Unexpected increased ferritin concentration in patients with anorexia nervosa

We enjoyed the recent review article on the clinical biochemistry of anorexia nervosa ¹ and editorial addressing hyperferritinaemia ² in the journal, but would like to add our observations regarding ferritin. Despite presumably reduced iron intake, we have observed and confirmed that raised ferritin concentration in the absence of infection/inflammation is quite common in patients with anorexia nervosa. ³

To confirm our observations, a data mining exercise was designed to extract ferritin data from our population of patients with anorexia nervosa. A registry of 72 consecutive patients with eating disorders presenting to the Women’s and Children’s Hospital from 2002 to 2012 were reviewed for iron studies, liver function tests, complete blood picture and markers of inflammation. A relative risk analysis was also performed against all iron studies performed on an adolescent population tested for whatever reasons between 2009 to 2011 inclusive (n = 467).

Of the 72 anorexia nervosa patients, 18 (25%) had increased ferritin (>200 µg/L). Of those 18, C-reactive protein was measured and was normal in 5. Thirteen had no C-reactive protein measured, but assessment of globulins, albumin/globulin ratio, erythrocyte sedimentation rate and white cell count suggested that inflammation was not significant.

A review of the liver function tests confirmed that the increased ferritin concentration was not due to hepatic injury as manifest in plasma enzyme activities. The anorexia nervosa patient pool showed normal mean corpuscular volume and haemoglobin concentration suggesting that iron deficiency was unlikely. Relative risk analysis showed that patients with anorexia nervosa were six times more likely to have ferritin greater than 200 µg/L than a similar population without anorexia nervosa.

These are not novel observations. A pubmed review dated 21 April 2013 using the search terms ‘ferritin’ and ‘anorexia nervosa’ yielded five papers describing ferritin measurements in patients with anorexia nervosa.^3–7 All reports studied less than 30 patients with anorexia nervosa. One report described reduced ferritin concentrations. ⁴ The others showed variable increases in ferritin concentration in patients with anorexia nervosa.

While we are unaware of the pathophysiology of this phenomenon, several factors present in anorexia nervosa may well affect ferritin concentrations.^3,7,8 Ferritin is a large protein composed of 24 subunits which is a mixture of heavier (H) and lighter (L) chains. The ratio of H:L depends on the tissue source of the ferritin and some forms of ferritin (e.g. liver) contain much more iron than others.^2,9 Serum ferritin concentration, widely used to assess iron stores, is known to be increased in liver injury, malignancy, inflammation and infection, thus complicating its interpretation. Furthermore, ferritin is also thought to have significant non-iron storage roles including protection from oxidative damage. ¹⁰ Given these roles, the significance of perturbations in ferritin concentrations in disease states including anorexia nervosa is uncertain.

Due to the catabolic nature of anorexia nervosa, other isoferritins may also be increased. ⁶ While iron studies are used to review iron stores within the body, measured ferritin with routine commercial assays may not be assessing liver ferritin specifically, but also other isoferritins.

Hepcidin, one of the key regulators of iron metabolism/homeostasis may be involved. A recent study by Papillard-Marechal et al. ³ shows a correlation between elevated hepcidin and ferritin concentrations in anorexia nervosa. Refeeding syndrome and acute malnourishment may induce stress in hepatocytes, thus increasing hepcidin and ferritin. ³

Approximately one quarter of the anorexia nervosa patients seen at our hospital have increased ferritin concentrations. While the clinical significance of this finding is uncertain, it should be recognised as an unexplained but expected anomaly.