Whether blood amyloid-β (Aβ) could be a peripheral biomarker of Alzheimer's disease (AD) remains in dispute. In the present study, we conducted a meta-analysis with 19 citations searched from Embase, PubMed, and the Cochrane Library database. Weighted mean difference (WMD) with 95% confidence intervals (CIs) was used to estimate the effect size. We firstly analyzed the plasma Aβ40, Aβ42, and Aβ42/Aβ40 ratio in AD and control group subjects. However, only a lower level of plasma Aβ42 was figured out in AD group subjects with weak statistical significance (WMD 1.82; 95% CI 0.59, 3.06; P = 0.004; I2 = 84%). We considered that the medical histories of control subjects could influence the biomarker ability of plasma Aβ. Therefore, subgroup analyses were then carried out based on a new recruiting criterion for control subjects, defining as no afflictions of any Aβ-related diseases. Surprisingly, AD group subjects showed a significant decrease in plasma Aβ42/Aβ40 ratio with low heterogeneity among studies (WMD 0.02; 95% CI 0.02, 0.02; P < 0.00001; I2 = 0%). Moreover, not only the Aβ42/Aβ40 ratio but also Aβ42 and Aβ40 were indifferent between AD and pseudo-control subjects which might be afflicted with Aβ-related diseases. This meta-analysis demonstrated that medical histories of control subjects were interference factors impeding plasma Aβ to be a biomarker of AD.
Keywords: Alzheimer’s disease (AD); Amyloid-β; Biomarker; Meta-analysis.