Abstract
In this issue of the Journal, Cheung et al. (1) present data on the perinatal mortality of twins and singletons from Sweden. A familiar picture of intersecting mortality curves by absolute gestational age emerges. For lower gestational ages, twins have lower mortality than do singletons but for higher gestational ages, twins have higher mortality. The picture persists after adjustment for known risk factors for low gestational age. The authors conclude that twins are initially healthier than singletons, and that they do not benefit as much as singletons do from longer gestational duration. Intersecting mortality curves are often seen when standardizing for gestational age or birth weight. To conclude that twins are initially healthier is to base a longitudinal interpretation on cross-sectional data. Things may not be as they appear.