2.1 Participants
This prospective study was designed as a double-blind, randomized, placebo-controlled trial and was approved by the Institutional Medical Ethics Committee of No. 1 Affiliated Hospital, Wenzhou Medical University (Approval document #2016045). Heart-kidney imbalance insomnia patients (HKIIPs) were recruited from The First Affiliated Hospital of Wenzhou Medical University from September 2018 to August 2019 through WeChat, flyers and bulletin boards at the hospital. All HKIIPs were asked to complete the magnetic resonance imaging (MRI) examination, polysomnography (PSG) and Pittsburgh Sleep Quality Index (PSQI) before and after the medication intervention.
The inclusion criteria for HKIIPs were as follows: (a) Aged 18- to 60- years-old, right-handed, male and female, in junior high school or above; (b) According to the DSM-V, insomnia was defined as having difficulty falling asleep, staying asleep or having refreshing sleep, occurring 3 or more times per week, for at least 3 months [28]; (c) A PSQI score of greater than or equal to 7 according to the criteria for insomnia described in CCMD-3 [29]; (d) A Disharmony of Heart and Kidney Scoring System score is greater than 9 [29]; (e)Free of any psychoactive medication or hypnotic or cognitive behavioral therapy for insomnia for at least 2 weeks before the study.
Major exclusion criteria are listed as follows: (a) The presence of any abnormal brain MRI findings; (b) Insomnia caused by changes in lifestyle or environmental factors; (c) A history of serious mental disorders (e.g., anxiety, depression or schizophrenia); (d) Suffering from a physical disorder that affects the central nervous system; (e) Having liver and kidney dysfunction; (f) Alcoholism and/or psychotropic drug addiction; (g) Pregnant, breastfeeding or menstruating women.
The brief flow diagram of the study process is shown in Fig. 1. A total of 80 HKIIPs who passed preliminary screening criteria were recruited to this study. Among these, 2 patients were excluded due to abnormal intracranial occupying signals in their MR images, as diagnosed by two physicians with over 20 years experiences in neuroradiology. In addition, 3 patients were excluded due to severe mental illness. Seventy-five eligible HKIIPs (26 males, 49 females) were included in the present study. HKIIPs were randomly divided into two groups: the JTW group (n = 36) and the placebo group (n = 39). A total of 25 patients were excluded based on the following criteria: no DTI data after treatment (JTW group = 12; placebo group = 11); lack of post-treatment PSQI and PSG data (JTW group = 0; placebo group = 2). Eventually, we obtained a sample of 50 HKIIPs (JTW group = 24, placebo group = 26; 19 men, 31 women).
Forty-one age- and sex-matched healthy controls (HCs; 16 males, 25 females) were also recruited from the local community through advertisements as mentioned above. The inclusion criteria for HC participants were as follows : good sleep quality (PSQI ≤ 5); no history of psychiatric or neurologic diseases; and normal conventional MRI findings.
All participants for examination and/or treatment were asked to sign an informed consent form.
2.2 Intervention
A double-blind placebo-controlled drug administration technique was used. Both experimental drugs (JTW) and matched placebos were made by the Kang-ren Pharmaceutical Factory (Shanghai China). JTW were produced as yellowish-brown granules, containing 1.1 g JTW soft extract consisted of Cortex Cinnamomi and Rhizome Coptidis at a ratio of 1:10. The placebos (containing only the excipients, i.e., corn starch, citric acid, lactose hydrate and caramel color) with the same appearance and color as JTW were packaged in identical plastic bags to ensure blinding. HKIIPs ingested 2 g JTW or placebo at 4 pm and 9 pm daily for a week according to groups, and followed the scheduled examination time at the termination of the medication intervention period.
2.3 Pittsburgh Sleep Quality Index
The Pittsburgh Sleep Quality Index (PSQI) used in this study is a retrospective self-rated questionnaire consisting of 19 items. [30] The PSQI was used in this study at the beginning and end of the medicine intervention. The scores obtained at the beginning were recorded as score 1, and the scores obtained at the end were recorded as score 2. A higher PSQI score means worse quality of sleep. Clinical efficacy was calculated according to the following formula: Clinical efficacy = score 2 – score 1.
2.4 Polysomnography
Polysomnography (PSG) was performed at baseline and 1 week later at the end of the medication intervention. All HKIIPs underwent full, in-laboratory PSG by study-certified technicians according to the guidelines of American Academy of Sleep Medicine. Changes in sleep quality were objectively measured by mean changes in all parameters, including total sleep time (TST), sleep efficiency (SE), sleep onset latency (SL), rapid eye movement latency (RL), awakening times (AT), arousal index (AI), rapid eye movement (REM) and 3 parts of non-rapid eye movement (N1,N2 and N3), from baseline to the end of the medication intervention. Clinical effects were assessed by the following formula: Clinical efficacy = parameters after treatment – parameters before treatment. During the recording period, all participants were asked to abstain from alcohol, caffeine and daytime naps.
2.5 MRI
Each participant signed a written consent form before undergoing the MRI examination. All MRI scans were carried out using a 3 T MRI scanner (Philips Achieva TX) with an 8-channel receive-only head coil. Additionally, in order to reduce the effect of scanner noise and head motion, all subjects laid in a supine position with ear plugs and foam pads. The DTI datasets were acquired along 16 gradient directions (b = 1000s/mm2), including five acquisitions without diffusion weighting (b = 0). The sequence parameters were as follows: repetition time (msec) / echo time (msec), 6800 / 93; 128 × 128 matrix; field of view, 256 mm × 256 mm; slice thickness = 3 mm, no gap; and 50 contiguous axial slices. Several other sequences were also scanned, such as T2-weighted images, T1-weighted images, T1- fluid attenuated inversion recovery (FLAIR) images, T2-FLAIR images, and all scans were inspected for motion artefacts and for the absence of pathologic findings by a neuroradiologist.
2.6 Data analysis
Data preprocessing. —DTI data were preprocessed using tools in FSL (FMRIB Software Library 5.0.9) as follows. Firstly, all DTI images were corrected for distortions caused by eddy current distortions and head motion by using an affine alignment of each diffusion-weighted image to the non-diffusion weighted (b = 0) image by the FMRIB’s Diffusion Toolbox. Secondly, FSL's Brain Extraction Tool (BET) was applied to generate a binary mask from the b0 image and remove any non-brain tissue (fractional intensity threshold = 0.2)[31]. 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 [32], which was used to characterize the microstructure with multiple diffusion measures better than that with FA measures alone. 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 thresholded (FA > 0.2) to generate a mean FA skeleton 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.
2.7 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 two-sample, two-tailed t tests.
A white-matter statistical analysis (covarying for age and sex) was performed, as described in the general linear model (GLM, https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/GLM) and permutation analysis of linear models (PALM, https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/PALM). To correct for multiple-comparisons, a threshold-free cluster enhancement with randomized (5000 permutations) nonparametric testing protocol was used to control for false discovery rates < 0.05 (family-wise error correction, FWE or FDR-correction) using PALM. In addition, we used a cluster-correction to assess possible correlations, using p values < 0.001 and a minimum cluster size of 50 voxels. The Johns Hopkins University (JHU) ICBM-DTI81 White Mater Labels and JHU white-matter tractography atlas were used to locate the specific fibers.