2026 ECNP Congress

PS04-4053

N-acetyl aspartate as a cognitive biomarker across the psychosis continuum

Belén Taulero Escalera — M. Ortuño, M. Pardo de Vera, I. Martínez, K. Fuller, P. Camprodon-Boadas, C. Wannan, M. Ptukha, C. Suo, V. Cropley, B. Moffat, C. Pantelis, G. Sugranyes, M. Rapado-Castro · Institute of Psychiatry and Mental Health- Hospital General Universitario Gregorio Marañón- UCM, Department of Child and Adolescent Psychiatry, Madrid, Spain

2026 ECNP Congress

Cognitive dysfunction is a central feature of psychotic disorders and of individuals at clinical or familial risk [1–3], yet its biological substrate remains incompletely understood. Magnetic resonance spectroscopy (MRS) allows in vivo quantification of brain metabolites such as N-acetyl aspartate (NAA), glutamate, choline, creatine and GABA, which are implicated in neuronal integrity, neurovascular coupling and synaptic transmission [4–8]. We first conducted a systematic review to clarify how these metabolites relate to cognition along the psychosis continuum, and then examined, within this framework, longitudinal NAA–cognition associations in a sample of early-onset psychosis (EOP) and matched controls.

Two researchers independently searched in several datebases, combining terms for cognition, MRS and psychosis-spectrum populations (at-risk states, first-episode psychosis, established non-affective and affective psychosis). We included full-text empirical studies of participants under 65 years that examined associations between MRS-derived metabolite levels in any brain region and performance on standardised neuropsychological tests. Data extraction covered clinical group, illness stage, MRS methodology, regions of interest, metabolites and cognitive domains. The review is registered in PROSPERO (CRD42024567183).

From 3333 records, 87 studies entered data extraction; 67% focused primarily on NAA, mostly using single-voxel proton MRS. Across schizophrenia, bipolar disorder, first-episode psychosis, high-risk states and healthy controls, higher NAA in dorsolateral prefrontal cortex (DLPFC), anterior cingulate cortex (ACC), hippocampus and centrum semiovale was generally associated with better executive functions, working memory, attention, learning and global cognition [9–17]. Glutamate tended to show negative associations with executive functions and memory [5,6,15], while choline and creatine were negatively related to executive functions and GABA positively to working memory and executive functions [7,8]. Overall, the review suggests that NAA is a plausible cognitive biomarker across the psychosis spectrum, but heterogeneity in design, acquisition and analysis, together with several null findings, limits causal inference.

Within this context, we conducted a longitudinal MRS study of NAA in DLPFC in children and adolescents with EOP and healthy controls, followed over five years. At five years, NAA concentrations were significantly lower in EOP than controls (p = 0.030). A composite index of DLPFC-related functions (global attention, working memory, executive functioning) remained higher in controls throughout follow-up. In controls, the index increased significantly from baseline to five years and from two to five years, and these gains were strongly linked to NAA levels. Between-group differences in the cognitive index at five years became significant when NAA was considered, and NAA moderated cognitive trajectories in a sex- and diagnosis-specific manner. In male controls, NAA influenced cognitive growth between two and five years, whereas in EOP, NAA variation was associated with changes in executive functions and attention, with pronounced effects in males and in schizoaffective disorder. These sex effects are consistent with previous evidence of sex-related NAA differences in healthy populations [18, 19].

Integrating review and cohort findings, NAA appears to index abnormal developmental trajectories of prefrontal cognition from EOP to established adult psychosis. This supports its potential use in early detection, prognosis and cognition-focused intervention strategies, whilst highlighting the need for longitudinal, harmonised MRS–cognition studies across the psychosis spectrum.

References

  1. Bombín, et al. (2013).
  2. Huang, et al. (2017).
  3. Morales, et al. (2012).
  4. Martínez, et al. (2016).
  5. Merritt, et al. (2016).
  6. Bustillo, et al. (2011).
  7. Turner, et al. (2015).
  8. Bekdash, R. (2019).
  9. Frey, et al. (2007).
  10. Steen, et al. (2005).
  11. Zabala, et al. (2007).
  12. Weinberger, D. R., & Bermam, K. F. (1996).
  13. Ohrmann, et al. (2007).
  14. Galińska, et al. (2007).
  15. Wang, et al. (2019).
  16. Brugger, et al. (2011).
  17. Stuss, D. T. (2006).
  18. Grachev, et al. (2000).
  19. Rubinow, et al. (1996).