Neighborhood conditions have long been recognized as social determinants of children's health, but their link to biological aging at the molecular level has remained poorly understood. Previous work mostly assessed cross-sectional associations between individual adverse factors, such as air pollution or poverty, and epigenetic age. Research on this topic already includes studies connecting genes and environment to adolescent behavior, as well as work on cellular aging through DNA methylation. However, no one had examined how a holistic neighborhood profile at birth shapes the long-term trajectory of epigenetic aging in childhood. Our work fills that gap by using multidimensional exposome profiles and longitudinal data.
We used data from the prospective Project Viva cohort, in which 570 children had at least one DNA methylation measurement at age 3, 8, or 13. For each child, we determined 22 neighborhood characteristics at the place of birth, covering socioeconomic deprivation, social fragmentation, population density, and environmental quality. Using self-organizing maps (SOM), we grouped participants into three neighborhood profiles based on these characteristics. We then calculated epigenetic age acceleration (EAA) using three epigenetic clocks: Horvath, Skin & Blood, and Wu. To identify EAA trajectories, we applied latent classes of linear mixed models, which allowed us to distinguish groups with stable, increasing, and decreasing (or low baseline) acceleration. The association between neighborhood profiles and trajectories was assessed with multinomial mixed models adjusted for covariates. In addition, we conducted an exploratory weighted quantile sum (WQS) analysis to estimate the joint contribution of the mixture of characteristics.
We identified three neighborhood profiles: Profile 1 (48.8% of participants), suburban areas with the least disadvantage; Profile 2 (29.5%), urban areas with high housing instability, a large share of single people, and proximity to a highway; and Profile 3 (21.8%), urban areas with high socioeconomic deprivation and air pollution. Children born in Profile 2 had significantly higher odds of belonging to the group with an increasing EAA trajectory on the Wu clock compared with children from Profile 1 (adjusted odds ratio 1.67, 95% CI: 1.03–2.69). For the Horvath clock, the association was positive but did not reach statistical significance (aOR 1.75, 95% CI: 0.96–3.17). Profile 3 showed no significant associations with any trajectory. In the exploratory WQS analysis, a one-decile increase in the mixture of adverse characteristics was associated with a 6% increase in the odds of an increasing Wu EAA trajectory (OR 1.06, 95% CI: 1.01–1.11). The largest contributions to this mixture came from traffic density (weight 0.17), housing instability (0.16), and the share of single people (0.11).
This work differs from neighboring studies in that it considers not individual factors but holistic configurations of neighborhood conditions and their link to longitudinal trajectories of epigenetic aging. Whereas previous work mostly examined static snapshots or the influence of genes on behavior, here a multidimensional exposome approach is applied and the dynamics of biological age are tracked throughout childhood. The results show that a specific profile of instability and proximity to traffic, rather than socioeconomic deprivation alone, may be associated with accelerated epigenetic aging. This refines our understanding of which aspects of the environment are most biologically significant and underscores the importance of the early life period as a window of plasticity.
In the future, studies could expand neighborhood assessment to multiple time points to account for changes in the environment during childhood. It is also necessary to examine how epigenetic age trajectories relate to actual clinical outcomes in adolescence and adulthood. Incorporating data on family conditions and individual stress biomarkers to identify mediators could be potentially useful. Developing epigenetic clocks more sensitive to child development and validating them in diverse populations will also be an important step.
The results can be used in urban planning and public health to justify measures aimed at reducing housing instability and limiting traffic near residential areas. They also give researchers a tool for identifying children at increased risk of accelerated biological aging.
The next step will be to test the reproducibility of the results in independent cohorts with a more diverse ethnic composition and a longer follow-up period. It is also necessary to study which biological pathways (for example, inflammation or the stress response) mediate the link between neighborhood and epigenetic aging.
The work directly addresses an unresolved problem in epidemiology: how exactly the social environment becomes embedded in human biology, and why some children are more vulnerable to adverse conditions than others. It also touches on the question of the clinical significance of epigenetic age in pediatrics, which remains open. In addition, the study raises the problem of mixing methylation measurement platforms (450K versus EPIC), which is a methodological challenge for longitudinal epigenetic research.
The Wu clock, which showed a significant association, was trained exclusively on pediatric samples, making it especially sensitive to developmental processes. In our study, only 30 children had all three methylation measurements, but the statistical models allowed data from all 570 participants to be used.
The idea that place of birth programs biological aging echoes dystopias in which social inequality is literally woven into characters' bodies, as in the film Gattaca. There, genetic destiny was determined at birth, whereas here neighborhood conditions may leave an epigenetic mark that unfolds over time.
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.