Grucza Richard A, Przybeck Thomas R, Cloninger C Robert
Department of Psychiatry, Washington University School of Medicine, St. Louis, MO 63110, USA.
Compr Psychiatry. 2005 May-Jun;46(3):214-22. doi: 10.1016/j.comppsych.2004.08.006.
The aim of this study was to determine whether personality might partially explain associations between sociodemographic factors and self-reported suicide attempts. This analysis was motivated by reports that certain personality traits are logical targets for intervention, whereas sociodemographic characteristics are not generally modifiable. Data were from a postal survey sent to community residents who were previously selected at random (N = 912). Age, gender, health-insurance status, education, self-reported health, and marital history were identified as relevant sociodemographic predictors of having made one or more lifetime suicide attempts. Risk associated with each of these variables was mediated by the personality traits of self-directedness (SD) and harm avoidance (HA). In a multiple logistic-regression analysis constrained to sociodemographic predictors, only young age, female sex, poor self-reported health, and Medicaid status remained as predictors of suicide attempts. When personality factors were added to the model, all of the sociodemographic predictors except Medicaid status were rendered nonsignificant or marginally significant. Risk associated with gender was primarily related to HA, risk associated with poor self-reported health was mediated by both HA and SD, and the risk associated with young age was primarily mediated by SD; the last was the largest mediation effect observed. In contrast, risk associated with receipt of Medicaid, presumed to indicate low socioeconomic status, was not mediated by personality. We conclude that risk associated with certain nonmodifiable demographic factors is often mediated by potentially modifiable intrapersonal factors, such as SD.
本研究的目的是确定人格是否可能部分解释社会人口学因素与自我报告的自杀未遂之间的关联。进行这项分析的动机是有报告称，某些人格特质是干预的合理目标，而社会人口学特征通常不可改变。数据来自一项邮寄调查，该调查发送给之前随机选取的社区居民（N = 912）。年龄、性别、医疗保险状况、教育程度、自我报告的健康状况和婚姻史被确定为有过一次或多次终生自杀未遂经历的相关社会人口学预测因素。与这些变量中的每一个相关的风险都由自我导向（SD）和避免伤害（HA）的人格特质介导。在一项限于社会人口学预测因素的多元逻辑回归分析中，只有年轻、女性、自我报告健康状况差和医疗补助状况仍然是自杀未遂的预测因素。当将人格因素添加到模型中时，除医疗补助状况外的所有社会人口学预测因素都变得不显著或接近显著。与性别相关的风险主要与HA有关，与自我报告健康状况差相关的风险由HA和SD共同介导，与年轻相关的风险主要由SD介导；最后一个是观察到的最大中介效应。相比之下，与接受医疗补助相关的风险（推测表明社会经济地位较低）并未由人格介导。我们得出结论，与某些不可改变的人口学因素相关的风险通常由潜在可改变的个人内部因素介导，例如SD。
