Aims/hypothesis: To investigate clinical and sociodemographic predictors of birthweight in singletons born to women with type 1 or type 2 diabetes.
Methods: Normally formed singleton live births and intrapartum stillbirths, born to women with pre-conception diabetes during 1996-2008, were identified from the population-based Northern Diabetes in Pregnancy Survey (n = 1,505). Associations between potential predictors and birthweight were analysed by multiple regression.
Results: Potentially modifiable independent predictors of increase in birthweight were pre-pregnancy care (adjusted regression coefficient [b] = 87.1 g; 95% CI 12.9, 161.3), increasing third-trimester HbA(1c) ≤7% (53 mmol/mol) (b = 310.5 g per 1% [11 mmol/mol]; 95% CI 246.3, 374.7) and increasing maternal BMI (b = 9.5 g per 1 kg/m(2); 95% CI 3.5, 15.5). Smoking during pregnancy (b = -145.1 g; 95% CI -231.4, -58.8), later gestation at first antenatal visit (b = -15.0 g; 95% CI -26.9, -3.0) and higher peri-conception HbA(1c) (b = -48.2 g; 95% CI -68.8, -27.6) were independently associated with birthweight reduction. Pre-pregnancy nephropathy (b = -282.7 g; 95% CI -461.8, -103.6) and retinopathy (b = -175.5 g; 95% CI -269.9, -81.0) were independent non-modifiable predictors of reduced birthweight, while greater maternal height was a non-modifiable predictor of increasing birthweight (b = 17.8 g; 95% CI 12.3, 23.2). Other predictors of birthweight increase were male sex, multiparity and increasing gestational age at delivery. Type or duration of diabetes, socioeconomic status and ethnicity were not associated with continuous birthweight.
Conclusions/interpretation: Poor glycaemic control before and throughout pregnancy is associated with abnormal fetal growth, with increasing peri-conception HbA(1c) predicting weight reduction and increasing third-trimester HbA(1c) predicting increased birthweight. Women with microvascular complications of diabetes may require increased surveillance to detect fetal growth restriction.