Townsend · Psychological medicine 2023 · systematic review and meta-analysis · n=36 studies (1335 subjects)

Brain glucose metabolism in schizophrenia: a systematic review and meta-analysis of 18 FDG-PET studies in schizophrenia.

Cited 70 times in the scientific literature.

Level 3 - non-randomized controlled study

Systematic review and meta-analysis of observational case-control neuroimaging studies

PubMed 35730361 · doi:10.1017/S003329172200174X · record verified 2026-08-28

What was done

The authors searched MEDLINE, PsychINFO, and EMBASE from 1980 through May 2021 for studies using 18F-FDG PET to compare regional brain glucose metabolism (absolute and relative CMRGlu) between patients with schizophrenia or first-episode psychosis and healthy controls. Effect sizes (Hedges' g) across the frontal, temporal, parietal, and occipital lobes, basal ganglia, and thalamus were pooled using random-effects meta-analysis. A subset of eight studies reporting voxel-based morphometry measures of FDG uptake was evaluated via signed differential mapping analysis.

What was found

Thirty-six studies with 1,335 subjects were included. Schizophrenia patients showed significantly lower absolute frontal glucose metabolism (Hedges' g = -0.74 ± 0.54, p = 0.01; I² = 67%) and relative frontal metabolism (g = -0.44 ± 0.34, p = 0.01; I² = 55%) compared to controls. Absolute frontal hypometabolism was pronounced in chronic patients (g = -1.18 ± 0.73) and medicated patients (g = -1.04 ± 0.26), but not statistically significant in first-episode (g = -0.09 ± 0.88) or drug-free patients. No metabolic differences were found in parietal, temporal, or occipital lobes, or the thalamus. Excluding outliers showed lower absolute basal ganglia metabolism in schizophrenia (g = -0.25 ± 0.24, p = 0.049; I² = 5%). Voxel-based analyses revealed lower uptake in the left anterior cingulate gyrus (Z = -4.143, p = 0.007) and left inferior orbital frontal gyrus (Z = -4.239, p = 0.02).

Why it matters

This review provides quantitative evidence supporting the hypofrontality hypothesis in schizophrenia, demonstrating that metabolic dysfunction is localized to frontal regions rather than generalized across the whole brain.

Limits

Moderate heterogeneity was observed in frontal metabolic analyses (I² = 55–67%). Differences were primarily evident in chronic and medicated cohorts, confounding disease pathology with antipsychotic exposure and illness duration. Voxel-based analysis was restricted to only eight studies.

Cited by