Department of Clinical and Social Sciences in Psychology, University of Rochester, 453 Meliora Hall, Rochester, NY 14607, United States of America.
Department of Clinical and Social Sciences in Psychology, University of Rochester, 453 Meliora Hall, Rochester, NY 14607, United States of America.
Schizophr Res. 2019 Jun;208:258-267. doi: 10.1016/j.schres.2019.01.046. Epub 2019 Feb 5.
Outcomes for people with schizophrenia-spectrum disorders (SSDs) are generally poor, making it important to understand risk states and illness transition. The network approach, which conceptualizes psychopathology as a network of causally interacting symptoms, may hold promise in this regard. Here, we present a network analysis of schizotypal personality traits (i.e., schizophrenia-like cognitive, perceptual, affective, interpersonal, and behavioral anomalies that may index one's vulnerability for a SSD) using an international sample. We analyzed data from 9505 participants between the ages of 14-70 who completed the Schizotypal Personality Questionnaire-Brief on TestMyBrain.org. In line with other research, we find that the network of schizotypal traits is densely connected, characterized by three communities of items-interpersonal (I), disorganized (D), cognitive-perceptual (CP)-with I and D features exhibiting the greatest centrality (z-scored M strength: I = 0.56, D = 0.29, CP = -0.84; expected influence: I = 0.54, D = 0.33, CP = -0.84) and predictability (M I = 0.37, D = 0.43, CP = 0.23). Importantly, within our sample, we found the estimated network to be replicable (Network Comparison Test: network structure difference: M = 0.304, p = .420; global strength difference: S = 0.904, p = .530), and estimates of node centrality to be stable (correlation-stability coefficient = 0.75). Further, we find network differences between certain groups differing in levels of SSD risk as a function of age (network structure: difference M = 0.562, p < .001; global strength difference: S = 3.483, p = .012) and ethnic minority status (global strength difference: S = 11.488, p = .004). Together, these findings demonstrate the utility of using network approaches to understand SSD risk states as well as the replicability of network findings on schizotypal personality traits and related SSD risk concepts.
精神分裂症谱系障碍（SSDs）患者的预后通常较差，因此了解风险状态和疾病进展非常重要。网络方法将精神病理学概念化为因果相互作用的症状网络，在这方面可能具有前景。在这里，我们使用国际样本对精神分裂症人格特质（即与 SSD 易感性相关的精神分裂症样认知、感知、情感、人际和行为异常）进行了网络分析。我们分析了在 TestMyBrain.org 上完成简短精神分裂症人格问卷的 9505 名年龄在 14 至 70 岁之间的参与者的数据。与其他研究一致，我们发现精神分裂症特质网络连接紧密，由三个项目社区组成-人际关系（I）、混乱（D）、认知感知（CP）-其中 I 和 D 特征具有最大的中心性（z 得分 M 强度：I=0.56，D=0.29，CP=-0.84；预期影响：I=0.54，D=0.33，CP=-0.84）和可预测性（M I=0.37，D=0.43，CP=0.23）。重要的是，在我们的样本中，我们发现估计的网络是可复制的（网络比较测试：网络结构差异：M=0.304，p=0.420；全局强度差异：S=0.904，p=0.530），并且节点中心性的估计值是稳定的（相关性稳定性系数=0.75）。此外，我们发现某些群体之间的网络差异，这些群体的 SSD 风险水平随年龄而变化（网络结构：差异 M=0.562，p<0.001；全局强度差异：S=3.483，p=0.012）和少数族裔地位（全球实力差异：S=11.488，p=0.004）。总的来说，这些发现表明使用网络方法来了解 SSD 风险状态以及在精神分裂症人格特质和相关 SSD 风险概念上复制网络发现是有用的。
