors state meaningfulness limited by heterogeneity.
Sahinbegovic E, Dallos T, Aigner E, et al. 2010. Musculoskeletal disease burden of hereditary hemochromatosis. Arthritis & Rheumatism. PMID 20722017. N=199. Joint pain in 72.4%, preceding the diagnosis by 9.0 ± 10.7 years; joint replacement in 16.1% at mean age 58.3.
Carroll GJ, Breidahl WH, Bulsara MK, et al. 2011. Hereditary hemochromatosis is characterized by a clinically definable arthropathy that correlates with iron load. Arthritis & Rheumatism. PMID 20954257. N=103. Typical arthritis in 24% of definite or probable cases; ferritin above 1,000 odds ratio 14.0; no link with heavy physical work.
Allen KJ, Gurrin LC, Constantine CC, et al. 2008. Iron-overload-related disease in HFE hereditary hemochromatosis. New England Journal of Medicine. PMID 18199861. 31,192 genotyped. Iron-related disease in 28.4% of male and 1.2% of female carriers of the double fault.
Lucas MR, Atkins JL, Pilling LC, et al. 2024. HFE genotypes, haemochromatosis diagnosis and clinical outcomes at age 80. BMJ Open. PMID 38479735. N=451,270. Joint replacements in 27.9% of affected men versus 17.1%; excess deaths present in the undiagnosed too.
Wijarnpreecha K, Aby ES, Panjawatanan P, et al. 2021. Hereditary hemochromatosis and risk of joint replacement surgery. European Journal of Gastroenterology & Hepatology. PMID 32118852. Meta-analysis, 5 cohorts, N=1,293,407. Ankle replacement roughly 9 times more likely; hip roughly 2.6 times; knee not significant.
Wang Y, Gurrin LC, Wluka AE, et al. 2012. HFE C282Y homozygosity and risk of total hip replacement for osteoarthritis. Seminars in Arthritis and Rheumatism. PMID 22209421. N=27,848 genotyped. Bilateral hip replacement odds ratio 5.86.
Kiely PD. 2018. Haemochromatosis arthropathy, a conundrum of the Celtic curse. Journal of the Royal College of Physicians of Edinburgh. PMID 30191911. The source of the three observations that break the iron model.
Dejaco C, Stadlmayr A, Duftner C, et al. 2017. Ultrasound verified inflammation and structural damage in hereditary haemochromatosis-related arthropathy. Arthritis Research & Therapy. PMID 29065925. N=87. Inflammation in 96.2% of affected patients but also 83.3% of unaffected ones; no score linked to pain or function.
Oppl B, Husar-Memmer E, Pfefferkorn S, et al. 2018. HFE hemochromatosis screening in patients with severe hip osteoarthritis. PLoS One. PMID 30427934. N=940 under 70. No excess of the gene fault or iron overload. A well-powered negative.
Andersson L, Powell LW, Ramm LE, et al. 2022. Arthritis prediction of advanced hepatic fibrosis in HFE hemochromatosis. Mayo Clinic Proceedings. PMID 35422339. N=112. Absence of arthritis ruled out advanced liver scarring 95 times in 100.
Barg A, Elsner A, Hefti D, et al. 2011. Total ankle arthroplasty in patients with hereditary hemochromatosis. Clinical Orthopaedics and Related Research. PMID 20665138. 16 patients, 21 implants, mean 5.3 years. Pain 6.7 to 1.9; function score 46 to 84.
Agarwal AR, Wang KY, Xu AL, et al. 2022. Postoperative complications in hereditary hemochromatosis undergoing total joint arthroplasty. Journal of the American Academy of Orthopaedic Surgeons. PMID 34932508. Matched cohort. More stiffness after hip replacement, more loosening after knee replacement.
European Association for the Study of the Liver. 2022. EASL Clinical Practice Guidelines on haemochromatosis. Journal of Hepatology. PMID 35662478. The clinical practice guideline. Diagnostic thresholds and treatment targets.
Barton JC, Parker CJ. HFE-Related Hemochromatosis. GeneReviews, contributed 2000, revised 2024. PMID 20301613. Treatment targets, what to avoid, family testing, and the statement that joint management does not differ.

Your body hoards iron for decades, and it attacks the wrong knuckles.