- Research article
- Open Access
Growth hormone (GH) dose-dependent IGF-I response relates to pubertal height gain
BMC Endocrine Disorders volume 15, Article number: 84 (2015)
Responsiveness to GH treatment can be estimated by both growth and ∆IGF-I. The primary aim of the present study was to investigate if mimicking the physiological increase during puberty in GH secretion, by using a higher GH dose could lead to pubertal IGFs in short children with low GH secretion. The secondary aim was to explore the relationship between IGF-I, IGFBP-3 and the IGF-I/IGFBP-3 ratio and gain in height.
A multicentre, randomized, clinical trial (TRN88-177) in 104 children (90 boys), who had received GH 33 μg/kg/day during at least 1 prepubertal year. They were followed from GH start to adult height (mean, 7.5 years; range, 4.6–10.7). At onset of puberty, children were randomized into three groups, to receive 67 μg/kg/day (GH67) given once (GH67x1; n = 30) or divided into two daily injection (GH33x2; n = 36), or to remain on a single 33 μg/kg/day dose (GH33x1; n = 38). The outcome measures were change and obtained mean on-treatment IGF-ISDS, IGFBP3SDS and IGF-I/IGFBP3 ratioSDS during prepuberty and puberty. These variables were assessed in relation to prepubertal, pubertal and total gain in heightSDS.
Mean prepubertal increases 1 year after GH start were: 2.1 IGF-ISDS, 0.6 IGFBP3SDS and 1.5 IGF-I/IGFBP3ratioSDS. A significant positive correlation was found between prepubertal ∆IGFs and both prepubertal and total gain in heightSDS. During puberty changes in IGFs were GH dose-dependent: mean pubertal level of IGF-ISDS was higher in GH67 vs GH33 (p = 0.031). First year pubertal ∆IGF-ISDS was significantly higher in the GH67vs GH33 group (0.5 vs −0.1, respectively, p = 0.007), as well as ∆IGF-ISDS to the pubertal mean level (0.2 vs −0.2, p = 0.028). In multivariate analyses, the prepubertal increase in ‘∆IGF-ISDS from GH start’ and the ‘GH dose-dependent pubertal ∆IGF-ISDS’ were the most important variables for explaining variation in prepubertal (21 %), pubertal (26 %) and total (28 %) gain in heightSDS.
TRN 88–177, not applicable 1988.
The dose-dependent change in IGFs was related to a dose-dependent pubertal gain in heightSDS. The attempt to mimic normal physiology by giving a higher GH dose during puberty was associated with both an increase in IGF-I and a dose-dependent gain in heightSDS.
Insulin-like growth factors (IGFs) have been used in the diagnosis of growth hormone (GH) deficiency, to monitor the impact of GH replacement therapy on growth and to assess treatment compliance and safety [1, 2]. Monitoring the impact of GH treatment on growth is mainly based on measurement of serum IGF-I levels, and less often on IGF-binding protein 3 (IGFBP3) levels and the IGF-I/IGFBP3 ratio. Rudman et al. were the first to report the relationship between short-term IGF-I increments and GH growth response . Further short-duration studies in prepubertal children conducted by different groups found an increase in IGF-I from baseline to be a reliable sign of greater growth in response to GH [4–7]. Only two studies reported results for multiple variables (IGF-I, IGFBP3 and their molar ratio). These were non-randomized, 1-year clinical trials in prepubertal children with and without GH deficiency (GHD) [8, 9]. They both observed an increase in IGFs during GH treatment as a sign of high GH sensitivity and treatment compliance. To our knowledge, the relevance of the variables IGF-I, IGFBP3 and the IGF-I/IGFBP3 ratio in relation to prepubertal, pubertal and total gain in height has not been previously reported in GH-treated children.
Under normal conditions, serum IGF-I level increases slowly during childhood before rising to a peak in puberty. This peak correlates with pubertal stage due to the action of sex steroids to increase GH secretion [10–13]. Factors explaining pubertal growth in response to GH therapy: gender, age, the difference between the child’s height standard deviation score (SDS) and midparental heightSDS (diffH-MPH) at the onset of puberty, and GH dose were identified from the KIGS observational study, but the IGF-I variable was not available in that study . There have only been two published randomized trials in GH-deficient pubertal children on GH treatment receiving different GH doses. Both reported a greater pubertal height gain in high GH-dose groups (50–100 μg/kg/d) accompanied by an increase in IGF-I [15, 16]. Our group has recently published results from a randomized study in non-GH-deficient children followed from early puberty to adult height (AH): data show the greater the increase in IGF-I, the greater the gain in height . In children with low GH secretion diagnosed with idiopathic isolated GHD (IIGHD), we have also reported that GH dosing, mimicking the physiological pubertal increase in GH secretion, has a dose-dependent effect on the gain in heightSDS until AH. In this randomized GH-treatment trial, pubertal height gainSDS and AHSDS were greater in children randomized to a high GH dose (67 μg/kg/d) than a standard dose (33 μg/kg/d) [18, 19]. The hypothesis of the present analysis of prospectively collected data in the aforementioned clinical trial was that the IGFs would follow the normal pubertal change in GH-deficient children receiving a higher, more physiological, GH dose during puberty. The secondary aim was to explore the relationship between serum IGF-I, IGFBP-3 and the IGF-I/IGFBP-3 ratio and gain in height until AH.
Patients and methods
The study (TRN number 88–177) was approved by the Ethics Committees at the Universities of Gothenburg, Lund, Linköping, Uppsala, Huddinge and Umeå and by the Karolinska Institute (dnr LU 426–1988). Informed consent was obtained from the children and their parents verbally with written registration date in the patients’ medical file.
The study was a nationwide, randomized, multicenter trial conducted from 1988 to 2009 and its design has been described previously [18, 19]. In brief, all children with IIGHD had received GH at a dose of 33 μg/kg for at least 1 year prior to the onset of puberty. After the development of clinical signs of puberty, the children were allocated randomly, without stratification, into three dose groups: 33 μg/kg once daily (GH33x1); 67 μg/kg once daily (GH67x1) or 33 μg/kg twice daily (GH33x2). The children were followed until AH (the observed height at a growth rate of < 1 cm during the preceding 12 months). They were seen at least once a year at a university hospital to monitor treatment safety and efficacy and at their local children’s hospital at 3-monthly intervals.
GH dose reduction
In line with the protocol, it was accepted that any patient could reduce GH dose or stop treatment if he/she was satisfied with the height development. The reduced doses used were 25 % less than the randomized dose (doses were 50 μg/kg/d and 25 μg/kg/d for the high and low dose groups, respectively). In the intention-to-treat (ITT) population the GH dose was reduced in 35 children and GH treatment was stopped in 5 children; this affected 44 % of children in the high dose and 29 % in the standard dose groups. For efficacy analyses, only IGF-ISDS levels obtained during treatment with the randomized dose were used, but for safety analyses all IGF-ISDS measurements were used.
Inclusion and exclusion criteria
All patients were diagnosed with IIGHD in the 1980s. GH deficiency was diagnosed based on a GH response cut off corresponding to “10 μg/L” [20, 21] in two GH provocation tests, mainly the arginine–insulin tolerance test (AITT). A positive response to GH treatment was also an inclusion criterion. This was ascertained by an increase in height velocity of at least 50 % during the first prepubertal year of GH treatment (33 μg/kg/d). IGF-I, IGFBP3 and the IGF-I/IGFBP3 ratioSDS were measured but were not required for diagnosis. Children with any hormone insufficiency in addition to GH or with significant chronic diseases or syndromes were excluded from the study.
The study population has been described previously . Briefly: a total of 149 (116 boys) short children (< −2SDS) with low GHmax during AITT (GHmaxAITT) were enrolled in the study between 1989 and 2000 and form the safety population. Of the enrolled children, 38 were excluded from efficacy analyses due to protocol violation/wrong inclusion: 1 child had a bone age (BA) delay of 3.6 years at GH start, 24 children were already pubertal (breast > 1, testes > 8 ml) at or within 1 year after GH start and 1 child was lost from follow-up. Moreover, 6 children who were born at a gestational age < 32 weeks and 6 adopted children with missing information at birth were excluded from the present analyses. Seven further children (6 boys) for whom most IGF-I data were missing were also excluded from the present analysis.
The remaining 104 children (90 boys) constituted the ITT population of whom 95 (82 boys) comprised the per-protocol (PP) population.
Of the 9 children (8 boys) not belonging to the PP population, 5 boys had stopped GH treatment prematurely (<2.25 years after study start and before AH was reached) and 4 children (3 boys) on GH67 decreased their GH dose to GH33x1
When the study group was divided into the groups to which they were later randomized, there were no differences in IGF levels between the groups. Moreover, there were no differences in IGF-I levels at baseline between boys and girls.
Serum IGF-I and IGFBP3 levels were measured using an IGFBP-blocked radioimmuno assay (RIA) with an excess of IGF-II for determination of IGF-I, and a specific RIA for IGFBP3 (Mediagnost GmbH, Tübingen, Germany). The intra-assay coefficients of variation (CVs) for the IGF-I assay were 11.1, 7.2 and 7.4 % at concentrations of 36, 204 and 545 μg/L, respectively; the interassay CVs for the same concentrations were 13.5, 8.8, and 9.9 %, respectively. For the IGFBP3 assay, the intra-assay CVs were 7.1, 7.3, and 7.9 % at concentrations of 1800, 3790 and 5776 μg/L, respectively; the interassay CVs for the same concentrations were 13.4, 10.5 and 14.1 %, respectively. Results were converted into SDS according to age, sex and pubertal stage, and the ratioSDS of IGF-I to IGFBP3 was calculated [12, 22]. IGF-I and IGFBP3 were analyzed before and after the initiation of GH treatment (+10 days, +1 month, +3 months, +1 year and annually thereafter). For the purpose of the present analyses, mean pubertal IGF-ISDS, IGFBP3SDS and IGF-I/IGFBP3 ratioSDS were calculated based on individual mean levels in the time period from 12 months after study start to treatment stop. The change in level was defined as the mean level (as calculated above) minus the level at study start. The prepubertal mean IGF-ISDS, IGFBP3SDS, and IGF-I/IGFBP3 ratioSDS were based on data collected in the time period from 1 year after GH start until the clinical onset of puberty. All samples from each individual were analyzed at the same time.
GHmax was assessed using both the AITT and a spontaneous 24-h GH secretion profile [20, 21]. GH concentrations were analysed with polyclonal antibodies, and for comparison, all measurements were converted to the WHO standard international reference preparation 80/505, even if measured with 66/127 .
Prepubertal gain in height SDS was defined as heightSDS at last pre-pubertal visit minus heightSDS at GH start using the childhood component of the total reference .
Total gain in height SDS was calculated using AHSDS (adult height in cm transferred into SDS for age 18 years) minus heightSDS at GH start, using the prepubertal childhood component of the total growth reference [23, 24].
There were no differences in gain in heightSDS between the two high-dose groups with GH67 given once or divided into twice daily GH33 (Table 1); therefore, the results from the high-dose groups are presented combined (GH67+33x2).
Normal or delayed infancy–childhood transition (ICT)
Familial short stature
Midparental height (MPH) was below −2 SDS  in 23 children (21 boys).
Small for gestational age
Twenty children (18 boys) were born small for gestational age with a birth lengthSDS and/or birth weightSDS < −2 SDS .
Statistical analyses were performed using the standard package SPSS version 20. Results are expressed as mean ± SD unless otherwise specified. Analyses concerning primary and secondary outcome variables were performed using non-parametric tests of the Wilcoxon type (Wilcoxon signed-rank test for within-group and Mann–Whitney U test for between-group comparisons). Safety analyses included all 149 children who received study drug. Analyses were performed for the ITT population and for boys in the PP population. Statistical significance was considered if p < 0.05.
Simple bivariate correlation analyses were performed using Pearson’s r.
Stepwise multiple regression analysis was used to analyse the influence of IGF-I variables on height gain and AHSDS. Data on birth characteristics and growth until 3 years of age, as well as baseline characteristics and prepubertal growth, were added as predictors in the analysis. Only variables entering the regressions below the significance level p < 0.05 were used. No correction was performed for multiplicity. Stepwise forward regression analyses were performed with p < 0.05 as entering criterion for predictors and p < 0.10 for exclusion after inclusion in an earlier step.
The characteristics of the study group have recently been reported including 7 more children  (see Table 1, Additional file 1: Table S1 for characteristics according to later randomization groups). Girls were younger than boys at GH start, 7.4 vs 10.2 years (p = 0.002) and also at study start, 12.3 vs 13.7 years (p = 0.001). Girls gained more heightSDS during prepuberty than boys, 2.2 vs 1.1 SDS, respectively (p = 0.0069) and less during puberty 0.2 vs 0.7, respectively (p = 0.029) (Table 1; Additional file 1: Table S2).
Comparison between children with DICT and normal ICT
Patients with DICT had significantly lower IGF-ISDS at GH start than patients with a normal ICT, −1.9 ± 2.0 vs −0.9 ± 0.9 (p = 0.014). See Additional file 1: Table S2 for baseline and study characteristics according to gender and ICT.
Per-protocol population (PP)
Comparing ITT and PP populations, similar results were found for the boys in both groups (data not shown); therefore, only the ITT population will be presented (Table 1).
Prepubertal study results
At GH start, mean IGF-ISDS was −1.2 for the total study group, mean IGFBP3SDS was −0.3 and mean IGF-I/IGFBP3 ratioSDS was–1.0, and values did not differ between the three groups who later constituted the randomization groups (Additional file 1: Table S2).
The observed prepubertal mean change (∆) in IGF-ISDS from GH start to the first year on GH therapy was 2.1 (Table 2). The corresponding ∆IGFBP3SDS was 0.6 (Table 3) and the ∆IGF/IGFBP-3 ratioSDS was 1.5 (Table 4). The change in IGF-ISDS relative to prepubertal mean level was 2.21, range −3.89 to 6.55.
The relationship between prepubertal IGF variables and gain in heightSDS for all children on GH33 (Tables 5, 6 and 7)
The group means for the three IGF variables at GH start were negatively correlated with prepubertal gain in heightSDS (IGF-ISDS r = −0.53, p = 0.001; IGFBP-3SDS r = −0.43, p = 0.001; IGF-I/IGFBP3 ratioSDS r = −0.31, p = 0.009) and with total gain in heightSDS (IGF-ISDS r = −0.37, p = 0.002; IGFBP3SDS r = −0.36, p = 0.002; IGF-I/IGFBP-3 ratioSDS r = −0.37, p = 0.001).
The first-year prepubertal ∆IGF-ISDS was positively correlated with the prepubertal gain in heightSDS (Fig. 1), but did not correlate with the ∆IGFBP3SDS or ∆IGF-I/IGFBP3 ratioSDS after 1 year. All three prepubertal first-year ∆IGF variables were positively correlated with the total gain in heightSDS (∆IGF-ISDS r = 0.42, p = 0.001; ∆IGFBP3SDS r = 0.29, p = 0.022 and ∆IGF-I/IGFBP3 ratioSDS r = 0.42, p = 0.001).
Pubertal study results
At study start, the group mean IGF-ISDS was 0.9, the group mean IGFBP3SDS was 0.4 and the group mean IGF-I/IGFBP3 ratioSDS was 0.5. There were no differences between the three randomization groups.
The highest IGF-ISDS was reached after 1 year in the study and the value subsequently decreased, i.e. the pubertal mean level was lower than the value at 1 year after randomization (see Table 2).
The pubertal IGF-ISDS for the total study group ranged from −2.1 to 3.5, with mean value being higher for the GH67group than for the GH33group, 1.1 vs 0.9, respectively (p = 0.031). The mean IGFBP3SDS for the total study group was 0.3, with a range −1.2 to 1.3. For IGF-I/IGFBP3 ratioSDS, the mean was 0.9, with a range of −1.3 to 2.4. There were no significant differences between the randomization groups for IGFBP3SDS or IGF-I/IGFBP3 ratioSDS.
Dose-dependent change (∆) in IGF-ISDS (Table 2), IGFBP3SDS (Table 3) and IGF-I/IGFBP3 ratioSDS (Table 4)
After the first year in the study, pubertal ∆IGF-ISDS was significantly greater in the group randomized to GH67 than in the group still receiving the GH33 dose, 0.5 vs −0.1, respectively (p = 0.007; Table 2).
A similar pattern was evident for the first year ∆IGF-I/IGFBP3 ratioSDS for GH67 vs GH33, 0.6 vs 0, respectively (p = 0.015; Table 4), but there was no change in IGFBP3SDS after 1 year for any group (Table 3).
When instead calculating the ∆SDS from study start to the mean pubertal level, again ∆IGF-ISDS was significantly greater for the GH67 vs GH33 group, 0.2 vs −0.2, respectively (p = 0.028), as shown in Fig. 2. When comparing the mean prepubertal and pubertal level of IGF-ISDS, IGFBP3SDS and IGF-I/IGFBP3 ratioSDS actually 47, 65 and 34 % respectively of children, equally for both dose groups, did not maintain the prepubertal level during puberty.
The relationship between IGF-ISDS, IGFBP3SDS, IGF-I/IGFBP3 ratioSDS and gain in heightSDS, (Tables 5, 6 and 7)
At randomization, the mean IGF-ISDS and IGFBP3SDS did not correlate with any of the height gain outcomes, whereas the mean IGF-I/IGFBP3 ratioSDS was negatively associated with total height gainSDS (r = −0.19, p = 0.048; Tables 5, 6 and 7).
For the individual, the pubertal ∆IGFSDS from randomization was significantly correlated with the pubertal gain in heightSDS for both GH33 and GH67, r = 0.32 (p = 0.003) and r = 0.24 (p = 0.026) respectively (Fig. 3a). For the total study group, the IGF-ISDS pubertal mean level correlated with the pubertal gain in heightSDS, r = 0.17 (p = 0.034; Fig. 3b). The range in gain in heightSDS was wide, and there were no significant differences between the dose groups; therefore, the results from the total study group were used for further analyses.
Disregarding randomization dose, the pubertal gain in heightSDS was positively correlated with the 1st year pubertal ∆IGF-ISDS (r = 0.26, p = 0.010) and also with the ∆IGF-ISDS to the pubertal mean level (r = 0.35, p = 0.001, Table 5).
The first year pubertal ∆IGFBP3SDS correlated with the pubertal gain in heightSDS (r = 0.20, p = 0.044), and the mean pubertal ∆IGFBP3SDS correlated with the pubertal gain in height level (r = 0.30, p = 0.002; Table 6). The attained mean pubertal level of IGFBP3SDS was not correlated with height gain outcomes. The pubertal ∆IGF-I/IGFBP3 ratioSDS was correlated (r = 0.22, p = 0.022) with the pubertal gain in heightSDS. The attained mean pubertal level of IGF-I/IGFBP3 ratioSDS correlated with the pubertal gain in heightSDS (r = 0.23, p = 0.021) and with total gain in heightSDS (r = −0.19, p = 0.048; Table 7).
Variance in pubertal ∆IGFs
In total, 39 % of the variation in pubertal ∆IGF-ISDS was explained by the variable ‘IGF-ISDS at study start’; the lower the value at study start, the greater the increase during puberty. Similarly, for pubertal ∆IGFBP3SDS, 28 % (also including age, +) was explained and for ∆IGF-I/IGFBP3 ratioSDS 40 % was explained.
Variance in gain in heightSDS with only IGF variables available (Table 8)
With only the IGF variables available, 26 % of the variation in pubertal gain in heightSDS could be explained by ‘pubertal (from study start to pubertal level) ∆IGF-ISDS’ and ‘first year prepubertal ∆IGF-I/IGFBP3 ratioSDS’. For total gain in heightSDS, 28 % of the variation was explained by ‘prepubertal IGF-ISDS first year after GH start’ and ‘pubertal ∆IGF-ISDS’. Regarding the variation in prepubertal gain in heightSDS, 21 % was explained by ‘IGF-ISDS at GH start’ (the lower the better).
Variance in gain in heightSDS with all variables available (Table 9)
The variables are presented in the order they entered in the analyses. For the pubertal gain in heightSDS, 46 % of the variation could be explained by ‘bone age at study start’ (−), ‘prepubertal years on GH’ (−), ‘∆IGF-ISDS 1 year after GH start’ (+) and ‘pubertal ∆IGF-ISDS (pubertal level from study start).
For the variation in total gain in heightSDS. 63 % could be explained by the variables ‘prepubertal gain in heightSDS’ (+), ‘bone age at GH start’ (−),’ ∆IGF/IGFBP3 ratioSDS 1 year after GH start’ (+) and ‘high GH dose’ (+).
In the analysis of prepubertal gain in heightSDS, 94 % of the variation was explained by ‘number of prepubertal years on GH’, ‘first year ∆heightSDS’, ‘first year ∆IGF-I/IGFBP3 ratioSDS’, ‘gender’ (girl +), ‘bone age at GH start’ (the greater the delay the better) and ‘first year ∆IGFBP3SDS’.
Changes in IGFs relate to prepubertal and pubertal height gain
The present analysis reported results for multiple variables including IGF-I, IGFBP3 and their ratio in IIGHD children from the start of GH treatment until AH, and the relationship of IGFs to both the prepubertal period when all participants received the same 33 μg/kg/d and the pubertal period when they were randomised to 33 or 67 μg/kg/d. The analysis focused on GH responsiveness as estimated by changes and obtained prepubertal and pubertal levels of IGF-ISDS, IGFBP3SDS and IGF-I/IGFBP3 ratioSDS. The main findings were: a significant dose-dependent (33 or 67 μg/kg/d) change in IGF-ISDS from randomization at onset of puberty to mean pubertal study level, and a positive correlation between this pubertal ∆IGF-ISDS and the pubertal gain in heightSDS. The greatest change was found in IGF-I followed by a less pronounced change in IGF-I/IGFBP3 ratio, while IGFBP3 values remained more stable. The GH dose given during puberty did not maintain the mean prepubertal IGF-I level, suggesting that some children may have benefitted from a higher dose in order to undergo a pubertal growth spurt of normal magnitude . The need for wide-ranging individual GH dosing in order to promote growth has previously been demonstrated during prepuberty , here we find similar requirements when studying IGFs and growth during the pubertal period. Actually, GH responsiveness in the present study group was so broad that the low dose was too high for some individuals while the high dose was too low for others in order to attain IGFs and pubertal height gain within normal range.
Prepubertal responsiveness to GH, reflected by a significant increase in IGFs after GH start [5, 28], has previously been reported to be of great importance for the short-term growth response; in the present analysis it was found to be important also for the total gain in height. This highlights the importance of individual GH dosing from treatment start, with the available prediction models presently being the best tools with which to estimate outcomes [29, 30]. Thus, when an increased GH dose during puberty induced a greater ∆IGF-ISDS, this resulted in a greater pubertal height gain than observed with a standard GH dose for most subjects. This finding has previously been reported in non-GHD subjects, with ∆IGF-ISDS found to be the most reliable variable correlating with pubertal gain in heightSDS . Thus, there is no principal difference in the IGF-I response and growth response between these two aetiologies of short stature, although the magnitude of GH responsiveness is higher in the GH-deficient than in the non-GH-deficient group [5, 31].
GH has effects on longitudinal bone growth both directly at the growth plate and locally mediated through IGF-I [32, 33]. The increase in IGF-I level as a response to GH treatment could be seen as a sign of GH responsiveness . GH/IGF-I responsiveness varies not only between individuals but also between tissues within an individual, e.g. more GH is needed to produce an effect on IGF-I production than for longitudinal bone growth .
Pubertal response in IGFs
In the present analyses, there was a significant difference in pubertal IGF-ISDS between randomization groups, with higher mean values in the GH67 than the GH33 group. This finding supports results from the study of Mauras et al.  who reported that by using higher GH doses during puberty (100 vs 42 μg/kg/d), a significantly higher pubertal gain in heightSDS was found, as well as a higher but non-significant IGF-I response in the high- relative to the low-dose group. The difference in importance of IGF-I in relation to our study may be due to differences in IGF-I references: we used our in-house reference with SDS considering gender, age and pubertal stage . In the study by Sas et al. using GH doses corresponding to 25 or 50 μg/kg/d until AH, the mean increase in IGF-ISDS after 1 year on GH was twice as high for the high- versus the low-dose group, although the difference did not reach significance. The relationship between IGF-ISDS and growth response was not given . Neither IGFBP3 nor IGF-I/IGFBP3 ratio was measured in these two published trials.
The multivariate analyses
In agreement with our findings in non-GH-deficient children , among the IGF-variables also in the present study, the pubertal ∆IGF-ISDS was found to be the most informative variable, and more important than the level per se, for explaining the variance in both pubertal and total gain in heightSDS . When all variables were allowed, bone age delay and less prepubertal years on GH in addition to ∆IGF-ISDS was positive for pubertal gain in height SDS . This may be explained by some remaining catch-up growth occurring during puberty in some study subjects, even though all children had been treated with GH for at least 1 year before randomization.
The explanatory variables for pubertal gain in height were in accordance with those identified within the KIGS observational study, except that bone age was selected in our study and chronological age in the KIGS study . This difference may be because bone age was estimated by a single radiologist in our trial which increased the quality and consistency of this variable.
For total gain in height in the present study, a GH dose high enough to result in a substantial prepubertal ∆IGF-I/IGFBP-3 ratioSDS was favourable. The high dose was set at 67 μg/kg/d even though the present results suggest that nearly 50 % of the children could have benefitted from an even higher dose. We need to remember that at the time of study design, there were limited data on GH treatment doses and safety, and no tools available for estimation of individual GH responsiveness such as the prediction models for GH growth response [30, 35].
Mean IGF-ISDS, IGFBP3SDS and IGF-I/IGFBP3 ratioSDS at GH start were in the low/normal range. Lower IGF levels at baseline were associated with greater changes after 1 year of GH therapy. We found a strong negative bivariate linear correlation between low baseline IGF-ISDS, IGFBP3SDS, their ratioSDS and both prepubertal and total gain in heightSDS. The prepubertal IGF-I response and its relationship to gain in height is already well known [3–5, 36]. It was previously reported that the observed prepubertal 1-year growth response makes it possible to estimate the full prepubertal gain in height in both children with GHD and in non-GH-deficient groups . Knowledge of the relationship between the change in IGFs and total gain in heightSDS can now be added, and highlights the importance of considering prepubertal GH responsiveness and achieving a greater first year ∆IGFs when selecting GH dose. The ∆IGF-ISDS, ∆IGFBP3SDS and ∆IGF/IGFBP3 ratioSDS 1 year after GH start mirrored individual GH responsiveness as measured by IGF-I generation. This supports results from 1-year non-randomized observations in GH-deficient and non-GH-deficient groups [8, 9, 36].
In our analyses, prepubertal IGF-ISDS did not correlate with gain in height. This is in contrast to the findings of the study by Cohen et al. who found a relationship between IGF-I level (the higher the better) and height gain in prepubertal children treated for 2 years with IGF-I-targeting GH doses . The difference in results may be explained by differences in study design, mainly in terms of inclusion criteria and dose ranges. However, in both studies the ∆IGF-ISDS was found in multivariate analyses to be the most informative variable explaining growth response.
The multivariate analyses
In multivariate regression analyses using only IGF variables, baseline IGF-ISDS alone explained 21 % of the variation in prepubertal gain in height. When adding auxological variables, the variation explained improved to 94 % with the following important variables: bone age delay at GH start (more delay better growth) and prepubertal years on GH (more years, more growth). Many prepubertal years on GH are also a sign of GHD (being young at diagnosis was associated with more severe GHD).
Different information from IGFBP3SDS and IGF/IGFBP3 ratioSDS
IGFBP3 and IGF-I/IGFBP3 ratio are not routinely monitored during GH treatment. IGFBP3 is less sensitive to short-term nutritional variations and diseases than IGF-I and could therefore be valuable when monitoring efficacy of and compliance with GH treatment . In addition, a more pronounced change in IGF-I relative to IGFBP3 results in an increased IGF-I/IGFBP3 ratio [22, 39], which can be seen as an indicator of increased IGF-I bioavailability . In the present study, changes in IGF-I/IGFBP3 ratioSDS followed changes in IGF-I, and correlated significantly with both pubertal and total gain in heightSDS. In the multivariate regression analysis, prepubertal ∆IGF-I/IGFBP3 ratioSDS was selected for explanation of variance in both prepubertal and total gain in heightSDS. ∆IGF-I/IGFBP3 ratioSDS may be a result of a synergistic effect of IGF-I and IGFBP3 during GH therapy , where IGFBP3 modulates the actions of IGF-I, as well as having an independent effect . To our knowledge, longitudinal observation of IGF-I/IGFBP3 ratioSDS during pubertal growth in a population with IIGHD has not previously been reported.
IGFs markers for safety
In the present study no dose-dependent adverse event involving carbohydrate metabolism was observed, which confirms previous studies with high GH doses [15, 16, 42, 43]. However, a modest association between increased circulating level of IGF-I and an increased risk of common cancers in adult has been reported [1, 44]. Most circulating IGF-I is bound to IGFBP3 and to the acid labile subunit (ALS) , therefore, the IGF-I/IGFBP3 ratioSDS could reflect the tissue availability of IGF-I and its correlation with free IGF-I [39, 40, 46]. Our data showed no dose-dependent differences in pubertal IGF-I/IGFBP3 ratioSDS, only a broad range of IGF-ISDS and IGFBP3SDS, highlighting that GH dosing needs to be individualized .
In the present study, stable IGF-ISDS and IGFBP3SDS concentrations were observed, which is a sign of good compliance [38, 47]. There were more than 10 samples for each patient and nearly 2000 samples in the total analysis. Only 8 single samples were excluded due to suspected poor compliance.
Poor compliance could be a factor that underestimates the study results; both in terms of IGF-I and growth responses. Pubertal teenagers are known to have the lowest treatment compliance due to many factors, not least psychological [48, 49]. In order to promote good compliance in our study, all participants were responsible for their own injections and were followed every third month. At each visit they were invited to discuss their treatment with their endocrine team .
The heterogeneity of the study group
The study group was heterogeneous, including patients with classic GHD, partial GHD and some short boys with low GH secretion in the late prepubertal period , and there was a broad range of IGF-ISDS and IGFBP3SDS at baseline. This reflects the reality in daily clinical practice and allows results to represent patients with a wider range of baseline levels of GH secretion and a broad range in GH responsiveness.
Limitations of the study
The study was designed before individual GH responsiveness was broadly considered. Children were randomized to weight-based dosing and 38 % of patients decided to reduce or stop GH treatment before AH was reached due to satisfaction with their attained height, which was accepted by the protocol. However, the combination of the limited number of patients in the treatment groups, the broad variability in their growth responses, and the premature stop/reduction of GH dose will underestimate the result of the study regarding changes in IGF-ISDS, IGFBP-3SDS and IGF-I/IGFBP-3 ratioSDS, as well as the dose-dependent effect on pubertal gain in heightSDS.
The onset of puberty was defined by clinical signs in the present trial, making it possible that some pubertal growth had occurred before randomization, thus leading to underestimation of the pubertal IGF-I response and gain in height.
Girls constituted a small group in the current analyses. The low number of girls in each dose group does not allow conclusions about use of an even higher GH dose during puberty in girls.
In the present analysis we studied GH responsiveness estimated on the obtained levels and change of IGF-ISDS, IGFBP3SDS and IGF-I/IGFBP-3 ratioSDS and the associated growth response in children with IIGHD randomized in puberty to different weight-based GH dose regimens. Thereby, the relationship of IGFs to gain in heightSDS during both the prepubertal and pubertal growth phases could be explored. The prepubertal increase in IGFs associated with a 33 μg/kg/d GH dose and the GH dose-dependent (33 or 67 μg/kg/d) pubertal increase in IGF-I, were both important variables that explained the total gain in height: the higher the prepubertal GH responsiveness, the greater the total gain in heightSDS. Our hypothesis that increased GH dose during puberty would result in a more pronounced IGF-I response and greater growth was found to be valid: the higher GH dose during puberty was followed by both higher IGF-ISDS and a greater gain in heightSDS than observed in patients receiving the lower GH dose. Thus, of great importance in the clinical setting: the individual who remains short at onset of puberty require a GH dose increase great enough to result in an increment in IGF-I in order to gain any heightSDS during puberty.
Arginine–insulin tolerance test
Body mass index
Delayed infancy–childhood transition, ICT >12 months after birth
The difference between the child’s HeightSDS vs MPHSDS
Growth hormone deficiency
GH, Standard dose 33 μg/kg once daily injection
GH, 33 μg/kg twice daily injection
GH, 67 μg/kg once daily injection
GH 33×2 and 67×1, the high-dose group
Maximum GH level during a spontaneous 24h GH profile
Maximum GH level during an AITT
Infancy–childhood transition, months
Insulin-like growth factor I
Idiopathic isolated growth hormone deficiency
Insulin-like growth factor-binding protein 3
International reference preparation
Standard deviation score
World Health Organization
Renehan AG, Zwahlen M, Minder C, O'Dwyer ST, Shalet SM, Egger M. Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet. 2004;363:1346–53.
Cutfield WS, Derraik JG, Gunn AJ, Reid K, Delany T, Robinson E, et al. Non-compliance with growth hormone treatment in children is common and impairs linear growth. PLoS One. 2011;6:e16223. doi:10.1371/journal.pone.0016223.
Rudman D, Moffitt SD, Fernhoff PM, McKenzie WJ, Kenny JM, Bain RP. The relation between growth velocity and serum somatomedin C concentration. J Clin Endocrinol Metab. 1981;52:622–7.
Albertsson-Wikland K, Hall K. Growth hormone treatment in short children: relationship between growth and serum insulin-like growth factor I and II levels*. J Clin Endocrinol Metab. 1987;65:671–8.
Kriström B, Jansson C, Rosberg S, Albertsson-Wikland K. Growth response to growth hormone (GH) treatment relates to serum insulin-like growth factor I (IGF-I) and IGF-binding protein-3 in short children with various GH secretion capacities. Swedish Study Group for Growth Hormone Treatment. J Clin Endocrinol Metab. 1997;82:2889–98.
Kamp GA, Wit JM. High-dose growth hormone therapy in idiopathic short stature. Horm Res. 1998;49:67–72.
Cohen P, Bright GM, Rogol AD, Kappelgaard AM, Rosenfeld RG, Group ANCT. Effects of dose and gender on the growth and growth factor response to GH in GH-deficient children: implications for efficacy and safety. J Clin Endocrinol Metab. 2002;87:90–8.
Tillmann V, Patel L, Gill MS, Whatmore AJ, Price DA, Kibirige MS, et al. Monitoring serum insulin-like growth factor-I (IGF-I), IGF binding protein-3 (IGFBP-3), IGF-I/IGFBP-3 molar ratio and leptin during growth hormone treatment for disordered growth. Clin Endocrinol (Oxf). 2000;53:329–36.
Scire G, Del Bianco C, Spadoni G, Cianfarani S. Growth hormone therapy does not alter the insulin-like growth factor-I/insulin-like growth factor binding protein-3 molar ratio in growth hormone-deficient children. J Endocrinol Invest. 2008;31:153–8.
Juul A, Bang P, Hertel NT, Main K, Dalgaard P, Jørgensen K, et al. Serum insulin-like growth factor-I in 1030 healthy children, adolescents, and adults: relation to age, sex, stage of puberty, testicular size, and body mass index. J Clin Endocrinol Metab. 1994;78:744–52.
Albertsson-Wikland K, Rosberg S, Karlberg J, Groth T. Analysis of 24-hour growth hormone profiles in healthy boys and girls of normal stature: relation to puberty. J Clin Endocrinol Metab. 1994;78:1195–201.
Löfqvist C, Andersson E, Gelander L, Rosberg S, Blum WF, Albertsson WK. Reference values for IGF-I throughout childhood and adolescence: a model that accounts simultaneously for the effect of gender, age, and puberty. J Clin Endocrinol Metab. 2001;86:5870–6.
Veldhuis JD, Metzger DL, Martha Jr PM, Mauras N, Kerrigan JR, Keenan B, et al. Estrogen and testosterone, but Not a nonaromatizable androgen, direct network integration of the hypothalamo-somatotrope (growth hormone)-insulin-like growth factor I axis in the human: evidence from pubertal pathophysiology and Sex-steroid hormone replacement 1. J Clin Endocrinol Metab. 1997;82:3414–20.
Ranke MB, Lindberg A, Martin DD, Bakker B, Wilton P, Albertsson-Wikland K, et al. The mathematical model for total pubertal growth in idiopathic growth hormone (GH) deficiency suggests a moderate role of GH dose. J Clin Endocrinol Metab. 2003;88:4748–53.
Mauras N, Attie KM, Reiter EO, Saenger P, Baptista J. High dose recombinant human growth hormone (GH) treatment of GH-deficient patients in puberty increases near-final height: a randomized, multicenter trial. Genentech, Inc., Cooperative Study Group. J Clin Endocrinol Metab. 2000;85:3653–60.
Sas TC, de Ridder MA, Wit JM, Rotteveel J, Oostdijk W, Reeser HM, et al. Adult height in children with growth hormone deficiency: a randomized, controlled, growth hormone dose–response trial. Horm Res Paediatr. 2010;74:172–81.
Kriström B, Lundberg E, Jonsson B. IGF-I and growth response to adult height in a randomized GH-treatment trial in short non-GH-deficient children. J Clin Endocrinol Metab. 2014;99:2917–24.
Albertsson-Wikland K, Alm F, Aronsson S, Gustafsson J, Hagenas L, Hager A, et al. Effect of growth hormone (GH) during puberty in GH-deficient children: preliminary results from an ongoing randomized trial with different dose regimens. Acta Paediatr Suppl. 1999;88:80–4.
Albertsson-Wikland K, Kriström B, Lundberg E, Aronson A, Gustafsson J, Hagenäs L, et al. Growth hormone dose-dependent pubertal growth: a randomized trial in short children with Low growth hormone secretion. Horm Res Paediatr. 2014;82:158–70.
Jansson C, Boguszewski C, Rosberg S, Carlsson L, Albertsson-Wikland K. Growth hormone (GH) assays: influence of standard preparations, GH isoforms, assay characteristics, and GH-binding protein. Clin Chem. 1997;43:950–6.
Albertsson-Wikland K, Rosberg S. Methods of Evaluating Spontaneous Growth Hormone Secretion. In: Ranke M, editor. Functional Endocrinologic Diagnostics in Children and Adolescents. Mannheim: J & J Verlag; 2010. p. 129–59.
Löfqvist C, Andersson E, Gelander L, Rosberg S, Hulthen L, Blum WF, et al. Reference values for insulin-like growth factor-binding protein-3 (IGFBP-3) and the ratio of insulin-like growth factor-I to IGFBP-3 throughout childhood and adolescence. J Clin Endocrinol Metab. 2005;90:1420–7. doi:10.1210/jc.2004-0812.
Albertsson-Wikland K, Luo ZC, Niklasson A, Karlberg J. Swedish population-based longitudinal reference values from birth to 18 years of age for height, weight and head circumference. Acta Paediatr. 2002;91:739–54.
Karlberg J. On the construction of the infancy-childhood-puberty growth standard. Acta Paediatr Scand Suppl. 1989;356:26–37.
Hochberg Z, Albertsson-Wikland K. Evo-devo of infantile and childhood growth. Pediatr Res. 2008;64:2–7.
Karlberg P, Taranger J, Engstrom I, Karlberg J, Landstrom T, Lichtenstein H et al. I. physical growth from birth to 16 years and longitudinal outcome of the study during the same age period. Acta Paediatr Scand Suppl. 1976;258:7–76.
Niklasson A, Ericson A, Fryer JG, Karlberg J, Lawrence C, Karlberg P. An update of the Swedish reference standards for weight, length and head circumference at birth for given gestational age (1977–1981). Acta Paediatr Scand. 1991;80:756–62.
Cohen P, Rogol AD, Howard CP, Bright GM, Kappelgaard AM, Rosenfeld RG, et al. Insulin growth factor-based dosing of growth hormone therapy in children: a randomized, controlled study. J Clin Endocrinol Metab. 2007;92:2480–6.
Ranke MB, Lindberg A, Chatelain P, Wilton P, Cutfield W, Albertsson-Wikland K, et al. Derivation and validation of a mathematical model for predicting the response to exogenous recombinant human growth hormone (GH) in prepubertal children with idiopathic GH deficiency. J Clin Endocrinol Metab. 1999;84:1174–83.
Albertsson-Wikland K, Kristrom B, Rosberg S, Svensson B, Nierop AF. Validated multivariate models predicting the growth response to GH treatment in individual short children with a broad range in GH secretion capacities. Pediatr Res. 2000;48:475–84.
Kristrom B, Albertsson-Wikland K. Insulin and IGF-I levels during 2 yrs of treatment with individual GH doses (17–100 mu g/kg/d) do not differ to standard dose (43 mu g/kg/d) in prepubertal GHD and ISS children. Hormone Research. Allschwillestrasse 10, CH-4009 Basel, Switsseland: KARGER; 2009.
Isaksson O, Jansson J-O, Gause I. Growth hormone stimulates longitudinal bone growth directly. Science. 1982;216:1237–9.
Green H, Morikawa M, Mxon T. A dual effector theory of growth‐hormone action. Differentiation. 1985;29:195–8.
Decker R, Nygren A, Kristrom B, Nierop AF, Gustafsson J, Albertsson-Wikland K, et al. Different thresholds of tissue-specific dose-responses to growth hormone in short prepubertal children. BMC Endocr Disord. 2012;12:26.
Ranke MB, Lindberg A. Predicting growth in response to growth hormone treatment. Growth Horm IGF Res. 2009;19:1–11.
Ranke M, Schweizer R, Elmlinger M, Weber K, Binder G, Schwarze C, et al. Relevance of IGF-I, IGFBP-3, and IGFBP-2 measurements during GH treatment of GH-deficient and non-GH-deficient children and adolescents. Horm Res Paediatr. 2001;55:115–24.
Kriström B, Dahlgren J, Niklasson A, Nierop AF, Albertsson-Wikland K. The first-year growth response to growth hormone treatment predicts the long-term prepubertal growth response in children. BMC Med Inform Decis Mak. 2009;9:1.
Buckway CK, Selva KA, Pratt KL, Tjoeng E, Guevara-Aguirre J, Rosenfeld RG. Insulin-like growth factor binding protein-3 generation as a measure of GH sensitivity. J Clin Endocrinol Metab. 2002;87:4754–65.
Juul A, Main K, Blum WF, Lindholm J, Ranke MB, Skakkebaek NE. The ratio between serum levels of insulin-like growth factor (IGF)-I and the IGF binding proteins (IGFBP-1, 2 and 3) decreases with age in healthy adults and is increased in acromegalic patients. Clin Endocrinol (Oxf). 1994;41:85–93.
Chen J-W, Højlund K, Beck-Nielsen H, Sandahl Christiansen J, Ørskov H, Frystyk J. Free rather than total circulating insulin-like growth factor-I determines the feedback on growth hormone release in normal subjects. J Clin Endocrinol Metab. 2005;90:366–71.
Mandel SH, Moreland E, Rosenfeld RG, Gargosky SE. The effect of GH therapy on the immunoreactive forms and distribution of IGFBP-3, IGF-I, the acid-labile subunit, and growth rate in GH-deficient children. Endocrine. 1997;7:351–60.
Coelho R, Brook CG, Preece MA, Stanhope RG, Dattani MT, Hindmarsh PC. A randomised study of two doses of biosynthetic human growth hormone on final height of pubertal children with growth hormone deficiency. Horm Res. 2008;70:85–8.
Riddick L, Alter C, Davis DA, Frane J, Lippe B, Bakker B. A stepwise increase in recombinant human growth hormone dosing during puberty achieves improved pubertal growth: a National Cooperative Growth Study report. J Pediatr Endocrinol Metab. 2009;22:623–8.
Popovic V, Mattsson AF, Gaillard RC, Wilton P, Koltowska-Häggström M, Ranke MB. Serum insulin-like growth factor I (IGF-I), IGF-binding proteins 2 and 3, and the risk for development of malignancies in adults with growth hormone (GH) deficiency treated with GH: data from KIMS (Pfizer International Metabolic Database). J Clin Endocrinol Metab. 2010;95:4449–54.
Frystyk J, Skjærbæk C, Dinesen B, Ørskov H. Free insulin-like growth factors (IGF-I and IGF-II) in human serum. FEBS Lett. 1994;348:185–91.
Skjaerbaek C, Vahl N, Frystyk J, Hansen T, Jorgensen J, Hagen C, et al. Serum free insulin-like growth factor-I in growth hormone-deficient adults before and after growth hormone replacement. Eur J Endocrinol. 1997;137:132–7.
Gelander L, Blum WF, Larsson L, Rosberg S, Albertsson-Wikland K. Monthly measurements of insulin-like growth factor I (IGF-I) and IGF-binding protein-3 in healthy prepubertal children: characterization and relationship with growth: the 1-year growth study. Pediatr Res. 1999;45:377–83.
Haverkamp F, Johansson L, Dumas H, Langham S, Tauber M, Veimo D, et al. Observations of nonadherence to recombinant human growth hormone therapy in clinical practice. Clin Ther. 2008;30:307–16.
Ba C, Kj T. Noncompliance in adolescents: A review. J Dev Behav Pediatr. 1989;10:207–15.
Westphal O, Lindberg A, Board SKN. Final height in Swedish children with idiopathic growth hormone deficiency enrolled in KIGS treated optimally with growth hormone. Acta Paediatr. 2008;97:1698–706.
Cianfarani S. Is high-dose growth hormone treatment during puberty worthwhile? Horm Res Paediatr. 2014;82:143.
The authors are thankful for the participation of all the children and their families; for the great care taken of the study population over the years by our Swedish paediatric colleagues, for our excellent paediatric nurses, and the staff members at Gothenburg Paediatric Growth Research Center (GP-GRC) for their long-term commitment to the study, not least Carola Pfeiffer-Mosesson and the monitor Eva Mölleborn, especially Lisbeth Larsson and Irene Leonardsson. The authors thanks to all investigators in study group: A.Stefan Aronson, Department of Paediatrics, Halmstad hospital; Jan Gustafsson, Department of Women’s and Children’s Health, Uppsala University; Lars Hagenäs, Department of Women’s and Children’s Health, Karolinska Institute; Sten-A Ivarsson, Department of Paediatrics, Lund University; Björn Jonsson, Department of Women’s and Children’s Health, Uppsala University; Martin Ritzén, Department of Women’s and Children’s Health, Karolinska Institute; Torsten Tuvemo, Department of Women’s and Children’s health, Uppsala University; Ulf Westgren, Department of Paediatrics, Lund University; Otto Westphal, Gothenburg Pediatric Growth Research Center, Department of Peditarics, Insitute of Clinical Science, Sahlgrenska Aacademy at University of Gothenburg; Jan Åman School of Health and Medical Sciences, Örebro University.
The authors also thank Harriet Crofts for valuable language editing of the manuscript.
This investigator-initiated and sponsored trial, TRN 88–177 began in 1988. Kabi/Pharmacia supported the study by paying for monitoring of patients, the additional clinical costs and provided free drugs for patients receiving more than 33ug/kg/day. The study was also supported by grants from the Swedish Research Council (7509), the Swedish Foundation for Pediatric GH Research and the Foundation Växthuset for Children as well as University Hospital Government Grants (ALF) in Gothenburg and Umea. Kabi/Pharmacia/Pfizer has not had any influence on the collection, analysis and interpretation of these data or in the writing of this manuscript.
EL has no competing interest, has not received any commercial source of funding.
BK has received Umeå University Hospital Government Grants (ALF), lecture and/or consultation honoraria from Merck Serono, NovoNordisk, Pfizer and Sandoz. BJ was employed by Pilan Data Management AB until he died in February 2015. KA-W received an unrestricted research grant from Pharmacia/Pfizer until 2005. This investigator-initiated and sponsored trial, TRN 88–177 began in 1988. Kabi/Pharmacia supported the study by paying for monitoring of patients, the additional clinical costs and provided free drugs for patients receiving more than 33ug/kg/day. The study was also supported by grants from the Swedish Research Council (7509), the Swedish Foundation for Pediatric GH Research and the Foundation Växthuset for Children as well as University Hospital Government Grants (ALF) in Gothenburg and Umea. Kabi/Pharmacia/Pfizer has not had any influence on the collection, analysis and interpretation of these data or in the writing of this manuscript.
The study was initiated (concept and design) in the 1980s by KAW, the PI who also served as local investigator for Gothenburg. Later in the 1990s, BK also became a local investigator for Umea. KAW, BK, EL made the clean file. BJ as biostatistician made all the statistical analyses. EL made the first draft. All authors made substantial intellectual contribution with analyses, interpretation of data and revising the manuscript for important intellectual content. All authors have given approval for the final version to be published.
Table S1. Baseline characteristics according to randomization group: at birth, at GH start and at study start. Table S2. Baseline characteristics according to gender and infancy–childhood transition (ICT): at birth, at GH start, at study start and at adult height. (PDF 222 kb)
About this article
Cite this article
Lundberg, E., Kriström, B., Jonsson, B. et al. Growth hormone (GH) dose-dependent IGF-I response relates to pubertal height gain. BMC Endocr Disord 15, 84 (2015). https://doi.org/10.1186/s12902-015-0080-8
- Gain in height
- IGF-I increment
- IGF-I level
- Ratio IGF-I/IGFBP3
- GH dose-dependent pubertal IGF-I response