Background/Purpose: Hyperuricemia is a known cardio-metabolic risk factor as well as risk factor for gout. Physical activity has been found to improve many other cardio-metabolic risk factors; however, the relationship between physical activity and serum uric acid levels is not well understood. It has been shown that physical activity improves insulin sensitivity, lowering fasting serum insulin levels, and that lower insulin levels improve uric acid excretion. It would follow that physical activity may be beneficial in lowering serum uric acid levels. We aimed to evaluate how objectively measured physical activity related with serum uric acid concentrations in the 2003-2004 National Health and Nutrition Examination Survey (NHANES).

Methods:  As part of the 2003-2004 NHANES serum uric acid was measured via a colorimetric enzymatic assay.  Physical activity was assessed with seven days of ActiGraph accelerometry (n=3475). Here we used restricted cubic splines to parametrically estimate the relationship of physical activity amounts and serum uric acid levels, while controlling for age, gender, ethnicity, alcohol use, body mass index (BMI), and renal function.

Results:   We found a borderline overall relationship of physical activity amounts and serum uric acid (df=4, p=0.519) and a significant effect of the non-linear portion over and above the additional impact of linear effect (df=3, p=0.0331). This non-linear relationship between physical activity and uric acid was U-shaped: below 760 kCal/day, uric acid was inversely related to physical activity amount; above 760 kCal/day, uric acid was positively related to physical activity amount.  The median physical activity amount was 403 kCal/day, with 75% of the study population averaging less than 604 kCal/day and 90% of the study population averaging less than 849 kCal/day. The mean serum uric acid of the study population was 5.36mg/dL with a standard deviation of 1.36 mg/dL.  Based on this relationship, the effects of physical activity on predicted serum uric acid ranged over 0.7 mg/dL (5.5-6.2 mg/dL).

Conclusion:  In a nationally representative sample of US adults, a non-linear, U-shaped, relationship between objectively measured physical activity and serum uric acid was found. This suggests possible competing mechanisms by which physical activity influences serum uric acid at lower and higher amounts and intensities of physical activity.