Exploring Neuroimaging-Genetic Co-Alteration Features of Auditory Verbal Hallucinations in Different Subjects for the Establishment of a Predictive Model
Imagine hearing voices when no one is around—clear, vivid, and often distressing. This is auditory verbal hallucination (AVH), a experience that touches millions of lives: over 70% of people with schizophrenia, up to 46% of those with borderline personality disorder (BPD), 11–62% of individuals with bipolar disorder (BP), and even 4.2% of healthy people report AVHs. Yet early misdiagnosis is common, delaying proper care and worsening physical, emotional, and safety risks—from treatment side effects to increased suicide or self-harm.
For years, researchers have struggled to unlock AVH’s roots. Because AVH is tied to human language and speech, no animal models exist—meaning we can only study it in living people. Now, a team of scientists from China’s leading mental health and neuroscience labs is proposing a breakthrough approach: combining neuroimaging (brain scans) and genomics (gene studies) to identify shared and unique AVH features across different groups—and build a predictive model to guide early, precise care.
The Problem: Why AVH Is Hard to Treat Early
AVH isn’t a “one-size-fits-all” condition. Its content, severity, and impact vary wildly: a person with schizophrenia might hear threatening commands, while a healthy individual could experience neutral or comforting voices. This diversity makes early diagnosis tricky—and misdiagnosis can lead to treatments that don’t work or cause harm.
Worse, AVH’s biological roots are complex. It involves both brain network disruptions (like faulty connections in the “salience network,” which helps prioritize important stimuli) and genetic susceptibility (genes that increase risk or affect treatment response). Until now, these pieces have been studied in isolation—not together.
The Breakthrough Idea: Neuroimaging + Genetics = A “Bridge” to Understanding
The research team—led by Lang-Lang Cheng (Wenzhou Seventh People’s Hospital) and Chuan-Jun Zhuo (Tianjin Mental Health Centre)—argues that AVH’s key lies in neuroimaging-genetic co-alterations: the link between a person’s genes, their brain structure/function, and their clinical symptoms.
Here’s the hypothesis:
- Genes (endogenous factors) and environment shape brain features (like the strength of connections in language or default-mode networks).
- Over time, these brain changes lead to clinical AVH (the voices people hear).
This “bridge” between genes, brain, and behavior could be the biomarker clinicians need to diagnose AVH early—before it escalates—and tailor treatments to a person’s unique biology.
What We Already Know: Clues from Neuroimaging and Genetics
Recent studies back this approach:
- Neuroimaging: fMRI and EEG show that AVH is tied to abnormal activity in the salience network (schizophrenia patients) and theta-wave changes (a brain rhythm linked to memory and attention). Healthy “voice hearers” even show unusual self-generated speech patterns on EEG.
- Genetics: The FOXP2 gene (linked to language development) increases AVH risk in schizophrenia. COMT and NRG1 genes affect how well treatments like transcranial direct current stimulation (tDCS) work for AVH.
- Machine learning: Algorithms using brain scan data already predict AVH in schizophrenia with 73.9% specificity—proof that pattern recognition can work.
The Proposed Research: A Pathway to Early, Precise Care
The team’s plan has two core parts:
- Map Common and Unique AVH Features: Study first-episode, untreated people with schizophrenia, BP, PTSD, BPD, or major depressive disorder (MDD)—and healthy individuals experiencing their first AVH. Use high-throughput sequencing (to analyze genes) and human connectome techniques (to map brain connections) to find features shared across all groups and those unique to each diagnosis. For example: Do schizophrenia and BPD share a faulty salience network, but differ in genetic risk factors?
- Build a Predictive Model: Use machine learning to turn these features into a tool for early diagnosis and treatment prediction. The model would help clinicians:
- Identify AVH before it becomes severe.
- Match patients to treatments (e.g., clozapine for schizophrenia, tDCS for BPD) based on their unique neuroimaging-genetic profile.
Why This Matters: From Research to Real-World Impact
The study’s biggest innovation is its focus on first-episode patients and healthy people. By capturing AVH at its earliest stages—before treatment or long-term brain changes—researchers can isolate the “root” features of AVH, not just its consequences. A 2-year follow-up will also track how brain and genetic markers change with treatment, helping identify precise targets (e.g., a specific brain network to stimulate with TMS).
For patients, this means:
- Early diagnosis: No more waiting for severe symptoms to get help.
- Precise treatment: Less trial and error with medications or therapies.
- Reduced suffering: Lower risks of side effects, suicide, or self-harm.
The Team Behind the Work
The study was led by researchers from:
- Wenzhou Seventh People’s Hospital (PNG-Lab, a leader in psychiatric neuroimaging-genetics).
- Tianjin Mental Health Centre (specializing in neuroimaging and comorbidity).
- Jining Medical University and Hebei Medical University (psychiatry and pharmacy expertise).
Funding comes from China’s top scientific agencies, including the National Natural Science Foundation of China and Tianjin Health Bureau.
The Road Ahead
While the proposed pathway is promising, the team acknowledges gaps—like how environmental factors (e.g., childhood trauma) interact with genes and brain networks. They invite global collaboration to refine the model and test it in diverse populations.
For now, the work offers a hopeful path forward: by combining two of neuroscience’s most powerful tools—brain mapping and genetics—we can move beyond guesswork to treat AVH with the precision it deserves.
This article is based on research published in the Chinese Medical Journal (2019) by Lang-Lang Cheng, Guo-Wei Wang, Yan-Chi Zhang, Gong-Ying Li, Hong-Jun Tian, Li-Na Wang, Xiu-Hai Sun, Chun-Hua Zhou, and Chuan-Jun Zhuo.
doi.org/10.1097/CM9.0000000000000385
Was this helpful?
0 / 0