【作者】 胡玲；

【导师】 赵星； 刘元元；

【作者基本信息】 四川大学 ， 流行病与卫生统计学， 2023， 硕士

【摘要】 背景抑郁和焦虑会影响患者的认知及社会功能,进而影响其工作和日常生活,对患者个人、家庭以及整个社会都会造成极大的经济负担。据世界卫生组织(World Health Organization,WHO)估计,2019年全球约有2.8亿人患有抑郁,约3.01亿人患有焦虑。近年来有许多研究发现,受教育程度与焦虑/抑郁存在关联,受教育程度较高是焦虑和抑郁的保护因素,提高受教育程度能降低焦虑/抑郁的疾病负担以及相关的经济损失。因此,有必要探索受教育程度与焦虑/抑郁关联机制,为制定预防焦虑和抑郁症状的策略和措施提供科学依据。目前关于受教育程度与焦虑/抑郁的中介研究,其关注的可改变中介因素较为单一,且研究人群多为青少年和老年人。鉴于探索受教育程度与焦虑/抑郁症状的关联机制可以识别可控的中间因素,可为制定人群层面的社会政策提供一定依据,具有公共卫生意义,尤其是在教育和经济欠发达的我国西南地区更具有实际应用价值。同时考虑到受教育程度与代谢综合征及其组成因素关联密切,而代谢综合征及其组成因素又是焦虑和抑郁的主要危险因素,且可对代谢综合征及其组成因素进行干预。本研究拟探讨西南区域自然人群受教育程度与焦虑/抑郁症状的关联机制,并进一步探索代谢综合征及其组成因素是否会介导受教育程度与焦虑/抑郁症状的发生,即“受教育程度→代谢综合征及其组成因素→焦虑症状/抑郁症状”,本研究将关注该路径是否能解释受教育程度与焦虑/抑郁症状发生风险的关联。中介分析是通过引入中介变量,用以识别或者解释暴露A与结局Y之间关系或作用机制的统计方法。传统中介分析和因果中介分析都是应用较为广泛的中介分析方法。传统中介分析是基于线性回归分析进行直接效应和间接效应的参数估计,目标估计量可能是因果效应也可能是条件关联,且模型依赖性大,当暴露变量和中介变量间存在交互作用或者模型为非线性时,传统中介分析不再适用。而基于因果推断的中介分析提供了关于效应的明确定义以及对效应估计的因果解释所需的所有前提假设,可以使用多种估计方法对效应大小进行估计,并且能增强受教育程度与焦虑/抑郁关联的因果解释。鉴于因果中介分析较传统中介分析的优势,故本研究采用因果中介分析方法探索受教育程度与焦虑/抑郁的关联机制。方法本研究依托国家重点研发计划项目“西南区域自然人群队列”(后简称西南队列),纳入同时参加了基线调查与重复调查研究的研究对象,共计7403名30-79岁成年人,通过对研究对象进行现场面对面问卷调查、体格检查及实验室生化检测收集受教育程度、焦虑/抑郁症状、代谢综合征及其组成因素以及协变量信息。本研究的受教育程度信息来自基线调查时问卷调查中的一般情况部分,焦虑/抑郁症状信息来自重复调查时问卷调查中的心理状况部分,采用抑郁症筛查量表(The 9-question patient health questionnaire,PHQ-9)评估调查对象的抑郁症状,采用广泛性焦虑自评量表(Generalized anxiety disorder 7,GAD-7)评估调查对象的焦虑症状,代谢综合征及其组成因素信息分别来自基线调查和重复调查中的体格检查和实验室生化检测数据。焦虑抑郁的流行情况分析以是否有焦虑症状或抑郁症状分组,描述研究人群中社会人口学特征、行为生活方式、健康状况等特征。通过logistic回归分析探索受教育程度与代谢综合征及其组成因素、焦虑症状发生风险以及抑郁症状发生风险的关联;代谢综合征及其组成因素与焦虑症状和抑郁症状的关联。随后本研究使用基于自然效应模型的因果中介分析方法,分别分析代谢综合征及其组成因素是否介导受教育程度与焦虑抑郁症状发生风险,以及代谢综合征组成因素是否共同介导受教育程度与焦虑抑郁症状发生风险。结果本研究纳入7403名研究对象,女性4496人,占60.7%,男性2907人占39.3%。年龄平均值为51.37(标准差:10.75),年龄为45-64岁4348人,占58.7%;汉族4579人占比61.9%;受教育程度为未正规上学或小学3540人,占47.8%,初中2062人,占27.9%,高中及以上1801人,占24.3%;在基线调查中,代谢综合征、肥胖、甘油三酯异常、高密度脂蛋白胆固醇异常、高血压、糖尿病患病率分别为34.8%、59.5%、36.5%、22.2%、47.1%、30.6%。无抑郁症状和有抑郁症状的人群在年龄、性别、婚姻状况、受教育程度、家庭年收入、工作状况、地区、城市农村户口、吸烟状况、饮酒状况、非静坐体力活动代谢当量、DASH饮食指数、代谢综合征、甘油三酯异常、高密度脂蛋白胆固醇异常方面差异具有统计学意义。无焦虑症状和有焦虑症状的人群在年龄、性别、民族、婚姻状况、受教育程度、家庭年收入、工作状况、地区、城市农村户口、吸烟状况、饮酒状况、非静坐体力活动代谢当量、DASH饮食指数、甘油三酯异常、高密度脂蛋白胆固醇异常方面差异具有统计学意义。基于logistic回归模型分析结果显示,与未正规上学及小学的相比,受教育程度为初中对抑郁症状发生有保护作用(OR:0.769,95%CI:0.606,0.932);受教育程度为初中对焦虑症状发生有保护作用(OR:0.720,95%CI:0.542,0.898)。与受教育程度为未正规上学及小学相比,受教育程度为初中与代谢综合征风险降低有关(OR:0.848,95%CI:0.701,0.995);受教育程度为初中与的高血压风险降低有关(OR:0.782,95%CI:0.651,0.913),高中及以上与的高血压风险降低有关(OR:0.815,95%CI:0.644,0.986);受教育程度为初中与的糖尿病风险降低有关(OR:0.865,95%CI:0.726,1.004),高中及以上与糖尿病的风险降低有关(OR:0.794,95%CI:0.616,0.972)。患代谢综合征与抑郁症状的风险增加有关(OR:1.283,95%CI:1.140,1.426),患代谢综合征与焦虑症状的风险增加有关(OR:1.172,95%CI:1.017,1.327);患甘油三酯异常与抑郁症状的风险增加有关(OR:1.384,95%CI:1.255,1.513),与焦虑症状的风险增加有关(OR:1.335,95%CI:1.196,1.474);患高密度脂蛋白胆固醇异常与抑郁症状的风险增加有关(OR:1.168,95%CI:1.021,1.315),与焦虑症状的风险增加有关(OR:1.172,95%CI:1.019,1.325)。基于自然效应模型的单中介分析结果表明,代谢综合征及其组成因素在受教育程度与焦虑及抑郁症状关联的中介效应没有统计学意义;多重中介分析结果表明,肥胖、甘油三酯异常、高密度脂蛋白胆固醇异常、高血压及糖尿病共同介导的中介效应没有统计学意义。效应修饰分析结果表明,在汉族人群中,单中介分析结果显示,代谢综合症和甘油三酯异常是介导受教育程度和抑郁症状关联的中介因素,与未正规上学及小学的相比,受教育程度为初中的人群可能会通过降低代谢综合症和甘油三酯异常的发生从而降低抑郁症状的发生风险,代谢综合征的间接效应为0.987(95%CI:0.973,1.000),甘油三酯异常的间接效应为0.985(95%CI:0.972,0.999),与未正规上学及小学的相比,受教育程度为高中可能会通过降低代谢综合症的发生从而降低抑郁症状的发生风险,代谢综合征的间接效应为0.981(95%CI:0.963,0.999);单中介分析发现甘油三酯异常是介导受教育程度和焦虑症状关联的中介因素,与未正规上学及小学的相比,受教育程度为初中可能会通过降低甘油三酯异常的发生从而降低焦虑症状的发生风险,甘油三酯异常的间接效应为0.982(95%CI:0.965,0.999);多中介分析结果显示,调查肥胖、甘油三酯异常、高密度脂蛋白胆固醇异常、高血压及糖尿病共同介导的中介效应没有统计学意义。结论本研究发现,未见代谢综合征及其组成因素介导受教育程度与焦虑/抑郁症状关联的中介作用。效应修饰分析显示,在汉族人群中,代谢综合症和甘油三酯异常是介导受教育程度和抑郁症状关联的中介因素,甘油三酯异常是介导受教育程度和焦虑症状关联的中介因素症状。我们的研究结果表明,在汉族人群中,较高的受教育程度可能通过降低代谢综合症和甘油三酯异常的发生从而降低抑郁症状的发生风险;较高的受教育程度可能通过降低甘油三酯异常的发生从而降低焦虑症状的发生风险。在汉族人群中重点控制代谢综合征和甘油三酯异常可以将较高的受教育程度对焦虑及抑郁症状的保护作用最大化。建议在汉族人群中加强对代谢综合征及甘油三酯异常的预防、管理和治疗,以预防焦虑抑郁的发生。更多还原

【Abstract】 Background: Depression and anxiety can affect the cognitive and social functioning of patients,which in turn affects their work and daily life,and can impose a significant economic burden on the individual,the family,and society as a whole.The World Health Organization(WHO)estimates that in 2019,approximately 280 million people worldwide suffer from depression and 301 million people suffer from anxiety.A number of studies in recent years have found an association between educational attainment and anxiety/depression,that higher educational is a protective factor for anxiety and depression,and that increased educational reduces the disease burden of anxiety/depression and the associated economic costs.Therefore,it is necessary to explore the mechanism of the association between educational and anxiety/depression to provide a scientific basis for the development of strategies and measures to prevent anxiety and depressive symptoms.Current studies on the mediation of educational attainment and anxiety/depression have focused on a single modifiable mediator,and the study populations are mostly adolescents and older adults.Given that exploring the association mechanism between educational attainment and anxiety/depression symptoms can identify controllable mediators,it can provide a basis for formulating population-level social policies with public health implications,especially in the less educated and economically developed southwestern region of China with practical applications.Also,considering that education level is closely associated with metabolic syndrome and its components,which are the main risk factors for anxiety and depression,and that interventions can be made for metabolic syndrome and its components.In this study,we intend to investigate the mechanism of the association between educational attainment and anxiety/depressive symptoms in a natural population in the southwest region,and further explore whether the metabolic syndrome and its components mediate the occurrence of educational attainment and anxiety/depressive symptoms,i.e.,"educational attainment → metabolic syndrome and its components → anxiety symptoms/depressive symptoms".This study will focus on whether this pathway explains the association between educational attainment and the risk of anxiety/depressive symptoms.Mediational analysis is a statistical method used to identify or explain the relationship or mechanism of action between exposure A and outcome Y by introducing mediating variables.Both traditional mediation analysis and causal mediation analysis are more widely used methods of mediation analysis.Traditional mediation analysis is based on linear regression analysis for parameter estimation of direct and indirect effects,and the target estimates may be causal effects or conditional associations,and the model dependence is large.Traditional mediation analysis is no longer applicable when there is an interaction between exposure and mediating variables or the model is nonlinear.In contrast,mediation analysis based on causal inference provides a clear definition of the effect and all the prerequisite assumptions needed for the causal interpretation of the effect estimates,allows the estimation of effect sizes using multiple estimation methods,and enhances the causal interpretation of the association between educational attainment and anxiety/depression.Given the advantages of causal mediation analysis over traditional mediation analysis,the present study used causal mediation analysis to explore the mechanisms underlying the association between educational attainment and anxiety/depression.Methods: The study was based on the "Southwest Regional Natural Population Cohort"(later referred to as the Southwest Cohort),a national key research and development project,and included a total of 7403 adults aged 30-79 years who participated in both the baseline survey and the repeat survey.Education,anxiety/depression symptoms,metabolic syndrome and its components,and covariates were collected through face-to-face questionnaires,physical examinations,and laboratory tests.In this study,information on education level was obtained from the general information section of the questionnaire at baseline,and information on anxiety/depression symptoms was obtained from the psychological status section of the questionnaire at repeat survey.The Generalized anxiety disorder 7(GAD-7)was used to assess the anxiety symptoms of the respondents,and information on the metabolic syndrome and its components were obtained from the physical examination and laboratory biochemical test data in the baseline and repeated surveys,respectively.The prevalence of anxiety-depression was analyzed in terms of the presence or absence of anxiety symptoms or depressive symptoms grouped to describe the characteristics of sociodemographic characteristics,behavioral lifestyle,and health status in the study population.The association of educational attainment with the metabolic syndrome and its components,the risk of anxiety symptoms,and the risk of depressive symptoms was explored by logistic regression analysis;the association of the metabolic syndrome and its components with anxiety symptoms and depressive symptoms.The study then analyzed whether the metabolic syndrome and its components mediated the risk of education and anxiety and depressive symptoms,and whether the components of the metabolic syndrome jointly mediated the risk of education and anxiety and depressive symptoms,respectively,using a natural effects model-based causal mediation analysis.Results: The study included 7403 study subjects,4496 females(60.7%)and2907 males(39.3%).The mean age was 51.37(standard deviation: 10.75),with 4348(58.7%)aged 45-64 years;4579(61.9%)were Han Chinese;3540(47.8%)had no formal schooling or elementary school,2062(27.9%)had middle school,and 1801(24.3%)had high school and above;in the baseline survey,metabolic syndrome,obesity,abnormal triglycerides,abnormal HDL cholesterol,hypertension,and diabetes prevalence were 34.8%,59.5%,36.5%,22.2%,47.1%,and 30.6%,respectively.The differences between those without and with depressive symptoms were statistically significant in terms of age,gender,marital status,education,annual household income,work status,region,urban-rural household,smoking status,alcohol consumption status,non-sedentary physical activity metabolic equivalent,DASH diet index,baseline metabolic syndrome,baseline triglyceride abnormalities,and baseline HDL cholesterol abnormalities.The differences between those without and with anxiety symptoms were statistically significant in terms of age,gender,ethnicity,marital status,education,annual household income,work status,region,urban-rural household,smoking status,alcohol consumption status,non-sedentary physical activity metabolic equivalent,DASH diet index,triglyceride abnormalities,and HDL cholesterol abnormalities.The results of logistic regression model-based analysis showed that compared with those who did not attend formal schooling and elementary school,education level of junior high school had a protective effect on the occurrence of depressive symptoms(OR: 0.769,95% CI: 0.606,0.932);education level of junior high school had a protective effect on the occurrence of anxiety symptoms(OR: 0.720,95% CI:0.542,0.898).Compared with non-formal schooling and elementary schooling,junior high school education was associated with a lower risk of metabolic syndrome(OR:0.848,95% CI: 0.701,0.995);junior high school education was associated with a lower risk of hypertension(OR: 0.782,95% CI: 0.651,0.913)and high school and above was associated with a lower risk of hypertension(OR: 0.815,95% CI: 0.815).(OR: 0.815,95%CI: 0.644,0.986);lower risk of diabetes mellitus was associated with middle school education(OR: 0.865,95%CI: 0.726,1.004)and high school and above(OR: 0.794,95%CI: 0.616,0.972).Having metabolic syndrome was associated with an increased risk of depressive symptoms(OR: 1.283,95%CI: 1.140,1.426),having metabolic syndrome was associated with an increased risk of anxiety symptoms(OR:1.172,95%CI: 1.017,1.327);having triglyceride abnormalities was associated with an increased risk of depressive symptoms(OR: 1.384,95%CI.1.255,1.513)and an increased risk of anxiety symptoms(OR: 1.335,95%CI: 1.196,1.474);having abnormal HDL cholesterol was associated with an increased risk of depressive symptoms(OR:1.168,95%CI: 1.021,1.315)and with an increased risk of anxiety symptoms(OR: 1.172.95%CI: 1.019,1.325).The results of the single mediator analysis based on the natural effects model showed no statistically significant mediating effect of the metabolic syndrome and its constituents in the association of educational attainment with anxiety and depressive symptoms;the results of the multiple mediator analysis showed no statistically significant mediating effect jointly mediated by obesity,abnormal triglycerides,abnormal HDL cholesterol,hypertension,and diabetes mellitus.The results of effect modification analysis showed that in the Han population,the results of single mediator analysis showed that metabolic syndrome and triglyceride abnormalities were mediators of the association between education level and depressive symptoms,and those with junior high school education may reduce the risk of depressive symptoms by reducing the occurrence of metabolic syndrome and triglyceride abnormalities compared with those with no formal schooling and elementary school,and the indirect 0.987(95% CI: 0.973,1.000)for metabolic syndrome and 0.985(95%CI: 0.972,0.999)for triglyceride abnormalities,compared with those who did not attend regular school and elementary school,and high school education may reduce the risk of depressive symptoms by reducing the occurrence of metabolic syndrome,and the indirect effect of metabolic syndrome was 0.981(95% CI: 0.963,0.999);single mediator analysis identified triglyceride abnormalities as a mediator mediating the association between educational attainment and anxiety symptoms,and compared to those with no formal schooling and elementary school,educational attainment of middle school may reduce the risk of anxiety symptoms by decreasing the occurrence of triglyceride abnormalities,with an indirect effect of triglyceride abnormalities of 0.982(95% CI: 0.965,0.999);results of the multiple mediator analysis showed no statistically significant mediating effect of investigating the joint mediation of obesity,triglyceride abnormalities,HDL cholesterol abnormalities,hypertension and diabetes mellitus.Conclusions: In this study,the results of this study found no mediating effect of metabolic syndrome and its constituents mediating the association between educational attainment and anxiety/depressive symptoms.Effect modification analysis revealed that metabolic syndrome and triglyceride abnormalities were mediators of the association between educational attainment and depressive symptoms and triglyceride abnormalities were mediators of the symptom of the association between educational attainment and anxiety symptoms in the Han population.Our findings suggest that in the Han population,higher educational attainment may reduce the risk of depressive symptoms by reducing the occurrence of metabolic syndrome and triglyceride abnormalities,and higher educational attainment may reduce the risk of anxiety symptoms by reducing the occurrence of triglyceride abnormalities.Focusing on the control of metabolic syndrome and triglyceride abnormalities in the Han population may maximize the protective effect of higher educational attainment on anxiety and depressive symptoms.It is recommended that prevention,management,and treatment of metabolic syndrome and triglyceride abnormalities be enhanced in the Han population to prevent the onset of anxiety and depression.更多还原