The RELAX Randomized Controlled Trial: Stretching versus Meditation for Nocturnal Muscle Cramps

Abstract

Background:

Muscle cramps are common among persons with cirrhosis and associated with poor health related quality of life (HRQOL). Treatment options are limited. We compared stretching or meditation in a randomized-controlled trial (RCT).

Patients:

We enrolled 98 patients with a history of >4 muscle cramps in the prior month from 7/22–7/23. We randomized patients 1:1 to stretching versus meditation for 35-days. Our primary outcome was the change in cramp severity measured by the Visual Analog Scale for cramps (VAS-cramps, scaled 0–10). Secondary outcomes included Patient global impression of change (PGIC), change in sleep quality, and global HRQOL measured using the EQ-5D and VAS-global HRQOL.

Results:

Overall, 48% of patients had cirrhosis, 40% had diabetes, 16% the median age was 63, most were women (67%), and 81% were college educated. Both arms experienced a reduction in cramp severity – a median of 1.44 (0.58–2.29) points for stretching and 1.97 (1.01–2.93) points for meditation. These changes were significant changes from baseline (p=0.001 for stretching, p<0.0001 for meditation) but these changes were equivalent between arms (p=0.4). The PGIC was improved: 1.33 (1.02–1.65) for stretching, 1.05 (0.70–1.41) for meditation, p-difference 0.2. Sleep was also improved for both. HRQOL did not change according to the Eq5D; according to the VAS, HRQOL rose for meditation by 6 (0.1–11.8) points but not for stretching. More patients recommended stretching than meditation (79.2% vs 55.3%,p=0.02).

Conclusion:

In a randomized trial, stretching and meditation both reduced cramp severity, improved sleep quality and global impression of change. While patients preferred stretching, there was no difference in effect between arms.

Registration:

(NCT05457322)

Lay summary

Muscle cramps are common and distressing, particularly for patients with cirrhosis. Many patients have cramps that do not respond to medical therapies. Further, many medical therapies for cramps are risky or costly. We investigated the role of stretching and meditation to improve cramp severity in a randomized trial. Both groups experience benefits but neither was better than the other. These low-cost safe non-drug treatments should be considered when treating muscle cramps.

Introduction

Muscle cramps are common and independently diminish quality of life.^{1, 2} Cramps cause pain, interfere with sleep, and limit mobility. They afflict 2 in 3 of patients with cirrhosis irrespective of disease severity.² However, cramps also affect 1 in 2 people over the age of 60 and among those receiving dialysis.^{3, 4} Patients with chronic liver disease are aging, have many comorbidities, and thus have many risk factors for muscle cramps. Efforts to improve their quality of life could include cramp reduction interventions.

Options for cramp therapies have important risks and are often derived from small, dated studies. Quinidine may be effective but is potentially toxic,⁵ albumin infusion is possibly effective but expensive and increases the risk of volume overload,⁶ and taurine is possibly effective but expensive and unregulated.⁷ Although magnesium is often prescribed, it is not effective.⁸ Better interventions and studies are needed. We previously showed that, compared to a water control, sips of pickle brine at cramp onset was more likely to abort the muscle cramp and reduced cramp severity.⁹ Sleep quality and health-related quality of life (HRQOL), however, was not improved. We hypothesized that the lack of cramp prevention explained the discrepancy between improved cramp severity and the inability to improve these secondary outcomes. Unfortunately, safe and effective treatments for cramp prevention remain limited.¹⁰ Hallegraeff et al randomized 80 participants aged>55 years to a stretching intervention and found that it reduced cramp severity by 1.3 points on a 10-point visual analog scale (VAS) and cramp frequency by 1.2 cramps/night.¹¹

We therefore conducted a randomized, controlled trial of night-time stretching. We compared stretching to meditation as an active control, an intervention commonly used by our group.¹² Herein we present the results of the RELAX trial of stretching versus meditation for lower body muscle cramps.

Methods

We conducted a randomized controlled trial of stretching versus meditation for the management of muscle cramps (Randomized EvaLuation of Stretching And Meditation eXperiment). We registered our trial in July 2022 prior to recruiting subjects (NCT05457322). The last subject was enrolled in July 2023. This study was approved after full review by the University of Michigan IRB (HUM00216488). All authors had access to these data and approved this manuscript.

Study population and recruitment

We sought to enroll adults, at least half of whom had cirrhosis (using clinical, histological, and imaging criteria) who had muscle cramps (“painful muscle spasms, cramps, or charley horses of the lower body that come on while resting”), that occurred >4 times in the prior month, and bothered the patient. We excluded persons with a history of cerebral palsy, stroke with paralysis, and multiple sclerosis. We did not exclude patients based on current or prior muscle cramp therapies. Patients were recruited by phone, email, or in-person. Patients could be enrolled from the University of Michigan, Beth Israel Deaconess Medical Center in Boston, Cedars-Sinai Medical Center in Los Angeles, Mayo Clinic in Rochester, or by voluntary contact through postings on clinicaltrials.gov or ihatedialysis.facebook.com. Informed consent was obtained electronically through SignNow or in-person.

Intervention

The trial lasted 35 days. After randomization, both groups were assigned to a 1-week training period where they were instructed to watch an instructional video training for their intervention followed by a 28-day period of outcome assessment. The stretching intervention was derived from a trial by Hallegraeff et al that involved 3 sets of exercises with breaks to stretch the calf and hamstring in sitting and standing positions.¹¹ They were instructed to do the exercises each night. The instructional video is available at: https://www.youtube.com/channel/UCf85x1_ATwI1tWZZZPL4F3g. As both stretching and meditation can be performed by anyone, we did not mention the intervention during the consent process, explaining that trial was evaluating a ‘Nonpharmaceutical Intervention for Nocturnal Muscle Cramps’ and labeled the interventions as exercises learned from audiovisual materials. The control we used was 10-minutes of guided meditation created by a licensed meditation therapist. The video used for this exercise can be accessed at: https://www.youtube.com/channel/UCylMExfaQtA8Oqjf2y4RKzw. At the conclusion of the study, all subjects had a disclosure visit where the design was explained, and the other trial arm was revealed.

Randomization and follow-up

All subjects were enrolled remotely. Patients were randomized 1:1 by drawing from a pool of 160 sealed envelopes and stratified by the presence of cirrhosis or not. All patients completed a baseline history to detail their sociodemographics, health history, and muscle cramp experience and concurrent cramp treatments. During the 28-day intervention phase, subjects were contacted on Study Days 10, 13, 16, 19, 22, 25, 28, 31, and 34 via automated or manual SMS text messages, phone calls, or emails. Questions included:

1. How many foot and leg muscle cramps did you have at night in the past 3 days? Please respond with a number. If zero, the subject will skip to question 3. If >0, the subject will continue:

2. On a scale from 0–10, where zero means no cramps and 10 means the worst imaginable cramps, how severe were your nighttime foot and leg muscle cramps in the past 3 days? Please respond with a number from 0–10.

3. Out of the last 3 days, how many days did you perform the (stretching or meditation) exercise? The subject only saw the treatment method applicable to their study group assignment.

Outcomes

Our primary aim was to assess if those in the intervention arm experienced a greater reduction in cramp severity than those in the control arm. This was measured using the change in Visual Analog Scale (VAS) for cramps between enrollment and day 35. The VAS was labeled scale that is numbered from 0 (means no cramps) to 10 (worst cramps imaginable). This measure was endorsed for muscle cramp trials by an exper panel led by Patel et al.¹³ Our secondary outcomes included:

1. Change in sleep quality based on the summary question from the Pittsburgh Sleep Quality Index (PSQI¹⁴) in which the participants reply how their sleep quality during the past month has been (a 5-point Likert from Very Good to Very Bad).

2. Change in patient global impression of change (PGIC). A one question survey coded from −3 to +3 in which the participants rate overall status as one of the following: Very Much Improved, Much Improved, Minimally Improved, No Change, Minimally Worse, Much Worse, Very Much Worse.

We had several exploratory outcomes:

1. The proportion of participants recommending their study arm after completion

2. Number of nights with foot and leg cramp severity <5 on the Visual Analog Scale for Cramps per subject.

3. Change in health-related quality of life (HRQOL) as assessed using the EQ-5D-3L.

Sample size derivation

To determine sample size for this trial, we used the effect size observed in a trial of stretching.¹¹ In the Hallegraeff et al trial of 80 subjects without liver disease, the authors detected a difference of cramp severity between stretching and no intervention of 1.3±1.1. This trial had 64 subjects analyzed with 16 dropouts. Allowing for a 3-fold greater variance of effect, we sought to enroll participants to achieve 80% power given a 5% alpha to detect a difference of 1.3 point on the VAS-cramps. We therefore enrolled 98 subjects, allowing for up to 20 dropouts.

Statistical analysis

Outcomes were analyzed in a modified intention-to-treat fashion where all subjects were analyzed according to their allocation. As the primary endpoint was patient reported, those who were lost to follow up or who withdrew could not be included in final analysis. Outcomes were compared between arms using ANCOVA and within arms using paired t-testing. Exploratory subgroups were selected based on clinical judgement regarding heterogeneity of cramp treatment effects: sex, cirrhosis, diabetes, and neuropathy.

Results

Study population

A flowchart of recruitment and enrollment activities is provided in Figure 1. Among those agreeing to discuss enrollment, cramp frequency <4 cramps/month was the most common exclusion. In total, 10 subjects were enrolled at non-Michigan sites: 4 from Cedars-Sinai, 3 from Mayo Rochester, 1 from Beth Israel Deaconess, 1 from clinicaltrials.gov, and 1 from ihatedialysis.facebook. After enrollment, loss-to-follow-up or disenrollment occurred in 9 subjects and 2 were excluded after serious (unrelated) hospitalizations. Baseline details for all subjects included are provided in Table 1. The sample was aged more than 62 years on average with an average, half had cirrhosis. Most (≥95%) reported muscle cramps that awoke them from sleep. HRQOL was poor with average Eq5D scores 0.82 and Global HRQOL VAS of 65–70. The stretching and meditation videos were watched an average of 5.4 and 12.3 times per participant, respectively.

Figure 1:

Consort diagram

Table 1 –

All: Baseline Characteristics of All Patients

		Stretching (n=53)	Meditation (n=45)
Demographics	Age – years (Median, IQR)	62.00 (18.00)	64.00 (15.00)
	Male	37.74%	31.11%
	White Race	94.34%	93.33%
	College education	77.36%	84.44%
Liver Disease History	Cirrhosis (%)	45.28%	51.11%
	Alcohol-related liver disease	7.55%	11.11%
	MASLD	24.53%	28.89%
	Ascites	20.83%	21.74%
	Hepatic Encephalopathy	20.83%	26.09%
	Liver cancer	0.00%	0.00%
Medical history	Dialysis	5.66%	4.44%
	Diabetes	32.08%	48.89%
	Neuropathy	26.42%	40.00%
	Diuretics	41.51%	44.44%
	Alcohol misuse	16.98%	15.56%
Prior Cramp management	Gabapentin/pregabalin	5.66%	11.11%
	Quinine	0.00%	2.22%
	L-Carnitine	0.00%	0.00%
	Baclofen	0.00%	4.44%
	Magnesium	11.32%	20.00%
	Vitamin E	3.77%	2.22%
	Taurine	0.00%	0.00%
	Potassium	3.77%	8.89%
	Tonic water	7.55%	13.33%
	Pickle juice	16.98%	28.89%
Cramp history	Cramps wake up from sleep	98.11%	93.33%
	Cramps/month	25.00 (32.00)	25.00 (35.00)
	Cramp duration (minutes)	17.00 (27.00)	15.00 (55.00)
	Calf cramps	86.79%	91.11%
	Thigh cramps	0.00%	0.00%
	Feet cramps	73.58%	80.00%
	Toes	67.92%	73.33%
Patient reported outcomes	VAS-Cramps	3.00 (6.00)	4.00 (6.00)
	Poor sleep	30.19%	26.67%
	Eq5d	0.82 (0.08)	0.82 (0.09)
	Global HRQOL VAS	65.00 (25.00)	70.00 (25.00)

HRQOL = Health-related quality of life, MELD-Na = Model for Endstage Liver Disease – Sodium, Metabolic-Dysfunction associated steatotic liver disease, VAS = visual analog scale

Outcomes

The differences in outcomes between arms is summarized in Table 2. Both arms experienced a reduction in cramp severity – a median of 1.44 points for stretching and 1.97 points for meditation. Within each arm, these changes were significant (p=0.001 for stretching, p<0.0001 for meditation) but these changes were equivalent between arms (p=0.4). Both arms experienced positive patient global impression of change (1.33 points for stretching, 1.05 points for meditation p-difference 0.2). Within each arm, sleep quality was improved (0.35 points for stretching, p=0.03; 0.37 points for meditation, p=0.04) but the changes were equivalent between arms (p-difference 1.0). More participants who stretched recommended their intervention than those who meditated (79.5% vs 55.3%, p=0.02). Global HRQOL did not change according to the Eq5d but was numerically higher according to the global HRQOL VAS for those who meditated (−0.06 points for stretching, 6.0 points for meditation; p = 0.06) and the within-arm change approached statistical significance for meditation (p=0.05). In subgroup analyses, there were no differences in the between-arm assessments according to sex or the presence of cirrhosis, diabetes, or neuropathy. Within groups, however, stretching appeared more effective for those without cirrhosis, diabetes, and neuropathy but there was no difference according to sex. Meditation was equally effective for men and women, cirrhosis or non-cirrhotic but was more effective for those with diabetes and without neuropathy.(Table 3) In the Supplmentary Table, we provide heterogeneity of effects for the duration of cramps. There were no significant differences beween the arms, however women had greater responses to both therapies and patients without cirrhosis appeared to benefit greatly from meditation.

Table 2:

Outcomes

	Stretching	Meditation	p-value
Primary Outcome
Change in cramp severity (scale 0/none −10/worst cramps)	-1.44 (−2.29, −0.58)	-1.97 (−2.93, −1.01)	0.4
Secondary outcomes
Patient Global Impression of Change (−3/very much worse to +3/very much improved)	1.33 (1.02, 1.65)	1.05 (0.70, 1.41)	0.2
Change in Sleep Quality (Higher is better)	0.35 (0.04, 0.67)	0.37 (0.01, 0.72)	1.0
Exploratory outcomes
Proportion recommending the intervention to the other group	79.2% (65.4%, 88.4%)	55.3% (39.5%, 70.1%)	0.02
End-of-trial cramp severity (Scale 0–10)	2.00 (1.32, 2.68)	2.11 (1.34, 2.87)	0.8
End-of-trial cramp duration	12.34 (7.22, 17.47)	13.63 (7.87, 19.39)	0.7
End-of-trial EQ5d (scale 0–1)	0.82 (0.79, 0.85)	0.83 (0.80, 0.87)	0.5
End-of-trial global HRQOL VAS (scale 0 to 100)	62.5 (56.4, 68.6)	71.2 (64.4, 78.1)	0.06
Change in global HRQOL VAS	−0.06 (−5.3, 5.1)	6.0 (0.1, 11.8)	0.1
Proportion of cramp-days with severity <5 on VAS	85.4% (72.4%, 92.9%)	81.6% (66.1%, 91.0%)	0.6

HRQOL = Health-related quality of life, VAS = visual analog scale

Table 3:

Subgroup analysis of changes in cramp severity

		Stretching	Meditation	P value
Sex	Men	−1.79 (−3.07, −0.51)	−2.25 (−3.86, −0.64)	0.7
	Women	−1.21 (−2.38, −0.04)	−1.85 (−3.08, −0.61)	0.5
Cirrhosis	Cirrhosis	−1.05 (−2.36, 0.26)	−2.37 (−3.71, −1.03)	0.2
	Non-cirrhosis	−1.71 (−2.88, −0.55)	−1.58 (−3.00, −0.16)	0.9
Diabetes	Diabetes	−1.13 (−2.52, 0.27)	−2.50 (−3.75, −1.25)	0.1
	Non diabetes	−1.59 (−2.71, −0.48)	−1.39 (−2.87, 0.10)	0.8
Neuropathy	Neuropathy	−0.75 (−2.45, 0.95)	−1.57 (−3.15, 0.00)	0.5
	Non-neuropathy	−1.74 (−2.79, −0.68)	−2.50 (−3.81, −1.19)	0.4
Advanced liver disease	Advanced	−0.29 (−2.55, 1.98)	−3.00 (−5.44, −0.56)	0.1
	Compensated	−1.63 (−2.57, −0.70)	−1.78 (−2.84, −0.72)	0.8

Adverse events

Adverse events were rare. Two patients reported discomfort with stretching, one hip and one in the knee. Both stopped the stretching for at least one night. One subject discontinued the stretching permanently and the other resumed. Two patients in the stretching arm were withdrawn from the study for prolonged hospitalizations: one developed severe COVID-19 and the other had decompensated congestive heart failure.

Discussion

The RELAX evaluated stretching compared to an active control of meditation for the treatment of muscle cramps. We found that both arms reduced cramp severity and resulted in positive changes in the patient’s health state without significant differences between interventions. While more people would recommend stretching over meditation, meditation may be more likely to improve global HRQOL.

Non-pharmacological solutions for in muscle cramp prevention

In the PICCLes trial, we previously found that sips of pickle juice reduced cramp severity (by 2.25 points on a VAS) while tap water sips did not (0.36 points) but pickle juice does not prevent cramps, improve sleep, or HRQOL. In RELAX, we find that both stretching, and meditation reduce cramp severity and improve sleep. In PICCLes, we found that the proportion of cramp-days with severity <5 on the VAS for cramps was 46% for pickle juice users and 35% for tap water users; in RELAX those proportions were substantially higher at 85% for stretchers and 82% for meditators. Given that there was no difference between the arms and there was no arm without an intervention, one critique is that we may have observed regression to the mean or a null effect. Though possible, there were no benefits of tap water in PICCLes (an inert control), and most subjects recommended their allotted intervention (though more 79% recommended stretching, compared to meditation, 55%). In RELAX, we found that stretching reduced cramp severity by 1.44 points on the VAS-Cramps 95%CI(−2.29, −0.58) and meditation reduced it by −1.97 95%CI(−2.93, −1.01). This magnitude of effect confirms and exceeds the 1.3 point reduction observed in the Hallegraeff et al trial which developed the protocol for older adults.¹¹ It is unclear why more subjects were more likely to recommend stretching than meditation despite the nominally larger effect size observed with meditation. Meditation may have stigma or people would be more comfortable recommending a physical intervention for a physical condition however additional research may be needed. Regardless, the concordance of results with a prior trial and having 79% of participants recommend stretching suggests the effect is clinically important. In RELAX, we enrolled patients with and without cirrhosis. As such, we believe that RELAX may have found two therapeutic interventions worthy of time-limited trials for patients suffering from muscle cramps regardless of the presence of cirrhosis. Given the duration of this trial – 35 days – we believe longer trials are needed. As we have shown that patients can be enrolled remotely, consented remotely, and monitored remotely, a large pragmatic decentralized trial may be feasible.

The RELAX population.

The participants in RELAX were uniform with respect to cramp frequency and severity. Suboptimal HRQOL was common – baseline Eq5D 0.8 and VAS-HRQOL of 65–70 – as was poor sleep (27–30%). In contrast to PICCLes, we enrolled patients with and without cirrhosis to produce results generalizable to the spectrum of cramp sufferers. This may reveal subtle heterogeneity of effects. For example, meditation appeared to be more effective for those with cirrhosis than without, those with diabetes than those without, and those with neuropathy than those without. We did not further characterize the etiology or electromyographic characteristics of the patient’s neuropathy and as such it is unclear whether specific neuropathies are more or less amenable to each therapy. Given the side-effects of pharmacological therapies for muscle cramps, clinicians wishing to trial a non-pharmacological intervention for patients encountered in practice may find these subgroup data informative. Further, we detail the prior treatment experience of the population in Table 1. Many patients were experienced with a variety of anti-cramp therapies, numerically more of whom were in the meditation arm. The protocol allowed for any combination of clinically prescribed cramp therapies and as such these data should be interpreted as providing the effect of stretching or meditation with or without concomitant interventions.

The cramp therapy landscape

Despite the high prevalence of cramps among patients with cirrhosis and chronic liver disease, very few large and controlled trials have been conducted.¹⁵ Among trials enrolling ≥30 subjects, Quinine, Taurine, and pickle juice have been evaluated.^{5, 9, 16} Quinine and taurine may prevent cramps but either carry a higher risk of adverse events or are an unregulated supplement. In a meta-analysis, 2 weeks of quinine (300–500mg) reduces cramp frequency by 2.45 cramps 95%CI(1.36–3.54), and slightly reduces cramp intensity on a 3-point scale by 0.12 95%CI(0.05–0.2) points.¹⁷ While many suggest tonic water, it only contains 83mg of quinine per liter and this is not sufficient to reach the doses studied in randomized trials. Although clinical trials of magnesium for patients with chronic liver disease are lacking, in a meta-analysis 4-weeks of magnesium is associated with an insignificant 0.18 cramp/week reduction (−0.89 to 0.49) in frequency and increase in cramp intensity of ≥moderate cramping (relative risk 1.33, 0.81–2.21).⁸ In one randomized trial of 49 subjects (30 of whom completed the protocol) who received Taurine or placebo for 2 weeks, Taurine resulted in 7 fewer cramps and a reduction in cramp intensity (on a 10-point VAS) of 1.4 points.¹⁶ In the context of these data, intervention options for cramps can progress from non-pharmacologic to pharmacologic in a stepped fashion. As a low-cost, widely available, and safe therapy, however, pickle juice could serve as a first-line therapy for patients with cramp-duration as the primary problem or consideration of stretching or meditation otherwise. Should these interventions prove unsatisfactory, a trial of pharmacological therapy could be considered. It should be noted that the mechanism of cramping has never been fully elucidated, prior effective trials of pickle juice or taurine did not determine the physiologic triggers for cramps, and that all cramp therapies are directed at reducing muscle or nerve hyperactivity. Just as pickle juice is used by athletes but only trialed in patients with cirrhosis or quinine is used for cramps in cirrhosis or not and stretching obtained the same effect size in this study as it did in a trial enrolling older participants without cirrhosis, the average effects of an intervention are likely generalizable broadly.

Contextual factors

These data must be interpreted in the context of the study design. First, this is short-term study and long-term benefits and risks are unknown. Second, cramps are a subjective outcome, however both arms resulted in reductions in cramp severity that were much greater than those observed with placebo in prior clinical trials.^{9, 16} Both stretching and meditation carried reductions in cramp-severity similar to those observed with stretching in a prior controlled trial.¹¹ Third, there was no inactive control arm because this was previously tested in PICCLes and we intented to assess superiority for physical versus non-physical therapies. Fourth, the mode of outcome assessment (SMS every three days) may have increased adherence to the intervention which must be considered when implementing clinically. Finally, remote enrollment facilitated recruitment but reduced our ability to phenotype patients based on laboratory or physical exam findings.

Conclusion

Both stretching and meditation were associated with large and equivalent reductions in cramp severity and improvements in the patients’ assessment of their health status. Participants may have preferred the stretching intervention but there was a slight improvement in HRQOL using a visual analog scale among those who meditated. These data support consideration of two non-pharmacological interventions for patients with frequent muscle cramps.

Data Transparency Statement:

Aggregated, deidentified data will be available with data use agreements though the University of Michigan