Abstract: BACKGROUND: In China, the patient population with osteonecrosis is large, and there is an urgent need to find new preventive targets to develop more effective treatment strategies. Metabolomics studies have shown that there is an association between human metabolites and osteonecrosis, but the causal relationship between blood metabolites and osteonecrosis has not yet been clarified.
OBJECTIVE: To investigate the causal relationship between blood metabolites and osteonecrosis through two-sample Mendelian randomization analysis.
METHODS: The public data of 486 blood metabolites (exposure factors) and osteonecrosis (outcome factors) were collected. Data of 486 blood metabolites were derived from a genome-wide association estimate for blood metabolites published in Nature Genetics in 2014, which covered 7 824 European adults. The single nucleotide polymorphism data for osteonecrosis were obtained from the FinnGen public database R11 dataset, containing information on a total of 431 614 samples and 21 306 430 single nucleotide polymorphism loci, with 1 788 cases of osteonecrosis and 429 826 controls, with all participants being of European descent. Mendelian randomization analysis (inverse variance weighting method, MR-Egger method, and weighted median method) was performed by Rstudio software, and then the heterogeneity test, horizontal pleiotropy test and Steiger directionality test were performed to ensure the robustness and reliability of the results.
RESULTS AND CONCLUSION: (1) Sixteen blood metabolites were identified as having a significant causal relationship with osteonecrosis (Pinverse variance weighting < 
Pfalse discovery rate < 0.05). (2) Eight blood metabolites increased the risk of osteonecrosis (including four known metabolites and four unknown metabolites), specifically pantothenate, beta-hydroxyisovalerate, hippurate, salicyluric glucuronide, X-08766, X-11452, X-12776 and X-14662. (3) Eight blood metabolites could reduce the risk of osteonecrosis (six known metabolites and two unknown metabolites), including cortisol, 1-palmitoylglycerol (1-monopalmitin), pyroglutamyl glycine, 2-stearoylglycerophosphocholine, p-cresol sulfate, ergothioneine, X-06307, X-12092. (4) The above results suggest that there is a causal relationship between 16 blood metabolites and osteonecrosis, which is expected to be a potential target for intervention in the occurrence and treatment of osteonecrosis in the future. (5) Despite the lack of relevant data from large-scale Asian populations at present, this study provides important reference value for the field of osteonecrosis in China based on European population data. In the future, domestic medical workers may be able to achieve precise intervention for osteonecrosis by regulating metabolite levels. In addition, based on the results of this study, relevant researchers can further explore the mechanism of action of metabolites in the treatment of osteonecrosis with traditional Chinese medicine, which not only helps to deepen the understanding of traditional Chinese medical therapies but also promotes the progress of integrated traditional Chinese and Western medicine research, driving the development of personalized treatment plans that are more suitable for the characteristics of the Chinese population.

Key words: blood metabolites, osteonecrosis, Mendelian randomization, causality, sensitivity analysis, genome-wide association study, single nucleotide polymorphism, genetics, engineered tissue construction