Adult height and health-related quality of life after growth hormone therapy in small for gestational age subjects

Abstract

Objective: To estimate health-related quality of life (HRQoL) in non-growth hormone deficient (GHD) small for gestational age (SGA) children before and after growth hormone (GH) treatment to adult height (AH).

Methods: This was a multicentre, two-arm trial. Following an initial 2-year double-blind study period, patients entered a 2-year extension period followed by treatment to AH. At baseline patients were randomised to GH (0.033 or 0.067 mg/kg/day) and continued treatment at that dose until AH. Height was assessed at baseline and 3-monthly intervals to AH (height velocity <2 cm/year). Height standard deviation score (SDS) before and after GH therapy was mapped onto estimated HRQoL scores up to AH.

Results: Of the 79 children randomised into the study 53 were non-GHD (defined as peak GH >20 mU/L [peak 24-h GH value and peak arginine tolerance test]). At baseline these children had a mean (mean [±SD]) height SDS of −3.2 (0.7), height velocity SDS −0.6 (1.2) and age, 8.1 (1.9) years. Estimated HRQoL scores were significantly (p < 0.001) increased from baseline at AH (ΔHRQoL, 95% CI) (0.033 mg/kg/day, 0.112 [0.092, 0.132]; 0.067 mg/kg/day, 0.115 [0.094, 0.136]). HRQoL was not different between treatment groups. A significant gain in AH, relative to an SGA reference population, was reported in GH-treated patients. Mean (95% CI) ΔAH SDS (0.033 mg/kg/day, +1.4 [1.1, 1.6]. 0.067 mg/kg/day, +1.7[1.4, 2.0]).

Limitations: The analysis assumes HRQoL can be mapped onto height SDS.

Conclusions: GH treatment in short children born SGA without signs of persistent catch-up growth was associated with significant improvement in HRQoL and normalisation of AH.

• adult height
• growth hormone
• health-related quality of life
• HRQoL
• SGA
• short stature
• small for gestational age

Introduction

It is commonly assumed that short stature is associated with psychological and social problems as well as physical difficulties[1]. In childhood, the physical limitations associated with short stature as well as the younger appearance often associated with it may lead to short-statured children being subject to juvenilisation and to exhibit clinically significant behavioural or emotional problems[2],[3]. Adult short stature may confer physical limitations with respect to the ability to drive a normal car, or play sports, as well as difficulties, or even discrimination in the job market[4],[5]. Thus, it is expected that short stature may negatively impact health-related quality of life (HRQoL). The concept of HRQoL evaluates physical, psychological and social well-being. In this way a person's HRQoL is defined by their perception of problems of health status in combination with the affective response to such problems[6].

Small for gestational age (SGA) is defined as a birth weight and/or birth length that is less than −2 standard deviation scores (SDS) from the mean[7]. SGA is associated with an increased risk of not reaching normal adult height (AH)[8]. The condition also carries an elevated risk of developing cardiovascular complications, obstructive pulmonary disease, type 2 diabetes mellitus, renal insufficiency and impaired reproductive functions[8–13].

Approximately 5% of all newborns are born SGA[8]. Of those born SGA, most achieve appropriate catch-up growth by 2 years of age, but about 15% do not catch up to a height above −2 SDS from the mean[8],[14]. With effective growth hormone (GH) treatment started early in childhood SGA children without catch-up growth can achieve normal height during childhood and adolescence[15], and 85% will achieve an AH that is within the normal height range for the population[7],[16]. Some SGA children also suffer from GH deficiency. Growth hormone deficient (GHD) children normally respond better to GH treatment than SGA children. Inclusion of children with both SGA and GH deficiency in the present study were considered inappropriate as this might overestimate the treatment effects for a ‘clean’ SGA population.

There is currently a lack of HRQoL data available for SGA children treated with GH, especially the type of HRQoL data which can be used as input in health economic models to guide decisions on reimbursement. This study attempts to make the needed HRQoL available by mapping pre-existing HRQoL relationship onto a pre-existing 15-year-long randomised clinical trial. From a methodological point of view this is a second best solution. However, it is problematic to await the results from a new 15-year-long clinical study before informed reimbursement decisions can be made.

The aim of this study was to estimate HRQoL in non-GHD SGA children before and after GH (Norditropin) treatment to AH.

Methods

GH-treated SGA subjects

This was a multicentre (four sites in the Netherlands), double-blind, randomised, two-arm trial comparing the effects of two dose regimens (0.033 and 0.067 µg/kg/day) of GH (Novo Nordisk A/S, Copenhagen, Denmark) to AH in children born SGA. The study comprised a 2-year initial trial period with a 2-year trial extension followed by treatment to AH. A total of 79 short children were enrolled in the trial. Data on these children have been previously reported[16],[17].

The main inclusion criteria for participation in the current study were SGA diagnosis (see definition below), chronological age (CA) between 3 and 11 years in boys and 3 and 9 years in girls at start of treatment, height (cm) SDS for CA < −2, height velocity (cm/year) SDS for CA ≤ 0, prepubertal (Tanner stage 1 for girls and testicular volume < 4 mL for boys), uncomplicated neonatal period and no spontaneous catch-up growth. Main exclusion criteria were presence of an endocrine/metabolic disorder, chromosomal abnormalities, growth failure caused by other disorders or syndromes and present or previous use of drugs that could interfere with GH treatment.

After completion of the original study a consensus definition of SGA was established by the Joint Committees of the European Society of Pediatric Endocrinologists (ESPE), Lawson Wilkins Pediatric Endocrinology Society (LWEPS) and Growth Hormone Research Society (GRS)[7]. This definition (birth weight and/or length ≤2 SDS, and height SDS at inclusion ≤2 [height SDS was calculated in accordance with Dutch reference population standards]) was applied to the data.

Furthermore, in the study, initiated in 1990, eligible patients for inclusion were short SGA patients with/without GH deficiency or who are partially GHD[16]. However, in order not to confuse the evaluation of GH treatment in SGA patients, with data from GHD children, the current report only includes data collected from 53 non-GHD SGA children.

As a result of the exclusion of GHD patients it can be expected that there will be a discrepancy in the AH reported in the present study compared with that reported in previous publications.

GH status was defined based on a 24-h GH profile and/or arginine tolerance test (ATT). On the basis of these tests, GH deficiency was defined as a GH value below 20 mU/L (peak 24-h GH value and peak ATT value). In the event that only one test was performed with a peak value <20 mU/L, or if no test result was available, the child was not included in the subgroup of non-GHD children as abnormal GH secretion could not be excluded.

After stratification for CA, the children were randomly and blindly assigned to one of two GH (0.033 or 0.067 µg/kg/day) dose groups. The trial inclusion period was between April 1991 and January 1993. Biosynthetic GH (Novo Nordisk A/S, Copenhagen, Denmark) was given subcutaneously once daily at bedtime with a pen injection system. Every 3 months the total GH dose was adjusted to the calculated body surface. To ensure the double-blind design, an equal volume of reconstituted preparation was used.

Trial objectives were to assess and compare the efficacy and safety of two doses of GH during long-term treatment. The rationale was to investigate optimal GH doses to improve height for short children born SGA and to assess whether the response to GH was sustained during long-term treatment. The efficacy of long-term GH treatment was evaluated by its effect on AH as assessed by AH (cm), AH SDS, and delta AH SDS. Height was measured at each clinic visit using a Harpenden stadiometer (four measurements were taken and the mean was used for analysis). Height was expressed as the SDS for CA (height SDS). AH was defined as height velocity <0.5 cm/year during the previous 6 months with bone age ≥15 years for girls and ≥16.5 years for boys. GH treatment was discontinued at AH or on the patient's decision on achieving near AH (height velocity 0.5–2 cm/year during the previous 6 months).

The trial was initiated in 1990. The study protocol was approved by the relevant ethics committees and written informed consent was obtained from the parents/guardians of each child.

Design and evaluation of HRQoL before and after GH treatment

HRQoL of the SGA population was evaluated by mapping height SDS scores, before and after GH therapy, onto the HRQoL scores shown in Table 1. The mapping table is based on HRQoL data, subdivided by height SDS scores, derived from a nationally representative population, the 2003 Health Survey for England (HSE 2003) (a nationally-representative survey of the general adult population in England (n = 14.057)[4]. The HRQoL was assessed using the EQ-5D, a standardised instrument used as a recognised measure of health outcome[18],[19]. The EQ-5D descriptive system comprises five dimensions of health (mobility, self-care, usual activities, pain/discomfort, anxiety/depression). A unique HRQoL health status evaluation (EQ-5D index) is identified with perfect health scoring 1 and death equal to 0. HRQoL results obtained with the EQ-5D questionnaire are also defined as ‘utilities’ which can be used in health economic cost-utility models. However, for consistency the term HRQoL will be used throughout the manuscript. Methodology associated with use of the questionnaire and HRQoL results from the HSE 2003 is documented elsewhere[4].

Table 1.  Mapping height standard deviation scores (height SDS) into EQ-5D.

Height SDS	Mean EQ-5D score
Below <−3.00	0.6866
−3.00 to <−2.50	0.7440
−2.50 to <−2.00	0.8009
−2.00 to <−1.50	0.8194
−1.50 to <−1.00	0.8314
−1.00 to <−0.50	0.8505
−0.50 to <0.00	0.8638
0.00 to <0.50	0.8808
0.50 to <1.00	0.8874
1.00 to <1.50	0.8891
1.50 to <2.00	0.8986
2.00 to <2.50	0.9010
2.50 and above	0.9259

The following assumptions were made: the same mapping table can be used at baseline and at AH; if short children born SGA are not treated with GH they will gain 0.3 height SDS until AH. This last assumption is based on the untreated control group followed in a study by Van Pareren (2003)[16].

Statistical analysis

All efficacy analyses were based on the intention-to-treat (ITT) analysis set. Efficacy outcomes were analysed by ANOVA. Normality of data was tested by plotting the participants’ residuals against their corresponding expected normal order. Homogeneity was checked by visual inspection of the variability within treatment groups using box-plots.

For endpoints related to AH, the statistical model accounted for the following likely confounding factors: GH treatment group (0.033 and 0.067 mg/kg/day); gender; treatment duration (years); age (years); bone age (years); target height (cm) and height SDS at start of treatment; peak stimulated GH level; insulin-like growth factor-1 at baseline and centre as random effects.

Results

A total of 53 children classified as SGA (born short and/or light for gestational age) and non-GHD were included in the present analysis. At baseline, children were very short (mean [SD] height SDS, −3.2 [0.7]) compared with their peers and pre-trial height velocity SDS (mean [SD] −0.6 [1.2]) was also below the mean growth rate for children with normal growth. Mean bone age (TW-II RUS blinded) was retarded by 0.5 years and 0 years (not retarded) in the 0.033 and 0.067 mg/kg/day groups, respectively. Both groups had similar clinical characteristics at the start of GH treatment (Table 2).

Table 2.  Baseline patient demographic characteristics.

Characteristic	GH treatment group
	0.033 mg/kg/day	0.067 mg/kg/day
n (male/female)	26 (22/4)	27 (16/11)
Birth data
Birth weight (g)	1841 (775)	1857 (765)
Birth weight SDS	−2.9 (1.3)	−2.7 (0.9)
Birth length (cm)	41.2 (5.1)	41.2 (5.7)
Birth length SDS	−3.9 (1.5)	−3.5 (1.4)
Gestational age (weeks)	37.2 (3.3)	36.5 (4.0)
Baseline data
Age at treatment start (years)	7.1 (2.2)	7.1 (2.4)
Bone age (years)	5.9 (2.2)	6.2 (2.8)
Height velocity (cm/year)	5.5 (0.7)	5.4 (1.4)
Height velocity SDS	−0.4 (1.3)	−0.8 (1.1)
Height (cm)	109 (13.0)	109 (13.8)
Height SDS	−3.2 (0.7)	−3.2 (0.7)
Bone age (years)	5.9 (2.2)	6.2 (2.8)
EQ-5D, baseline	0.726 (0.047)	0.726 (0.0407)

Data are shown as mean (SD).

HRQoL

Using the mapping table (Table 1), EQ-5D scores were calculated based on height SDS scores at baseline and at AH. As shown in Table 3 there was a significant increase in EQ-5D score from baseline to end of treatment of 0.112 and 0.115 respectively for the low and high dose (p < 0.001 for both groups). The difference in improved EQ-5D between treatment groups was not significant. When compared with an untreated control group (assumed to achieve +0.3 in AH SDS from baseline) the EQ-5D improvement was estimated to be 0.01, which was significantly lower than the EQ-5D improvements in both treatment groups (p < 0.001 for both groups). A 0.01 difference in HRQoL is generally considered below the 0.03 minimal important clinical difference[20],[21].

Table 3.  Estimated height and EQ-5D scores at baseline and end of study.

Parameter	GH treatment group
	0.033 mg/kg/day	0.067 mg/kg/day
Height
Adult height SDS, mean (95% CI)	−1.8 (−2.1, −1.5)	−1.5 (−1.7, −1.2)
Δ Adult height SDS, mean (95% CI)	+1.4 (1.1, 1.6)	+1.7 (1.4, 2.0)
	p < 0.05	p < 0.05
HRQoL
EQ-5D, adult height, mean (95% CI)	0.837 (0.828; 0.846)	0.841 (0.829; 0.853)
EQ-5D, Change from baseline, mean (95% CI)	0.112 (0.092; 0.132)	0.115 (0.094; 0.136)
	p < 0.001	p < 0.001

AH

In all, 38 children completed the study and reached AH (height velocity ≤2 cm/year) after a mean (min–max) GH treatment period of 9.04 (4.98–13.26) years (Table 4). The remaining children either stopped as satisfying height was reached, albeit less than 2 cm/year (n = 6), or withdrew from treatment for other reasons (n = 9). None withdrew because of adverse events. Mean bodyweight (kg) increased during the treatment period from 17 to 57 kg (Table 4).

Table 4.  Number of patients treated with growth hormone (GH), number (%) of patients achieving adult height, and mean body weight (kg) by treatment duration (years).

Treatment duration	Number of patients treated with GH, % (n)	Number of patients achieving adult height % (n)	Mean weight for treated children, kg (SD)
≤1 year	100 (53)	0 (0)	16.95 (5.1)
≤2 year	100 (53)	0 (0)	20.91 (5.9)
≤3 year	98.11 (52)	0 (0)	24.61 (7.1)
≤4 year	96.23 (51)	0 (0)	29.11 (8.6)
≤5 year	94.34 (50)	1.89 (1)	33.99 (10.2)
≤6 year	90.57 (49)	9.43 (5)	39.10 (11.0)
≤7 year	83.02 (46)	13.21 (7)	42.71 (10.6)
≤8 year	69.81 (37)	30.19 (16)	44.82 (10.7)
≤9 year	47.17 (30)	39.62 (21)	47.13 (10.2)
≤10 year	35.85 (20)	45.28 (24)	49.15 (8.5)
≤11 year	28.30 (16)	52.83 (28)	53.28 (8.2)
≤12 year	18.87 (11)	64.15 (34)	55.41 (8.2)
≤13 year	7.55 (6)	69.81 (37)	57.58 (8.6)
≤14 year	1.89 (1)	71.70 (38)	56.00 (NA)
≤15 year	0.00 (0)	71.70 (38)	NA

The ITT analyses (n = 38) showed that the mean AH SDS (95% CI) was −1.8 (−2.1, −1.5) and −1.5 (−1.7, −1.2) for patients in the 0.033 and 0.067 mg/kg/day GH dose groups, respectively. The mean improvement in AH SDS (95% CI) was +1.4 (1.1, 1.6) and +1.7 (1.4, 2.0) for patients in the 0.033 and 0.067 mg/kg/day GH dose groups, respectively. The following factors were found to have a significant effect on outcome: gender, duration of treatment, baseline bone age, target height and age, height SDS and maximum GH peak at treatment start.

A trend in favour of the higher dose in mean improvement in AH SDS between the two doses was observed (p = 0.11). When the two treatment groups were compared with a matching untreated control group who were assumed to gain +0.3 in AH SDS from baseline the results were also significant for both doses (p < 0.01).

Discussion

Results of this study show that long-term, continuous GH treatment of short children born SGA results in a significant improvement in AH. The improvements in AH were mapped into a significant concomitant improvement in HRQoL. The improvements in HRQoL are substantially above the considered minimal important clinical difference associated with the EQ-5D instrument (∼0.03)[20],[21].

The SGA group was short at start of treatment with a mean baseline height SDS of −3.2 yet they achieved a height within the normal range (−2 SDS to 2 SDS) in adulthood following GH treatment. The standardised AH showed a trend in favour of the children who received 0.067 mg/kg/day over those who received 0.033 mg/kg/day. This dose difference corresponds to the estimated dose differences obtained in a recent meta-analysis[22]. Furthermore, there was a significant improvement in estimated HRQoL from baseline in both treatment groups; HRQoL was similar between treatment groups. The AH reported for the cohort of short SGA children in the present study is, as expected, below that previously reported by Bannink et al. (2007)[23]. This is due to the omission of GHD children from the present analysis. Non-GHD SGA children are recognised to be less sensitive to the growth-promoting effect of GH[24].

In this study, a retrospective evaluation of HRQoL data mapping previously published data onto height SDS, showed a significant improvement in estimated HRQoL between baseline and AH in GH-treated SGA subjects. In the original prospective study using the full dataset (including GHD SGA children), a significant improvement in HRQoL in SGA subjects treated with long-term GH therapy was observed[17]. The TNO AZL Children's Quality of Life-short stature (The Netherlands Organization for Applied Scientific Research Academic Medical Centre Children's Quality of Life – Short Stature – TACQOL-S), a disorder specific questionnaire designed to assess the impact of short stature on quality of life for children aged 5–15 years[17], and the Child Health Questionnaire (CHQ)[25] were used. These tests evaluated HRQoL in the two groups of adolescents born SGA without spontaneous catch-up growth; a GH-treated group (n = 24; mean [SD] age 14.2 [1.2] years) treated with GH for a mean (SD) duration of 8.8 [1.7] years and an untreated group (n = 15, mean age 14.2 [1.1] years). The GH-treated group scored better HRQoL on all TACQOL-S scales with significantly better scores for ‘physical abilities’ (p = 0.002), and ‘contact with adults’ (p = 0.002). The CHQ showed near normal HRQoL in adolescents after long-term GH therapy, although no significant differences were evident between the GH-treated and untreated subjects. Another study in the same patient group indicated a reduction in problem behaviour in short SGA children treated with GH[26].

The HRQoL results in the present study are based on the pivotal assumption that HRQoL can be mapped onto height SDS scores. Although criticism may be levelled at the study design since HRQoL scores were estimated and not directly obtained from study participants, the indirect assessment of HRQoL applied in the present study provides a reasonable method of obtaining HRQoL data in this special population. First, it should be considered that when the study was initiated (April 1991) HRQoL was a new science not commonly investigated in clinical trials and the questionnaires included in the original study do not meet the requirement of today's regulatory and reimbursement agencies. New HRQoL data from a new prospective long-term clinical study would take another 10–15 years to obtain. Second, HRQoL data from young children are inherently unreliable as the questionnaires are difficult to complete and the validity of baseline data especially is often questionable[27]. Thus, mapping HRQoL data onto clinical data may provide more reliable results in this special population. Finally, It should also be considered that the most compelling HRQoL data would be provided by a new randomised controlled clinical trial comparing GH treatment with no treatment, an approach that is currently ethically unjustifiable given the known clinical benefits of GH in this population and was equally so when this study was initiated.

This analysis represents a simple conservative approach to estimate the HRQoL benefits in SGA children treated with GH. Although the analysis only mapped height to HRQoL, other potential issues may increase or decrease the HRQoL. For example, children born SGA are thought to have an increased risk of diabetes and cardiovascular events⁸ and there is evidence to show that women of short stature have a two-fold risk of caesarean sections when giving birth, compared with taller patients (p < 0.001)[28]. Moreover, there is a considerable impact of short stature on social factors such as academic achievement, social class and risk of suicide. The early school progress of boys has been reported to be influenced by height. A study of 2,848 children aged 5–12 years found that of boys who were experiencing difficulty at school, those of short stature were more likely to have to repeat a grade[29]. It was suggested that this was an example of a societal perception of childhood ability that was biased against smaller children[29]. Slow growth in infancy has also been linked to a lower income in later life. A study of 4,630 boys concluded that irrespective of the social class into which they were born, those who grew slowly between birth and 1 year of age had poor educational achievements, a lower occupational status and a lower income than those who grew more rapidly[30]. A strong inverse relationship has been shown between height and risk of suicide. In a population of 1,299,177 Swedish men between the age of 18 and 49, there was a two-fold increased risk of suicide in short men than tall men. A 5 cm increase in height was found to be associated with a 9% decrease in suicide risk[31]. A further study linked short birth length to the increased risk of violent suicide attempts[32].

Conclusion

The results of this study suggest that children born SGA without signs of persistent catch-up growth treated with long-term GH to AH may experience, in addition to a significant improvement in height, an associated relevant improvement in HRQoL. However, future prospective clinical studies should include HRQoL instruments as the current analysis rests on the pivotal assumption that HRQoL can be mapped onto height.

Transparency

Declaration of funding: This clinical study was sponsored by Novo Nordisk.

Declaration of financial/other relationships: Christian Born Djurhuus, Torsten Christensen and Kirsten Joens are all employees of Novo Nordisk A/S Denmark.

Acknowledgements: We thank Penny Butcher, PhD of Watermeadow Medical plc for medical writing services financially supported by Novo Nordisk.

Notes

*Norditropin is the registered trademark of Novo Nordisk A/S, Denmark.