Birth weight references for twins

OBJECTIVE: We sought to formulate fetal and birth weight references for twi ns from longitudinal data. STUDY DESIGN: This historic cohort study was based on 1831 pregnancies of t wins born alive at greater than or equal to 28 weeks' gestation from Baltim ore, Maryland; Miami, Florida; Charleston, South Carolina; and Ann Arbor, M ichigan. RESULTS: When we compared singletons and twins, the percentiles of twins fe ll substantially below the 10th percentile of singletons by 28 weeks' gesta tion, below the 50th percentile by 30 weeks' gestation, and below the 90th percentile by 34 weeks' gestation. The difference at the 50th percentile wa s 147 g (10%) at 30 weeks' gestation, 242 g (14%) at 32 weeks' gestation, 3 47 g (17%) at 34 weeks' gestation, 450 g (19%) at 36 weeks' gestation, 579 g (22%) at 38 weeks' gestation, and 772 g (27%) at 40 weeks' gestation. CONCLUSION: This new reference demonstrates that, although the overall patt ern of fetal growth is slower for twins versus singletons from about 30 wee ks' gestation, well-grown twins and singletons do not differ as much as pre viously believed.