TL;DR
This large-scale Chinese cohort study of nearly 5 million women reveals that both pre-conception maternal Body Mass Index (BMI) and Fasting Plasma Glucose (FPG) are independently associated with an increased risk of preterm birth (PTB). Crucially, the study highlights a synergistic effect: the combination of abnormal BMI and FPG significantly amplifies PTB risk, suggesting that optimizing both weight and glucose levels before pregnancy is essential for better birth outcomes.
What Problem Does This Paper Address?
Preterm birth (PTB) remains a leading cause of infant mortality and morbidity worldwide. Identifying modifiable risk factors is critical for effective prevention strategies. While maternal pre-pregnancy Body Mass Index (BMI) and blood glucose levels (like Fasting Plasma Glucose, FPG) have been individually linked to adverse pregnancy outcomes, the precise nature of these associations, especially their combined effects and dose-response relationships, remains an area needing deeper exploration. Specifically, previous research on BMI and PTB has yielded conflicting results, and the impact of pre-pregnancy glucose levels, particularly in combination with BMI, has not been thoroughly investigated in large populations. This study aims to fill this gap by examining the independent and combined associations of pre-conception BMI and FPG with PTB risk in a massive cohort of Chinese women.
Methods
The researchers conducted a population-based retrospective cohort study. They analyzed data from nearly 5 million reproductive-aged women who participated in China's National Free Pre-Pregnancy Checkups Project between 2013 and 2016 and subsequently had a singleton delivery before December 2017.
Key methodological aspects include:
- Large-Scale Data: The sheer size of the cohort (over 4.9 million women) provides substantial statistical power to detect associations and explore complex relationships.
- Retrospective Cohort Design: This design allows for the examination of exposures (pre-conception BMI and FPG) and their subsequent outcomes (PTB) over time.
- Categorical and Dose-Response Analysis: The study employed both categorical analyses (e.g., underweight, normal weight, overweight, obesity for BMI; hypoglycemia, normal glucose, pre-diabetes, diabetes for FPG) and sophisticated dose-response modeling using restricted cubic splines (RCS). RCS models are particularly valuable for visualizing non-linear relationships and identifying thresholds.
- Stratified Analysis: To understand the interplay between BMI and FPG, the researchers stratified their analyses, examining BMI's effect within different FPG categories and vice-versa.
- Multivariable Adjustment: The models were adjusted for a comprehensive list of potential confounders, including maternal age, education, ethnicity, occupation, parity, smoking and alcohol consumption, history of adverse pregnancy outcomes, hypertension, thyroid dysfunction, anemia, infectious disease, and sex of the child. This rigorous adjustment helps isolate the effects of BMI and FPG.
Datasets
The study utilized data from the National Free Pre-Pregnancy Checkups Project (NFPCP) in China. This project collected data from reproductive-aged women across 31 provinces and municipalities. The dataset included information on pre-conception health examinations, lifestyle factors, and pregnancy outcomes for over 5.1 million women who had live births. After exclusions, the final analysis included 4,987,129 women. No new datasets were introduced by this study.
Key Insights and Ideas
- Synergistic Risk: The most striking insight is the synergistic effect of combined abnormal BMI and FPG on PTB risk. This means that having both a high BMI and elevated glucose levels is more dangerous than the sum of their individual risks, highlighting the importance of addressing both factors concurrently.
- Non-Linear Relationships: The study effectively demonstrates that the relationship between BMI and PTB, and FPG and PTB, is not always linear. The dose-response curves reveal U-shaped and J-shaped patterns, indicating that both extremes (underweight/low glucose and overweight/high glucose) can be detrimental, but the specific shape of the risk curve depends on the other factor.
- Context Matters: The risk associated with a particular BMI category changes significantly depending on the FPG level, and vice-versa. For example, underweight women with hypoglycemia face a different risk profile than underweight women with normal glucose.
- Underweight and Hypoglycemia are Not to Be Ignored: While overweight and obesity often receive more attention, this study underscores that underweight and hypoglycemia, particularly when combined, also significantly increase PTB risk. This challenges a singular focus on combating obesity.
Key Results
- Prevalence of PTB: Out of 4,987,129 women, 339,662 (6.81%) experienced preterm deliveries.
- Independent Associations:
- Compared to normal weight/normal glucose women:
- Underweight and normal weight women with hypoglycemia had increased PTB risk (aORs 1.24 and 1.16, respectively).
- Underweight, overweight, and obesity women with normal glucose had increased PTB risk (aORs 1.09, 1.06, and 1.08, respectively).
- All BMI groups were significantly associated with PTB among women with pre-diabetes or diabetes.
- Dose-Response Patterns:
- BMI vs. PTB (stratified by FPG): U-shaped curves were observed in normal glucose and pre-diabetes women. J-shaped curves were seen in diabetes women (risk increased with BMI). An L-shaped curve (risk decreased with BMI) was noted in hypoglycemia women.
- FPG vs. PTB (stratified by BMI): U-shaped curves were observed in normal weight, overweight, and obesity women. An L-shaped curve was seen in underweight women.
- Highest Risk Group: Obese women with diabetes exhibited the highest estimated risk of PTB.
Limitations and Future Work
The authors acknowledge several limitations:
- Lack of Pregnancy Data: The study could not adjust for maternal weight gain, use of hypoglycemic medication during pregnancy, or pregnancy-specific complications like gestational hypertension or pre-eclampsia. This omission could lead to biased results.
- PTB Subtypes Not Differentiated: The study did not distinguish between spontaneous and iatrogenic PTB, which might have different underlying causes and risk factors.
- Potential Selection Bias: As a retrospective cohort study, there's a possibility of selection bias in the study population.
- Rural Dominance: While covering urban and rural areas, the majority of participants were from rural China, which might limit the generalizability of findings to other populations.
Future work suggested includes:
- Further assessment of the impact of maternal weight gain during pregnancy on PTB.
- Detailed exploration of the associations between BMI and FPG with specific subtypes of PTB (spontaneous vs. iatrogenic).
- Investigating the potential biological mechanisms linking underweight/hypoglycemia and PTB, such as micronutrient deficiencies or decreased plasma volume.
- Evaluating the cost-effectiveness of universal pre-conception care programs in China.
Summary updated