Abstract
Osteoporosis is a widespread skeletal disorder characterized by reduced bone mineral density (BMD) and increased fracture risk. While dietary factors are known to influence bone health, the causal role of specific macronutrients-particularly carbohydrates-remains uncertain. This study aimed to investigate the causal effect of carbohydrate intake on BMD and to identify plasma metabolites that may mediate this association. We conducted a bidirectional two-sample Mendelian randomization (MR) analysis using genome-wide association study summary data from large-scale cohorts. Genetic variants associated with carbohydrate, fat, and protein intake were used as instrumental variables to evaluate their effects on total BMD. To investigate potential mediators, we performed two-step MR mediation analysis to identify plasma metabolites underlying the carbohydrate-BMD relationship. Pleiotropy and heterogeneity were assessed using MR-Egger regression and weighted median approaches. Genetically predicted carbohydrate intake was significantly associated with lower total BMD (odds ratio = 0.89, 95% confidence interval: 0.80-0.98, P = .021), while no causal effects were observed for fat or protein intake. Mediation analysis identified 3 plasma metabolites-mannose-to-trans-4-hydroxyproline ratio, caffeine-to-theobromine ratio, and mannonate-as significant intermediaries. These metabolites jointly explained up to 27% of the total effect, suggesting that metabolic pathways related to mannose and caffeine may contribute to carbohydrate-induced bone loss. Our findings provide genetic evidence that carbohydrate intake adversely affects BMD through specific metabolic intermediates. These insights highlight the potential of metabolomic profiling to inform targeted nutritional strategies for osteoporosis prevention and bone health optimization.