不同家庭負載之思覺失調症病患的神經認知表現: 比較不同神經精神疾病或一般認知能力之多基因分數的預測

Abstract

背景與目的 認知功能障礙是思覺失調症的一個重要特徵，全基因組關聯研究發現有少數的基因位點同時和思覺失調症和認知功能顯著相關，但這些位點僅能解釋遺傳率的一小部分，思覺失調症患者之認知功能表現的基因架構目前仍不清楚。首先，此研究想探討思覺失調症患者的認知功能是否可以被思覺失調症、其他同樣有認知功能障礙的神經精神疾病相關的疾病位點，抑或是來自一般人和認知功能相關的位點所組成的基因分數解釋，並比較在不同家族負載之思覺失調症病患此相關性的強度；再者，分別在單發性家庭和多發性家庭比較不同程度持續注意力表現的思覺思調症患者之多基因分數。 方法 單發性家庭的樣本來自台灣思覺失調症三元體基因體計畫 (Schizophrenia Trio Genomics Research in Taiwan, S-TOGET)，其中納入1,649位思覺失調症患者與3,298位無思覺失調症的雙親，而多發性家庭的樣本來自台灣思覺失調症連鎖分析研究 (Taiwan Schizophrenia Linkage Study, TSLS)，此研究的收案條件為一個家庭中至少要有兩個患有思覺失調症的手足，從中納入581位患病手足及479位無病雙親。利用精神疾病微陣列晶片 (PsychChip) 做基因定型，且病患會接受神經認知功能測驗，分別為柯能氏持續表現作業 (Continuous Performance Test, CPT) 及威斯康辛卡片分類測驗 (Wisconsin Card Sorting Test, WCST)。利用柯能氏持續表現作業的表現將病患分成三組，分別為沒有認知功能障礙組、中度認知功能障礙組，以及重度認知功能障礙組，並比較他們的多基因分數之負載，之後，我們將神經功能測驗的多個指標利用驗證性因素分析集合成四個潛在變項 (CPT、WCST1、WCST2、Neurocognitive Performance)。最後，思覺失調症、其他神經精神疾病 (躁鬱症、阿茲海默症、泛自閉症障礙、注意力不足過動症)，以及一般認知能力 (教育程度、一般認知能力) 多基因分數的權重從相應的全基因組關聯研究之統合分析而來。 結果 多發性家庭的思覺失調症患者在柯能氏持續表現作業和威斯康辛卡片分類測驗的表現都較單發性家庭的患者差。在七種多基因分數下，思覺失調症的疾病狀態在單發性家庭中可以被思覺失調症、躁鬱症、泛自閉症障礙、教育程度、一般認知能力的多基因分數解釋；在多發性家庭中則可被思覺失調症的多基因分數解釋。在單發性家庭的病人中，只有教育程度和一般認知能力的多基因分數和WCST2潛在變數正相關。再者，單發性家庭中沒有認知功能障礙這組病患有最高的思覺失調症以及一般認知能力多基因分數。 結論 思覺失調症的神經認知表現最能被來自一般人和教育程度及一般認知能力之相關位點所組成的多基因分數解釋，而不是來自思覺失調症病人及其他神經精神疾病和疾病相關位點的多基因分數解釋。而病人的認知功能可能受修飾基因影響，且多發性家庭的認知表現可能涉及更多罕見遺傳變異，思覺失調症認知障礙的基因架構未來仍需進一步的研究。

Background Despite a few genetic variants overlap between neurocognitive deficits and schizophrenia (SZ) revealed by genome-wide association studies (GWAS), the genetic architecture influencing patients’ neurocognitive performance remains unclear. This study aimed to (1) examine whether the neurocognitive performance in SZ patients could be explained by the polygenic risk score (PRS) derived from schizophrenia versus that derived from other neuropsychiatric disorders or neurocognitive traits; (2) to examine the magnitude of the association of the PRS with the neurocognitive performance in schizophrenia patients from different familial loadings; and (3) to compare the PRS among three subgroups of schizophrenia patients classified by their magnitude of impairment in sustained attention. Methods Participants were 1649 sporadic cases and 3298 parents without SZ in simplex families from Schizophrenia Trio Genomics Research in Taiwan. For multiplex families with at least two SZ siblings, there were 581 co-affected probands and 479 parents without SZ from Taiwan Schizophrenia Linkage Study. All were genotyped using PsychChip and only patients underwent Continuous Performance Test (CPT) and Wisconsin Card Sorting Test (WCST). Patients was categorized into three groups based on their magnitude of impairment in sustained attention to compare their PRS. Confirmatory factor analysis of a four-latent model structure was performed to capture patients’ performance on CPT and WCST. Meta-analyses GWAS data of SZ, bipolar disorder (BD), Alzheimer's disease (AD), autism spectrum disorders (ASD), attention deficit hyperactivity disorder (ADHD), educational attainment (EA), and general cognitive ability (GCA) were used to derive corresponding PRS. Results SZ patients in multiplex families had worse scores than simplex ones on most CPT and WCST indices. Among the seven PRS, the phenotype of schizophrenia could be predicted by SZ-PRS, BD-PRS, ASD-PRS, EA-PRS, and GCA-PRS in simplex families and by SZ-PRS in multiplex families. Only EA-PRS and GCA-PRS were significantly associated with higher WCST2 factors among patients with schizophrenia in simplex families. Furthermore, no impairment group in simplex families had the highest GCA-PRS and SZ-PRS. Conclusions The neurocognitive performance of schizophrenia patients was best explained by the general cognitive abilities PRS derived from healthy individuals rather than the schizophrenia and other neuropsychiatric disorders PRS derived from patients with neuropsychiatric disorders. Neurocognitive deficits in schizophrenia patients may involve modifier genes. Other genetic architecture underlying schizophrenia’s cognitive impairment warrants further investigation.