Abstract Body: Introduction: Gut microbiome (GMB) has been implicated in the cholesterol-lowering effects of statins. Observational studies and animal models suggest statins increase GMB diversity and alter GMB composition. However, secondary analyses of a rosuvastatin trial (N=20/arm) did not support these findings. This trial was limited in demographics (Norwegian women) and GMB measurement (16s rRNA sequencing, which lacks data on species and function). We sought to emulate a randomized trial of statins on GMB in a longitudinal cohort with greater demographic inclusion and higher-resolution GMB sequencing.
Hypothesis: Statins increase GMB diversity and alter GMB composition and function.
Methods: We selected statin initiator and non-initiator-matched participants from the Baltimore Longitudinal Study of Aging. Eligible participants had to have 2+ GMB samples (collected between 2013 and 2019), be statin naïve at start of collection, have non-missing matching covariate data, and be statin eligible at their baseline visit (using 2013 ACC/AHA Guidelines). We used a 1:1 matching algorithm on the baseline propensity score, with nearest-neighbor matching on age, BMI, education, LDL-c, and ASCVD lifetime risk category (Yes/No/Prior ASCVD event); and exact matching on sex, diabetes status, smoking, and race. We measured GMB features (diversity, composition, and functional genes) using shotgun sequencing. We tested associations of statin initiation with post-treatment GMB alpha diversity and with changes in GMB beta diversity using ANCOVA g-computation, and with differential abundance of microbial species and functional genes using ANCOM-BC2. We adjusted models for the baseline matching covariates; alpha diversity models were further adjusted for baseline alpha diversity.
Results: We identified N=30 matched pairs (N=60 total; 40% female, 30% Black). Matching covariates were well-balanced (all absolute standardized mean differences <0.1). Samples were also balanced on baseline GMB features. Among treated participants, statins lowered LDL-c (B: -40.7, 95%CI: -53.8 to -27.6), but did not affect other lipids. Statins were not associated with GMB diversity. Statins decreased Bifidobacterium pseudocatenulatum and altered 10 functional genes (mostly increased fatty acid biosynthesis) (all p <0.05); however, these associations were not significant at FDR <0.20.
Conclusions: Using an emulated randomized trial, statins did not alter GMB diversity, but may induce changes in GMB function.