*4.3. Statistical Analysis*

All data were assessed for normality using the Shapiro-Wilk test. Population characteristics (maternal age, parity, gestational age, mode of delivery, baby sex, birth weight and placenta weight) were tabulated. Data are presented as mean ± SD if normally distributed or median and interquartile range (IQR) if non-parametric. Potential confounders, including maternal age, mode of delivery, birth weight and placental weight were assessed with univariate analysis for each of the outcome variables prior to group comparisons being completed. None of these factors were independently associated with changes in outcomes measures. Statistical di fferences in placental creatine and GAA content, mRNA and protein expression between groups were established with either t-tests or Wilcoxon rank-sum, as appropriate. Baby sex was considered as a covariate for each of these analyses. Correlations between creatine and GAA content, mRNA and protein expression data, with maternal characteristics and birth outcomes were determined using the Spearman rank correlation coe fficient. *p* ≤ 0.025 was considered statistically significant for direct comparisons and *p* ≤ 0.004 for correlations after a Benjamini-Hochberg adjustment for false-positives due to multiple comparisons. All analyses were undertaken using SPSS ® (Version 23, IBM Corporation, 2015, Armonk, NY, USA).
