Division of Psychology, Nottingham Trent University, Burton Street, Nottingham, UK.
Alcohol Alcohol. 2010 Nov-Dec;45(6):563-72. doi: 10.1093/alcalc/agq052. Epub 2010 Sep 27.
To identify population-based clinical and demographic correlates of alcohol use dimensions.
Using data from a population-based sample of Great Britain (n = 7849), structural equation modelling (SEM) was used to identify associations between demographic and clinical variables and two competing dimensional models of the Alcohol Use Disorders Identification Test (AUDIT).
A two-factor SEM fit best. In this model, Factor 1, alcohol consumption, was associated with male sex, younger age, lower educational attainment, generalized anxiety disorder (GAD) and suicide attempts. Factor 2, alcohol-related problems, was associated with the demographic variables (to a lesser extent) and to a wider range of clinical variables, including depressive episode, GAD, mixed anxiety and depressive disorder, obsessive compulsive disorder, phobia, suicidal thoughts and suicide attempts. The one-factor SEM was associated with demographic and all assessed clinical correlates; however, this model did not fit the data well.
Two main conclusions justify the two-factor approach to alcohol use classification. First, the model fit was considerably superior and, second, the dimensions of alcohol consumption and alcohol-related problems vary considerably in their associations with measures of demographic and clinical risk. A one-factor representation of alcohol use, for instance, would fail to recognize that measures of affective/anxiety disorders are more consistently related to alcohol-related problems than to alcohol consumption. It is suggested therefore that to fully understand the complexity of alcohol use behaviour and its associated risk, future research should acknowledge the basic underlying dimensional structure of the construct.
确定基于人群的酒精使用维度的临床和人口统计学相关性。
使用来自英国基于人群的样本（n=7849）的数据，采用结构方程模型（SEM）来确定人口统计学和临床变量与两种竞争性的酒精使用障碍识别测试（AUDIT）维度模型之间的关联。
双因素 SEM 拟合效果最佳。在这个模型中，第一个因素是饮酒量，与男性、年龄较小、受教育程度较低、广泛性焦虑障碍（GAD）和自杀企图有关。第二个因素是与人口统计学变量（在较小程度上）和更广泛的临床变量有关的酒精相关问题，包括抑郁发作、GAD、混合性焦虑和抑郁障碍、强迫症、恐惧症、自杀念头和自杀企图。单因素 SEM 与人口统计学和所有评估的临床相关性有关；然而，这个模型与数据拟合得不太好。
两个主要结论证明了酒精使用分类的双因素方法是合理的。首先，模型拟合度显著提高，其次，饮酒量和酒精相关问题的维度在与人口统计学和临床风险的测量指标的关联方面存在很大差异。例如，将酒精使用表示为一个因素将无法认识到情感/焦虑障碍的测量指标与酒精相关问题的关系比与饮酒量的关系更为一致。因此，为了充分理解酒精使用行为及其相关风险的复杂性，未来的研究应该承认该结构的基本潜在维度结构。
