Patients diagnosed as SIVD were enrolled in this clinical trial. Inclusion criteria were as follows: (1) age 50-80 years old; (2) complaint of cognitive impairment lasting for at least 3 months; and (3) vascular dementia diagnosis, according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, with a Mini-Mental State Examination(MMSE) score 10-26; Montreal Cognitive Assessment(MOCA) score <26; and Clinical Dementia Rating(CDR) score 1-2. All patients meeting the clinical criteria underwent brain MR imaging (MRI). The MRI inclusion criteria were as follows: (1) moderate to severe white matter lesions (score ≥2, according to the Fazekas rating scale); or multiple (≥3) small supratentorial subcortical infarcts (3-20 mm in diameter); or small infarcts strategically located in the caudate nucleus, globus pallidus, or thalamus; (2) absence of hemorrhages, cortical and watershed infarcts, hydrocephalus, and white matter lesions from specific causes (e.g, multiple sclerosis). Exclusion criteria included the patients cannot complete neuropsychological testing (e.g., severe aphasia, physical disabilities), or experienced new strokes within 3 months before enrollment; small vessel disease due to inheritance or inflammation; schizophrenia or a score >17 on the Hamilton Depression Scale (HAMD); cancer; clinically significant systemic diseases (e.g., cardiovascular, respiratory); peripheral vascular disease; use of donepezil and memantine that may affect cognitive functioning; Refusal to sign informed consent.
RIC and Control Treatment Procedures
Participants were randomly divided into two groups according to a random number. The experimental group underwent five brief cycles of RIC (bilateral upper limb compression at 200 mmHg) for 5 minutes followed by reperfusion for another 5 minutes, which performed twice daily over 6 consecutive months. The control group underwent the same process, but at a pressure of 60 mmHg which caused no restriction on the blood flow of the upper limb. RIC was carried out by an electric inflation auto-control device (patent number ZL200820123637.X, RenQiao IPC-906D, China), which is similar to the blood pressure measurement. Participants could abort the RIC treatment at any time if they did not feel well.
Special personnel unassociated with the study were responsible for randomization and allocation of RIC instruments. Researcher personnel and all participants were blinded to treatment assignment.
During the entire 6-month study, both groups of patients continued taking their standard medications, including anti-platelet, anti-hypertensive, anti-diabetic, anti-homocysteine, and lipid control agents.
Treatment Compliance Guarantee
The data of each treatment were recorded by the RIC device and sent via the internet to our researchers’home computer in real time. The investigators scanned the participants’ therapy compliance routinely. Only when the time and frequency per day achieved the standard, the treatment could be qualified. If abnormal conditions occured in the course of the treatment, the researcher would contact the patient or family members in time. If necessary, the patients could complete the treatment with the assistance of family members. Participants who did not complete the treatment for seven consecutive days were excluded.
The primary outcome measures for assessing RIC efficacy in improving cognition were neuropsychological assessments. These assessments included five domains: memory, language, attention, executive function, and orientation. Immediate memory, delayed memory, and recognition memory were tested with the Hopkins Verbal Learning Test-Revised (HVLT-R) . Language usage and category fluency were tested with the Controlled Oral Word Association Test (COWAT) . Attention and executive function were tested with the Trail Making Test A and B (TMT-A and TMT-B) , and the Symbol Digit Modalities Test (SDMT) . Visuospatial processing was examined with the Judgment of Line Orientation (JLO) . Cognitive tests were administered at baseline and 6 months later.
Measurements of Inflammation
High-sensitive C-reactive protein (hs-CRP) was measured in plasma by using commercially available turbidimetric immunoassay kits (MedicalSystem Biotechnology Co., Ltd., Ningbo, China), following the manufacturer’s instructions.
MRI examinations of the head were done using a 3.0T whole body system (Discovery MR750; General Electric, Milwaukee, WI, USA). Cube FLAIR (156 slices; repetition time (TR), 6000 ms; echo time (TE), 144 ms; echo train length, 200; slice thickness, 1.2 mm; in-plane resolution, 1 mm2) was used for lesion detection. High-resolution sagittal three-dimensional (3D) T1-weighted images were acquired using a brain volume (BRAVO) sequence (156 slices; TR, 8.14 ms; TE, 3.17 ms; inversion time (TI), 450 ms; flip angle, 12°; slice thickness, 1.2 mm; in-plane resolution, 1 mm2). The high-resolution images were used to calculate brain volume. Also, two-dimensional (2D) echo-planar diffusion tensor images (DTI) were acquired (48 slices; TR, 5000 ms; TE, 60.6 ms; slice thickness, 3 mm; in-plane resolution, 2 mm2; 50 non-collinear diffusion gradients [b = 1000 s/mm2]). Three non-diffusion-weighted images (b = 0 s/mm2) were used for measuring white matter tract integrity.
Analysis of White Matter Lesions
Demarcations of interest regions and measurement of white matter lesion volume (WMLV) on cube FLAIR images were performed manually by using MRIcro software (http://www.mccauslandcenter.sc.edu/mricro/mricro/mricro.html).
We extracted the mean diffusion indices of the Whole-Brain White Matter (WBWM), White Matter Lesion (WML) and Normal-Appearing White Matter (NAWM) (WBWM minus lesion regions). DTI data were first processed along the following pipeline using the FMRIB Software Library (FSL) 5.0: (1) Eddy current correction: this step applied an affine transformation on the raw diffusion data to correct for image distortion caused by eddy current, and it corrected for misalignment between volumes caused by head motion. (2) Brain extract: voxels outside brain tissue were filtered out using the brain extract toolbox (BET) in FSL. Then, a linear least-squares fitting algorithm was carried out to fit the tensor, and the three eigenvalues, Mean Diffusivity (MD), and Fractional Anisotropy (FA) were calculated from the tensor.
Brain lesions were manually segmented from the 3D T2 FLAIR images and saved as a binary mask for each subject. To separate the three different types of diffusion indices (WBWM, WML, and NAWM), we first segmented the 3D T1 weighted images (T1-WI) into gray matter tissue, white matter tissue, and cerebrospinal fluid using a unified segmentation method carried out in SPM8 (http://www.fil.ion.ucl.ac.uk/spm/software/spm8/). Then the white matter tissue of each subject was binarized to create a whole white matter mask using a threshold of 50%. The T1WI was then affinely co-registered with the b0 images (diffusion images without exerting a gradient field). T2 FLAIR images were rigidly co-registered with the T1 images, and then were transformed into the diffusion space along with the lesion mask using the deformation parameters between T1 and b0 images. The normal white matter mask for each subject was calculated by subtracting the whole white matter mask from the white matter lesion mask. Finally, the mean diffusion metrics (FA and MD) of each mask of each subject were calculated by averaging the values of all voxels with this mask.
Baseline data homogeneity between the two groups was analyzed with two independent samples t-test, X2 test, Fisher’s exact test and multiple linear regression analyses. Data from the pre-treatment and post-treatment periods were analyzed by paired t-test to evaluate the effect of the treatments in the group. Due to the small sample size, covariance analysis was used to compare the effect of the two groups.
All analyses were done with SPSS (IBM SPSS Statistics for Windows, Version 22.0). All hypothesis tests were two-tailed, and P values < 0.05 were considered significant.