Introduction
Small for gestational age (SGA) defines infants whose birth weight or crown-heel length is 2 standard deviations (SDs) below the mean for their appropriate gestational age (AGA). Approximately 3-5% of neonates are born SGA. This cohort is estimated to account for 1 in 5 children (and adults) who are short in stature. SGA is associated with poor outcomes in terms of growth and metabolism. Understanding population distributions of growth outcomes will aid consideration of treatments such as growth hormone in a paediatric population. Data from the ‘Growing Up in Ireland’ (GUI) longitudinal study is explored.

Materials and MethodsBirth height was standardised according to gestational age and sex. Childhood growth measures were standardised for sex. Distributions of outcome variables were examined using robust graphical methods. Univariate poisson regression was employed to examine the relative risk of SGA on quintiles of growth outcomes at 9 months, 3 and 5 years of age. A sex-adjusted height below two standard deviations was identified as a significant cut-score from the literature. Univariate logistic regression was employed to examine the likelihood of this outcome in the SGA group. R statistical programming was used for analysis.
ResultsThe GUI cohort consisted of 12,121 children in Ireland, 56% of whom were female. Those born SGA were 29%, 16% and 12% more likely to fall in a height decile below that of their average for gestational age peers at 9 months, 3 and 5 years respectively (significant at an alpha level of 0.95). At 5 years of age the SGA group demonstrated a likelihood of falling 2 SD below the sex-adjusted mean, five-fold over their peers (OR: 5.40, 95% CI: 2.86, 10.20). Quantile analysis was undertaken and graphical visualisations were plotted to aid interpretation of growth outcomes for the SGA group in this dataset. A sankey diagram (figure 1) is used to track the distribution of participants with SGA across quintiles of height through 5 years of age.

ConclusionSmall for Gestational Age neonates represent between 3 to 5% of a given population. In Ireland, data from the GUI study demonstrated a 3.5% prevalence. This analysis suggest that subgroups of SGA infants may exhibit worse growth outcomes than their population matched peers. Future analysis may employ machine learning techniques at a population level to identify those who may benefit from growth hormone therapy.

Introduction
Small for gestational age (SGA) defines infants whose birth weight or crown-heel length is 2 standard deviations (SDs) below the mean for their appropriate gestational age (AGA). Approximately 3-5% of neonates are born SGA. This cohort is estimated to account for 1 in 5 children (and adults) who are short in stature. SGA is associated with poor outcomes in terms of growth and metabolism. Understanding population distributions of growth outcomes will aid consideration of treatments such as growth hormone in a paediatric population. Data from the ‘Growing Up in Ireland’ (GUI) longitudinal study is explored.

Materials and MethodsBirth height was standardised according to gestational age and sex. Childhood growth measures were standardised for sex. Distributions of outcome variables were examined using robust graphical methods. Univariate poisson regression was employed to examine the relative risk of SGA on quintiles of growth outcomes at 9 months, 3 and 5 years of age. A sex-adjusted height below two standard deviations was identified as a significant cut-score from the literature. Univariate logistic regression was employed to examine the likelihood of this outcome in the SGA group. R statistical programming was used for analysis.
ResultsThe GUI cohort consisted of 12,121 children in Ireland, 56% of whom were female. Those born SGA were 29%, 16% and 12% more likely to fall in a height decile below that of their average for gestational age peers at 9 months, 3 and 5 years respectively (significant at an alpha level of 0.95). At 5 years of age the SGA group demonstrated a likelihood of falling 2 SD below the sex-adjusted mean, five-fold over their peers (OR: 5.40, 95% CI: 2.86, 10.20). Quantile analysis was undertaken and graphical visualisations were plotted to aid interpretation of growth outcomes for the SGA group in this dataset. A sankey diagram (figure 1) is used to track the distribution of participants with SGA across quintiles of height through 5 years of age.

ConclusionSmall for Gestational Age neonates represent between 3 to 5% of a given population. In Ireland, data from the GUI study demonstrated a 3.5% prevalence. This analysis suggest that subgroups of SGA infants may exhibit worse growth outcomes than their population matched peers. Future analysis may employ machine learning techniques at a population level to identify those who may benefit from growth hormone therapy.