Validity Of Short Forms Of The Center For Epidemiological Studies Depression Scale In Chinese Adolescents

Objective:To establish suitable for Chinese adolescents Center for Epidemiologic Studies Depression Scale(CES-D) of Chinese shorter version, and to assess its reliability and validity, while the initial establishment of the simple version of adolescent depression screening demarcation points. Methods:2821 urban and rural middle school students 2580(14.20 ± 1.70 years) completed 20 items the center for epidemiological Depression Scale(CES-D) and the Beck Depression Inventory version 2 Chinese version(BDI-II-C) of test; Then CES-D scores from the test results in the CES-D≥20 points, a total of 322 adolescents(14.14 ± 1.83 years), and CES-D < 20 points 312 non-depressive disorder adolescents(14.14±1.8 years) using school-age children affective disorder and schizophrenia questionnaire(K-SADS) interviews with the American Diagnostic and Statistical Manual of Mental Disorders 4th Edition(DSM-IV) criteria for making the diagnosis of major depressive disorder. In order to establish the Chinese short forms of CES-D, First, confirmatory factor analysis by comparing several different domestic and short forms of CES-D and full-scale factor structure in our young people fit indices; Fitting selected factor structure than CES-D full-scale’s short forms of CES-D;Then the results of the diagnosis based on clinical interviews with major depressive disorder, and Calculate and compare the sensitivity of screening out all the short forms of CES-D and full-scale on the delimitation of the respective sub-optimal screening for adolescent depression and depressive disorders, specificity, area under the receiver operating characteristic curve(ROC) area under curve(AUC), Establishing the best suited use to short forms of CES-D of Chinese adolescents. To determine the test-retest reliability, in the first eight weeks after Surveying 902 selected normal students and through interviews with young people who diagnosed major depressive disorder(n = 66) for two weeks after the completion of the CES-D re-test. Result:1.Construct validity:confirmatory factor analysis showed that, caompared CES-D full-scale four-factor model fit indices( χ2/df=14.49,CFI=0.926, TLI=0.905, RMSEA=0.049),in 9 different short forms of the CES-D:10 items two-factor model, 10 items three factors model, 11 items four-factor model and 13 items three-factor model of short forms are better than the CES-D full-scale, all proposed and indicators χ2/df=12.74-17.07,CFI=0.950-0.965,TLI=0.918-0.942,RMSEA=0.046-0.054;and better than full-scale of short forms of CES-D, to compared expect Cross-Validation Fit Indices(ECVI),10 items threefactor model short forms of CES-D’s ECVI(0.10) is more close to Saturated model(0.02);2.Clinical validity and cutoff:(1)screening with depressive disorder, compared with CES-D of full- scale(AUC = 0.88, sensitivity = 0.97, specificity = 0.74),10 items three- factor model short forms of CES-D’s specificity(0.78) is better than full-scale and other three short forms of CES-D,and AUC(0.87) quite with full-scale and other three short forms of CES-D, sensitivity(0.92) equal with all short forms of CES-D;(2) screening depression disorder,compared with the CES-D of full-scale(AUC = 0.91, sensitivity = 0.94,specificity = 0.79); 10 items three- factor model short forms of CES-D’s sensitivity(1.00) is better than full-scale and other three short forms of CES-D,and its specificity(0.75),AUC(0.89) are not significant with full-scale;3.Criterion-related validity : non-clinical sample of adolescents and clinical samples CES-D10 total score and fullscale CES-D total scores were correlated positively(r=0.95, 0.95 respectively;ps<0.01), with BDI-II scores were positively correlated(r=0.70,0.56 respectively;ps<0.01);4.Discrimination validity: depression and mild depression and normal non-depression disorder adolescents of three groups of CES-D10 total score scores were significantly different(F=126.68,p<0.001).Pairwise comparison showed were significantly(Fs=7.90-257.11,ps<0.001);5.Reliability: In the non-clinical and clinical samples of adolescents, short form of CES-D10 internal consistency coefficient(Cronbach α) were 0.75 and 0.81 for each entry and short forms of CES-D10 total score correlation coefficients were 0.43-0.69 and 0.26-0.77 respectively. test-retest correlation coefficients were 0.56(P<0.01)and 0.63(P<0.01) respectively; Conclusion:The 10 items and three factors of Chinese shorter form of CES-D10 has good and validity in Chinese adolescents, can be used as screening of adolescent depression tools.