Fetal growth restriction is associated with an increased risk of developing insulin resistance and type 2 diabetes in adulthood. In addition, 10-20% of children born small for gestational age (SGA) do not achieve a normal final height. The purpose of this study was to investigate insulin sensitivity and endocrine status in SGA children, compared with that in children born appropriate for gestational age (AGA). Furthermore, within the SGA group, we aimed to relate postnatal growth to anthropometric, biochemical, and endocrine parameters. Eighty-two SGA children (with a mean age of 8.6 +/- 3.5 yr) and 53 short-AGA children (with a mean age of 9.3 +/- 3.3 yr) were studied. A case-control study was carried out in 26 SGA and 26 short-AGA subjects. For each SGA subject, we selected a short-AGA child matched for sex, age (within 1 yr), pubertal status, body mass index (within 0.5 kg/m(2)), and height (within 0.25 z-score). Children's statures were corrected for their midparental height, and SGA children were subdivided into 2 groups: catch-up growth (CG) group (children with corrected height with at least 0 z-score); and non-CG (NCG) group (subjects with corrected height with less than 0 z-score). Comparing SGA with short-AGA subjects, no significant differences in fasting insulin, fasting glucose/insulin ratio, homeostasis assessment model for insulin resistance, and homeostasis assessment model-beta-cell values were observed. SGA children showed significantly reduced levels of glucose (4.4 +/- 0.6 vs. 4.9 +/- 0.6 mM, P < 0.0001), total cholesterol (160.1 +/- 28.8 vs. 171.8 +/- 28.5 mg/dl, P = 0.02), and high-density-lipoprotein cholesterol (53.3 +/- 12.1 vs. 58 +/- 11.4 mg/dl, P = 0.02). The analysis of the subjects selected for the case-control study confirmed that SGA children did not have significant differences in the indices of insulin sensitivity but showed significantly lower glucose levels (4.4 +/- 0.7 vs. 4.9 +/- 0.4 mM, P < 0.005). Subdividing the SGA group into CG (n = 25) and NCG (n = 57) children, we found that NCG children showed significantly higher levels of TSH (2.5 +/- 1.3 vs. 1.9 +/- 0.6 mU/liter, P = 0.002). Our data indicate that SGA children do not have altered insulin sensitivity when compared with auxologically identical AGA subjects but show a significant reduction of glucose concentrations. Whether the lower glucose levels are attributable to an early phase of augmented insulin sensitivity, as previously reported in animal models, has to be established. The finding of higher TSH concentrations in SGA children with blunted CG suggests that intrauterine reprogramming might involve thyroid function, which, in turn, might affect postnatal growth and cholesterol metabolism, eventually increasing the risk of cardiovascular disease.