Development And Validation Of A Model For Predicting The Six-year Risk Of Cognitive Impairment Among Chinese Older Adults

Backgroud:Modest declines in cognitive function are reported in younger adults but cognitive function rapidly worsens in the sixth decade of age and thereafter throughout the remaining adult lifespan.Currently,the prevalence of mild cognitive impairment(MCI)and dementia are 20.8%and 5.1%among Chinese older adults aged>65,corresponding to about 36.61 million and 8.98 million older adults who are suffering from MCI and dementia.In community settings,about 5%of people with MCI will progress to dementia within one year.Therefore,it’s essential to identify risk factors and high-risk population of cognitive impairment in improving the life quality of older adults and promoting healthy aging.However,a simple tool for predicting the risk of cognitive impairment was not adequately studied among Chinese older adults.Object:On the basis of a large cohort study of older adults,we intend to eatablish a risk prediction tool of cognitive impariment for Chinese older adults aged>65,and assess the incidence of cognitive impairment and identify its key risk factorsMethod:In Chinese Longitudinal Healthy Longevity Survey,we included 10,066 older adults aged≥65(mean age,83.2±11.1 years),without cognitive impairment at baseline in the 2002-2008 cohort and 9354 older adults(mean age,83,5±10.8 years)in the 2008-2014 cohort for model development and validation,respectively.We assessed the cognitive function of older adults according to the Mini-Mental State Examination(MMSE).Demographic,medical,and lifestyle information was also collected for the development of the Cox proportional hazards regression model,which included key risk factors selected via the least absolute shrinkage and selection operator(LASSO)combined with evidence reported in previous studies.The model was finally displayed through a nomogram,which was used for the establishment of Excel-based risk calculation tool of cognitive impairment.We validated the nomogram internally with 2000 bootstrap resamples and externally in a later cohort.Area under the receiver operating characteristic curve(AUC)and calibration curve were applied to evaluate the predictive discriminative ability and accuracy of the model.The net benefit of the model was evaluated by decision curve analysis.Results:1.The risk for incident cognitive impaiment is pretty high among Chinese older adults aged≥65,corresponding to incidence densities of 43.9/1000 person-years and 50.8/1000 person-years calculated using CLHLS cohort data from 2002 to 2008 and 2008 to 2014,respectively.Totally,the incidence density increases with age and is higher in female2.In this study,we included eight key predictors based on a LASSO selection procedure combined with evidence reported in previous studies,including age,baseline MMSE,activities of daily living(ADL)and instrumental activities of daily living(IADL)score,chewing ability,visual function,history of stroke,watching TV or listening to the radio,and growing flowers or raising pets.The final results revealed that hazard ratio(HR)and 95%confidence interval(CI)is 1.09(1.07,1.10)for the linear association between age and cognitive impairment,and 0.97(0.96,0.99)for their cubic association;HR(95%CI)for baseline MMSE is 0.95(0.94,0.97),for ADL and IADL score is 1.03(1.02,1.04);HR(95%CI)are 1.12(1.02,1.24),1.21(1.08,1.36)and 1.29(1.04,1.61)for poor chewing ability,poor visual function and history of stroke;compared with older adults who never watching TV or listening to the radio,the HR(95%CI)are 0.89(0.78,1.00)and 0.84(0.75,0.94)for those who sometimes and often do that;compared with older adults who never growing flowers or raising pets,the HR(95%CI)are 0.83(0.68,1.01)and 0.72(0.60,0.86)for those who sometimes and often do that.3.The prediction model of cognitive impairment established in this study has a good performance,with AUCs of 0.891 and 0.867 for internal and external validation,respectively The calibration curves showed good consistency between predicted and observed risk of cognitive impairment.4.Nomogram-based visualization tool and Excel-based risk calculation tool of cognitive impairment were developed on the basis of prediction model in this study,we can quickly assess the risk of cognitive impairment with easily ascertained variables.Conclusion:1.The risk of cognitive impairment is pretty high among Chinese older adults.We should take active interventive measures to help reduce or postpone the incident of cognitive impairment.2.Based on the findings of this study,older adults can reduce or postpone the incident of cognitive impairment via watching TV or listening to the radio,growing flowers or raising pets,or maintain independence of ADL and IADL,good chewing ability,good visual function,and etc.3.The prediction model established in this study was validated with good reliability,the nomogram and Excel-based risk calculator can be used as convenience tools for risk assessment of cognitive impairment.Public health practitioners or physicians may not only quickly assess the future risk of cognitive impairment for older adults,but also provide individualized instructive advices on preventive interventions for cognitive impairment on the basis of predictors input.