This study was conducted as a component of the “Comprehensive Prevention and Treatment of Key Diseases of the elderly in Zhongshan City” initiative, utilizing a random sampling methodology. Participants were selected from five community centers in the Zhongshan Torch Development Zone between May and October 2023. Of the 497 elderly individuals aged 60 years or older who underwent mammography, 92 were identified as having MCI. Ultimately, the study included a total sample size of 52 individuals with MCI and 55 healthy elderly controls (CN). The subjects with MCI were selected by applying Paterson’s diagnostic criteria of MCI (15, 16): Memory loss within the last year; Age 60 years and above; Right-handed; Normal hearing and eyesight; A Mini Mental Status Examination Scale (MMSE) score ≤ 17 points (illiterate), ≤20 points (education level of primary school), or ≤ 24 points (education level of secondary school or above) respectively; A Activities of daily living (ADL) score ≥ 60 points; physician-confirmed diagnosis of not AD; Signature of informed consent. The inclusion criteria for healthy controls were as follows: Age 60 years and above; Right-handed; Normal hearing and eyesight; MMSE score >17 points (illiterate), >20 points (education level of primary school), or >24 points (education level of secondary school or above) respectively; ADL score ≥ 60 points; physician-confirmed diagnosis of not AD; Signature of informed consent. Exclusion criteria are as follows: a current or history of alcohol/drug abuse; central nervous system trauma; were taking medication which may affect cognitive function; had or previously had any health condition that may influence cognitive function. The study protocol was approved by the Medical Ethics Committee of Zhongshan third people’s hospital (approval number: SSYLL20220401). All participants were informed about the contents of the respective study and gave their written consent to participate.

The study utilized a self-designed questionnaire to gather demographic information from respondents, encompassing variables such as gender, age, education level, marital status, and area of specialization. Cognitive function was assessed using the MMSE (17), which evaluates various cognitive domains including orientation, registration, attention, calculation, language, and recall through a total of 30 items scored on a binary scale. ADL were evaluated to assess subjects’ ability to perform daily tasks, with a scale consisting of 10 questions related to daily activities (18), the higher the score the worse the self-care ability.

All participants fasted overnight (not less than 8 h), and blood was collected in EDTA-treated 5 mL vacutainer tubes and gently inverted and mixed. After collection, the specimens were placed in universal transport boxes, stored at 4°C, and transported to laboratory within 4 h of collection. Plasma and blood cell pellets were separated from EDTA-treated blood samples by centrifugation at 3000 rpm for 10 min at 4°C, stored at −80°C for later processing.

The TIANamp Genomic DNA kit was used to extract blood cell pellets sample DNA. The DNA concentration was detected by an ultra-micro spectrophotometer and confirmed DNA integrity through agarose gel electrophoresis.

The MassARRAY SNP genotyping platform (Agena Bioscience, San Diego, CA, United States) was employed for genotyping SNPs within the BIN1 and APOE genes. Primers were designed utilizing the Agena primer design tool¹. The polymerase chain reaction (PCR) primers were diluted to a concentration of 100 μM, and the PCR primer mixture was prepared in accordance with the 1EXT 200 protocol. The extension primer mixture was formulated using a 1:25 dilution of the primers. Following the finalization of the extension primer mixture configuration, 2 μL of each primer were further diluted at a 1:25 ratio for subsequent mass spectrometric analysis. The proportion of extension primers for each individual locus was adjusted based on the results of the tests. In preparation for PCR amplification, a PCR master mix was formulated in a 1.5-ml Eppendorf tube, achieving a total volume of 4.1 μL. This mix comprised 1.850 μL of water, 0.625 μL of PCR buffer containing 15 mM MgCl2, 0.325 μL of additional MgCl2, 0.100 μL of deoxyribonucleoside triphosphate (dNTP) mix, 0.1000 μL of primer mix, and 0.200 μL of HotStarTaq polymerase (Qiagen, Hilden, Germany). Utilizing an 8- or 12-channel pipette, 4 μL of the master mix was dispensed into each well of a 384-well plate, followed by the addition of 1 μL of genomic DNA at a concentration of 20 ng/μl. The samples were subsequently mixed and subjected to centrifugation at 1,000 rpm for 1 min. To mitigate evaporation during the PCR process, the 384-well plates were sealed with a protective film. The PCR amplification protocol was executed as follows: an initial denaturation at 94°C for 5 min; 45 cycles consisting of denaturation at 94°C for 20 s, annealing at 56°C for 30 s, and extension at 72°C for 1 min; a final extension at 72°C for 3 min; and a hold at 4°C. The PCR products were subjected to treatment with shrimp alkaline phosphatase (SAP) to eliminate free dNTPs from the reaction mixture. A reaction mixture consisting of 1.53 μL of water, 0.17 μL of SAP buffer, and 0.3 μL of SAP was prepared in a 1.5-ml Eppendorf tube. Subsequently, 2 μL of this reaction mixture was added to each well of a 384-well PCR plate containing 5 μL of the PCR product. The SAP reaction was conducted in a PCR machine under the following conditions: 37°C for 20 min, followed by 85°C for 5 min, and a final hold at 4°C. Following the alkaline phosphatase treatment, a single-base extension reaction was performed. After the PCR process, Na+, Mg2+, K+, and other salt ions were removed using cation exchange resin to minimize the occurrence of excessive peaks in the mass spectrometry analysis spectrum, which could compromise the accuracy of the results. A microvolume of the purified sample was subsequently loaded onto the SpectroCHIP using a MassARRAY nanodispenser to facilitate the preparation of a co-crystalline film of the chip matrix and the sample.

MRI plays a crucial role in both research on MCI and in clinical studies, particularly in the assessment of significant hippocampal abnormalities through structural MRI (19). The MRI data utilized in this study were obtained using a 3.0 Tesla Vida Siemens MRI Scanner (Siemens, Erlangen, Germany) equipped with a 64-channel head coil. To minimize head movement and scanner noise, foam pads and headphones were employed during data collection. High-resolution structural images were acquired using a three-dimensional magnetization-prepared rapid gradient-echo (3D MPRAGE) T1-weighted sequence. The scanning parameters included a repetition time (TR) = 2000 ms, an echo time (TE) = 2.26 ms, an inversion time (Tl) =900 ms, a specific slice thickness = 1 mm, interslice gap = 0 mm, Voxel size = 1 × 1 × 1 mm3, number of slices =176, flip angle (FA) = 8°, field of view (FOV) =256mmx256mm, matrix =256×256 and length of the time course = 4.06 min. Structural MRI data were analyzed using voxel-based morphometry, with each scan being visually inspected by experienced neuroradiologists to identify and exclude any artifacts or gross anatomical abnormalities, The neuroradiologists were blinded to genetic information. Structural image preprocessing was conducted using the SPM12² and CAT12³ toolbox in MATLAB (version R2016b, The MathWorks, Inc., Natick, MA, USA), which involved segmenting into gray matter, white matter, and cerebrospinal fluid. Subsequently, the images were non-linearly normalized to MNI space using the DARTEL algorithm. Finally, the segmented GM images were smoothed with an 8 mm full-width-half-maximum (FWHM) isotropic Gaussian kernel. To mitigate boundary effects between gray matter and white matter, an absolute threshold masking of 0.1 was implemented in this study. In this study, we will concentrate on the hippocampus and its subregions as subcortical regions of interest within the brain. Drawing upon previous research (20), and acknowledging the inherent complexity, we utilized the shape and anatomical variation of the hippocampal region, along with the probability map of hippocampal cellular structure, to delineate it into the following five subregions: Cornu Ammonis (CA), Dentate Gyrus (DG), Entorhinal Cortex (EC), Subiculum (Subc), and the hippocampal-amygdaloid transition area (HATA). To mitigate partial volume effects, the subfields were aggregated as follows: CA1, CA2, and CA3 were combined into CA; CA4 and DG were merged into DG (21). Subsequently, a total of 10 hippocampal subregion masks for both the left and right hemispheres were generated, and the gray matter volumes of these 10 regions were extracted for intergroup comparison. Eventually, 41 individuals with MCI and 45 healthy controls underwent MRI scans for the study.

The data entry was conducted using Epidata3.1, while statistical analysis was performed using SPSS 26.0. The Hardy–Weinberg genetic equilibrium test was applied to the subjects’ genotypes using the χ2 test for goodness of fit. Data that exhibited normal distribution were presented as mean ± standard deviation and analyzed using ANCOVA. Conversely, data that did not adhere to normal distribution were presented as median (upper quartile, lower quartile) and analyzed using nonparametric statistical methods. Categorical data were analyzed using the chi-squared test. The comparison of hippocampal subfield volumes between individuals with MCI and control groups was conducted through analysis of covariance, with adjustments made for age, gender, education level, and APOEε4 as covariates. Additionally, the examination of hippocampal subfield volumes differences among individuals with MCI with varying BIN1 genotypes was carried out using a general linear model, with adjustments made for age, gender, education level, APOEε4, and total intracranial volume (TIV) as covariates. A two-tailed test was used, and a difference of p < 0.05 was considered statistically significant.

A total of 107 community-dwelling older adult participants aged 60 and above, diagnosed with MCI (n = 52) and CN (n = 55), were included in the study. There were no significant differences observed between the MCI and CN groups in terms of mean age (72.98 ± 7.38 vs. 73.35 ± 6.85, p > 0.05), gender distribution (male: female,9/43 vs. 17/38, p > 0.05), educational levels (12/18/25 vs. 17/15/20, p > 0.05), and APOEε4 (1/51 vs. 3/52, p > 0.05). The mean MMSE score was significantly higher in the MCI group compared to the CN group (9.00 vs. 5.00, p < 0.05), as depicted in Table 1.

Genotype data for the rs10200967, rs1060743 and rs4663093 polymorphisms in the BIN1 gene for MCI (n = 52) and CN (n = 55) were tested for Hardy–Weinberg Equilibrium. There were no significant differences between the observed genotype frequencies and allele distribution at these loci, as shown in Table 2. The distribution of genotype frequencies at each locus showed that the genotype at the BIN1 locusrs10200967 was predominantly CT in both MCI and CN, with 51.9% in MCI and 45.4% in CN. The rs1060743 genotype was predominantly AG, with 48.1% in MCI and 49.1% in CN. The rs4663093 genotype was predominantly CC, with 69.2% in MCI and 72.7% in CN.

In the 41 MCI and 45 CN participants included in the hippocampal subfield volumes MRI analysis, the Comparisons of the hippocampal subfield volumes were performed using a general linear model procedure, adjusting for age, gender, education level, and APOEε4 status as covariates. The findings revealed that the lCA, lDG, lHATA, lSubc, rCA, rDG, rSubc, lHIP and rHIP were also significantly decreased in the MCI group compared to the CN group (both p < 0.05), while the lEC, rEC, rHATA and TIV were non-significant decreased in the two groups (p > 0.05), as shown in Figure 1.

A covariate ANOVA was done with age, gender, education level, APOEε4 status and TIV to compare the structural differences in the hippocampal subfield volumes between different genotypes of BIN1 rs10200967, rs1060743 and rs4663093 in the MCI group. The results showed statistically significant differences in IHATA among different genotypes (p < 0.05), and there were no significant differences in lCA, lDG, lEC, lSubc, rCA, rDG, rEC, rHATA, rSubc, lHIP and rHIP according BIN1 rs10200967 genotypes (p > 0.05). We did not find any significant results with respect to differences in lCA, lDG, lEC, lHATA, lSubc, rCA, rDG, rEC, rHATA, rSubc, lHIP and rHIP among the BIN1 rs1060743 and rs4663093 genotypes (p > 0.05), As shown in Table 3.