Effects of different frequencies of repetitive transcranial magnetic stimulation on sleep disorders and depression in patients with Parkinson’s disease: a systematic review and network meta-analysis

Parkinson’s disease (PD), which primarily affects middle-aged and older adults, is the world’s second most common neurodegenerative disorder (Rocca, 2018; GBD 2016 Neurology Collaborators., 2019). In addition to its hallmark motor symptoms such as tremor, rigidity and bradykinesia, PD is also characterized by a wide range of non-motor impairments, including mood disturbances, cognitive decline and sleep disorders (Hendricks and Khasawneh, 2021; Kirmani et al., 2021). The exact pathogenesis of PD remains unclear; current evidence suggests that non-motor symptoms predominantly arise from diminished dopaminergic transmission within the midbrain–limbic and midbrain–cortical systems (Moore et al., 2008). Most individuals with PD experience sleep problems early in the disease course or even before overt motor signs appear (Barone et al., 2009). Common sleep disorders in PD include rapid eye movement sleep behavior disorder (RBD), insomnia, restless legs syndrome, sleep-related breathing disturbances and excessive daytime sleepiness (Chahine et al., 2017). These disturbances may result from side effects of dopaminergic medications, neurodegenerative changes in brain structures that regulate sleep and nocturnal motor symptoms (French and Muthusamy, 2016). The regulation of sleep relies on the comprehensive function of multiple brain regions and various neurotransmitters, including dopamine, serotonin, norepinephrine, and other PD-related neurotransmitters (Stefani and Högl, 2020). These neurotransmitters not only regulate sleep disorders but may also be associated with cognitive dysfunction in PD (Yeung and Cavanna, 2014; Maggi et al., 2021; Malhotra, 2022). Emotional disorders involve depression, anxiety, etc. Recent studies have shown that there is a strong correlation between sleep quality and depressive and anxious emotions in PD patients, and the severity of sleep disorders is related to the degree of depression (Kay et al., 2018; Rana et al., 2018). Depression is another prevalent non-motor feature of PD and often has an insidious onset that precedes typical motor manifestations (Langston, 2006; Simonetta-Moreau, 2014). In patients with PD, depressive episodes frequently co-occur with anxiety, irritability, sadness and pessimism about the future, all of which can worsen sleep disturbances (van der Hoek et al., 2011). Depression and sleep disorders frequently co-occur in PD, each exacerbating the other in a self-perpetuating cycle. The close link between sleep dysfunction and depression leads to loss of mobility, reduced functional independence and severe impairments in mood and daily quality of life (Hariz and Forsgren, 2011; Zhao et al., 2021). These challenges also place a heavy burden on families and healthcare systems, since the global economic cost of PD reached an estimated $52 billion in 2017 and is projected to exceed $80 billion by 2040 as populations continue to age (Dorsey et al., 2018). Current management of PD relies mainly on dopaminergic pharmacotherapy, but long-term use of these agents carries risks of adverse effects and may even accelerate neurodegeneration (Jiménez-Urbieta et al., 2015; Seppi et al., 2019). It is therefore critical to explore non-pharmacological interventions that carry fewer risks and can help alleviate both sleep disturbances and depression in PD patients.

Repetitive transcranial magnetic stimulation (rTMS) is a non-invasive neuromodulation technique grounded in electromagnetic induction; a pulsed magnetic field is applied to the skull surface to induce weak electrical currents in targeted brain regions (Chail et al., 2018). Depending on the frequency of the stimulation pulses, rTMS is classified as either low-frequency (≤1 Hz, low frequency rTMS, LF-rTMS) or high-frequency (>1 Hz, high frequency rTMS, HF-rTMS) stimulation (Xia et al., 2022). Previous studies have demonstrated that its therapeutic effects arise from bidirectional modulation of cortical excitability: low-frequency rTMS reduces excitability, while high-frequency rTMS enhances it (Cirillo et al., 2017). However, direct comparisons across frequencies are scarce, so this study employed a network meta-analysis (NMA) to evaluate how different rTMS frequencies affect sleep disorders and depression in PD patients and to identify the optimal stimulation frequency for clinical use.

This systematic review was conducted in accordance with Preferred Reporting Items for Systematic Evaluation and Meta-Analysis statement (PRISMA) guidelines (Hutton et al., 2015) and the Cochrane Handbook for Systematic Reviews of Interventions to ensure methodological rigor. The protocol was registered with PROSPERO under registration number CRD42024614337.

This study employed a NMA to evaluate and compare the effects of rTMS at various frequencies, combined with conventional therapy, on sleep disorders and depressive symptoms in patients with Parkinson’s disease. Compared with conventional therapy alone, all rTMS frequencies significantly improved PSQI, PDSS, and HAMD scores, with 10 Hz rTMS appearing to be the most effective for both sleep and mood. We then stratified stimulation by pulse count (600 pulses vs. >600 pulses). In the 600-pulse group, which did not include 10 Hz stimulation, 5 Hz rTMS yielded the greatest benefit; in the >600-pulse subgroup, 10 Hz rTMS produced the most pronounced improvements.

This study employed a NMA to assess and compare the effects of different rTMS frequencies, each combined with conventional therapy, on sleep disorders in PD patients. The findings demonstrated that, relative to conventional treatment alone, all rTMS frequencies significantly improved PSQI, PDSS and HAMD scores, with 10 Hz rTMS plus standard therapy emerging as the most effective intervention for each outcome. There are certain differences in the findings regarding the impact on sleep between this study and that of Cristini et al. (2025). Cristini’s research revealed that LF-rTMS can enhance subjective sleep quality in PD patients, yet the evidence for HF-rTMS improving sleep quality is insufficient. This discrepancy may stem from Cristini’s study not precisely categorizing HF-rTMS by frequency, potentially leading to interference from mixing different frequency groups. Additionally, the patients in the HF-rTMS group in that study had relatively mild sleep issues, which could have contributed to a ceiling effect. Regarding depression, a previous meta-analysis (Zhou et al., 2018) indicated that 5 Hz rTMS is most effective in alleviating depressive symptoms. Upon comparison, we found that the studies included in that meta-analysis were self-controlled before-and-after designs, and the number of included studies was limited. Currently, it is believed that abnormal discharges in the subthalamic nucleus (STN) of PD patients are transmitted through the cortical–striatal–thalamic circuit, leading to disruptions in the sleep–wake cycle. Repetitive TMS is one of the most widely applied neurostimulation modalities: high-frequency rTMS (HF-rTMS), defined as stimulation above 1 Hz has been shown to induce long-term excitatory effects (Valero-Cabré et al., 2017), whereas low-frequency rTMS (LF-rTMS), defined as 1 Hz or below, is expected to produce inhibitory effects and elicit long-term depression (Romero et al., 2002).

Low frequency rTMS reduces sleep fragmentation by attenuating abnormal beta oscillations (20–30 Hz) in the thalamus, subthalamic nucleus (STN) and motor cortex via long-term depression (LTD) (Chen et al., 1997). Simultaneous stimulation of the prefrontal cortex (PFC) increases δ-wave (1–4 Hz) power and prolongs slow-wave sleep. LF-rTMS also upregulates striatal dopamine D₂-receptor expression, enhances dopaminergic signaling, and alleviates Parkinson’s disease-associated REM sleep behavior disorder (RBD) (Ahmed et al., 2012). Moreover, cortical rTMS promotes the release of dopamine and pineal melatonin, increases brain serotonin and norepinephrine levels, and elevates serum GABA–neurotransmitters critical to the sleep–wake cycle–thereby improving sleep quality and reducing daytime somnolence (Strafella et al., 2003; Feng et al., 2019). High-frequency (HF) rTMS activates the dorsolateral prefrontal cortex (DLPFC) and anterior cingulate cortex (ACC) via long-term potentiation (LTP), inhibits the noradrenergic arousal system in the locus coeruleus (LC), and ameliorates excessive daytime sleepiness (Lefaucheur et al., 2020). Prior studies have shown that HF-rTMS over the parietal lobe enhances deep sleep and sleep efficiency while reducing nocturnal awakenings in PD patients, suggesting the parietal cortex as a key target for deepening subsequent sleep by decreasing Stage I and increasing Stage IV sleep (van Dijk et al., 2009). HF-rTMS also augments cortical excitability, improves cerebral blood flow, and promotes endogenous dopamine release, thereby modulating excitation within the direct and indirect striatal–pallidal pathways, which may further alleviate sleep disturbances (Chou et al., 2015; Qin et al., 2018). The superior efficacy of 10 Hz rTMS observed here may reflect dose-dependent neuroplastic changes: higher pulse counts strengthen neural network connectivity and induce sustained synaptic potentiation, enhancing neuromodulatory potential (Zhou et al., 2018; Anil et al., 2023). This dose dependency is supported by our subgroup analysis, which indicates that stimulation dosage differentially affects outcomes in PD patients.

In our network meta-analysis of depressive symptoms, hyporeactivity of the left DLPFC has been implicated in PD-related depression (Mottaghy et al., 2002). Clinical protocols therefore aim to increase left DLPFC excitability while inhibiting right DLPFC activity: LF-rTMS to the right DLPFC reduces cortical excitability, diminishing negative affect and trans-synaptically activating the hypoactive left DLPFC (Grimm et al., 2008). Conversely, HF-rTMS elicits release of dopamine, serotonin (5-HT), glutamate, and brain-derived neurotrophic factor (BDNF). Because depression in PD involves deficits in dopaminergic and serotonergic systems, rTMS may improve mood through dual-transmitter regulation (Strafella et al., 2001). HF-rTMS targeting the DLPFC also modulates prefrontal–limbic functional connectivity (e.g., amygdala, ACC) by enhancing local neuronal excitability, inhibiting aberrant default mode network (DMN) activity, and strengthening frontal regulation of limbic regions (Lefaucheur et al., 2020). Furthermore, the observed correlation between sleep quality and mood, namely, PD patients with poor sleep exhibit more severe depression than those with normal sleep, suggests that amelioration of sleep disturbances may contribute to improvements in depressive symptoms. Patients with neurodegenerative diseases often exhibit higher rates of depression than the general population (Tandberg et al., 1998). Dopaminergic dysfunction is hypothesized to underlie the strong association between poor sleep quality and depression severity in Parkinson’s disease (PD). In healthy individuals, sleep deprivation elicits a compensatory increase in central dopamine levels; however, PD-related dopamine deficits may impair this adaptive response, thereby exacerbating depressive symptoms (Kay et al., 2018). Moreover, improvements in HAMD anxiety scores have been positively correlated with PSQI improvements, suggesting that enhanced sleep quality is associated with reduced anxiety (Huang et al., 2018).

The use of dopaminergic drugs may also affect the efficacy of rTMS. Previous studies (Fierro et al., 2008) found that 10 Hz rTMS only enhances cortical inhibition during drug withdrawal in PD patients, whereas the improvement in cortical inhibition during medication use is comparable to that of the drugs themselves. All subjects included in this study were on dopaminergic drugs during the trial period. The reason for the divergence may be related to the large number of subjects included in this study–all of whom were randomized controlled trials–as well as differences in intervention methods and targets. Fierro et al. (2008) used 10 Hz, 500-pulse stimulation over the M1 region, while most studies in this review used 10 Hz, 1200-pulse stimulation, with the stimulation targets mostly being the DLPFC, which may also account for the differences in results. The stage of PD is another factor affecting the efficacy of rTMS. Flamez et al. (2016) found that LF-rTMS did not significantly improve motor function in PD patients, possibly because all subjects in their study were late-stage PD patients (H&Y ≥ 3). In this study, most subjects in the included literature were in H&Y stages 1–3 (only one was a late-stage patient; after sensitivity analysis, the results remained unchanged; see Supplementary Figures 10, 11), which also explains the differences between this study’s results and those of previous studies. Meanwhile, previous studies (Cong et al., 2022) have also shown that the degree of sleep disturbance and depression in PD patients is positively correlated with H&Y stage. Late-stage PD patients have extensive neurodegenerative lesions, and local stimulation may not be able to regulate distant pathological networks.

Regarding adverse events reported in this study, only a small number of subjects experienced transient dizziness, headache, or scalp numbness, which resolved after rest and allowed them to complete the trial. This indicates the safety of rTMS treatment for PD patients and supports its clinical application. The funnel plot results showed overall symmetry but with a small number of scatter points outside the funnel, so the findings should be interpreted with caution, and more high-quality studies are needed for future verification.

Several limitations should be acknowledged. First, rTMS target regions and pulse counts varied across the included studies, limiting the generalizability of our findings. Second, previous studies (Chung et al., 2019) have indicated that high-level estrogen exposure during HF-rTMS stimulation can enhance the neuroplasticity effect of the prefrontal cortex, suggesting that gender may also influence stimulation outcomes. This study includes a mixed-gender sample from the literature, making subgroup analysis impossible. Third, the severity of sleep disturbance correlates positively with age in PD, yet all participants in the analyzed studies were over 60 years old, precluding age-stratified subgroup analyses. Therefore, future research can focus on more personalized designs for rTMS stimulation targets, pulse counts, gender, and age to provide references for clinical applications.

In summary, this analysis demonstrates the potential of different rTMS frequencies to ameliorate sleep disturbances and depressive symptoms in PD patients. Notably, 10 Hz rTMS emerged as the most effective intervention for both outcomes. These results provide clinicians and researchers with valuable guidance for managing non-motor symptoms in PD.

The author(s) declare that financial support was received for the research and/or publication of this article. This study was funded by Guizhou Provincial Science and Technology Support Programme Project (NO:²⁰²³general179).

The original contributions presented in this study are included in this article/, further inquiries can be directed to the corresponding author. Supplementary material

YX: Writing – review & editing, Writing – original draft, Data curation, Software. HW: Methodology, Investigation, Supervision, Writing – review & editing, Conceptualization. XH: Validation, Data curation, Methodology, Supervision, Conceptualization, Writing – review & editing. WS: Investigation, Conceptualization, Data curation, Software, Methodology, Writing – review & editing. YL: Validation, Writing – original draft, Data curation, Methodology, Writing – review & editing.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnagi.2025.1623917/full#supplementary-material↗