TL;DR
This study reveals that advanced maternal age (over 35) is non-linearly associated with increased risks of various adverse pregnancy outcomes in Chinese women. These risks are particularly pronounced for nulliparous women, underscoring the importance of considering both age and parity when advising on optimal childbearing timing.
What Problem Does This Paper Address?
While it's generally accepted that advanced maternal age can lead to adverse pregnancy outcomes, the precise nature of this relationship and how it varies across different outcomes and maternal characteristics remains an area needing further clarification. Previous research often treated maternal age as a simple categorical variable (e.g., in 5-year increments) and focused on Western populations. This study aims to address these gaps by:
- Characterizing the relationship as continuous and non-linear: Understanding if the risk increases steadily or if there are specific "inflection points" where risk escalates more sharply.
- Investigating a diverse range of adverse outcomes: Moving beyond just a few common complications to a broader spectrum.
- Examining the role of parity: Determining if the risks associated with maternal age differ significantly between women having their first child (nulliparous) and those who have had previous births (multiparous).
- Focusing on a Chinese population: Recognizing that sociodemographic factors and clinical practices can influence these associations, making population-specific data crucial.
The increasing trend of delayed childbearing globally, including in China, makes this research highly relevant for public health guidance and clinical practice.
Methods
The researchers employed a prospective multicenter cohort study design, utilizing data from the University Hospital Advanced Age Pregnant (UNIHOPE) cohort.
Key methodological aspects include:
- Restricted cubic splines: This statistical technique was used to model the nonlinear relationships between maternal age (treated as a continuous variable) and adverse pregnancy outcomes. This allows for a more nuanced understanding of risk escalation compared to simple linear models or categorical age groups.
- Multivariable log-binomial regression: This was used to estimate the adjusted relative risks (RRs) and 95% confidence intervals (CIs) for the associations between maternal age and adverse outcomes. This method is suitable for binary outcomes and provides risk estimates adjusted for various potential confounders.
- Interaction terms: To assess whether parity modified the association between maternal age and adverse outcomes, interaction terms were included in the regression models. If significant interactions were found, subgroup analyses were performed.
- Adjustment for confounders: The models were adjusted for a comprehensive list of potential confounding factors, including ethnicity, education, occupation, income, pre-pregnancy BMI, smoking/alcohol consumption, parity, method of conception, and year of delivery.
- Sensitivity analysis: Missing data was handled using multiple imputation, and sensitivity analyses were conducted to assess the robustness of the findings.
Datasets
The study utilized data from the University Hospital Advanced Age Pregnant (UNIHOPE) cohort, a prospective multicenter study conducted in eight public hospitals across seven cities in China between July 2016 and June 2021. The final analysis included 18,495 singleton pregnant women. This dataset is a newly established cohort designed to investigate pregnancy complications among women of advanced maternal age, with a portion of younger women included for comparison.
Key Insights and Ideas
- Non-linear risk escalation: The study strongly suggests that the risk of adverse pregnancy outcomes doesn't simply increase linearly with age. Instead, there appear to be specific "inflection points" where the risk begins to climb more steeply, particularly around ages 35.6-40.4. This provides a more refined understanding of "advanced" maternal age.
- Parity matters significantly: The finding that risks are more pronounced in nulliparous women is a critical insight. This implies that first-time mothers who delay childbearing face a disproportionately higher risk compared to multiparous women of the same age. This highlights the need for tailored counseling.
- Broad spectrum of affected outcomes: The research confirms increased risks across a wide array of outcomes, including gestational diabetes mellitus (GDM), hypertensive disorders of pregnancy (HDP), preeclampsia, placenta accreta spectrum (PAS), placenta previa, cesarean delivery, preterm birth (PTB), large for gestational age (LGA), macrosomia, and congenital anomalies. This comprehensive view underscores the pervasive impact of maternal age.
- COVID-19 influence: Preliminary findings suggest that the risks for GDM and PAS might have been more pronounced during the COVID-19 pandemic compared to the pre-pandemic period, hinting at potential environmental or behavioral influences on these age-related risks.
Key Results
- Inflection points: The study identified inflection points for various outcomes, with some risks (e.g., PAS, LGA) starting to increase significantly around 35.6-36.4 years, while others (e.g., GDM, PTB, macrosomia, HDP, PE, congenital anomaly) showed steeper increases around 39.6-40.4 years.
- Nulliparous vs. Multiparous: The risks for HDP, PE, PP, and cesarean delivery were significantly higher in nulliparous women compared to multiparous women across various age groups. Similarly, GDM and PAS risks were more pronounced in nulliparous women.
- Age categories: Compared to women younger than 35, women aged 40-44 had higher risks for PAS, PP, LGA, and macrosomia. Women aged ≥45 had the highest risks for GDM, HDP, PE, cesarean delivery, and PTB.
- Specific outcome risks: For instance, the prevalence of GDM increased from 18.3% in women <35 to 44.3% in women ≥45. Similarly, cesarean delivery rates rose from 41.5% to 84.4% across these age groups.
Limitations and Future Work
The authors acknowledge several limitations:
- Hospital-based cohort: The study included only women from public referral hospitals, which might lead to biased effect estimates if associations differ in non-referral or private settings.
- Residual confounding: Despite adjusting for many factors, residual confounding from unmeasured sociodemographic or lifestyle factors (e.g., physical activity, diet) cannot be entirely ruled out.
- Limited scope for very young mothers: The cohort was designed for advanced maternal age, so associations for very young mothers (<20 years) could not be examined.
- Selection bias: Women with higher adherence to antenatal examinations were more likely to be included, potentially introducing selection bias.
- Self-reported data: Smoking and alcohol consumption data relied on self-reporting, which is prone to under-reporting.
- Limited power for older age groups: Statistical power to detect differences in risks for some outcomes, especially for the ≥45 age group, was limited.
Future research should focus on replicating these findings in other populations, exploring the interaction effects of parity, ethnicity, and pre-existing comorbidities more deeply, and investigating the mechanisms underlying these age-related risks. Further studies are also needed to understand the impact of the COVID-19 pandemic on these associations.
Summary updated