A Multicenter, Randomized, Double-Blind, Placebo-Controlled Trial to Evaluate the Metabolic and Respiratory Effects of Growth Hormone in Children With Cystic Fibrosis
OBJECTIVE. Positive effects of growth hormone therapy on growth, nutritional status, and lung function have been observed in patients with cystic fibrosis, but the current evidence is based on unblinded studies that involved a small number of patients. This trial was designed as a multicenter, randomized, placebo-controlled, double-blind study to assess the efficacy and safety of 2 dosages of growth hormone in cystic fibrosis.
METHODS. Sixty-three dystrophic patients with cystic fibrosis were randomly assigned for 24 weeks to 1 of 3 treatment arms: growth hormone dosage of 0.11 IU/kg body weight per day, growth hormone dosage of 0.21 IU/kg body weight per day, or placebo. The 24-week double-blind period was followed by an open treatment period of 24 weeks. The primary outcome measure was the change in forced expiratory volume in 1 second in percentage predicted from baseline. Secondary outcome measures were changes in height, weight, and exercise tolerance.
RESULTS. Height, growth velocity, and growth factors (insulin-like growth factor 1 and insulin-like growth factor–binding protein 3) increased significantly in both treatment groups, whereas weight gain did not differ between the growth hormone groups and placebo. A trend toward improvement in absolute forced vital capacity was observed in patients who received the higher growth hormone dosage, whereas forced expiratory volume in 1 second did not change significantly with growth hormone treatment. Maximal oxygen uptake during peak exercise increased significantly in treated patients. There were no significant differences in the frequency or severity of adverse effects or in the incidence of abnormalities in glucose metabolism.
CONCLUSIONS. These data suggest that in the group investigated, growth hormone therapy was well tolerated and had positive metabolic effects but did not result in short-term improvement of lung function in patients with cystic fibrosis.
Cystic fibrosis (CF) is caused by defects in the cystic fibrosis transmembrane regulator (CFTR) protein, which functions as a cyclic adenosine monophosphate–regulated chloride channel. The CFTR defect results in a multisystem disorder with the dominant clinical features being chronic lung disease and exocrine pancreatic insufficiency. Advances in medical treatment, including intensive antibiotic treatment and aggressive nutritional support, have resulted in improved prognosis with increasing numbers of patients surviving into adult life.1,2 However, normal growth and body weight are not achieved in all patients.3 Catabolism has been recognized as a poor prognostic marker in these patients, because it is correlated with reduced life expectancy.4–6
Malnutrition is a major cause of growth retardation and results from loss of exocrine pancreatic function, increased energy requirements, and systemic anorexia in patients with severe pulmonary disease manifestation. Chronically increased concentrations of inflammatory cytokines such as tumor necrosis factor α and interleukin 1 and 6 affect the production and the secretion of growth hormone (GH) or growth factors but can also inhibit growth at the tissue level as a result of GH resistance.7,8 Overall, patients with CF have been shown to be shorter than predicted by their genetic target range.3,5
Recombinant human GH (rhGH) as an anabolic adjunct to stimulate growth has been studied in a number of trials that reported benefits on growth, nutritional status, and lung volume.7–22 All published trials were conducted unblinded and involved a small number of patients. This study was undertaken as the first double-blind, placebo-controlled trial to determine whether the addition of GH to standard therapy has beneficial metabolic and pulmonary effects in patients with CF.
The trial was designed as a multicenter, randomized, double-blind, placebo-controlled, parallel-groups study to compare the efficacy and the safety of 2 fixed dosages of rhGH or placebo in patients with CF. The study was conducted as a 24-week double-blind study with a 24-week open-label treatment period in 12 German CF centers between January 2001 and February 2004. The study was approved by the institutional review boards of the participating centers. Written informed consent was obtained in all cases.
Patients who were included in this trial had an established CF diagnosis that was based on a sweat chloride concentration >60 mmol/L and/or 2 disease-causing CFTR mutations, a bone age of 8 to 18 years, dystrophy defined as a BMI <10th and/or body weight <3rd percentile despite a high caloric intake (>120% of the recommended dietary allowance) according to a 3-day food-intake diary. Exclusion criteria were an acute pulmonary exacerbation in the 4 weeks before entering the trial; diabetes (fasting plasma glucose > 126 mg/dL); liver cirrhosis with hypoalbuminemia; serum creatinine > 120 μmol/L; inability to perform exercise and lung-function testing; history of malignancy; suspected noncompliance; participation in any other clinical trial during active treatment phase; pregnancy or lactation; and treatment with GH, anabolic steroids, or systemic corticosteroids within 12 months before the start of the study. Eligible patients underwent a screening visit with clinical assessment and lung-function testing. The patient's body composition was determined by means of bioelectrical impedance analysis (BIA) measured by BIA 2000-M (Data Input, Frankfurt, Germany). Fat mass, lean body mass, and extracellular and total body water were recorded. Radiograph of the patient's left wrist was performed for bone age according to Greulich and Pyle.23 Patients were also instructed to complete a 3-day food-intake history. Patients who fulfilled the entry criteria were entered into the trial 1 week after the screening visit and reassessed at 12-week intervals thereafter.
Standing height was recorded as the means of 3 independent measurements by use of a wall-mounted Harpenden stadiometer (Holtain Ltd, Crymych, Wales). Body weight was measured on a standardized scale (Soehnle S 20, Murrhardt, Germany). Height and body weight values were compared with the standards of Reinken and van Oost.24 BMI was assessed by the standards of Rolland-Cachera et al.25
Patients were assigned to 1 of the 3 treatment arms for 24 weeks in a double-blind manner with 0.070 mg/kg body weight per day (∼0.21 IU/kg body weight per day) somatropin (subsequently referred to as higher dosage), 0.039 mg/kg body weight per day (∼0.11 IU/kg body weight per day) somatropin (subsequently referred to as lower dosage), or placebo. Recombinant somatropin from Escherichia coli K12 (Genotropin; Pfizer, Karlsruhe, Germany) was supplied in 2-chamber cartridges with powder and solvent for injection using a Genotropin Pen. For safety reasons, the maximum daily GH dosage was limited to 3.4 mg. After the end of the double-blind treatment period, patients who were on GH therapy were maintained on their current GH dosage for an additional 24 weeks; patients in the placebo group were randomly assigned to either the low or the high GH-dosage treatment regimen.
Assessments at the 12-week intervals included pulmonary-function measurements with a spirometer as well as blood sampling to measure growth factors (insulin-like growth factor 1 [IGF-1] and IGF-binding protein 3 [IGF-BP3]). In addition glucose, cholesterol, triglycerides, electrolytes, creatinine, serum urea nitrogen, albumin, liver function test (aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and bilirubin), and immunologic parameters (red blood cells, white blood cells, thrombocytes, and immunoglobulins) were determined at these time points. An oral glucose tolerance test was performed at study entry as well as every 24 weeks during treatment. The patient received 1.75 g/kg body weight (maximum 75 g) of oral glucose, and plasma levels were determined immediately before glucose intake as well as after 120 minutes.
At the 24-week intervals, exercise testing using a bicycle ergometer26 was performed. Initial work rate was set at 20 W and increased every minute by increments of 15 W for patients who were between 125 and 150 cm or 20 W for patients who were taller than 150 cm. Ten-watt increments were used for patients with a forced expiratory volume in 1 second (FEV1) percentage predicted of 30% or less. Heart rate, O2 consumption (V̇o2), CO2 production, and ventilatory volumes were measured continuously while the patient was breathing through a mouthpiece. The exercise test was terminated at subjective exhaustion. A Borg scale was used to assess subjective exertion at each exercise testing.27
A patient diary card that contained information on fever >38.5°C, increased cough or sputum, intake of additional and/or other antibiotics, days off school/work, and hospitalization was dispensed/collected at each visit. Quality of life was assessed every 12 weeks using a disease-specific questionnaire.28 Clinical safety variables (physical assessment, systolic and diastolic arterial blood pressure, and heart rate) were determined at each study visit.
The study was conducted according to a predefined analysis plan using a 3-stage group sequential adaptive design with sample size adjustments after planned interim analyses. For the purpose of efficacy analysis, 2 data sets were to be defined: the intention-to-treat (ITT) sample and the per-protocol (PP) sample. The ITT sample included all patients who had received at least 1 dose of study medication and who had been assessed at least once after baseline. The PP sample, being a subset of the ITT sample, included patients who did not violate the study protocol or terminate treatment prematurely.
The primary end point was the mean change from baseline in FEV1 at the end of the 24-week double-blind treatment period. Secondary objectives were changes in exercise capacity, body weight (lean body mass), lung function, hospitalization, days off school or work, and quality of life.
At the first interim analysis, the 2 null hypotheses H01 and H02 were to be rejected and the study stopped if the analysis for FEV1 yielded a P value of <.00026. If H01 or both H01 and H02 could not be rejected, then the study was to be continued with a recalculated sample size based on the effect size estimation of the interim analysis. At the second confirmatory analysis, null hypotheses that could not yet be rejected at the first interim analysis were to be rejected if the test statistic that is based on the inverse normal method exceeded the critical value 2.454.29,30
For the hypotheses that could not be rejected at the second stage, the sample size was recalculated again and the analogous procedure was conducted at the third (final) analysis by using the critical value 2.004. This procedure preserved the overall type I error rate of α = .025. For confirmatory hypothesis testing at the interim analysis as well as at the final analysis, the nonparametric Mann-Whitney U Wilcoxon rank sum W test (1-sided) was used. All other group comparisons are hypothesis generating in nature (ie, P values that resulted from statistical tests were interpreted as exploratory). The statistical analysis was conducted by ClinResearch GmbH (Cologne, Germany) using SPSS (SPSS, Chicago, IL) and SAS (SAS Institute, Cary, NC) software packages.
The study was powered to show a significant effect of GH on FEV1. The first interim analysis was planned with 12 patients in each treatment arm and conducted with 14 patients in the low-dosage group, 12 patients in the high-dosage group, and 10 patients in the placebo group (36 patients according to ITT). The second interim analysis was planned with 24 patients in each treatment arm and conducted with 20 patients in the low-dosage group, 20 patients in the high-dosage group, and 19 patients in the placebo group (59 patients according to ITT). The second interim analysis revealed that a total number of 300 patients would be needed to achieve significance for FEV1. Because recruitment of 300 patients for this study was not feasible in the participating centers, additional recruitment was stopped at this time point.
Seventy-one patients were available for the final statistical analysis. Four patients refused study participation before randomization. Sixty-four of the 67 randomly assigned patients who took study medication at least once were contained in the safety data set. The primary analysis of efficacy according to the ITT analysis was based on 63 patients (24 female, 39 male) who were randomly assigned, took study medication, and were assessed at least once after baseline. The baseline characteristics for the ITT study population are summarized in Table 1. In addition, a PP analysis of a reduced set of variables was performed on 51 patients (19 lower dosage, 16 higher dosage, and 16 placebo).
FEV1 was primarily analyzed as change from baseline in SD scores. Compared with baseline values, FEV1 did not change significantly during the 6-month period (Table 2). In addition, no difference in percentage change from baseline FEV1 was observed in GH-treated compared with placebo-treated patients with CF (P = .362; Fig 1). There was no correlation between baseline lung function and changes in FEV1 during the study period (data not shown). Although a trend toward an increase in absolute forced vital capacity was seen both in the lower- and higher-dosage groups (Table 2, Fig 2), this difference was not statistically significant compared with placebo for the low-dosage group (P = .49) and failed to reach statistical significance for the high-dosage group (P = .068). Flows at lower lung volumes (maximal expiratory flow at 50% of vital capacity and maximal midexpiratory flow) showed no systematic changes during the study period (data not shown). No major changes were observed in FEV1 during the open treatment phase in any of the groups (mean change ± SD: 0.6 ± 2.4% for all groups). The mean increase in forced vital capacity in the 24 weeks of open trial was 3.1 ± 5.5% with no significant differences between the groups.
Growth and Weight
Absolute growth velocity increased significantly in both the higher- (P = .025) and the lower-dosage groups (P = .018; Fig 3, Table 2). In parallel, growth velocity in SD scores for chronological age increased in both treated groups (P < .01). The change in height in percentage from baseline was negatively correlated with age in GH-treated patients (r = 0.61; P < .0001; Fig 4). This correlation was not observed in patients who received placebo. Growth velocity during the 24 weeks of open treatment was 4.7 ± 3.3 cm/year for the total study population without significant differences between the groups.
Absolute weight gain and changes in BMI were not significantly different between treatment groups (Table 2, Fig 5). Changes in weight in percentage from baseline were negatively correlated with age (r = 0.32; P < .05) and positively correlated with baseline lung function (r = 0.49; P = .001) (Fig 6). Similar relationships were not observed in the placebo group. Body composition reflected by body fat and lean body weight was very similar at baseline in the 3 groups (Table 1). Lean body mass increased in all 3 groups. Fat mass remained unchanged in the placebo group and decreased marginally in the 2 treatment groups (Table 2). None of these differences between groups was statistically significant.
IGF-1 and IGFBP-3
IGF-1, the metabolic mediator of anabolic GH action and growth-stimulating factor, increased in treated patients and remained unchanged in the placebo group (P < .002 for both treatment groups; Table 2). In parallel, IGFBP-3, the main binding protein of IGF-1, increased, whereas concentrations decreased in the placebo group (P < .05 for treated patients).
At baseline, there were no significant differences between the treatment groups in work rate, peak heart rate, maximal V̇o2, CO2 production, O2 saturation, minute volume, tidal volume, and subjective exertion as assessed by the Borg scale at the end of the exercise test. A nonsignificant increase in work rate was observed in both treatment groups, whereas work rate was essentially unchanged in the placebo group (Table 2). Subjective exertion as monitored on the Borg scale was similar in the various treatment groups; there were no significant differences in O2 saturation both at baseline and during peak exercise between the groups.
Absolute V̇o2 increased in both the lower- and higher-dosage groups and remained unchanged in the placebo group (P = .05 versus placebo; Fig 7). These differences were more pronounced in the PP analysis (P = .009 for the lower-dosage group and P = .02 for the higher-dosage group). During the open treatment period, both mean work rate (6.1 ± 16.6 W) and mean V̇o2 at peak exercise (86.9 ± 220.4 mL/min) increased in patients who previously received placebo.
The assessment of the quality-of-life questionnaire revealed no major differences among the treatment groups. There were no statistically significant differences in the number of days off school or work as well as the number of hospitalization days among the 3 groups.
Compliance was assessed by diaries and vial counts returned at regular study visits. According to this analysis, 19 (95.0%) of 20 patients in the higher-dosage group, all 22 (100.0%) patients in the lower-dosage group, and 19 (90.5%) of 21 patients in the placebo group confirmed the regularly received study medication.
Twelve (60.0%) of 20, 14 (63.6%) of 22, and 13 (59.1%) of 22 patients in the higher-dosage, lower-dosage, and placebo groups, respectively, experienced at least 1 adverse event (AE) during the double-blind study period. The most frequently observed single category was pulmonary exacerbations, affecting 7 (35.0%) of 20, 6 (27.3%) of 22, and 4 (18.2%) of 22 patients in the higher-dosage, lower-dosage, and placebo groups, respectively, with no significant differences among the groups.
All treatment groups were approximately equally affected by severe AEs (SAEs), namely 4 (20.0%) of 20, 5 (22.7%) of 22, and 4 (18.2%) of 22 patients in the higher-dosage, lower-dosage, and placebo groups, respectively. The distribution of SAEs was similar to that of nonserious AEs, with respiratory infections requiring antibiotic therapy being the most frequently reported SAE (Table 3).
Fasting blood glucose levels (high dosage: 91.9 ± 7.0 mg/dL; low dosage: 95.9 ± 12.0 mg/dL; placebo: 87.6 ± 10.1 mg/dL) as well as stimulated glucose levels (high dosage: 117.0 ± 62.2 mg/dL; low dosage: 109.9 ± 31.6 mg/dL; placebo: 108.0 ± 37.9 mg/dL) after oral glucose tolerance testing were almost identical at baseline. No significant changes in fasting glucose (high dosage: 96.8 ± 13.4 mg/dL; low dosage: 101.2 ± 15.2 mg/dL; placebo: 88.8 ± 13.7 mg/dL) and stimulated glucose concentrations (high dosage: 120.7 ± 59.5 mg/dL; low dosage: 128.2 ± 39.5 mg/dL; placebo: 116.1 ± 23.3 mg/dL) were seen after the 24-week double-blind, placebo-controlled period. GH treatment did not result in an inappropriate acceleration of bone age (data not shown).
To our knowledge, this is the first study to assess the effect of GH therapy in patients with CF in a double-blind, placebo-controlled design. GH treatment increased growth velocity and improved maximal V̇o2 during peak exercise but had no significant effect on pulmonary function. Therefore, GH demonstrated positive metabolic effects but did not result in short-term improvement in pulmonary status.
The increase in height and growth velocity that was observed in this study is in concordance with previous unblinded studies.9–12 Younger children grew better in response to GH therapy than did older patients. This study therefore provides additional evidence that linear growth rate can be increased with GH therapy in patients with CF. Nevertheless, it can be debated whether the dosages used in this study are sufficient to optimize growth in this patient population. IGF-1 and IGFBP-3 levels were normalized in both groups but did not exceed the reference ranges. This study provides no clear evidence for a dosage effect, because no significant differences were observed in any of the outcome variables between the 2 treatment groups. Dosages ranging from 0.16 to 0.35 mg/kg body weight per week have been used in previous trials, and the higher-dosage group in this study therefore received the highest dosage that has been given to patients with CF to date. This dosage is also higher compared with studies that were performed in other populations without primary GH deficiency.31–34 Although it cannot be excluded that a more pronounced metabolic effect can be achieved with even higher dosages, this study provides no evidence to support this hypothesis.
Weight gain was observed in all groups in this study, but no significant differences were detected between GH-treated and placebo-treated patients. The mean yearly increase in weight was 4.4 and 5 kg in the 2 previous randomized trials, which is similar to the overall weight gain that was observed during this trial.14,16 It is interesting that placebo-treated patients in this study also demonstrated a similar weight gain, which may reflect a nonspecific effect driven by study participation and more intense nutritional counseling. This underlines the need to assess treatment effects in a blinded study to clarify whether changes are directly related to therapy. Although no significant difference was observed in body weight, a trend toward higher lean body mass and lower fat mass was observed in treated patients and likely reflects the anabolic effects of GH. Whether shifts in body mass are associated with improvements in clinical status in CF is unclear. In addition, techniques that assess body composition, such as BIA, have not been thoroughly validated as an outcome parameter for clinical trials in the CF population.
A previous study described impressive weight gains after GH treatment in patients who had CF and received nutritional support via enteral tube feeding.21 Enteral tubes are often placed if weight gain is insufficient in patients with CF. Enteral tube feeding has been demonstrated to benefit most but not all of the patients with nutritional failure.35 Only 3 patients who were included in this study (all in the lower-dosage group) received enteral tube feeding as part of their routine medical care. This may reflect a selection bias for the patients who were included in this study, because enteral tube feeding is usually offered to patients who have CF with inadequate caloric intake. Because of the low number of patients who received enteral tube feeding, we could not perform a meaningful subgroup analysis. The inclusion criteria of this trial required patients to have a high caloric intake independent of the way that these calories were administered. As reflected by the previous study that focused on enteral tube–fed patients alone, weight gain may be more pronounced in these patients. Whether a high-calorie diet that is provided for extended periods, as is the case for enteral tube feeding, will improve the anabolic potency of GH could not be addressed by this study.
Pulmonary function in percentage predicted did not change significantly with GH therapy in patients with CF, although a trend toward an increase in absolute lung volumes was seen in the higher-dosage group, which supports the findings of previous unblinded trials.14,16 Whether an increase in lung volume that parallels an increase in body size can be considered a positive finding is debatable. Most of the previous treatment approaches that directly targeted the lung did not use growth as an outcome parameter, and no long-term data are available to clarify whether increases in height and absolute lung function will improve the subsequent course of the disease process. Because the primary effects of GH in CF are metabolic rather than pulmonary, the potential respiratory benefits of GH treatment may not translate into short-term increases in lung function. In addition, the lack of short-term decline in FEV1, seen in recent trials in patients with CF, including this study, have made it difficult to use lung function as a primary outcome parameter.36,37 Whether GH therapy has a stabilizing effect on lung function over time is much harder to assess and would require a longer treatment and observation period. Long-term studies to assess treatment effects on evolution of lung function over time have become challenging, with overall improvements in CF therapy that have resulted in an annual decline in pulmonary function of <2%.38,39
Positive effects of GH therapy on work load and maximal V̇o2 during exercise were observed in this trial. Maximal V̇o2 during exercise has been reported to predict mortality in patients with CF and significant lung disease, and improvements in exercise capability may increase physical activity levels and eventually patients’ outcome.40–43 An inactive lifestyle with lower physical activity has been proposed to be partially responsible for the more rapid decline of female patients during puberty.39 Whether an intervention such as GH therapy will positively affect physical activity and whether this would result in better outcome for patients with CF cannot be answered with this study design.
Compared with previous studies, this trial included older patients with CF and more advanced pulmonary disease. The mean FEV1 in these patients is considerably lower than the mean FEV1 of the overall CF population at a similar age, and this may reflect the well-described link between nutritional failure and poor lung function. The positive correlation between baseline pulmonary function and weight gain would suggest that patients with poorer lung function are unlikely to benefit from GH therapy. In addition, benefits on growth were more pertinent in younger patients. Therefore, treatment can be expected to be more efficacious if started in younger patients with preserved pulmonary function.
No SAEs of GH that could be directly attributed to GH therapy were observed in this trial. Because GH may increase blood glucose level, concerns had been raised that GH therapy would trigger diabetes in patients with CF. Random blood glucose levels or glycosylated hemoglobin concentrations are not sufficient to diagnose abnormal glucose metabolism in CF, and only glucose tolerance testing can be considered adequate to address this question.44 Similar to previous trials, none of the patients developed glucose impairment as investigated by oral glucose tolerance test during the study. This trial confirms the evidence from previous trials that GH therapy seems to be safe in children with CF with regard to glucose metabolism.
This short-term study provides evidence for positive metabolic effects of GH therapy in dystrophic patients with CF. Although growth velocity and exercise capacity as well as biochemical marker of anabolism improved with treatment, no direct effects on pulmonary function could be observed. Long-term studies in a larger cohort of patients are needed to clarify whether GH therapy will have a positive effect on lung-function decline in patients with CF. In addition, IGF-1 therapy may be a plausible alternative to GH that warrants additional study. This study suggests that the effect of GH treatment may be better if less severely affected and younger patients are included. As an end point, a combination of morphologic (eg, body composition), functional (eg, exercise capacity), pulmonary (eg, FEV1), and biochemical markers might be useful to assess the efficacy of treatment.
Investigators for the German Cystic Fibrosis Growth Hormone Study were Michael Barker (Aachen), Dirk Schnabel (Berlin, principal investigator), Doris Staab (Berlin), Volker Stephan (Bochum), Ernst Rietschel (Koeln), Antje Schuster (Düsseldorf), Lutz Nährlich (Erlangen), Felix Ratjen (Essen), Hans-Georg Posselt (Frankfurt), Joachim Kühr (Freiburg), Sabine Brömme (Halle/Saale), Manfred Ballmann (Hannover), and Matthias Griese (München).
- Accepted December 12, 2006.
- Address correspondence to Felix Ratjen, MD, PhD, Division of Respiratory Medicine, Hospital for Sick Children, 555 University Ave, Toronto, Ontario, Canada M5G 1X8. E-mail:
Financial Disclosure: This study was supported by Pharmacia. Dr Wollmann is a former employee of Pharmacia.
- ↵Cystic Fibrosis Foundation. Patient Registry 2005 Annual Report. Bethesda, MD: Cystic Fibrosis Foundation; 2006. Available at: www.cff.org/research/ClinicalResearch/PatientRegistryReport. Accessed March 27, 2007
- ↵Sharma R, Florea VG, Bolger AP, et al. Wasting as an independent predictor of mortality in patients with cystic fibrosis. Thorax.2001;56 :746– 750
- ↵Hardin DS. GH improves growth and clinical status in children with cystic fibrosis: a review of published studies. Eur J Endocrinol 2004;151(suppl 1):S81–S85
- Hardin DS, Sy JP. Effects of growth hormone treatment in children with cystic fibrosis: the National Cooperative Growth Study experience. J Pediatr. 1997;131(pt 2):S65–S69
- Alemzadeh R, Upchurch L, Mc Carthy V. Anabolic effects of growth hormone treatment in young children with cystic fibrosis. J Coll Am Nutr.1998;17 :419– 424
- ↵Hütler M, Schnabel D, Staab D, et al. Effect of growth hormone on exercise tolerance in children with cystic fibrosis. Med Sci Sports Exerc.2002;34 :567– 572
- Schibler A, von der Heiden R, Birrer P, Mullis PE. Prospective randomised treatment with recombinant human growth hormone in cystic fibrosis. Arch Dis Child.2003;88 :1078– 1081
- ↵Greulich WW, Pyle SI. Radiographic Atlas of Skeletal Development of Hand and Wrist. 2nd ed. Stanford, CA: Stanford University Press; 1959
- ↵Godfrey S. Exercise Testing in Children: Applications in Health and Disease. London, United Kingdom: WB Saunders; 1974
- ↵Quittner, AL, Sweeny, S, Watrous M, et al. Translation and linguistic validation of a disease-specific quality of life measure for cystic fibrosis. J Pediatr Psychol.2000;25 :403– 414
- ↵Conway SP, Morton A, Wolfe S. Enteral tube feeding for cystic fibrosis. Cochrane Database Syst Rev. 2000;(2):CD001198
- ↵Que C, Cullinan P, Geddes D. Improving rate of decline of FEV1 in young adults with cystic fibrosis. Thorax.2006;61 :155– 157
- Pianosi P, Leblanc J, Almudevar A. Peak oxygen uptake and mortality in children with cystic fibrosis. Thorax.2005;60 :50– 54
- Hebestreit H, Kieser S, Rudiger S, et al. Physical activity is independently related to aerobic capacity in cystic fibrosis. Eur Respir J.2006;28 :734– 739
- ↵Moorcroft AJ, Dodd ME, Morris J, Webb AK. Individualised unsupervised exercise training in adults with cystic fibrosis: a 1 year randomised controlled trial. Thorax.2004;59 :1074– 1080
- Copyright © 2007 by the American Academy of Pediatrics