Schizophrenia and Psychosis After Traumatic Brain Injury: Research Insights from Ting-Yi Chu, MD

A new meta-analysis showed significant associations between traumatic brain injury (TBI) and psychosis, finding a pooled hazard ratio of 3.55 for the 11 studies included in analysis.¹ Further subgroup analysis of the studies showed increased risk across groupings of sex and age, with highest risk among adults and moderate-to-severe TBI cases. Psychiatric Times sat down with physician and lead investigator Ting-Yi Chu, MD to learn more.

Psychiatric Times: What background in the field indicated that there may be a relationship between TBI and schizophrenia/psychosis?

Ting-Yi Chu, MD: The inspiration for this study grew out of a clinical observation rather than a purely academic question. In geriatric medicine and public health, we frequently emphasize fall prevention in the elderly because of the risk of traumatic brain injury. At the same time, I was learning about the cognitive and behavioral disturbances associated with dementia, including psychotic symptoms such as delusions and hallucinations. It struck me that aging, dementia, and TBI all independently contribute to cognitive decline, and when these factors converge in the same patient, the clinical picture becomes remarkably complex. This overlap raised a question in my mind: could TBI itself serve as a risk factor for schizophrenia or other psychotic disorders, distinct from the cognitive decline we typically attribute to aging or dementia?

This question felt particularly urgent because awareness of the distinction between dementia-related behavioral changes and emerging primary psychotic disorders may be limited—not only among the general public, but also within the medical community. If these conditions are being conflated or underrecognized, a meaningful proportion of TBI patients with emerging psychosis may be falling through diagnostic cracks.

From a scientific standpoint, 2 prior meta-analyses, Molloy et al (2011) and Yau et al (2024), had identified a significant association between TBI and psychosis, but several questions remained open. No clear dose-response relationship between TBI severity and psychosis risk had been established, and demographic and clinical modifiers such as age, sex, and follow-up duration had not been systematically examined. Our study was designed to address these gaps through a more stratified analysis. We used both odds ratios and hazard ratios—one captures cumulative risk, the other captures how risk unfolds over time.

PT: How can psychiatric clinicians utilize your findings in their practice?

Chu: I believe the most immediate takeaway is that TBI should be treated as a relevant risk factor in the psychiatric history, much like family history or substance use. When a patient presents with new-onset psychotic symptoms, asking about prior head injuries, including concussions that may have seemed minor at the time, could provide valuable diagnostic context.

One challenge I have observed is that TBI patients tend to remain primarily within neurology or rehabilitation pathways. When behavioral or psychiatric symptoms emerge, both clinicians and families may attribute these changes to the expected consequences of brain injury itself, rather than considering the possibility of a distinct psychiatric disorder. Even when psychiatric care is involved, the focus often leans toward pharmacological management rather than comprehensive evaluation and intervention such as cognitive behavioral therapy or structured psychosocial support. This can create what I think of as a gray zone in medical coverage: the patient is being treated for TBI, but an emerging psychotic disorder goes unrecognized or inadequately addressed.

Our findings point to several practical considerations. First, psychiatric screening should be part of routine follow-up for TBI patients, particularly those with moderate-to-severe injuries. Second, monitoring should extend well beyond the acute recovery period: our data show that risk continues to rise even beyond 5 years post-injury. Third, both male and female TBI survivors are at elevated risk; surveillance should not be limited to male patients, as earlier literature might suggest.

Ultimately, this does not require sophisticated tools—it requires awareness. The clinical instinct to consider psychosis as part of the differential, even when the presentation might initially be attributed to TBI-related cognitive decline, can make a meaningful difference.

PT: Can you share more on the relationship between factors like age, TBI severity, and follow-up with psychosis?

Chu: Certainly. I want to begin by noting that our subgroup analyses should be interpreted as exploratory rather than definitive. The number of studies reporting stratified data was limited, and some subgroups were based on only 2 or 3 studies. With that caveat, several patterns emerged that I believe warrant attention.

Regarding age, adults with TBI showed a stronger association with subsequent psychosis than minors (pooled odds ratio (OR) of 2.99 versus 1.25), and this difference was statistically significant. However, even the modest elevation in minors reached significance, suggesting that TBI may elevate risk even when baseline rates of psychosis are inherently low. The smaller effect in children may partly reflect the rarity of psychotic disorders in pediatric populations. Children may also benefit from greater neuroplasticity, which could serve as a partial protective factor, though this needs further study. What concerns me clinically is the follow-up implication: psychosis typically emerges in late adolescence or early adulthood, so a child who sustains a TBI at age 10 may not manifest symptoms for another decade. Standard pediatric follow-up may not capture these delayed presentations.

For TBI severity, we observed what appears to be a dose-response gradient. Moderate-to-severe TBI carried higher risk estimates (OR = 5.42; hazard ratio (HR) = 8.35) compared to mild TBI (OR = 3.59; HR = 4.45). The difference between groups did not reach statistical significance, so I would call this a suggestive trend rather than a confirmed finding. Importantly, even mild TBI—including concussion—conferred significant risk. This matters clinically because mild TBI is far more common and often goes unmonitored after the acute phase.

Perhaps the most striking finding relates to follow-up duration. In our hazard ratio analyses, the pooled estimate rose from 1.58 for studies with less than one year of follow-up to 7.43 for those exceeding 5 years, and this difference was statistically significant. One might argue that longer observation simply allows more cases to accumulate. But hazard ratios inherently account for follow-up time, which suggests a genuine escalation in risk, not just case accumulation. This pattern is consistent with models of progressive neurodegeneration after TBI, including tau accumulation and hippocampal atrophy, which unfold over years.

Taken together, these patterns suggest that a one-size-fits-all approach to post-TBI psychiatric monitoring is unlikely to be sufficient. Age at injury, severity of trauma, and time since injury all appear to matter, and follow-up strategies should be tailored accordingly.

PT: How does your work highlight the need for monitoring and thorough assessment around TBI and psychiatric issues?

Chu: The core message is that TBI-associated psychosis is not exclusively an acute or subacute phenomenon—it can emerge years, even decades, after the initial injury. This has direct implications for how we structure follow-up.

Psychosis tends to develop gradually. In differential diagnosis, it is often (and reasonably) considered only after more immediately life-threatening conditions have been ruled out. That is appropriate. But this reasonable hierarchy should not lead to psychotic disorders being permanently overlooked. What I believe is most needed is not a new screening tool, but a shift in awareness: clinicians managing TBI patients should keep psychosis on their radar, even when symptoms appear years after the injury and even when other explanations seem available.

There are also feasible, low-cost approaches to improving detection. Older adult patients, for example, often already undergo periodic cognitive assessments. Adding a brief psychiatric screening component to these existing assessments could help identify emerging psychotic symptoms without significantly increasing clinical burden. Psychotic symptoms, when present, tend to be clinically conspicuous; the challenge is not that they are subtle, but that clinicians may not be thinking to look for them in the context of TBI.

Additionally, TBI patients frequently move between care settings: from the emergency department, to neurosurgery or neurology, to rehabilitation, and eventually to primary care. Psychiatric follow-up can easily be lost in these transitions. Building psychiatric screening into long-term TBI management as a standard component, rather than relying on individual clinicians to remember, could help close this gap.

PT: Is there a proposed neurobiological basis for the relationship between TBI and psychosis/schizophrenia?

Chu: Yes, and our findings are broadly consistent with the emerging framework. The proposed mechanisms operate across multiple time scales.

In the acute phase, TBI induces diffuse axonal injury and demyelination, disrupting neuronal integrity and setting the stage for downstream pathology. In the subacute to medium-term phase, progressive damage—particularly in the temporal and parietal lobes—has been linked to deficits in cognition, emotional processing, and semantic comprehension, all of which are implicated in psychosis. Cortical thinning in prefrontal regions may further contribute to executive dysfunction. In the chronic phase, tau protein accumulation has been correlated with late-onset neuropsychiatric symptoms, and hippocampal atrophy may disrupt dopaminergic pathways, which are a core mechanism in schizophrenia.

Our finding of a time-dependent increase in risk fits this multiphase model: the rising hazard over longer follow-up periods is more consistent with progressive neurodegeneration than with a single acute insult leading to immediate psychosis.

However, I want to be transparent about the limitations. Our meta-analysis was based on clinical outcome data and could not directly assess neurobiological processes. The mechanisms I have described are a plausible framework, but direct causal evidence linking specific post-TBI neuropathology to psychosis onset remains limited. Future studies integrating neuroimaging biomarkers—such as longitudinal hippocampal volume tracking or PET-based tau imaging—with psychiatric outcome monitoring would be invaluable in bridging the gap between epidemiology and neurobiology.

PT: Did your analysis indicate relationships between TBI and any other psychotic disorders?

Chu: Our meta-analysis focused specifically on schizophrenia and psychosis broadly defined following TBI. We included studies using various diagnostic frameworks (ICD-8, ICD-9, ICD-10, and DSM criteria) which allowed us to capture a relatively broad spectrum of psychotic presentations. This diversity in coding reflects the long time span of the included studies and the limited amount of available research; restricting to a single diagnostic system would have shrunk an already small pool of eligible studies.

We did not separately analyze individual psychotic disorders beyond schizophrenia, such as schizoaffective disorder or brief psychotic disorder, because the included studies rarely provided that level of detail. It is well established that TBI is associated with a wide range of psychiatric outcomes beyond psychosis, including depression, PTSD, anxiety, and mood disturbances.^2,3 Our study contributes one piece of this larger picture.

That said, if I am being candid about where I think the most important next step lies, it is not in expanding horizontally to other psychiatric outcomes using the same methodology. It is in addressing the question of reverse causality. In our paper, we acknowledged that prodromal symptoms of schizophrenia (impaired attention, impulsivity, cognitive decline, social dysfunction) may themselves increase the likelihood of sustaining a TBI, rather than TBI causing subsequent psychosis. This is a fundamental challenge to causal inference in this field, and one our study could not fully resolve.

What I would like to explore next is essentially the mirror image of the present study: examining the risk of TBI among individuals with schizophrenia or psychosis, with particular attention to the prodromal period. If we can better characterize whether preexisting psychiatric vulnerability contributes to TBI risk, we would be in a much stronger position to disentangle the bidirectional relationship between these conditions. I believe this would be a more meaningful next step than simply applying the same framework to additional outcomes.

Dr Chu is a postgraduate physician at Taipei Medical University Hospital, Taipei, Taiwan. Her research interests include neuropsychiatric outcomes following traumatic brain injury and the intersection of neurology and psychiatry.

References

1. Chu TY, Wu CC, Xu ZX, et al. The risk of schizophrenia or psychosis following traumatic brain injury: a systematic review and meta-analysis. J Psychiatry Res. 2026;197:275-288.

2. Palou Martinez Y, Arrey Agbor DB, Panday P, et al. Mood disorders in the wake of traumatic brain injury: a systematic review. Cureus. 2024;16(6):e62524.

3. Duarte DC, Duarte JC, Gonzalez AAO, et al. Psychiatric disorders in post-traumatic brain injury patients: a scoping review. Heliyon. 2023;9(1):e12905.