Associations between estimated foetal weight discordance and clinical characteristics within dichorionic twins: The NICHD Fetal Growth Studies

Background Birthweight discordance is well studied, with less known about longitudinal inter-twin differences in foetal growth. Objective To examine inter-twin per cent differences in EFW (EFW%), head (HC%) and abdominal circumference (AC(%)), and femur length (FL%) across gestation in dichorionic twin gestations and explore associated characteristics. Methods Foetal biometrics were assessed by ultrasound and EFW calculated at <= 6 study visits among women with dichorionic twin pregnancies enrolled in the NICHD Fetal Growth Studies cohort (US, 2012-2013). Inter-twin per cent difference was defined: ([Size(larger twin) - Size(smaller twin)]/Size(larger twin) x 100). Linear mixed models evaluated per cent differences in foetal biometrics at 15 weeks and their change per week overall and by maternal/neonatal characteristics in unadjusted and adjusted models. Results In 140 pregnancies, inter-twin per cent differences increased across gestation for EFW (0.18%/week, 95% confidence interval [CI] 0.10, 0.27), HC (0.03%/week, 95% CI 0.00, 0.06), and AC (0.03%/week, 95%CI -0.01, 0.08) but decreased for FL (-0.03%/week, 95% CI -0.09, 0.02). After adjustment, change in EFW% difference across gestation differed by pre-pregnancy body mass index (BMI [kg/m(2)]; underweight [P-interaction = .022); and conception method (in vitro fertilisation [IVF], intrauterine insemination, ovulation induction medication, donor egg/embryo, none; P-interaction = .060). While EFW% difference increased with normal pre-pregnancy BMI (0.24%/week, 95% CI 0.12, 0.37), little change was noted with pre-pregnancy obesity (0.01%/week, 95% CI -0.15, 0.17). EFW% difference increased in conceptions without fertility treatments (0.23%/week, 95% CI 0.11, 0.34) but not IVF conceptions (-0.00%/week, 95% CI -0.16, 0.16). Similar patterns of differences across gestation were noted for HC% by conception method (P-interaction = .026) and AC(%) by pre-pregnancy BMI (P-interaction = .071); changes in HC% differed by parity (nulliparous, multiparous; P-interaction = .004). Conclusions EFW% difference increased across gestation in dichorionic twins, but remained stable with pre-pregnancy obesity or IVF conception, patterns mirrored for HC and AC. Research is needed to understand pathologic versus physiologic differential twin growth trajectories.