Development of a microRNA clock for gestational age in the general population
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Introduction
DNA methylation (DNAm) has shown promise as a biological marker of gestational age (GA). Here, we aimed to characterize how plasma circulating miRNAs – another epigenetic mechanism regulating gene expression – associate with GA at birth, and to construct a miRNA-based GA clock (miRClock-GA).
Methods
We leveraged 2083 umbilical cord plasma-derived circulating miRNAs from Generation R (N=1695). First, we performed linear regressions to identify miRNome-wide significant miRNAs associated with GA. Second, we applied elastic net regression to construct miRClock-GA. These steps were validated in Gen3G (N=213). Finally, we computed age acceleration (miRClock-AA) and evaluated association of miRClock-GA and miRClock-AA with child developmental outcomes up to 17 years of age, including comparisons with DNAmClocks.
Findings
We identified 123 miRNAs associated with GA, with miR-150-5p showing the strongest positive association ( B =0.244, SE=0.036, P=2.3e-11) and miR-373-3p the strongest negative association ( B =-0.255, SE=0.065, P=8.6e-5). MiRClock-GA correlated consistently with GA in Generation R train (r range =0.62-0.72) and test sets (r range =0.45-0.52) and in the independent validation cohort (r Gen 3 G =0.33). Correlations between miRClock-GA and DNAmClocks were weak to moderate (r range =0.28-0.42). MiRClock-AA explained significant variance in birthweight and childhood BMI beyond clinical GA.
Conclusions
This study reveals widespread associations between circulating miRNAs and GA, supports miRClock-GA as a consistent, well-performing biological marker of GA, with miRClock-AA predicting birthweight and childhood BMI beyond GA itself. Our findings provide a broader perspective on the potential utility of miRNAs as early markers of development.
Funding
E.U. Horizon Europe Research and Innovation Programme (FAMILY,No.101057529); European Research Council (TEMPO,No.101039672). Full funding in Acknowledgements.
Research in context
Evidence before this study
Epigenetic clocks have emerged as powerful tools for assessing individual differences in biological ageing. Existing epigenetic clocks are typically constructed using DNA methylation (DNAm) data and are developed to estimate whether an individual’s epigenetic age deviates relative to their chronological age. In adults, higher predicted biological epigenetic age than chronological age (accelerated epigenetic ageing) is an established indicator of mortality risk and other age-related morbidities. MicroRNAs are another epigenetic mechanism associated with chronological age and age-related phenotypes across species such as C. elegans , mice, primates and humans. Recently, a plasma-derived, cell-free circulating miRNA clock was developed in a sample of older European adults, with accelerated miRNA age predicting biological age-related conditions such as frailty and multi-system blood biomarkers. However, miRNA-based estimations of biological gestational age at birth have not been previously done and their relationship to child health outcomes are unknown.
Added value of this study
To gain a broader perspective on the role of epigenetic markers in biological gestational age at birth and the potential utility of miRNAs as early markers of (altered) development, we leveraged cord blood plasma circulating miRNA data collected at birth in a large population-based cohort. First, we identified 123 miRNAs associated with gestational age. Second, we used this information to construct a miRNA-based epigenetic clock for gestational age (miRClock-GA) and validated this clock in an independent cohort. MiRClock-GA showed moderate correlations with both gestational age and existing DNAm-based gestational clocks. Third, we found that miRClock-GA correlated cross-sectionally with birthweight, as well as prospectively with several adaptive, behavioural, cognitive, and growth outcomes measured up to age 17y, after taking into account maternal influences such as maternal education level, smoking, pre-pregnancy BMI, and age. Accelerated ageing of miRClock-GA (miRClock-AA) explained variance in birthweight and childhood BMI beyond clinical GA, after accounting for maternal influences.
Implications of all available evidence
Our study provides the first large-scale evidence that circulating miRNAs in cord blood plasma show widespread associations with gestational age at birth and can be used to derive a robust biological marker of GA (MiRClock-GA), which explains a substantial proportion of the variation in clinical GA. MiRClock-AA provides information beyond GA in some child outcomes, such as BMI. These findings extend adult research, supporting the potential of miRNAs as promising markers for early risk assessment and health monitoring in a developmental context.
