| 摘要 | 精神分裂症患者存在广泛的认知缺损，其中语义和注意缺损尤为严重。语义缺损是预测个体是否会发展成精神分裂症的关键认知指标。注意缺陷主要体现在对目标刺激的选择性注意，和对干扰刺激的控制能力上。同时，选择性注意和干扰控制与语义加工间也存在着紧密的联系。患者的一些特征在普通人群中也能够观察到，这些特征被称为分裂型特质。已有研究表明普通人群中分裂型特质明显的个体也存在语义和注意加工异常。分裂型特质包括认知-知觉性、人际性和无组织性维度，目前尚不清楚（1）分裂型特质的哪些维度和语义加工、选择性注意或者干扰控制存在关联；（2）这种关联是否存在跨时间的稳定性。
为此，本研究考察了分裂型特质及其不同维度与语义、选择性注意和干扰控制的关联及其稳定性。在时间点1招募70名健康大学生，使用分裂型特质问卷测量其分裂型特质水平，并使用语义启动范式（研究一）、视觉双刺激oddball范式（研究二）和Stroop范式（研究三）来分别考察被试的语义加工、选择性注意和干扰控制特征，记录被试的脑电，并在三个月后完成重测（时间点2）。
研究一发现显著的语义启动效应：被试在语义无关条件的反应时大于语义相关条件（F(1,42) = 43.19, p < .01, η2 = .51）；语义无关条件的N400平均波幅大于语义相关条件（F(1,42) = 210.16, p < .01, η2 = .83）。被试的反应时（语义无关：r = .83, p < .01; 语义相关：r = .74, p < .01），N400潜伏期（语义无关：r ≥ .48, p < .01）和平均波幅（语义无关：r ≥ .75, p < .01; 语义相关：r ≥ .56, p < .01）重测信度较好。在时间点2，被试的认知-知觉性（r = .45, p <.01）和无组织性（r = .31, p = .04）与语义无关条件的N400潜伏期显著相关，即认知-知觉性和无组织性越高，语义加工效率越差。
研究二发现被试目标刺激的P300平均波幅显著大于标准刺激（F(1,50) = 10.51, p < .01, η2 = .17）。P300潜伏期（标准刺激：r ≥ .41, p < .01; 目标刺激：r ≥ .24, p < .05）和平均波幅（标准刺激：r ≥ .46, p < .01; 目标刺激：r ≥ .60, p < .01）重测信度较好。被试的人际性与目标刺激条件的P300平均波幅显著负相关（时间点1：r = .25, p = .04; 时间点2：r = .27, p = .04），即人际性越高，选择性注意的分配越差。
研究三发现显著的Stroop效应：被试字色不一致的反应时显著大于字色一致条件（F(1,50) = 135.42, p < .01, η2 = .73）；字色不一致的N450平均波幅显著小于字色一致条件（F(1,50) = 10.51, p < .01, η2 = .17）。被试的反应时（字色不一致：r ≥ .59, p < .01; 字色一致：r ≥ .59, p < .01）和N450平均波幅（字色不一致：r ≥ .53, p < .01; 字色一致：r ≥ .54, p < .01）重测信度良好。在时间点2，被试的人际性与字色不一致的N450潜伏期显著负相关（r = - .30, p = .03），即人际性越高，排除干扰信息和冲突解决越差。
比较三个实验范式和ERP成分的重测结果，发现ERP平均波幅的重测信度高于潜伏期；P300的重测信度高于N400和N450。
综上所述，分裂型特质与选择性注意的关联在亚临床群体中稳定存在，反映选择性注意的P300成分最有潜力成为精神分裂症谱系的生物学标记。文章对分裂型特质与语义和注意加工的关联，以及三个常见的实验范式的重测信度进行了讨论。 |
| 其他摘要 | Patients with schizophrenia have a wide range of cognitive impairments, of which semantic and attention deficits are particularly serious. Semantic processing a key cognitive indicator that predicts whether an individual will develop schizophrenia. Attention impairments in schizophrenia mainly reflect in selective attention to the target stimuli and interference control to the deviant stimuli. Meanwhile, selective attention and interference control are involved with semantic processing. Some characteristics of patients can also be observed in the general population called schizotypal traits. Studies showed that individuals with schizotypal traits also have abnormal semantic and attention processing. Schizotypal traits generally aggregate into three dimensions, namely the cognitive-perceptual, the interpersonal and the disorganised dimensions. However, it remains unclear (1) which dimensions of the schizotypal traits are related to semantic processing, selective attention or interference control; (2) whether these relationships are stable across time.
To solve the above questions, the present study investigated the relationship and the stability of schizotypal traits and their different dimensions with semantic processing, selective attention, and interference control. At time 1, 70 healthy college students were recruited. The schizotypal Personality Questionnaire (SPQ) was used to measure the schizotypal traits of each participant and the semantic priming task (Study 1), the visual two-stimulus oddball task (Study 2) and the Stroop task (Study 3) were used to measure semantic processing, selective attention and interference control, respectively. EEG data were recorded. Participants were retested after three months.
Study 1 found significant semantic priming effect at both time points: the reaction times (RTs) were faster on semantically related words than unrelated ones (F(1,42) = 43.19, p < .01, η2 = .51); the N400 amplitudes of unrelated words were more negative than related ones (F(1,42) = 210.16, p < .01, η2 = .83). RTs (unrelated: r = .83, p < .01; related: r = .74, p < .01), N400 latenices (unrelated: r ≥ .48, p < .01) and mean amplitudes (unrelated: r ≥ .75, p < .01; related: r ≥ .56, p < .01) had high retest reliability between two time poins. At Time 2, the cognitive-perceptual (r = .45, p <.01) and the disorganised (r = .31, p = .04) dimensions were correlated with N400 latencies of unrelated words, indicating low semantic processing efficiency in individual with high cognitive-perceptual and disorganised traits.
Study 2 found that P300 mean amplitudes for the target stimuli were greater than those for standard stimuli at two time points (F(1,50) = 10.51, p < .01, η2 = .17). P300 latencies (standard: r ≥ .41, p < .01; target: r ≥ .24, p < .05) and P300 amplitudes (standard: r ≥ .46, p < .01; target: r ≥ .60, p < .01) both showed good retest reliability. The interpersonal dimension was negatively correlated with P300 amplitudes for targets at both time points (Time 1: r = .25, p = .04; Time 2: r = .27, p = .04), reflecting a significant and stable correlation between schizotypal traits and selective attention.
Study 3 found significant Stroop effect at both time points: RTs were faster on congruent words than incongruent ones (F(1,50) = 135.42, p < .01, η2 = .73); N450 amplitudes for incongruent words were more negative than those for congruent ones (F(1,50) = 23.24, p < .01, η2 = .32). RTs (incongruent: r ≥ .59, p < .01; congruent: r ≥ .59, p < .01) and N450 amplitudes (incongruent: r ≥ .53, p < .01; congruent: r ≥ .54, p < .01) showed high test-retest reliability. At time point 2, the interpersonal dimension was negatively correlated with N450 latencies for incongruent words (r = - .30, p = .03), indicating low interference control and conflict resolution in individuals with higher interpersonal traits.
The study also compared the reliability of three tasks and ERP components over a three-month interval. The retest reliabilities of ERP amplitudes were higher than latencies. Besides, the test-retest reliabilities of P300 were higher than N400 and N450, suggesting that P300 had the highest temporal stability among different ERP components.
In conclusion, the relationship between selective attention and schizotypal traits could be reliably observed in subclinical groups. P300, which reflects selective attention, has the most potential to become a biological marker for the schizophrenia spectrum. The study discussed the relationship of schizotypal traits with semantic and attention processing, and the retest reliability of three common ERP components. |
修改评论