Faraone S V, Kremen W S, Lyons M J, Pepple J R, Seidman L J, Tsuang M T
Harvard Medical School Department of Psychiatry, Massachusetts Mental Health Center, Boston, USA.
Am J Psychiatry. 1995 Sep;152(9):1286-90. doi: 10.1176/ajp.152.9.1286.
Numerous studies suggest that the nonschizophrenic relatives of schizophrenic patients exhibit psychiatric and other features that discriminate them from normal comparison subjects. These features have been put forth as "spectrum" phenotypes that may be variant manifestations of the schizophrenia genotype. However, most of these studies do not address a key measurement question: does the diagnostic accuracy of these spectrum classifications warrant their use in genetic linkage studies of schizophrenia?
The authors reviewed 30 studies of putative indicators of the schizophrenic genotype: schizotypal personality disorder, eye tracking dysfunction, attentional impairment, auditory evoked potentials, neurological signs, neuropsychological impairment, and allusive thinking.
Although each of 42 measures of these indicators discriminated the relatives of schizophrenic patients from the normal comparison subjects, a diagnostic accuracy analysis suggested that only six of these would improve the informativeness of genetic linkage data.
Many proposed spectrum phenotypes for schizophrenia may not be useful for linkage analysis because of high false positive rates (poor specificity). Future work aimed at describing and developing phenotypes for linkage analysis should assess the diagnostic accuracy of proposed measures.
大量研究表明，精神分裂症患者的非精神分裂症亲属表现出一些精神方面及其他特征，这些特征将他们与正常对照对象区分开来。这些特征已被提出作为“谱系”表型，可能是精神分裂症基因型的变异表现形式。然而，这些研究大多未解决一个关键的测量问题：这些谱系分类的诊断准确性是否足以使其用于精神分裂症的基因连锁研究？
作者回顾了30项关于精神分裂症基因型假定指标的研究：分裂型人格障碍、眼动跟踪功能障碍、注意力损害、听觉诱发电位、神经体征、神经心理学损害及隐喻性思维。
尽管这些指标的42项测量中的每一项都能将精神分裂症患者的亲属与正常对照对象区分开来，但诊断准确性分析表明，其中只有6项能提高基因连锁数据的信息量。
由于假阳性率高（特异性差），许多提出的精神分裂症谱系表型可能对连锁分析无用。未来旨在描述和开发用于连锁分析的表型的工作应评估所提出测量方法的诊断准确性。
