Night Light Exposure Sharply Raises Heart Disease Risk in 89,000-Person Study

Disruption of the body's circadian rhythms is an established pathway to cardiovascular harm. Blood pressure, heart rate variability, platelet activation, and vascular endothelial function all follow circadian cycles, and even short-term circadian disruption in humans raises blood pressure, heart rate, and inflammation. Light at night is one of the most potent disruptors of circadian rhythms, suppressing melatonin and shifting the biological clock — yet evidence linking personal light-at-night exposure to hard cardiovascular outcomes has been limited to small cohorts or satellite-based outdoor lighting estimates.

This prospective cohort study used approximately 13 million hours of individual light-exposure data collected from wrist-worn light sensors worn by 88,905 UK Biobank participants for one week between 2013 and 2016. Factor analysis identified two distinct temporal clusters of light exposure: daytime (7:30 AM–8:30 PM) and nighttime (12:30 AM–6:00 AM). Participants were divided into four night-light percentile groups (0–50th, 51st–70th, 71st–90th, 91st–100th). Incident cardiovascular events — coronary artery disease (CAD), myocardial infarction (MI), heart failure (HF), atrial fibrillation (AF), and stroke — were tracked through NHS records over 9.5 years of follow-up to November 2022.

The findings were striking and dose-dependent. Relative to the darkest-night group, individuals with the brightest nights faced adjusted hazard ratios (aHRs) of 1.32 for CAD, 1.47 for MI, 1.56 for HF, 1.32 for AF, and 1.28 for stroke. These associations survived adjustment for a comprehensive set of cardiovascular risk factors including physical activity, diet quality, smoking, alcohol, sleep duration, socioeconomic deprivation, urbanicity, shift work, and polygenic risk scores. Importantly, no equivalent risk gradient was observed for daytime light, suggesting the nighttime window is uniquely hazardous.

Subgroup analyses revealed meaningful effect modification by sex and age. Females showed larger night-light associations for HF (P for interaction = .006) and CAD (P for interaction = .02). Younger participants in this cohort (aged roughly 40–60) had stronger associations for HF (P = .04) and AF (P = .02). These patterns may reflect sex-based differences in melatonin sensitivity or hormonal modulation of circadian pathways, and age-related differences in light exposure habits or biological vulnerability.

Key limitations include the observational design, precluding causal inference, and the use of only one week of light data, which may not fully reflect habitual long-term exposure. The cohort is predominantly White and healthier than the general UK population ('healthy volunteer bias'), potentially underestimating true population-level effects. Light sensor data captured total illuminance, not blue-light-specific metrics known to drive circadian disruption most potently. Despite extensive covariate adjustment, residual confounding — for example, from unmeasured comorbidities or indoor light source types — cannot be excluded.