Your Ferritin is Normal.
But Is It Optimal?

A client comes to see me. She has been tired for three years. Her GP has run blood tests twice. Both times she was told everything was normal. She isn't anaemic. Her thyroid is fine. Her B12 is fine. She is simply — inexplicably, according to the tests — exhausted.

I look at her ferritin. It is 19 µg/L.

By the NHS reference range, this is normal. The conventional lower threshold for women is somewhere between 10 and 20 µg/L depending on which laboratory is running the test. She has cleared it. Not by much, but she has cleared it. Normal.

I have seen this particular version of "normal" more times than I can count. And I have watched what happens when ferritin is brought up to 70–100 µg/L over three to four months. The fatigue lifts. The brain fog clears. The hair stops falling out at the rate it was. The thyroid conversion — which requires iron as a cofactor — starts working properly again.

This is not anecdote. There is a growing body of peer-reviewed evidence supporting functional ferritin thresholds well above the conventional lower limit. And there is also a legitimate counter-argument that deserves to be heard rather than dismissed. Both matter — because understanding the debate clarifies exactly what is being measured, what the number means, and what to do with it.

The debate — and why both sides have a point

The American Society of Haematology recently proposed raising the lower ferritin threshold to 30 µg/L for both men and women, and creating a new diagnosis: iron deficiency without anaemia. This has split opinion sharply.

The sceptical position is making a valid argument about population screening and the risk of over-medicalisation. The screening position is making a valid argument about reference range methodology and the clinical consequences of under-recognition. They are, to a significant degree, arguing about different things.

The wrong question — and the right one

Both the sceptical and the screening positions are asking the same underlying question: at what ferritin level does a statistically significant proportion of the population develop iron deficiency? This is a reasonable question for screening policy. It is not the question a clinician should be asking about an individual patient.

These are different questions, and they produce different answers. A ferritin of 22 µg/L may be entirely adequate for one person and functionally insufficient for another — depending on their iron demands, their metabolic rate, their thyroid status, their gut absorption capacity, their training load, and a dozen other variables that a single blood marker cannot capture in isolation.

This is precisely why ferritin should be read in context — alongside haemoglobin, serum iron, TIBC, transferrin saturation, and the full clinical picture — rather than as a standalone number with a pass/fail threshold attached.

What ferritin actually does — beyond haemoglobin

The sceptical position focuses heavily on anaemia as the endpoint — and fatigue in the absence of anaemia as an unreliable symptom. This framing misses something important. Iron's role in physiology extends far beyond red blood cell production, and the consequences of sub-optimal iron stores manifest in systems that are well upstream of the point at which haemoglobin drops.

The point is not that every person with a ferritin of 22 µg/L will have all of these consequences. The point is that "normal" ferritin means the population average was not anaemic — it does not mean iron stores are adequate to support the full range of iron-dependent physiological processes.

The reference range problem — a methodological flaw worth understanding

The sceptical camp's strongest argument is about over-medicalisation. Their weakest argument, in my view, is their defence of the existing reference range.

Normal reference ranges in laboratory medicine are typically derived by taking a sample of "apparently healthy" individuals and identifying the central 95% of results. The 2.5th and 97.5th percentiles become the lower and upper limits of normal. This approach has a fundamental vulnerability: if the population used to derive the range contains a significant proportion of undiagnosed iron-deficient individuals — which, given global prevalence, is almost certain for ferritin — then the lower limit of normal is shifted downward by the very condition the range is trying to identify.

This is not a fringe argument. Several haematology researchers have made it explicitly: the lower ferritin reference range is methodologically circular. It was derived from a population that included iron-deficient people, making it an unreliable benchmark for identifying iron deficiency.

What the functional ranges actually say

The question I ask is not "is this ferritin normal?" but "at what ferritin level does this person's physiology work properly?" The research literature — across thyroid conversion, hair follicle cycling, neurotransmitter synthesis, and mitochondrial function — points consistently to functional thresholds well above the conventional lower limit.

What this means in practice

The sceptical camp is right that we shouldn't screen healthy asymptomatic populations with a single ferritin measurement and prescribe iron to everyone below 30. Iron is not a benign supplement — over-supplementation has genuine risks, particularly for those with haemochromatosis or inflammatory conditions where ferritin elevation reflects pathology rather than sufficiency.

The screening camp is right that we've been missing real pathology in real people for decades, using a reference range that was built with a methodological flaw at its foundation, in a population where iron deficiency is the most prevalent micronutrient deficiency globally.

The clinical reality sits between both positions — and it requires individual assessment rather than population policy.

When I see ferritin in a comprehensive blood chemistry panel, I look at it alongside haemoglobin, haematocrit, MCV, serum iron, TIBC, and transferrin saturation. I look at the clinical picture — the fatigue, the hair loss, the cold intolerance, the cognitive symptoms. I look at the thyroid panel, because low ferritin and apparent hypothyroid symptoms often appear together in people whose TSH is technically within range. I look at the OAT for mitochondrial function markers. I look at the full picture, not the single number.

If a client's ferritin is 22 µg/L and they have fatigue, hair loss, poor thyroid conversion, and a clinical history consistent with iron insufficiency, I don't reassure them that their ferritin is normal and send them on their way. I address the iron stores — through diet, through targeted supplementation where indicated, and through retesting to confirm restoration rather than assuming it.

If their ferritin is 22 µg/L and they have no symptoms, good energy, no hair loss, and their thyroid is converting well, I note it, monitor it, and don't medicalise it.

That is the distinction the population debate cannot make. The individual clinical assessment can.

The deeper issue — what "normal" has come to mean

The ferritin debate is a specific instance of a larger problem in conventional laboratory medicine. Reference ranges tell us what is statistically common. They were never designed to tell us what is physiologically optimal. When we use them as though they are the same thing, we end up in the position of reassuring people that a result is normal when what we actually mean is that it falls within the central 95% of a population sample — a sample that may itself have been iron-deficient, vitamin D-insufficient, or metabolically compromised when the range was set.

The question that conventional ranges ask is: how does this compare to the average? The question functional medicine asks is: does this level support optimal physiology in this person? They are different questions. They produce different answers. And the gap between them is often where chronic, complex, unexplained symptoms live.

A ferritin of 19 µg/L is normal. But it is not necessarily optimal. For many people — particularly women of reproductive age, people with thyroid conversion issues, people with unexplained fatigue and hair loss — it is not remotely close to optimal. And the fact that it clears a population-derived reference range should not be the end of the clinical conversation. It should be the beginning.