Objective Given the limitations of current treatments for Alzheimer’s disease (AD), this study aims to comprehensively evaluate the therapeutic efficacy of human umbilical cord mesenchymal stem cells (hUCMSCs) in AD mouse models through a systematic review and meta-analysis. Additionally, we explore the impact of transplantation dose and route on treatment outcomes to identify the optimal window for clinical application. Methods In accordance with the PRISMA guidelines, we systematically searched four major databases to identify randomized controlled trials involving hUCMSCs in AD mouse models. We used the standardized mean difference (SMD) to synthesize effect sizes and performed subgroup analyses based on pre-defined transplantation routes and doses. Results A total of 13 studies were included in the analysis. The meta-analysis revealed that hUCMSCs transplantation significantly improved spatial learning and memory in AD model mice, with a marked reduction in escape latency (SMD = −2.55; 95% CI: −3.34 to −1.75; I 2 = 77.9%, random-effects model). Additionally, it significantly lowered brain β-amyloid levels (SMD = −5.34; 95% CI: −7.21 to −3.47; I 2 = 80.3%), increased brain-derived neurotrophic factor (SMD = 4.25; 95% CI: 3.18 to 5.31), and reduced neuronal apoptosis (SMD = −4.96; 95% CI: −6.52 to −3.41). Subgroup analyses further revealed that efficacy was significantly dose- and route-dependent. For cognitive improvement, intravenous injection of medium to high doses (≥1 × 10 6 cells) was most effective and robust. For amyloid-β clearance, low-dose administration via intravenous, lateral ventricle, and cortical routes showed significant efficacy, whereas bilateral hippocampal injection did not yield significant benefits. Conclusion Human umbilical cord mesenchymal stem cells can improve behavioral and pathological outcomes in AD mouse models via multiple mechanisms of action. The intravenous route using medium to high doses emerges as a critical factor for achieving optimal effects, providing important evidence and informing future experimental design and clinical translational research.
Si et al. (Mon,) studied this question.