【作者】 胡燕；

【导师】 叶冬青； 范引光；

【作者基本信息】 安徽医科大学 ， 流行病与卫生统计学， 2021， 博士

【摘要】 目的在修订和评价简版青少年自闭特质自评问卷(the Short Self-reported Adolesc-ent Version of Autism Spectrum Quotient,AQ-SSAV)的基础上,描述青少年高自闭特质的流行现状,探讨其与心理行为之间的关联,同时探讨高自闭特质青少年的共情倾向和情绪面孔加工特点。方法采取分层整群抽样的方法选择安徽省阜阳、合肥和芜湖在校初、高中2,644名学生,分别从敏感性、代表性、内部一致性等指标对简版青少年自闭特质自评问卷进行修订,结合因子分析法、信效度检验对问卷加以评价。另外选取安徽省芜湖、六安、阜阳、合肥4个地市共7,521名初、高中学生作为调查对象,收集人口统计学特征、自闭特质、抑郁症状、自伤行为、共情能力等信息,描述高自闭特质的检出情况,通过Χ~2检验比较不同特征间高自闭特质的分布特征,并运用多因素Logistic回归模型分析自闭特质与上述心理行为的关联。使用自举(Bootstraping)程序和PROCESS软件检验共情能力在自闭特质与抑郁症状关联中的中介效应和调节效应。采用病例对照设计选取高自闭特质、一般群体、低自闭特质群体共96名学生,运用多维度共情测验和情绪面孔自由观看任务测定两组被试的共情倾向和情绪面孔加工特点,最终纳入83名被试的测试结果进行分析。结果通过条目分析和因子降维,最终确定2个维度共计22个条目的AQ-SSAV,方差累积贡献率为52.29%,总问卷的重测信度、Cronbach’sα系数和分半信度系数分别是0.752、0.843和0.859,内容效度中平均S-CVI(S-CVI/Ave)为0.98,与述情障碍总问卷和各维度的关联均有统计学意义。同时,研究中验证了自闭特质在人群中的连续分布特性,并给出了高和低自闭特质青少年群体的划界值分别为≥59分和≤37分。通过横断面调查得出青少年高自闭特质群体的检出率约为14.4%,且检出率在不同年级、居住地、父母文化程度、留守经历、兄弟姐妹个数、自评成绩水平、自评家庭经济情况之间差异均有统计学意义,但在不同性别和不同家庭形态间差异无统计学意义。4种不同程度抑郁症状在不同自闭特质水平间的检出情况差异有统计学意义(Χ~2=455.696,P<0.001),自闭特质水平与抑郁症状的检出有正向关联,控制混杂因素之后,多因素Logistic回归模型分析显示,普通组和高自闭特质与抑郁症状的发生仍呈正向关联(OR值分别为1.879和3.936,P值均<0.001);无自伤、偶发自伤和多发自伤在不同自闭特质水平间的检出情况差异有统计学意义(Χ~2=21.937,P<0.001),单因素Logistic回归模型分析显示,高自闭特质与NSSI的发生呈正向关联(OR值分别为1.393,P值<0.001)。横断面调查结果还提示,共情能力在不同自闭特质水平之间的差异均有统计学意义(F值范围为21.97～210.37,P值均<0.001),且随着自闭水平的升高,共情能力总体呈下降趋势;但个人痛苦维度和情感共情维度的得分却随着自闭水平的升高而上升。控制混杂因素之后,Bootstraping程序中介效应结果显示,个人痛苦维度在自闭特质与抑郁关联中起部分中介作用,中介效应为17.66%;调节作用显示“自闭特质×个人痛苦”和“自闭特质×(共情性关心+个人痛苦)”的回归系数有统计学意义(β=-0.0114,t=-3.1749,P=0.0015;β=-0.0093,t=-3.4792,P=0.005),且调节效应显著(R~2=0.2176,P<0.001;R~2=0.2179,P<0.001)。多维共情测试结果提示高自闭特质组、普通组和低自闭特质组在正性情绪图片和负性情绪图片的认知共情判断、正性和负性外显情感共情体验均无统计学意义。但三组在对负性情绪图片的认知强度等级中,高自闭特质组得分低于对照组(F=2.722,P=0.009),比较组间总的认知共情强度等级得分,差异依然有统计学意义(F=2.196,P=0.022);在对负性情绪图片内隐情感共情体验进行组间比较时,高自闭特质组得分高于对照组,且差异有统计学意义(F=2.494,P=0.006)。在情绪面孔自由观看任务中,眼部的注视时间在3组被试6种情绪之间差异有统计学意义(F=4.343,P=0.001)。以情绪喜的眼部注视时间为参照进行协方差分析,怒、悲、惊、恐、厌的眼部注视时间与之差异均有统计学意义,注视时间趋势图可见三组被试均在情绪怒的眼部注视时间最长。三组被试鼻子注视时间的差异有统计学意义(F=5.898,P=0.009),低自闭特质组注视时间最长,高自闭特质组最短。三组被试不同情绪间对于嘴部注视时间的差异比较有统计学意义(F=5.522,P<0.001),从趋势图看出喜的情绪图片中嘴部注视时间最长。三组被试不同情绪类别间非核心区域注视时间差异比较有统计学意义(F=3.119,P=0.009),高自闭特质群体在非核心区域的注视时间最长。结论(1)《简版青少年自闭特质自评问卷》的修订符合心理测量学要求,可以用来进行青少年自闭特质水平的流行病学调查,建议以≥59分作为高自闭特质的划界值。(2)青少年高自闭特质的检出率约为14.4%,高自闭特质与抑郁症状的关联有统计学意义,与自杀相关心理行为的关联无统计学意义;随着自闭水平的升高,共情能力总体呈下降趋势;但个人痛苦维度和情感共情维度的得分却随着自闭水平的升高而上升。中介和调节效应显示个人痛苦和情感共情的高得分在自闭特质与抑郁水平之间发挥了保护作用。(3)高自闭特质青少年对他人情绪强度的认知、尤其是对负性情绪强度的认知与普通人群之间的差异有统计学意义;高自闭特质青少年在眼睛和嘴部两个核心区域投注的注视时间和注视次数均少于对照组;对非核心区域的投注和加工时间、次数要多于对照组。高自闭特质青少年怒、恐、惊、悲等负性情绪面孔的加工时间比正性情绪面孔的加工时间长。更多还原

【Abstract】 ObjectiveBased on the revision and evaluation of the Short Self-reported Adolescent Version of Autism Spectrum Quotient(AQ-SSAV),the aims are to describe the prevalence of adolescents with high autism traits,and to explore the relationship between high autism traits and psychological behavior,and also to explore the tendency of adolescents’empathy and the characteristics of emotional face processing of the individuals with high autistic trait.MethodsA total of 2,644 junior and senior high school students from Fuyang,Hefei and Wuhu of Anhui Province were selected by stratified cluster sampling.The questionnaire was revised from sensitivity,representativeness,internal consistency and other indicators,and was evaluated by factor analysis and reliability and validity test.Then using the same method,a total of 7,521 middle school students from Wuhu,Lu’an,Fuyang and Hefei of Anhui Province were selected as the survey subjects.The information of demographic characteristics,autistic traits,depressive symptoms,NSSI behavior and empathy ability were collected.The detection of high autistic traits was described,Chi Square tests were used to compare the distribution characteristics of high autistic traits between different characteristics,and multivariable Logistic regression models were used to analyze the relationships between autistic traits and the above psychological behavior.Bootstrap program and PROCESS software were applied to test the mediating and moderating effect of empathy in the association between autistic traits and depressive symptoms.Using case control design,a total of 96 students from high autistic traits,general groups and low autistic groups were selected,and the common tendency and emotional face processing characteristics of the two groups were measured using Multifaced Empathy Test(MET)and Emotional Face Processing task(EFP task),and 83subjects were finally included for the final analysis.ResultsThrough item analysis and factor analysis,AQ-SSAV was finally divided into 2dimensions with 22 items in total.The cumulative contribution rate of variance was52.29%.The retest-reliability,Cronbach’sαcoefficient and split reliability coefficient of the total questionnaire were 0.752,0.843 and 0.859,respectively.The mean S-CVI(S-CVI/AVE)in the content validity was 0.98,which was statistically significant in relation to the total alexithymia questionnaire and all dimensions.At the same time,the continuous distribution of autistic traits in the population was verified in this study,and the demarcation values of high and low autistic traits in adolescents were≥59 points and≤37 points respectively.By cross-sectional survey high adolescent autistic traits group detection rate was 14.4%.There were statistically significant differences among different grades,places of residence,parents’education level,left-behind experience,number of brothers and sisters,self-rated achievement level and self-rated family economic situation.However,there was no statistical significance among different genders and different family forms.Four kinds of depressive symptoms in different detection condition between autistic traits level difference was statistically significant(Χ~2=455.696,P<0.001).The level of autistic traits was positively correlated with the occurrence of depressive symptoms.After controlling for confectioners,the multivariate Logistic regression model showed that ordinary and high autistic traits were still positively correlated with the occurrence of depressive symptoms(OR=1.879 and OR=3.936,P<0.001).There were statistically significant differences in the detection of None NSSI,occasional NSSI and multiple NSSI among different levels of autistic traits(Χ~2=21.937,P<0.001).Single-factor Logistic regression model analysis showed that high autistic traits were positively correlated with the occurrence of NSSI(OR=1.393,P<0.001).By cross-sectional survey,here were statistically significant differences in empathy among different autistic traits(F=21.97～210.37,P<0.001),and with the increase of autistic traits,the overall trend of empathy decreased.However,the scores of personal distress and emotional empathy increased with the increase in the level of autism.After controlling for confounding factors,the mediating effect of personal distress dimension between autistic traits and depression was 17.66%.The regression coefficients of"autistic trait×personal distress"and"autistic trait×(empathy concern+personal distress)"were statistically significant(β=-0.0114,t=-3.1749,P=0.0015;β=-0.0093,t=-3.4792,P=0.005),and the moderating effect was significant(R~2=0.2176,P<0.001;R~2=0.2179,P<0.001).In the Emotional Face Processing task(EFP task),there was a statistically significant difference in the fixation time among the three groups among different emotions(F=3.644,P=0.008).Covariance analysis was conducted with the eye fixation time of emotional joy as the reference,and the difference of eye fixation time of anger,sadness,surprise,fear and disgust was statistically significant.The trend chart of eye fixation time showed that the eye fixation time of three groups of subjects was the longest in emotional anger.The difference of nose fixation time among the three groups was statistically significant(F=5.898,P=0.009).The fixation time of the low autistic trait group was the longest,and the high autistic trait group was the shortest.The three groups of subjects with different emotions had statistically significant differences in the gaze time of the mouth(F=5.522,P<0.001).The trend chart showed that the gaze time of the mouth was the longest in the happy emotion picture.There were statistically significant differences in the fixation time in the non-core area among the three groups of subjects in different emotional categories(F=3.119,P=0.009),and the fixation time in the non-core area was the longest in the group with high autistic traits.Conclusions(1)The version of the Self-assessment Questionnaire of Adolescent Autism Traits conformed to the requirements of psychodeoting and can be used to conduct epidemiological surveys on the level of adolescent autism traits.It was suggested that≥59 points should be recommended as the demarcation value of high autism traits.(2)The detection rate of high autistic traits in adolescents was about 14.4%,and there was statistically significant correlation between high autistic traits and depressive symptoms and NSSI,but no statistically significant correlation between high autistic traits and suicide-related psychological behaviors.High autistic individuals showed different directions in cognitive empathy and affective empathy,and the mediating and moderating effects showed that high scores in personal pain and affective empathy played a protective role between autistic traits and depression levels.(3)There was a significant difference between adolescents with high autistic traits and the general population in their cognition of other people’s emotional intensity,especially their cognition of negative emotional intensity;The duration and frequency of fixation in the two core areas of eyes and mouth of the adolescents with high autistic traits were less than those of the control group.The fixation time in non-core areas were more than those in the control group.The addressing time of the faces with negative emotions,such as anger,fear and sadness,was longer than that of the faces with positive emotions.更多还原