1 | INTRODUCTION
Twin pregnancies are associated with a threefold to sevenfold increase in perinatal morbidity and mortality compared to singleton pregnancies,1 mainly due to higher rates of preterm birth and discordant growth.2 Intertwin size discordance has been reported to be an independent risk factor for stillbirth and short-term adverse neonatal outcomes especially the smaller twin of birthweight discordance of 20 % or more suffering from a neurodevelopmental disadvantage.3
Twin birth weight (BW) discordance is attributable to various genetic or environmental factors containing differences in nutrient supply associated with variation in placental mass, placental insufficiency and umbilical cord insertion. These impact one twin more than the other in a sub-optimal intrauterine environment, which of twins contributed to short- 4 and long-term perinatal outcomes,5,6 especially the effects on neural development (e.g. induced neurological morbidity).7
Several studies have assessed the link between metabolic alterations with BW discordance and abnormal infant neurodevelopment. However, most have focussed on circulating factors in blood– often highly dynamic and not reflective of the cumulative time in utero .8,9 A downregulation of amino acids, including valine, tryptophan, isoleucine, and proline, occurred in the cord blood plasma of selective fetal growth restriction (FGR) in monochorionic-diamniotic (MCDA) twins.10 Disrupted essential amino acids, such as methionine, phenylalanine and tyrosine in cord plasma were correlated with phenotypic growth discordance of selective FGR.11 Gut microbial dysbiosis and variation in the faecal metabolome of neonatal twins have also been correlated with long-term neurobehavioural development in selective FGR.12
The concentration of endogenous compounds and environmental compounds in hair samples is maintained in an ordered temporal manner as hair grows and therefore the neonatal hair metabolome is considered reflective of the time in utero . Maternal exposure in the hair metabolome has been explored to reflect the lower language ability in offspring with 373 infant-mother included.13
We previously demonstrated the feasibility of using the neonatal hair metabolome to accurately generate a longitudinal picture of the intrauterine environment over pregnancy.14 However, to date, no study has attempted to identify specific metabolic differences in the newborn hair metabolome associated with discordant in utero growth in dichorionic (DC) twin pregnancies, or explored the link between metabolic variation in hair at birth with later neuro-developmental outcomes.
This study aimed to define hair metabolic perturbations associated with intrauterine growth discordance of DC twin pregnancies and to assess the link between metabolic variation in the hair at birth with infant neurodevelopment at 2-3 years of age.
2 | MATERIALS AND METHODS
2.1 | Experimental Design
This is a cohort-based case-control study, within a twin cohort recruited at the Peking University Third Hospital (Clinicaltrials.gov Identifier: NCT03220750) between September 2017 and December 2018. Chorionicity was established at booking ultrasound prior to 14 weeks’ gestation. A total of 201 DCDA twin pregnancies were recruited between 14-28 weeks of gestational age. The exclusion criteria included maternal factors, such as chronic diseases, obstetric complications and delivery complications; fetal factors included major congenital anomalies, or major fetal structural anomalies, or aneuploidy, other adverse twin pregnancy outcomes, and those lost to follow-up. DC twin neonates were classified into DCDA-C (DCDA twins with birth weight concordance) and DCDA-D (DCDA twins with birth weight discordance) according to consensus-based diagnostic criteria of a within-pair difference in BW exceeding 25%, with one twin <10th centile.15 The gross examination of placenta was executed carefully by our obstetricians immediately after delivery including discerning the intertwin membrane to confirm chorionicity (determined by antenatal ultrasound), the placental territory and cord insertions attributed to each twin. Maternal and fetal clinical characteristics were measured within 24 hours after delivery.