While treatment non-adherence is a well-established relapse risk factor, broader contributors require investigation. Studying chronic, high-risk schizophrenia cohorts in real-world lower middle-income country (LMIC) settingsremains methodologically challenging.

To explore multidimensional factors moderating treatment trajectories in individuals with chronic schizophrenia at high risk of relapse, comparing relapsers and non-relapsers in a real-world LMIC setting.

In this 12-month prospective naturalistic study, 56 clinically stable adults with chronic schizophrenia (≥2 relapses within 3 years), receiving long-acting injectable and/or oral antipsychotics were followed up monthly. Relapse was defined as worsening psychotic symptoms requiring re-hospitalisation. Linear mixed-effects models examined trajectories of Positive and Negative Syndrome Scale (PANSS) total and factor domains (positive, negative and disorganised) and tested group*time*moderator interactions for neurological soft signs (NSS), quality of life (QOL), aberrant salience, anxiety, resilience, self-esteem, social adversity and childhood trauma.

Twenty-three participants (41%) relapsed; overall attrition was 59%, with 21% of discontinuations relapsing. Most relapses (87%) involved non-adherence and substance misuse. No group*time interaction effects were found for PANSS domains. Poorer overall QOL (F(4, 31 = 4.0), p = 0.009) was associated with worse negative symptom trajectory in the relapse group, while more pronounced NSS was associated with worse positive symptom trajectory in the non-relapse group (F(2, 23 = 8.4), p = 0.002).

Conducting longitudinal research in chronic, high-risk schizophrenia populations in LMICs is challenging. Although exploratory and hypothesis-generating, QOL and NSS may represent clinically meaningful moderators of treatment course and potential targets for intervention and future research.

Progressive brain structural MRI changes are described in schizophrenia and have been ascribed to both illness progression and antipsychotic treatment. We investigated treatment effects, in terms of total cumulative antipsychotic dose, efficacy and tolerability, on brain structural changes over the first 24 months of treatment in schizophrenia.

A prospective, 24-month, single-site cohort study in 99 minimally treated patients with first-episode schizophrenia, schizophreniform and schizoaffective disorder, and 98 matched healthy controls. We treated the patients according to a fixed protocol with flupenthixol decanoate, a long-acting injectable antipsychotic. We assessed psychopathology, cognition, extrapyramidal symptoms and BMI, and acquired MRI scans at months 0, 12 and 24. We selected global cortical thickness, white matter volume and basal ganglia volume as the regions of interest.

The only significant group × time interaction was for basal ganglia volumes. However, patients, but not controls, displayed cortical thickness reductions and increases in white matter and basal ganglia volumes. Cortical thickness reductions were unrelated to treatment. White matter volume increases were associated with lower cumulative antipsychotic dose, greater improvements in psychopathology and cognition, and more extrapyramidal symptoms. Basal ganglia volume increases were associated with greater improvements in psychopathology, greater increases in BMI and more extrapyramidal symptoms.

We provide evidence for plasticity in white matter and basal ganglia associated with antipsychotic treatment in schizophrenia, most likely linked to the dopamine blocking actions of these agents. Cortical changes may be more closely related to the neurodevelopmental, non-dopaminergic aspects of the illness.