Iron deficiency is neglected in women's health
BMJ 2002; 325 doi: https://doi.org/10.1136/bmj.325.7373.1176/a (Published 16 November 2002) Cite this as: BMJ 2002;325:1176
- D Hugh Rushton, honorary senior lecturer (rushton{at}btinternet.com),
- Robin Dover, clinical research officer,
- Anthony W Sainsbury, senior veterinary officer,
- Michael J Norris, principal lecturer,
- Jeremy J H Gilkes, consultant dermatologist,
- Ian D Ramsay, consultant endocrinologist
- School of Pharmacy and Biomedical Sciences, University of Portsmouth, Portsmouth PO1 2DT
- Cancer Research UK, London WC2A 3PX
- Institute of Zoology, Zoological Society of London, London NW1 4RY
- School of Pharmacy and Biomedical Sciences, University of Portsmouth
- Lister Hospital, London SW1W 8RH
EDITOR—The data by Waalen et al show that 38% of menstruating women living in San Diego are iron deficient using the accepted cut-off point of <20% for transferrin saturation as an indicator of deficiency. These data support our hypothesis that haematological distributions contain a large proportion of iron deficient women.1
By focusing on the mean haemoglobin distribution, Waalen et al lose sight of a fundamental part of our work: why should women have lower reference limits for haemoglobin and serum ferritin concentrations than men?2 Our hypothesis for this was that a significant number of women are iron deficient. Mean values mask anomalies in distribution and are inappropriate to evaluate the issues we have raised. Since women menstruate, men would be expected to have higher mean haemoglobin concentrations because they can attain higher upper limits, so the 10g/l difference found by Waalen et al was not unexpected.
A similar situation occurs in menstruating but not in non-menstruating non-human primates.3 To support their case Waalen et al cite one menstruating non-human primate example—unfortunately this paper included data from infants, adolescents, and female animals half the age of the males, and it is therefore unreliable.4 Data analysis of a sex difference with age is more complex than simply presenting mean values from uncontrolled populations.
Close inspection of the ferritin data in part C of Waalen et al's study shows a large number of women “falling off” the lower end of the distribution. The haemoglobin difference between the sexes was of the range 15-20 g/l, which represents a significant proportion (39%) of the female distribution having a serum ferritin concentration below the lowest male value. In part B, the transferrin saturation data for 26-55 year old menstruating women show a significant difference in haemoglobin concentration of 15-25 g/l with a transferrin saturation below 20%, the accepted cut-off point for iron deficiency. In postmenopausal women (part D) 29% were affected. Ten per cent were classified as normal (haemoglobin concentration>120g/l) but were clearly iron deficient on the basis of having transferrin saturations <10%.
Iron deficiency is a significant and seemingly neglected factor in women's health. Why should the lower reference level be sex dependent? What compelling evidence is there to support this apparent anomaly? In this context men are non-menstruating women, haematologically speaking.