Comparing allostatic load scoring methods and timing: Their links to preterm birth outcomes

Document Type

Article

Department

Obstetrics and Gynaecology; School of Nursing and Midwifery, Pakistan

Abstract

Background: Allostatic load is a well-recognized indicator of cumulative physiological stress during pregnancy, with scores derived from biomarker algorithms that hold promise for identifying women at risk of preterm birth. However, there is currently no consensus on the most accurate and reliable method for aggregating these biomarkers into a composite allostatic load score, and the comparative predictive performance of various scoring approaches—along with the influence of assessment timing—remains insufficiently explored.
Objectives: To compare multiple allostatic load scoring algorithms’ ability to predict preterm birth among pregnant women and assess the impact of biomarker assessment timing on these associations.
Study Design: This prospective longitudinal cohort study (2018–2020) enrolled 1576 healthy women (10–19 weeks’ gestation), reassessed 1140 at 22–29 weeks from four Pakistani antenatal clinics. Six allostatic load biomarkers were measured with seven scoring algorithms applied to derive allostatic load scores. Their associations with preterm birth were assessed using logistic regression, with model comparisons made by Akaike Information Criterion and Area under the Curve.
Results: Average allostatic load scores across algorithms ranged from 0.16 to 6.13, with 13.1% of participants experiencing preterm birth. Six of seven allostatic load scoring algorithms—highest quartile, two-sided quartiles, highest decile, extreme z-scores, clinical limits, and extreme sextiles—were significantly associated with preterm birth when assessed in early pregnancy, though associations weakened at 22–29 weeks. Clinical limits scoring showed the strongest association with preterm birth early in pregnancy but lost significance later.
Conclusion: The association between allostatic load and preterm birth varies by scoring algorithm and timing, with early clinical limits assessment offering robust prediction. While extreme sextiles, highest quartile and decile algorithms predict preterm birth at 22–29 weeks, they lack standardized cut-offs. Early assessment and allostatic load scoring using clinical limits offers a robust, objective tool to assess risk of preterm birth. Standardizing cut-offs for extreme sextiles and highest quantile could produce allostatic load measures that are less sensitive to timing and can complement clinical thresholds as effective screening tools.

Comments

Pagination is not provided by author/publisher

Publication (Name of Journal)

AJOG Global Reports

DOI

10.1016/j.xagr.2026.100655

Link to Full Text

Share

COinS