creased in the third trimester. When adjusted for fat-free mass (FFM), TEE decreased in all BMI groups toward the end of gestation. Using multiple regression analysis, the change in TEE throughout the course of gestation was related to prepregnancy BMI and percent body fat as well as weight gain and increase in FFM. These variables accounted for 33 percent of the variance in 24-hour TEE, primarily from change in BMR. Physical activity accounted for very little net increase in TEE and actually decreased in all groups with advancing gestation.
As discussed in Chapter 3, there is wide variation in maternal metabolic response to pregnancy. Maternal pregravid insulin sensitivity may vary up to two- to three-fold, depending on factors such as obesity, level of fitness, and genetic make-up. Over the course of pregnancy a 40-60 percent decrease in insulin sensitivity occurs, depending on pregravid metabolic status (Catalano et al., 1993, 1999). For example, a 50 percent decrease in insulin sensitivity in both a thin athletic woman and an obese sedentary woman with type 2 diabetes may represent a two-fold or greater quantitative change in insulin sensitivity between them by the end of gestation. In the last 12 weeks of pregnancy, when fetal weight increases on the average from 1.0 kg to 3.5 kg, decreased insulin sensitivity increases the availability of energy to support fetal growth (Hytten and Chamberlain, 1991).
Although these changes in insulin sensitivity occur over a matter of months, compared to years in nonpregnant individuals, the same physiological associations detected in some large epidemiological studies among nonpregnant individuals may exist during pregnancy as well. For example, Swinburn et al. (1991), in a 3.5-year longitudinal study, showed that Pima Indians who were insulin resistant (measured using the euglycemic clamp technique) gained less weight than individuals who were insulin sensitive (3.1 versus 7.6 kg, p = 0.0001). The percent change in weight per year was correlated with glucose utilization (r = 0.34, p = 0.0001). The same could be true of obese vs. non-obese women. In fact, there is some preliminary data showing that, at least in early pregnancy, changes in maternal BMR and fat accretion are inversely related to the changes in insulin sensitivity in a small number of subjects (Catalano et al., 1998). Whether increased energy intake in obese insulin-resistant women during pregnancy has a greater effect on maternal and fetal fat accretion than in non-obese women remains to be determined.
Although there are no direct mechanistic effects relating leptin and adiponectin to GWG, both adipocytokines have been correlated with various