O'Neill · Biological psychiatry 2020 · Cross-sectional case-control study · n=174

Parsing the Heterogeneity of Brain Metabolic Disturbances in Autism Spectrum Disorder.

Cited 25 times in the scientific literature.

Level 4 - case-series / case-control

Case-control neuroimaging study

PubMed 31427037 · doi:10.1016/j.biopsych.2019.06.010 · record verified 2026-08-28

What was done

Near-whole-brain proton magnetic resonance spectroscopy was conducted to measure levels of N-acetyl compounds, glutamate plus glutamine, creatine plus phosphocreatine, and choline compounds in 78 children and adults with autism spectrum disorder (ASD) and 96 typically developing controls (total n = 174). The authors assessed the effects of ASD diagnosis and severity on metabolite levels, examining moderation by age, sex, and IQ.

What was found

The abstract reports no numerical values, effect sizes, or test statistics. Qualitatively, ASD diagnosis and severity were associated with reduced levels of multiple metabolites in white matter and perisylvian cortex, and elevated levels in the posterior cingulate. Regional age-related metabolite declines differed between ASD and controls (showing both slower and faster decreases depending on region). Sex differences in metabolite levels were smaller in ASD than in typically developing participants, and lower IQ was associated with ASD-specific metabolite decreases across multiple brain regions.

Why it matters

This study extends magnetic resonance spectroscopy across near-whole-brain coverage in a comparatively large cohort, demonstrating that neurometabolic abnormalities in ASD are regionally heterogeneous and significantly moderated by age, sex, and cognitive level.

Limits

The abstract contains no numeric values, effect sizes, or confidence intervals. The cross-sectional design limits inferences about true longitudinal developmental trajectories. Proton magnetic resonance spectroscopy measures aggregate tissue metabolite concentrations, which cannot isolate specific cellular mechanisms such as neuronal loss versus neuroinflammation or mitochondrial dysfunction.

Cited by