INTRODUCTION Fetal macrosomia is the most important complication in infants of women with diabetes, whether preconceptional or gestational. Its occurrence is related to certain maternal and fetal conditions and negatively affects maternal and perinatal outcomes. The definitive diagnosis is made at birth if a newborn weighs >4000 g.

OBJECTIVE Identify which maternal and fetal conditions could be macrosomia predictors in infants born to Cuban mothers with gestational diabetes.

METHODS A case-control study comprising 236 women with gestational diabetes who bore live infants (118 with macrosomia and 118 without) was conducted in the América Arias University Maternity Hospital, Havana, Cuba, during 2002–2012. The dependent variable was macrosomia (birth weight >4000 g). Independent maternal variables included body mass index at pregnancy onset, overweight or obesity at pregnancy onset, gestational age at diabetes diagnosis, pregnancy weight gain, glycemic control, triglycerides and cholesterol. Fetal variables examined included third-semester fetal abdominal circumference, estimated fetal weight at ≥28 weeks (absolute and percentilized by Campbell and Wilkin, and Usher and McLean curves). Chi square was used to compare continuous variables (proportions) and the student t test (X ± SD) for categorical variables, with significance threshold set at p <0.05. ORs and their 95% CIs were calculated.

RESULTS Significant differences between cases and controls were found in most variables studied, with the exception of late gestational diabetes diagnosis, total fasting cholesterol and hypercholesterolemia. The highest OR for macrosomia were for maternal hypertriglyceridemia (OR 4.80, CI 2.34–9.84), third-trimester fetal abdominal circumference >75th percentile (OR 7.54, CI 4.04–14.06), and estimated fetal weight >90th percentile by Campbell and Wilkin curves (OR 4.75, CI 1.42–15.84) and by Usher and McLean curves (OR 8.81, CI 4.25–18.26).

CONCLUSIONS Most variables assessed were predictors of macrosomia in infants of mothers with gestational diabetes. They should therefore be taken into account for future studies and for patient management. Wide confidence intervals indicate uncertainty about the magnitude of predictive power.

KEYWORDS Fetal macrosomia, fetal diseases, gestational diabetes, risk factors, risk prediction, Cuba