Hacker News Reader

Comment by jasonvorhe | Hacker News Reader

Good luck staying misinformed then. "Trust the science".

metareview CRD420251034611 "non-significant reduction in sleep onset latency" means blue light has been conclusively observed to not delay sleep - it doesn't mean that blue light has not yet been observed to delay sleep, we instead know conclusively that it does not.

registered trial: https://clinicaltrials.gov/study/NCT01855126 - RCT where shining a bunch of blue light on old people's eyeballs didn't help them stay awake longer.

something i learned from this journey is, going to bed too early is actually a more common clinical issue than staying up too late. like old people go to the doctor, and that's the issue that they have, they feel like they sleep way too much.

Well, we know that conclusively in the study population of 65+ year olds who suffer from insomnia and suffer from excessive daytime sleepiness

So you are saying we need to conduct trials on a representative set of bayesnets of say.. 200k and then it'll be conclusive?

I think that’s uncharitable. The study has a very specific goal of assessing whether blue light can keep seniors with problems staying awake from falling asleep early. It’s neither obvious nor intended by the authors that conclusions about this population generalize further.

Especially in the context of the flux and night shift, which are intended to help people with the opposite problem (who want to sleep sooner rather than stay awake later), it’s an important distinction.

The placebo effect is a hell of a thing, and is especially potent with sleep interventions.

Reply on news.ycombinator.com