1Hz rTMS over left DLPFC rewired the coordination with hippocampus in insomnia patients: A pilot study

As the most common sleep disorder among adults, insomnia disorder is associated with substantial deleterious effects on mental and physical health and quality of life [[1]Stein M.B. et al.Genome-wide analysis of insomnia disorder.Mol Psychiatr. 2018; 23: 2238-2250Google Scholar], such as impaired memory consolidation and executive function [[2]Wardle-Pinkston S. Slavish D.C. Taylor D.J. Insomnia and cognitive performance: a systematic review and meta-analysis.Sleep Med Rev. 2019; 48: 101205Google Scholar]. The hyperarousal hypothesis is a robust framework for the conceptualization of insomnia etiology, and provides a potential target for intervention [[3]Kalmbach D.A. et al.Hyperarousal and sleep reactivity in insomnia: current insights.Nat Sci Sleep. 2018; 10: 193Google Scholar]. Specifically, disturbed sleep in insomnia patients is associated with greater cortical metabolism [[4]Nofzinger E.A. et al.Functional neuroimaging evidence for hyperarousal in insomnia.Am J Psychiatr. 2004; 161: 2126-2128Google Scholar]. Therefore, repetitive transcranial magnetic stimulation (rTMS) over the left dorsal lateral prefrontal cortex (DLPFC) has been widely used in the treatment of insomnia disorders, and subjective sleep improvements are commonly reported. However, findings based on objective sleep measures (polysomnography, PSG) were not consistent, and the underlying mechanisms of rTMS therapy for insomnia remain unknown [[5]Herrero Babiloni A. et al.The effects of non-invasive brain stimulation on sleep disturbances among different neurological and neuropsychiatric conditions: a systematic review.Sleep Med Rev. 2021; 55: 101381Google Scholar]. In the current study, we employed a double-blind, sham-controlled experimental design. Twenty active sessions of 1Hz rTMS on the left DLPFC in real group and 20 sham sessions in sham groups were carried out over four consecutive weeks. Both subjective sleep data indicated by Pittsburgh Sleep Quality Index (PSQI) and Insomnia Severity Index (ISI) and objective sleep data measured by polysomnography (PSG) were collected and compared between pre- and post-treatment. To shed some light on the potential durability of rTMS treatment for insomnia, the PSQI and ISI were acquired one-month later after their last rTMS session. Although the stimulation target is commonly a single brain region, converging evidence suggests that the beneficial effects of rTMS may be mediated via distributed networks. Thus, the undirected and directed interaction changes of the left DLPFC-centered brain circuits were examined by resting-state functional connectivity (RSFC) and Granger causality analysis (GCA) between pre- and post-treatment. The study was approved by the local Institutional Review Board and was registered in the Chinese Clinical Trial Registry (No. ChiCTR2100042449). According to DSM-5 criteria, we enrolled 44 insomnia patients (28 female; mean age: 43.8 ± 10.3 years; mean ± SD) at the Second Hospital of Hebei Medical University, Shijiazhuang, China. Twenty-two age-, gender- and education-matched healthy controls were enrolled as well. Written and informed consent was obtained from all participants. Insomnia patients were separated into real (n = 22) or sham groups (n = 22) randomly. Then, 20 active or sham sessions of left DLPFC rTMS were delivered over 4 consecutive weeks (5 times/week; Fig. 1. A). Insomnia disorder patients were assessed before and after real or sham rTMS treatment, including behavioral scales (PSQI, ISI, MoCA, MMSE, BAI, BDI), PSG recording (Grael 4K system, Australia) and MRI scanning (Philips Achieva 3.0T, Netherland). Due to personal reasons, the number of people who completed pre- and post-treatment PSG data collection was 17 (real group) and 18 (sham group). Active low-frequency rTMS was administered using a MagPro R30 TMS stimulator with a Fig. 1-shaped coil (MagVenture, Denmark) with the "5-cm rule" to locate the left DLPFC [[6]Yuan K. et al.Potential neural mechanism of single session transcranial magnetic stimulation on smoking craving.Sci China Inf Sci. 2020; 63: 1-3Google Scholar]. The rTMS stimulation was delivered at 1Hz, and the stimulus intensity was set at 80% of the resting motor threshold (RMT). Sham rTMS is also carried out as the coil is turned away from the skull at 90°. No side effects were reported during or after brain stimulation. The sleep stage classification was conducted by employing YASA algorithm on the PSG data [[7]Vallat R. Walker M.P. An open-source, high-performance tool for automated sleep staging.Elife. 2021; 10: e70092Google Scholar]. Then quantitative variables were calculated, such as sleep efficiency (SE), non-rapid eye movement sleep stages (NREM3) sleep duration, and sleep onset latency (SOL). RSFC and GCA were used to identify the rTMS-induced changes of the left DLPFC functional circuits in insomnia patients. Further technical details are provided in the Supplementary material. Two-way ANOVA revealed significant "treatment × time" interaction effect in both subjective sleeping measures (PSQI: F = 15.40, p = 0.0003; ISI: F = 44.92, p = 0.0001) and objective sleeping measurements (SE: F = 32.45, p = 0.0001; NREM3: F = 7.987, p = 0.0079; SOL: F = 9.356, p = 0.0044). In the real rTMS group, subjective sleep improvement was found by showing reduced PSQI (t = 5.97, p = 0.0001) and ISI (t = 6.54, p = 0.0001) (Fig. 1. B). Similarly, objective sleep improvement was found, as evidenced by increased sleep efficiency (t = 3.55, p = 0.0027), NREM3 sleep duration (t = 2.20, p = 0.043), and shorter sleep onset latency (t = 2.58, p = 0.02) (Fig. 1. C). One-month follow-up revealed a significant improvement in subjective sleeping measurements (PSQI, ISI) compared with the pre-treatment period (PSQI: t = 3.88, p = 0.0011; ISI: t = 6.59, p = 0.0001) (Fig. 1. D). However, those findings mentioned above were not detected in the sham group (p > 0.05). Neither real nor sham rTMS produced significant differences in MMSE, MoCA, BDI and BAI scores between pre- and post-rTMS treatment. Meanwhile, reduced RSFC between the left DLPFC and right superior frontal gyrus (SFG) was observed in insomnia patients after real rTMS treatment relative to baseline (FWE corrected, p < 0.05), which approached the average level of healthy controls (Fig. 1. F). Furthermore, GCA revealed disrupted functional coupling from the left hippocampus to the left DLPFC after real treatment (FWE corrected, p < 0.05) (Fig. 1. G). The RSFC changes of the left DLPFC-middle temporal lobe circuit were significantly positively associated with sleep improvement (i.e., PSQI changes) in the real group (r = 0.763, p = 0.0001, Fig. 1. E). Our pilot study validated the effectiveness and investigated the durability of rTMS therapy on insomnia disorders by demonstrating both subjective and objective sleep quality improvement. Moreover, we explored the possible mechanisms of the rTMS therapeutic effect for insomnia by using fMRI. The 1Hz rTMS over left DLPFC showed the potential to normalize the hyperexcited functional connections between the left DLPFC-right SFG cortex in insomnia patients (Fig. 1. F). Further, we observed disrupted cortico-hippocampal interactions in insomnia patients, and rTMS may have rewired these impaired connections (Fig. 1. G). This may supports its use as a novel intervention for memory consolidation deficits in insomnia patients [[8]Girardeau G. Lopes-dos-Santos V. Brain neural patterns and the memory function of sleep.Science. 2021; 374: 560-564Google Scholar]. Finally, our findings revealed the remote effects occurred in rTMS treatment for insomnia disorder [[9]Castrillon G. et al.The physiological effects of noninvasive brain stimulation fundamentally differ across the human cortex.Sci Adv. 2020; 6: eaay2739Google Scholar], which should be taken into consideration when using rTMS for brain disorders. In conclusion, we reported therapeutic improvement in sleep and potential underlying mechanisms of 1HZ rTMS therapy over the left DLPFC in insomnia disorders, which were associated with the modulation of the left DLPFC centered pathway. The lack of significant improvement in memory tasks in the current study leaves the possible link between the cortico-hippocampal interaction and memory performance unresolved. Thus, our findings should be considered preliminary and need to be replicated. Meanwhile, more accurate location of targets, advanced stimulation patterns and predictions of treatment outcomes [[10]Liu S. et al.Brain responses to drug cues predict craving changes in abstinent heroin users: a preliminary study.Neuroimage. 2021; 237: 118169Google Scholar] are also encouraged to improve rTMS therapy. This work was supported by National Natural Science Foundation of China (Grant Nos. 81871426, 81871430). The authors report no other conflicts of interest. We thank all the participants who volunteered their time to take part in this research. We are appreciated for the help in English writing contributed by Peter Manza from the Laboratory of Neuroimaging, NIAAA, NIH. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. The following is the Supplementary data to this article: Download .docx (.03 MB) Help with docx files Multimedia component 1