An Evening Routine for Longevity: What the Light Data Shows

Brightest-tenth night light exposure tracked with 21-34% higher mortality across 88,905 people. An evening routine built on the light and meal data.

Among 88,905 people wearing wrist light sensors, those in the brightest tenth of night-time light exposure had 21% to 34% higher all-cause mortality risk than those with dark nights. For cardiometabolic mortality the gap was 33% to 46%. The evening variable with the strongest mortality evidence behind it is not what you do. It is how dark it is.

Why does light at night matter more than most evening habits?

Light in the middle of the biological night suppresses circadian amplitude and shifts circadian phase, both of which independently predicted mortality risk in the 2024 PNAS analysis. The timing detail is striking: modelling light exposure in half-hour blocks across 24 hours, the peak risk sat between 02:30 and 03:00, where the brightest tenth carried 56% to 67% higher cardiometabolic mortality risk than the darkest half.

Short sleep partially explained the night-light association but did not eliminate it. The relationship held after adjustment for age, sex, ethnicity, physical activity, smoking, alcohol, socioeconomic factors, urbanicity, baseline cardiometabolic health, and after excluding shift workers.

Meal timing is the second evening input with real numbers. In the NutriNet-Santé cohort of 103,389 French adults, eating the last meal after 9pm was associated with 28% higher cerebrovascular disease risk compared with eating before 8pm, with each additional hour of delay associated with an 8% increase. Worth being precise: for overall cardiovascular disease the same comparison gave a hazard ratio of 1.13 with a confidence interval crossing 1, so the finding was specific to cerebrovascular events and strongest in women, who made up 79% of the cohort. Diet quality, not just timing, also tracks with long-term outcomes.

What does the evidence support in an evening routine?

  1. Darkness through the small hours, the window between roughly 01:00 and 06:00 where the mortality association concentrates
  2. An earlier last meal, given the association between eating after 9pm and cerebrovascular risk
  3. A consistent bedtime that protects a consistent wake time, since regularity was the stronger mortality predictor in the same cohort family
  4. Caffeine already cleared, a standard coffee having a modelled cutoff 8.8 hours before bed
  5. Dimmer rather than banned screens, since intensity and timing are what the light data measures

The circadian phase finding adds a wrinkle worth knowing. Both unusually early and unusually late circadian timing predicted higher mortality risk, 16% to 30% for early phase and 13% to 20% for late. Earlier is not automatically better. Consistent is.

The bottom line

  • Brightest-tenth night light exposure was associated with 21% to 34% higher all-cause mortality across 88,905 people
  • The risk peak sat between 02:30 and 03:00, at 56% to 67% higher cardiometabolic mortality for the brightest tenth
  • Eating the last meal after 9pm was associated with 28% higher cerebrovascular risk in 103,389 adults, though overall cardiovascular risk did not reach significance

Frequently asked questions

Is light at night bad even if I sleep through it?
The PNAS analysis measured personal light exposure whether or not participants woke. Adjusting for short sleep partially attenuated the association but did not remove it, and adjusting for long sleep or sleep efficiency did not attenuate it at all, which suggests the circadian effect operates partly independently of sleep disruption.
How late is too late to eat?
The NutriNet-Santé cohort compared a last meal after 9pm with one before 8pm and found 28% higher cerebrovascular disease risk in that comparison. It is observational, drawn from one French cohort that was 79% women, and needs replication before it becomes a rule.

Sources

  1. Windred DP et al., 2024. Brighter nights and darker days predict higher mortality risk. PNAS 121(43):e2405924121.
  2. Palomar-Cros A et al., 2023. Dietary circadian rhythms and cardiovascular disease risk in the prospective NutriNet-Santé cohort. Nature Communications 14:7899.
  3. Gardiner C et al., 2023. The effect of caffeine on subsequent sleep: A systematic review and meta-analysis. Sleep Medicine Reviews 69:101764.
  4. Windred DP et al., 2024. Sleep regularity is a stronger predictor of mortality risk than sleep duration. Sleep 47(1):zsad253.

Hydra's editorial team. We write from primary research and cite every study we rely on, with its year and journal. We quantify claims wherever the evidence allows and state plainly where it does not: where a finding rests on a small or single trial, we say so in the text rather than in a footnote.