Short stature is a common clinical issue in pediatric endocrinology with a diverse range of underlying causes, including both pathological conditions and constitutional factors. Although growth hormone deficiency (GHD) is a well-established and treatable cause of short stature, a subset of children exhibit significant growth failure despite having normal peak growth hormone (GH) levels on stimulation tests. This discrepancy reveals the limitations of current diagnostic approaches and draws attention to underrecognized GH dysfunctions, such as bioinactive GH. Bioinactive GH denotes a phenotype in which circulating GH is immunoreactive but biologically ineffective; it is not a distinct therapeutic intervention but rather an identifiable subgroup among children receiving recombinant human growth hormone (rhGH) (1–10).

Bioinactive GH is a rare but clinically important condition characterized by circulating GH that is immunoreactive but biologically inactive due to structural abnormalities of the GH molecule, often associated with pathogenic variants in the GH1 gene. These abnormal GH isoforms fail to effectively bind or activate the GH receptor, resulting in impaired activation of downstream signaling pathways, including JAK2–STAT5 signaling, which is essential for insulin-like growth factor-1 (IGF-1) production and linear growth. Despite normal GH secretion, affected children typically have significantly reduced serum IGF-1 and Insulin-like Growth Factor Binding Protein 3 (IGFBP-3) levels, along with a poor correlation between GH and IGF-1 concentrations, leading to short stature due to functional impairment of GH action (11–14). Initially described by Kowarski in 1978 (2) and further characterized at the molecular level by Takahashi in the 1990s (3), bioinactive GH remains underdiagnosed, with its true prevalence unknown due to limited awareness and the absence of standardized diagnostic tests. Recent reviews highlight the continued lack of standardized diagnostic tools and emphasize the need for harmonized international diagnostic criteria (15).

Previous studies have consistently shown that children with bioinactive GH respond well to exogenous rhGH therapy, despite normal endogenous GH secretion (15–17). Their growth patterns are largely comparable to those of patients with classical GHD, and in some cases, an even greater response has been observed (17). This paradox arises because exogenously administered rhGH retains full biological activity, effectively compensating for the functional defect of the endogenous hormone. Collectively, these findings indicate that bioinactive GH represents a GH-responsive phenotype with a pathophysiological profile more closely resembling GHD than other non-GHD etiologies of short stature.

Accordingly, the present study compares patients with bioinactive GH to two well-established non-GHD short stature cohorts: small for gestational age (SGA) and idiopathic short stature (ISS) (6, 8). Both SGA and ISS represent heterogeneous populations characterized by normal GH secretion but variable responsiveness to rhGH, thereby providing suitable comparator groups for evaluating the growth potential of bioinactive GH. Children born SGA typically experience intrauterine growth restriction (18–22) and demonstrate highly variable patterns of postnatal catch-up growth, whereas ISS is a diagnosis of exclusion applied to children with unexplained short stature in the absence of identifiable systemic, endocrine, or genetic abnormalities (8).

We hypothesize that children with bioinactive GH—representing a distinct subset of non-GHD short stature patients identified through IGF-1 generation testing—will demonstrate a stronger growth response to rhGH therapy, as measured by annual growth velocity, changes in height z-score, and final height z-score, compared with children diagnosed with SGA or ISS.

Data were analyzed using IBM SPSS Statistics Version 22.0 (26). Normality was assessed with the Shapiro–Wilk and Kolmogorov–Smirnov tests, together with evaluation of skewness and kurtosis. Parametric data (mean ± SD) were compared using one-way ANOVA with Bonferroni post-hoc tests, while non-parametric data (median, Q1–Q3) were analyzed with the Kruskal–Wallis test, followed by Mann–Whitney U tests with Bonferroni correction when significant. Within-group longitudinal changes were assessed using repeated-measures ANOVA with Greenhouse–Geisser correction for normally distributed variables, followed by Bonferroni-adjusted pairwise comparisons. Time-point multiplicity was controlled using Bonferroni correction across time points. A p-value < 0.05 was considered statistically significant.

Height, growth velocity, and biochemical measurements were included in year-specific analyses only when complete data for that time point were available. Complete-case analyses were performed without imputation, and missingness was assumed to be missing completely at random (MCAR) or missing at random (MAR). The reduction in sample size in later treatment years reflects patients who either discontinued rhGH therapy, transferred follow-up to other institutions, or had incomplete documentation for that interval.

In this study, out of 170 short-statured patients (height ≤ –2 z-score) who demonstrated a normal GH response (≥10 ng/mL) to clonidine or L-dopa stimulation tests and received rhGH treatment at our center, a total of 109 patients classified as bioinactive GH (n = 8), SGA (n = 27), and ISS (n = 74) according to the definitions provided in the Methods section were selected and included in the analysis. The selection of patients is summarized in Figure 1 .

Baseline characteristics are shown in Table 1 . Sex distribution was similar descriptively across groups (bioinactive GH: 38% male; SGA: 37% male; ISS: 40% male).

BW and BW z-score were significantly lower in the SGA group compared with both the bioinactive GH and ISS groups (p < 0.001). The mean BW was 2148 ± 466 g in the SGA group, 3105 ± 349 g in the bioinactive GH group, and 2947 ± 595 g in the ISS group. The mean BW z-score was –2.75 ± 0.55 in the SGA group, –0.81 ± 0.91 in the bioinactive GH group, and –0.76 ± 0.76 in the ISS group.

At treatment initiation, 25% of the bioinactive GH group, 44% of the SGA group, and 56% of the ISS group were pubertal.

No statistically significant differences were observed among the groups in terms of age, height age, BA, or maternal height. Paternal height differed significantly among groups (p = 0.017), being lowest in the bioinactive GH group (160.6 ± 3.79 cm) compared with SGA (165.9 ± 6.50 cm) and ISS (168.2 ± 6.80 cm). Similarly, target height z-score was significantly lower in the bioinactive GH group (–2.28 ± 0.66) than in the ISS group (–1.49 ± 0.84; p < 0.001).

Biochemical parameters related to the GH-IGF axis before treatment are shown in Table 2 . All patients in the bioinactive GH group had IGF-1 and IGFBP-3 z-scores below –2 z-scores. Both the absolute levels and z-scores of IGF-1 and IGFBP-3 in this group were markedly lower compared with the SGA and ISS groups (p < 0.001 for all comparisons).

There was no significant difference in pre-treatment GV among the groups (bioinactive GH: 3.7 ± 0.8 cm/year; SGA: 4.8 ± 1.7 cm/year; ISS: 4.7 ± 1.5 cm/year; p = 0.172).

During the first treatment year, GV increased substantially in all groups, with the highest mean GV observed in the bioinactive GH group (9.1 ± 1.4 cm/year) compared with SGA (8.2 ± 1.9 cm/year) and ISS (8.4 ± 1.9 cm/year; p = 0.504). GH doses were similar across groups (mean 0.032–0.033 mg/kg/day). Approximately half of the bioinactive GH group were pubertal at this stage, compared with 63% in the SGA group and 80% in the ISS group.

In the second year, the bioinactive GH group again showed the highest GV (8.6 ± 2.2 cm/year) relative to SGA (6.7 ± 2.1 cm/year) and ISS (7.5 ± 1.9 cm/year; p = 0.10). GH dose was comparable across groups (p = 0.31), and the proportion of pubertal patients increased in all cohorts, with ISS showing the highest rates.

In the third year, mean GV remained higher in the bioinactive GH group (8.2 ± 2.2 cm/year) than in the SGA (6.4 ± 1.7 cm/year) and ISS (6.9 ± 2.1 cm/year) groups (p = 0.372). GH doses did not differ significantly, and most patients in all groups were pubertal by this stage.

Final height analysis showed a greater total height gain in the bioinactive GH group (40.7 ± 12.5 cm) compared with SGA (28.2 ± 15.4 cm) and ISS (29.2 ± 11.0 cm), although this difference was not reach statistical significance (p = 0.071).

Longitudinal changes in height z-score from the initiation of rhGH therapy to final height are summarized in Table 3 . At baseline, mean height z-scores were similar across groups (bioinactive GH: –3.14 ± 0.33; SGA: –3.02 ± 0.60; ISS: –2.92 ± 0.62; p = 0.572).

After the first year of treatment, height z-scores improved in all groups (bioinactive GH: –2.33 ± 0.52; SGA: –2.48 ± 0.62; ISS: –2.59 ± 0.67; p = 0.434). Incremental gains were maintained in the second and third treatment years, with no significant differences between groups (p = 0.704 and p = 0.960, respectively).

At final height, mean z-scores reached –1.30 ± 0.77 in the bioinactive GH group, –2.11 ± 0.96 in the SGA group, and –1.79 ± 0.95 in the ISS group (p = 0.091). Within-group analyses confirmed significant improvements in height z-scores over time in all cohorts (p < 0.001). Corresponding GH doses and pubertal distribution are presented in Table 4 .

Figure 2 illustrates longitudinal changes in height z-scores across the three study groups.

Yearly and cumulative changes in height z-score (Δ height z-score), together with corresponding GH doses and pubertal distribution, are presented in Table 5 .

At the end of the first year, the mean Δ height z-score in the bioinactive GH group was 0.81 ± 0.37, higher than both the ISS group (0.33 ± 0.38, p = 0.024) and the SGA group (0.54 ± 0.48). GH dosing was comparable across groups, and the stronger response in the bioinactive GH group occurred despite a lower proportion of pubertal patients.

In the second year, the Δ height z-score remained higher in the bioinactive GH group (0.48 ± 0.25) compared with SGA (0.26 ± 0.26) and ISS (0.38 ± 0.38), though differences were not significant (p = 0.347). In the third year, Δ height z-scores were similar among groups.

Cumulatively over the first three years, the bioinactive GH group showed the highest total Δ height z-score (1.39 ± 0.21), compared with 1.05 in both SGA and ISS (p = 0.152). From baseline to final height, Δ height z-score was greatest in the bioinactive GH group (1.86 ± 0.85) versus SGA (0.85 ± 0.96) and ISS (1.21 ± 1.02), approaching statistical significance (p = 0.073).

Figure 3 shows the annual and cumulative Δ height z-score in the three groups—bioinactive GH, SGA, and ISS—following rhGH therapy.

Among the 100 patients who reached final height, 57 (57%) achieved a normal height (height > −2 z-score). In the bioinactive GH group, all 6 patients (100%) reached normal height, and 5 (83%) achieved their target height. In the SGA group, 11 out of 25 patients (44%) reached normal height and 9 (36%) reached target height, whereas in the ISS group, 40 out of 69 patients (58%) attained normal height and 29 (42%) reached target height.

Comparisons of target height z-scores and observed height z-scores at baseline, the third year of treatment, and final height are presented in Table 6 . At baseline, the difference between target and observed height z-scores (Δ height z-score_target–observed) was significantly higher in all groups, indicating a marked initial height deficit relative to genetic potential. The largest discrepancy was observed in the ISS group (1.41 ± 0.90), followed by the SGA (1.30 ± 0.70) and bioinactive GH (0.94 ± 0.77) groups.

After three years of rhGH therapy, all groups demonstrated significant improvement, with the gap between target and observed height z-scores narrowing considerably. In the bioinactive GH group, this difference became negative by year 3 (−0.45 ± 0.83) and further declined to −0.92 ± 0.87 at final height, indicating that most patients exceeded or approached their genetic potential. The SGA and ISS groups also showed progressive narrowing of the target–observed gap, reaching near parity by final height (−0.09 ± 0.84 and −0.02 ± 0.91, respectively), though less pronounced than in the bioinactive GH group.

Within-group changes over time were statistically significant in all three groups (p < 0.001).

This study provides one of the most detailed comparative analyses to date of growth outcomes in children with bioinactive GH versus two major non-GH-deficient short stature populations: SGA and ISS. All participants exhibited normal peak GH responses on stimulation testing yet demonstrated distinct patterns of responsiveness to rhGH therapy. Children classified as having bioinactive GH based on IGF-1 generation test results demonstrated a more pronounced and consistent growth response than both the SGA and ISS groups, reflected in annual growth velocity, increases in height z-score, and attainment of higher final height z-scores. These findings highlight the clinical relevance of bioinactive GH as a distinct phenotype and reinforce the diagnostic utility of IGF-1 generation testing in selected patients with unexplained short stature.

A notable strength of this investigation is the direct, protocol-consistent comparison of three clinically meaningful subgroups treated in a single tertiary center, ensuring standardized dosing, monitoring, and anthropometric assessments. This methodological uniformity minimizes variability related to treatment approach or measurement techniques, thereby increasing the reliability of between-group comparisons. The consistent superiority of growth outcomes in the bioinactive GH group further supports the validity of this classification, despite the absence of confirmatory molecular or bioactivity assays.

Children with bioinactive GH form a biologically distinct and highly treatment-responsive subgroup of non-GHD short stature. Our findings highlight the diagnostic value of IGF-1 generation testing in this context. Future prospective, multicenter studies integrating genetic and bioactivity confirmation are essential to refine diagnostic criteria and establish internationally applicable guidelines for clinical management.