The Ferritin Threshold

A ferritin of 4.5 µg/L. A ferritin of 7. A ferritin of 11. A ferritin of 13. All read as “in range” by clinicians reading the reference interval printed on the lab report — often after two, five, ten years of symptoms attributed to anxiety, depression, or the ordinary tiredness of being alive.

Behind Every Test, an Industry names iron as the case where the medical system gets it right — the bilan martial, a real multi-marker functional assessment, corrected for inflammation by concomitant CRP, capable of answering the question of tissue iron adequacy. That claim stands. It is what makes iron the anchor case of that essay, and it is why the failure this editorial names is subtler, and in its own way more revealing, than the four other cases traced there.

The tool is honest. Its routine institutional use is not.

The threshold

The World Health Organization’s cutoff for iron deficiency in non-pregnant adult women is a serum ferritin of 15 µg/L. Many clinical laboratories print reference ranges that start below this — 10, 12 µg/L, depending on the platform and the population that generated the interval. In routine primary care, ferritin is often the only marker of the panel actually ordered. Transferrin saturation, soluble transferrin receptor, CRP to correct for the inflammatory rise of ferritin, reticulocyte haemoglobin — these are technically available; they are also, in most non-specialist settings, not requested.

The scientific literature has been arguing for higher functional thresholds for over a decade. Mei and colleagues at the US CDC (American Journal of Clinical Nutrition, 2017) proposed 25 µg/L as the healthy-adult floor, on the basis of the value at which iron-dependent haematological markers begin to shift. Pasricha and coauthors, in the Lancet 2021 seminar on iron deficiency, argue for 30 µg/L as the diagnostic threshold, and several national gastroenterology and haematology guidelines have quietly adopted this figure. Auerbach and colleagues, from the clinical practice of intravenous iron therapy, have argued that symptomatic iron deficiency without anaemia often requires treatment at ferritin values up to 50 µg/L, sometimes higher in athletes and menstruating women whose functional demand is elevated. In restless legs syndrome, the treatment threshold favoured by International Restless Legs Study Group consensus documents is around 75 µg/L. In cognitive and mood symptoms, several smaller studies point to functional improvement above 100 µg/L.

The distance between the routine cutoff (10–15) and the functional thresholds proposed by the specialty literature (30–100) is where the failure of the test lives. It is not a failure of biochemistry. It is a failure of interpretation.

The category the panel does not name

Iron deficiency anaemia is the recognised endpoint. Haemoglobin falls, the complete blood count declares itself, the reticulocyte response confirms. Iron deficiency without anaemia is what the routine panel does not name. Ferritin can be at 8 while haemoglobin is still 130. The stores are depleted; the compensatory erythropoiesis is still holding. The patient is severely deficient in the sense that matters biologically — reduced tissue iron for oxidative phosphorylation, for dopaminergic neurotransmission, for thyroid peroxidase, for immune function — while the anaemia panel returns “normal.”

The clinical entity exists in the literature under the name iron deficiency without anaemia (IDWA). Its symptoms are precisely those that primary care most consistently attributes to anxiety, depression, deconditioning, or the ordinary fatigue of being alive: unexplained tiredness disproportionate to activity, hair loss, restless legs at night, tachycardia and palpitations, brain fog, cold extremities, breathlessness on stairs, brittle nails, unusual cravings for ice or earth. The populations most affected are menstruating women — particularly those with heavy menstrual bleeding — endurance athletes, regular blood donors, patients with coeliac or inflammatory bowel disease, and adolescents through their growth spurt.

The routine reading of the panel does not detect IDWA because IDWA presents with a normal haemoglobin and a ferritin that the reference range calls normal.

What the patient community already knows

A recurring pattern in patient discussions online, across communities organised around chronic fatigue, fibromyalgia, dysautonomia, POTS, menopause, PCOS, hair loss, and unexplained anxiety, is the following sequence. Symptoms present. Bloodwork returns. The clinician reads the ferritin at the top of a two-digit number and pronounces it in range. Two to eight years pass. The patient encounters, often by accident — a friend, a naturopath, a chance conversation, an online forum — the possibility that ferritin has a functional threshold above the population reference range. A repeat draw, or an old value reviewed with new eyes, reveals a ferritin at 4, 7, 11, 13. Treatment — oral iron for months, often subsequently intravenous iron — produces improvement in weeks that the patient describes as recovering a decade of lost life.

The consistency of this narrative across thousands of individual accounts is data of a specific kind. It does not carry the epistemic force of a randomised controlled trial. It carries a different epistemic force: it maps the population of patients failed by the routine reading of an honest test. The treatment worked. The test could have identified the deficiency at any of the earlier draws. The clinical practice did not.

The industries behind the threshold

Behind Every Test, an Industry traces the industries that shape which nutrients get tested at all. The ferritin threshold traces a distinct question: which industries have an interest in a threshold set at the bottom of the population distribution rather than at the level associated with function?

The dietary iron industries — the red-meat lobby, mandatory fortification of milled grain across roughly eighty-five countries, prenatal supplementation, infant formula — carry a public-health message whose institutional coherence depends on inadequacy being rare. A threshold at 15 µg/L reports low iron in the small tail of the distribution and validates the fortification framework: the system is working; the exceptions are individual and correctable with dietary advice. A threshold at 30 or 50 would report deficiency in a substantial fraction of premenopausal women in industrialised populations. It would surface an ambient failure the fortification framework does not, in fact, address, and it would trigger treatment costs the health systems have not budgeted for.

Downstream, the intravenous iron market — ferric carboxymaltose (Injectafer, Ferinject), ferric derisomaltose (Monoferric), ferumoxytol (Feraheme) — is expensive per unit and requires specialist infrastructure. A low ferritin threshold contains referral. A patient at ferritin 12 is told to eat spinach. A patient at ferritin 4, hospitalised for anaemic collapse, is referred. The threshold is the gate that separates the two.

The industries do not, in most cases, actively lobby for a low threshold. They benefit from its inertia. And a technical detail compounds it: the reference range on the lab printout is generated by each laboratory from its local population’s percentile distribution — which, in a population where iron deficiency is widespread, is itself depressed. A ferritin of 12 sits inside the population’s own lower percentile. The reference range is the population, not the biology. And the population is, on this specific nutrient, deficient.

The physiological right

Every person has the right to have their iron status assessed against the value associated with tissue function, not the value associated with the bottom of the local population’s distribution.

The distinction is not academic. It determines whether five years of unexplained fatigue in a thirty-five-year-old woman are absorbed into her personality, treated with an SSRI, or reversed by a course of iron. It determines whether a teenager’s declining school performance is read as adolescence, or as the visible face of a deficiency a single reinterpreted test could have caught. It determines whether the hair loss, the cold extremities, the restless legs, the tachycardia at rest are named as iron insufficiency or dispersed across four specialties none of whom will look at ferritin again.

The right to a threshold that reflects biological need is a physiological right in the same sense as the right to be measured at all. The most honest test in the field, read at the wrong number, is functionally equivalent to no test. Even the best test in modern medicine can be defeated by the value chosen to interpret it.