| Mild cognitive impairment(MCI)is a transitional state between normal aging and dementia.Amnestic mild cognitive impairment(aMCI)is a subtype of MCI,of which memory is impaired.aMCI has the characteristics of high incidence,high prevalence and high conversion rate to dementia.Therefore,early identification of aMCI is of great significance for effective intervention to prevent the occurrence and development of dementia timely.Compared to paper-and-pencil cognitive screening tests,computerized cognitive screening tests have many advantages,such as high interest,standardized management,long-term storage and so on,which can reduce the restrictions of professionals on screening.However,there is no computerized cognitive screening test designed for aMCI in China.Objectives(1)To design and develop the Computerized Semantic Paired Associative Learning Test(CSPALT)for Chinese older adults with aMCI,based on the theory of controlled learning,semantic interference and memory binding,as well as the design of the Miami Test of Semantic Interference and Learning(MITSI-L);(2)To examine the discriminant validity of CSPALT in distinguishing between older adults with aMCI and older adults with cognitively normal(CN);the concurrent validity of CSPALT compared with common cognitive scale;and the test-retest reliability of CSPALT.MethodsBased on the theory of controlled learning,semantic interference and memory binding,as well as the design of MITSI-L,the content and the process of CSPALT were designed and the test documents of CSPALT were written;the program of CSPALT were written and debugged;Through the pre-experiment,CSPALT was improved further.The older adults aged 60 years old and above were recruited.According to the golden standard diagnosis of CN and aMCI,used Mini-Mental State Examination(MMSE),Montreal Cognitive Assessment(MoCA),Auditory Verbal Learning TestHuashan version(AVLT-H),Clinical Dementia Rating(CDR),Activity of Daily Living(ADL),15-item Geriatric Depression Scale(GDS-15),the older adults who met the inclusion and exclusion criteria were diagnosed and divided into two groups: the CN group(n1 = 55)and the aMCI group(n2 = 50).Subjects in the two groups accepted CSPALT.At an interval of one month,10 older adults selected from the CN group accepted CSPALT again.The Receiver Operating Characteristic(ROC)curve and the Area Under the Curve(AUC)of ROC were used to examine the discriminant validity of CSPALT according to golden standard diagnosis.Pearson correlation was used to evaluate the concurrent validity of CSPALT by comparing with MMSE,MoCA,and AVLT-H Long-term Delayed Recall.The Intra-class Correlation Coefficient(ICC)was used to evaluate the test-retest reliability of CSPALT.SPSS 23.0 and MedCalc 19.2 were used to analyze the data.Results1.Design and development of CSPALT:CSPALT is a computerized cognitive screening test with touch screen interaction,including “Personal Information”,“Test Instruction”,“Test Practice”,“Set Up the Test”and “Formal Test”,allowing to learn and recall list A and list B containing 12 semantic word-pairs,assessing cognitive function by the correct number of CSPALT and the mean correct time of CSPALT.2.The recruitment of subjects:In this study,a total of 164 subjects were recruited.According to the inclusion and exclusion criteria,52 older adults were excluded.112 subjects met the golden standard diagnosis of CN and aMCI,and accepted CSPALT.7 older adults were excluded due to invalid data.105 subjects(number of CN group = 55,number of aMCI group = 50)were included in this study.One month later,10 subjects in the CN group accepted CSPALT again.115 CSPALT data were successfully collected.3.Socio-demographic data and cognitive function:The median age in the aMCI group was significantly higher than that in the CN group(p < 0.001).The average scores of MMSE,MoCA and AVLT-H Long-term Delayed Recall in the aMCI group were significantly lower than those in the CN group(p < 0.001).4.The correct number of CSPALT and the mean correct time of CSPALT:The correct number of CSPALT and its subtests in the aMCI group were significantly lower than those in the CN group(p < 0.001).The mean correct time of CSPALT and its subtests in the aMCI group were significantly higher than those in the CN group(p < 0.05).5.The screening effect of CSPALTThe optimal cut-off of the correct number of CSPALT was established for the entire sample as well as for different age levels.The cut-off of 33 was optimal for the entire sample yielding a sensitivity of 86.00% and a specificity of 83.64% for differentiating aMCI and CN,and AUC was 0.908(95% CI = 0.836～0.956);For subjects aged under 75 years old,the correct number of CSPALT less than 34 yielded a sensitivity of 83.33% and a specificity of 83.33% for aMCI,and AUC was 0.873(95%CI = 0.754～0.948);For subjects aged 75 years old and above,the correct number of CSPALT less than 34 yielded a sensitivity of 86.84% and a specificity of 84.62% for aMCI,and AUC was 0.929(95% CI = 0.821～0.982).There was a moderate positive correlation between the correct number of CSPALT and the score of MMSE(r = 0.461,p < 0.001);There was a strong positive correlation between the correct number of CSPALT and the score of MoCA(r = 0.667,p < 0.001);There was a strong positive correlation between the correct number of CSPALT and the score of AVLT-H Long-term Delayed Recall(r = 0.700,p < 0.001).In a sub-sample of 10 older adults selected from the CN group at an interval of one month,a very strong positive correlation in the correct number of CSPALT was observed between the two time-points(r = 0.819,p < 0.001);a very strong positive correlation in the mean correct time of CSPALT was observed between the two timepoints(r = 0.904,p < 0.001).ConclusionsThe CSPALT designed and developed in this study can effectively distinguish older adults with normal cognition from older adults with aMCI,and has high sensitivity,specificity and test-retest reliability.It is a simple,efficient,interesting and acceptable computerized cognitive screening test for the identification of older adults with aMCI. |
PDF Full Text Request