Blue light and sleep — what the research says
Short answer
Limited evidenceLight at the wavelengths produced by phones and screens (around 480 nm) does suppress melatonin and shift circadian phase. But the effect from typical evening screen use is smaller than commonly suggested — light intensity matters more than colour, and the arousing content on the screen probably affects sleep more than the light itself.
Key points
- Specialised retinal cells (ipRGCs) are most sensitive to blue-green light around 480 nm. This is the photoreceptor that signals circadian information.
- Bright light in the evening — regardless of colour — delays your circadian rhythm. The colour story is one input, not the whole story.
- Most modern phones run at 100–300 lux at typical viewing distance. Indoor lamp lighting can be similar. Both produce modest, real effects.
- Blue light filtering apps (Night Shift, blue blockers) reduce melatonin suppression by ~30% in lab studies. Whether this matters subjectively is unclear.
- The biggest evening sleep disruptor for most people is content arousal (work email, news, social media), not screen wavelength.
The mechanism is real
In 2002, researchers identified a third class of photoreceptors in the retina — intrinsically photosensitive retinal ganglion cells (ipRGCs) — distinct from the rods and cones used for vision. These cells contain melanopsin and are most responsive to light at ~480 nm (blue-green).
ipRGCs project directly to the suprachiasmatic nucleus (the circadian master clock) and to other arousal-related brain areas. When they fire, they signal “it's daytime, be alert, suppress melatonin.”
Exposing ipRGCs to bright blue-wavelength light in the evening therefore produces measurable melatonin suppression and a small phase delay.
The effect size is smaller than headlines
The classic study cited in alarmist articles (Chang et al, 2015) compared 4 hours of reading on an iPad versus a printed book before bed. The iPad group showed:
- ~22% reduction in evening melatonin (statistically significant).
- ~10 min longer sleep onset latency.
- Slight reduction in REM the following night.
These are real effects, but modest — and they came from 4 hours of close reading at maximum brightness. The typical evening of glancing at a phone before bed produces a much smaller dose-response.
Intensity matters more than wavelength
Total photon flux at ipRGC-relevant wavelengths is what determines the effect. A dim screen at 100 lux for 30 minutes produces less melatonin suppression than a bright (300+ lux) overhead room light for the same time — even though the room light has less blue content per photon.
For most people in most modern homes, the bigger evening light dose comes from overhead room lighting, not from the phone. Dimming the room is a higher-leverage intervention than blue-light apps.
Do blue light blockers work?
Amber-tinted glasses and phone “night shift” modes both reduce blue-wavelength output. Lab studies show ~30% less melatonin suppression. Two questions remain:
- Does this translate to better subjective sleep? Studies are mixed. Some show small improvements; others find no effect.
- Compared to just dimming the screen or reducing screen time? Probably similar overall.
Arousal — the bigger story
For most people, the problem with screens at bedtime isn't the photons. It's the content:
- Work email keeps you in problem-solving mode.
- News and social media spike cortisol and arousal.
- Engaging entertainment delays the wind-down process.
- Phone notifications trigger micro-arousals.
Switching from a phone to a paper book at 10pm produces a much larger improvement than enabling night mode on the phone.
Practical guidance
- Dim overhead room lighting in the last 2 hours before bed. Bigger lever than screen filters.
- Use night-shift / blue-blocker modes if convenient. Modest help.
- Switch from arousing content to calming content before bed. Bigger lever than wavelength.
- If sleep is a serious problem, remove the phone from the bedroom entirely.