Introduction

Insomnia is one of the most prevalent subjective complaints of sleep-disordered patients worldwide, and it is reported that approximately 4–20% of the population suffers from insomnia (Morin et al. 2011; D. Riemann et al. 2011; Roth et al. 2011). The main characteristic of insomnia is that people with insomnia usually have chronic difficulty falling asleep, difficulty staying sleep, and/or early morning awakening. Insomnia is associated with an increased risk of psychiatric disorders and the quality of an affected patient’s life is severely reduced (Baglioni et al. 2011; Kyle et al. 2010; Morin et al. 2015). Moreover, chronic insomnia is a predisposing factor for type II diabetes, metabolic syndrome and cardio-cerebrovascular diseases (Knutson et al. 2007; M. Li et al. 2014). Although numerous studies have focused on insomnia, the mechanisms underlying the occurrence of insomnia remain poorly understood. Insomnia is often accompanied by other somatic symptoms. These signs and symptoms are diverse and reflect the malfunctioning or imbalance of various body systems, which could be either causes or consequences of insomnia. According to the signs and symptoms of patients, traditional Chinese medicine divides insomnia into many different patterns. The heart-kidney imbalance is one of the main patterns, often accompanied by red tongue, thin coating, fine and rapid pulse, dizziness, tinnitus, seminal emission, night sweating and so on. (Poon et al. 2012).

Current guidelines set cognitive behavioral therapy for insomnia as the preferred treatment for patients who suffer from insomnia (Buysse. 2013; Schutte-Rodin et al. 2008). However, poor adherence may diminish its therapeutic effect (Matthews et al. 2013). Many people cannot address their insomnia without taking a combination of medications. Besides, most therapies only transiently alleviate the symptoms of insomnia instead of delivering a complete cure. The most widely used medications for insomnia are benzodiazepines and benzodiazepine receptor agonists (Buysse. 2013). Still, these therapies incur numerous side effects, including depression, poor memory or forgetfulness, drowsiness and impaired work productivity. In addition, their use is heavily limited by developed tolerance and increased dependency with long-term treatment (Buysse. 2013; Dieter Riemann et al. 2015). So far, guidance for clinicians in choosing the best treatment remains limited, and the search for better treatments is underway. In recent years, increasing evidence has shown that traditional Chinese medicine can be effectively applied in the treatment of insomnia (Singh and Zhao, 2017; Yeung et al. 2012; Zhang et al. 2019). It was reported that Jiao-tai-wan (JTW), made of Rhizome Coptidis and Cortex Cinnamomi at a ratio of 10:1, has a remarkable effect in treating of insomnia linked to an incompatibility between heart and kidney (Huang et al. 2020; Sun et al. 2020; Yue et al. 2016). Experimental studies have shown that JTW can not only regulate levels of neurotransmitters in the blood, such as orexin A (brain tissue), 5-hydroxy tryptamine and γ-aminobutyric acid, which are all therapeutic targets for insomnia (Abad and Guilleminault, 2018; Huang et al. 2020), but it may also regulate immune cytokines, such as interleukin-6 and tumor necrosis factor, which are upregulated by sleep loss (Zou et al. 2017).

Diffusion-tensor imaging (DTI) is a non-invasive method which is widely used to assess alterations in white matter integrity. The decrease of fractional anisotropy (FA) value means the destruction of white matter integrity. Although several studies have investigated white matter structures in people with insomnia, the results of these studies remained inconsistent. For instance, Spiegelhalder et al. (Spiegelhalder et al. 2014) reported a reduced white matter integrity of bilateral anterior internal capsule in primary insomnia, while Bresser et al. (Bresser et al. 2020) only found decreased FA values in the right limb of the anterior internal capsule. Kang et al. (Kang et al. 2018) showed low white-matter integrity between the inferior frontal gyrus and left thalamus in insomnia patients. Li et al. (S. Li et al. 2002). Finally, a tensor model was fitted locally to each voxel using DTIfit (FSL) and maps of FA and MD were calculated.

Tract-based spatial statistics —The voxel-wise statistical analysis of DTI data was based on tract-based spatial statistics (TBSS) from the FMRIB Software Library (S. M. Smith et al. 2006). The FA images from all subjects (n = 116) were aligned to a common target FA image (1 × 1 × 1 mm3 MNI152 FMRIB58_FA standard space) using a non-linear registration. Next, a mean FA image was created and a mean FA skeleton map thresholded to the standard value of 0.2 was generated, representing the centers of all tracts. Subsequently, each participant’s aligned FA images were projected onto the mean FA skeleton for statistical analysis. For better assessment, a TBSS analysis was repeated for MD maps.

Statistical analysis

Differences in age, PSQI and objective sleep measures on polysomnography between HKIIPs and HCs, JTW and placebo group were analyzed by two-sample, two-tailed t tests. Differences in the proportion of females and males between above groups were determined by a two- tailed Pearson χ2 test. Differences in clinical efficacy of medication between JTW and placebo group were assessed by ANCOVA.

A white-matter statistical analysis (covarying for age and sex) was firstly conducted using a general linear model in the FSL randomize tool. Then, 5000 permutation analysis of linear models were carried out with MATLAB software. Finally, threshold-free cluster enhancement (TFCE) was applied to define significant clusters (p < 0.05) and correct multiple comparisons for family-wise error (FWE). Cluster-correction (CC, p < 0.01 and a minimum cluster size of 50 voxels) was used, when FWE correction showed no significant clusters. The Johns Hopkins University (JHU) ICBM-DTI81 White Mater Labels and JHU white-matter tractography atlas were used to locate the specific fibers.

Results

Demographic characteristics

Table 1 presents the characteristics of the participants in this study. The differences between HC group and HKIIPs group were not significantly different for age (p = 0.641) or sex distribution (p = 0.816). Furthermore, HKIIPs had higher PSQI scores than HC participants (12.5±3.0 versus. 2.9±1.4; p < 0.001).

Table 1 Demographics and clinical characteristics of all Participants
Full size table

Whole brain white matter difference

Comparison between HCs and HKIIPs

As shown in Fig. 2 and Table 2, the tract-based spatial statistics analysis demonstrated that when compared with HCs, HKIIPs showed significantly higher FA values in the left corticospinal tract (CST) (Fig. 2A, FWE, p < 0.05 ), and significantly lower FA values in several other white matter tracts, including the bilateral superior longitudinal fasciculus (SLF), bilateral anterior limb of internal capsule (ALIC), bilateral anterior and posterior corona radiata (ACR and PCR), bilateral body of corpus callosum (BCC), left cingulum, left posterior thalamic radiation (PTR), right superior corona radiata (SCR), and right CST (Fig. 2B, FWE, p < 0.05). In addition, the tract-based spatial statistics analysis showed lower MD values in HKIIPs at several fibers, such as bilateral SLF, bilateral PCR, bilateral cingulum, bilateral cerebral peduncle, bilateral CST, bilateral ALIC, bilateral posterior limb of internal capsule, bilateral PTR, bilateral SCR, left ACR, and left BCC (Fig. 2C, FWE, p < 0.05).

Fig. 2
figure 2

Tract-based spatial statistics analysis showing increased or decreased fractional anisotropy (FA) and mean diffusivity (MD) values in different white matter (WM) tracts of heart-kidney imbalance insomnia patients (HKIIPs). Green represents the mean FA or MD skeleton across all participants. Red yellow depicts the WM tracts whose FA or MD values were significantly changed (family-wise error correction, FWE, p < 0 .05 or cluster-correction, CC, p < 0.01 and cluster size > 50 voxels). (A) WM tracts with increased FA values in HKIIPs compared with health control (HC) (FWE, p < 0 .05); (B) WM tracts with decreased FA values in HKIIPs compared with HC (FWE, p < 0 .05); (C) WM tracts with decreased MD values in HKIIPs compared with HC (FWE, p < 0 .05); (D) WM tracts with decreased FA values in JTW group compared with placebo group (CC, p < 0.01 and cluster size > 50 voxels).

Full size image
Table 2 Results of white matter analysis
Full size table

Clinical efficacy after medication

The characteristics of the JTW group and the placebo group are shown in Table 1. There were no significant differences in gender, age, PSQI score or PSG between the two groups at baseline. PSQI and PSG regulated by ANCOVA were showed in Table 3. PSQI score of HKIIPs decreased significantly after treatment with JTW, but no changes were observed for PSG. Additionally, in Fig. 2D, HKIIPs showed decreased FA values in the left CST after JTW treatment, while the white matter structure of the placebo group remained unchanged (CC, p < 0.01 and cluster size > 50 voxels).

Table 3 Differences in clinical efficacy of medication
Full size table

Discussion

Previous studies mainly focused on the changes of white matter in patients with primary insomnia. In this study, we assessed the differences in white matter integrity between HKIIPs and HCs, in addition to the curative effect of JTW on white matter integrity in HKIIPs.

Analysis of white matter changes in HKIIPs

Our findings showed increased FA values in the left CST of HKIIPs compared with HCs. We speculate that this may be associated with increased cortical excitability. The hyperarousal theory of insomnia assumes that increased cognitive and physiological arousal leads to difficulty in initiating or maintaining sleep (Perlis et al. 1997). A study of cortical excitability indicated that chronic insomnia patients showed a globally increased excitability, with larger motor evoked potential sizes to stimulation compared with control participants (van der Werf et al. 2010). In addition, it is reported that a lower corticospinal excitability is related to a smaller CST volume (Lepley et al. 2020). Considering that the CST is of paramount importance in the somatic motor system and increased FA values can be attributable to improved axon density and fiber coherence, we boldly speculated that the increased FA of CST in patients with insomnia was associated with increased cortical excitability. Autonomic, neuroendocrine, electrophysiological, neuroimmunological and neuroimaging findings confirm increased arousal levels in patients with primary insomnia (Bonnet and Arand, 2010; D. Riemann et al. 2010). Insomniacs showed nocturnal elevations of norepinephrine (M. Irwin et al. 2003). Norepinephrine not only has neuroprotective and anti-inflammatory effects, but also can promote the differentiation of oligodendrocytes, and then promote the formation of myelin (Galea et al. 2003; Ghiani et al. 1999; Madrigal et al. 2007). With the formation of myelin, the FA value of neural fibers may increase (Beaulieu. 2002). Conversely, our study also found that the FA and MD values of a large number of fiber bundles in bilateral brain were decreased in HKIIPs. One possible explanation for our findings is the presence of insomnia-associated neuroinflammation involving white matter. A meta-analysis showed that insomnia is associated with increased systemic inflammation markers (M. R. Irwin et al. 2016). In addition, the proinflammatory interleukin IL-6 was highly expressed in the brain of sleep-disordered mice (Zhu et al. 2012). Higher levels of circulating inflammatory markers are correlated with lower FA values in the brain (Rodrigue et al. 2019). Neuroinflammation and low FA values have also been reported in mouse models of craniocerebral trauma (Missault et al. 2019). Thus, we believe that the decrease in FA and MD in the white matter of HKIIPs may be due to neuroinflammation. It is worth noting that the FA values of bilateral CST in HKIIPs showed inconsistent trends, whereby the FA value of right CST was decreased, while the left showed the opposite pattern. This may be due to the fact that the amount by which the FA value increased, due to the excitability in the dominant hemisphere, was much greater than the amount by which it decreased due to neuroinflammation.

The therapeutic effect of JTW

We found that the FA value in the left CST was decreased in HKIIPs after JTW treatment. This may be attributed to the central inhibitory effect of JTW leading to decreased cortical excitability. JTW antagonizes central excitability by regulating the levels of various neurotransmitters and hormones associated with insomnia (Sun et al. 2020). It plays a sedative and hypnotic role by increasing gamma-aminobutyric acid content and receptor expression in the brain of rats. It also increases the level of 5-hydroxytryptamine and decreases the level of norepinephrine in the brain by inhibiting the hypothalamic–pituitary–adrenal axis (HPA) of insomnia rats, thus exerting its therapeutic effect on insomnia (Huang et al. 2020). Our findings may serve to elucidate the underlying therapeutic mechanism of JTW.

PSQI and PSG results

The PSQI score of HKIIPs decreased significantly after treatment with JTW, reflecting the effectiveness of JTW in treating insomnia. The PSQI score was used to assess sleep quality and severity of sleep disorders in the past one month, while the PSG was mainly used to identify the sleep phases, such as REM, NREM, and wakefulness (Chen et al. 2018; Zeng et al. 2020). The PSG results showed no significant difference between the JTW and placebo groups after treatment, which could be due to the little effect of JTW on the sleep structure, or short-term treatment that has no effect on the structure of sleep. Better results might be got if giving a longer course of treatment.

Limitations

This study has several limitations. Firstly, the sample size for the HKIIPs was too small. Secondly, the duration of treatment was not long enough, which may lead to no difference in PSG changes between JTW and placebo after treatment.

Conclusion

Our study demonstrates that JTW treatment for one week did not affect objective sleep measures while the effects on subjective measures deserve further research. The FA value of multiple white matter in the whole brain of the HKIIPs decreased, while the FA value of the left CST increased. After one week of JTW treatment, the increased FA value of left CST decreased.