Aims/hypothesis: To investigate the incidence of foetal macrosomia (i.e. birth weight >90th percentile) in a non-selected nationwide cohort of women with Type I (insulin-dependent) diabetes mellitus in The Netherlands and to identify risk indicators predictive for macrosomia.
Methods: We conducted a prospective nationwide cohort based survey regarding the outcome of Type I diabetic pregnancy in The Netherlands. Data of 289 women who gave birth to a live singleton infant without major congenital malformations at more than or equal to 28 weeks of gestation are shown.
Results: The incidence of foetal macrosomia was very high (48.8%), with 26.6% of infants weighing more than 97.7th percentile. Glycaemic control during pregnancy was good (i.e. mean HbA(1c) <or=7.0%), in almost all (84%) women. Multiple logistic regression analysis resulted in a predictive model for macrosomia that incorporated five variables: third trimester HbA(1c) (Odds Ratio [95% Confidence Interval]: (1.6[1.1-2.4]), absence of third trimester severe hypoglycaemia (3.0[1.2-7.3]), the use of insulin lispro (3.1[0.9-10.4]), weight gain during pregnancy (1.1[1.0-1.2]) and non-smoking (2.8[0.9-9.3]). Third trimester HbA(1c) was the most powerful predictor for the occurrence of macrosomia, but its predictive capacity was weak (explained variance <5%).
Conclusion/interpretation: Despite apparent good glycaemic control, the incidence of foetal macrosomia in this non-selected prospective nationwide cohort of 289 Type I diabetic women was very high. Third trimester HbA(1c) was the most powerful predictor, but its predictive capacity was weak. Thus, future research should focus on new more detailed glucose monitoring techniques (such as a continuous glucose monitoring system) as well as to alternative factors to reduce macrosomia.