Higher uric acid levels positively correlated with fructose in the overall pattern of PCOS
As stated, elevated serum uric acid levels were accompanied with increased fructose levels in women with PCOS. Therefore, using RCS models, we further explore the relationship between serum fructose and uric acid levels (Figure 1a; Supplementary Figure 1). Remarkably, there was a linear correlation between uric acid and fructose levels in women with PCOS (P overall < 0.001,P non-linear = 0.30; Figure 1a), whereas no correlation between serum uric acid and fructose levels in controls (P overall = 0.712,P non-linear = 0.43; Figure 1a).
Since there is no clear cut-off value for serum fructose levels in clinical practice, we divided women with PCOS according to the quartile of serum fructose levels. The serum uric acid levels were substantially higher combined with increasing quartiles of serum fructose levels and multivariate ANOVA confirmed differences in serum uric acid levels at different fructose levels (P < 0.001; Figure 1b).
Considering that clinical, biochemical, and endocrine characteristics may change with age,21 we further stratified women with PCOS by age (< 25, 25-29, 30-34, and ≥ 35 years). Interestingly, there was no linear or non-linear correlation between serum uric acid and fructose levels when age was less than 25 years. In contrast, there was a non-linear association between serum uric acid and fructose levels in women with PCOS between 25 and 29 years of age (P overall = 0.001, P non-linear = 0. 01). After 30 years of age, there was a significant linear correlation between serum uric acid and fructose levels in women with PCOS (Figure 1c).