Liu et al. respond to "isolating preterm birth to assess its impact".

We thank Drs. Savitz and Werner (1) for their thoughtful comments about isolating the causal impact of early delivery on neonatal health outcomes. In our article (2), we sought to provide an estimate of the “pure outcome” of shortened gestation resulting from the acute event of unintentional injury during pregnancy and to contrast that outcome with that of preterm birth from other (predominantly chronic) causes of preterm birth. Ultimately, we hoped to gain insight into the possible role of adaptation to a suboptimal intrauterine environment, which has been invoked as playing a role in reducing adverse sequelae among fetuses who are delivered preterm (3). We agreewith all of the comments of Savitz andWerner but wish to emphasize 2 points made in our paper on which they did not comment. First, we speculate that an acute traumatic event will often lead to rapid labor (or labor induction) that precludes sufficient time to administer a full course of antenatal corticosteroids (4).Unfortunately, our data source does not contain any information on drugs administered to the mother during labor and delivery, and thus we are unable to confirm that speculation. Second, the smaller birth weight for gestational age that we observed among non-injury-associated preterm births versus their injury-associated counterparts provides important evidence for the chronic nature of (most) non-injuryassociated preterm births.