Does a blue-light filter on my phone really affect my sleep?
A blue-light filter on your phone probably has little effect on your sleep. If you really want to improve your sleep through light, switch your home lighting to warm yellow in the evening instead of cool white.
A blue-light filter app on your phone probably does not help. An observational study found no link between using such an app and better sleep. And an experiment involving Facebook use on a tablet showed that filtering blue light alone produced no sleep improvement: only the combination of removing blue light and eliminating distracting social content had a visible effect.
Blue-light filter glasses paint a somewhat mixed picture. Six smaller RCTs split neatly into thirds: three found an improvement, three found nothing. The certainty of that evidence is rated 'very low' by a Cochrane analysis. Two separate experiments with employees did find improvements in sleep and even in work performance, but those groups were small and the context specific. Evening types appeared to benefit more than morning types.
One thing stands out: the choice of your home lighting seems to have more influence than any filter on your screen. Cool-white LED or fluorescent light suppresses the production of the sleep hormone melatonin considerably more strongly than warm-white light. If you set your LED lamp to a warm colour temperature in the evening (around 2100K), that suppression drops to virtually zero. That is more effective than most blue-light filter lenses, of which only two with a brown tint performed genuinely well in the same laboratory study. Standard lenses do very little.
One further caveat: the Cochrane review found no evidence that blue-light filter glasses reduce eye strain. Side effects were rare, but were present in a proportion of users: think mild mood changes or headaches.
Based on a Cochrane review (6 RCTs, 148 participants), two small experimental employee studies (n=63 and n=67), an observational smartphone study, a tablet experiment, a meta-analysis of intraocular lens implants (8 studies, 1007 patients) and a laboratory study of 52 lamps. All human studies are small; the evidence for the app specifically is observational.