The study is claiming correlation.

It's listed right in there. Search for "lux".

Only 13% of the population works nights, yet people with high levels of exposure to light at night were far more than 13% of the sample.

The percentiles of 50-60, 60-70, 70-80, and 80-90, however, are obviously not shift workers (unless they're shift workers working in the dark); you compare the lowest percentile of daylight light exposure to the highest percentile of nighttime light exposure, and see that there just aren't that many shift workers present in the study. About 13% of the UK population is considered to be nighttime workers. We can safely assume nighttime workers aren't represented in the 50%-90% percentile, because there simply aren't enough of them to go around, and would not be statistically significant.

N>88,000, so it has a huge sample size. Most people have the same patterns of sleep and nighttime/daytime habits week to week.

"Correlation is not causation!" is one of the sillier things that people say when a study comes out with meaningful, interesting data. If nobody ever finds correlations in the first place, then it won't be possible to figure out causation - and we certainly do want to eventually find more causation for things like heart attacks, don't we?

I agree.

I see it constantly when a study comes out all of these statistics students come out of the wood work and say "Correlation doesn't mean causation" without any other thought.

The study clearly claims only a correlation and absolutely, correlations help focus attention to try and find causal links in the future.

I agree, correlation certainly hints at a relationship.

However, I think you are just being triggered by past traumas, because nobody said that here.

They grouped the population into percentiles with 0-50% having quite low exposure (0.62 lux median, range of 0-1.21) vs the 90%-100% percentile being 105 lux (range of 48.3-6400). You can compare the daylight light exposure to the nighttime and basically see what a shift worker would be.

But they already deconfounded for shift workers, so that's irrelevant. And they also showed the amount of light exposure for both night and day.

It turns out they took both rural and urban samples too and there was no meaningful difference between the two populations. Exposure to indoor overhead light is probably far more significant than what streams in your window with the drapes shut.

Good catch, although there are plenty of studies and even metastudies on that topic (eg https://pmc.ncbi.nlm.nih.gov/articles/PMC11129786/ ); generally speaking, the healthiest people are those who work "normal" day shifts with "normal" hours (full time; not part time; no overtime).

Isn't that just a rough proxy for wealth? Outside of the coke-fueled hedge fund manager, the poor are more likely to have over-time/shift-work/irregular work hours. Your standard office drone can have a consistent work/sleep schedule.

It's not, because the study specifically disambiguated for socioeconomic status.

The effect was also seen outside of the night worker population (13% of the population).

I've always wondered about people who say that... there's enough light (particularly when it's not a new moon) to see to go the bathroom, but maybe my eyes adjust to the light better?

Unless you use shutters so the room is completely dark

You must live in the city.

I live in a rural area.