Factors Influencing Adult Height in Short SGA Children
We present AH results of 170 short children born SGA who were treated in a multicentre GH trial. GH treatment improved height SDS, resulting in an AH of −1·8 SDS in boys and −1·9 SDS at AH in girls, corresponding to a TH-corrected AH of −1·1 SDS in boys and −1·3 SDS in girls. In contrast to our hypothesis, spontaneous catch-up growth after birth was negatively related to total height gain SDS during GH treatment. During puberty, height SDS declined with −0·4 in boys and −0·5 in girls. This resulted in a lower total height gain SDS and thus lower AH SDS than expected at the onset of puberty.
Growth hormone has become a frequently applied growth promoting therapy in short children born SGA. According to a systematic review in 2009, high-quality trials on AH in short SGA children treated with GH are scarce. Also, a potential limitation of growth studies is the effect of dropouts on the validity of outcome measures. Current practice is to base an analysis only on patients with complete follow-up. This can result in biased data, whereas the analysis by Lin's method results in a higher accuracy. We therefore used both methods and showed that our group of uncensored cases was a good representation of the total group.
Median total height gain in our GH-treated short SGA children was 1·1 SDS, bringing 73% of the children into the normal AH range and/or their TH range. These results are in line with data described by Van Pareren and Dahlgren. Although the need for an early start of GH treatment has been demonstrated, there is a group of short statured children that come to medical attention at an age ≥8 years. GH treatment in our older short SGA children resulted in a median height gain of 0·8 SDS, showing that adolescent children can still have impressive catch-up growth with GH treatment. These data are in agreement with the 0·9 SD increase reported in such children by Dahlgren et al.
As there is a wide variation in GH-induced growth response in short SGA children, several studies have searched for predictors of the growth response to GH treatment. Characteristics found to be related were as follows: age and height at start of GH treatment, TH and IGFBP-3 levels, explaining approximately 40% of the variability. As GH treatment is increasingly applied, it is important to investigate new factors which can influence the growth response.
Many short children born SGA show spontaneous catch-up growth after birth but not to such an extent that a height within the normal range and/or TH range is reached. We found that 42% of the children showed spontaneous catch-up growth ≥0·5 SDS in the years after birth. Particularly, children who were born premature and those with a smaller size at birth showed a greater spontaneous catch-up growth. We hypothesized that children with a greater spontaneous catch-up growth after birth have a larger potential of responding to GH treatment. Surprisingly, we found that spontaneous catch-up growth was negatively associated with the total height gain SDS during GH. One explanation might be that a greater spontaneous catch-up growth results in a height closer to TH, possibly reducing the potential for response to GH. Indeed, the distance to TH at start of GH was negatively correlated with spontaneous catch-up growth. Although spontaneous catch-up growth after birth was negatively correlated with the total height gain SDS, we did find that it was positively associated with AH SDS. This is probably the result of a greater height SDS at start of GH treatment because children with a greater spontaneous catch-up growth were taller at start of GH.
In 68% of the children, height SDS declined during puberty. Possible explanations for this decline might be a relative early onset of puberty, a reduced pubertal height gain (cm), an acceleration of bone maturation and/or a different pubertal growth pattern. Our GH-treated SGA children did, however, not start puberty at a younger age compared to normal statured Dutch children born AGA. Boys and girls started their puberty even significantly later. The average pubertal height gain in cm was reduced compared to the pubertal height gain of AGA boys and girls, which might be due to the later start of puberty. Children with delayed pubertal development also demonstrate a lower height velocity during their growth spurt. Furthermore, although there was no acceleration in bone maturation during puberty, BA at the onset of puberty was moderately advanced. These findings together might contribute to the early growth deceleration occurring from mid-puberty. Lazar et al. described a similar phenomenon in untreated short SGA children, so this phenomenon seems to be related to SGA.
Although children who started GH treatment at a younger age showed a greater total height gain SDS to AH, they in particular showed a decline in height SDS during puberty. Dahlgren et al. described similar findings in short SGA children. If treated for >2 years before the onset of puberty, there was a loss in AH SDS (−0·2 SDS) compared to those treated for <2 years. Our findings show that the height SD score before the onset of puberty is not a reliable predictor of AH, and physicians should therefore be careful in extrapolating prepubertal SD scores to predict AH because this can result in an overestimation of AH.
In 30% of the short SGA children, height at the onset of puberty was below 140 cm and in these children puberty was postponed by adding 2 years of GnRHa to GH treatment. Our data demonstrate that children who received GnRHa, due to their poor AH expectation at the onset of puberty, attained a comparable AH as children who did not. This is in line with recently published data of our group investigating combined GH/GnRHa treatment. Combined GH/GnRHa treatment resulted in a longer duration of puberty and a higher pubertal height gain compared to children who received GH only, particularly in girls. This indicates a beneficial effect of GnRHa treatment, probably as a result of delayed BA development. In the present trail, BA was not determined at start of GnRHa and during GnRHa treatment, because treatment with GnRHa in addition to GH was not randomized and depended on absolute height at start of puberty. No model is known to accurately predict AH at the onset of puberty, and we used a pragmatic cut-off level; children with a height below 140 cm at the onset of puberty were not expected to reach a normal AH. Also, our present study was not designed to investigate pubertal postponement by GnRHa vs no postponement in a randomized design. Despite these limitations, our study suggests that when GH-treated short SGA children are still short at the onset of puberty, they may benefit from combined GH/GnRHa treatment.
In conclusion, GH treatment improves AH in short children born SGA, even when started after the age of 8 years. However, these results should not lead to postponement of starting GH treatment until an older age, because a younger age at start results in a better AH despite some height loss during puberty. In contrast to our hypothesis, children with greater spontaneous catch-up growth after birth show a lower total height gain SDS during GH. Furthermore, height SDS declines from mid-puberty, due to a marked early deceleration of growth velocity.
Discussion
We present AH results of 170 short children born SGA who were treated in a multicentre GH trial. GH treatment improved height SDS, resulting in an AH of −1·8 SDS in boys and −1·9 SDS at AH in girls, corresponding to a TH-corrected AH of −1·1 SDS in boys and −1·3 SDS in girls. In contrast to our hypothesis, spontaneous catch-up growth after birth was negatively related to total height gain SDS during GH treatment. During puberty, height SDS declined with −0·4 in boys and −0·5 in girls. This resulted in a lower total height gain SDS and thus lower AH SDS than expected at the onset of puberty.
Growth hormone has become a frequently applied growth promoting therapy in short children born SGA. According to a systematic review in 2009, high-quality trials on AH in short SGA children treated with GH are scarce. Also, a potential limitation of growth studies is the effect of dropouts on the validity of outcome measures. Current practice is to base an analysis only on patients with complete follow-up. This can result in biased data, whereas the analysis by Lin's method results in a higher accuracy. We therefore used both methods and showed that our group of uncensored cases was a good representation of the total group.
Median total height gain in our GH-treated short SGA children was 1·1 SDS, bringing 73% of the children into the normal AH range and/or their TH range. These results are in line with data described by Van Pareren and Dahlgren. Although the need for an early start of GH treatment has been demonstrated, there is a group of short statured children that come to medical attention at an age ≥8 years. GH treatment in our older short SGA children resulted in a median height gain of 0·8 SDS, showing that adolescent children can still have impressive catch-up growth with GH treatment. These data are in agreement with the 0·9 SD increase reported in such children by Dahlgren et al.
As there is a wide variation in GH-induced growth response in short SGA children, several studies have searched for predictors of the growth response to GH treatment. Characteristics found to be related were as follows: age and height at start of GH treatment, TH and IGFBP-3 levels, explaining approximately 40% of the variability. As GH treatment is increasingly applied, it is important to investigate new factors which can influence the growth response.
Many short children born SGA show spontaneous catch-up growth after birth but not to such an extent that a height within the normal range and/or TH range is reached. We found that 42% of the children showed spontaneous catch-up growth ≥0·5 SDS in the years after birth. Particularly, children who were born premature and those with a smaller size at birth showed a greater spontaneous catch-up growth. We hypothesized that children with a greater spontaneous catch-up growth after birth have a larger potential of responding to GH treatment. Surprisingly, we found that spontaneous catch-up growth was negatively associated with the total height gain SDS during GH. One explanation might be that a greater spontaneous catch-up growth results in a height closer to TH, possibly reducing the potential for response to GH. Indeed, the distance to TH at start of GH was negatively correlated with spontaneous catch-up growth. Although spontaneous catch-up growth after birth was negatively correlated with the total height gain SDS, we did find that it was positively associated with AH SDS. This is probably the result of a greater height SDS at start of GH treatment because children with a greater spontaneous catch-up growth were taller at start of GH.
In 68% of the children, height SDS declined during puberty. Possible explanations for this decline might be a relative early onset of puberty, a reduced pubertal height gain (cm), an acceleration of bone maturation and/or a different pubertal growth pattern. Our GH-treated SGA children did, however, not start puberty at a younger age compared to normal statured Dutch children born AGA. Boys and girls started their puberty even significantly later. The average pubertal height gain in cm was reduced compared to the pubertal height gain of AGA boys and girls, which might be due to the later start of puberty. Children with delayed pubertal development also demonstrate a lower height velocity during their growth spurt. Furthermore, although there was no acceleration in bone maturation during puberty, BA at the onset of puberty was moderately advanced. These findings together might contribute to the early growth deceleration occurring from mid-puberty. Lazar et al. described a similar phenomenon in untreated short SGA children, so this phenomenon seems to be related to SGA.
Although children who started GH treatment at a younger age showed a greater total height gain SDS to AH, they in particular showed a decline in height SDS during puberty. Dahlgren et al. described similar findings in short SGA children. If treated for >2 years before the onset of puberty, there was a loss in AH SDS (−0·2 SDS) compared to those treated for <2 years. Our findings show that the height SD score before the onset of puberty is not a reliable predictor of AH, and physicians should therefore be careful in extrapolating prepubertal SD scores to predict AH because this can result in an overestimation of AH.
In 30% of the short SGA children, height at the onset of puberty was below 140 cm and in these children puberty was postponed by adding 2 years of GnRHa to GH treatment. Our data demonstrate that children who received GnRHa, due to their poor AH expectation at the onset of puberty, attained a comparable AH as children who did not. This is in line with recently published data of our group investigating combined GH/GnRHa treatment. Combined GH/GnRHa treatment resulted in a longer duration of puberty and a higher pubertal height gain compared to children who received GH only, particularly in girls. This indicates a beneficial effect of GnRHa treatment, probably as a result of delayed BA development. In the present trail, BA was not determined at start of GnRHa and during GnRHa treatment, because treatment with GnRHa in addition to GH was not randomized and depended on absolute height at start of puberty. No model is known to accurately predict AH at the onset of puberty, and we used a pragmatic cut-off level; children with a height below 140 cm at the onset of puberty were not expected to reach a normal AH. Also, our present study was not designed to investigate pubertal postponement by GnRHa vs no postponement in a randomized design. Despite these limitations, our study suggests that when GH-treated short SGA children are still short at the onset of puberty, they may benefit from combined GH/GnRHa treatment.
In conclusion, GH treatment improves AH in short children born SGA, even when started after the age of 8 years. However, these results should not lead to postponement of starting GH treatment until an older age, because a younger age at start results in a better AH despite some height loss during puberty. In contrast to our hypothesis, children with greater spontaneous catch-up growth after birth show a lower total height gain SDS during GH. Furthermore, height SDS declines from mid-puberty, due to a marked early deceleration of growth velocity.
Source...