失独者抑郁症状的潜在类别分析

摘要

目的：对失独者抑郁症状进行潜在类别分析，了解失独者抑郁症状的个体差异，以及不同人口学特征对潜在类别的影响。方法采取分层随机整群抽样的方法，于2014年9月—2015年12月对长沙市249例失独者进行抑郁自评量表（SDS）评分，对失独者的抑郁状况进行潜在类别分析。并分析性别、年龄、婚姻状况、家庭所在地对潜在类别的影响。结果共发放问卷249份，剔除30份未作答问卷，回收219份有效问卷，有效回收率为88.0%。本研究探索了1~5个潜在类别模型，随着分类数逐渐增加，AIC、BIC、aBIC 不断减少，Entropy 在4个类别时已最大，但是 LMR 不显著，且 BIC 增大，因此，3个类别的分类模型最为合理，其中 C1、C2、C3的类别概率分别为0.428、0.305、0.267。根据 C1、C2、C3在 SDS 20个条目上的条件概率，将其分别命名为高抑郁组（42.9%）、生理-悲观组（31.5%）、低抑郁乐观组（25.6%）。性别、年龄、婚姻状况对潜在类别的影响，差异均无统计学意义（ P >0.05）；家庭所在地对潜在类别的影响，差异有统计学意义（ P <0.05）。结论失独者的抑郁症状有明显分类特征，可以分为 3个潜在类别，分别为高抑郁、生理-悲观、低抑郁乐观，且家庭所在地可影响潜在类别。%Objective To perform latent class analysis among the only - child - lost persons,to explore the individual differences of the depression symptom among the only - child - lost persons,and to explore the effects of different demographic characteristics on the latent class. Methods 249 only - child - lost persons were selected as study subJects by using stratified cluster random sampling method from September 2014 to December 2015,the latent class analysis was used to identify the depression,and cases were asked to complete the self - rating depression scale(SDS). And the effect of gender,age,marital status and home location on the latent class was analyzed. Results A total of 249 questionnaires were distributed,among which 30 had no answer,219 were effective,the effective recovery rate was 88. 0% . This study explored 1 - 5 latent class models. The values of AIC,BIC and aBIC dereased with the increase of number of classes. The value of Entropy would be the largest when the latent class was a four - classes model. However,at that time,the value of LMR was not significant and the value of BIC was larger than that in the three - classes model. In summary,three - classes model was the most reasonable model. And the probability class of C1,C2,C3 were 0. 428,0. 305,0. 267 respectively. According to conditional probability of C1,C2, C3 on 20 items of SDS,the three - classes could be named as high - depression group(42. 9% ),physical - pessimistic group (31. 5% ),low depression optimistic group(25. 6% ). There was no significant difference in the effect of gender,age,marital status on the latent class among groups(P > 0. 05),and there was significant difference in the effect of home location on the latent class among groups(P < 0. 05). Conclusion The only - child - lost persons have obvious grouping features,which can be divided into three latent classes:high - depression,physical - pessimistic,low depression optimistic. The home location can influence latent class.