2.3. Data analysis
For all descriptive statistics (e.g., baseline clinical and demographic
characteristics, 6MMP:6TGN ratios, etc.), categorical variables were
summarized as counts and percentages. Continuous variables were
summarized as medians with first to third quartiles, or as means with
standard deviation, as appropriate.
All inferential statistical testing was two-sided with an alpha level
set to 0.05 (P<0.05). A paired t-test was used if the
data was normally distributed, and parametric testing was appropriate.
If non-parametric testing was appropriate, a Wilcoxon matched-pairs
signed rank test was used. For secondary endpoints with categorical or
binary variables (e.g., differences in incidence of symptom reduction),
a Chi-square test or Fisher’s exact test was used, as appropriate. For
paired continuous variables (e.g., before and after allopurinol in the
same patient), a paired t-test was used when assumptions for parametric
testing were met (i.e., normal distribution of data), while a Wilcoxon
matched-pairs signed rank test was used when non-parametric testing was
more appropriate. Finally, for non-paired continuous data, a t-test was
used when parametric testing was appropriate, and a Mann-Whitney test
was used as the non-parametric alternative. All statistical testing was
performed in SAS JMP v16 (SAS, Cary, NC), while figures were created
using GraphPad Prism v9.5 (GraphPad, San Diego, CA) and using R v4.1.0
(R Foundation for Statistical Computing), RStudio (RStudio), and the
ggplot2 package v3.4.2.11