Paediatric Iron Deficiency Interpreter
Paediatric Iron Deficiency Interpreter
Ferritin is the best single test for iron deficiency in a child and it is also an acute-phase reactant, so a normal ferritin with a raised CRP does not exclude deficiency — it means the test has been neutralised. This page reads ferritin and transferrin saturation against the age-banded decision thresholds three bodies publish, which disagree: the WHO uses 12 and 15 µg/L, the AAP and ASH use 20 and 30 ng/mL. All three are printed. What is not printed is an age-banded paediatric ferritin reference interval, because no such thing is soundly published — for that, ask your own laboratory.
Ferritin, saturation and the CRP
Ferritin + CRP + age → verdictA 3-year-old with pallor and poor appetite, drinking about a litre of cow’s milk a day. Ferritin 14 µg/L, transferrin saturation 18 per cent, CRP 2 mg/L and the child is well. Haemoglobin 98 g/L, which is below the WHO cut-off of 110 g/L for 24 to 59 months.
Three bodies, three sets of thresholds, and no reference interval at all
AAP (2026) — at or below 20 ng/mL in young and school-aged children; at or below 30 ng/mL in adolescents and menstruating individuals.
ASH (2026) — at or below 20 ng/mL in children aged 9 months to 4 years; at or below 30 ng/mL in adults and menstruating or pregnant individuals; at or below 50 ng/mL in high-risk groups including heavy menstrual bleeding; and, with inflammation, ferritin below 100 ng/mL or transferrin saturation below 20 per cent.
Units — 1 ng/mL = 1 µg/L for ferritin. No conversion sits between these numbers; the gap between 12 and 20 is disagreement.
Not published — a universal age-banded paediatric ferritin reference interval. Use your own laboratory’s.
- an acute-phase reactant
- the whole difficulty. Ferritin rises with inflammation independently of iron stores, so the error runs one way: a deficient child can have a normal ferritin, but a replete child cannot have a low one. That asymmetry is what makes a low ferritin diagnostic and a normal ferritin uninformative without a CRP beside it
- 12 against 20
- the disagreement, in one line. The WHO’s thresholds are set to define population prevalence with high specificity; the paediatric and haematology guidelines set theirs to catch treatable individuals, including those with iron deficiency without anaemia. Neither is wrong for its own purpose, and for an individual child in a clinic the higher, more recent threshold is the one this page follows
- iron deficiency without anaemia
- the reason the thresholds moved up. Stores empty long before the haemoglobin falls, and the intervening state is associated with impaired neurodevelopment in the first years and with fatigue, poor concentration and reduced exercise tolerance later. Waiting for anaemia means waiting through the part of the illness that is treatable and reversible
- transferrin saturation
- serum iron divided by total iron binding capacity, and a weaker test in a child than it looks. Serum iron swings diurnally, falls with any acute illness independently of stores, and rises for hours after an oral iron dose — so one saturation is a snapshot of a moving number. Repeat it fasting, in the morning, off iron for 24 hours, before acting on it
- 500 mL of cow’s milk
- the commonest cause of iron deficiency in a toddler, roughly. Above about that volume a day, milk both displaces iron-rich food and causes occult gastrointestinal blood loss, and correcting it is often most of the treatment. The figure is a practical rule rather than a threshold from a trial
Worked example
A 3-year-old with pallor and poor appetite, drinking about a litre of cow's milk a day. Ferritin 14 µg/L, transferrin saturation 18 per cent, CRP 2 mg/L and the child is well. Haemoglobin 98 g/L, which is below the WHO cut-off of 110 g/L for 24 to 59 months.
CRP is normal and the child is well, so none of the three inflammation branches applies and the ferritin can be read at face value
Age 2 to 4 years, so the adolescent threshold of 30 does not apply
Ferritin 14 is at or below 20, the AAP and ASH threshold for a child of this age → iron deficiency
Note what the WHO would say: 14 µg/L is above its threshold of 12 µg/L for a child under 5, so this result is iron deficiency by one document and not by another. 14 ng/mL and 14 µg/L are the same number, so there is no conversion hiding in that disagreement
The haemoglobin of 98 g/L is below the WHO 2024 cut-off of 110 g/L for 24 to 59 months, so this is iron deficiency anaemia rather than iron deficiency alone
The cause is in the history: about a litre of cow's milk a day is well above the volume at which milk displaces iron-rich food and causes occult gastrointestinal blood loss
So: reduce the milk, treat with oral iron at the locally recommended dose, and recheck both the haemoglobin and the ferritin afterwards — a haemoglobin that has recovered while the ferritin has not means the stores are still empty and the course is not finished
Ferritin thresholds for iron deficiency — the three published sets
| Group | WHO 2020 | AAP 2026 | ASH 2026 |
|---|---|---|---|
| 6 to 23 months | Below 12 µg/L | At or below 20 | At or below 20 (9 months to 4 years) |
| 2 to 4 years | Below 12 µg/L | At or below 20 | At or below 20 |
| 5 to 11 years | Below 15 µg/L | At or below 20 | — |
| 12 years and over, not menstruating | Below 15 µg/L | At or below 30 | At or below 30 |
| Menstruating adolescent | Below 15 µg/L | At or below 30 | At or below 30; at or below 50 with heavy menstrual bleeding |
| With infection or inflammation | Below 30 µg/L under 5; below 70 µg/L from 5 | Ferritin may be falsely normal or high | Below 100, or transferrin saturation below 20% |
What a ferritin and a CRP mean together
| Ferritin | CRP normal | CRP raised |
|---|---|---|
| Low for age | Iron deficiency. Find the cause and treat | Still iron deficiency. Inflammation cannot lower a ferritin, only raise it — so a low one with inflammation is beyond argument |
| Between the age threshold and 100 | Iron deficiency not demonstrated; repeat if the suspicion is strong | Iron deficiency likely. The ordinary threshold is invalid here; the raised threshold applies |
| 100 or more | Iron deficiency unlikely | Uninformative about stores. Read the transferrin saturation instead, and repeat once the CRP has settled |
A normal ferritin is only reassuring if the CRP is normal too
Iron deficiency is the commonest nutritional deficiency in childhood and ferritin is the single best test for it, because serum ferritin tracks total body iron stores closely across most of the range. The difficulty is that ferritin is also an acute-phase reactant, rising with inflammation independently of iron status. The error runs in one direction only, and that asymmetry is the most useful thing to hold onto: inflammation can push a deficient child’s ferritin up into the normal range, but nothing pushes a replete child’s ferritin down. A low ferritin is therefore diagnostic whatever the CRP is doing. A normal ferritin means very little unless a CRP was measured alongside it and was normal.
The thresholds themselves are a source of genuine confusion, because three respected bodies publish three different sets and they do not agree. The WHO’s 2020 guideline defines iron deficiency at a ferritin below 12 µg/L under five years and below 15 µg/L from five, raising those to 30 and 70 µg/L respectively where there is infection or inflammation. The 2026 American Academy of Pediatrics clinical report uses 20 ng/mL or below for young and school-aged children and 30 ng/mL or below for adolescents and menstruating individuals. The 2026 American Society of Hematology guideline matches those and adds 50 ng/mL or below for high-risk groups such as heavy menstrual bleeding, and, in people with inflammation, a ferritin below 100 ng/mL or a transferrin saturation below 20 per cent. Because ferritin in ng/mL and in µg/L are numerically identical, none of that gap is a unit artefact. The WHO’s numbers are set for population surveillance, where specificity matters most; the clinical guidelines’ numbers are set to catch treatable individuals, including the large group with iron deficiency and no anaemia at all.
That last group is why the thresholds moved upward. Iron stores empty long before the haemoglobin falls, and the intervening state is not asymptomatic: it is associated with impaired neurodevelopment in the first years of life, and with fatigue, poor concentration, reduced exercise tolerance and restless legs in older children and adolescents. Treating only once the haemoglobin has fallen means treating at the end of a process rather than during it. It is also why the ferritin, not the haemoglobin, is the test to recheck after treatment: a haemoglobin that has normalised while the ferritin has not means the stores are still empty and the course is not finished.
One thing this page deliberately does not print is an age-banded paediatric ferritin reference interval. No such thing is soundly and universally published — what exists are the decision thresholds above, which are a different kind of number, and laboratory-specific intervals that depend on the assay and the local population. If the question is what a particular ferritin means against a reference range rather than against a guideline threshold, the answer is your own laboratory’s interval, issued with the result. The haemoglobin bands are not restated here either: the anaemia severity classifier already implements the WHO 2024 cut-offs with the paediatric age bands and the altitude adjustment. For the rest of the panel, the iron studies panel interpreter, the functional versus absolute iron deficiency interpreter, the ferritin correction for inflammation calculator and the transferrin saturation calculator each take a piece.
Frequently asked questions
What ferritin level means iron deficiency in a child?
It depends which guideline you open, and the three main ones genuinely disagree. The 2026 AAP clinical report and the 2026 ASH guideline both use 20 ng/mL or below for young and school-aged children and 30 ng/mL or below for adolescents and menstruating individuals, with ASH adding 50 ng/mL or below for high-risk groups such as heavy menstrual bleeding. The WHO’s 2020 guideline uses lower numbers — below 12 µg/L under five years and below 15 µg/L from five — because its thresholds are designed for population surveillance rather than individual diagnosis. Since 1 ng/mL and 1 µg/L are the same thing for ferritin, none of that difference is a unit conversion. This page follows the higher, more recent clinical thresholds, and tells you where the other verdict would differ.
Can a child have iron deficiency with a normal ferritin?
Yes, and inflammation is the usual reason. Ferritin is an acute-phase reactant, so any acute or chronic inflammatory illness raises it independently of iron stores and can lift a genuinely deficient child’s result into what looks like the normal range. That is why a CRP should be measured with every paediatric ferritin. If the CRP is raised, the ordinary thresholds do not apply: the current haematology guideline uses a ferritin below 100 ng/mL, or a transferrin saturation below 20 per cent, to diagnose iron deficiency in the presence of inflammation, and the WHO raises its own thresholds from 12 to 30 µg/L under five years and from 15 to 70 µg/L from five. The error only runs one way, though — inflammation cannot make a ferritin falsely low, so a low ferritin is iron deficiency whatever the CRP shows.
Is there a published age-banded reference interval for ferritin in children?
Not a universal one, and this page will not print an invented one. What is published are decision thresholds — the guideline numbers above, which answer "should this child be treated for iron deficiency?" rather than "is this result within the normal distribution for this age?" Actual reference intervals for paediatric ferritin are assay-specific and population-specific, and the interval that applies to a given result is the one your laboratory issues with it. If you need a reference interval rather than a decision threshold, ask the laboratory that ran the assay; do not borrow one from a textbook or from another laboratory’s report.
Does a normal transferrin saturation rule out iron deficiency in a child?
No, and a single saturation is weaker evidence than most people assume. Transferrin saturation is serum iron divided by total iron binding capacity, and serum iron — the numerator — has a pronounced diurnal rhythm with morning values well above afternoon ones, falls transiently with any acute illness independently of iron stores, and rises for several hours after an oral iron dose. So a saturation measured in the afternoon can read low in a replete child, and one measured a few hours after the morning iron dose can read normal in a deficient one. Repeat it fasting, in the morning, at least 24 hours after any iron, before acting on it. Ferritin remains the better first test in a child with a normal CRP.
Why has the ferritin threshold for children gone up?
Because the target of treatment has changed from iron deficiency anaemia to iron deficiency itself. Stores empty well before the haemoglobin falls, and that intervening state — iron deficiency without anaemia — is common in toddlers and in menstruating adolescents and is not benign: it is associated with impaired neurodevelopment in the first years and with fatigue, poor concentration, reduced exercise tolerance and restless legs later. Thresholds set to define anaemic populations were therefore too low to identify the children who would benefit from treatment, and the 2026 guidelines moved them to 20 ng/mL in children and 30 ng/mL in adolescents. The practical consequence is also that the ferritin, not the haemoglobin, is the test that tells you when treatment can stop.
What should I check first in a toddler with iron deficiency?
The milk intake. Excessive cow’s milk is the commonest cause of iron deficiency in a toddler and it works two ways at once: it displaces iron-rich foods from the diet, and it causes occult gastrointestinal blood loss. Roughly 500 mL a day is the volume above which this becomes a problem. After that, ask about the rest of the diet, about coeliac disease and other malabsorption, and about any overt blood loss. In an older child or one whose intake is unremarkable, iron deficiency needs a cause found rather than just corrected. And recheck the ferritin as well as the haemoglobin after treatment, because a normalised haemoglobin with a still-low ferritin means the stores have not been replaced.
Related calculators
References
- World Health Organization. WHO guideline on use of ferritin concentrations to assess iron status in individuals and populations. Geneva: World Health Organization; 2020. ISBN 978-92-4-000012-4. Table 1: iron deficiency at a serum ferritin below 12 µg/L in infants and children aged 0–23 months and 24–59 months, and below 15 µg/L from 5 years; in the presence of infection or inflammation, below 30 µg/L under 5 years and below 70 µg/L from 5 years.
- Powers JM, Heeney MM, Hord J, et al; American Academy of Pediatrics Section on Hematology-Oncology, Committee on Nutrition, and American Society of Pediatric Hematology-Oncology. Prevention, screening, diagnosis, and treatment of iron deficiency and iron deficiency anemia in infants, children, and adolescents: clinical report. Pediatrics. 2026;158(1):e2026077414. doi:10.1542/peds.2026-077414. Ferritin at or below 20 ng/mL is consistent with iron deficiency in young and school-aged children, and at or below 30 ng/mL in adolescents and menstruating individuals. Screening with a full blood count and ferritin at 9–12 months in breastfed infants, 15–18 months in formula-fed infants after transition to cow or plant milk, and at least one year post-menarche or by age 14 in menstruating adolescents. Verbatim: "Ferritin is an acute-phase reactant and therefore can be elevated in a patient who is anemic but also has an acute or chronic inflammatory condition."
- Powers JM, Lim MY, Achebe MO, et al. American Society of Hematology 2026 guidelines for diagnosis of iron deficiency. Blood Adv. 2026. doi:10.1182/bloodadvances.2025015950. Ferritin at or below 20 ng/mL in children aged 9 months to 4 years; at or below 30 ng/mL in adults and in menstruating or pregnant individuals; at or below 50 ng/mL in high-risk groups including heavy menstrual bleeding; and, in people with inflammation, a ferritin below 100 ng/mL or a transferrin saturation below 20%.
- World Health Organization. Guideline on haemoglobin cutoffs to define anaemia in individuals and populations. Geneva: World Health Organization; 2024. The cut-offs referred to by the haemoglobin input on this page: 105 g/L at 6–23 months, 110 g/L at 24–59 months, 115 g/L at 5–11 years and 120 g/L at 12–14 years. The severity bands are not restated here because this site’s anaemia severity classifier already implements them.
Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.