NEW ORLEANS — The proportion of gestational diabetes cases attributable to overweight and obesity totaled 46% in a population-based study of more than 20,000 women from seven U.S. states.
The data, from the Centers for Disease Control and Prevention's Pregnancy Risk Assessment Monitoring System (PRAMS), were used to generate a population-based estimate of the contribution of prepregnancy overweight and obesity to the development of gestational diabetes mellitus (GDM). Shin Y. Kim reported the results at the annual scientific sessions of the American Diabetes Association.
“If we assume that the relationship between GDM and obesity and overweight is causal and no other confounders exist, then a large proportion of GDM cases are potentially preventable,” said Dr. Kim of the CDC's division of reproductive health.
She and her associates analyzed PRAMS data from the seven states that had implemented the 2003 revised birth certificate, which distinguishes GDM from diabetes that existed prior to pregnancy. The surveillance system collects data via a questionnaire from mothers of newborns 2-6 months after delivery. A total of 22,767 women with complete chart information who did not have pre-existing diabetes were included.
The overall GDM prevalence was 4%, ranging from 3.1% in Florida to 5% in Ohio. (The other five states were Nebraska, South Carolina, Utah, Washington, and New York, excluding New York City.) More than 70% of the women with GDM had a prepregnancy body mass index of at least 25 kg/m2, compared with 44.9% of the women who did not have GDM during pregnancy, Ms. Kim reported.
The GDM prevalence was 0.7% for women classified as underweight (body mass index 13-18.4 kg/m2) prior to pregnancy, 2.3% for those with normal weight (BMI 18.5-24.9), 4.8% for overweight women (25-29.9), 5.5% for those who were obese (30-34.9), and 11.5% for extremely obese women (35-64.9). With normal weight used as the reference group, the unadjusted relative risks of developing GDM were 2.1, 2.4, and 5 for women who were overweight, obese, and extremely obese, respectively.
“The probability of GDM increases with increasing BMI, with no clear BMI threshold below which a dose-response relationship was not evident,” Ms. Kim said.
The relative risks did not change after adjustment for maternal age, race/ethnicity, marital status, or parity. Once adjusted, the proportions of gestational diabetes cases attributable to overweight, obesity, and extreme obesity were 15.4%, 9.7%, and 21.1%, for a total of 46.2%. “In other words, if all women with a BMI of 25 or greater had a GDM risk equal to that of women in the normal BMI category, nearly half of GDM cases could be prevented. Lifestyle interventions to reduce BMI have the potential to lower GDM risk,” she commented.
There are a few possible reasons for why overweight/obesity contributed to only about half of GDM cases, Ms. Kim said in a follow-up interview.
“First, prepregnancy weight was self-reported, and women tend to underreport their weight. This may have led us to underestimate the contribution of overweight and obesity to the fraction of GDM attributable to weight. Also, there may be a race/ethnic difference in the relationship between BMI and GDM risk, and our analysis overrepresents non-Hispanic white women compared to the general population,” she noted.
If the analyses had been done using data representing the entire U.S. population, she continued, the study might have generated a larger estimate of the proportion of GDM cases associated with a high BMI. “Physical activity also contributes to GDM risk, and we had no data on physical activity levels in our study population. In addition, BMI is not a perfect measure of body fat, but we use it often because it can easily be obtained. If we had used lean women as our reference group, the [attributable proportion] would have been much higher. This is because the GDM risk increased in a nearly linear fashion as BMI increased.”
However, she said, the 46% estimate from this study is “higher than other non-population-based estimates found in the literature, and the dose-response relationship is consistent with estimates found in the general population with type 2 diabetes.”
Ms. Kim stated that she had no disclosures to make.