Efficacy and Safety of Pulse Magnetic Therapy System in Insomnia Disorder: A Multicenter, Randomized, Double-Blind, Placebo-Controlled Trial

Study design and setting

This multi-center, double-blind, sham-controlled randomized clinical trial included a 1-week baseline observation, a 4-week PMTS or sham treatment, and a 1-week follow-up observation. The study design and protocol were approved by the ethics committees of three centers: the First Affiliated Hospital of Jinan University (Ethics Approval number: 2018-009), the Third Affiliated Hospital of Guangzhou Medical University (Ethics Approval number: 2019-009), and Shenzhen mental Hospital (Ethics Approval number: 2019-001-01). The protocol was registered at Chinese Clinical Trial Register (https://www.chictr.org.cn; identification number: CTR1900024627).

Participants

Sample size was calculated using nQuery^TM Advanced (https://www.statsols.com/; Statsols, San Diego, CA, USA). The expected effect size was set as 0.5 according to previous studies that suggested a medium to high level of effect size [21,22]. With a two-tailed α=0.05, 1-β=0.80, and a drop-out rate=20%, a sample size of 77 participants per group (154 in total) was necessary.

Participants were enrolled at the three research centers or via social media advertisements. Participants predetermined by their clinicians as being potentially suitable and having interest in the current study underwent a set of screening, including a structural clinical interview, diagnostic scales (Insomnia Severity Index [ISI], Pittsburgh Sleep Quality Index [PSQI], Generalized Anxiety Disorder-7 [GAD-7], and Patient Health Questionnaire-9 [PHQ-9]), blood pressure and heart rate measurements, and a single-night at-lab polysomnography (PSG). A total of 153 participants were included in the final sample (Figure 1). All participants provided written informed consent.

Figure 1.

Consolidated Standards of Reporting Trials enrollment flow diagram. AEs, adverse events.

The inclusion criteria were: 1) meeting the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition [23] diagnostic criteria for insomnia disorder (F51.01); 2) being aged 18 to 65 years; and 3) signing the informed consent form.

The exclusion criteria were: 1) insomnia associated with any physical or mental disorders which was caused by substance use; jetlag or work shifts; 2) untreated obstructive sleep apnea syndrome (apnea–hypopnea index >15/h) or periodic limb movement of sleep (Periodic Limb Movement of Sleep Index >15/h), which was determined by overnight PSG, or other severe physical health issues (e.g., cancer, AIDS, cardiovascular, and liver and renal dysfunction); 3) moderate to severe anxiety or depression (GAD-7 >9, PHQ-9 >9); 4) employments that are not suitable for the test; 5) pregnancy; 6) history of epilepsy; 7) current or past neural modulation treatment; 8) implanted with metal or magnets; 9) participated in other clinical studies in 3 months; and 10) other issues determined by the researchers.

Enrolled participants were randomized 1:1 to the PMTS or sham group according to a predetermined random-number table-generated allocation sequence. This randomization sequence was generated by an independent third-party using SAS 9.4 (www.sas.com; SAS Institute, Cary, NC, USA), and was concealed from participants, investigators, and statisticians.

Interventions

PMTS version 1.0 (BestCare & SuMian BioTech Co., Ltd., Dongguan, China) is a portable therapeutic system designed for noninvasive sleep improvement. It consists of a portable pulsed magnetic field generator positioned under the head as the participant goes to bed, and a wrist sleep monitor tracking real time sleep stage. It is designed to generate pulsed magnetic fields of 1 Hz as the participant goes to sleep. The generation of magnetic pulses is controlled by algorithm according to the sleep-stage data, so that the stimulation is stopped after falling asleep.

In the current study, the PMTS group received active PMTS intervention every night over the 4 weeks of treatment. The sham group received fake intervention, in which all the procedures were the same except that the generators did not actually output pulsed magnetic fields. To ensure that the sham group was blind to their grouping, the sham generators emitted mild heat to simulate the working mode of the active generators.

Outcome measurements

The primary outcome was the changes in total score of the ISI, a 7-item symptomatic measure of insomnia that has good validity, reliability, and sensitivity to short-term insomnia changes, assessed at the end of each week. Usually, a total score of 7 or less indicates no clinically significant (insomnia) [24], while according to Chung et al. [25], 9 is suggested to be a better cutoff point for Chinese populations. Therefore, the current study defined a total ISI score of ≤9 as response. The response rate was calculated by dividing the number of responses by the number of participants.

Secondary outcomes included PSQI scores at the end of baseline and week 4, and weekly sleep diary-derived values for sleep latency, sleep efficiency, actual sleep time, and wake after sleep onset (WASO).

Statistical methods

Data were analyzed using IBM SPSS Statistics for Windows, Version 26.0 (IBM Corp., Armonk, NY, USA). A multiple imputation approach was used to impute missing values. General linear models were used to test group×time interactions and post-hoc analysis between-group and between-time effects for the ISI, PSQI, and sleep diary-derived outcomes. Generalized estimating equations were conducted for response rates over time. A Pearson’s χ² test or Fisher’s exact test was conducted for baseline characteristics and adverse events (AEs). All tests were two-sided, with a test level of α=0.05.

Primary outcomes

Results of the primary outcome, the ISI scores, were presented in Figure 2. The ISI scores of the PMTS group and the sham group were 7.13±0.50, 11.07±0.51 at week 4, respectively. There was a significant group×time interaction (F_{3.214, 485.271}=24.25, p<0.001, η_p²=0.138). Simple effect analysis revealed that the PMTS group improved weekly during the whole study (changes from baseline: Cohen’s d=1.07, 1.69, 1.92, 2.21, and 2.38). The sham group also improved significantly within the first week (changes from baseline: Cohen’s d=0.95), but then only improved very mildly afterward (changes from baseline: Cohen’s d=0.96, 1.07, 1.12, and 1.21). Analysis of group effect indicated that the PMTS group began to overperform the sham group since the 2nd week (post-hoc between-group p=0.592, 0.652, <0.001, <0.001, and <0.001; Cohen’s d=0.09, 0.07, 0.59, 0.73, 0.86, and 0.98 at the end of each week).

Figure 2.

The ISI scores of the PMTS group and the sham group. An ISI score ≤9 indicated no clinically significant insomnia disorders. Error bars indicated ±95% CI. ^*p_{between-group}<0.05. ISI, Insomnia Severity Index; PMTS, Pulsed Magnetic Therapy System; CI, confidence interval.

In addition, an ISI ≤9 was defined as not having clinically significant insomnia disorder, or as response (Table 2). Results showed a similar interactive between group and time (χ²=17.69, p=0.003). The PMTS group had outperformed the sham group since week 2, while nearly 70% of the participants who received PMTS treatment did not show clinically significant insomnia disorder at the end of the intervention.

Table 2.

Response rate at the end of each week

Secondary outcomes

A significant group×time interaction (F_1,151=18.05, p<0.001, η_p²=0.107) was found on PSQI scores (Figure 3A). Specifically, both groups improved after intervention compared to baseline (PMTS group: Cohen’s d=1.58; sham group: Cohen’s d=0.80), while the PMTS group experienced a greater improvement (between-group Cohen’s d=0.15, 0.58 for baseline and week 4, respectively).

Figure 3.

PSQI scores and weekly sleep diary-derived value of the PMTS group and the sham group. A: PSQI score. B: Sleep latency. C: Sleep efficiency. D: Actual sleep time. E: Wake after sleep onset. Error bars indicated ±95% CI. ^*p_{between-group}<0.05; ^**p_{between-group}<0.01; ^***p_{between-group}<0.001. PMTS, Pulsed Magnetic Therapy System; PSQI, Pittsburgh Sleep Quality Index; CI, confidence interval.

Weaker but still significant group×time interactions were found for all the diary-derived indexes (Figure 3B-E): sleep latency (F_{3.822, 577.116}=5.01, p=0.001, η_p²=0.032), actual sleep time (F_{3.851, 581.466}=3.60, p=0.007, η_p²=0.023), sleep efficiency (F_{3.937, 594.482}=4.28, p=0.002, η_p²=0.028), and WASO (F_{3.312, 500.085}=2.94, p=0.028, η_p²=0.019).

AEs

Participant-reported AEs were compared between groups (Table 3). There were 26 and 42 cases of AEs related to the research medical device in the experimental group and the control group, respectively. Among them, 67 cases were mild, 1 case was moderate, and none of them were severe in severity. There was no significant difference in the incidence of AEs between groups (p=0.052).

Table 3.

Participant-reported adverse events (AEs)

Limitations and generalizations

The current findings must be considered in light of several issues. First, although the two groups were highly comparable in ISI and PSQI at baseline, we did find a tendency that dairy-derived indexes were worse for the PMTS group. This may cast some doubt on the homogeneity of our sample; however, the significant find of the group×time interactions strongly supported the efficacy of the treatment. Second, various populations, such as the elderly, children and adolescents, pregnant women, and patients with other physical or mental disorders, were not included in the current study. Although studies concerning pulsed magnetic therapy have generally found a good tolerance for such techniques among different populations [29], considering the need for at-home usage, it must be highly cautious to apply PMTS to these populations. Third, considering the majority (93.5%) of the participants with mild to moderate insomnia disorder at baseline, the conclusions of current study should not be directly drawn on severe insomnia disorder, which is found to be closely associated with cortical hyperarousal [18,30]. PMTS is therefore expected to be more efficient than psychological treatments for severe insomnia disorder [30], but still needs further research. Fourthly, objective outcome was not included like PSG or actigraphy and follow-up duration was too short to evaluate the durability or sustainability of efficacy of PMTS.

Conclusion

The current study provides compelling evidence that a portable pulsed magnetic therapy system, PMTS, is of good efficacy and safety among adults with mild to moderate insomnia disorder. The positive effect of PMTS on insomnia patients could last for at least 1 week. Considering its portability and user friendliness, PMTS has great promise as an at-home atnight treatment for insomnia disorder.