Abstract
BACKGROUND: The impact of methylation gestational age (GAmAge; a biomarker of fetal maturity) at birth on childhood blood pressure (BP) trajectories is unknown. METHODS: This cohort study included 500 boys and 440 girls with data on cord blood DNA methylation and BP at 3 to 15 years of age. Systolic BP (SBP) and diastolic BP percentiles were calculated based on clinical guidelines. Time-series K-means clustering identified 4 distinct SBP and diastolic BP percentile trajectories: high-steady, high-decrease, normal-increase, and normal-steady. GAmAge was estimated using an existing pediatric epigenetic clock. Extrinsic age acceleration was calculated as residuals of associations between GAmAge and chronological gestational age. Intrinsic age acceleration was calculated using the same method adjusting for cord blood cell compositions. RESULTS: Extrinsic age acceleration and intrinsic age acceleration were inversely associated with repeated measures of BP percentiles. Significant inverse associations were observed between extrinsic age acceleration and SBP percentiles in boys (β=-2.02; P=0.02) but not in girls (β=-0.49; P=0.58). Both extrinsic age acceleration and intrinsic age acceleration were inversely associated with SBP percentiles in girls born preterm (<37 weeks; β(EAA)=-2.95; β(IAA)=-3.00; P<0.05). Compared with the normal-steady SBP trajectory, significant inverse associations were observed between intrinsic age acceleration and high-steady, high-decrease, and normal-increase SBP trajectories in boys (odds ratio, 0.73-0.81; P<0.03), and significant positive associations were observed for high-decrease and normal-increase SBP trajectories in girls (odds ratio, 1.26-1.38; P<0.01). Significant sex differences were observed (P(sex-interaction)<2×10(-16)). CONCLUSIONS: GAmAge acceleration at birth was inversely associated with child BP, and such association was more pronounced in boys than in girls. Our findings may shed new light on the developmental origins of high BP and sex differences in cardiovascular risk.