We did a thorough meta-analysis using data from many neuroimaging studies to look for neuroimaging biomarkers that predict cognitive deterioration in persons with preclinical Alzheimer's disease (AD).

We conducted a thorough search of PubMed, MEDLINE, and Embase between 2010 and 2024 to identify research that addressed neuroimaging indicators in preclinical AD populations. A total of 37 studies with 5,200 participants, aged between 45 and 80 years, were considered. The primary outcomes were changes in hippocampal volume, cortical thickness, and Aβ deposition, evaluated by structural MRI, FDG-PET, and amyloid PET imaging. Random-effects meta-analyses were used to investigate the relationship between these neuroimaging biomarkers and future cognitive impairment throughout a 5-year follow-up period.

The meta-analysis found that decreased hippocampus volume was significantly associated with a higher risk of cognitive impairment (pooled effect size = -0.37, 95% CI: -0.43 to -0.31, p < 0.001). Cortical thinning, particularly in the entorhinal cortex, significantly predicted cognitive deterioration (effect size = -0.29, 95% CI: -0.35 to -0.22, p < 0.001). Larger amyloid deposition on PET scans was associated with increased risk of moderate cognitive impairment or Alzheimer's disease (odds ratio = 3.21, 95% CI: 2.43 to 4.25, p < 0.001). FDG-PET studies found decreased glucose metabolism in AD-related areas (posterior cingulate and precuneus) as a feasible early indication (effect size = -0.26, 95% CI: -0.31 to -0.19, p < 0.001).

Our meta-analysis demonstrates that hippocampal shrinkage, cortical thinning, and amyloid buildup are important neuroimaging indicators for predicting cognitive deterioration in preclinical AD. These results support the use of neuroimaging markers in therapy trials and risk assessment systems to identify high-risk patients early in the illness process.