Objective To construct a dynamic framework for bidirectional transitions of mild cognitive impairment(MCI),quantifying both rare reversion and high-risk progression trajectories in cognitive dynamics.Methods Patients diagnosed with MCI at baseline from 2005 to 2022 and completed at least two follow-up visits were selected from the Alzheimer's Disease Neuroimaging Initiative(ADNI),and a retrospective cohort was constructed.Demographic information,APOEε4 genotype,and neuropsychological scales data were collected.Longitudinal cognitive assessments were functionally reconstructed using multivariate functional principal component analysis(MFPCA),with functional principal components(FPCs)extracted based on cumulative variance contribution rate(PVE＞90％).Functional multi-state Markov models were developed to estimate inter-state transition intensities,year to year transition probabilities,and covariate effects.Results Among 1,019 MCI patients(4,657 follow-up visits),93(9.1％)reverted to normal cognition,while 359(35.2％)progressed to Alzheimer's disease(AD).Longitudinal trajectory analysis revealed significant heterogeneity:progressive MCI＞stable MCI＞reverted MCI in the first functional principal component(MFPC1)scores.The transition intensity for MCI reversion(0.020)was approximately one-fourth of the AD progression risk(0.086),but the post-reversion cognitive re-impairment intensity was 0.138.Reduced MFPC1(HR=0.993,95％Cl:0.991,0.995)and elevated MFPC2(HR=1.004,95％Cl:1.001,1.007)were closely associated with MCI reversion.Conclusion MCI exhibits marked heterogeneity in longitudinal cognitive trajectories.Although reversion is rare,reversed patients remain at high risk of cognitive re-impairment.

Objective To construct a dynamic framework for bidirectional transitions of mild cognitive impairment(MCI),quantifying both rare reversion and high-risk progression trajectories in cognitive dynamics.Methods Patients diagnosed with MCI at baseline from 2005 to 2022 and completed at least two follow-up visits were selected from the Alzheimer's Disease Neuroimaging Initiative(ADNI),and a retrospective cohort was constructed.Demographic information,APOEε4 genotype,and neuropsychological scales data were collected.Longitudinal cognitive assessments were functionally reconstructed using multivariate functional principal component analysis(MFPCA),with functional principal components(FPCs)extracted based on cumulative variance contribution rate(PVE＞90％).Functional multi-state Markov models were developed to estimate inter-state transition intensities,year to year transition probabilities,and covariate effects.Results Among 1,019 MCI patients(4,657 follow-up visits),93(9.1％)reverted to normal cognition,while 359(35.2％)progressed to Alzheimer's disease(AD).Longitudinal trajectory analysis revealed significant heterogeneity:progressive MCI＞stable MCI＞reverted MCI in the first functional principal component(MFPC1)scores.The transition intensity for MCI reversion(0.020)was approximately one-fourth of the AD progression risk(0.086),but the post-reversion cognitive re-impairment intensity was 0.138.Reduced MFPC1(HR=0.993,95％Cl:0.991,0.995)and elevated MFPC2(HR=1.004,95％Cl:1.001,1.007)were closely associated with MCI reversion.Conclusion MCI exhibits marked heterogeneity in longitudinal cognitive trajectories.Although reversion is rare,reversed patients remain at high risk of cognitive re-impairment.

Objective:This study analyzes the expression level of miR-1180-3p and constructs the regulatory network of relevant ceRNA by integrating the DNA methylation and gene expression profile of hepatocellular carcinoma from the Cancer Genome Atlas (TCGA).Methods:Firstly, the expression level of miR-1180-3p in hepatocellular carcinoma and adjacent tissues was analyzed by TCGA database, and the differential expression of lncrna and mRNA was screened. Secondly, the LncBase database and the TargetScan database were used to predict the relationship between miR-1180-3p and lncRNA and mRNA, and the DNA methylation-mediated lncRNA was screened by the DNA methylation profile of lncRNA. Finally, Cytoscape software was used to construct miR-1180-3p relevant ceRNA network, and WebGestalt website was used to perform GO and KEGG analysis of related mRNA in ceRNA.Results:Compared with patients with low expression of miR-1180-3p (mean overall survival duration, 5.69 ± 0.35 years), patients with high expression of miR-1180-3p had shorter overall survival time (mean overall survival duration, 3.99 ± 0.47 years), indicating that the high expression of miR-1180-3p was hepatocellular carcinoma risk factor affecting the prognosis ( HR = 1.28, 95% CI = 1.1 ~ 1.5, P < 0.01). A miR-1180-3p related ceRNA regulatory network was constructed in this study, which contained 2 lncRNAs (F11-AS1 and LINC01511) and 37 mRNAs. Conclusion:This study has successfully constructed miR-1180-3p relevant ceRNA regulatory network, and DNA methylation-mediated F11-AS1 and F11-AS1/miR-1180-3p/C11of54 ceRNA regulatory axis has played an important role in the occurrence and development of hepatocellular carcinoma.