Magnesium citrate monotherapy improves restless legs syndrome symptoms and multiple suggested immobilization test scores in an open-label pilot study

Abstract

Study Objectives:

An estimated 3% of the population has clinically significant restless legs syndrome. Given the limited pharmacological options in the arsenal, there is a need for a therapeutic agent with a better side effect profile.

Methods:

Twelve treatment naive adults (10 women and 2 men with a median age of 41.5 [32–48.5] years) with primary restless legs syndrome were recruited in our open-label pilot study; magnesium citrate 200 mg was administered daily for 8 weeks. Serum magnesium levels, International Restless Legs Syndrome Study Group Rating Scale, Kohnen quality of life scale, and multiple suggested immobilization tests (three 1-hour tests) were performed before and after supplementation. Paired t tests and Wilcoxon signed-rank tests were used for data analysis. Pearson and Spearman’s analyses assessed the association between magnesium levels and restless legs syndrome variables.

Results:

Participants had a significant reduction in International Restless Legs Syndrome Study Group Rating Scale scores (6.67 [2.33–11] P = .006) and improved Kohnen quality of life scores (8.5 [2.09–14], P = .014) without notable differences in serum magnesium levels (P = .3). The median periodic limb movements during wakefulness index (30.40 [5.20, 122.40] to 8.63 [0.32, 17.47] P = .043) and self-reported discomfort score (19 [14, 30.5] to 6 [0, 8] P = .0010) of all 3 multiple suggested immobilization test trials also demonstrated improvement. Serum magnesium levels negatively correlated with multiple suggested immobilization test self-reported scores and the periodic limb movements during wakefulness indices.

Conclusions:

Despite the limitations of open-label design, our study’s positive results indicate the need for a placebo-controlled trial with a larger sample size.

Clinical Trial Registration:

Registry: ClinicalTrials.gov; Name: The Effect of Magnesium Citrate Supplementation in Restless Legs Syndrome (RLS); URL: https://clinicaltrials.gov/ct2/show/study/NCT04462796; Identifier: NCT04462796.

Citation:

Gorantla S, Ravisankar A, Trotti LM. Magnesium citrate monotherapy improves restless legs syndrome symptoms and multiple suggested immobilization test scores in an open-label pilot study. J Clin Sleep Med. 2024;20(8):1357–1361.

BRIEF SUMMARY

Current Knowledge/Study Rationale: There is a need for adjunct therapeutic agents to treat chronic persistent restless legs syndrome with a better side effect profile. The evidence for the efficacy of magnesium supplementation in restless legs syndrome is inadequate.

Study Impact: Our pilot study, despite the limitations of the study design, provided positive results and set the stage for further investigation. Magnesium supplementation may be a viable therapeutic option, either as an adjunct treatment or standalone therapy, pending future randomized controlled trials.

INTRODUCTION

Restless legs syndrome (RLS) is a common sleep-related movement disorder that negatively affects sleep and quality of life. Current treatment options for chronic persistent RLS (gabapentinoids, dopamine agonists, and opioids) all have potentially serious side effects. Patients with RLS receiving pregabalin and gabapentin enacarbil experience significantly increased incidence of daytime sleepiness (odds ratio 4.67, 95% confidence interval 1.23–17.66; odds ratio 3.56, 95% confidence interval 1.77–7.14, respectively).¹ Dopamine agonists have limited utility due to the risk of augmentation.² Opiates are not indicated as first-line medications but are useful in refractory RLS.³ Although iron supplementation may reduce the severity of symptoms, it is often insufficient as monotherapy.⁴

Over the last decade, none of the investigated supplements gathered adequate evidence for routine use in clinical practice, except for iron in patients with RLS and low iron stores. In this context, it is critical to investigate supplemental magnesium’s therapeutic effect as it is a commonly used and widely available supplement. A survey of 1,544 Dutch participants showed that 7% of the general population and 25% of athletes take it regularly for general health and sports performance enhancement.⁵ Magnesium supplements are typically well tolerated, and adverse effects are rarely seen at doses less than 350 mg of daily elemental magnesium.⁶

The studies that investigated the effectiveness of magnesium in RLS yielded mixed results.⁷ A handful of noncontrolled studies (3 case series and 4 case reports) have shown RLS symptom alleviation with magnesium supplementation using only self-report measures.⁷ A small placebo-controlled study conducted in 2005 (364.8 mg elemental magnesium per day for 28 days, with 15 participants in each arm) did not show significant improvement in RLS symptoms or polysomnographically measured periodic limb movements of sleep.⁸ In 2023, Jadidi et al published a single-blinded placebo-controlled study with 75 patients with RLS, 25 of whom received 250 mg of magnesium oxide for 2 months.⁹ The magnesium group showed improvement in RLS and sleep self-reported metrics: Pittsburgh Sleep Quality Index (18.28 [3.080] to 5.92 [2.59], P = .001) and International Restless Legs Syndrome Study Group Rating Scale (IRLS) (30.96 [7.71] to 16 [5.88], P = .001) scales.⁹ However, this study did not include an assessment of periodic limb movements of sleep nor an objective measure of motor symptom burden, such as the multiple suggested immobilization test (m-SIT).

One key tool for objectively and subjectively assessing RLS severity is the SIT. Since its development in 1988, the SIT has evolved over the years.^10–13 The initial SIT was a single 60-minute immobilization challenge administered around 9:00 pm.¹³ Given the propensity of RLS symptoms to fluctuate, the m-SIT was developed to capture the variations in symptom onset and course in the evening with three 1-hour tests separated by an interval of 1 hour between the tests.¹² The m-SIT is a validated test that assesses at least 3 diagnostic criteria of RLS (the urge to move, the provocation of symptoms at rest, and the increase in severity in the evening) with self-reported scales.

Given the side effect burden and limitations of existing pharmacologic management of chronic persistent RLS and the favorable side effect profile of magnesium combined with compelling yet inadequate existing data, we conducted a prospective pilot study to assess the impact of magnesium citrate administration on RLS symptoms and objectively measured motor activity on m-SIT. The m-SIT findings add robustness to the study methodology.

METHODS

Adults diagnosed with RLS by a Sleep Neurologist according to the International Classification of Sleep Disorders, third edition diagnostic criteria were recruited from OSF HealthCare neurology and sleep clinics. Participants included in this study were: (1) treatment naive, (2) on stable doses of medications that negatively impact RLS and limb movements (eg, antidepressants) during the study period, (3) not on gabapentinoids, opioids, or dopamine agonists. Exclusion criteria included: (1) current use of magnesium supplements, (2) history of chronic kidney disease, (3) untreated obstructive sleep apnea, (4) diarrhea and gastrointestinal disorders, and (5) current pregnancy. Participants with iron deficiency were offered iron supplementation; they could choose to take iron before or after participation in this study but could not take both simultaneously, as magnesium interferes with iron absorption. We registered our study on clinicaltrials.gov (NCT04462796). We obtained Institutional Review Board approval at OSF HealthCare and the University of Illinois College of Medicine at Peoria. All participants provided informed consent.

Participants were instructed to take magnesium citrate 200 mg (Puritan’s Pride, Oakdale, NY) with food at dinner time daily for 8 weeks. The tolerable upper limit of elemental magnesium supplementation for adults (both males and females) is 350 mg.⁶ We used 200 mg of elemental magnesium as magnesium citrate because of its superior bioavailability compared to magnesium oxide.^14–17 Historically, more than 2,000 mg of magnesium citrate in a single dose has been tried to relieve occasional constipation.

Before starting magnesium and again at the end of the 8-week treatment period, participants completed questionnaires, m-SIT, and a blood draw for serum magnesium levels. The questionnaires were the IRLS¹⁸ (mild [score 1–10], moderate [score 11–20], severe [score 21–30], very severe [score 31–40]) and Kohnen quality of life scale¹⁹ (0–60, higher scores are associated with more severe symptom burden).

The m-SIT was performed as 3 trials, each lasting 1 hour with a gap of 1 hour between trials. The participants rested in a bed reclined at a 45-degree elevation with their legs extended at 170 degrees while limb movements were monitored with electromyography leads on both tibialis anterior muscles. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications, version 3 limb movement criteria were used to score and quantify limb movements during wakefulness.²⁰ The average periodic limb movements during wakefulness (PLMW) index was calculated to reflect the objective limb movements that span the entire evening. A conventional polysomnogram electroencephalogram montage was used to monitor wakefulness, and RLS symptom severity was monitored every 10 minutes. Discomfort score (DS) ranges from 0 to 10 (maximum score per hour is 60), verbally communicated to the sleep lab technician by the participant every 10 minutes during the test. All the study participants completed the 1-hour trials without interruptions. The test’s time frame (test start time varied from 5:00–7:00 pm and ended at 10:00 pm to midnight) was modified according to the participant’s availability and convenience, and the same time windows were used before and after supplementation for each participant. All measures were reassessed on day 56 (8 weeks).

In the landmark validation study using m-SIT to assess the severity of RLS symptoms, the mean PLMW index differed significantly across groups [controls (n = 10) had a mean PLMW index of 6.12 ± 11.19 events/h, patients with RLS (n = 19) on medication demonstrated a markedly higher mean PLMW index of 29.94 ± 24.1 events/h (P = .0012) and patients with RLS who were off medications for 3 days displayed a substantially elevated mean PLMW index of 67.62 ± 52.5 events/h (P < .0001)]. Using PLMW index > 10 events/h as a cut-off for detecting RLS (while off medication when compared to controls), 3-test m-SIT at 4:00–8:00 pm and 6:00–10:00 pm has an area under the curve of 0.93 and 0.90, respectively, with similar sensitivity (79%) and specificity (90%).¹² As expected, all healthy controls scored zero at all tested times on the m-SIT self-reported discomfort scales in the validation study. The self-reported DS carried better discriminatory value; 100% sensitivity and specificity were noted for tests conducted at 6:00 pm + 8:00 pm + 10:00 pm and 4:00 pm + 6:00 pm + 8:00 pm + 10:00 pm using a cut-off score 2.¹²

The Shapiro-Wilk test was used to determine the data distribution, and the paired t test or Wilcoxon signed-rank test was used to compare the outcome measure before and after magnesium supplementation. Pearson and Spearman’s analyses assessed the correlation between magnesium, self-reported RLS scales, and m-SIT scores.

RESULTS

Thirteen participants enrolled in the study, but 1 participant withdrew after completing baseline scales because of time commitment constraints. Of the remaining 12 participants, 41% (n = 5; IRLS scores were 25, 25, 23, 23, 21) had severe, and 59% (n = 7; IRLS scores were 19, 17, 17, 17, 16, 13, 10) had moderate RLS. Four participants had % iron saturation less than 25%, and 2 participants’ ferritin level was less than 75 mg/dl. None of the participants were taking iron supplements during the study. Half of the cohort had obstructive sleep apnea and were adherent to PAP therapy. Eight out of 12 completed both m-SITs. All participants took magnesium throughout the full 8-week period (mean adherence to magnesium was 94%). Although participants were instructed not to start new medications during the study, 1 participant had to start atomoxetine in week 8. The remaining cohort characteristics are described in Table 1.

Table 1.

Sample characteristics.

Variable
Age, years, mean ± standard deviation	42.4 ± 14.2
Sex, % female	83
Serum magnesium level at baseline, mg/dl, median [IQR]	1.85 [1.8–2.1]
IRLS score, median [IQR]	18 [16.7–23]
Kohnen QOL score, median [IQR]	19.5 [16.2–22.2]
Disease duration, years, median [IQR]	5 [2–16.2]
Serum ferritin (n = 12), ng/ml, median [IQR]	85 [58–102]
% Iron saturation (n = 11), median [IQR]	26 [21.5–31.7]

Normal serum magnesium level is 1.7–2.2 mg/dl. IQR = interquartile range, IRLS = International Restless Legs Syndrome Study Group Rating Scale, QOL = quality of life.

As described in Table 2 and Table 3, IRLS (6.67 [2.33–11], P = .006), Kohnen quality of life (8.5 [2.09–14.91], P = .014), and m-SIT scores showed significant improvement. All self-reported m-SIT DS showed significant improvement: m-SIT trial 1 DS (20 [17, 32] to 7 [0.25, 12.5], P = .028), m-SIT trail 2 DS (26 [15, 39] to 5.5 [0.75, 8.25], P = .012), and m-SIT trial 3 DS (21 [14, 35] to 6 [0, 8], P = .043). The median m-SIT DS of the three trials also improved substantially with magnesium citrate supplementation (19 [14, 30.5] to 6 [0, 8] P = .0010). PLMW index in the middle trial (19.1 [6.8, 93.9] to 1.95 [0, 8.13], P = .028) and average PLMW index (30.40 [5.20, 122.40] to 8.63 [0.32, 17.47], P = .043) showed statistically significant improvement. The PLMW indices from trial 1 and trial 3 showed an improvement trend. Serum magnesium levels did not change significantly. One participant with RLS had no PLMs during wakefulness on m-SIT, both at baseline and the follow-up.

Table 2.

Comparison of IRLS, Kohnen QOL, serum magnesium level before and after magnesium citrate supplementation (paired t test).

Outcome Measure	Mean ± SD		Mean Difference	95% CI of Mean Difference		P
	Pre	Post		Lower	Upper
IRLS (n = 12)	18.83 ± 4.73	12.17 ± 4.59	6.67	2.33	11.00	.006*
QOL (n = 12)	19 ± 7.24	10.5 ± 7.04	8.50	2.09	14.91	.014*
Mg (n = 11)	1.91 ± 0.23	1.95 ± 0.16	0.05	0.05	0.14	.320

^*= P value < .05. CI = confidence interval, IRLS = International Restless Legs Syndrome Study Group Rating Scale, Mg = magnesium, QOL = quality of life, SD = standard deviation.

Table 3.

Comparison of medians of multiple suggested immobilization test self-reported and objective measures (PLMW) before and after magnesium citrate supplementation.

Outcome Measure	Median (IQR)		P (Wilcoxon Signed-Rank Test)
	Pre	Post
DS trial 1 (n = 8)	20 (17, 32)	7 (0.25, 12.5)	.028*
DS trial 2 (n = 8)	26 (15, 39)	5.5 (0.75, 8.25)	.012*
DS trial 3 (n = 8)	21 (14, 35)	6 (0, 8)	.043*
Median DS of 3 trials	19 (14, 30.5)	6 (0, 8)	.001*
PLMW trial 1 (n = 8)	8.9 (0, 85.9)	0 (0, 15.9)	.465
PLMW trial 2 (n = 8)	19.1 (6.8, 93.9)	1.95 (0, 8.13)	.028*
PLMW trial 3 (n = 8)	33.6 (9.2, 190.6)	11.9 (0, 50.6)	.063
Median PLMW of 3 trials	30.40 (5.20, 122.40)	8.63 (0.32, 17.47)	.043*

^*= P value < .05. DS = discomfort score, IQR = interquartile range, PLMW = periodic limb movements during wakefulness (events/h).

Table 4 illustrates a high negative correlation between serum magnesium level and self-reported DS of m-SIT trials 1 and 2 (P < .05) at baseline. There was a statistically insignificant negative correlation between serum magnesium levels with quality of life and m-SIT trial 3. PLMW index from trial 2 correlated negatively with serum magnesium levels; the remaining indices showed insignificant but negative correlation. We performed a subgroup analysis to assess the impact of magnesium citrate supplementation on participants with iron deficiency (low ferritin and/or low % iron saturation). We observed similar symptom improvement trends to those of the entire group, with statistical distinction limited by low sample size (n = 5).

Table 4.

Correlation between serum magnesium level with IRLS, QOL, and m-SIT trial metrics at baseline.

Variables	Correlation Coefficient (95% CI)	P
Mg level and IRLS	.133 (−.477 to .657)	.680
Mg level and QOL	−.195 (−.691 to .427)	.544
Mg level and DS trial 1	−.720 (−.922 to −.211)	.012*
Mg level and DS trial 2	−.711 (−.937 to −.319)	.005*
Mg level and DS trial 3	−.567 (−.871 to .050)	.069
Mg level PLMW trial 1	−.246 (−.747 to .432)	.465
Mg level PLMW trial 2	−.613 (−.891 to 0)	.045*
Mg level PLMW trial 3	−.259 (−.753 to .420)	.441

^*= P value < .05. CI = confidence interval, DS = discomfort score, IRLS = International Restless Legs Syndrome Study Group Rating Scale, Mg = magnesium, PLMW = periodic limb movements during wakefulness, QOL = quality of life.

One participant reported a change in stool consistency without a change in frequency; another one experienced diarrhea when 400 mg of magnesium citrate was accidentally taken for 2 days.

DISCUSSION

Our open-label study demonstrated improvement in self-reported and objective RLS measures with 200 mg magnesium citrate supplementation for 8 weeks. The effect size, as measured by IRLS change, is smaller when compared with a recent study (−6 vs −14)⁹ and an average reduction in IRLS of 6.58 points has been noted with placebos in the systematic review and meta analysis of RLS trials.²¹ In the Jadidi et al study, the placebo group (Pittsburgh Sleep Quality Index from 16.6 [2.9] to 10.5 [3.5], P = .001, and IRLS scales from 31 [4.2] to 24 [8], P = .001).⁹ Thus, we cannot exclude a placebo effect on our self-reported measures. However, we also observed a substantial reduction in the median PLMW index during the m-SIT (pre 30.4 [5.2, 122.4] to post 8.63 [0.32,17.4], P = .043). Objectively measured PLMs are much less susceptible to the placebo effect in patients with RLS.²¹ The gradual increase in PLMW index in the evening (8.9→19.1→33.6) supports the expected circadian variation of RLS. However, the self-reported DS scores did not show this pattern. Allen et al demonstrated the same phenomenon of PLMW rise in a study with 28 patients with RLS.²² These findings suggest that the placebo effect is unlikely to fully explain our observed beneficial effects of magnesium citrate supplementation.

Patients with RLS often use magnesium despite the lack of robust evidence for efficacy. According to the National Health and Nutrition Examination Survey that analyzed the data from 5,021 adults, 68% consumed less than the recommended daily allowance of 310–420 mg per day and 19% less than 50% of recommended daily allowance.²³ Another larger study from 2008 (n = 18,177) reiterated similar findings with an estimated 60% of the United States adults with less than required dietary magnesium.²⁴ However, magnesium deficiency has not been consistently demonstrated in patients with RLS.²⁵ The mechanism by which magnesium may improve RLS is currently unknown. Magnesium supplementation may exert efficacy on RLS even in the absence of serum deficiency by reducing excitability within the CNS.²⁶ Magnesium, an essential cofactor in more than 300 enzymes, competes with calcium on N-methyl-d-aspartate receptors and inhibits excitatory glutaminergic neurotransmission in the central nervous system.^26,27 N-methyl-d-aspartate receptor has the specific binding site for magnesium at the channel pore, and it blocks the ion flow at resting membrane potential.²⁶ Magnesium displacement at the binding site facilitates glutaminergic neurotransmission and excitotoxicity.²⁶ Excessive glutaminergic neurotransmission has been implicated in the pathophysiology of RLS.²⁸

The significant negative correlation of serum magnesium levels with some of the m-SIT scores is an interesting finding, aligning with our hypothesis. All variables, except for IRLS, showed a negative correlation. In this context, it is important to note that less than 1% of magnesium is distributed in extracellular fluids, and the majority is stored in bone (53%), muscle (27%), and soft tissue (19%),²⁹ such that blood testing does not capture the status of total body stores. The magnesium loading test, collection of 24-hour urine excretion followed by an oral or parenteral magnesium load, provides a better estimation of magnesium status, but it is impractical.³⁰

The body tightly regulates serum magnesium levels due to its direct influence on neuromuscular activity and bone formation.³⁰ This regulatory mechanism likely explains why serum magnesium levels did not elevate despite 8 weeks of supplementation in our study. Similarly, the absence of substantial circadian variation in serum magnesium levels can be attributed to this phenomenon. Møller et al’s investigation into the diurnal variation of serum and urine magnesium levels in healthy adults (n = 10), conducted every 3 hours for 24 hours, did not reveal clinically significant fluctuations.³¹ However, intracellular magnesium concentrations fluctuate in response to changes in energy and metabolic demands driven by circadian rhythms.³²

Apart from the open-label design, a few other limitations exist in this study. The sample size was small, and the m-SIT was completed by only 66% (8 out of 12 participants). The 3-test m-SIT necessitates a 6-hour time commitment, leading to 1 individual dropping out solely due to time constraints. Discomfort induced by RLS symptoms during the initial immobilization test prompted withdrawal from follow-up testing for the remaining 3 participants.

CONCLUSIONS

Magnesium citrate 200 mg supplementation appears to have a beneficial effect on self-reported and objective RLS symptoms, and this dosage was tolerated well without major side effects. A larger placebo-controlled study is warranted.

DISCLOSURE STATEMENT

All authors have read and approved the manuscript. This work was performed at OSF HealthCare Illinois Neurological Institute. OSF HealthCare Illinois Neurological Institute funded this study. The manufacturer of the magnesium supplement used in this study was not involved in any portion of this study. Dr. Trotti is a member of the Board of Directors of the American Academy of Sleep Medicine; any opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the American Academy of Sleep Medicine. The authors report no conflicts of interest.

ABBREVIATIONS

IRLS

International Restless Legs Syndrome Study Group Rating Scale

m-SIT

multiple suggested immobilization test

PLMW

periodic limb movements during wakefulness

RLS

restless legs syndrome