according to a case-control study drawn from the prospective NICHD Fetal Growth Studies-Singleton cohort.
Women who went on to develop GDM had higher HbA1c levels, and measuring this factor improved prediction when added to traditional GDM risk factors, Stefanie N. Hinkle, PhD, of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) and her associates reported in.
showed that GDM-associated fetal overgrowth begins before GDM diagnosis, which suggests a need to identify GDM earlier in pregnancy (Int J Epidemiol. 2018 Feb;47:25–25l).
HbA1c is already used to screen for type 2 diabetes mellitus outside of pregnancy. Previous studies that examined its potential utility for GDM have focused on high-risk patients, examined an HbA1c threshold of 5.7%, used GDM during the first trimester only as the outcome, or had other limitations. There is little research on HbA1c levels and GDM in population-based samples.
Dr. Hinkle and her associates conducted a secondary analysis ofthat involved 2,334 low-risk pregnancies among nonobese women and 468 low-risk pregnancies among obese women (n = 2,802) at 12 U.S. centers. The women were recruited during gestational weeks 8-13 and then followed until the end of the pregnancy. The researchers did a nested GDM case-control study of 107 GDM cases and 214 matched non-GDM controls.
GDM cases had higher HbA1c levels throughout their pregnancies (P less than .03). The researchers found a linear association between HbA1c at enrollment and GDM risk (P = .001). Women with a first trimester HbA1c level of 5.7% had an odds ratio of 2.73 for GDM, compared with women at the median level of 5.2%.
In the adjusted model, for each increment of 0.1% at enrollment, women had an OR of 1.22 for GDM (P less than .001). For every 0.1% difference between HbA1c levels at enrollment and the second visit (24-29 weeks), the OR was 1.21 (P = .04). When the researchers excluded women who were obese, had smoked, had prior GDM, or had a hematologic disorder, the OR per 0.1% increase was similar (OR, 1.23; 95% confidence interval, 1.10-1.38).
A potential optimal cutoff point is 5.1%, which had a sensitivity of 47% (95% CI, 34%-60%) and a specificity of 79% (95% CI, 62%-88%). At 5.7%, which is used as the cutoff for prediabetes in nonpregnant women, the sensitivity was 21% (95% CI, 8%-36%) and the specificity was 95% (95% CI, 91%-99%).
When the model was added to conventional risk factors such as age, race/ethnicity, being overweight or obese before pregnancy, family history of diabetes, previous GDM, and nulliparity, the area under the curve of HbA1c levels at enrollment increased from 0.59 to 0.65.
Robert Atlas, MD, chair of obstetrics and gynecology at Mercy Medical Center, Baltimore, said in an interview, “This is just the first study that needs to be replicated in different patient populations. No one has looked at a continuum of HbA1c and what value above puts you at an increased risk. I think this is a very powerful study that sets the stage for further investigation into how to utilize HbA1c in a better way than we’ve ever looked at it before.”
The study was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development intramural funding and by American Recovery and Reinvestment Act funding. Dr. Hinkle and her associates had no relevant financial disclosures. Dr. Atlas had no relevant financial disclosures.
SOURCE: Hinkle SN et al. .