【作者】 邹涛；

【导师】 姚树桥；

【作者基本信息】 中南大学 ， 精神病与精神卫生学， 2006， 博士

【副题名】系列评估量表中文版的建立和整合模型的构建

【摘要】 目的：(1)建立与抑郁易感性——应激模型研究相关系列量表的中文版，并分析其信效度；(2)应用建立的量表，在中国大学生人群中对抑郁的易感性(认知／人际)——应激理论模式进行实证分析；(3)探索中国大学生的抑郁易感性因素，在考虑文化因素影响的基础上，初步构建大学生的易感性——应激整合模型。方法：(1)相关量表英文版由加拿大McGill大学心理系Abela教授提供给本研究使用。由研究者们翻译为中文，McGill大学双语学家回译，再修正。647名大学生完成13个量表的初次测试。3-4周后重测，每个量表重测人数在120-130人之间。信度检验采用Cronbach’sα系数、条目间平均相关系数和重测信度(内部一致性系数)。效度检验包括内容效度和结构效度(总分与各因子间相关系数及验证性因素分析)。(2)在一月后进行追踪测查，测查量表为学生日常生活和学习应激量表(General Social and Academic Hassles Scale for Students，GSASHS)和流调中心用抑郁量表(Center for Epidemiologic Studies Depression Seale，CES-D)，完成人数601人。然后，用多层复回归分析进行抑郁症状的预测分析；路径分析和结构方程模型用于易感性-应激模型的探索和整合结果：(1)除BHRQ量表外，其余量表的α系数均在0.7以上，其中CSQ、RSQ、CES-D、GSASHS、AI更是在0.90以上，而BHRQ的条目间相关均数(MIC)为0.43；各量表内部因子(分量表)间α系数，除DAS分量表外，均大于0.6，而DAS分量表的MIC也在可接受范围(0.17-0.28)；各量表重测信度指标内部一致性系数(ICC)都具有显著意义，大于0.6；(2)各量表总分与其分量表有较强的相关(r＞0.6)，验证性因素分析显示：CSQ四子结构；DAS八因子结构；ASQ五因子结构；RSQ三因子结构；以及GASHSS三因子结构；AI7因子结构；ICS四因子结构；以及CESD的4因子结构，其CFI、IFI，GFI均大于或等于0.90，而RMSEA也都在＜0.1以下。(3)认知／人际易感因素与应激的交互作用进入回归方程，除自我批评×应激外，对抑郁的发作均有预测作用，增强了抑郁发作的可能(P＜0.00)。(4)路径分析显示：文化因素→不安全依恋→负性认知→过分思虑→抑郁是一条重要的作用路径，过分思虑是一个关键的中介(β=0.12，P＜0.01)。(5)整合的易感性——应激模型有较好的拟合指数，GFI=0.95，CFI=0.94，IFI=0.94，RMSEA=0.085。(6)文化影响因素对抑郁作用的回归方程不具有显著意义(ΔR=0.00ΔF=0.96，P=0.43)。结论：(1)引进的与抑郁的易感性——应激研究相关的13个量表有较好的同质性信度和重测信度，验证性因素分析结果表明CSQ四子结构；DAS八因子结构；ASQ五因子结构；RSQ三因子结构；以及GASHSS三因子结构；AI7因子结构；ICS四因子结构；以及CESD的4因子符合该量表的理论构想。(2)在中国大学生人群的研究，支持抑郁的易感性——应激理论模式，抑郁与应激的交互作用能够增强对抑郁的预测。(3)整合的路径分析表明，文化因素→不安全依恋负性认知→过分思虑→抑郁是一条重要的作用路径，过分思虑是一个关键的中介。(4)在考虑文化因素的基础上，初步的建立和验证了一个符合中国大学生人群的易感性——抑郁的整合模型，为大学生抑郁的干预和治疗提供了理论依据。(5)文化影响因素对抑郁的作用是间接的，不具有直接的预测作用。更多还原

【Abstract】 Objective: (1) To develop Chinese translation of a set ofassessments scales about diathesis-stress of depression and evaluate theirreliability and the validity. (2) To test the validity of the Vulnerability-Stress model of depression in Chinese undergraduates. (3) To explore theintegrated Vulnerability-Stress model of depression in Chineseundergraduates and take culture factors into account.Method: (1) The original version of a set of assessments scalesabout diathesis-stress of depression was provided by Abela professor inMcGill University and these scales were translated into Chinese byinvestigators. Then, the two bilingual translators in McGill universitymodified these translations, and a final versions were approved by bothof them after repeated back-translation.647 undergraduate studentsfinished the Chinese version of 13 scales in time one. The reliabilityanalysis included the Cronbach’s coefficient, the mean inter-itemcorrelation coefficient and test-retest reliability (intra-class correlationcoefficient).The validity analysis included construct validity (the totalscale and subscale inter-correlations and several indexes of themaximum-likelihood confirmatory factor analysis). (2) A follow-upassessment in 1 month later, the GSASHS (General Social and AcademicHassles Scale for Students) scale and CES-D (Center for EpidemiologicStudies Depression Scale) scale were completed. Then, vulnerabilityfactors for depression symptoms were used to predict depressionsymptoms by Multiple Hierarchical Regression (MHR). Path Analysisand Structural Equation Model (SEM) were used to explore integratedVulnerability-Stress model of depression in Chinese undergraduates.Result: (1) The set of scales about diathesis-stress of depressionexhibited Cronbach’s alpha coefficient was more than 0.7 and the meaninter-item correlation than 0.6(F was statistics significantly). (2) TheGFIs(≥0.90),CFIs(≥0.90), IFIs(≥0.90) and RMSEAs(＜0.1) of the setof scales all met the adequacy of fit standards., which respectively,confirmatory analysis for four-factor structure of the CSQ, eight-factorstructure of the DAS, five-factor structure of ASQ, three-factor structureof RSQ, three-factor structure of GASHSS, seven-factor structure ofAI,four-factor structure of ICS and four-factor structure of CESD.(3) vulnerability-stress interaction was enter into regression equation, Theresults show that the vulnerability-stress interaction provide incrementalpredictive validity to depressive symptoms (sigΔF=0.00). (4)Pathanalysis show that culture factor→insecurity attachment→negativecognition→rumination→depression was a important pathways to causedepression (β=0.12,P＜0.01). (5) The structural equation model forintegrated Vulnerability-Stress model of depression analysis indicatedthe fit indices: GFI=0.95, CFI=0.94, IFI=0.94, RMSEA=0.085.(6)culturefactor as predict variable was entered into regression equation forpredictor depression but it was no statistics significantly (ΔR=0.00ΔF=0.96,P=0.43).Conclusion: (1) The set of scales about diathesis-stress ofdepression exhibited rightly internal consistency. The GFIs,CFIs, IFIsand RMSEAs of the set of scales all met the adequacy of fit standards,which confirmatory four-factor structure of the CSQ, eight-factorstructure of the DAS, five-factor structure of ASQ, three-factor structureof RSQ, three-factor structure of GASHSS, seven-factor structure ofAI,four-factor structure of ICS and four-factor structure of CESD,respectively, and supported these scales of construct validity. The set ofassessment scales was appropriate for assessing diathesis-stress ofdepression in china. (2) The results of this study in Chineseundergraduates affirm the theory of Vulnerability-Stress model ofdepression, the vulnerability-stress interaction could incrementalsignificantly power to depressive symptoms. (3) Path analysis show thatculture factor→insecurity attachment→negative cognitionrumination→depression was a important pathways to cause depression(4)To take culture factors into account and A integratedVulnerability-Stress model of depression in Chinese undergraduates wasprovided and to be confirmed. (5) Culture factor was indirect effect todepression and could not incremental predictive validity to futuredepression.更多还原