Longitudinal Associations Between Maternal Glucose Levels and Ultrasonographic Fetal Biometrics in a Shanghai Cohort

IMPORTANCE: Evidence on the timing of fetal growth alterations associated with gestational diabetes or on the association of the maternal glycemic trajectory with fetal growth during pregnancy remains lacking. OBJECTIVE: To examine the associations between maternal glucose levels and offspring intrauterine growth. DESIGN, SETTING, AND PARTICIPANTS: This cohort study used data from 4574 eligible pregnant women and their offspring in the Shanghai Maternal-Child Pairs Cohort collected from April 10, 2016, to April 30, 2018. Group-based trajectory modeling was used to classify fasting plasma glucose levels during pregnancy into 3 glycemic trajectories (trajectory 1, consistently normal glucose levels in all 3 trimesters; trajectory 2, hyperglycemia only in late pregnancy; and trajectory 3, hyperglycemia in all 3 trimesters [ie, consistently high glucose levels]). Statistical analysis was performed from April 25, 2020, to October 1, 2021. EXPOSURES: Gestational diabetes, which was defined using the results of an oral glucose tolerance test. MAIN OUTCOMES AND MEASURES: Longitudinal fetal biometrics during gestational weeks 11 to 40 and birth outcomes were obtained from medical records. Pregnancy was partitioned into 3 periods (<24, 24-34, and >34 weeks’ gestational age). The differences in offspring growth (log-transformed) and maternal glucose levels were compared using generalized linear mixed models. RESULTS: A total of 4121 pregnant women had oral glucose tolerance test results (mean [SD] age, 28.8 [4.1] years), 3746 of whom had glycemic trajectory data (mean [SD] age, 28.6 [4.1] years); 983 women (23.8%) had gestational diabetes. Throughout the pregnancy period and compared with the women without gestational diabetes or with women in the trajectory 1 group, the fetal biometrics for the women with gestational diabetes or for those in the trajectory 3 group were significantly higher (except for biparietal diameter), with an estimated increase in fetal weight in the group with gestational diabetes (β = 1.82; 95% CI, 1.03-2.61) and in the trajectory 3 group (β = 1.50; 95% CI, 0.54-2.47; P = .002). Fetal biometric alterations among women with gestational diabetes appeared before 24 weeks’ gestational age, with neonatal birth weight significantly higher than in the group without gestational diabetes at 40.4 g (95% CI, 9.8-71.1 g) along with an increased risk of large size for gestational age (odds ratio, 1.36; 95% CI, 1.05-1.75) and macrosomia (odds ratio, 1.47; 95% CI, 1.12-1.94). However, pregnant women in the trajectory 2 group manifested significantly reduced fetal biometrics, and abdominal circumference was significantly augmented after 34 weeks’ gestational age (increase, β = 1.92; 95% CI, 0.87-2.99). CONCLUSIONS AND RELEVANCE: In this cohort study, pregnant women who received a diagnosis of gestational diabetes in midpregnancy or had hyperglycemia during all 3 trimesters showed an association with altered fetal growth patterns, including increased estimated fetal weight that appeared before 24 weeks’ gestational age, increased birth weight, and the risk for large size for gestational age and macrosomia.