Abstract
BACKGROUND: Hyperuricaemia (HUA) is a common metabolic disease that has become a global public health burden. The Planetary Health Diet Index for the United States (PHDI-US), a sustainable dietary pattern emphasizing a plant-based diet, has great potential for chronic disease prevention and control, but its relationship with HUA is unclear. METHODS: This study was based on nationally representative NHANES data. Multivariate logistic regression and restricted cubic spline models were used to assess the association between raw PHDI-US and HUA. The key dietary component Dairy was identified by machine learning, the optimized PHDI-Dairy index was constructed, and the association between dairy intake and HUA risk was verified in an independent self-built cohort. RESULTS: Primitive PHDI-US was significantly negatively correlated with the risk of HUA, but its protective effect was weaker than traditional dietary indices such as AHEI, AEI, and MEDI. Machine learning results show that Dairy is the most critical component in PHDI-US composition. The PHDI-Dairy index, constructed after optimizing the weight of dairy products, was more negatively correlated with the risk of HUA, and the protection efficiency was better than that of other dietary indices. External validation further confirmed that increased dairy intake was associated with a low risk of HUA. CONCLUSION: PHDI-US was negatively correlated with the risk of HUA. After optimizing the weighting of dairy products, PHDI-Dairy demonstrated a significantly enhanced protective effect, outperforming both the original PHDI and other dietary indices. This result highlights its potential as a nutritional intervention tool for chronic diseases, contributing to both public health and environmental sustainability.