Intended for healthcare professionals

Letters

Reference limits for haemoglobin and ferritin

BMJ 2001; 323 doi: https://doi.org/10.1136/bmj.323.7316.806 (Published 06 October 2001) Cite this as: BMJ 2001;323:806

If it's not broken, don't fix it

  1. Anne-Louise M Heath, postdoctoral fellowa,
  2. Susan Fairweather-Tait, head of division,
  3. Mark Worwood, professor
  1. Nutrition and Consumer Science Division, Institute of Food Research, Norwich Research Park, Colney, Norwich, NR4 7UA
  2. Department of Haematology, University of Wales College of Medicine, Cardiff, CF14 4XN
  3. Southern Community Laboratories, Dunedin, New Zealand 9001
  4. Department of Human Nutrition, University of Otago, Dunedin, New Zealand 9005
  5. School of Pharmacy and Biomedical Sciences, University of Portsmouth, Portsmouth PO1 2DT
  6. Imperial Cancer Research Fund, London WC2A 3PX
  7. Institute of Zoology, Zoological Society of London, London NW1 4RY
  8. Lister Hospital, London SW1W 8RH

    EDITOR—Rushton et al suggest that reference ranges for men should be used to assess iron status in women of childbearing age.1 They do, however, make some incorrect assumptions and do not consider the practical implications of such a change.

    Women must have sufficient iron stores to prevent iron deficiency from menstrual blood loss or pregnancy. However, one in 150 people in the United Kingdom are homozygous for the C282Y mutation of the HFE gene, which is associated with haemochromatosis.2 Although the clinical penetrance of this genotype is low, widespread measures to increase the intake of iron in younger women will also increase the intake of men and postmenopausal women. It is therefore important that any changes in lower limits of indices of iron status are firmly supported by clinical and experimental evidence.

    Rushton et al are incorrect in assuming that different lower limits for ferritin are used for detecting iron deficiency in young men and women. Although reference ranges (95%) in healthy young adults differ—for example, 35-220 µg/l for men and 9-136 µg/l for women2—the limit for iron deficiency used by clinicians is around 15 µg/l for both men and women, a value originally established by determining the highest value found in patients with iron deficiency anaemia.3

    The national dietary and nutritional survey reported a median value for haemoglobin concentration in British women of childbearing age of 132 g/l. Increasing the lower cut-off point to 130 g/l, as used for men, would therefore define half the premenopausal adult female population of the United Kingdom as anaemic. Zhu and Haas found that in women with serum concentrations of ferritin <16 µg/l and haemoglobin >120 g/l, iron supplementation increased serum concentrations of ferritin but not haemoglobin.4 We also have to ask how the iron intake of all these women would be increased. A recent dietary intervention study showed that highly motivated people with mild iron deficiency can improve iron status through diet but that supplements are a more practical option.5 Supplements do, however, produce unpleasant side effects in a notable proportion of individuals, and any programme entailing the use of supplements is likely to have a detrimental effect on the wellbeing of a notable number of women.

    We believe that there is no evidence to support reclassification of haemoglobin and serum concentrations of ferritin in women to normal values for men. Furthermore, we are unable to see how such a move could result in a positive outcome for women's health and welfare with no adverse effects.

    Footnotes

    • a annelouise.heath{at}bbsrc.ac.uk

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.

    Differences in haemoglobin concentrations reflect physiological differences

    1. Ian M Morison, haematologist (ian.morison{at}sclabs.co.nz),
    2. Elaine L Ferguson, lecturer
    1. Nutrition and Consumer Science Division, Institute of Food Research, Norwich Research Park, Colney, Norwich, NR4 7UA
    2. Department of Haematology, University of Wales College of Medicine, Cardiff, CF14 4XN
    3. Southern Community Laboratories, Dunedin, New Zealand 9001
    4. Department of Human Nutrition, University of Otago, Dunedin, New Zealand 9005
    5. School of Pharmacy and Biomedical Sciences, University of Portsmouth, Portsmouth PO1 2DT
    6. Imperial Cancer Research Fund, London WC2A 3PX
    7. Institute of Zoology, Zoological Society of London, London NW1 4RY
    8. Lister Hospital, London SW1W 8RH

      EDITOR—Rushton et al argued that a high prevalence of iron deficiency among women has resulted in the use of haemoglobin reference ranges that are lower than ideal, and that male reference ranges should be used when assessing the iron status of women.1

      We have used two independent sources of data to show that sex differences in haemoglobin concentrations reflect real physiological differences rather than the prevalence of iron deficiency. We have used specific exclusion criteria (ferritin, age, pregnancy) to ensure that the two groups were comparable and iron replete. The first data set, extracted from our community laboratory database, consists of 612 men and 1327 women aged 20-45 years, all of whom have a serum ferritin value within the narrow range of 80-100 µg/l (indicating adequate iron status). The 10th, 25th, 50th, 75th, and 90th centiles for men and women were 139, 146, 152, 157, 163, and 123, 129, 135, 141, 146 g/l, respectively.

      To confirm that these differences also existed in a community based population, we provide data from a random nutrition survey of 4636 New Zealand adults.2 From the survey population we selected all non-pregnant participants aged 20-45 years with a ferritin concentration of 50-100 µg/l and a normal concentration of C reactive protein (to exclude participants with anaemia due to chronic inflammatory conditions). The 10th, 25th, 50th, 75th, and 90th centiles for men and women were 140, 146, 150, 157, 160 (n=110) and 127, 131, 138, 142, 147 g/l (n=252), respectively. These results confirm our conclusion that the distribution of haemoglobin concentrations is substantially lower in women than in men.

      This analysis also shows that, depending on the choice of the lower limit of haemoglobin for men (usually 130-135 g/l), 25-50% of women would be incorrectly classified as anaemic if a male reference range was used. This would have a profound public health impact owing to indiscriminate diagnosis of poor iron status in women.

      With respect to ferritin concentrations, we agree that marked sex differences in ferritin distributions in a “normal” population occur because the iron requirements of menstruating women are higher and intakes lower than their male counterparts, resulting in lower iron stores among women.3 For ferritin, the lower limit of the reference range is more difficult to extract from population studies and should perhaps instead be based on biological definitions of iron deficiency, which include depletion of bone marrow iron stores. The appropriate ferritin cut-off value to use for defining iron deficiency is under debate,3-5 but most would agree that sex specific cut-off points are inappropriate for ferritin when defining iron deficiency.

      References

      1. 1.
      2. 2.
      3. 3.
      4. 4.
      5. 5.

      Authors' reply

      1. Hugh Rushton, honorary senior lecturer (rushton{at}btinternet.com),
      2. Michael J Norris, senior lecturer,
      3. Robin Dover, clinical research officer,
      4. Anthony W Sainsbury, senior veterinary officer,
      5. Jeremy J H Gilkes, consultant dermatologist,
      6. Ian D Ramsay, consultant endocrinologist
      1. Nutrition and Consumer Science Division, Institute of Food Research, Norwich Research Park, Colney, Norwich, NR4 7UA
      2. Department of Haematology, University of Wales College of Medicine, Cardiff, CF14 4XN
      3. Southern Community Laboratories, Dunedin, New Zealand 9001
      4. Department of Human Nutrition, University of Otago, Dunedin, New Zealand 9005
      5. School of Pharmacy and Biomedical Sciences, University of Portsmouth, Portsmouth PO1 2DT
      6. Imperial Cancer Research Fund, London WC2A 3PX
      7. Institute of Zoology, Zoological Society of London, London NW1 4RY
      8. Lister Hospital, London SW1W 8RH

        EDITOR—Heath et al acknowledge that women should be iron replete. The potential benefits far outweigh the possible overload affecting a very small proportion in whom monitoring may be employed. 1 2 They say that the study by Zhu and Haas was too short and that both tissue iron status and body iron stores were still improving simultaneously and had not yet reached a steady state.3 It was no surprise therefore to see an increase in the ferritin concentration and no rise in the haemoglobin concentration.

        The contribution of Morison and Ferguson is welcomed, but their analysis of the data is flawed. They considered the female population to be iron replete, but the first data set was uncontrolled for the presence of chronic infection or inflammatory disease, and the subjects were unmatched for age or weight. In the second set, they chose a lower ferritin concentration than in the first. The differences in haemoglobin concentrations were substantially lower, but not significant, in women compared with men. They agree that sex specific cut-off points for ferritin are inappropriate. In young people the median values only begin to diverge as menses exerts its effect: 4-6 years (28 µg/l men v 27 µg/l women), 7-10 years, (29 v 32), 11-14 years (28 v 28), and 15-18 years (44 v 23).4

        We believe that a similar case exists for equality of lower haemoglobin values. Hobbs showed 40 years ago that the frequency distribution of haemoglobin concentration in response to iron supplementation in women paralleled that of men (figure). He concluded that the normal female range is not physiological and that iron should be given to all patients with a haemoglobin concentration below 13.6 g/100 ml (136 g/l).

        Figure1

        Frequency distribution of haemoglobin concentrations in response to iron treatment in anaemic men and women compared with “normal” women (reproduced with permission from Hobbs JR. Iron deficiency after partial gastrectomy. Gut 1961;2:141-9)

        Neither Heath et al nor Morison and Ferguson present any data to address the fact that no other mammals (including menstruating non-human primates) exhibit a sex difference for haemoglobin and red blood cell count (ferritin is not assessed in veterinary medicine). They also do not explain why these variables approximate in postmenopausal women and aged matched men, or why there are no significant differences in children. A healthy, iron replete, 30 year old woman has the same haem synthesis as one aged 60.5 What biological evidence is there to employ different reference limits for premenopausal and postmenopausal women, other than widespread iron deficiency due to menses?

        Our aim was to highlight the scale of the underestimation of iron deficiency in menstruating women and the deleterious effects on women's health.2 The practicalities of ensuring that women are iron replete are a separate issue.

        Footnotes

        • A longer version of this letter is available on bmj.com.

        References

        1. 1.
        2. 2.
        3. 3.
        4. 4.
        5. 5.
        View Abstract