Ferritin Levels in the Cerebrospinal Fluid and Restless Legs Syndrome: Effects of Different Clinical Phenotypes

RESTLESS LEGS SYNDROME/PERIODIC LIMB MOVEMENT DISORDER Ferritin Levels in the Cerebrospinal Fluid and Restless Legs Syndrome: Effects of Different Clinical Phenotypes Christopher J. Earley, MD, PhD1; James R. Connor, PhD2; John L. Beard, PhD3; Stacey L. Clardy2; Richard P. Allen, PhD1 Department of Neurology, Johns Hopkins School of Medicine, Baltimore, MD; 2Department of Neurosurgery, M.S. Hershey Medical Center, Hershey, PA; 3Department of Nutrition, Pennsylvania State University, College Park, PA 1 Study Objective: To determine whether patients with restless legs syndrome (RLS) and controls differ in regard to levels of ferritin and transferrin in the cerebrospinal fluid (CSF) when samples are collected at night, to determine whether patients with early-onset and late-onset RLS show a different outcome for CSF values, and to determine whether the CSF ferritin level correlates with disease severity. Design: Collection of CSF and plasma; assessment of disease severity using objective (periodic limb movements) and subjective (Johns Hopkins Restless Legs Severity Scale) measures of severity. Setting: General Clinical Research Center. Participants: Thirty subjects with idiopathic RLS (15 early- and 15 lateonset RLS) and 22 age- and sex-matched controls. Intervention: N/A. Results: Nighttime CSF ferritin levels were lower in the total RLS group compared with controls. Further assessment found that the early-onset (less than 45 years of age) but not the late-onset (greater than or equal to 45 years of age) RLS group had significantly lower CSF ferritin levels compared with controls. There was a strong correlation between the age of symptom onset and CSF ferritin values (r = 0.64): the earlier the age, the lower the ferritin level. A regression analysis showed that both sex and RLS subtype had significant effects on the CSF ferritin level, with women with early-onset RLS having substantial lower values than men with late-onset RLS. A comparison between these nighttime CSF values and previously published daytime samples suggests that diurnal changes may have effects on the findings. Conclusions: This study is distinct in showing that the degree of the CSF-ferritin effect is best defined by the clinical phenotypes of sex and age of symptom onset and by the time of day that samples are collected. Keywords: Circadian, restless legs syndrome, CSF ferritin, early-onset, late-onset, gender, iron Citation: Earley CJ; Connor JR; Beard JL et al. Ferritin levels in the cerebrospinal fluid and restless legs syndrome: Effects of different clinical phenotypes. SLEEP 2005;28(9): 1069-1075. in selected brain regions.3,4 The findings of decreased CSF ferritin and increased CSF transferrin levels in patients with RLS compared with unaffected individuals also support the notion of a relative decrease in brain iron status.5 Studies have nonetheless shown a good correlation between low body stores of iron, as determined by serum ferritin levels and symptom severity, with oral iron treatment capable of improving symptoms for patients with reduced body iron stores.6,7 Under condition of normal body iron stores, however, serum markers of iron status appear to be poor predictors of the clinical condition and brain iron status.5 One of the primary characteristics of the syndrome is the occurrence of the symptoms predominantly at night or during sleep.8 Intriguingly, systemic iron metabolism also varies over the day, with a marked drop in serum iron at night.9 Given the aforementioned diurnal changes in symptoms and serum iron, and the fact that all previous comparisons of serum and CSF used daytime sampling, it would be important to explore the relationship between serum and CSF from nighttime samples and their relation to the clinical picture. The phenotypic differences between an early- and late-onset of RLS symptoms have been established primarily for the degree of genetic susceptibility. The first family history study showed greater rates of familial occurrence of RLS for probands whose RLS started before age 45 years.2 A subsequent segregation analyses reported model fits for a dominant single gene for patients with RLS starting by age 30 years and no genetic model for RLS patients with symptoms starting after 30 years of age.10 One study showed a positive correlation between age and symptom severity for patients with RLS symptoms starting before age 45 but not for those with later onset of symptoms.2 This would suggest that the early-onset phenotype identifies a subset of RLS patients for whom symptoms gradually gets worse with age, whereas the INTRODUCTION TWO BIOLOGICALLY SIGNIFICANT FACTORS ABOUT THE NEUROLOGIC DISORDER OF RESTLESS LEGS SYNDROME (RLS) NOW APPEAR TO BE WELL ESTABLISHED. First, abnormalities in iron metabolism are central to the pathology of RLS.1 Second, age of onset of the disorder determines genetic susceptibility.2 It seems likely that the genetic susceptibility reflects expression of the primary biologic factors, and, thus, age of onset may relate to the degree of abnormality in iron metabolism. Finding such a relationship would support the primary role of the iron pathology in RLS and could also open new approaches for identifying patients with RLS and evaluating the genetics of RLS. The primary role of iron metabolism has been documented by studies demonstrating low brain iron concentrations for RLS patients. Both magnetic resonance imaging quantification of brain iron and autopsy studies have shown a reduced iron concentration Disclosure Statement This was not an industry supported study. Dr. Allen has received research support from GlaxoSmithKline; and has received honoraria and consultant payments from GlaxoSmithKline, Pfizer, Boehringer Ingelheim, Sepracor, Norvatis, Orion Pharma, and Schwarz Pharma. Drs. Earley, Connor, Beard, and Clardy have indicated no financial conflicts of interest. Submitted for publication March 2005 Accepted for publication May 2005 Address correspondence to: Christopher J. Earley, Johns Hopkins Bayview Medical Center, 5501 Hopkins Bayview Circle, RM 1B-82, Baltimore, MD 21224; Tel: 410 550 1044; Fax: 410 550 3364; E-mail address: cearley@jhmi. edu SLEEP, Vol. 28, No. 9, 2005 1069 CSF ferritin and RLS—Earley et al late-onset phenotype does not show a significant progression of symptoms with age. The age at which symptoms start may thus define 2 pathologically distinct RLS populations. A previous study failed to find age-of-onset differences in CSF ferritin or transferrin levels, but the samples size was small.5 Additionally, these samples were taken in the morning when RLS symptoms are abated rather than during the night when RLS symptoms are most prominent. Therefore this study was designed to examine nighttime serum and CSF iron variables in relation to disease severity and to determine whether the age at which symptoms started has an effect on those variables. METHODS This was an Institutional Review Board-approved protocol. To be el (...truncated)