Statistical analysis
Statistical analyses were performed using Statistical Package for Social Sciences version 24 (IBM Corp., Armonk, NY). According to the Kolmogorov-Smirnov test, continuous variables were divided into normally and non-normally distributed variables. Normally distributed continuous variables were reported as mean ± standard deviation, whereas non-normally distributed continuous variables were reported as median (quartile spacing). Differences between the PCOS and control groups were examined using an independent sample t-test (normally distributed variables) or the Mann-Whitney U test (non-normally distributed variables). One-way analysis of variance (ANOVA) with Tukey’s or Dunnett’s post-hoc test (two-sided) was conducted for multi-group comparisons. Logistic regression analysis was used to determine the odds ratios (OR) with 95% confidence intervals (CI) for various characteristics associated with hyperuricemia in PCOS. All tests were two-sided, and statistical significance was defined as P< 0.05.
Restricted cubic splines (RCS) were conducted using R software (version 4.0.2, using packages “segmented,” “splines,” “Hmisc,” “rms,” and “ggplot2”) to assess the relationship between serum uric acid and fructose levels in PCOS. We prespecified three knots located at the 10th, 50th, and 90th percentiles of serum uric acid, as recommended by Stone and Koo.17 The RCS is a smoothly joined sum of polynomial functions, which can avoid inappropriate linearity assumptions.18 The advantages of RCS include its ability to relate to the natural shape of the relationship and the detected sensitivity of nonlinear relationships.19, 20