Lenzenweger Mark F, Johnson Matthew D, Willett John B
Department of Psychology, State University of New York at Binghamton, Binghamton 13902-6000, USA.
Arch Gen Psychiatry. 2004 Oct;61(10):1015-24. doi: 10.1001/archpsyc.61.10.1015.
The long-term stability of personality pathology remains an open question. Its resolution will come from prospective, multiwave longitudinal studies using blinded assessments of personality disorders (PD). Informative analysis of multiwave data requires the application of statistical procedures, such as individual growth curve modeling, that can detect and describe individual change appropriately over time. The Longitudinal Study of Personality Disorders, which meets contemporary methodological design criteria, provides the data for this investigation of PD stability and change from an individual growth curve perspective.
Two hundred fifty subjects were examined for PD features at 3 different time points using the International Personality Disorders Examination during a 4-year study. Stability and change in PD features over time were examined using individual growth modeling.
Fitting of unconditional growth models indicated that statistically significant variation in PD features existed across time in the elevation and rate of change of the individual PD growth trajectories. Fitting of additional conditional growth models, in which the individual elevation and rate-of-change growth parameters were predicted by subjects' study group membership (no PD vs possible PD), sex, and age at entry into the study, showed that study group membership predicted the elevation and rate of change of the individual growth curves. Comorbid Axis I psychopathology and treatment during the study period were related to elevations of the individual growth trajectories, but not to rates of change.
From the perspective of individual growth curve analysis, PD features show considerable variability across individuals over time. This fine-grained analysis of individual growth trajectories provides compelling evidence of change in PD features over time and does not support the assumption that PD features are traitlike, enduring, and stable over time.
人格病理学的长期稳定性仍是一个悬而未决的问题。其解决方案将来自前瞻性、多波次纵向研究，该研究需对人格障碍（PD）进行盲法评估。对多波次数据进行有效分析需要应用统计程序，如个体生长曲线建模，以能恰当检测和描述个体随时间的变化。符合当代方法学设计标准的人格障碍纵向研究，为从个体生长曲线角度对PD稳定性和变化进行的这项调查提供了数据。
在一项为期4年的研究中，使用国际人格障碍检查表在3个不同时间点对250名受试者的PD特征进行检查。使用个体生长建模来检查PD特征随时间的稳定性和变化。
无条件生长模型的拟合表明，个体PD生长轨迹的高度和变化率在不同时间存在统计学上的显著差异。进一步的条件生长模型拟合显示，个体高度和变化率生长参数由受试者的研究组成员身份（无PD与可能有PD）、性别和研究开始时的年龄预测，结果表明研究组成员身份可预测个体生长曲线的高度和变化率。研究期间的共病轴I精神病理学和治疗与个体生长轨迹的高度有关，但与变化率无关。
从个体生长曲线分析的角度来看，PD特征随时间在个体间表现出相当大的变异性。对个体生长轨迹的这种细粒度分析提供了令人信服的证据，证明PD特征随时间发生了变化，并不支持PD特征是特质性的、持久的且随时间稳定的这一假设。
