Response to Comment on: Hernandez et al. Patterns of Glycemia in Normal Pregnancy: Should the Current Therapeutic Targets Be Challenged? Diabetes Care 2011;34:1660–1668

We appreciate Ryan's (1) thoughtful comments. The intent of our pooled analysis was to describe a normal pattern of glycemia in nonobese pregnant women according to the published data (2). We demonstrated that, similar to the Hyperglycemia and Adverse Pregnancy Outcomes (HAPO) study (3), glucose concentrations in normal pregnancy are lower than were previously appreciated. If the treatment goal for pregnancies affected by diabetes is to mimic patterns of glycemia in normal pregnancy, then a contemporary understanding of normo- or euglycemia is warranted. On the basis of our review, we suggested that prospective trials test whether 1- and 2-h postprandial (PP) targets of <122 and <110 mg/dL can attenuate incidence of fetal macrosomia and large-for-gestational-age (LGA) infants.

Ryan's concern that fasting blood glucose (FBG) was 10 mg/dL lower in our pooled analysis compared with the HAPO trial (70.9 ± 7.8 vs. 80.9 ± 6.9 mg/dL) (2,3) is well-taken. It is possible that the FBG was higher in HAPO because the women were overweight (mean BMI 27.7 ± 5 kg/m²) in comparison with a mostly normal-weight population in our pooled analysis. Our own recent data (4) showed a mean plasma FBG at 28 weeks gestation of 76 ± 3 mg/dL in normal-weight women. However, the pooled data are simply descriptive means. We do not suggest that the lower FBG target ± 1 SD be tested in an interventional trial because HAPO has already determined, based on outcome data, that a mean FBG ≥92 mg/dL confers a 1.75-fold increased risk of LGA. Because we do not have such powerful outcome data for 1- and 2-h PP therapeutic targets, we recommend that PP targets be prospectively tested.

Ryan argued against testing lower PP targets because the Maternal-Fetal Medicine Units Network trial (5) already demonstrated that adopting conventional targets to treat gestational diabetes resulted in a low frequency (7.1%) of LGA. However, it is important to recall that that trial excluded women with fasting hyperglycemia (≥95 mg/dL), a strong predictor of LGA. Further, the majority of women (426/462) attained a median 2-h PP glucose in the range of 97.3–107.8 mg/dL across meals, which met our suggested lower target of <110 mg/dL, potentially accounting for the low frequency of LGA.

We acknowledge the concern for an increased risk of small-for-gestational-age (SGA) if lower PP targets are adopted. However, in the context of obesity and gestational diabetes, SGA is less likely to occur when compared with lean women with type 1 diabetes who are at higher risk for placental insufficiency. The data of Langer and Mazze (6) suggested that mean 24-h glucose between 87–104 mg/dL minimizes SGA and LGA incidence. Our pooled mean 24-h glucose + 1 SD of 98 mg/dL supports this range.

We acknowledge that adopting lower therapeutic glucose targets will not guarantee lower macrosomia/LGA rates and that other nutrients, such as maternal triglycerides, likely contribute to excess fetal fat accretion, as supported by our data (4). To be clear, we do not suggest that lower PP targets be adopted based on our review. However, we do suggest that lower targets be prospectively tested to appropriately determine whether they might decrease the rate of LGA infants and, more importantly, excess neonatal adiposity—an even stronger predictor of childhood obesity.

• © 2011 by the American Diabetes Association.