The prevalence of gestational diabetes mellitus (GDM) has been increasing over the past three decades, a large retrospective cohort study from the US has shown.

Using data from the National Hospital Discharge Survey database between 1979 and 2010, a cohort of 124,352,115 pregnant women aged 15-49 years, were identified retrospectively. Of these, 3,337,284 developed GDM (2.7%).

The prevalence of GDM over time was addressed by collating data into five-year groupings for maternal age, and infant birth period. The effect of temporal changes in GDM prevalence was calculated looking at these two variables. Demographic data were available on marital and social status, chronic and gestational hypertension, multiple pregnancy and preeclampsia. However, data on maternal obesity and smoking were not recorded. So for this part of the analysis separately collected national data were used for comparative purposes.

Between 1979-81 and 2008-10 there was a highly significant increase in prevalence of GDM, from 0.3% to 5.8% of deliveries.

The prevalence of GDM rose through successive five-year periods studied from 1980 onwards. For example, in women aged 25-29 years the prevalence rose from 0.2% of deliveries in 1980 to 6.2% in 2010.

This increase was observed across all age groups. GDM prevalence also rose with increasing maternal age through all time periods. So, for example, the GDM prevalence among 15-19 year olds in 2010 was 1.6% but for 40-44 year olds it was 14.3%.

A complex regression analysis, taking into account the incremental effect of GDM, calculated that there was a slight increase in stillbirth rates of 0.7% in women with GDM over time (i.e. 1979-81 to 2008-2010); risk ratio (RR) 0.007 (95% CI: 0.007-0.008). Over the same period, a similar analysis showed a decrease in large for gestational age/macrosomia rates in women with GDM; RR -1.84 (95% CI:-1.80 to -1.88).

‘The decreased prevalence of large for gestational age and macrosomic infants in the setting of increasing GDM rates supports current screening and diagnostic criteria,’ the authors say.

This study has its weaknesses; it had to use imputed data on maternal weight and the criteria for diagnosing GDM has changed over the past 30 years. However, the very large numbers studied, and the long time period surveyed, do lend its findings considerable weight.

What the authors were not able to do was firmly associate the rising rates of GDM with its most likely fellow traveller, rising rates of maternal obesity. However, they conclude:

‘The prevalence of GDM has been increasing in the US over the past three decades. The observed period effects are consistent with the trend towards increasing pre-pregnancy BMI.’

Dr Chris Barclay

Lavery JA, Friedman AM, Keyes KM et al. Gestational diabetes in the United States: temporal changes in prevalence rates between 1979 and 2010. BJOG 2017;124:804-13