In this retrospective study, we analyzed data on 3250 children over 20,120 born between 2004 and 2014 in the province of Trieste, with a 4-year auxological follow-up by primary care pediatricians. To our knowledge, this is the first study in Italy that comprehensively considered the prevalence of short stature in children born SGA and the first in Europe, which considered SGA children who qualify for rhGH treatment at 4 years of age.
The SGA prevalence in our study (3.6%) was similar in previous studies in Finland (3.1%)  and Japan (3.5%) , lower than that reported in Sweden (5.5%)  and in a previous Italian study in the Chieti province (14.5%) in which – however – a different definition of SGA was used (<− 1.28 SDS) .
Catch-up growth was recorded earlier in our sample compared to previous reports: at the age of 1 year, 91% of children born SGA had a height > − 2 SDS (vs. 68% in Japan  and 87% in Sweden ), at the age of 2 the rate was 94% (vs. 89% in Japan and 87% in Sweden), at the age of 4 the rate was 97% (vs. 88% in Japan and 92% in Sweden at the age of 5); term babies had catch-up growth earlier than pre-term, as reported in previous studies [20,21,22]. These data seem to justify the European indication for waiting up to the age of 4 years, since there is a slight possibility of spontaneous catch-up growth between 2 and 4 years of age, not only in children born prematurely, as previously reported [23, 24], but – according to our data – also in those born at term. While it is well known that SGA infants differ in postnatal growth, the complex mechanism underlying is difficult to unveil. Different anomalies in the GH-IGF1 axis had been described [25, 26], although circulating concentrations of GH, IGF-I, IGF-BP3 were found not to be predictive of subsequent growth . MicroRNAs have been described as novel biomarkers for the early identification of catch-up growth in SGA infants, in particular miR-576-5p – with whom insulin, IGF-1, PDGFR-B, and mTOR signaling pathways are associated – seems to contribute to the regulation of postnatal growth and potentially influence the risk for cardiometabolic diseases associated in SGA children . An ultrasound parameter, the evaluation of neonatal bone maturation by studying Béclard’s nucleus - has been recently found to be a predictive factor of SGA height gain during the first year of life .
Remarkably, the median height of children born SGA at 4 years of age in our cohort was lower than the general population (− 0.52 SDS), and children born SGA-WL had a median height (− 0.73 SDS) that was lower compared to SGA-L and SGA-W (− 0.54 and − 0.23 SDS, respectively), as previously reported , although this difference did not reach statistical significance.
The prevalence in the general pediatric population of children born SGA who qualify for rhGH treatment at 4 years of age in our cohort was 1:3250, smaller than the hypothetical one (1:417)  and than previously reported (1:1800 in Japan at 3 years of age) , but still very far from the stated prevalence of children treated with rhGH with SGA indication in Italy (0.37/100,000) . Intriguingly, all children with short stature at the age of 4 years were born at term, and if we had considered the age of 2 years (as in the USA) or 3 years (as in Japan), no children would have qualified for rhGH treatment, considering all the remaining criteria. A prolonged follow-up in children born SGA is crucial throughout all infancy, even after the age of 2 years and in term babies. Recent studies confirmed that monogenic conditions (mostly related to the growth plate) play an important role in short stature, especially among SGA children [30, 31]. For instance, mutations in the IHH gene are associated in children born SGA with short stature (with nonspecific skeletal abnormalities) that may appear later in infancy; these children seem to have a good response to the rhGH treatment . Genetics studies will probably answer the question related to the highly variable response to rhGH therapy in SGA children , and other short stature conditions.
This study has some limitations, mainly due to its retrospective design. Anthropometric parameters were assessed by different physicians and not standardized; 13/20 primary care pediatricians could not be involved, leading to an analyzed sample which is not completely representative of the entire population of Trieste province; since data were anonymized, we could not retrieve any information about concomitant diseases that can interfere with growth (such as suspected or recognized genetic syndromes) or other additional data (such as ethnicity of the enrolled children).
Nevertheless, this study has a sample size comparable to that of previous studies on this topic, reflects real-life clinical practice and provides data that were not reported before and useful in clinical practice (follow-up in SGA children) and in guiding public health interventions (prevalence of children who qualify for the rhGH treatment according to SGA indication).