Abstract
PURPOSE: Previous observational studies have demonstrated that Low back pain (LBP) often coexists with Alzheimer's Disease (AD), however, the causal relationship remains unclear. The purpose of this study is to explore the causal relationship between AD and LBP through Mendelian randomized analysis. METHODS: Instrumental variables (IVs) were derived from the genome-wide association study (GWAS) of AD. Information regarding Instrumental variables (IVs) in LBP was extracted from a GWAS database. MR-Egger, weighted median, inverse variance weighted (IVW)and Weighted mode were used to evaluate the causal effects. Cochran's Q test and MR-Egger intercept were applied to detect the heterogeneity and horizontal pleiotropy, respectively. Outliers were found and remove based on MR-PRESSO analysis to mitigate the effect of horizontal pleiotropy on the results. Deleting each genetic variant applying the leave-one-out analysis can help evaluate the robustness of results. Finally, MR-PRESSO Raw and Outlier-corrected were used to enhance the credibility of the results. RESULTS: IVW assessment provided strong evidence that AD is positively associated with LBP (Odds Ratio (OR)= 1.046, 95% confidence interval(CI) = 1.023-1.070, P=5.8×10(-5)). There was no heterogeneity in our study (p > 0.05). The results of the pleiotropy test indicated that there was no pleiotropy in our IVW analysis (p > 0.05). MR-Egger, Weighted median, Weighted mode analysis results are consistent with our IVW analysis results. There is a genetic relationship between Alzheimer's disease and low back pain. CONCLUSION: This study provides evidence for a causal relationship between AD and LBP. It emphasizes the necessity of improving pain management for patients with Alzheimer's disease in clinical practice. However, the data were derived from a European population, which may limit the generalizability of the results to other populations.