Statistical analysis
Statistical analysis was performed using an SPSS 26.0 software package
and SAS 9.4. Characteristics of index and subsequent pregnancies were
compared using paired t-tests for parametrically distributed data and
Wilcoxon signed-rank tests for non-parametrically distributed data.
Differences in proportions of dichotomous variables between index and
subsequent pregnancies were analysed using McNemar’s test. Chi-squared
tests were used to calculate the relative risk (RR) of adverse outcomes
in subsequent pregnancies. Characteristics of subsequent pregnancies
with and without LGA, with and without SGA, and with and without the
composite neonatal outcome were compared using independent t-tests for
parametrically distributed data, Mann-Whitney U tests for
non-parametrically distributed data and chi-squared tests for
dichotomous variables.
Binomial regression analysis was performed to estimate the RR of
recurrent SGA, LGA or composite ANO in a subsequent pregnancy. Potential
factors identified on univariate analysis were included in the model and
backward stepwise removal was performed in order to identify independent
predictors of each outcome of interest and their adjusted RR.
Results are expressed as mean and standard deviation for parametric data
and median and interquartile range for non-parametric data, unless
otherwise stated. Critical significance is taken at 5%.