In model 2, inadequate GWG remained
significant after controlling for maternal and
infant characteristics. The risk of infant mortality
caused by inadequate GWG was reduced
57% after adding controls to the model.
Birth weight and gestational age were added
to model 3. Infants born to women with inadequate
GWG had odds of infant death that
were 48% higher than the odds for the reference
group of infants. The risk of infant
mortality caused by inadequate GWG was
reduced 29% after adding gestational age and
birth weight. Inadequate GWG remained statistically
significant in all models.