Abstract The relationship between gut microbiota dysbiosis and bone mineral density (BMD) in hemodialysis patients, mediated through immune-metabolic pathways, remains to be fully elucidated. In this single-center prospective cross-sectional study, 165 maintenance hemodialysis patients were included to evaluate the independent association between gut microbiota composition and BMD, quantify the mediating roles of immune markers and gut-derived metabolites, and assess the effect modification by nursing-modifiable factors. Fecal samples underwent 16S rRNA sequencing and functional prediction. Inflammatory cytokines (IL-6, TNF-α), gut-derived metabolites (indoxyl sulfate, butyrate), and BMD via dual-energy X-ray absorptiometry (DXA) were measured. Gut microbiota community structure significantly differed across BMD tertiles ( R 2 = 0.033, P = 0.003). After full adjustment, principal coordinate 1 (PCoA-PC1, beta-diversity) was negatively associated with femoral neck BMD, while the Shannon diversity index showed a positive association (both P < 0.05). We identified 15 differentially abundant genera between high and low BMD groups. Functional prediction revealed short-chain fatty acid pathways were positively associated with BMD, while indole/p-cresol pathways showed negative associations. Mediation analysis demonstrated that immune markers and gut-derived metabolites collectively explained 45.71% of the microbiota-BMD relationship. Nursing factors significantly modified this association, with the negative relationship strengthened by low fiber intake, severe constipation, proton pump inhibitor use, and inadequate dialysis (Kt/V < 1.4). In conclusion, gut microbiota dysbiosis is independently associated with lower BMD in hemodialysis patients, partially mediated through immune-inflammatory pathways and gut-derived metabolites. Dietary fiber optimization, constipation management, prudent proton pump inhibitor prescribing, and dialysis adequacy represent actionable nursing targets to mitigate gut-mediated bone loss in this vulnerable population.
Cheng et al. (Mon,) studied this question.