Clark Duncan B, Jones Bobby L, Wood D Scott, Cornelius Jack R
Center for Education and Drug Abuse Research, University of Pittsburgh, and Western Psychiatric Institute and Clinic, PA 15213, USA.
Addict Behav. 2006 Jun;31(6):995-1009. doi: 10.1016/j.addbeh.2006.03.016. Epub 2006 May 3.
Substance use disorders (SUDs) may be characterized by onset age, severity, substance type, course, and outcomes. SUD phenotypes in the literature typically consider each of these features in isolation. Conceptual frameworks and data collection procedures for assessing SUD phenotypes are increasingly "diachronic" in approach, providing for characterizations "throughout time". The recent availability of statistical procedures for the identification of latent classes offers the possibility of developing SUD phenotypes integrating these developmental features. This article illustrates the utilization of SAS-TRAJ mixture modeling to characterize variations in SUD symptom trajectories to define phenotypes.
The subjects were 332 adult males with SUDs. Their course of symptoms from early adolescence through middle adulthood was retrospectively determined. Symptom trajectories were defined by the number of DSM-IV SUD symptoms by year of age. SAS-TRAJ mixture models identified trajectory classes. Model development, evaluation, and selection using this approach are discussed.
Among these men with SUDs, six trajectory classes were identified, including groups characterized by early-onset and severe SUD symptoms persisting into adulthood, an early-onset group similar in adolescence but improving in adulthood, and other groups with symptoms emerging later with varying degrees of severity and persistence. The SUD trajectory classes were significantly different on comorbid psychopathology, particularly childhood disruptive behavior disorders.
The results present a new method for the comprehensive depiction of heterogeneity in SUD symptoms. Future studies may determine the extent to which SUDs phenotypes based on the course of symptom development inform etiology, prevention and treatment research.
物质使用障碍（SUDs）可通过发病年龄、严重程度、物质类型、病程和结局来表征。文献中的SUD表型通常孤立地考虑这些特征中的每一个。评估SUD表型的概念框架和数据收集程序在方法上越来越“历时性”，提供“贯穿时间”的特征描述。最近用于识别潜在类别统计程序的可用性为整合这些发育特征来开发SUD表型提供了可能性。本文说明了利用SAS-TRAJ混合模型来表征SUD症状轨迹的变化以定义表型。
研究对象为332名患有SUDs的成年男性。回顾性确定了他们从青春期早期到成年中期的症状病程。症状轨迹由按年龄逐年统计的DSM-IV SUD症状数量来定义。SAS-TRAJ混合模型识别轨迹类别。讨论了使用这种方法进行模型开发、评估和选择。
在这些患有SUDs的男性中，识别出六个轨迹类别，包括以持续到成年期的早发性和严重SUD症状为特征的组、在青春期相似但成年期有所改善的早发性组，以及其他症状出现较晚且严重程度和持续时间各不相同的组。SUD轨迹类别在共病精神病理学方面存在显著差异，尤其是儿童期破坏性行为障碍。
研究结果提出了一种全面描述SUD症状异质性的新方法。未来的研究可能会确定基于症状发展过程的SUD表型在多大程度上为病因学、预防和治疗研究提供信息。
