Predictors of large and small for gestational age birthweight in offspring of women with gestational diabetes mellitus

Aim To identify predictors of large and small for gestational age in women with gestational diabetes mellitus. Methods A retrospective audit of clinical data analysed for singleton births in women diagnosed with gestational diabetes by Australasian Diabetes in Pregnancy Society guidelines from 1994 to 2009. Exclusions were: incomplete data, delivered at  4 weeks pre-delivery. We assessed: pre-pregnancy BMI, ethnicity, total maternal weight gain, weight gain before and after treatment initiation for gestational diabetes, HbA1c at gestational diabetes presentation and treatment modality (diet or insulin) and smoking. Birthweight was assessed using customized percentile charts (large for gestational age > 90th; small for gestational age < 10th percentile). Multiple regression analyses were undertaken; statistical significance was p < 0.05. Results There were 1695 women first seen at (mean ± sd) 28.1 ± 5.3 weeks gestation (range 6–39). Ethnic mix was South-East Asian 36.7%, Middle Eastern 27.6%, European 22.4%, Indian/Pakistani 8.6%, Samoan 1.9%, African 1.5% and Maori 1.1%. Therapy was diet 69.1% and insulin 30.9%. Mean total weight gain was 12.3 ± 6.1 kg, the majority (10.6 ± 6.0 kg), gained before dietary intervention. There were 7.9% small for gestational age and 15.2% large for gestational age births. Significant independent large for gestational age predictors were: weight gain before intervention, pre-pregnancy BMI, weight gain after intervention and treatment type, but not HbA1c or smoking. Significant small for gestational age predictors were: weight gain before intervention, weight gain after intervention, but not pre-pregnancy BMI, HbA1c or smoking. Conclusion Conventional treatment for gestational diabetes mellitus concentrates on management of blood glucose levels. The trends identified here emphasize the need to also address pregnancy weight gain stratified by pre-pregnancy BMI.