Abstract: BACKGROUND: Osteoarthritis is a complex disease closely related to metabolic abnormalities. However, previous studies only involved limited blood lipid indicators and did not conduct more comprehensive blood lipid profile analysis. An in-depth exploration of the causal relationship between the seven items of blood lipids and osteoarthritis will not only help understand the pathogenesis of osteoarthritis, but also provide new research directions and clinical basis for its prevention and treatment.
OBJECTIVE: To explore the causal relationship between blood lipids and osteoarthritis.
METHODS: The genome-wide association analysis statistical data of 7 items of blood lipids and osteoarthritis from the IEU OPEN GWAS database were used to summarize, and significant single nucleotide polymorphisms were used as instrumental variables. The causal relationship between seven items (serum total cholesterol, triacylglycerol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, apolipoprotein B, apolipoprotein AI and apolipoprotein A1) of blood lipids and osteoarthritis (osteoarthritis, knee or hip osteoarthritis, knee osteoarthritis and hip osteoarthritis) was determined through two-sample Mendelian randomization analysis. The inverse variance weighting was the main effect, and the MR-Egger regression method and the weighted median method were the supplementary effects. Bonferroni correction and reverse Mendelian randomization analysis could ensure validity. Multivariable Mendelian randomization analysis was used to further eliminate confounding factors. A significant causal relationship between seven items of blood lipids and osteoarthritis was obtained to ensure the robustness of the analysis. Co-localization analysis was used to once again ensure the robustness of the causal relationship and identify significantly influencing gene loci, making the evidence of causality more complete.
RESULTS AND CONCLUSION: (1) In the two sample Mendelian randomization analysis, the results from inverse variance weighting indicated negative correlations between osteoarthritis and the following serum lipids: total cholesterol (OR=0.937 2, 95%CI=0.885 6-0.991 9, P=0.025), low-density lipoprotein cholesterol (OR=0.959 4, 95%CI=0.923 6-0.996 6, P=0.033), high-density lipoprotein cholesterol (OR=0.911 2, 95%CI=0.833 5-0.996 2, P=0.04), apolipoprotein B (OR=0.926 7, 95%CI=0.887 7-0.967 4, P=0.000 5), and apolipoprotein AI (OR=0.951 2, 95%CI=0.911 0-0.993 1, P=0.023). Additionally, total cholesterol (OR=0.892 3, 95%CI=0.843 1-0.944 3, P=0.000 08), triglycerides (OR=0.938 5, 95%CI=0.884 7-0.995 6, P=0.035), and apolipoprotein B (OR=0.911 6, 95%CI= 0.865 9-0.959 7, P=0.000 4) were negatively associated with knee or hip osteoarthritis. For knee osteoarthritis specifically, total cholesterol (OR=0.898 3, 95%CI=0.841 2-0.959 3, P=0.001), high-density lipoprotein cholesterol (OR=0.881 2, 95%CI=0.794 7-0.977 0, P=0.016), and apolipoprotein B (OR=0.919 0, 95%CI=0.869 8-0.971 0, P=0.002) also showed negative correlations. Lastly, with respect to hip osteoarthritis, total cholesterol (OR=0.864 5, 95%CI=0.797 5- 0.937 3, P=0.000 4), low-density lipoprotein cholesterol (OR=0.925 6, 95%CI=0.879 5-0.974 1, P=0.003), and apolipoprotein B (OR=0.888 8, 95%CI=0.817 6- 0.966 3, P=0.005) exhibited negative correlations. No statistically significant differences were found in the reverse Mendelian randomization analysis. (2) In the multivariable Mendelian randomization analysis, the results from inverse variance weighting indicated a negative correlation between high-density lipoprotein cholesterol and osteoarthritis (OR=0.942 7, 95%CI=0.896 1-0.991 8, P=0.022). Additionally, total cholesterol (OR=0.799 8, 95%CI=0.647 8-0.987 6, 
P=0.037) and high-density lipoprotein cholesterol (OR=0.865 1, 95%CI=0.7781-0.961 9, P=0.007) were also negatively associated with knee osteoarthritis. (3) Colocalization analysis revealed that total cholesterol and low-density lipoprotein were significantly associated with osteoarthritis at single nucleotide polymorphisms rs13107325 (H4 posterior probability=99.9%). (4) These findings, using international databases and non-Asian populations, provide valuable insights for early clinical diagnosis, understanding the pathogenesis, and researching prevention and treatment of osteoarthritis in Chinese biomedicine and the Chinese population.

Key words: osteoarthritis, serum metabolites, lipids, Mendelian randomization, colocalization analysis, causal relationship, engineered tissue construction