Abstract
BACKGROUND: Several previous observational studies have shown that abnormal sphingomyelin metabolism may be implicated in the pathogenesis of Alzheimer's disease. To determine the causal relationship between sphingolipid abundance and gut microbiota abundance at the genetic level, we conducted a Mendelian randomization (MR) investigation. METHODS: We first used the TwoSampleMR and MRPRESSO packages for conducting two-sample MR studies. Second, we utilized random effect inverse variance weighting (IVW) as the principal method of analysis and used MR‒Egger, the weighted median, the simple mode and the weighted mode as supplementary methods. Finally, we performed tests for heterogeneity and horizontal pleiotropy. These analyses were also conducted to evaluate the impact of individual SNPs on the outcomes of our analysis. A Bonferroni-corrected threshold of p = 2.4e-4(0.05/211) was considered significant, and p values less than 0·05 were considered to be suggestive of an association. RESULTS: The results showed that sphingolipid levels were suggestively associated with the abundance of 6 gut microbiota taxa. Specifically, two taxa were positively correlated with sphingolipid levels, including the family Alcaligenaceae (p = 0.006, OR 95% CI = 1.109 [1.030-1.194]) and the species Ruminococcus callidus (p = 0.034, OR 95% CI = 1.217 [1.015-1.460]). In contrast, negative correlations were observed with the abundances of 4 gut microbiota taxa, including the genus Flavonifractor (p = 0.026, OR 95% CI = 0.804 [0.663-0.974]), the genus Streptococcus (p = 0.014, OR 95% CI = 0.909 [0.842-0.981]), the species Bacteroides caccae (p = 0.037, OR 95% CI = 0.870 [0.763-0.992]), and the species Haemophilus parainfluenzae (p = 0.006, beta 95% CI = -0.269 [-0.462, -0.076]). The results presented a normal distribution, with no anomalous values, heterogeneity, or horizontal pleiotropic effects detected. CONCLUSIONS: This two-sample MR study revealed a potential causal relationship between sphingomyelin levels and gut microbiota abundance.