Ingram DG, Al-Shawwa B. Serum ferritin in the pediatric sleep clinic: what’s normal anyway? J Clin Sleep Med. 2019;15(11):1699–1700.
We read with great interest the study by Dr. Rosen and colleagues in a recent issue of the Journal of Clinical Sleep Medicine.1 They examined the effect of iron therapy in children who had restless legs syndrome (RLS) and serum ferritin levels less than 50 ng/mL (which they defined as biochemical evidence of iron deficiency). Although overall, iron supplementation did increase serum ferritin levels and decrease median RLS scores compared to baseline, there was no correlation between the observed change in symptom scores and serum ferritin levels.
In pediatric sleep medicine, we routinely measure serum ferritin levels as a surrogate of iron stores in the body when assessing children with RLS, periodic limb movement disorder (PLMD), or simply self-reported restless sleepers.2 This last category is also an area of active research, with Dr. DelRosso characterizing a proposed new diagnostic category of “restless sleep disorder,”3 in which children who do not qualify for the diagnosis of RLS/PLMD but who have self-reported restless sleep and daytime consequences may respond to iron supplementation.
So, are children who present to a pediatric sleep center actually different in terms of their ferritin levels compared to those children in the general population? Using unpublished data, we examined ferritin levels in children who underwent a sleep study at our pediatric sleep center (excluding those who were found to have sleep apnea). We grouped them according to age (ages 1 to 5 years, 6 to 11 years, and 12 to 19 years), and mean ferritin levels were 24.7 (95% confidence interval 23.4–26.0, n = 535), 31.4 (29.4–33.3, n = 411), and 35.0 (29.6–40.4, n = 112). These levels correspond very closely to general population levels found in national surveys,4 as astutely mentioned in Dr. Rosen’s introduction. Clearly, if we were to define iron deficiency as less than 50 ng/mL, most children in the United States would fall into that category.
We, along with most of the pediatric sleep medicine practitioners (anecdotally), agree that children with RLS, PLMD, or restless sleep clinically improve with supplemental iron. Why, then, do ferritin levels not consistently correspond with symptom improvement? One possibility is serum ferritin does not accurately reflect brain iron stores, and therefore will not reflect the resulting symptoms. This is supported by prior studies demonstrating significant differences in cerebrospinal fluid ferritin level but not in serum ferritin levels between adults with or without RLS.5,6
Finally, we suggest care be taken with the terminology we use when discussing ferritin levels with our patients and families. Telling a family that their child’s ferritin level of 20 or 30 is “low” or represents “iron deficiency,” is simply not accurate and may result in excessive parental worry about underlying pathology. Instead, we suggest describing a ferritin level in the 20s or 30s as simply “suboptimal” for the restless sleeper, but not representing iron deficiency. Although following a serum ferritin level in a given patient with RLS/PLMD may be helpful to ensure that they are not becoming iron overloaded with supplementation, the data do not substantiate it as a predictor of clinical improvement.
All authors have seen and approved the final manuscript. No funding was secured for this study. Dr. Ingram has served as a consultant for Jazz Pharmaceuticals. Otherwise, the authors have no financial interests or conflicts of interested relevant to this manuscript.