Objective: We tested the hypothesis that gestational hypertension may have a more benign effect on neonatal outcomes in twin compared with singleton pregnancies, because the elevated blood pressure in twin pregnancies may partly or merely reflect the extra demand for blood supply.
Methods: A retrospective cohort study of 102,988 twin and 5,523,797 singleton live births using the U.S. birth cohort linked birth and infant death data sets, 1998-2000. Main outcomes are relative risks (RRs) of adverse neonatal outcomes: preterm birth, intrauterine growth restriction (less than the third percentile), low 5-minute Apgar score (less than 4), and neonatal death comparing gestational hypertensive with no-event healthy pregnancies for twins and singletons.
Results: For singletons, crude RRs (95% confidence intervals) comparing gestational hypertensive with healthy pregnancies were 2.23 (2.20-2.25) for preterm birth (17.4 compared with 7.8%), 2.49 (2.45-2.53) for intrauterine growth restriction (7.4 compared with 3.0%), 1.33 (1.21-1.45) for low 5-minute Apgar score (2.6 compared with 2.0 per 1,000), and 1.07 (0.96-1.19) for neonatal death (1.9 compared with 1.8 per 1,000), respectively. For twins, the corresponding RRs were much lower or showed reversed associations: 1.21 (1.19-1.24) (63.6 compared with 52.4%), 1.04 (0.98-1.11) (16.4 compared with 16.4%), 0.32 (0.23-0.46) (4.1 compared with 12.7 per 1,000), and 0.21 (0.14-0.30) (3.6 compared with 17.2 per 1,000), respectively. The adjusted odds ratios showed a similar risk pattern in twin compared with singleton pregnancies after controlling for maternal race, age, education, marital status, parity, smoking, alcohol use, perinatal care use, and mode of delivery.
Conclusion: Gestational hypertension has a much more benign effect on neonatal outcomes in twin compared with singleton pregnancies. There might be a need for twin- or multiple fetus-specific recommendations for hypertension management in pregnancy, but further interventional studies are needed to test the hypothesis.
Level of evidence: II-2.