Quantity-effect correlation between water intake and serum uric acid in US adults: a cross-sectional study based on NHANES data.

Background: Hyperuricemia (HUA) imposes a substantial clinical and public health burden, affecting 43 million adults in the U.S. and 170 million in China. Increased hydration is often recommended as a preventive measure for HUA, but the quantity-effect correlation between water intake and serum uric acid (SUA) is unclear. This study aimed to investigate the quantity-effect correlation between water intake and SUA. Methods: This cross-sectional study evaluated the data of 15,174 participants from the National Health and Nutritional Examination Surveys (NHANES) 2009-2018. The SUA levels (mg/dL) of all the participants were tested, and their daily water intake (g) was estimated based on the average of two 24-hour dietary recalls. To more accurately assess individual water consumption, the ratio of water intake to body weight (g/kg) was employed as a treatment variable. Multivariate linear regression models were used to evaluate the association between the two variables. Generalized additive models with smooth curve fittings were also applied. A Sobel-Goodman analysis was conducted to determine whether osmolality served as a mediator between water intake and SUA levels. Results: The final 15,174 participants had a mean age of 44.28±16.33 years (49.0% male), with SUA quartiles showing significant differences in sex, age, and lifestyle factors. The multivariate regression analyses showed that water intake/body weight was negatively correlated with SUA after adjustment for other confounders. Subgroup analyses stratified by sex showed that this negative correlation was more significant in females. Further, a nonlinear negative correlation between water intake/body weight and SUA was observed in the generalized additive model (P<0.001), showing an L-shaped curve (inflection points: 7.591 and 33.57 g/kg for total plain water consumption/body weight and total moisture intake/body weight, respectively). Plasma osmolality mediated the relationship between water intake/body weight and SUA, with mediating effects of 9.14% and 5.84% for total plain water consumption/body weight and total moisture intake/body weight, respectively. Conclusions: This study found a negative correlation between water intake/body weight and SUA, particularly in females. Moreover, the negative relationship exhibited nonlinear characteristics, following an L-shaped curve. Plasma osmolality was found to play a mediating role in the reduction of SUA by water intake. These findings support personalized water intake guidance-targeting intake below inflection points (7.591 g/kg for total plain water, 33.57 g/kg for total moisture) may optimize SUA reduction (especially in females) and inform non-pharmacological HUA management in clinical practice.