Screens before bed: what is actually going on

The idea is that screen light suppresses melatonin and delays sleep. There is a real mechanism there, and the size of the effect from a phone at arm's length is smaller than the amount of attention it gets.

Night shift modes are worth using and are unlikely to be the thing that fixes your sleep. If you have turned one on and nothing changed, that is not surprising.

Displacement: the hour you spent scrolling is an hour you did not spend asleep. This is arithmetic rather than physiology and it is almost certainly the largest effect.

Arousal: the content is engineered to hold attention, which is close to the opposite of what winding down means. An argument in a comment section is not a neutral way to spend the last hour.

• Displacement: you went to bed later. Simple and large.
• Arousal: you arrived in bed activated rather than settled.
• Light: real, and smaller than the other two.

If light were the main mechanism, a filter would solve it. Since it mostly is not, the useful change is about when the phone stops rather than what colour it is.

A hard stop at a set hour addresses displacement and arousal together, which is why it outperforms every setting you can turn on.

This subject is full of confident claims with thin support, including from people selling things. We are people selling something, so treat this page the same way.

What is reasonable to say: going to bed an hour later makes you more tired, and an hour of engaging content immediately beforehand does not help you settle. Anything more precise than that is more precise than the evidence.