Salmon · Cortex; a journal devoted to the study of the nervous system and behavior 2024 · Narrative review · n=?

Cerebral glucose metabolism in Alzheimer's disease.

Cited 28 times in the scientific literature.

Level 5 - mechanism / opinion, no new human data

Narrative review synthesizing existing neuroimaging literature without systematic search or quantitative pooling

PubMed 39141935 · doi:10.1016/j.cortex.2024.07.004 · record verified 2026-08-27

What was done

This narrative review synthesizes literature on 18F-fluoro-deoxy-glucose positron emission tomography (FDG-PET) across preclinical, prodromal, and clinically probable Alzheimer's disease (AD). It evaluates FDG-PET patterns in pathologically confirmed cases, utility in differential diagnosis, automated and machine learning analysis methods, cognitive-metabolic correlates (including episodic memory and anosognosia), and interactions between glucose metabolism and amyloid/tau biomarkers.

What was found

The abstract reports no numerical values, diagnostic accuracy statistics, or effect sizes. Qualitatively, it reports that posterior brain hypometabolism is a sensitive indicator of AD that predicts progression in prodromal stages, while hippocampal metabolism shows paradoxical variability. Machine learning and automated analysis techniques yielded variable accuracy and reliability. Episodic memory was linked to default mode network and Papez circuit metabolism. Intriguingly, increased metabolism was sometimes observed in association with amyloid/tau deposition, and preserved glucose metabolism was frequently detrimental rather than compensatory. Limbic-predominant metabolic changes were often tied to non-AD profiles such as limbic-predominant age-related encephalopathy (LATE).

Why it matters

It outlines how FDG-PET captures biological and clinical heterogeneity across AD stages and aids in differentiating AD from other neurodegenerative conditions like LATE.

Limits

As a narrative review, it lacks systematic search methodology, structured study quality appraisal, and meta-analytic pooling. The abstract provides no quantitative metrics (such as sensitivity or specificity values) and highlights inconsistency across automated analysis techniques and machine learning algorithms.

Cited by