Circadian Light: Managing Blue Light and Morning Sun

Key Takeaways

For most of human history the question of circadian light answered itself: the sun came up, you woke; it went down, you slept, and the only light after dark was fire, which is dim and red. The body clock evolved under that regime and still expects it. What has changed in the last century is that we spend our days in rooms lit to a fraction of outdoor brightness and our evenings bathed in light of exactly the colour the clock reads as morning.

This article is about the physiology that makes that a problem, the numbers that define a solution, and the large new datasets that show what is at stake. It is deliberately heavy on lux values and light on product recommendations, because the single most effective intervention — go outside in the morning, and make the evening dark — costs nothing.

The Photoreceptor You Did Not Know You Had

Until 2001 it was assumed that light reached the body clock through the rods and cones. Two action-spectrum studies published that year, one in 72 volunteers and one in 22, measured how effectively light of each wavelength suppressed night-time melatonin. Both found a peak around 446–480 nm, in the blue, with a spectral shape that matched neither rods nor cones but fitted a previously unknown opsin. That pigment was melanopsin, housed in a small set of intrinsically photosensitive retinal ganglion cells that project directly to the suprachiasmatic nucleus.

The practical implications follow directly. The clock cares about blue-enriched light. Daylight is enormously rich in it; so are white LEDs and screens. Candlelight and incandescent bulbs are not. And because these cells integrate light over long periods and are far from the fovea, you do not need to look at a light source for it to count — ambient light at eye level is what matters.

How Far Light Moves the Clock

The definitive human phase-response curve to light was built in 2003: 21 entrained volunteers were kept under constant conditions and given a single 6.7-hour pulse of roughly 10,000 lux at different circadian times. The curve has a peak-to-trough amplitude of about five hours. Light falling before the core body temperature minimum — in the late evening and early night — delays the clock. Light falling after it — in the morning — advances it. And, contrary to earlier assumptions, there was no extended daytime "dead zone" where light does nothing: daytime light continued to produce small shifts.

Five hours from a single exposure is far more than any drug achieves. Melatonin's own phase-response curve, by comparison, moves the clock by an hour or so under optimal timing. This is why every serious circadian protocol puts light first and the tablet second, a hierarchy laid out in the Circadian Optimization Protocol.

How Little Evening Light It Takes

The most unsettling study in this field was published in 2019. Fifty-five young adults were exposed, on separate evenings, to light between 10 and 2,000 lux for five hours before their usual bedtime, and their melatonin was measured. The median light level that produced half-maximal suppression was 24.6 lux — dimmer than most living rooms. At 10, 30 and 50 lux, apparent melatonin onset was delayed by 22, 77 and 109 minutes respectively.

The individual variation was the real finding. The half-maximal threshold ranged from 6 lux to 350 lux between participants — a more than fifty-fold spread. One person's melatonin is halved by a bedside lamp; another's barely notices a bright kitchen. You do not know which you are, there is no consumer test for it, and the safe assumption is the sensitive one.

An earlier crossover study compared reading on a light-emitting e-reader with reading a printed book for the hours before bed. The e-reader lengthened the time to fall asleep, reduced evening sleepiness, suppressed melatonin, pushed circadian timing later and left readers less alert the next morning. The exact magnitudes are in the full paper; the direction is unambiguous.

The Numbers: Day, Evening, Night

In 2022 an international expert group published consensus recommendations for indoor light, expressed in melanopic equivalent daylight illuminance — lux weighted for the melanopsin system, measured vertically at the eye.

The daytime target is the one most people miss without realising. A room that feels bright may deliver 100–150 lux melanopic at the eye, which the clock registers as a dim, ambiguous signal. The result is a flattened day-night contrast: the clock gets too little light when it should get a lot, and too much when it should get almost none.

What Happens When You Give the Clock Natural Light

In 2013 researchers sent a group of adults camping for a week in summer with no electrical light. Under natural light alone, biological night — the period of melatonin secretion — began at sunset and ended before wake, just after sunrise. In the modern environment the same people had received far less daytime sunlight, far more light after sunset, and their clocks ran later. The later a person's chronotype, the larger the advance that camping produced.

A follow-up in 2017 showed the shift happens fast. A week of winter camping produced an earlier biological night and sleep; a single weekend of summer camping achieved about 69% of the shift seen after a full week. The clock is not stubborn. It is simply being given the wrong information.

The Population Data: Daylight, Night Light and Outcomes

Three large analyses from the UK Biobank, two of them using a week of wrist-worn light-sensor data in tens of thousands of people, have moved this from physiology to epidemiology.

Daylight and mood. Across more than 400,000 participants, the median time spent outdoors in daylight was 2.5 hours a day. Each additional hour was associated with lower odds of lifetime major depression, antidepressant use and low mood, greater happiness, easier waking, less tiredness, fewer insomnia symptoms and an earlier chronotype. The longitudinal analyses pointed the same way, with smaller effects.

Night light and type 2 diabetes. In 84,790 people followed for about eight years, those in the brightest 10% of night-time light exposure had a 53% higher risk of incident type 2 diabetes than the darkest half (hazard ratio 1.53), with a dose-response across intermediate groups. The association was independent of polygenic risk, and the difference between a bright night and a dark night was roughly equivalent to the difference between low and moderate genetic risk.

Light and mortality. In 88,905 people with 3,750 deaths over eight years, brighter days predicted lower all-cause mortality — 17 to 34% lower in the brightest decile — while brighter nights predicted higher mortality, 21 to 34% higher in the brightest decile. The associations were strongest for cardiometabolic deaths.

A separate prospective study of 43,722 women found that sleeping with a television or light on in the room was associated with a 17% higher risk of gaining 5 kg or more and a 33% higher risk of becoming obese over five to six years, after adjustment for sleep duration, diet and activity. Exposure here was self-reported, and all of these are observational findings: they cannot prove that light causes the outcomes. But the direction matches the physiology precisely, and the sensor-based studies are hard to explain away as recall bias.

Morning Light as Therapy

Bright light in the morning has moved beyond circadian hygiene into treatment. A 2025 meta-analysis of 11 randomised trials in 858 people with non-seasonal depression found that adding bright-light therapy raised remission from 23.5% to 40.7% — odds ratio 2.42 — and response from 38.6% to 60.4%. An earlier meta-analysis found a standardised effect of −0.62, larger in trials of two to five weeks. Blinding is imperfect in light trials, which tempers the certainty, but the effect size is in the range of antidepressant medication.

For delayed sleep phase, the picture is more complicated and more interesting. A randomised trial in 40 young adults crossed bright morning light against dim light and melatonin against placebo, with everyone also advancing their rise time gradually. Over two weeks, the melatonin onset advanced by about two hours — in all four groups. The gradual rise-time advance alone did the work. Light and melatonin mattered for maintaining the shift over three months of follow-up, not for creating it. The lesson: a fixed, gradually earlier wake time is itself a light intervention, because it determines when your eyes first meet the day.

Blue-Blocking Glasses: The Honest Verdict

Amber lenses are the product most associated with "circadian light", and the evidence for them is the thinnest part of this article.

A 2023 Cochrane review identified 17 randomised trials of blue-light-filtering spectacles. No meta-analysis was possible. For sleep quality, six trials with 148 participants between them gave very low-certainty evidence: three positive, three null. Not one trial measured serum melatonin. Sixty-five percent were at high risk of detection bias. Adverse events were infrequent but included headache, low mood and depressive symptoms.

The positive trials are small. A randomised crossover in 14 people with insomnia symptoms found that amber lenses worn for two hours before bed improved insomnia severity and increased sleep time by actigraphy over one week. A narrative systematic review of 29 publications concluded there was substantial evidence for shorter sleep onset in sleep disorders, jet lag and shift work — but it pooled nothing and the authors' confidence outran the trials.

The reasonable position: amber lenses reduce the blue component of light reaching the melanopsin cells, which is mechanistically sound. If you cannot control your evening environment — a shared living room, a night shift, a hotel — they are a plausible fallback. They are not a substitute for turning the lights down, and no trial has shown they are.

The Other Clock Signals Light Interacts With

Light is dominant but not alone, and two laboratory findings put the others in perspective. Caffeine equivalent to a double espresso, taken three hours before habitual bedtime, delayed the melatonin rhythm by about 40 minutes — roughly half the delay produced by three hours of bright evening light. And exercise has its own phase-response curve: in 99 adults, moderate treadmill exercise at 07:00 or between 13:00 and 16:00 advanced the clock, while exercise between 19:00 and 22:00 delayed it, with an amplitude comparable to bright light of the same duration. Evening coffee and evening training both pull in the same direction as evening screens. How to line all of them up is the subject of the protocol article.

What To Actually Do

1. Within 30 minutes of waking, get outside for 20–30 minutes. Overcast counts. Through a window does not count nearly as much: glass and distance cut melanopic light sharply.
2. Work near a window, or add a bright light at eye level during the day. The target is 250 lux melanopic at the eye; a 10,000 lux light box at arm's length for 30 minutes in the morning is the therapeutic version.
3. From three hours before bed, drop below 10 lux. Turn off overhead lights. Use low, warm lamps. Set screens to their warmest, dimmest setting, or stop using them.
4. Sleep in darkness. Below 1 lux. Blackout curtains or a sleep mask; no standby lights at eye level.
5. Keep your wake time fixed. It is the hinge on which your daily light exposure turns.
6. Use melatonin only to move a late clock, and time it to the afternoon. The dosing argument is in Melatonin: Optimal Dosing and the Low-Dose Debate.

Frequently Asked Questions

Is it really blue light, or just brightness? Both. Melanopsin is most sensitive in the blue, so blue-enriched light has more circadian punch per lux. But at high enough intensity any white light suppresses melatonin, and dim red light at very low lux barely registers. That is why the recommendations use melanopic EDI rather than plain lux.

My phone has a night-mode filter. Is that enough? It helps, by shifting the spectrum away from the melanopsin peak. Brightness still matters, and the 2019 dose-response study found measurable suppression at 10 lux. Warm filter plus minimum brightness plus distance from the face is the combination.

Does morning light work through a window? Weakly. Glass cuts intensity and most windows are not in your line of sight. Ten minutes outside is worth more than an hour at a desk by a window.

How long until I notice a difference? The camping studies suggest days, not weeks. In the delayed-phase trial, two hours of advance took two weeks of a gradually earlier rise time.

Are light boxes safe? At 10,000 lux and UV-filtered, yes for most people. Anyone with bipolar disorder, retinal disease or on photosensitising medication should use them under supervision.

Where does this fit with supplements for sleep? Upstream of all of them. Magnesium L-threonate, L-theanine and the compounds in the deep-sleep protocols work on sleep depth and arousal. None of them moves a clock that light has set to the wrong time.

Related Research

• The Circadian Optimization Protocol: Light + Melatonin + Timing
• Melatonin: Optimal Dosing and the Low-Dose Debate
• The Circadian Optimization Protocol
• Melatonin
• Deep Sleep Architect