Willem Van der Does A J, Dingemans P M, Linszen D H, Nugter M A, Scholte W F
Department of Psychiatry, Leiden University, Oegstgeest, The Netherlands.
J Nerv Ment Dis. 1995 Nov;183(11):681-7. doi: 10.1097/00005053-199511000-00002.
A sample of young patients with schizophrenia was studied in a longitudinal design to investigate the stability of an earlier reported four-dimensional symptom model (positive symptoms, negative symptoms, disorganization, and depression). Symptoms were measured in an acute phase soon after hospitalization, in a remitted stage 3 months later, and in a "mixed" phase 1 year after the second assessment. A different factor structure was found at each measurement, but a stable four-dimensional structure could be defined that closely resembled the factor structure of the first assessment. The symptom dimensions were uncorrelated, except disorganization and depression. Whether meaningful subtypes could be defined using statistical criteria also was investigated. With cluster analysis, four subtypes were defined that could be characterized as follows: rapid treatment responders, slower responders, and two relapse groups. Although some indication of the validity of these subtypes was found, it is concluded that there is a high risk of creating artificial boundaries with this approach.
对一组患有精神分裂症的年轻患者进行了纵向研究，以调查先前报道的四维症状模型（阳性症状、阴性症状、紊乱和抑郁）的稳定性。在住院后不久的急性期、3个月后的缓解期以及第二次评估后1年的“混合”期对症状进行测量。每次测量时发现了不同的因子结构，但可以定义一个与首次评估的因子结构非常相似的稳定的四维结构。除了紊乱和抑郁外，症状维度不相关。还研究了是否可以使用统计标准定义有意义的亚型。通过聚类分析，定义了四种亚型，其特征如下：快速治疗反应者、反应较慢者和两个复发组。尽管发现了这些亚型有效性的一些迹象，但得出的结论是，用这种方法创建人为界限的风险很高。
