Comparison of long-acting versus short-acting growth hormone therapy in children with idiopathic short stature: a retrospective cohort study with advanced causal inference methods

Abstract

To compare the efficacy and safety of polyethylene glycol-modified recombinant human growth hormone (PEG-rhGH) versus daily rhGH in children with idiopathic short stature (ISS) during 12–24 months of follow-up and to propose individualised treatment strategies incorporating mid-parental height standard deviation score (MPH-SDS) correction for genetic potential. This single-centre retrospective cohort study enrolled patients aged 4–14 years with ISS consecutively between January 2018 and December 2024. Treatment assignment reflected real-world physician–guardian shared decision-making after counselling regarding observation, daily rhGH and PEG-rhGH, and prescribed starting doses averaged 0.17 ± 0.03 mg/kg/week for PEG-rhGH and 0.23 ± 0.04 mg/kg/week for daily rhGH. Propensity score matching (1:1, calliper = 0.2) and inverse probability treatment weighting controlled for confounding. Linear mixed-effects models estimated treatment × time interactions, and marginal structural models adjusted for time-dependent confounding. Multiple sensitivity analyses, including E-value calculation, negative control outcomes and sex-stratified subgroup analyses, were conducted. Among 596 patients (187 PEG-rhGH, 213 daily rhGH and 196 controls), 12-month height SDS changes were 0.82 ± 0.31, 0.65 ± 0.28 and 0.18 ± 0.21, respectively. Adjusted mean differences versus controls were 0.64 SDS (95% confidence interval [CI]: 0.58–0.70) for PEG-rhGH and 0.47 SDS (95% CI: 0.41–0.53) for daily rhGH (both p < 0.001). Direct comparison showed PEG-rhGH superior by 0.17 SDS, increasing to 0.31 SDS after MPH-SDS adjustment (95% CI: 0.24–0.38, p < 0.001), exceeding the preset minimal clinically important difference of 0.3 SDS. At 24 months, height SDS improvements were 1.38 ± 0.42, 1.09 ± 0.38 and 0.31 ± 0.27, respectively. Sex-stratified analyses confirmed consistent effects in boys and girls (interaction p = 0.72). Children with lower MPH-SDS showed greater benefit from PEG-rhGH. Safety profiles were comparable. Growth hormone treatment significantly improved height SDS in children with ISS, and PEG-rhGH demonstrated marginal superiority over daily rhGH that reached clinical significance after MPH-SDS correction. Children with lower genetic height potential preferentially benefit from long-acting formulations.