Brighter nights are linked to heart remodelling and higher cardiovascular risk

People exposed to more light at night had measurably different hearts — thicker, stiffer, weaker-contracting muscle — and went on to develop cardiovascular disease at higher rates, according to a UK Biobank study using wrist-worn light sensors [s1]. Over 8 to 10 years of follow-up, participants with high nighttime light exposure had a 29% higher risk of heart failure, 33% higher risk of stroke, 24% higher risk of heart attack, 26% higher risk of cardiovascular death and 15% higher risk of atrial fibrillation than those with no measured nighttime light [s1].

The exposure here is personal light at night — what a wrist sensor recorded around each sleeper, indoor and outdoor light combined — rather than the satellite-measured outdoor "light pollution" of streetlamps and signage, which is a related but separate literature. The mechanism the study points to is circadian disruption: light at the wrong time desynchronises the body clock, and the analysis found much of the heart effect ran through a familiar downstream pathway — shorter sleep [s1].

What was measured

The core analysis followed 11,071 UK Biobank participants who wore an accelerometer with a built-in light sensor for seven days between 2013 and 2015, then had cardiac magnetic resonance imaging about three years later [s1]. Nighttime light exposure above 3 lux — roughly the brightness of deep twilight, well below normal room lighting — was estimated for each person and related to detailed measures of heart structure and motion [s1]. A larger group of 73,286 participants was followed for later cardiovascular events and death [s1].

Wrist light sensors are the strength of this design and its limitation. They capture actual exposure at the person, not a modelled estimate from where someone lives — but they cannot tell a bright bedroom from a lit street, and a week of wear is assumed to represent years of habit.

What the hearts looked like

After adjustment, high nighttime light exposure was associated with the pattern cardiologists call concentric hypertrophy — a heart that thickens and remodels under strain [s1]. Compared with no measured nighttime light, high exposure was linked to 2.4% greater left-ventricular mass indexed to height (95% CI 1.6% to 3.1%), 1.5% greater mean wall thickness (1.0% to 2.0%), and a 1.9% lower myocardial contraction fraction (1.1% to 2.7%) — thicker walls doing proportionally less work [s1].

Measures of the heart's deformation with each beat — how much the muscle shortens and thickens — were also worse: global circumferential strain was 2.7% lower in absolute terms (95% CI 1.7% to 3.7%), radial strain 2.9% lower (1.8% to 3.9%) and longitudinal strain 2.8% lower (1.7% to 3.9%) [s1]. The changes were not confined to the main pumping chamber: the right ventricle showed larger end-diastolic (0.9%, 95% CI 0.1% to 1.8%) and end-systolic (1.6%, 0.5% to 2.8%) volumes, and the left atrium a larger maximum volume (2.3%, 0.7% to 3.9%) and lower emptying fraction (1.1%, 0.3% to 1.9%) [s1].

Each of these is small, and each showed a linear dose-response — more light, more remodelling [s1]. The authors report that shorter sleep duration statistically mediated a substantial share of the associations, between 24% and 49% depending on the measure [s1]. That is a clue to mechanism but also a caution: much of what looks like a light effect may be a sleep-loss effect that light helps cause.

The outcomes, and an independent replication

In the larger follow-up sample, the structural changes were matched by hard events: the 29% higher heart failure risk, 33% higher stroke risk, 24% higher heart attack risk, 15% higher atrial fibrillation risk and 26% higher cardiovascular mortality already noted, each with a confidence interval clear of no effect [s1].

A separate UK Biobank analysis, published in JAMA Network Open in 2025, reached the same conclusion from the incidence side [s2]. Among 88,905 participants, those with the brightest nights (91st to 100th percentile) versus the darkest (0 to 50th) had adjusted hazard ratios of 1.47 for myocardial infarction (95% CI 1.26 to 1.71), 1.56 for heart failure (1.34 to 1.81), 1.32 for coronary artery disease (1.18 to 1.46), 1.32 for atrial fibrillation (1.18 to 1.46) and 1.28 for stroke (1.06 to 1.55) — robust to adjustment for smoking, diet, physical activity, sleep duration, socioeconomic status and polygenic risk [s2].

The circadian story extends beyond the heart. A 2024 UK Biobank study of 84,790 people found type 2 diabetes risk rising step by step with brighter nights: adjusted hazard ratios of 1.29 (95% CI 1.14 to 1.46), 1.39 (1.24 to 1.57) and 1.53 (1.32 to 1.77) across ascending light bands relative to dark nights, with the gap between bright and dark nights comparable to the gap between low and moderate genetic risk [s3].

How much weight it bears

The consistency is real, but so is a shared weakness: all three studies draw on the same cohort, UK Biobank, and largely the same one-week light-sensor substudy [s1][s2][s3]. Agreement across analyses of the same volunteers is not the same as replication in an independent population, and UK Biobank participants are healthier and less deprived than the general population. These are observational associations; reverse causation (early, undiagnosed heart disease disrupting sleep and light habits) and residual confounding cannot be excluded, and the mediation through short sleep means light is at most part of the causal chain [s1].

What lifts the finding above correlation-of-the-week is the coherence: a plausible mechanism (circadian disruption), a measured structural intermediate (cardiac remodelling), a dose-response, and matching hard outcomes across cardiovascular disease and diabetes [s1][s2][s3]. That is the shape evidence takes when an exposure is doing something, even before a trial proves it.

What to watch

Whether the associations replicate in cohorts outside UK Biobank and with longer light monitoring, and whether any interventional test — dimming or timing light before sleep — moves the intermediate markers. Until then the honest summary is that light at night is a credible, biologically coherent cardiovascular risk signal built on one large but singular dataset, not a settled cause.

This article is informational and does not constitute medical advice.

Sources

• [s1] Dai J, Dai W, Heianza Y, Qi L, "Nighttime light exposure and cardiac structure and function," European Heart Journal, published online 9 September 2026. https://doi.org/10.1093/eurheartj/ehag563
• [s2] Windred DP, Burns AC, Rutter MK, et al., "Light Exposure at Night and Cardiovascular Disease Incidence," JAMA Network Open, vol. 8, no. 10, 1 October 2025. https://doi.org/10.1001/jamanetworkopen.2025.39031
• [s3] Windred DP, Burns AC, Rutter MK, et al., "Personal light exposure patterns and incidence of type 2 diabetes: analysis of 13 million hours of light sensor data," The Lancet Regional Health – Europe, vol. 42, 5 June 2024. https://doi.org/10.1016/j.lanepe.2024.100943