A large British study linked neighborhood deprivation, crime, and air pollution to accelerated aging, but only the newest epigenetic clocks picked up the signal
A large British study linked neighborhood deprivation, crime, and air pollution to accelerated aging, but only the newest epigenetic clocks picked up the signal
On 16 September, a team from Britain, the Netherlands, and Spain published a preprint examining how 30 features of the residential environment relate to biological aging in 3307 adults from the British Understanding Society cohort. Five age measures and four analytic approaches converged on one finding: the strongest responses to deprivation, crime, low social cohesion, lack of green space, air pollution, and heat came from the newest clocks, those that estimate the pace of aging rather than accumulated biological age.
The study was written by a group led by Gergő Baranyi of University College London and Catherine Tonne of the Barcelona Institute for Global Health and posted on medRxiv, a preprint server without peer review. The authors drew on the Understanding Society panel, which has surveyed 40,000 British households annually since 2009, and linked 30 environmental indicators across six domains (deprivation, crime, noise, air quality, heat, and infrastructure) to the home addresses of 3307 participants. DNA methylation was measured only in white participants, so the results apply to that group.
Biological age is estimated from DNA methylation, the chemical marks that accumulate on DNA over a lifetime without altering the gene sequence but regulating gene activity. The Horvath and Hannum clocks were trained to predict chronological age. PhenoAge was trained on mortality risk. DunedinPoAm and the updated DunedinPACE, currently the most widely used tool in gerontology, were trained on declining organ function observed in participants of the New Zealand Dunedin cohort; DunedinPACE was specifically calibrated on twenty years of those longitudinal measurements, which is why it translates the same methylation marks into a rate: how many biological years pass per calendar year.
The Hannum clock showed almost no response to the environment. Horvath and PhenoAge both showed associations with distance from sports centers and distance from betting shops, but these two distances nearly always coincide (correlation 0.83 out of 1), and joint modeling ruled out betting shops as a spurious signal: the only association that survived for Horvath was that greater distance from a sports center corresponded to older biological age on that clock. PhenoAge responded more strongly than first-generation clocks, and DunedinPACE responded most strongly of all: its effect size for air pollution was 2.5 to 3 times larger than for DunedinPoAm.
Instead of the conventional urban/rural split, the authors clustered participants by the actual profile of their environment, producing an urban cluster (2241 people, 73%) and a rural cluster (1066, 27%). By the Horvath clock, urban residents were 0.385 years younger than rural ones, while by DunedinPACE they were aging faster by 0.011 pace-years per calendar year. Two years earlier, one of the study's co-authors, Meena Kumari, had tested aging in the same cohort using the official urban/rural classification and found no difference. The difference appeared only when the grouping was built from statistical features of the environment: the administrative "urban" category lumps together dissimilar populations, whereas the data-driven cluster more precisely captures those who are aging faster.
Different mechanisms likely underlie these associations: chronic stress in the case of deprivation and crime, inflammation in the case of air pollution, disrupted thermoregulation in the case of heat, and the effect of low social cohesion is partly explained by higher smoking rates and lower physical activity. The design is cross-sectional (a single time point, with no long-term follow-up), so reverse causation is possible: people who are already aging faster may be more likely to end up living in worse neighborhoods. The newer clocks have shown the same advantage in predicting mortality as well, based on a 24-year trajectory in the InCHIANTI cohort: the more precisely a clock is calibrated, the more clearly it detects the imprint of where a person lives.