Glycemic profiles on the oral glucose tolerance test and birthweight-for-gestational-age distribution in pregnancies with gestational diabetes: a historical cohort study
Introduction Gestational diabetes mellitus encompasses heterogeneous oral glucose tolerance test profiles. We examined whether threshold-defined phenotypes, abnormal-value location within the same count, and continuous glucose concentrations were associated with the birthweight-for-gestational-age distribution. Material and methods This single-center historical cohort included 4013 singleton pregnancies with gestational diabetes mellitus from 3922 women at a Chinese maternity hospital (2022–2025); adjusted complete-case analyses included 3898 pregnancies from 3815 women. Exposures were threshold-defined phenotypes, exact patterns within count strata, and continuous fasting, 1-hour, and 2-hour glucose concentrations. The primary outcome was eight ordered sex- and gestational-age-specific birthweight categories; large for gestational age was secondary. Models adjusted for maternal age, prepregnancy body mass index, parity, gestational age at testing, and delivery year. Results Birthweight distributions differed across phenotypes (global P < 0.001), predominantly in the upper tail. Adjusted probabilities above the 90th percentile were 8.6% for isolated post-load hyperglycemia, 16.1% for isolated fasting hyperglycemia, and 19.7% for combined hyperglycemia. Within the one- and two-abnormal-value strata, distributions differed by abnormality location (global P < 0.001 and P = 0.010). Compared with 1-hour only, fasting only was associated with a 7.3-percentage-point higher probability above the 90th percentile (95% confidence interval 3.4–11.2); the difference between fasting-plus-1-hour and 1-hour-plus-2-hour patterns was 9.1 percentage points (95% confidence interval 2.8–16.3). In the joint continuous model, each standard-deviation increase in fasting plasma glucose was associated with a common odds ratio of 1.37 for a higher birthweight category (95% confidence interval 1.29–1.46) and a large-for-gestational-age risk ratio of 1.43 (95% confidence interval 1.32–1.56). Associations for 1-hour and 2-hour glucose were smaller. Conclusions Among pregnancies with gestational diabetes mellitus, fasting-involved profiles were associated with higher birthweight-for-gestational-age categories and greater upper-tail probabilities. Patterns with the same number of abnormal values had different distributions, indicating that count alone incompletely characterizes fetal-size heterogeneity. These findings support preserving component-specific oral glucose tolerance test information but do not establish individual predictive utility or justify phenotype-directed management.
Authors
- Tianyi Wei
- Mengdi Xue
- Zhifa Jiang (ORCID: https://orcid.org/0000-0002-1817-6413)
- Zhongjun Li
- Xiangyun Ye
- Xiaojing Lin
Institutions
- Jinan University (CN)
- Dongguan People’s Hospital (CN)
- Southern Medical University (CN)
Publication Details
- Journal
- Frontiers in Endocrinology
- Published
- 2026-11-19
- DOI
- https://doi.org/10.3389/fendo.2026.2016532
- Primary Topic
- Gestational Diabetes Research and Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00