Objectives: This study assessed the effects of maternal smoking on birth outcomes among singletons and twins.
Methods: An algorithm was developed to link twins with their siblings in the 1995 Perinatal Mortality Data Set. A random-effects logistic regression model was then used to estimate the association between maternal smoking and several adverse outcomes for a random sample of singletons and for all twins with available maternal smoking information.
Results: The algorithm successfully linked sibling pairs for 91% of the twin sample. Maternal smoking was associated with a significantly increased risk of low birthweight, very low birthweight, and gestation of less than 33 weeks for both singletons and twins and with an increased risk of gestation of less than 38 weeks, infant mortality, and placental abruption for singletons. Among smokers, negative impacts on the risk of low birthweight, very low birthweight, and extreme premature delivery were significantly higher for women carrying twins.
Conclusions: Some of the negative effects of smoking on low birthweight and preterm delivery are greater for twins than for singletons. Women carrying twins should be warned that smoking increases their already high risk of serious infant health problems.