For the primary analyses, descriptive statistics for each outcome of interest will be computed separately by exposure cohort. For growth and bone development metrics, mean and SD will be presented for continuous metrics related to height, weight, and BMI, including change from baseline to end of follow-up in standing height, height percentile, height SDS, height velocity SDS, weight, weight percentile, weight SDS, BMI, BMI percentile, and BMI SDS. The proportion of participants whose height SDS at the end of follow-up is more than 1 or 2 standard deviations lower than their baseline height SDS will also be reported.
Tanner staging, a categorical maturation and pubertal development metric, will be described using counts and percentages by exposure cohort, and by sex within each exposure cohort. The cumulative incidence and IR of bone fractures and neurological adverse events, overall and by type, will be estimated within each exposure cohort.
Outcomes of interest will be compared between exposure cohorts as exploratory analyses. Conventional linear regression models will be fit on the PS-matched sample for continuous outcomes, including change in height SDS, change in weight SDS, age at PHV, and age at Tanner stage progression. To compare bone fractures and neurological adverse events, generalised linear regression models will be fit on the PS-matched sample.
In order to increase sample size and statistical power, the primary and exploratory analyses are proposed to be performed on a pooled sample of participants from Europe (UK and EU countries) and North America (US and Canada).