This preprint assessed whether a cord-blood DNA methylation signature of birthweight discovered in population studies also applies to a clinically high-risk group and whether the signal reflects birthweight per se or is driven by shared genetic, familial, maternal, or obstetric factors. The authors used a monocorionic (MC) twin design to probe independence from shared factors because MC co-twins are genetically identical and share a single placenta, yet can differ markedly in growth due to unequal placental sharing.
The analysis used data from the Twinlife study comprising 67 MC twin pairs (134 neonates). The cohort was heavily preterm: 97% of neonates were born preterm. The study population thus represents a high-risk clinical setting in which unequal placental resource distribution can produce discordant fetal growth.
The investigators computed a methylation profile score for birthweight (MPSBW) using weights from 835 CpG sites previously reported to associate with birthweight in the general population. The MPSBW summarizes cord-blood methylation at these CpGs into a single score intended to capture the epigenetic signature linked to birthweight.
The authors examined the association between birthweight and MPSBW across the sample. In the overall analysis, they report that a 500-gram increase in birthweight was associated with a 0.256 standard deviation (SD) increase in MPSBW (p < 0.005). This demonstrates that the methylation signature identified in population cohorts also correlates with birthweight in this high-risk, predominantly preterm MC twin sample.
To isolate effects not attributable to factors shared by co-twins (such as inherited DNA sequence, maternal environment that is the same for both fetuses, and shared obstetric conditions), the investigators performed within-pair analyses. The within-pair model evaluates whether differences in birthweight between co-twins correspond to differences in MPSBW. The within-pair result showed a similar magnitude of association: a 0.226 SD increase in MPSBW per 500 g increase in birthweight (p < 0.05). Because MC co-twins are genetically identical and share many prenatal exposures, this finding supports that the methylation–birthweight relation reflects individual-specific influences on growth rather than only shared familial or maternal factors.
The investigators further adjusted analyses for a polygenic score for birthweight (PGSBW) to test whether common genetic variation captured by the polygenic score explained the observed epigenetic associations. The reported results indicate that adjustment for PGSBW did not account for the methylation–birthweight association, suggesting that the MPSBW signal is not driven by the common genetic variation summarized by that polygenic score.
Source data for the main figures and statistical analyses are provided in Supplementary Data file 1. The Twinlife study individual-level data are not publicly available because of privacy constraints but can be requested from the corresponding investigator (B.T.H.) and shared for replication within the secure environment of Leiden University Medical Center; timelines for access depend on request complexity. The study received ethical approval from the Ethics Committee of Leiden University Medical Center. The authors declared no competing interests. Funding sources listed include Hartstichting, Leiden University Medical Center, the European Research Council, and the European Union’s Horizon Europe programme.
In a clinically relevant high-risk population—monochorionic twin pregnancies with a high rate of preterm birth—cord-blood DNA methylation summarized by an 835-CpG MPSBW associated with birthweight at the individual level. Crucially, a within-pair MC twin analysis, together with adjustment for a polygenic score for birthweight, indicates that the methylation–birthweight association persists independent of shared genetic, familial, maternal, and shared obstetric factors. The authors conclude that cord-blood methylation captures individual differences in fetal growth in this high-risk setting.
Notes and limitations reported in the source
This manuscript is a preprint and has not undergone peer review; the authors caution it should not be used to guide clinical practice. Additional methodological details, supplementary analyses, and data supporting figures are provided in the manuscript’s supplementary materials. The Twinlife data access is controlled and requires approval and secure-environment use at the originating center for replication requests.