The Effect of Parenteral or Oral Iron Supplementation on Fatigue, Sleep, Quality of Life and Restless Legs Syndrome in Iron-Deficient Blood Donors: A Secondary Analysis of the IronWoMan RCT

Background: Besides anemia, iron deficiency may cause more subtle symptoms, including the restless legs syndrome (RLS), the chronic fatigue syndrome (CFS) or sleeping disorders. Objective: The aim of this pre-planned secondary analysis of the IronWoMan randomized controlled trial (RCT) was to compare the frequency and severity of symptoms associated with iron deficiency before and after (intravenous or oral) iron supplementation in iron deficient blood donors. Methods/Design: Prospective, randomized, controlled, single-centre trial. (ClinicalTrials.gov: NCT01787526). Setting: Tertiary care center in Graz, Austria. Participants: 176 (138 female and 38 male) whole-blood and platelet apheresis donors aged ≥ 18 and ≤ 65 years with iron deficiency (ferritin ≤ 30ng/mL at the time of blood donation). Interventions: Intravenous iron (1 g ferric carboxymaltose, n = 86) or oral iron supplementation (10 g iron fumarate, 100 capsules, n = 90). Measurements: Clinical symptoms were evaluated by a survey before iron therapy (visit 0, V0) and after 8–12 weeks (visit 1, V1), including questions about symptoms of restless legs syndrome (RLS), chronic fatigue syndrome (CFS), sleeping disorders, quality of life and symptoms like headaches, dyspnoea, dizziness, palpitations, pica and trophic changes in fingernails or hair. Results: We found a significant improvement in the severity of symptoms for RLS, fatigue and sleep quality (p < 0.001). Furthermore, a significant decrease in headaches, dyspnoea, dizziness and palpitations was reported (p < 0.05). There was no difference between the type of iron supplementation (intravenous versus oral) and clinical outcome data. Conclusion: Iron supplementation in iron-deficient blood donors may be an effective strategy to improve symptoms related to iron deficiency and the wellbeing of blood donors.


Introduction
An imbalance between iron requirements and iron supply results in iron deficiency (ID), the most prevalent nutritive disorder worldwide 1 . It is estimated that two billion people are affected 2 and iron deficiency is the main cause for anemia besides other nutritional causes including vitamin B12 and folate deficiency 3 . Besides anemia, iron deficiency may have other, more subtle adverse effects on an individual such as a restless legs syndrome (RLS) [4][5][6] , one of the most prevalent neurologic diseases with different symptoms including an urge to move the legs, sometimes accompanied with unpleasant sensations, worsening at rest, and improving with movement 7 , a chronic fatigue syndrome (CFS) [8][9][10][11] or decreased physical and cognitive performance, especially in children and premenopausal women [12][13][14] . Furthermore, iron deficiency is associated with sleeping disorders 6 , trophic changes of fingernails 15 and hair 16 , symptoms of pica 17,18 and obstetric complications including low birth weight, preterm delivery and stillbirth 19 .
As the largest part of total body iron is bound to hemoglobin in red cells 3 , each whole blood donation (about 500 ml) results in an iron loss of 200 to 300 mg 20 and each apheresis donation (plasma, platelets) in a loss of 20 to 25 mg 21 . Given the maximum number of donations per year (e.g. in Austria 4-6 whole blood donations, up to 26 platelet or 45 plasma donations), blood donations may substantially compromise the iron store of individual donors who are indeed often iron deficient 17, 18, 22-31 . In the IronWoMan trial, we randomized iron deficient blood donors to a single high dose of intravenous ferric carboxymaltose or to oral iron for 10 weeks 32 . In this pre-planned secondary analysis, we aimed to assess the prevalence and severity of different clinical symptoms related to iron deficiency before and after iron therapy with intravenous or oral iron.

Materials and Methods
The study was carried out according to GCP guidelines and conformed to the Declaration of Helsinki. The protocol was approved by the Austrian Agency for Health and Food Safety (AGES) and by the Ethical Committee of the Medical University of Graz. It was registered at the European Clinical Trials Database (EudraCT no. 2013-000327-14) and ClinicalTrials.gov (NCT01787526).

Donor characteristics:
The study population consisted of healthy male and female whole blood and platelet apheresis donors aged ≥18 and ≤65 years, fulfilling the Austrian criteria for blood donation, with iron deficiency (ferritin level ≤30 ng/ml) but without anemia at the time of blood donation. Exclusion criteria were hemochromatosis, acute infection, pregnancy or lactation, a history of anaphylaxis to intravenous iron or other substances, and signs or symptoms suggestive for acute or chronic gastrointestinal or excessive gynecological bleeding.

Study design:
The clinical trial was a prospective, randomized, controlled single center study. Recruitment started in June 2014 and was completed in June 2016. The detailed study protocol was previously published 33 . In brief, ferritin screening was done from remaining serum for routine infectious disease testing in donors who had given general informed consent to participate in clinical studies. Donors with a serum ferritin ≤30 ng/ml were invited by telephone to participate in the study in the apheresis unit of the Department of Blood Group Serology and Transfusion Medicine. Before the first visit, a participant code was generated and a randomization for either intravenous (IV) or oral iron was performed.
During the first visit (visit 0, V0), written informed consent was obtained, inclusion/exclusion criteria were checked and a pregnancy test in premenopausal women was performed.
A case report form was filled out by the study participants including questions about: The questions were developed based on published criteria for RLS 34 , fatigue 35 , and sleeping behaviour 36,37 and also included questions on symptoms known to be associated with iron deficiency such as pica.
The study participants received either a single dose of IV ferric carboxymaltose (1 g in a short infusion, Ferinject ® , manufacturer Vifor Pharma, Austria) or 100 capsules of oral ferrous fumarate in combination with ascorbic acid given over 8-12 weeks (each capsule containing 100 mg of bivalent iron and 20 mg ascorbic acid, Ferretab ® capsules, manufacturer G.L. Pharma GmbH, Austria; total dose 10 g, assuming a 10% resorption rate).
At visit 1 (V1) after 8 to 12 weeks, the participants completed a second questionnaire containing additional questions about adverse events and compliance with study medication. After V1, all biochemical results were summarized in a written report for each study participant and were sent by mail (including recommendations in case of specific findings).
A schematic study flowchart is given in Figure 1.

Restless legs syndrome:
RLS was assumed when the three specific RLS questions of the questionnaire were answered with "Yes" by the study participant. Adapted from the essential diagnostic criteria by Allen et al 34 , we asked 1) for uncomfortable sensations in the legs accompanied by an urge to move the legs, 2) if the urge to move or unpleasant sensations begin or worsen during periods of rest and relieve by movement and 3) if the urge to move or the unpleasant sensations are worse in the evening or night or during the day. Additionally we asked how often the symptoms occurred.

Quality of sleep:
To Further different questions (i.e. hair loss, brittle nails, headaches) were asked with the possibility to answer with "Yes" or "No".
The study participants subjectively indicated their quality of life with 5 differences on a Likert type style scale ranging in five steps between "very poor" and "very good".

Statistical analysis:
This analysis followed the intention-to-treat principle. Descriptive statistics of the data are presented as median and quartile or absolute and relative frequencies, depending on the type of data. Differences between V0 and V1 are analyzed either by the Wilcoxon signed-rank test or by the Mc Nemar´s test. The significance level was set to alpha=0.05. SPSS statistical software (version 25.0; IBM Corp, Armonk, NY, USA) was used to perform the statistical data analysis. Since the questionnaires used in this study were translated by the study team, their structure was examined before further use. Usability of each questionnaire was evaluated by factorial structure (confirmatory factor analysis) and fit indices. Only questionnaires which met the criteria were analyzed. These analyses were performed with R software (version 3.5.0) using the package "Multidimensional Item Response Theory, mirt" 38 .

Results
Between June 2014 and June 2016, a total of 467 blood donors (370 women, 97 men) with a serum ferritin ≤30 ng/ml were invited to participate in the study. 291 declined to participate and 176 were included (138 women, 78%, 38 men, 22%). Total study duration for each participant was 8 to 12 weeks. In August 2016, the last participant completed the study. Demographic data of the study participants are presented in Table 1.

Donor Characteristics
Age, mean (SD), years 36 (13) 46 (13) Weight, mean (SD), kg 69 (11) 82 (16) Body mass index, mean (SD) 24 (4) 26 (4) Total calculated blood volume, mean The primary hypothesis of the study was that intravenous iron is superior to oral iron in improving iron stores. The primary endpoint of the study was the difference of the transferrin saturation after iron therapy between the two treatment groups. All the biochemical data have been published previously 32  Comparing the frequency of symptoms of the 18 study participants plus the 13 participants with RLS only at V0, a significant improvement of RLS symptoms between baseline and after IV or oral iron was determined (p<0.001). Results are shown in Figure 2.

Figure 2.
Bar diagram showing the number of study participants (x-axis) who indicate the frequency of their restless legs syndrome (RLS) at the two time points before (V0, dark bars) and after IV or oral iron therapy (V1, bright bars) on the y-axis. The decrease of the RLS occurrence between V0 and V1 was statistically significant (p<0.001, n=31).

Fatigue
Each of the five scales (general, physical, emotional, mental, vigor) and the global evaluation for fatigue improved significantly after iron therapy (p<0.001). Results are shown in Figure 3.   before (V0, dark bars) and after iron therapy (V1, bright bars). For all 4 quality of sleep items there was a statistically significant improvement between V0 and V1 (p<0.05 for all, n=160).

Different symptoms
The presence of different symptoms possibly related to iron deficiency at the two visits (ie. brittle nails, hair loss, pica etc.) is shown in Figure 5. between very poor and very good (y-axis) before (V0, dark bars) and after iron therapy (V1, bright bars). (no statstically significant differences, n=168).

Further results -type of iron treatment and ferritin levels at visit 1
There was no significant difference between IV and oral iron therapy for all reported outcomes (data not shown). There was also no significant difference between subgroups with lower or higher ferritin levels at visit 1 (data not shown).

Discussion
Besides anemia, iron deficiency may have more subtle adverse effects. In this pre-planned secondary analysis of the IronWoMan RCT 32 testing high-dose IV iron vs. a standard oral regime in iron deficient blood donors, we aimed to evaluate if symptoms known to be associated with iron deficiency improved after iron supplementation.

Restless legs syndrome (RLS)
RLS is one of the most prevalent neurologic diseases 39 . An association between RLS and iron deficiency has been proposed and iron treatment is regularly prescribed in RLS. Magnetic resonance imaging studies revealed a deficiency of iron in different areas of the brain 40 and this was confirmed by direct neuro-pathologic demonstration of low levels of iron in the substantia nigra of affected patients 41 . A relation between low serum ferritin and symptoms of RLS 5,6 and an improvement of RLS symptoms with iron treatment has also been demonstrated 4,42 . A recently published review concluded that of all known parameters associated with RLS including dopamine, glutamate, serotonin and ferritin levels, the most consistent finding appears to be iron deficiency 7 . Thus, iron replacement is suggested for RLS patients with a ferritin level <75 ng/ml 43,44 and according to recent clinical practice guidelines intravenous iron is even effective for treatment of moderate to severe RLS in patients with ferritin levels <300 ng/ml 45 .
In our study population of relatively young and healthy but iron deficient blood donors, the frequency of RLS at V0 was 21%, which is higher than the reported prevalence of RLS between 5 and 15% in the general population [46][47][48] . This high prevalence in blood donors is comparable to the numbers reported by others (Birgegard et al: 18%, Ulfberg et al: 18%, Burchell et al: 17%) 30,49,50 . We also compared the severity of self-reported RLS symptoms at baseline (V0) and after iron therapy and found a significant improvement after iron supplementation (p=0.001). These results are consistent with the findings of a double-blind, placebo-controlled study by Wang et al. where a statistically significant improvement in RLS patients with low serum ferritin levels (15-75 ng/ml) after oral iron treatment was demonstrated 42 .

Fatigue
All domains of reported fatigue symptoms in our study indicated a significant improvement after iron treatment (p<0.001). In contrast to our findings, two Swiss randomized controlled trials in blood donors did not show a clinical benefit by either oral or intravenous iron supplementation. However, Waldvogel et al. supplemented 80 mg oral iron per day for only 4 weeks and Fontana et al. gave 800 mg intravenous iron once but no detailed results have yet been published 51,52 . In line with our results, a recent review including 18 trials enrolling 1170 non-anemic iron deficient adults found that iron supplementation led to a reduction of fatigue 10 . Furthermore a meta-analysis reported that iron may be effective in reducing fatigue in iron deficient patients without anemia 53 . Sleep disorders often arise as a consequence of or in combination with RLS 5,6,39,49 . In our study, we evaluated the quality of sleep with the Jenkins Sleep Evaluation Questionnaire and classified the answers with the Jenkins Scale 37 . All 4 items: difficulty falling asleep, waking up several times per night, difficulty staying asleep, and waking up feeling tired and worn out after usual amount of sleep showed a clear and significant improvement after iron therapy. Our results support previous findings that in iron deficient individuals with sleep disorders, iron therapy improves symptoms 5, 6 .

Other symptoms
Unspecific symptoms including pica, brittle nails, hair loss and a plain tongue were reported more often before than after iron therapy (statistically nonsignificant). In the literature, the association of iron deficiency with these symptoms also remains controversial 15,16,54,55 .
Pica is not considered specific for iron deficiency 56,57 . However, pica for ice responds rapidly to treatment with iron 57 . Bryant et al. found a significantly higher percentage of pica among blood donors with iron deficiency (11%) compared to iron sufficient donors (4%). Furthermore, pica symptoms resolved completely with oral iron supplements 18 . In our analysis we found a nonsignificant improvement after iron supplementation but only 6% of participants reported pica before and 3% after iron therapy.
Headache, dyspnea and dizziness improved significantly in our study. However, this may also be explained with the high percentage of anemic blood donors after the last blood donation and before iron therapy (73% of women and 82% of men) with a mean Hb of 11.7 g/dl and with the higher mean Hb of 13.6 g/dl after iron supplementation 32 . Interestingly, there was no difference between IV and oral iron therapy and the attained ferritin concentration at visit 1, although our study was certainly underpowered for such analyses.

Strengths and limitations
Strengths of our study are the selective inclusion of iron deficient blood donors, the excellent compliance and the low number of participants lost to follow up.
However, our study has some limitations: first, it was performed in a single European academic institution. Second, the follow up was limited to 3 months. Third, questionnaires are subjective and fourth, and most importantly, we did not include a placebo group. Therefore, we were unable to assess the natural course of current clinical practice (not using any iron supplements) and the influence of the hemoglobin increase per se. However, the very high incidence of iatrogenic anemia at baseline (75%) and a recent RCT 58 suggest that the recovery to pre-donation hemoglobin concentrations and iron stores typically takes substantially longer than the whole blood inter-donation interval.

Conclusions
Our findings support the hypothesis that iron deficiency can cause clinically relevant subjective symptoms in otherwise healthy, iron deficient blood donors and that these symptoms may be improved by iron treatment. Our study was relatively small for the evaluation of clinical endpoints. The fact that we were able to see a significant improvement in several of the analyzed clinical symptoms suggests that benefits outweigh potential risks of iron treatment and is of clinical relevance. Interestingly, the approach of iron substitution (oral or IV) was not crucial for the improvement of symptoms. Therefore, blood donation services should identify donors at risk for iatrogenic iron deficiency and offer iron supplementation accompanied by the extension of the donation interval to maintain the health of each blood donor, especially in premenopausal women, the highest risk group.