Discriminating schizophrenia disease progression using a P50 sensory gating task with dense-array EEG, clinical assessments, and cognitive tests

ABSTRACT

Background: Schizophrenia affects approximately 10% of the world’s population. Early detection of schizophrenia may significantly delay its progression. Although sensory gating deficits are reported in schizophrenia, it remains challenging how sensory gating deficits can be used with other metrics for risk detection and early diagnosis.

Research design and methods: Using EEG, the authors examined effects of sensory gating on the performance of 136 participants in a P50 sensory gating task, including patients with first-episode schizophrenia (FESZ), ultra-high risk (UHR) individuals, high-risk (HR) individuals, and age- and sex-matched healthy controls (HCs). The authors also explored the differences among all groups using clinical assessments and cognitive tests.

Results: Compared with HCs, HR, UHR and FESZ groups showed significant P50 suppression impairment. Furthermore, EEG source localization analyses identified successively stronger activation in prefrontal and anterior temporal regions in the HR, UHR and FESZ groups than in the HC group. Moreover, brain connectivity (HC < HR < UHR < FESZ) in the gamma band of P50 components was increasingly enhanced in accordance with the level of psychosis risks.

Conclusions: These findings suggest that EEG source imaging techniques, brain network dynamics, and behavioral tests, can help clearly distinguish different stages of schizophrenia, and may assist in the early diagnosis of schizophrenia.

KEYWORDS:

• Dense-array EEG
• early detection
• EEG source localization
• functional connectivity
• schizophrenia
• sensory gating

Article highlights

• Schizophrenia is considered to be a neurodevelopmental disorder associated with four different stages, including the risk stage, the ultra-high-risk or prodrome stage, the psychosis stage, and the chronic disability stage.

• Early detection and early intervention in schizophrenia may yield substantial improvements in outcomes.

• Dense-array electroencephalography (EEG), source imaging techniques, and brain network dynamics, together with behavioral tests, can help clearly distinguish the different stages among healthy individuals, those with psychosis risk, and those with schizophrenia.

• Auditory processing deficits, depicted by the P50 sensory gating index may be further explored along with clinical assessment scales and cognitive test results in designing biomarkers for deterioration (risk) detection and early diagnosis of neurodevelopmental disorders, including schizophrenia.

• Individuals at a more advanced stage of schizophrenia showed more severe cognitive impairments and poorer clinical performance.

• Compared with the healthy controls (HCs), individuals with psychosis risk (healthy risk (HR), ultra HR (UHR)) and patients with first episode schizophrenia (FESZ) exhibited significantly impaired P50 suppression (in the order of HR < UHR < FESZ), demonstrated successively stronger activation (HR < UHR < FESZ) in the temporal and frontal cortices, and showed increasingly enhanced functional connectivity (HR < UHR < FESZ) in the gamma band in the P50 components.

• Using the machine learning algorithm XGboost, we attained higher classification accuracy in the four groups (FESZ, UHR, HR, and HC) with combined features extracted from behavioral tests and the EEG test than by only using features from behavioral examinations. The sensitivity, and specificity of the four groups were also higher when using the combined features. The results indicate that the P50 index in combination with clinical and cognitive indices, may be a good biomarker for schizophrenia diagnosis and risk detection.

This box summarizes key points contained in the article.

Acknowledgments

We thank Dr. Joshua Ewen for his comments on the paper.

Declaration of interest

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Author contributions

Conceptualization, Chuanyue Wang and Jicong Zhang; methodology, Yu Luo; formal analysis, Yu Luo,Xiaohui Zhao, Qi Chang, and Hua Wang; resources, Chuanyue Wang and Changming Wang; writing—original draft preparation, Yu Luo; writing—review and editing, Yu Luo, Xiaohui Zhao and Jicong Zhang; supervision, Jicong Zhang and Chuanyue Wang; funding acquisition, Jicong Zhang and Chuanyue Wang. All authors reviewed and revised the manuscript and provided their approval of the final version of the manuscript. All authors agree to be accountable for all aspects of the work.

Reviewer disclosures

A peer reviewer on this manuscript has received manuscript or speaker’s fees from Astellas, Dainippon Sumitomo Pharma, Eli Lilly, Elsevier Japan, Janssen Pharmaceuticals, Kyowa Yakuhin, Meiji Seika Pharma, Mitsubishi Tanabe Pharma, MSD, Novartis, Otsuka Pharmaceutical, Shionogi, Tsumura, Wiley Japan, and Yoshitomi Yakuhin, and research grants from Eisai, Mochida Pharmaceutical, and Meiji Seika Pharma. Peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.

Additional information

Funding

This work was supported by the National Key Research and Development Program of China (Grant number: 2016YFF0201002); the Psychosis Project of the Beijing Key Laboratory (grant number Z161100002616017); the National Natural Science Foundation of China (grant numbers 61301005, 61572055 and 81501155); the Graduate Innovation Foundation for Beihang University (grant number YCSJ-01-2015-11); the Hefei Innovation Research Institute, Beihang University; and the Thousand Young Talent Plan Station (to J Zhang) between Beihang University and Jiangsu Yuwell Medical Equipment and Supply Co. Ltd.