weight gain predicts SGA. Both LGA and SGA are themselves markers of neonatal morbidity. The literature on preterm birth is more ambiguous because of a less-extensive body of epidemiologic evidence, a nonlinear (U-shape) relationship between GWG and preterm birth that is modest in magnitude, and uncertainty about biologic mechanisms. Even when GWG is measured in a way that takes into account the shortened duration of pregnancy associated with GWG with preterm births, results are subject to some uncertainty. The U-shaped association of GWG with preterm birth is harder to interpret than the monotonic dose-response gradient with birth weight for gestational age, postpartum weight retention, and childhood obesity and may reflect distinct causal processes on the low and the high end of GWG.

Among the most important long-term child outcomes are obesity and its sequelae, chiefly cardio-metabolic consequences and neurodevelopmental disorders. Observational evidence is growing that GWG predicts childhood adiposity after adjusting for prepregnancy BMI and other key factors, although some older studies do not show this association. The evidence for neurodevelopmental outcomes is dependent on inferences from intermediate endpoints of fetal size and duration of gestation.

The need for randomized trials is especially important because, to date, there is no appropriate animal experimental model for GWG, thus making it difficult to meet one of the criteria—biological plausibility—that epidemiologists often use to support causal inference. Nevertheless, as reviewed in this chapter, pathways involving insulin resistance and fetal hyperglycemia may underlie associations of GWG with both fetal growth and subsequent obesity in the offspring. At the other end of the spectrum, reduced GWG is associated with lower fetal growth and preterm birth, which themselves are associated with later central obesity, insulin resistance, and metabolic syndrome. Very little current evidence, however, suggests that inadequate or low GWG predicts obesity-related outcomes in children.

1. Causal inferences relating GWG to childhood outcomes are tenuous as a result of the paucity of experimental studies.

2. Epidemiologic support for an association between gestational weight gain and stillbirth is weak; there are few methodologically sound studies.

3. Many epidemiologic studies are consistent in showing a linear, direct relationship between GWG and birth weight for gestational age. Thus, lower GWG predicts SGA, and higher GWG predicts