Introduction

Short stature, defined as height at or below two standard deviations below the mean for sex and age, is a common reason for referral to pediatric endocrinologists. The initial assessment entails ruling out any underlying medical conditions or genetic syndromes that could hinder normal growth. This is frequently followed by the assessment of growth hormone (GH) secretion and its efficacy as measured by insulin-like growth factor 1 (IGF1) levels, as well as by conducting GH stimulation tests. The past decade has witnessed the expansion of the prescription of recombinant human growth hormone (rhGH) to a broader range of indications and greater usage beyond simply replacing growth hormone deficiency (GHD) to encompass hormonal augmentation therapy in situations where deficiency has not been established and idiopathic short stature (ISS) is defined [1].

GH is an anabolic hormone that determines lean body mass. It stimulates whole-body anabolism with protein accretion occurring primarily in skeletal muscle, while also exerting tissue-specific effects on other body tissues. IGF1 is synthesized within the liver and released into the bloodstream in response to GH stimulation, with its concentrations serving as an indicator of GH status [2]. In addition to its impact on increasing linear growth, IGF1 facilitates muscle growth through various pathways [3, 4].

The secretion of IGF1 and its activity can be influenced by nutritional status in young children with undernutrition-depressing IGF1 production [5]. Furthermore, food deprivation may dampen the myogenic effects of IGF1 stimulation, as demonstrated in animal models [6]. Treatment with rhGH increases IGF1 levels, therefore, monitoring both IGF1 and growth constitutes an acceptable parameter of therapeutic safety and efficacy [7, 8]. However, there is a lack of data regarding the impact of rhGH treatment on body composition in relation to IGF1 levels.

In January 2018, our Pediatric Endocrine Unit implemented the analysis of body composition by means of bioimpedance analysis (BIA) as part of the standard intake assessment of subjects referred for endocrine consultation [9]. The escalating rates of obesity have led to a significant upsurge in cardiovascular risk factors in youth [10,11,12]. In addition, low muscle mass may increase the risk for cardiovascular complications, even in the absence of obesity [13]. While the body mass index (BMI) percentile is the most common tool used for identifying children at risk, it lacks the capacity to identify the ones with low muscle mass whose BMI percentiles may be within the normal range. Our group described the role of body composition assessment in identifying cardiovascular risk factors in pediatric patients with overweight/obesity, celiac disease, and non-classic congenital adrenal hyperplasia [14,15,16]. In the current study, we investigated the role of IGF1 levels in the determination of body composition components in children and adolescents treated with rhGH for short stature.

Methods

Study design and population

This real-life, observational study included pediatric participants (aged 5–18 years) followed for either GHD or ISS at the Institute of Pediatric Endocrinology in Dana-Dwek Children’s Hospital. Each patient underwent at least one body composition assessment BIA during routine follow-up between January 2018 and August 2023.

Querying the BIA database yielded a list of patients with the sole diagnosis of either GHD or ISS. The electronic medical records of those children and adolescents were examined, and sociodemographic and clinical data at the time of BIA assessment were retrieved. Excluded were patients with genetic syndromes, multiple pituitary hormone deficiency, a history of malignancy treated with chemotherapy and/or radiotherapy, and medical conditions and/or medications that could adversely affect growth. Sociodemographic characteristics, perinatal history, medical conditions, medications for, and family history of cardiovascular disease risk factors (obesity, hypertension, type 2 diabetes, and dyslipidemia) were retrieved from the hospital’s electronic medical files. Clinical data, anthropometric measurements, vital signs, pubertal staging, and laboratory evaluation were extracted at the time of BIA assessments. The study protocol was approved by our medical center’s Institutional Review Board which waived informed parental consent (065-18-TLV). The data were handled in accordance with the principles of good clinical practice.

Definitions of study variables

Socioeconomic position (SEP) grading

The SEP as determined by home address (SEP cluster and SEP index) was based upon the Israel Central Bureau of Statistics’ Characterization and Classification of Statistical Areas within Municipalities and Local Councils by Socio-Economic Level of the Population 2015 [17].

Growth hormone axis evaluation

Patients were referred for a GH provocation test after completing laboratory tests to exclude chronic medical conditions, such as hypothyroidism, active celiac disease, chronic renal insufficiency, and inflammatory bowel disease. Patients with a peak GH level lower than 7.5 ng/mL were referred for a second provocation test. Sex hormone priming of testosterone or estradiol was administered prior to testing in boys older than 13 years and girls older than 11 years, respectively [18]. The diagnosis of isolated GHD required evidence of insufficient GH reserve in two stimulation tests performed with different stimuli (glucagon, clonidine, and arginine) and with no evidence of other hypothalamic-pituitary dysfunction [19]. Patients diagnosed with GHD were referred for brain magnetic resonance imaging (MRI) to assess the structure of the pituitary gland and rule out any underlying structural abnormalities or lesions that could impact the management of GHD before the initiation of rhGH therapy.

rhGH therapy

Definitions of diagnoses and therapeutic doses of rhGH according to indication have evolved since the original FDA indication for GHD in 1985 [18]. In 2003, the FDA expanded rhGH use to the treatment of ISS, defined as height z-score ≤−2.25 in a healthy child with sufficient GH reserve [18]. Of note, the EMEA has not granted approval for this indication, resulting in limited global adoption. Nevertheless, the Israeli Ministry of Health has granted approval for it, following the guidelines published in 2016 [8]. The decision for rhGH treatment of children and adolescents diagnosed as having ISS was made jointly by the patient, the parents, and the treating pediatric endocrinologist. The initial dosage was determined according to the 2016 guidelines with subsequent adjustments made according to the findings of clinical and laboratory assessments. The rhGH dosage was recorded during the BIA assessment, and the total cumulative rhGH dosage was determined by multiplying the recommended average daily dose (in μg/kg/day) by the total number of days the treatment had been administered, with the final result being converted from micrograms to grams.

IGF1 monitoring

The goal of rhGH treatment is to promote linear growth and help individuals achieve a normal height for their sex and age. The dosage is adjusted according to the individual’s response to treatment and other clinical considerations, such as growth velocity and pubertal stage [1]. Adherence and IGF1 production in response to rhGH dose changes is monitored by periodic measurement of serum IGF1 levels. Furthermore, the interpretation of IGF1 levels with respect to sex, age, and pubertal development stages serves to aid in the prevention of hyper-augmentation of the GH-IGF1 axis, with the goal of avoiding an IGF1 exceeding two z-scores [8]. In the current study, IGF1 levels were converted to sex- and age-appropriate z-scores according to normal reference ranges in order to allow standardization between individuals [20, 21].

Clinical characteristics

Height, weight, and BMI calculated as weight (kg) divided by height (m) squared were converted to sex- and age-specific standard deviation scores (z-scores) by means of CDC 2000 growth charts [22]. The routine clinical evaluation in the endocrine clinic also included the calculation of mid-parental height (paternal height [cm] + maternal height [cm] ± 13 cm/2), from which the mid-parental height z-scores were derived [23]. Delta height z-scores were calculated as the difference between the patient’s current height z-score and the mid-parental height z-score. Weight status was defined according to BMI z-scores as follows: underweight as BMI percentile ≤5th percentile (z-score ≤−1.645), overweight as BMI percentile ≥85th and <95th percentiles (1.04 ≤ z-score <1.645), and obesity as BMI percentile ≥95th percentile (z-score ≥1.645) [24, 25].

Prematurity was defined as gestational age (GA) < 37 weeks. Birth weight z-scores were calculated by PediTools Electronic Growth Chart Calculators based on the Fenton growth chart for preterm infants [26]. Appropriate birth weight for gestational age (AGA) was defined as corrected birth weight z-scores between −1.645 to 1.645, small for gestational age (SGA) as birth weight z-scores <−1.645, and large for gestational age (LGA) as birth weight z-scores >1.645.

Systolic and diastolic blood pressure percentiles according to sex, age, and height percentile were determined via an online age-based pediatric blood pressure calculator [27]. The pubertal stage was graded according to Marshall and Tanner [28, 29]. Puberty onset was defined as genitalia Tanner stage 2 in boys with a testicular volume >3 mL and breast bud appearance in girls. Full puberty was indicated when pubertal signs matched Tanner stage 5.

Body composition assessment

Body composition was assessed by BIA (Tanita Body-Composition Analyzer, Tanita MC-780 MA, and GMON Professional Software) [30]. The BIA covered both the entire body as well as segmental analyses (trunk, upper and lower limbs) for fat and muscle. Measurements took place during routine clinic visits, preferably in the morning (8:00 AM to 1:00 PM), with subjects in a fasting, non-exerted state. The process involved standing barefoot on the analyzer and gripping the handles, and it took roughly 1 minute per subject. The BIA report contained data on whole body and segmental analysis of fat and muscle: fat percentage (FATP), fat mass, and muscle mass. Calculated variables included appendicular skeletal muscle mass (ASMM, the sum of limb muscle mass) and muscle-to-fat ratio (MFR = ASMM/fat mass). Z-scores for ASMM and MFR were determined by BIA pediatric reference curves [31].

Statistical analysis

Data were analyzed by IBM SPSS software (IBM SPSS Statistics for Windows, Version 29; IBM Corp., Armonk, NY). Continuous data were presented as mean ± standard deviation (normal distribution) or median [interquartile range] (skewed distribution), and as number and percentage for categorical variables. Differences in continuous data between groups were examined by independent-sample t-tests (normally distributed data) or Mann–Whitney U-tests (skewed data). Fisher’s exact and Chi-squared tests were used to examine the differences in categorical data, as appropriate. Generalized estimating equation (GEE) linear models were used to explore the contribution of clinical factors (sex, study group, age, height z-score, delta height z-score, cumulative rhGH dose, and IGF1 z-score) to the body composition components (FATP, MFR z-score, and ASMM z-score). The statistical findings are based on all cases with valid data for all the variables in the model. The GEE model was chosen in order to produce regression estimates when analyzing repeated measures. A p-value of ≤0.05 was considered significant.

Results

The analyses included 305 BIA reports of 135 pediatric patients (89 boys, 65.9%), of whom 64 were diagnosed with GHD and 71 with ISS (Fig. 1). All patients in the GHD group underwent brain MRI as part of the routine evaluation of GHD, and the findings were abnormal in 11 (17.2%): ectopic pituitary and/or abnormal stalk (n = 4), small pituitary gland (n = 3), pituitary microadenoma (n = 1), Rathke cleft cyst (n = 1), and Chiari 1 (n = 2). Twenty-eight girls (60.8%) underwent karyotype analysis, all with normal findings.

Fig. 1
figure 1

The study flowchart. Growth hormone deficiency; ISS, idiopathic short stature; MPHD, multiple pituitary hormone deficiency

Full size image

Sociodemographic and medical background characteristics of patients stratified by study group (GHD or ISS) are presented in Table 1. The mean age of the entire cohort was 10.7 ± 3.0 years (range 5–17.2 years), of whom 74 (54.8%) were prepubertal and 61 (45.2%) were pubertal (Tanner stages 2–5). Their collective SEP was above average: the median SEP cluster was 8 (range 1 to 10), with a median SEP index of 1.378 (range −1.719 to 2.765). The patients’ median gestational age was 40 (range 25 to 42), with 20 (14.8%) having been born preterm. Most of the cohort were born AGA (n = 104. 77%), with a normal mean birth weight that had a wide standard deviation [−0.91 ± 1.30]. Eight children (5.9%) had celiac disease with negative celiac serology under a gluten-free diet, and 5 children (3.7%) were diagnosed with ADHD and treated by stimulants. There were no significant group differences in sociodemographic characteristics or medical background. Family history of cardiovascular risk factors (obesity, hypertension, type 2 diabetes, and dyslipidemia) also did not differ between the groups. Of note, while no significant group difference was found in the systolic blood pressure percentile, the diastolic blood pressure percentile was significantly higher in the GHD group (p = 0.050).

Table 1 Sociodemographic and clinical characteristics of patients with growth hormone deficiency (GHD) or idiopathic short stature (ISS)
Full size table

The genetic height potential, as expressed by the MPH z-score, was taller in the GHD group compared to the ISS group (p < 0.001). At their first BIA, the subjects in the GHD group were significantly taller than those in the ISS group (p < 0.001). There was a significant association between the height z-score and the MPH z-score (r = 0.319, p < 0.001) of all the patients in the entire cohort. The achievement of the genetic potential, as expressed by delta height z-scores, did not differ between the groups.

At first BIA, the vast majority of the cohort (n = 116; 85.9%) was in the normal weight category. However, patients in the GHD group had significantly higher BMI z-scores compared to those in the ISS group (p < 0.001). Comparisons between body composition components of the two study groups revealed higher fat percentage, higher truncal fat percentage, and lower MFR z-score in the GHD group (p = 0.010, p = 0.027, and p = 0.050, respectively). The muscle component (represented by the ASMM z-score) of the entire cohort was −0.40 ± 0.99 and it did not differ between the groups. Figure 2A–C displays the distribution of body composition components (FATP, MFR z-scores, and ASMM z-scores) in the GHD and ISS groups at the first BIA assessment.

Fig. 2
figure 2

Violin plots for the distribution of body composition components and insulin-like growth factor 1 in pediatric patients with growth hormone deficiency and those with idiopathic short stature at first BIA. A Fat percentage (FATP); B appendicular skeletal muscle mass (ASMM) z-score; C Muscle-to-fat (MFR) z-score; D IGF1 z-score

Full size image

The clinical characteristics, anthropometric measurements, and body composition components of the patients at first BIA evaluation, stratified according to diagnosis and according to rhGH treatment at first BIA, are presented in Table 2. At first BIA, 90 (66.7%) patients had already been treated by rhGH. In non-treated patients, the FATP and truncal FATP were significantly higher in the GHD group compared to ISS group (p = 0.002 and p = 0.008, respectively), but not significant in the rhGH-treated patients. The rhGH-treated patients in both the GHD and ISS groups had lower truncal FATP (p < 0.001, p = 0.004, respectively). Finally, the ASMM z-score was higher in the rhGH-treated patients in both the GHD and ISS groups (p < 0.001 for both).

Table 2 Characteristics of patients with growth hormone deficiency (GHD) or idiopathic short stature (ISS) at first bioimpedance analysis assessment (BIA), classified as treated versus not treated with rhGH at first BIA
Full size table

In 239 (78.4%) of BIA assessments, patients had been treated by rhGH, for a mean treatment duration of 3.7 years, without significant difference between the groups (p = 0.903). Forty-three (67.2%) patients with GHD had received a median dose of 31 mcg/kg/dy [IQR 27.0, 37.5], while 47 (66.2%) patients with ISS had received 40 mcg/kg/dy [IQR 33.0, 43.0], p < 0.001. The mean IGF1 z-score at BIA assessment for the entire cohort was −0.10 ± 0.97, and the distribution of IGF1 z-scores of the GHD and ISS groups at their first BIA assessment is presented in Fig. 2D. The IGF1 z-score was significantly higher in both the GHD and ISS patients who were treated by rhGH (p < 0.001 for both groups). Only one patient diagnosed with GHD had an elevated IGF1 z-score of 2.67 at 7 months of treatment, which subsequently normalized during follow-up after rhGH dose adjustment.

GEE linear models were applied for identification of the contributing factors to the body composition components (Table 3). A higher FATP was associated with female sex (p < 0.001), GHD diagnosis (p < 0.001), lower height z-score (p = 0.036), and greater delta height z-score (p = 0.041). SEP index, age, cumulative rhGH dose, and IGF1 z-scores were not significant in the GEE model for FATP. A higher MFR z-score was associated with female sex (p = 0.049), an ISS diagnosis (p = 0.005), a higher height z-score (p = 0.004), and a smaller delta height z-score (p = 0.048). SEP index, age, cumulative rhGH dose, and IGF1 z-scores were not significant in the GEE model for the MFR z-scores. A higher ASMM z-score was associated with female sex (p < 0.001), older age (p < 0.001), higher height z-score (p = 0.003), and higher IGF1 z-score (p = 0.021). SEP index, study group, delta height z-score, and cumulative dose rhGH were not significant in the GEE model for the ASMM z-score. Females demonstrated a greater increase in IGF1 z-score over time after rhGH cumulative dose adjustment (β = 0.44, SE = 0.16, p = 0.006).

Table 3 Generalized estimating equations (GEE) linear models applied to evaluate contributory factors for body composition parameters: fat percentage (FATP), muscle-to-fat ratio (MFR) z-score and appendicular skeletal muscle mass (ASMM) z-score
Full size table

Discussion

To the best of our knowledge, our study is the first to examine the relationship between IGF1 levels and body composition components of fat and muscle in pediatric patients with ISS under treatment with rhGH. We found that individuals diagnosed with GHD exhibited increased adiposity characterized by a higher fat percentage, a higher truncal fat percentage, and a compromised muscle-to-fat ratio z-score compared to those with ISS. Of note, the muscle component (ASMM z-scores) did not differ between the GHD and ISS groups. Given that the IGF1 levels in patients play a significant role in determining their muscle mass, we propose that the rhGH treatment that is administered with the aim of restoring normal IGF1 levels may potentially mitigate the detrimental metabolic impact associated with increased adiposity.

There is no argument that height has a strong heritable trait [32], as had been supported by our results demonstrating the link between patients’ height z-scores and their mid-parental height z-scores, even for patients under rhGH treatment. Notably, our ISS patients were of shorter stature compared to those with GHD, a finding that was consistent with a shorter familial height. As a result, the comparison of delta height z-score, which indicates the achievement of genetic height during treatment, showed no difference between the two groups. These observations are in line with the objectives of rhGH treatment being different for each group: the primary goal of rhGH treatment for GHD is to attain the individual’s genetic height potential, while the primary focus of the treatment for ISS is upon achieving a height within or above the third percentile. Of note, patients with ISS were treated with a higher dose of rhGH compared to patients with GHD and attained similar IGF1 levels. This may imply a variable degree of GH resistance in patients with ISS, which may limit the effect of rhGH therapy [33].

Genetic predisposition has also been identified in body composition parameters [34], and genetic influence may contribute to the hereditary nature of metabolic disorders [35]. Both of the current study groups exhibited a similar low prevalence of familial obesity and metabolic syndrome components, suggesting similarity in metabolic risk. Most of our patients with GHD had already been treated by rhGH, and they were characterized by a higher fat percentage, a higher truncal fat percentage, and higher diastolic blood pressure percentiles than our patients in the ISS group. The statistical model we applied found that the diagnosis of GHD was related to increased adiposity, after adjusting for cumulative rhGH dose. These findings point towards a metabolic role of endogenous GH and raise several questions regarding the extent of reduction in adiposity, the time required for such reduction, and whether normal adiposity levels are attainable.

Increased adiposity, especially in visceral locations, has been reported in both adult and pediatric patients with GHD prior to treatment initiation [36,37,38]. Studies in adults found a marked reduction in total body fat content under treatment with rhGH [38,39,40,41] and, more specifically, in visceral adiposity [42,43,44], attributed to the lipolytic effect of GH. Poor compliance with treatment, marked by frequent and extended breaks, reportedly led to deteriorating metabolic parameters [45]. A reduction in adiposity had been noted mostly in the first months of treatment, with contradicting reports on the sustainability of achievement in pediatric patients [46, 47]. Examining the body composition of adults who had received rhGH treatment during childhood for isolated GHD could contribute valuable insight into issue.

The primary concern with increased adiposity is its association with cardiovascular disease risk. This link has primarily been shown in adults, with a limited number of studies indicating a similar association in pediatric populations [15]. In the current study, we observed a higher diastolic blood pressure percentile in patients with GHD but none in the systolic blood pressure percentile. This could herald an increased metabolic risk which could stem from an unfavorable body composition. The MFR z-score, which accounts for the muscle-to-fat ratio adjusted for sex and age, was lower in our GHD group. We had earlier demonstrated the strong predictive value of the MFR z-score in the detection of cardiometabolic risk in pediatric patients with obesity, type 1 diabetes, and non-classic congenital adrenal hyperplasia [14, 16, 48]. We now consider that employing therapeutic approaches that incorporate rhGH and that aim to address both fat and muscle components of body composition may decrease potential metabolic risks in a pediatric GHD population.

Sexual dimorphism could influence the way a medical condition is presented, progresses, and responds to treatment. There are variations in body composition between the sexes already at birth, with both males and females exhibiting comparable fat mass while males typically exhibit greater length and a higher proportion of lean mass [49]. As individuals progress into adulthood, boys undergo an increase in muscle mass with a concurrent decrease in fat, whereas females exhibit a lower propensity for muscle development and are more likely to accumulate peripheral fat [50]. In line with expectations, our study identified female sex as a contributing factor for increased adiposity. Surprisingly, however, female sex was identified as a contributor to a favorable sex- and age-adjusted muscle mass (ASMM z-score), as well as to a better muscle-to-fat ratio z-score.

The acquisition of skeletal muscle is affected by a complex interplay between genetic, hormonal, and environmental factors [51]. Of note, all of our patients are provided with nutritional counseling and recommended to routinely participate in physical activity according to the WHO guidelines for the pediatric population [52]. Since lifestyle habit questionnaires were not part of this study, we cannot comment upon the possibility that behavioral differences between the sexes account for more advantageous sex- and age-adjusted muscle mass among the females. Differences in the response to rhGH therapy had been observed between adult men and women with GHD [53] and sex hormone-dependent anabolic effects of GH could partially explain some of our findings in the pediatric population. The exact biological mechanisms by which sex steroid patterns could be a contributing factor, however, remain obscure. One report in the pediatric population suggested that sex hormones may lead to distinct GH responsiveness by influencing the production of IGF1 [54]. This is supported by our finding of greater responsiveness to rhGH therapy in females as manifested by a greater increase in IGF1 z-scores over time.

We demonstrated that sex- and age-adjusted IGF1 levels related to muscle mass but not to adiposity. It has been previously reported that IGF1 plays a critical role in the regulation of protein metabolism by enhancing the transport of amino acids into muscle cells, thus supporting a continuous supply of components for protein synthesis [3, 6]. Furthermore, it inhibits the degradation of pre-existing proteins within muscle tissue. By doing so, IGF1 makes a substantial contribution to the preservation and expansion of skeletal muscle, influencing both muscle size and strength [55, 56]. IGF1 has been reported to also possess metabolic and health-promoting capabilities, such as atheroprotective, neuroprotective, and insulin-like effects [57]. However, the myotrophic effect of IGF1 may be attenuated by insulin resistance in the case of increased adiposity [4]. Given that the cumulative dose of rhGH in our patients did not have a direct impact on body composition parameters, it can be inferred that the individual responsiveness to rhGH, mediated by endogenous IGF1 secretion, is the more likely factor influencing muscle mass.

The study has several limitations, such as its cross-sectional design, which precludes our ability to establish a causal relationship between exposure and outcomes. Additionally, it may not be representative of the pediatric GHD and ISS populations, since it includes only youth who visit our facilities. Lastly, the reliance on self-reported family histories of cardiovascular risk factors may introduce potential bias due to accuracy and recall issues. The primary strength of this study derives from the standardized and comprehensive assessment by trained medical experts, which supports the reliability of measurements, including those of blood pressure, anthropometrics, pubertal status, and body composition assessment. Another strength lies in the uniformity of medical care provided by the same multi-professional team in a single hospital-based tertiary center and the comprehensiveness of patient follow-up, including medical nutrition therapy and promotion of a healthy lifestyle.

Conclusions

The findings of our study demonstrate an inclination towards greater adiposity in young individuals with GHD compared to those with ISS. This is manifested by higher fat percentages, truncal fat percentages, and an altered muscle-to-fat ratio z-score, while the muscle component remained unaffected. IGF1 levels were not found to be significantly associated with body fat percentage, suggesting a mechanism not linked to IGF1 in the determination of body fat. Given the pivotal role of IGF1 in muscle mass regulation, we consider that rhGH treatment targeting normal IGF1 levels may offer a potential approach to mitigate the increased adiposity in patients with GHD by means of promoting muscle growth.