direct causal relationship between GWG and preterm delivery guided the committee’s approach to decision analysis in Chapter 7, which weighed the trade-offs of GWG with and without taking into account preterm delivery as an outcome.
The IOM (1990) report recommendations for GWG focused largely on avoiding inadequate GWG and the short-term consequences of low fetal growth and prematurity (see Chapter 1). Since that time, the emergence of epidemic obesity in the U.S. population has raised the possibility that excessive weight gain may also be harmful. A small number of recent studies have addressed the relationship between GWG and adiposity at birth, markers of childhood obesity and cardiometabolic sequelae of childhood obesity. The following discussion summarizes the committee’s review of the evidence for associations between GWG and neonatal body composition, infant weight gain, breastfeeding initiation, and other long-term outcomes.
As previously explained (see Fetal Growth section in this chapter), GWG is directly associated with fetal growth as measured by birth weight for gestational age. For long-term adiposity-related outcomes, however, it is important to measure not only weight (and length) at birth but also body composition. As mentioned in the chapter introduction, it has been hypothesized that relative amounts of adiposity and lean mass in fetal and neonatal life are important in setting long-term cardio-metabolic trajectories. Catalano and colleagues performed a series of studies examining the relationships between various maternal characteristics and neonatal body composition as measured by total body electrical conductivity (a method no longer in use). One set of studies compared infants who were born at term to overweight/obese women (pregravid BMI > 25 kg/m2; n = 76) with those born to lean/average weight women (n = 144) (Sewell et al., 2006). As expected, weight gain was higher among lean/average (mean 15.2 kg) than overweight/obese (13.8 kg) women. Among the overweight/obese women, stepwise regression analyses that included pregravid weight as a covariate revealed that the higher the GWG, the more the newborn fat mass. The authors did not report a correlation among the lean women, presumably because the associated p-value exceeded 0.05. In another study, which combined data from diabetic and nondiabetic pregnant women (total n = 415), GWG was directly associated with both lean and fat mass at birth (Catalano and Ehrenberg, 2006). The latter results are consistent with those of Udal et al. (1978), who found a direct association between GWG and the sum of 8 neonatal skinfold measurements among 109 nondiabetic mothers