Insomnia Disorder is characterized by persistent symptoms, including difficulty initiating or maintaining sleep, or early morning awakening, that meet formal diagnostic criteria. These symptoms must also cause clinically significant distress and lead to impairments in social, occupational, or other important areas of functioning. The International Classification of Diseases, 11th Revision (ICD-11), categorizes the disorder into three subtypes: Chronic Insomnia Disorder (CID), Short-Term Insomnia Disorder (SID), and other unclassified hyposomnia. The persistence of symptoms for 3 months distinguishes chronic from short-term conditions (The Lancet, 2019). Insomnia is a complex and prevalent global health issue, affecting about 10–15% of adults (Buysse, 2013). Each year, about 27% of the population experiences SID, with roughly 72% of cases resolving on their own. A subset of individuals progresses to CID (LeBlanc et al., 2009). In China, population-based studies report a 15% prevalence of CID in the general population, rising to 35.9% among adults aged 60 and older (Lu et al., 2019). The disorder severely compromises quality of life and imposes a substantial societal and economic burden. Current medical interventions are primarily pharmacological therapies and Cognitive Behavioral Therapy (CBT) (Krystal et al., 2019; Winkelman, 2015). However, both drug-based treatments and CBT (Voelker, 2019; Kay-Stacey and Attarian, 2016; Morin, 2017) have documented limitations. Developing effective, targeted therapeutic strategies for specific insomnia subtypes is now more critical than ever.

Although SID and CID share similar symptoms, they are underpinned by distinct pathophysiological mechanisms and clinical profiles. SID typically stems from a state of acute hyperarousal, whereas CID is characterized by a failure to inhibit wakefulness, resulting in conditioned arousal (Vargas et al., 2020). Consequently, they should not be conceptualized as different stages of a single disorder, nor can CID be adequately explained by a unitary hyperarousal model. Elucidating their unique neurophysiological substrates is therefore critical, as it will pave the way for more individualized and effective interventions, ultimately improving patient care and long-term management.

Transcranial magnetic stimulation (TMS) provides a non-invasive method for assessing cortical excitability in vivo (Kobayashi and Pascual-Leone, 2003). When a single TMS pulse is applied to the primary motor cortex (M1), it elicits a motor evoked potential (MEP) in contralateral muscles, enabling functional evaluation of the corticospinal conduction (Di Lazzaro et al., 2001). The amplitude of this MEP serves as an index of cortical excitability. A key metric derived from this technique is the resting motor threshold (RMT), defined as the minimum stimulus intensity required to produce an MEP of >50 μV in at least five of ten consecutive trials. The RMT is believed to reflect the excitability of a central core of neurons within M1, influenced by neuronal membrane properties and non-NMDA receptor-mediated glutamatergic neurotransmission. RMT typically increases with damage to the corticospinal tract but decreases in states of cortical hyperexcitability (Rossini et al., 2015). Recently, TMS has been employed to investigate cortical excitability across various sleep disorders (Lanza et al., 2015). While existing studies consistently report altered cortical excitability compared to healthy controls, the field remains limited. The heterogeneity of sleep disorders, combined with a small number of studies, precludes a comprehensive understanding. Crucially, it remains unknown whether these changes in excitability are related to the specific neural mechanisms underlying different subtypes of sleep disorders.

Functional near-infrared spectroscopy (fNIRS) is a non-invasive optical imaging technique that measures cortical changes in oxy-hemoglobin (Oxy-Hb) and deoxy-hemoglobin (Deoxy-Hb) during brain activity, thereby detecting dynamic functional changes throughout cognitive processing (Liu et al., 2024). It offers a favorable balance between temporal and spatial resolution (Kohl et al., 2020), alongside practical advantages such as low cost and high portability. These characteristics make fNIRS particularly suitable for a wide range of experimental settings and populations, including studies conducted in dynamic or naturalistic environments (Pinti et al., 2020). Consequently, fNIRS is utilized increasingly in the auxiliary diagnosis and clinical evaluation of neuropsychological disorders. The Verbal Fluency Task (VFT) is a well-established paradigm for assessing cognitive function in individuals with neurological and psychiatric disorders (Takahashi et al., 2011). It is frequently employed in conjunction with fNIRS to monitor functional cognitive deficits (Zhang et al., 2023). The VFT engages multiple cognitive domains, including memory, language, attention, and executive function (Itakura et al., 2017). The neural correlates of the VFT involve a network of prefrontal regions, encompassing Broca’s area, the bilateral dorsolateral prefrontal cortex, and other prefrontal areas (Da et al., 2024).

This study aims to elucidate the neural profiles underlying distinct subtypes of insomnia. We hypothesize that compared to patients with SID, those with CID display reduced cortical excitability and brain function. This neural pattern is proposed to associate with a decline in prefrontal inhibitory control, which in turn contributes to their symptom of difficulty inhibiting wakefulness. To test this hypothesis, we will employ TMS to assess cortical excitability and use fNIRS to characterize brain activation and functional connectivity during a VFT. By analyzing the relationships between these neural alterations, our findings seek to provide a foundation for targeted neuromodulatory interventions for insomnia.

This study employed TMS to assess cortical excitability and fNIRS to measure hemodynamic responses during the VFT in patients with SID and CID, exploring the association between these neural metrics. To the best of our knowledge, it represents the first multimodal investigation to combine fNIRS signals with TMS-derived electrophysiological biomarkers in a population with insomnia disorder.

Research has consistently posited insomnia as a condition of central nervous system hyperarousal (Riemann et al., 2010). Evidence for this model includes PET studies demonstrating elevated whole-brain metabolism (Nofzinger et al., 2004) and event-related potential (ERP) research indicating heightened cortical arousal in individuals with insomnia during both wakefulness and sleep (Bastien et al., 2008; Yang and Lo, 2007). These findings suggest that insomnia may be characterized by altered cortical excitability. Subsequent investigations have directly probed this hypothesis. For instance, Huang et al. (2012) examined median nerve somatosensory evoked potentials (SEPs) in patients with primary insomnia (PI) and observed increased cortical excitability compared to healthy controls. In another study using TMS, van der Werf et al. (2010) found that while patients with CID exhibited a similar RMT to controls, they showed greater MEP recruitment. Building upon this work, our study reveals that patients with SID exhibit a significantly lower RMT in the left hemisphere than those with CID, indicating that cortical excitability is higher in SID.

The excitability of cortical neurons is governed by the balance between GABA-mediated inhibition and glutamate-mediated excitation (Isaacson and Scanziani, 2011). In the context of PI, previous research suggests that heightened excitability within the parietal cortex may be driven by a reduction in GABA coupled with a compensatory increase in glutamate (Winkelman et al., 2008). This model of neurotransmitter imbalance is further supported by studies of sleep deprivation in healthy individuals, which demonstrate a shift toward cortical hyperexcitability—an “activating” pattern believed to reflect changes in glutamatergic synaptic function (Kreuzer et al., 2011; Placidi et al., 2013). Given that the RMT serves as a global measure of corticospinal excitability and is dependent on glutamatergic activity (Paulus et al., 2008), it provides a valuable proxy for this neurochemical state. We posit that an imbalance in the ratio of excitatory to inhibitory neurotransmitters relates to insomnia. Furthermore, this imbalance may differ between its short-term and chronic manifestations. We emphasize that these interpretations are theoretical models proposed for future validation.

Our findings reveal that patients with CID exhibited significantly reduced Oxy-Hb concentrations in the bilateral DLPFC, left mPFC, and right TL during the VFT compared to those with SID. This aligns with a growing body of neuroimaging evidence implicating prefrontal dysfunction in CID patients. For instance, a recent fNIRS study by Zhou et al. (2024) also reported diminished prefrontal activation in CID patients relative to HCs. Further corroborating this, resting-state fMRI research by Dai et al. (2014) identified lower brain activity in the bilateral frontal lobes of insomnia patients. Structural deficits have also been observed; Joo et al. (2013) found significantly smaller grey matter volumes in multiple regions, including the bilateral DLPFC and medial frontal gyrus, in CID patients compared to HCs.

The prefrontal cortex (PFC) is a central brain structure responsible for integrating internal and external environmental information and is critical for maintaining the brain’s resting-state network. A key PFC subregion, the DLPFC, is particularly vital for higher-order cognitive processes, including emotional regulation, working memory, executive function, and language (Asgharian Asl and Vaghef, 2022). The functional integrity of the PFC is closely linked to sleep; for instance, an fMRI study by Altena et al. (2010) demonstrated reduced activation in the medial and inferior PFC of patients with CID during a VFT, providing objective evidence of PFC dysregulation in insomnia. Consequently, targeting PFC excitability has emerged as a promising therapeutic avenue. Clinical studies have shown that neuromodulation techniques like high-frequency repetitive transcranial magnetic stimulation (rTMS) and anodal transcranial direct current stimulation (tDCS), which enhance PFC excitability, can improve sleep outcomes (Zhou et al., 2020; Nardone et al., 2013). Therefore, normalizing PFC function is a significant therapeutic objective for insomnia patients, and assessing PFC activity may serve as a valuable biomarker for evaluating treatment efficacy.

This study observed no significant difference in the left TL activation between the two groups. This finding aligns with recent fNIRS research, which also reported no significant difference in brain activation or functional connectivity changes in the TL of patients with CID during a VFT compared to HCs (Zhou et al., 2024). One plausible interpretation is that patients with CID employ compensatory neural strategies; they may heighten left TL activity to counterbalance cognitive deficits, thereby sustaining normal levels of behavioral performance.

The VFT is a recognized measure of executive function, sensitive to impairments in the frontal and temporal cortices and linked to language processing (He et al., 2018; Xiang et al., 2021). Specifically, in the left hemisphere, this frontotemporal area includes Broca’s area, a region critical for language production (Amunts and Zilles, 2012). Building on this, recent research indicates that individuals with Major Depressive Disorder (MDD) exhibit reduced activation in the left frontal lobes alongside increased activation in the right frontal lobes compared to HCs (Yang J. et al., 2024). This observed hypoactivity in the left frontal lobe during VFT performance suggests a decline in its functional capacity. As a result, the heightened activity in the right hemisphere may represent a compensatory mechanism to counterbalance left-hemisphere deficiencies (Balconi et al., 2017). Interestingly, our findings indicate that during the task, the left mPFC was most activated in patients with SID, whereas the right mPFC was most activated in those with CID. While these hemispheric activation patterns in the frontal lobes align with established research about MDD, the underlying lateralization mechanisms specific to insomnia pathology require further elucidation.

Hemispheric asymmetry is considered a fundamental neural characteristic that reduces redundancy and enhances processing efficiency. This specialization facilitates parallel and flexible information processing, enabling organisms to adapt to complex neurocognitive demands (Hartwigsen et al., 2021). A classic manifestation of this division is left-hemisphere dominance for language and logical reasoning, in contrast to right-hemisphere dominance for visuospatial and emotional processing (Demaree et al., 2005; Hagoort and Indefrey, 2014; Chen et al., 2019; Goel, 2019). Converging evidence from EEG and fMRI studies indicates that depressive disorders are characterized by a disruption of this typical hemispheric activity, particularly in frontal and parietal regions. This disturbance is potentially linked to aberrant lateralization underlying cognitive and emotional functions (Grimm et al., 2008; Bruder et al., 2017).

The distinct activation profiles observed in the left mPFC and DLPFC between the two groups suggest their potential utility as neural markers for differentiating SID from CID. Consequently, interventions targeting these regions in the CID group may hold promise for alleviating insomnia symptoms and enhancing cognitive function. These findings underscore the need for more in-depth investigation into the clinical application of hemispheric asymmetry as a basis for diagnostic and therapeutic strategies in insomnia.

Notably, in individuals with SID, the RMT in the left M1 was negatively correlated with Oxy-Hb levels in the left mPFC during the VFT, suggesting that greater cortical excitability in the left hemisphere is linked to increased activation of the left mPFC during this cognitive task. In contrast, no such correlation was observed in the chronic CID group. Given that insomnia pathophysiology involves dysregulated arousal, these findings indicate that patients with SID are in a state of hyperarousal, marked by elevated cortical excitability and brain activation. Conversely, patients with CID show lower levels of both, suggesting a different physiological state. Specifically, SID appears to be triggered by hyperarousal, potentially driven by ongoing activation of the orexin system (Saper et al., 2010). The orexin system plays a fundamental role in stabilizing the sleep–wake cycle. According to the hypothalamic “Flip-Flop Switch” model proposed by Saper et al. (2005), sleep and wakefulness are two mutually exclusive, stable states that switch rapidly. This bistability arises from mutual inhibition between the wake-promoting monoaminergic systems and the sleep-promoting ventrolateral preoptic nucleus (VLPO). Orexin neurons stabilize wakefulness within this circuit by boosting monoaminergic activity, which in turn enhances cortical excitability and inhibits VLPO neurons (De Luca et al., 2022).

While CID seems to persist in association with a failure of prefrontal inhibitory control, all components of the arousal system send robust projections to the prefrontal cortex, with a particular dominance in the medial prefrontal region. From there, they project back downward to the basal forebrain, hypothalamus, and brainstem components of the arousal system (Aston-Jones and Cohen, 2005; Hurley et al., 1991). Within this circuit, reciprocal excitation rapidly enhances arousal, while mutual inhibition sustains a low-arousal state. However, a decline in the inhibitory function of the frontal cortex may also lead to an increased arousal state. Thus, chronic insomnia may be better characterized by an inability to inhibit wakefulness at sleep onset, reflecting a maladaptive dissociation between wake- and sleep-promoting systems (Buysse et al., 2011).

The proposed neural profiles may further explain the differential efficacy of pharmacotherapy between patient groups, with drugs typically being more effective for individuals with SID than for those with the CID. The most effective conventional insomnia medications, Benzodiazepine receptor agonists, function by enhancing GABAergic activity (Gottesmann, 2002). As the primary inhibitory neurotransmitter, GABA plays a fundamental role in regulating sleep–wake cycles (Siegel, 2004), and a deficit in GABAergic inhibition is strongly implicated in the hyperarousal state of insomnia (Nofzinger et al., 2004). Beyond conventional drug classes, the most recent development in insomnia treatment involves dual (DORAs) and selective (SORAs) orexin receptor antagonists. Research indicates that antagonism of the orexin receptor 2 (OX2R), which is critical for maintaining arousal, promotes sleep in a dose-dependent manner (Mogavero et al., 2023). Specifically, DORAs augment total sleep time mainly by promoting REM sleep, without affecting or even suppressing non-REM sleep (Clark et al., 2020).

Insomnia is increasingly recognized as a disorder of aberrant brain network connectivity (Nardone et al., 2020). For instance, resting-state fMRI studies using the mPFC as a seed point have identified weakened functional connectivity (FC) with the bilateral temporal lobes and left parietal lobe in insomnia patients. Dysfunctional connectivity in these regions can adversely affect the mood, memory, and cognitive function of patients with insomnia (Spiegelhalder et al., 2015). It is particularly evident in the DLPFC, where disrupted connectivity serves as a marker of cognitive decline (Vartanian et al., 2014). Recent fNIRS research has further linked specific connectivity deficits to clinical severity, showing that decreased connectivity between the left PFC and right DLPFC correlates with higher PSQI scores and poorer sleep quality in patients with SID (Wu et al., 2023). Our findings build upon this foundation by revealing a key distinction between subtypes of insomnia. We observed that patients with CID exhibited significantly lower average FC than the SID group, a difference most pronounced in pathways from the right DLPFC to the left DLPFC and left mPFC. Within the SID group, we identified a specific neurophysiological signature; diminished FC between the left DLPFC and TL, as well as between the left TL and mPFC, was correlated with elevated cortical excitability (as measured by RMT). Collectively, these results delineate distinct neural profiles for different insomnia subtypes. The SID is underpinned by reduced integrity of a left-hemisphere frontotemporal network, which is associated with heightened cortical excitability. In contrast, the progression to CID is characterized by a broader decline in functional connectivity, particularly within the bilateral prefrontal cortex.

Intra- and inter-hemispheric connectivity are thought to subserve complementary functions. Intra-hemispheric connectivity reflects functional specialization and corpus callosum inhibition, underpinning processes such as language, reasoning, and attention. In contrast, inter-hemispheric connectivity facilitates signal transmission and information integration between the hemispheres, as observed in motor control and coarse spatial processing (Gazzaniga, 2000; Wan et al., 2022). Mediated by the corpus callosum, these two patterns correspond to specialized and integrated information processing, respectively (Hartwigsen et al., 2021). We posit that the in patients with SID, a functional discoordination within the left frontotemporal region during task performance may link to diminished intra-hemispheric connectivity, which in turn triggers compensatory activation. At this stage, the division of labor between the hemispheres, mediated by corpus callosum inhibition, remains relatively intact. In contrast, patients with CID appear to experience a loss of this specialized left-hemisphere dominance. A concomitant reduction in the inhibitory function of the corpus callosum allows for greater reliance on the right hemisphere to complete the task, resulting in the observed decrease in interhemispheric connectivity. The alterations in both intra- and inter-hemispheric connectivity observed in our study may therefore reflect a pathological reconfiguration of this balance in insomnia, a compelling premise for future research. This finding is particularly significant given that similar hemispheric alterations are observed in MDD, where they predict depressive traits (Yang Y. et al., 2024). Therefore, our work provides evidence of aberrant functional hemispheric asymmetry in insomnia, advancing the pathophysiological understanding of this disorder from a novel, systems-level perspective of hemispheric lateralization.

Investigating the neural mechanisms of distinct insomnia subtypes enables a more precise, pathophysiology-driven approach to treatment. Unlike traditional pharmacotherapy, which broadly inhibits neural excitability, TMS offers the capacity for direct and bidirectional modulation of both cortical excitability and functional connectivity. Current TMS protocols for insomnia typically involve low-frequency stimulation of the right DLPFC to reduce hyperexcitability (Herrero Babiloni et al., 2021). However, our findings suggest that for patients with CID, a combined protocol incorporating high-frequency stimulation to the left DLPFC may yield superior outcomes by restoring prefrontal functionality. This hypothesis warrants rigorous validation in future clinical trials.

This study is subject to several limitations. Firstly, the absence of a healthy control group hinders a definitive attribution of the observed neural changes specifically to the duration of insomnia and prevents the establishment of an absolute pathological deviation from a normative baseline. Secondly, future studies should incorporate neuronavigated, bilateral RMT measurements to investigate potential interhemispheric excitability differences with greater precision. Thirdly, the spatial coverage of our fNIRS system was limited; future investigations should target whole brain regions, especially the motor cortex. Fourthly, the reliance on subjective questionnaires for participant selection introduces potential bias, as it excludes objective measures like polysomnography. Finally, given the established interplay between insomnia and emotional states, incorporating standardized anxiety and depression assessments in future research would enable a more comprehensive evaluation.

Our results reveal that patients with CID demonstrate lower cortical excitability, decreased brain activity, and reduced task-related functional connectivity compared to the SID group. These findings suggest that suppressed blood oxygen metabolism in the cerebral cortex may be related to the pathogenesis of CID. Notably, in SID patients, higher cortical excitability was significantly correlated with greater activation of the left mPFC. Furthermore, cortical excitability in SID patients was significantly correlated with the degree of disrupted connectivity strength between the left frontal and temporal lobes. Consequently, future research should employ fNIRS and TMS to support clinical screening and optimize therapeutic interventions for insomnia.