Comment on: Hernandez et al. Patterns of glycemia in normal pregnancy: should the current therapeutic targets be challenged? Diabetes Care 2011;34:1660-1668.

The article by Hernandez et al. (1) makes a useful contribution to documenting normoglycemia in pregnancy. Although the current target glucose levels in managing gestational diabetes (GDM) are for the most part based on case-control studies, the author's statement that macrosomia still occurs if these targets are met is belied by the Maternal-Fetal Medicine Units Network report of a 7.1% large-for-gestational-age (LGA) rate when these targets are attained (2). Two more important concerns exist: the veracity of the data and using 1 SD above a weighted mean as a proposed target for glycemic control in GDM.

Based on their review of data from 255 pregnant women, the weighted mean fasting blood glucose was 70.9 ± 7.8 mg/dL, yet the Hyperglycemia and Adverse Pregnancy Outcomes study with 23,316 women reported a mean ± SD fasting plasma glucose of 80.9 ± 6.9 mg/dL. This 10 mg/dL difference is not explained in the article and raises concerns about the reliability of the data from the 12 studies with small numbers used versus data from over 23,000 women—at least for the fasting glucose level. It is of note that when it comes to picking a fasting target glucose, the authors use the International Association of Diabetes in Pregnancy Study Groups diagnostic threshold of 92 mg/dL as a therapeutic target, 13 milligrams above their logical level of 79 mg/dL (weighted mean +1 SD) if they wish to be consistent.

The authors dismiss the use of +2 SD as target values associated with normalcy with suggestions for an upper limit target of weighted mean +1 SD in the possible hope that this will result in lower rates of macrosomia. Such a move will likely increase the small-for-gestational-age (SGA) rate (3) with no guarantees that the LGA rate will decline; the long-term consequences of SGA may well be worse than LGA (4). The authors rightly point out that factors other than glucose may drive LGA, and it is noteworthy that over time controlling glucose has had positive results in terms of perinatal mortality, but the lack of impact on LGA is striking (5). Perhaps such a focus on glucose in GDM will have negative outcomes.

I would certainly concur with the authors’ call for prospective studies to specifically test therapeutic targets in GDM, but changing target levels of glycemia to those suggested by the authors, 122 and 110 mg/dL at 1 and 2 h postprandially, based on data that provided a fasting glucose that is inconsistent with that found in the general pregnant population and using a 1-SD cutoff is problematic. If the treatment studies reported to date can normalize LGA rates (2) using targets of 95, 140, and 120 mg/dL for fasting, 1 h, and 2 h postprandially, respectively, it should take a prospective intervention study to provoke a change. Rather than normoglycemia, maybe we should focus on euglycemia in its true etymological meaning: good glucose.