Growth Hormone Stimulation Testing Patterns Contribute to Gender Disparities in Growth Hormone Treatment

Abstract

Abstract Introduction: Growth hormone (GH) registries demonstrate that males outnumber females 2:1 for all indications combined and 3:1 for the idiopathic short stature indication. The aim of this study was to determine if gender disparities in GH treatment are due to differences in rates of stimulation testing and/or GH prescribing. Methods: Retrospective chart review was performed including children aged 2-16 years seen for short stature or poor growth in 2012-2019 at a large tertiary referral center. Children previously diagnosed with GHD were excluded. Continuous variables, reported as medians [IQR], were compared by Wilcoxon rank sum test and categorical variables by Chi-squared test. A two-tailed p-value <0.05 defined statistical significance. Results: Of 10,125 children seen for evaluation of short stature or poor growth (35% [3542] females [F], 65% [6583] males [M]), 1,245 underwent GH stimulation testing (30% [379] F, 70% [866] M). A larger proportion of males than females were tested (M 13.2%, F 10.7%; p <0.001). Amongst the entire study population, females had lower height Z-scores than males (F -1.98 [-2.46, -1.44], M -1.80 [-2.24, -1.31]; p<0.001). This difference persisted in those who proceeded to GH stimulation testing (F -2.52 [-3.00, -2.04], M -2.18 [-2.6, -1.81]; p<0.001) and GH treatment (F -2.62 [-3.11, -2.07], M -2.19 [-2.60, -1.81; p<0.001). Mean difference between height Z-score and mid-parental height (MPH) Z-score for the entire population did not differ by sex (F -1.52 [-2.17, -0.87], M -1.52 [-2.04, -0.97]; p=0.76), but the difference was greater in females among those who underwent GH stimulation testing (F -1.95 [-2.57, -1.40], M -1.79 [-2.32, -1.32]; p=0.009) and started GH treatment (F -1.93 [-2.58, -1.48], M -1.80 [-2.30, -1.32]; p=0.016). Peak stimulated GH levels were similar for males and females (F 9.6 [6.0, 13.6] ng/mL, M 9.4 [6.1, 13.2] ng/mL, p=0.62). The proportion of children prescribed GH after stimulation testing did not differ by gender (F 55% [208], M 56% [488]; p=0.63). This finding did not change upon sub-analysis by peak stimulated GH concentration groups (peak GH concentrations <7 ng/mL, 7-10 ng/mL, and >10 ng/mL). Conclusion: The male predominance among children seen for subspecialist evaluation of short stature was compounded by a greater proportion of those males subsequently undergoing GH stimulation testing despite less severe short stature. Although females who underwent GH stimulation testing had greater height deficit from their genetic potential than tested males, peak stimulated GH concentrations and GH prescription rates were similar by sex. Thus, gender disparities in GH treatment occur at the subspecialist referral and stimulation testing, but not GH prescription, steps. Further, GH stimulation test results failed to account for the more severe shortness among tested females, yet another limitation identified with such testing.