| Aim Subjective cognitive decline(SCD)was defined as the presence of cognitive complaints in the absence of pathological neuropsychological testing.It is believed that SCD occurs at the very early stage of Alzheimer's disease,even before mild cognitive impairment(MCI).However,not all SCD individuals would progress to MCI or AD,due to the heterogeneity in SCD cohorts.Therefore,long-term follow-up studies are required to assess the role of SCD in predicating disease progression.This study is based on the Chinese Longitudinal Aging Study(CLAS)which started on 2011,especially the SCD cohort.The aims are: 1)to identify the structural MRI biomarker of SCD and its relationship with cognition;2)to explore the features for AD in the SCD cohort a 7-years mean follow-up interval;and 3)to evaluate the impact of different daily activities on cognitive prognosis.Our study helps to investigate the preclinical neuroimaging mechanism,explore the key factors for predicating cognitive decline and provide evidence for effective intervention models at early stage of AD.Methods1?In the cross-sectional study at baseline,we collected standard T1-weighted MRI data from different cohorts(SCD,MCI and health control(HC)),and obtained the volumes of the hippocampus and amygdala.Then we evaluated the pattern and extent of asymmetry in hippocampus and amygdala.Furthermore,we also investigated the relationship between the altered brain regions and cognitive function.2?In terms of the longitudinal study,we proposed a classification model to discriminate progressive SCD(p SCD)(n=24)from stable SCD(s SCD)(n=52).Our model uses multi-type features including clinical information,physical disease,lifestyle,psychological measurement and Free Surfer-derived MRI features.To respect the interaction between the considered features,Relief algorithm and cost sensitive support vector machine(SVM)were wrapped for feature selection and classification.The classification model was validated by nested leave-one-out cross validation.3?To evaluate the impact of different daily activities on cognitive prognosis,three daily activity(physical activity,intellectual activity and leisure activity)and their different combination are further considered as into the model,respectively.The performance of the activities or their combinations on preventing cognitive decline were identified by comparing their effects on the classification rates of each model.Results1.Neuroimaging features at baseline: 111 SCD,30 MCI,and 67 HC are enrolled.Significant differences were found across the three groups in the volume and asymmetry of both hippocampus and amygdala(P<0.05).Controlled by age,gender,education level,depression symptoms,anxiety symptom,somatic disease and lifestyle in terms of smoking,both SCD and MCI groups showed significant decreased right hippocampal and amygdala volume than HC group.For asymmetry pattern,a ladder-shaped difference of left-larger-than-right asymmetry was found in amygdala with MCI>SCD>HC,and an opposite asymmetry of left-less-than-right pattern was found with HC>SCD>MCI in hippocampus.The correlation between the right hippocampus and amygdala with MMSE and Mo CA scores were found.2.Longitudinal study: Of total 223 features,5 features were selected as the most contributive features in our study,including stroke history,years of education,baseline Mo CA score,volume of left amygdala,and right white matter in banks of superior temporal sulcus(wmrhbankssts).Their combination achieves a satisfactory prediction with an Area Under the Curve(AUC)score of 0.80.Decreased gray matter volume in left amygdala and increased white matter volume adjacent to right bankssts were both found in p SCD.3.Analysis of daily activities on cognitive prognosis: Of the 6 different combination(three activity alone and three in pair),only physical activity combined with the other two activity could improve the specificity of the model,indicating an effect to prevent cognitive decline.Conclusions1.Our results supported that SCD individuals are biologically distinguishable from healthy controls,and this may relate to cognitive impairment.Moreover,different levels of asymmetry in hippocampus and amygdala might be a potential clinical diagnosis index.2.By incorporating clinical and neuroimage data,it becomes possible to predict MCI at SCD stage with machine learning techniques.A potential biomarker(increased white matter volume in adjacent banks of the superior temporal sulcus)was revealed,providing a valuable and objective early diagnostic tool for understanding the pathology of AD in its prodromal phase.3.Our study supports the important role of physical activity in preventing cognitive decline. |
PDF Full Text Request