26 September 2024 in European Psychiatry
Background: We need to better understand the risk factors and predictors of medication-related weight gain to improve metabolic health of individuals with schizophrenia. This study explores how trajectories of antipsychotic medication (AP) use impact body weight early in the course of schizophrenia. Methods: We recruited 92 participants with first-episode psychosis (FEP, n = 92) during their first psychiatric hospitalization. We prospectively collected weight, body mass index (BMI), metabolic markers, and exact daily medication exposure during 6-week hospitalization. We quantified the trajectory of AP medication changes and AP polypharmacy using a novel approach based on meta-analytical ranking of medications and tested it as a predictor of weight gain together with traditional risk factors. Results: Most people started treatment with risperidone (n = 57), followed by olanzapine (n = 29). Then, 48% of individuals remained on their first prescribed medication, while 33% of people remained on monotherapy. Almost half of the individuals (39/92) experienced escalation of medications, mostly switch to AP polypharmacy (90%). Only baseline BMI was a predictor of BMI change. Individuals in the top tercile of weight gain, compared to those in the bottom tercile, showed lower follow-up symptoms, a trend for longer prehospitalization antipsychotic treatment, and greater exposure to metabolically problematic medications. Conclusions: Early in the course of illness, during inpatient treatment, baseline BMI is the strongest and earliest predictor of weight gain on APs and is a better predictor than type of medication, polypharmacy, or medication switches. Baseline BMI predicted weight change over a period of weeks, when other traditional predictors demonstrated a much smaller effect.
4 August 2026 in Schizophrenia Research
Predicting long-term outcomes in first-episode schizophrenia (FES) remains difficult, despite being especially important early in the illness, when timely intervention is most critical. It also remains unclear how much data from the initial phase of illness is required to improve prognostic accuracy. We analysed 68 FES patients assessed at baseline (V1; mean 0.5 years post-onset), one-year follow-up (V2; mean 1.2 years), and outcome (V3; mean 4.9 years). Elastic-net models were trained to predict three V3 outcomes - negative symptoms (PANSS Negative factor; Wallwork/Fortgang), global functioning (GAF), and quality of life (WHOQOL-BREF psychological domain) - using either V1 predictors alone (23 variables) or V1 + V2 combined (43 variables). Performance was evaluated using nested cross-validation on held-out data. With V1 + V2 predictors, all three outcomes were predicted at statistically significant levels: PANSS Negative R2 = 0.22 (driven by log(DUP), PANSS Negative at V1/V2, and PANSS Disorganised at V2); WHOQOL-BREF Psychological Health R2 = 0.22 (driven by WHOQOL Psychological Health and GAF at V2); and GAF R2 = 0.14 (driven by GAF, PANSS Positive, WHOQOL Psychological Health at V2, and hospitalisation burden). With V1 predictors alone, only PANSS Negative showed meaningful predictive power (R2 = 0.15); GAF and WHOQOL-BREF did not outperform the intercept-only baseline. Long-term functioning and quality of life in FES cannot be predicted from first-episode data alone; at least one year of follow-up is required, suggesting post-onset changes shape these outcomes. Negative symptoms are an exception: comparatively stable after initial treatment and predictable from baseline, with past symptomatology along with DUP selected as predictors - indicating stronger persistency and predictability than in the other two investigated outcomes.