The perinatal environment has a major influence on long-term health and disease risk. Preterm birth alters early-life environment and is associated with altered metabolic function in adulthood. Whether preterm birth per se or the early nutritional interventions used to support growth in preterm infants underpins this association is unknown. Lambs born preterm, following dexamethasone induction of labour, or spontaneously at term were randomised to receive nutrient supplementation, analogous to the milk fortifier used clinically or water as a control for the first 2 weeks after birth. Thereafter, nutrition was not different between groups. Growth was monitored, and the glucose–insulin axis function was assessed in juvenile (4 months) and adult life (14 months). Early nutrition influenced adult metabolic function and body composition to a greater extent than preterm birth. In supplemented females, arginine-stimulated insulin secretion was increased in preterm but reduced in term-born juveniles compared with controls (repeated-measures ANOVA P<0·01). In supplemented preterm males, adult weight, ponderal index (PI) and fasting insulin concentrations were elevated compared with preterm controls (weight, 75 (sem 3) v. 69 (sem 2) kg; PI, 48·0 (sem 2·1) v. 43·7 (sem 1·7) kg/m3; fasting insulin, 0·19 (sem 0·02) v. 0·10 (sem 0·02) ng/ml). Conversely, supplemented term-born males had reduced adult weight, PI and fasting insulin concentrations compared with term-born controls (weight, 64 (sem 2) v. 70 (sem 2) kg; PI, 44·4 (sem 1·8) v. 48·2 (sem 1·7) kg/m3; fasting insulin, 0·09 (sem 0·02) v. 0·14 (sem 0·02) ng/ml; all group×supplement interactions P<0·05). Adult metabolic health may reflect both gestational age at birth and early nutrition. Human studies are urgently needed to investigate the adult sex-specific health implications of neonatal nutritional strategies.

Twins achieved through in-vitro fertilisation often undergo a viability ultrasound at 6–9 weeks of gestation. The presence of inter-twin crown-rump length discordance at this stage is not an uncommon finding; however the clinical significance of this is unknown. We analyzed 218 dichorionic twin pregnancies, producing two live fetuses > 24 weeks gestation, to determine whether inter-twin discordance (≥ 85th centile) in the mid-first trimester was associated with birthweight discordance (> 20%), or small for gestational age (< 10th centile). The incidence of birthweight discordance and small for gestational age infants were determined, with no increased risk found for the discordant population. This may provide some reassurance to treating clinicians.