help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism , doi:10.1210/jc.2007-1381
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Ridder, M. A. J.
Right arrow Articles by Hokken-Koelega, A. C. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Ridder, M. A. J.
Right arrow Articles by Hokken-Koelega, A. C. S.
Related Collections
Right arrow Neuroendocrinology and Pituitary
Right arrow Pediatric Endocrinology
The Journal of Clinical Endocrinology & Metabolism Vol. 93, No. 2 477-483
Copyright © 2008 by The Endocrine Society

Prediction Model for Adult Height of Small for Gestational Age Children at the Start of Growth Hormone Treatment

Maria A. J. de Ridder, Theo Stijnen and Anita C. S. Hokken-Koelega

Dutch Growth Foundation (M.A.J.d.R., A.C.S.H.-K.), 3001 KB Rotterdam, The Netherlands; Department of Epidemiology and Biostatistics (M.A.J.d.R.) Erasmus Medical Center-University Medical Center, 3000 CA Rotterdam, The Netherlands; Department of Pediatrics, Division of Endocrinology (A.C.S.H.-K.), Sophia Children’s Hospital, Erasmus Medical Center-University Medical Center Rotterdam, 3000 CB Rotterdam, The Netherlands; and Department of Medical Statistics (T.S.), Leiden University Medical Center, 2300 RC Leiden, The Netherlands

Address all correspondence and requests for reprints to: M. de Ridder, Dutch Growth Foundation, P.O. Box 23068, 3001 KB Rotterdam, The Netherlands. E-mail: m.deridder{at}erasmusmc.nl.


   Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
Context: GH treatment is approved for short children born small for gestational age (SGA). The optimal dose is not yet established.

Objective: Our objective was to develop a model for prediction of height at the onset of puberty and of adult height (AH).

Design and Setting: Two GH studies were performed in short SGA children.

Patients/Intervention: A total of 150 SGA children with height SD scores (SDS) less than –2, age 3 yr or older, no signs of catch-up growth, available height at the onset of puberty, and at least 1 yr of GH treatment before the onset of puberty were studied. In one study, patients were randomly assigned to either 0.033 or 0.067 mg/kg·d; in the other study all received 0.033 mg/kg·d. In 71 children, AH was reached.

Main Outcome Measures: Height SDS at the onset of puberty and AH SDS were calculated.

Results: Determinants positively related to height SDS at the onset of puberty were: height SDS at the start; target height SDS; and GH dose, whereas age at the start and female gender were negatively related. Positively related to AH SDS were: height SDS and chronological age – bone age at the start; target height SDS; and GH dose, whereas serum IGF binding protein (IGFBP)-3 SDS at the start was negatively related. There was a significant interaction between GH dose and IGFBP-3 SDS, indicating a smaller GH dose effect for higher levels of IGFBP-3. The final model explained 57% of the variance in height SDS at the onset of puberty and 41% of AH SDS.

Conclusions: The prediction model for height SDS at the onset of puberty and AH SDS of short SGA children treated with GH provides useful information about the expected long-term growth. Because GH dosage is one of the determinants, the model aids in determining the optimal GH dose for each child.


   Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
Children born small for gestational age (SGA) with persistent short stature can be effectively and safely treated with GH (1, 2, 3, 4, 5, 6). Patient characteristics found to be related to short-term response were: chronological age (CA) or bone age (BA) at the start, weight SD scores (SDS) at the start, and midparental height SDS (7, 8). For long-term response, target height (TH) SDS, height and weight SDS at the start, CA – BA at the start, and maximum GH response to a stimulation test were identified as significant factors (4, 9).

In Europe, the recommended dose for the approved indication (The European Agency for the Evaluation of Medicinal Products, 2003) is 0.035 mg/kg·d. In the United States, the Food and Drug Administration-approved indication describes a dose of 0.070 mg/kg·d. Several studies found a dose effect on growth response during the first treatment years (7, 8, 10, 11, 12, 13, 14, 15). For the long-term growth response, the dose effect was smaller but still significant (4, 5). The optimal dose for individual short SGA children is not yet established. Some investigators stated that a dose of 0.033 mg/kg·d results in significant gains in long-term growth, with IGF-I levels in the normal range and at lower costs (16). Others argued that the lower dose might be sufficient for children without extremely short stature (above –3 SDS) but that shorter or older children might better start with a higher dose (≥0.050 mg/kg·d), with tapering of the dose per kg when the absolute GH dose (in milligrams) is maintained over the years (5).

In the present study, we developed a model to predict height at the onset of puberty and adult height (AH) for short children born SGA who will start GH treatment. Determinants were various baseline characteristics and GH dose. The predictions from this model can be used as information about the expected AH for an individual child to decide on the prescribed GH dose.


   Subjects and Methods
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
Subjects

We used data from two GH trials in short SGA children (8, 17). These studies included children with birth length SDS for gestational age below –2.00 (18), height SDS for CA at the start below –2.00 (19), and height velocity SDS for CA below zero (19, 20), to exclude children with spontaneous catch-up growth. In both studies, CA at the start had to be 3 yr or older. Study 1 (8) included prepubertal children, defined as Tanner breast stage I for girls and a testicular volume less than 4 ml for boys (21), and age younger than 12 yr in boys and younger than 10 yr in girls. Study 2 (17) comprised a prepubertal group, aged 3–8 yr, and a group aged 8–14 yr. Both studies excluded children who had had a complicated neonatal period, endocrine or metabolic disorders, chromosomal disorders, growth failure caused by other disorders (emotional deprivation, severe chronic illness, or chondrodysplasia), or syndromes (except for Silver-Russell syndrome), as well as children who used or had used drugs that could interfere with GH treatment.

In study 1, children were randomly assigned to either 0.033 mg GH/kg·d (= 1 mg/m2·d, n = 41) or 0.067 mg GH/kg·d (2 mg/m2·d, n = 38). The prepubertal children in study 2 were randomized into either a GH group or a control group that remained untreated for 3 yr and started treatment afterward. In the older age group, all children were treated from the start. The GH dose in study 2 was 0.033 mg/kg·d. The inclusion period for study 1 was between April 1991 and January 1993, and for study 2 between October 1996 and December 1998.

For the present analysis, we selected children who were prepubertal for at least 1 yr after the start of GH treatment and had started or completed puberty at the time of analysis (December 2006). Birth year for boys had to be before 1993 and for girls before 1994. AH was used of boys born before 1988 and girls born before 1989.

Assays

Serum IGF-I and IGF binding protein (IGFBP)-3 levels were measured in one central laboratory, using a specific RIA (22). The intraassay coefficient of variation was 4%, and the interassay coefficient of variation was 6%. IGF-I and IGFBP-3 values were converted into SDS (23).

Measurements

Standing height was measured every 3 months using a Harpenden stadiometer. The mean of four measurements was used for analysis. Heights were converted into height SDS (19). AH SDS was computed using the reference values of adults (age > 20 yr). BA at the start was determined according to the Tanner and Whitehouse (24) radius, ulna, short bones score. At each visit, pubertal stages were assessed according to the method of Tanner and Whitehouse (21). The onset of puberty was defined as a breast stage II for girls and a testicular volume more than or equal to 4 ml for boys.

AH was defined as the height reached when height velocity had decreased less than 0.5 cm during the previous 6 months, and a BA of 15 yr or older for girls and 16.5 yr or older for boys.

Imputation of missing values and truncation of extreme values

Missing values were imputed using multiple imputation (25, 26). We generated five imputed data sets using the procedure SAS Proc MI (27). Variables with a nonnormal distribution were transformed to normality during the imputation procedure.

To restrict the influence of outlying values, the lowest values of the outcome and of the determinants were truncated to the first percentile and the highest values to the 99th percentile.

Development of the prediction model

The potential determinants were: 1) initial characteristics: sex, birth length SDS, and birth weight SDS (18), gestational age, and TH SDS (19); and 2) characteristics at the start of GH treatment: CA, height SDS, weight SDS, body mass index SDS (19), BA and CA – BA, maximum GH response to GH stimulation tests, serum IGF-I SDS, and IGFBP-3 SDS and GH dose.

We first developed separate models for height SDS at the onset of puberty and for AH SDS, using forward selection with an inclusion criterion of P < 0.05.

Next, we constructed a model for both outcomes (height SDS at the onset of puberty and AH SDS), using repeated measurements analysis. We started with all determinants selected in the two separate models as covariables, and the interaction terms for each determinant with time (0 = onset of puberty, 1 = AH), to allow for different effects on the two outcomes. Nonsignificant (P > 0.05) terms were excluded stepwise. After this selection, possible interactions between GH dose and each determinant in the model were tested.

Analysis of residuals

Using the model obtained by the previously described procedure, we calculated predicted outcomes and residuals (observed outcome minus predicted outcome). A possible relation between these values was assessed by examining the scatter plots, and by fitting the linear regression with the absolute values of the residuals as dependent and the predictions as independent variable.

Internal validation

For internal validation of the derived model, we used bootstrap techniques (28, 29). This is an important step in the procedure of development of a prediction model, needed to make the model less dependent from the data set. The predictive performance of a model applied to other data sets than the set on which it is derived, will be lower. Bootstrapping evaluates this difference by taking many random samples from the original data set. The models derived on these samples are consequently applied to the sample and to the original data set, and the performances are compared. The mean of the differences, called the optimism, is used to correct the residual SD and the R2 of the original model. Another result from bootstrapping is the shrinkage factor for the estimated coefficients of the model, which can be used to obtain the final prediction formula, corrected for overoptimism. For the internal validation of our model, we used 200 bootstrap samples.


   Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
The study group consisted of 150 children, of whom 115 (35 in study 1, 80 in study 2) received GH treatment in a dose of 0.033 mg/kg·d (dose 1) and 35 in a dose of 0.066 mg/kg·d (dose 2). Characteristics are listed in Table 1Go. At the onset of puberty, 129 (86%) of the children had a height above –2.00 SDS. A total of 71 children had AH available (38 children with dose 1 and 33 children with dose 2). Of them, 55 (77%) reached an AH above –2.00 SDS.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Descriptives

 
For height SDS at the onset of puberty, the significant determinants were height SDS at the start, TH SDS, GH dose (positive effects), age at the start, gender, and IGF-I SDS at the start (negative effects). For AH SDS, the significant determinants were height SDS at the start, TH SDS, GH dose, CA – BA at the start (positive effects), and IGFBP-3 SDS at the start (negative effect).

So, for the repeated measurements analysis, fitting one model for both outcomes, we started with candidate predictors TH SDS, GH dose, gender, age, CA – BA, height SDS, IGF-I SDS, and IGFBP-3 SDS (all measured at the start of treatment), and the interactions of these variables with time. In the backward selection, IGFBP-3 SDS was significant, but IGF-I SDS was removed because it had no significant contribution to the model. IGFBP-3 SDS was correlated with IGF-I SDS (r = 0.47; P < 0.0001) but proved to be a stronger determinant. The interactions of height SDS at the start, TH SDS, and GH dose effects with time (onset of puberty or AH) were not significant, indicating that the effects of these determinants are equal for both outcomes. The effects of age at the start and gender were only significant for height SDS at the onset of puberty, whereas IGFBP-3 SDS was only significant for AH SDS. For CA – BA, the effect on height at the onset of puberty was much smaller than the effect on AH SDS. There was a significant interaction between GH dose and IGFBP-3 SDS. This indicated that there was not one constant dose effect, but the dose effect was depending on the value of IGFBP-3 SDS at the start. A higher IGFBP-3 was related to a smaller effect of GH dose. The relation between the GH dose effect and the level of IGFBP-3 is illustrated in Fig. 1Go. We plotted the outcome (height SDS at the onset of puberty in Fig. 1AGo and AH SDS in Fig. 1BGo), adjusted for gender, TH SDS, height SDS, age, and CA – BA at the start, against the IGFBP-3 SDS value. Through these scatter plots, separate splines were drawn for cases treated with dose 1 and cases treated with dose 2. The distance between the splines, which is decreasing with increasing value of IGFBP-3 SDS, represents the dose effect on the outcome. The dose effect is plotted at the bottom of each figure. The plots also show that there is only a minority (15%) of children with IGFBP-3 below –2.5 SDS, in which the GH dose effect is substantial.


Figure 1
View larger version (12K):
[in this window]
[in a new window]

  		FIG. 1. Illustration of the effects of GH dose in relation to IGFBP-3 levels (SDS). Plotted are height SDS at the onset of puberty (A) and AH (B), both adjusted for gender, TH SDS, height SDS, age, and CA – BA at the start against IGFBP-3 SDS, dose 1 (+) and dose 2 (0). The distance between the splines indicates the dose effect and is plotted at the bottom.

 
The model explained 74% of the variance in height SDS at the onset of puberty and 68% of the variance in AH SDS. After correction for overfitting, these percentages were 57 and 41%. Table 2Go shows the results of the repeated measurements model.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Results of the final model for prediction of height at the onset of puberty and AH

 
To obtain prediction formulas, we refitted the model using for each outcome only the relevant terms (P < 0.20) and applied the shrinkage factor determined by bootstrapping. The prediction formulas are also given in Table 2Go.

In Table 3Go and Fig. 2Go, examples of predictions for four children are shown, presuming that either 0.033 mg GH/kg·d (dose 1) or 0.067 mg GH/kg·d (dose 2) is given. The predicted height SDS at the onset of puberty and the predicted AH SDS were plotted, both with 95% prediction interval. A reference line was drawn at –2 SDS, and the TH range (defined as the TH SDS ± 1.3) of the child was marked. For each child, AH SDS is lower than height SDS at the onset of puberty. This decrease in height SDS during puberty is mainly related to age at the start (for late starters the decrease is lower) and, to a lesser extent, to gender (more decrease for males) and IGFBP-3 SDS (more decrease for higher levels). At the right of each plot, the total gain in height SDS during treatment is plotted. By applying the prediction formula, it is expected that child 1 will achieve an AH above –2 SDS, with any of the GH dosages. For both dosages, predicted AH lies within the TH range, and the total gain in height SDS during treatment will be more than 1.5. Treatment with a GH dose of 0.033 mg/kg·d is likely to be sufficient for this child. Child 2 has a predicted AH below –2 SDS, even when dose 2 is given. However, with any dose, the predicted AH lies within the TH range. Dose 1 will result into a gain in height SDS of 1.34, whereas for dose 2, the predicted gain is 1.49 SDS. The small difference (1 cm) in the predicted growth response between the two doses is related to the relatively high serum level of IGFBP-3 (0 SDS). Therefore, for child 2, it does not seem reasonable to prescribe the higher dose. On the contrary, for child 3, the dose effect is expected to be substantial. With dose 1, an AH below –2 SDS is predicted, whereas the prediction is –1.23 SDS with dose 2, so within the normal range and within the TH range. This may justify prescribing a dose of 0.067 mg GH/kg·d. For child 4, aged 8 yr, the prospects are rather low, and the effect of a higher dose is limited, again because of the relatively high level of IGFBP-3 (0.5 SDS). The predicted gain in height SDS is less than one, with either dose. One may doubt whether GH treatment will be beneficial for this child.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Examples of predictions

 

Figure 2
View larger version (19K):
[in this window]
[in a new window]

  		FIG. 2. Examples of predictions for four children (also see Table 3Go). Predicted values at the onset of puberty and at AH and 95% prediction interval are plotted for doses 1 (connected with solid lines) and 2 (connected with dashed lines). The TH range (TH SDS ± 1.3) is marked. At the right, the gain in height SDS is plotted.

 

   Discussion
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
We have developed a model to predict height SDS at the onset of puberty and AH (AH SDS) for children born SGA with persistent short stature, in case GH treatment is considered. For height SDS at the onset of puberty, the model explains 57% of the variance and for AH SDS 41%. These percentages are corrected for overfitting, to give a realistic presentation of the performance of the model in general practice.

Our model includes generally available factors as TH SDS and height SDS and BA at the start of treatment. Serum IGFBP-3 levels are not always available for short children born SGA. When we removed IGFBP-3 SDS as a potential determinant, the model included IGF-I SDS as a significant determinant (estimated coefficient = –0.08; P = 0.03). This model had an 11% lower explained variance compared with our final model including IGFBP-3 SDS. There was no significant interaction between IGF-I SDS and GH dose.

We found a significant interaction between GH dose and IGFBP-3 SDS, with a negative coefficient. This means that the effect of a higher dose is depending on the level of IGFBP-3 of the child: a lower IGFBP-3 SDS at the start is related to a larger dose effect. For a child with an IGFBP-3 SDS of –2, the expected difference in AH is 0.65 SDS (~4.6 cm) when treated with 0.067 mg GH/kg·d compared with 0.033 mg GH/kg·d. For a child with an IGFBP-3 SDS of zero, this difference is only 0.11 SDS (0.8 cm). So, estimating the effect of a higher dose of GH treatment for an individual child requires a measurement of serum IGFBP-3 at the start of treatment. Therefore, we recommend the measurement of IGFBP-3 and calculation of its SDS in SGA children for whom GH treatment is considered.

When a model was constructed only for the outcome height at the onset of puberty (the first step in our procedure), IGF-I SDS was selected as a significant predictor. However, in the model only for the outcome AH, IGFBP-3 SDS was superior over IGF-I SDS as a predictor. There was a significant correlation between IGFBP-3 SDS and IGF-I SDS. Also in the repeated measurement model, in which we applied backward selection and started with both IGF-I SDS and IGFBP-3 SDS as candidate predictors, IGF-I SDS was not significant and, therefore, not included in the final model.

Internal validation of the prediction model was performed by bootstrapping. With this procedure, the model is corrected for overfitting and will give valid predictions with correct confidence limits for new patients. It is not necessary to perform an external validation. Applying the model to data of an independent cohort is still interesting.

Ranke et al. (7) developed prediction models for short-term growth response to GH treatment in short children born SGA. For both first and second-year responses (in cm/yr), the predicting factors were: age at the start (negatively related), weight SDS at the start, midparental height SDS, and GH dose (all positively related). In a Swedish study (9), a regression model for AH SDS was presented with paternal height SDS, height SDS at the start (both positively related), age at the start, and maximal GH response during provocation tests (both negatively related) as predictors. The Swedish study group differed from our group with respect to the percentage of children with GH deficiency (37% with a maximal GH response < 5.3 µg/liter vs. 21% in our study group) and the age range (2.5–15.1 vs. 3.0–11.2 yr in our study group). In addition, dosage of GH treatment was only randomized during puberty, and no significant influence of GH dose on the AH was found.

The interpretation of a predicted value for AH SDS depends on the objective of GH treatment. First, the goal might be to achieve an AH within the normal range (above –2 SDS). Second, the aim could be an AH in the TH range, usually defined as the TH ± 1.3 SDS. Third, a substantial gain in height SDS during treatment might be regarded as a reasonable goal. For an individual child, when the AH predicted for a GH dose of 0.033 mg/kg·d is unsatisfactory, a higher dose should be considered. Our study included only two fixed dosages (0.033 or 0.066 mg/kg·d). Therefore, for estimating results of intermediate doses, we have to specify the type of relation between GH dose and the effect on height SDS. Two assumptions are reasonable: a linear relation, meaning that each extra milligram of GH has the same additive effect; or a log linear relation, used in some studies (30, 31), meaning that the effect of an extra mg GH decreases with increasing dose. For example, the effect of 0.040 mg/kg·d compared with 0.035 mg/kg·d is larger than the effect of 0.065 mg/kg·d compared with 0.070 mg/kg·d. According to our model, for a child with IGFBP-3 SDS at the start of –2, the effect of 0.066 mg/kg·d compared with 0.033 mg/kg·d is 0.65 SDS. If the effect of GH dose is linear, the effect of a dose in the middle between these two doses, so 0.050 mg/kg·d, is half of 0.65, so 0.325 SDS. However, if the effect of GH dose is log linear, the effect of a dose of 0.05 mg/kg·d is 0.38 SDS. However, for doses between 0.033 and 0.066 mg/kg·d, the difference between the two assumptions (linear or log linear) is small, and, for simplicity, a linear relation can be assumed. Only for doses outside this range (extrapolation of dose effect), the difference might become large. Therefore, we recommend the following. For a child born SGA, the predicted AH SDS for a GH dose of 0.033 mg/kg·d should be computed. If this prediction is satisfactory, according to the considerations mentioned previously, the child is treated with this dose. If not, the predicted AH SDS for dose 0.066 mg/kg·d is computed. If this prediction is still relatively low, the benefit of any GH treatment should be doubted. If this prediction is in the desired range, the child is treated with this double dose. If the prediction is too far above the aimed AH SDS, the dose can be computed using the following formula: GH dose (mg/kg·d) = 0.033 *[aimed AH SDS – 0.11 – 0.66 *height SDS – 0.12 *TH SDS + 0.11 *IGFBP-3 SDS – 0.21 *(CA – BA)]/(0.15 – 0.27 *IGFBP-3 SDS) + 1.

For example, a child with a height SDS of –3.86, age 7.35 yr, BA 6 yr, IGFBP-3 SDS –2.46, and TH SDS –1.57 has a predicted AH SDS for dose 0.033 mg/kg·d of –2.07, so still outside the normal range. If a dose of 0.066 mg/kg·d will be used, the predicted AH SDS is –1.26. If the aimed AH SDS is –1.5, the required GH dose is computed as 0.056 mg/kg·d.

The prediction formulas given in Table 2Go, as well as the formula for calculation of the suitable GH dose, described previously, will be implemented in the next update of the Growth Analyser (www.growthanalyser.org).

In conclusion, the presented model predicts the expected AH of a short child born SGA who will be treated with GH. This may help in determining the optimal GH dose for each child.


   Footnotes
 
Disclosure Statement: The authors have nothing to declare.

First Published Online November 13, 2007

Abbreviations: AH, Adult height; BA, bone age; CA, chronological age; IGFBP, IGF binding protein; SDS, SD score; SGA, small for gestational age; TH, target height.

Received June 21, 2007.

Accepted November 5, 2007.


   References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Clayton PE, Cianfarani S, Czernichow P, Johannsson G, Rapaport R, Rogol A 2007 Management of the child born small for gestational age through to adulthood: a consensus statement of the International Societies of Pediatric Endocrinology and the Growth Hormone Research Society. J Clin Endocrinol Metab 92:804–810[Abstract/Free Full Text]
  2. Hokken-Koelega A, van Pareren Y, Arends N, Boonstra V 2004 Efficacy and safety of long-term continuous growth hormone treatment of children born small for gestational age. Horm Res 62(Suppl 3):149–154
  3. de Zegher F, Ong KK, Ibanez L, Dunger DB 2006 Growth hormone therapy in short children born small for gestational age. Horm Res 65(Suppl 3):145–152
  4. Van Pareren Y, Mulder P, Houdijk M, Jansen M, Reeser M, Hokken-Koelega A 2003 Adult height after long-term, continuous growth hormone (GH) treatment in short children born small for gestational age: results of a randomized, double-blind, dose-response GH trial. J Clin Endocrinol Metab 88:3584–3590[Abstract/Free Full Text]
  5. de Zegher F, Hokken-Koelega A 2005 Growth hormone therapy for children born small for gestational age: height gain is less dose dependent over the long term than over the short term. Pediatrics 115:e458–e462
  6. Cutfield WS, Lindberg A, Rapaport R, Wajnrajch MP, Saenger P 2006 Safety of growth hormone treatment in children born small for gestational age: the US trial and KIGS analysis. Horm Res 65(Suppl 3):153–159
  7. Ranke MB, Lindberg A, Cowell CT, Wikland KA, Reiter EO, Wilton P, Price DA, KIGS International Board 2003 Prediction of response to growth hormone treatment in short children born small for gestational age: analysis of data from KIGS (Pharmacia International Growth Database). J Clin Endocrinol Metab 88:125–131[Abstract/Free Full Text]
  8. Sas T, de Waal W, Mulder P, Houdijk M, Jansen M, Reeser M, Hokken-Koelega A 1999 Growth hormone treatment in children with short stature born small for gestational age: 5-year results of a randomized, double-blind, dose-response trial. J Clin Endocrinol Metab 84:3064–3070[Abstract/Free Full Text]
  9. Dahlgren J, Albertsson-Wikland A 2005 Final height in short children born small for gestational age treated with growth hormone. Pediatr Res 57:216–222[Medline]
  10. Boguszewski M, Albertsson-Wikland K, Aronsson S, Gustafsson J, Hagenas L, Westgren U, Westphal O, Lipsanen-Nyman M, Sipila I, Gellert P, Muller J, Madsen B 1998 Growth hormone treatment of short children born small-for-gestational-age: the Nordic Multicentre Trial. Acta Paediatr 87:257–263[CrossRef][Medline]
  11. Chatelain P, Job JC, Blanchard J, Ducret JP, Oliver M, Sagnard L, Vanderschueren-Lodeweyckx M 1994 Dose-dependent catch-up growth after 2 years of growth hormone treatment in intrauterine growth-retarded children. Belgian and French Pediatric Clinics and Sanofi-Choay (France). J Clin Endocrinol Metab 78:1454–1460[Abstract]
  12. de Zegher F, Albertsson-Wikland K, Wilton P, Chatelain P, Jonsson B, Lofstrom A, Butenandt O, Chaussain JL 1996 Growth hormone treatment of short children born small for gestational age: metanalysis of four independent, randomized, controlled, multicentre studies. Acta Paediatr Suppl 417:27–31[Medline]
  13. de Zegher F, Maes M, Gargosky SE, Heinrichs C, Du Caju MV, Thiry G, De Schepper J, Craen M, Breysem L, Lofstrom A, Jonsson P, Bourguignon JP, Malvaux P, Rosenfeld RG 1996 High-dose growth hormone treatment of short children born small for gestational age. J Clin Endocrinol Metab 81:1887–1892[Abstract]
  14. de Zegher F, Butenandt O, Chatelain P, Albertsson-Wikland K, Jonsson B, Lofstrom A, Chaussain JL 1997 Growth hormone treatment of short children born small for gestational age: reappraisal of the rate of bone maturation over 2 years and metanalysis of height gain over 4 years. Acta Paediatr Suppl 423:207–212[Medline]
  15. de Zegher F, Albertsson-Wikland K, Wollmann HA, Chatelain P, Chaussain JL, Lofstrom A, Jonsson B, Rosenfeld RG 2000 Growth hormone treatment of short children born small for gestational age: growth responses with continuous and discontinuous regimens over 6 years. J Clin Endocrinol Metab 85:2816–2821[Abstract/Free Full Text]
  16. Johnston LB, Savage MO 2004 Should recombinant human growth hormone therapy be used in short small for gestational age children? Arch Dis Child 89:740–744[Abstract/Free Full Text]
  17. Arends NJ, Boonstra VH, Mulder PG, Odink RJ, Stokvis-Brantsma WH, Rongen-Westerlaken C, Mulder JC, Delemarre-Van de Waal H, Reeser HM, Jansen M, Waelkens JJ, Hokken-Koelega AC 2003 GH treatment and its effect on bone mineral density, bone maturation and growth in short children born small for gestational age: 3-year results of a randomized, controlled GH trial. Clin Endocrinol (Oxf) 59:779–787[CrossRef][Medline]
  18. Usher R, McLean F 1969 Intrauterine growth of live-born Caucasian infants at sea level: standards obtained from measurements in 7 dimensions of infants born between 25 and 44 weeks of gestation. J Pediatr 74:901–910[CrossRef][Medline]
  19. Roede MJ, van Wieringen JC 1980 Growth diagrams 1980; Netherlands third nation-wide biometric survey. Tijdschr Soc Gezondheidsz 63(Suppl):1–34
  20. Rikken B, Wit JM 1992 Prepubertal height velocity references over a wide age range. Arch Dis Child 67:1277–1280[Abstract/Free Full Text]
  21. Tanner JM, Whitehouse RH 1976 Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Arch Dis Child 51:170–179[Abstract/Free Full Text]
  22. Hokken-Koelega AC, Hackeng WH, Stijnen T, Wit JM, de Muinck Keizer-Schrama SM, Drop SL 1990 Twenty-four-hour plasma growth hormone (GH) profiles, urinary GH excretion, and plasma insulin-like growth factor-I and -II levels in prepubertal children with chronic renal insufficiency and severe growth retardation. J Clin Endocrinol Metab 71:688–695[Abstract/Free Full Text]
  23. Rikken B, van Doorn J, Ringeling A, Van den Brande JL, Massa G, Wit JM 1998 Plasma levels of insulin-like growth factor (IGF)-I, IGF-II and IGF-binding protein-3 in the evaluation of childhood growth hormone deficiency. Horm Res 50:166–176[CrossRef][Medline]
  24. Tanner JM, Whitehouse R, Cameron N, Marshall W, Healy M, Goldstein H 1983 Assessment of skeletal maturity and prediction of adult height (TW2-method). 2nd ed. London: Academic Press
  25. Rubin DB 1987 Multiple imputation for nonresponse in surveys. New York:Wiley-VCH
  26. Schäfer JL 1997 Analysis of incomplete multivariate data. New York: Chapman & Hall/CRC
  27. SAS Online Doc 9.1.3 2006. Cary, NC: SAS Institute Inc.
  28. Harrell Jr FE, Lee KL, Mark DB 1996 Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 15:361–387[CrossRef][Medline]
  29. Efron B, Tibshirani R 1993 An introduction to the bootstrap. New York:Chapman & Hall/CRC
  30. Blethen SL, Compton P, Lippe BM, Rosenfeld RG, August GP, Johanson A 1993 Factors predicting the response to growth hormone (GH) therapy in prepubertal children with GH deficiency. J Clin Endocrinol Metab 76:574–579[Abstract]
  31. Ranke MB, Lindberg A, Chatelain P, Wilton P, Cutfield W, Albertsson-Wikland K, Price DA 1999 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. KIGS International Board. Kabi Pharmacia International Growth Study. J Clin Endocrinol Metab 84:1174–1183[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J. Clin. Endocrinol. Metab.Home page
B. Kristrom, A. S. Aronson, J. Dahlgren, J. Gustafsson, M. Halldin, S. A. Ivarsson, N.-O. Nilsson, J. Svensson, T. Tuvemo, and K. Albertsson-Wikland
Growth Hormone (GH) Dosing during Catch-Up Growth Guided by Individual Responsiveness Decreases Growth Response Variability in Prepubertal Children with GH Deficiency or Idiopathic Short Stature
J. Clin. Endocrinol. Metab., February 1, 2009; 94(2): 483 - 490.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Ridder, M. A. J.
Right arrow Articles by Hokken-Koelega, A. C. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Ridder, M. A. J.
Right arrow Articles by Hokken-Koelega, A. C. S.
Related Collections
Right arrow Neuroendocrinology and Pituitary
Right arrow Pediatric Endocrinology

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals