Light and the body clock

Section 01

The mismatch

Human biology evolved under a light-dark cycle set by the sun. For most of our history the daily signal was strong and unambiguous, with bright, blue-enriched daylight through the middle of the day giving way through amber dusk to a night that was close to dark. The body’s master clock uses that swing in intensity and spectrum to coordinate timing across almost every organ system, from hormone release and immune regulation to the structure of sleep and the clarity of thought.

Modern life has flattened the swing. A large survey of daily activity in the United States found that people spend around 90 per cent of their time indoors¹. The share varies with latitude, season and occupation, but wherever it has been measured the direction is the same. Indoors, the light that reaches the eye tends to be too dim during the day to give the clock a clear signal, and too bright in the evening to let it settle.

In 2022 an international group of researchers led by Brown and colleagues published consensus recommendations for healthy indoor light exposure². They expressed their thresholds in melanopic equivalent daylight illuminance (melanopic EDI), a measure of the stimulus that light gives the eye’s circadian photoreceptors rather than how bright it looks. Those thresholds are the yardstick for the rest of this paper.

Why going outside is not the whole answer

The obvious response is that people should spend more time outdoors. Morning daylight is the strongest signal the clock receives, and when people get enough of it the effect can be striking. In one study, a week of camping under natural light with no electric lighting moved participants’ melatonin onset about two hours earlier and closed most of the gap between early and late types³.

The difficulty is that strong entrainment of this kind is a by-product of the environment rather than a deliberate habit. A knowledge worker in London or Tokyo may travel underground, work in a deep-plan office and come home after dark for much of the year, and advice to get more morning light runs into the constraints of building design and working hours. If the problem lies in the environment, the most reliable remedy is to change the light a person receives in the places where they already spend their days.

Section 02

Too little light by day

On a clear day, outdoor daylight delivers between about 10,000 and 100,000 lux at ground level, and even an overcast sky provides 1,000 to 10,000 lux⁴. Indoor lighting standards were written for a different purpose. The European standard EN 12464-1 asks for at least 500 lux on the task area for typical office work⁵. That figure is measured on the desk rather than at the eye, and it says nothing about spectrum or the circadian system, although newer frameworks such as the WELL Building Standard have begun to add circadian criteria expressed in melanopic terms⁶.

Offices

A study of Indian office buildings by Green Business Certification Inc. found that 64 per cent of workspaces were below recommended illuminance, some at under 100 lux at the workstation, and that only 10 per cent met the recommended melanopic threshold in the morning⁷. Light dosimetry, the use of wearable sensors to record the light that reaches a person’s eyes, tells a similar story. A 2025 meta-analysis of dosimetry studies in offices found many workers below 250 melanopic EDI for most of the working day⁸, and even a modern London office with daylight-responsive lighting recorded personal exposure that varied a great deal with where people sat and how they behaved⁹.

Some of the causes are structural. Deep floor plates push desks away from windows, and blinds lowered once against a passing glare often stay down for the rest of the day, a pattern the GBCI study identified as one of the main reasons buildings with good daylight potential delivered little daylight in practice⁷. Its most telling finding is that 76 per cent of occupants said they were satisfied with their lighting despite these conditions⁷. The circadian system sends no conscious signal when it is under-stimulated. People do not feel short of light in the way they feel hungry or cold, and the consequences build up over weeks and months in ways that are hard to trace to any single cause.

Homes and home offices

The evidence from homes is thinner but points the same way. A 2021 systematic review from the MRC Centre for Environment and Health screened 4,043 records and found only 28 studies that met its criteria, a sign of how little attention home lighting has received¹⁰. Daylight at home appears to matter. In a small crossover study, greater access to circadian-effective daylight at home was associated with earlier sleep onset, more regular sleep and better mood¹¹. The move to remote and hybrid work has enlarged the problem. In 2023, 35 per cent of employed people in the United States did some or all of their work at home on the days they worked, up from 24 per cent in 2019¹², and without a commute many may lose the incidental outdoor light that travel once provided.

Care homes

The most severe deficits occur in care homes, where residents with limited mobility spend almost all their time indoors. In one study of 66 nursing home residents, the median light level through the day was just 54 lux, and residents spent a median of only 10.5 minutes a day above 1,000 lux¹³. Dosimetry in Norwegian dementia units has since found median vertical illuminance below 300 lux in living rooms even in summer¹⁴. The UK’s National Institute for Health and Care Research has called for a multi-site randomised trial of light therapy in care homes¹⁵.

Section 03

Too much light at night

The second half of the mismatch comes after dark. The Brown consensus recommends no more than 10 melanopic EDI from at least three hours before bedtime, and no more than 1 melanopic EDI during sleep². Ordinary domestic lighting can exceed those levels with ease. In a laboratory study, typical room light before bedtime suppressed melatonin, delayed its onset in 99 per cent of participants and shortened the melatonin period by about 90 minutes compared with dim light¹⁶. Dose-response work has shown that light of around 100 lux at night produces close to half the maximum effect on melatonin¹⁷.

Large cohort studies link this exposure to health. In the HEIJO-KYO cohort of older Japanese adults, bedroom light above about 3 lux at night was associated with higher rates of dyslipidaemia, abdominal obesity and raised night-time blood pressure, and bedroom light above about 5 lux with a higher incidence of depressive symptoms over follow-up¹⁸. A 2025 meta-analysis of eight observational studies covering more than 205,000 people found that higher exposure to light at night was associated with 22 per cent higher odds of depression, although results varied a good deal between studies¹⁹.

The UK Biobank adds scale. Burns and colleagues analysed wrist-worn light sensor data from 86,772 participants and found that greater night-time light exposure was associated with a higher risk of psychiatric disorders, while greater daytime exposure was associated with a lower risk, each independent of the other²⁰. In a prospective analysis of 88,905 participants, people in the brightest fifth for night-time light had higher risks of coronary artery disease, heart attack, heart failure, atrial fibrillation and stroke than those in the darkest fifth²¹. The adjusted hazard ratios ranged from 1.28 to 1.56, which in plain terms means risks between 28 and 56 per cent higher.

These are observational findings. People who sleep in brighter rooms may differ in other ways, and the studies can adjust for such differences only as far as their data allow. Correlation is not causation. Even so, the direction holds across countries and study designs, and it fits what laboratory work shows about how night-time light acts on the clock.

The wider cost

Circadian disruption sits within a wider set of health costs linked to poor sleep. In a cohort of more than 81,000 adults with pre-diabetes, insomnia was associated with a 28 per cent higher risk of developing type 2 diabetes²². The International Agency for Research on Cancer classifies night shift work as probably carcinogenic to humans (Group 2A), with circadian disruption among the proposed mechanisms, a classification that concerns night shift work rather than ordinary evening light²³.

These figures describe severe disruption in people with insomnia, shift work or chronic short sleep. They do not capture the much larger group who sleep a normal number of hours but whose daily rhythm is weakened by dim days and bright evenings, and the sections that follow explain why that group matters.

Section 04

A primer on the body clock

Circadian rhythms are among the oldest timing systems in biology. Cyanobacteria, some of the earliest forms of life, keep a rhythm of about 24 hours, and versions of a cellular clock run in plants, insects and mammals alike. In humans, a clock operates in nearly every nucleated cell. Our understanding of it was built over three centuries through a handful of decisive experiments.

Bünning’s result mattered because the period was close to, but not exactly, 24 hours. A clock that matched the length of the day could be mistaken for a passive response to light and dark, whereas a clock that drifts can only be explained by an internal mechanism. The molecular explanation came from fruit flies. The protein made by the period gene builds up through the night, enters the cell nucleus and switches off its own gene, and as levels fall the gene switches back on and the cycle begins again²⁷. Further genes, including Clock in mammals, revealed interlocking loops that give the clock both stability and the ability to reset³¹. When the Nobel Assembly recognised this work, it noted that a large proportion of the body’s genes are regulated by the biological clock²⁹.

Where light comes in

The clocks in the body’s tissues need a conductor, and the conductor needs a signal from the outside world. For most of the twentieth century that signal was assumed to come through the rods and cones that form images, until the discovery of intrinsically photosensitive retinal ganglion cells, or ipRGCs, which contain a photopigment called melanopsin. Russell Foster’s group at Oxford was central to this work²⁸. Melanopsin is most sensitive to short-wavelength light, with a peak at about 480 nanometres in the blue-cyan part of the spectrum³².

These cells play little part in forming the image we see. They measure the intensity and spectrum of ambient light and send that information to the suprachiasmatic nucleus (SCN), a paired cluster of about 20,000 neurons in the hypothalamus that acts as the body’s master pacemaker³³. Left without time cues, the human SCN runs a little long, at an average period of 24.18 hours³⁴. Because this does not match the solar day, the clock has to be corrected every day. That daily correction is called entrainment, and light is its strongest driver. Bright light in the morning moves the clock earlier, while bright light in the evening moves it later.

Once entrained, the SCN coordinates timing across the body. It governs the release of melatonin, which rises in darkness to mark the biological night and is suppressed by light. The timing of that evening rise, known as dim-light melatonin onset, is the standard laboratory marker of circadian phase. It also shapes the rise in cortisol soon after waking and the daily cycle of core body temperature, which under controlled conditions swings by about 1 degree Celsius³⁵.

Key concepts

Chronotype describes a person’s preferred timing for sleep and activity. It is in part inherited and shifts across life, becoming later in adolescence and earlier in older age³⁶. Circadian amplitude is the strength of the daily oscillation, the difference between peak and trough in markers such as melatonin, body temperature or the pattern of rest and activity. Timing and amplitude carry most of the argument that follows.

Section 05

Entrainment is a spectrum

It is tempting to think of circadian health as binary, with clocks that either work or have failed, and of light as a treatment reserved for the second group. The evidence describes something different. Entrainment sits on a continuous spectrum. At one end is a strong rhythm with consistent sleep timing and a clear melatonin onset, and at the other a flattened rhythm with fragmented sleep and peripheral clocks out of step. Most people sit somewhere in between, with clocks that are running but without much strength.

Amplitude, mood and cognition

The largest study of this question used accelerometer data from 91,105 UK Biobank participants to derive relative amplitude, a measure that compares a person’s most active ten hours of the day with their least active five hours of the night³⁷. Each one-fifth step down in relative amplitude was associated with 6 per cent higher odds of lifetime major depressive disorder and 11 per cent higher odds of bipolar disorder, along with greater mood instability, more loneliness, lower happiness and slower reaction times, after adjustment for age, sex, deprivation, smoking, activity level, body mass index and other factors³⁷.

This was not a study of people with diagnosed illness. It was a general-population sample, and the gradient ran right through it. Someone in the second fifth for amplitude is not unwell, but on average they report a little less happiness and react a little more slowly than someone in the top fifth. The study was cross-sectional, so it cannot show whether weaker rhythms cause these outcomes or result from them, and both directions are plausible.

The CARE study extended the finding to thinking and memory, using a wearable-derived measure of circadian amplitude that correlated moderately with melatonin amplitude in healthy volunteers (r = 0.46) to examine cognition in 1,703 Chinese adolescents and 92,202 UK Biobank adults³⁸. Higher amplitude was associated with better executive function in the adolescents and with better reasoning and memory in the adults.

Beyond moving the clock

Most research on light has focused on shifting the clock’s timing, which matters for jet lag and for people whose sleep runs very late. In a 2025 paper in the journal Sleep, Walch and colleagues argued that strengthening the rhythm deserves equal attention, a distinction that changes who might benefit³⁹. Timing interventions are relevant to people with clear timing problems, whereas amplitude is relevant to anyone whose rhythm has been compressed by dim days and bright evenings. On the evidence in Sections 2 and 3, that group may include most indoor workers.

Many signals, one rhythm

The associations reach beyond the brain. In 2025 the American Heart Association published a scientific statement on the role of circadian health in cardiometabolic health and disease risk⁴⁰. Light sets the master clock, but clocks in the liver, gut and muscle also respond to when a person eats and moves. In a randomised trial in people with metabolic syndrome, confining eating to a daily window reduced HbA1c by about a tenth of a percentage point, a modest but measurable change in glucose control⁴¹. Among nearly 61,000 UK Biobank participants, the regularity of the sleep schedule was a stronger predictor of mortality risk than its length, an observational finding that points to consistency as a health variable in its own right⁴².

Roenneberg and Merrow describe entrainment as the product of many time cues acting at once³⁶, and modelling work suggests that when those cues are aligned the resulting rhythm is more robust than any single cue can achieve⁴³. That is a modelling result rather than population data, but it supports an approach in which light is the anchor and regular routines reinforce it.

Age

Amplitude tends to decline with age, as the lens of the eye yellows and passes less short-wavelength light and as older adults spend less time outdoors. Randomised trials described in Section 7 show that bright light can partly restore circadian function in older people, with measurable effects on mood and sleep^44,45. A person does not need to be ill to stand to gain from a stronger rhythm.

Section 06

What each wavelength does

The circadian system does not respond to light as a single quantity. Visible light runs from about 380 nanometres (violet) to 780 nanometres (red). The visual system relies on cones and rods, while the circadian system relies on melanopsin, whose sensitivity peaks near 480 nanometres³². Because the two systems weight the spectrum in different ways, a lamp can look warm and dim yet still stimulate the clock, or look bright and white yet stimulate it much less than expected. Melanopic EDI, defined in CIE S 026, captures this by weighting light according to melanopsin’s sensitivity rather than the eye’s sense of brightness³⁰.

Blue-enriched white light

Light rich in the 460 to 490 nanometre region is the most potent driver of circadian responses. In a laboratory comparison at equal photon counts, 460 nanometre light suppressed melatonin more than 550 nanometre light and produced greater alertness, both reported and measured by EEG⁴⁶. In care homes, upgraded lighting that combined blue-enriched daytime light with low-level night lighting was associated with a 43 per cent lower rate of falls across 758 residents and more than 126,000 resident-days⁴⁷. That study was not a randomised trial, so other differences between settings may have contributed.

Broad-spectrum bright light

Many landmark studies used broad-spectrum white light rather than light tuned toward blue. The standard clinical protocol for seasonal affective disorder is 10,000 lux from a light box for about 30 minutes in the morning⁴⁸. Broad-spectrum sources work because at high intensity they deliver enough melanopic stimulation even when the spectrum is not optimised for melanopsin, at the cost of needing more total light than a source weighted toward the blue-cyan region.

Warm amber light

In the evening the goal reverses. The circadian system should receive as little short-wavelength stimulation as possible so that melatonin can rise, and the consensus recommends evening light with a spectrum depleted in short wavelengths². Warm light below about 3,000 kelvin concentrates its energy in the amber and red part of the spectrum, where melanopsin is least sensitive, and so delivers less circadian stimulation for the same visual brightness. Compared with reading a printed book, reading on a light-emitting device before bed suppressed melatonin, delayed circadian timing and lengthened the time taken to fall asleep⁴⁹.

Red and near-infrared light

Light at the red and near-infrared end of the spectrum, from about 650 to 1,000 nanometres, does not stimulate melanopsin and has no direct effect on the clock. It may act on tissue through a separate process called photobiomodulation, in which red and near-infrared photons are absorbed by cytochrome c oxidase, an enzyme in the mitochondria. Work from Glen Jeffery’s group at UCL has shown that 670 nanometre light can shift the redox state of this enzyme in the retina⁵⁰, and that a single exposure improved colour contrast sensitivity in people for up to a week⁵¹. These are early findings. Evidence in healthy people is limited, and the right doses have not been established.

The daylight benchmark

Natural daylight has a broad spectrum that changes across the day, with strong blue content at midday and warmer light at dawn and dusk. A clear day outdoors delivers tens of thousands of lux⁴, while office standards ask for around 500 lux on the desk⁵. The gap between the light the clock evolved under and the light most people now receive indoors is measured in orders of magnitude, and no indoor lamp closes it. What indoor light can do is reach the consensus threshold of at least 250 melanopic EDI at eye level at the times of day when it matters most².

Section 07

The evidence for light as an intervention

This section follows the usual hierarchy of evidence, starting with meta-analyses of randomised controlled trials, then landmark individual trials, then large population studies. Most of the studies used bright light therapy delivered on fixed protocols.

How to read the effect sizes

Meta-analyses often report standardised effects such as Hedges’ g or the standardised mean difference (SMD). By the conventions used in behavioural science, a value of about 0.2 is small, 0.5 is medium and 0.8 is large. A medium effect means the average person in the treated group ends up better off than about two-thirds of the control group. For outcomes such as depression or agitation, where lower scores are better, a negative value indicates improvement.

Meta-analyses

The most comprehensive pooled analysis in dementia, by Aini and colleagues, combined 24 randomised trials with 1,074 participants and found medium-sized improvements in sleep quality, circadian amplitude, depression and agitation⁵². Other reviews agree on sleep but not on everything else. Fong and colleagues found fewer night-time awakenings but no significant effect on agitation or depression⁵³. In mood disorders the pooled effect of bright light is large for seasonal depression⁵⁴ and small to medium for non-seasonal depression⁵⁵, and a randomised trial in non-seasonal major depression found bright light effective both on its own and alongside the antidepressant fluoxetine⁵⁶. The table summarises the main pooled results.

Landmark trials

The Riemersma-van der Lek trial, published in JAMA in 2008, tested bright light and melatonin, alone and together, in 189 residents of 12 group care facilities in the Netherlands for up to three and a half years, the longest controlled light study in the literature⁴⁴. Whole-day light of about 1,000 lux slowed cognitive decline by a relative 5 per cent, reduced depressive symptoms by 19 per cent and slowed the increase in functional limitations by 53 per cent.

The Lieverse trial, published in 2011, was double-blind and placebo-controlled. In 89 older adults with non-seasonal major depression, three weeks of bright pale blue light at 7,500 lux improved depression scores by 21 per cent more than a dim red placebo. Evening cortisol fell by 37 per cent, the rise in melatonin became 81 per cent steeper, and the benefits persisted three weeks after treatment ended⁴⁵. Both trials show durable effects with measurable changes in the hormones that track the clock. The Lieverse trial received donated equipment from a lighting manufacturer, a common arrangement in this field.

Alertness and cognition in healthy adults

The evidence extends beyond clinical groups. A meta-analysis of 28 studies in healthy people found that daytime electric light improved alertness and cognitive performance and reduced EEG alpha activity, a physiological marker of drowsiness, but found no pooled effect on mood⁵⁹. In a controlled crossover study of 104 office workers, four weeks of blue-enriched white light improved self-reported alertness, performance and night-time sleep quality⁶⁰. In a 14-week controlled study in a UK call centre, staff on a floor switched to very high colour temperature lighting reported a 36.8 per cent improvement in concentration and a 28.1 per cent improvement in alertness relative to an identical control floor⁶¹. Those outcomes were self-reported, and the study was funded by an educational grant from a lighting manufacturer. In a crossover study of 19 people working simulated night shifts, circadian-informed lighting produced about half as many attention lapses by the final night as a dim control⁶².

Not every workplace study is positive. A dynamic lighting scheme designed to follow natural daylight did not improve alertness or cognition in well-rested office workers⁶³, and a field study in a working office found lower sleepiness under brighter midday light but no clear effect on task performance or measured sleep⁶⁴. In healthy, rested adults the effects appear smaller and less consistent than in clinical groups.

Population-level evidence

The UK Biobank offers a scale that no trial can match. In more than 400,000 participants, more time spent in outdoor light was associated with lower odds of major depressive disorder, less antidepressant use and greater happiness⁶⁵, and the later sensor study found daytime light associated with lower psychiatric risk independent of night-time exposure²⁰. These studies cannot establish cause, since time outdoors travels with physical activity, social contact and vitamin D, but their scale and their agreement with the trial evidence give them weight.

The studies cited here come mostly from Europe, North America and East Asia. Consistency across these settings is expected, because the SCN and melanopsin are shared across human populations, and the scarcity of controlled studies from Africa, South Asia, the Middle East and Latin America is a gap in the literature rather than evidence that light acts in another way there.

What the evidence supports

Taken together, the evidence supports a clear but bounded claim. Bright light produces small to medium improvements in sleep quality, circadian amplitude, mood and behavioural symptoms across several populations, with the strongest and most replicated findings in older adults and people living with dementia. Evidence for alertness and cognition in healthy adults is consistent in direction but less mature, and some well-designed studies have found no effect. Most of this evidence comes from fixed protocols such as 30 minutes at 10,000 lux or whole-day light at about 1,000 lux. Personalised, adaptive protocols have not yet been tested in published randomised trials.

Section 08

Measuring is not the same as changing

Consumer sleep technology has grown up around measurement, with rings, watches and bedside apps that turn sleep stages, heart rate and movement into nightly scores. That information has value. A person who learns that they wake at 3am most nights, or that their deep sleep has declined over several months, knows something they did not know before. The data alone do not change the environment that caused the problem, though. The user still has to interpret them, form a plan and keep to it for weeks.

Sustained behaviour change through a screen is hard. Eysenbach’s law of attrition described how many of the people who start digital health programmes stop using them over time⁶⁶. A systematic review of app-based interventions for chronic disease found a pooled dropout rate of 43 per cent⁶⁷, and a meta-analysis of 117 trials of digital mental health interventions found that engagement was associated with features such as human support and personalisation⁶⁸.

Light boxes sit at the other end of the market. They deliver the intervention that the trials in Section 7 support, but at a fixed intensity and spectrum, at a prescribed time and distance, and with no knowledge of how the user slept or how much light the room already provides. They work for people motivated enough to sit in front of one every morning, and the attrition literature suggests that an intervention asking less of the user is more likely to be used by everyone else.

Aryal is designed to close the loop between the two. It senses sleep without anything worn on the body, reads the light already present in the room and adjusts its own light across the day in response. A person can open the companion app to look at their sleep and circadian trends, but the light programme runs whether they do or not.

Section 09

Personalisation and the Circadian Fingerprint

The trials in Section 7 gave every participant the same light at the same time. Whether timing and dose matched to the individual could do better is a fair question, and the case for trying rests on well-established differences between people.

Chronotype is the first. People differ in their natural timing by several hours, and the clock’s response to light depends on when in a person’s internal day the light arrives³⁶, so a morning pulse that suits an early type may land at the wrong biological time for a late one. Sensitivity is the second. In a study of evening light, the amount needed to suppress melatonin varied by about 50-fold between individuals⁶⁹. Age adds further variation. Younger adolescents show stronger melatonin suppression by evening light than older adolescents⁷⁰, and in a small study the suppression seen in children was about twice that seen in adults⁷¹. Recent light history may matter as well, since a person who spent the day outdoors may respond less to the same indoor light than someone who spent it in a dim office.

A fixed protocol is, at best, tuned to the average participant, and for people at either end of these distributions it may deliver light at the wrong time, intensity or duration. The closest published analogue to a tailored approach is a randomised crossover trial in 46 people living with dementia. Lighting tailored to deliver a high circadian stimulus during the day improved scores on the Pittsburgh Sleep Quality Index from 10.3 to 6.7, where scores above 5 indicate poor sleep, and also improved depression and agitation⁷².

The Circadian Fingerprint

The Circadian Fingerprint is Aryal’s term for the personal circadian profile it builds over time. Each night Aryal senses sleep without contact, noting when it begins and ends and how it is structured. By day it reads the light already in the room, and over time it notes whether a given day’s light was followed by better or worse sleep. As nights accumulate it learns the person’s habitual timing, how their working days differ from their days off and how they respond to light, and it uses that profile to shape the next day’s light.

Aryal uses a neurosymbolic approach, which combines knowledge drawn from the published research with learning from each person’s own data, to build that profile and to narrow its suggestions to the few that matter rather than multiplying them. The profile is not static. It changes as a person’s life does, with a new work schedule, the turn of the seasons, travel across time zones or the gradual changes that come with age. The consequence is the one this paper has been building toward. The environment adapts to the person, and the person does not have to do anything.

The validation gap

Each component of this approach rests on published evidence. What has not yet been tested in a published randomised trial is the integrated system, a personalised protocol informed by continuous sleep data. Controlled studies that compare personalised light with fixed protocols and with no added light are the obvious next step, and Aryal intends to contribute to them.

Section 10

Aryal across the 24-hour day

Aryal is a portable daylight lamp with contactless sleep sensing built in, designed to sit on a bedside table, a desk or a shelf and to move with its owner from room to room. The design is domestic rather than clinical. An optical diffuser softens the light for home use, while the output is designed to reach the consensus daytime threshold of at least 250 melanopic EDI at eye level at the intended distance². Aryal does not try to replicate sunlight. Its aim is to deliver the part of daylight that matters most to the clock, at the right time and in sufficient amount.

Morning

The circadian system is most responsive to bright, blue-enriched light in the hours after waking, when such light suppresses residual melatonin and promotes alertness⁴⁶. Small studies have found that light in the final half hour of sleep improves alertness and performance after waking⁷³, and that dawn-simulating light is associated with higher morning cortisol and better mood and wellbeing⁷⁴. Aryal’s morning programme delivers blue-enriched white light at or above 250 melanopic EDI, timed to the person’s actual wake time as sensed the night before.

Daytime

Sustained light through the day supports circadian amplitude and daytime alertness, as the office trials in Section 7 suggest⁶⁰, and for people working at home daytime light may be poorer still. During working hours Aryal keeps melanopic levels raised and adjusts its output to the light already in the room, giving more when the room is dim and less when daylight is plentiful.

Evening and night

In the evening the aim reverses, and the consensus recommends no more than 10 melanopic EDI in the three hours before bed². Aryal’s evening programme moves by degrees from neutral white toward warm amber, much as natural dusk does, leaving enough light to read and talk by without the short-wavelength content that holds back melatonin¹⁶. During sleep the circadian system needs darkness, and Aryal senses sleep without emitting any light. If it detects that the person has woken for a prolonged period, it can provide a brief, very dim warm light, kept below the consensus ceiling of 1 melanopic EDI for sleep².

The cumulative effect

No single part of this cycle is transformative on its own. The value is meant to build over weeks and months, moving a person from moderate toward robust entrainment, the shift that the UK Biobank data associate with better mood and sharper cognition³⁷. Whether a personal device produces that shift in everyday life is what the validation work in Section 9 needs to show.

Where Aryal is used

Because it is portable, Aryal can follow a person through the day, from the bedroom in the morning to a desk or home office during working hours and back again in the evening. It can travel too. In one study, three days of morning bright light before eastward travel advanced the clock by about 2.1 hours, compared with 0.6 hours under dim light⁷⁵. A personal light source can be carried into a hotel room, a rented flat or a temporary office, places where the built environment is hard to change.

Section 11

Safety and regulatory position

Aryal is a consumer wellness product. It is not a medical device, it is not registered with any medical device authority, and it is not intended to diagnose, treat, cure or prevent any disease. Its suggestions are framed as personal guidance for general circadian health, are designed never to make diagnostic statements or give treatment advice, and draw on published evidence with oversight from scientific advisers.

Light

Bright light is safe for most people, but it is not trivial. Aryal’s light output must meet IEC 62471, the international standard for the photobiological safety of lamps, which classifies light sources by their potential to harm the eye and skin. At bedside and desk distances its output does not exceed the intensities used in published clinical studies of bright light. Side effects reported in clinical studies of bright light therapy are mild and short-lived in most cases, the commonest being headache, eye strain and nausea, and they tend to settle within the first few days⁴⁸.

Bright light is not suitable for everyone. People with retinal disease and people taking medicines that increase sensitivity to light should seek medical advice before using bright light, as should people with bipolar disorder, in whom bright light at the wrong time can trigger mania⁴⁸. That last risk is one reason the timing of light matters as much as its strength.

Radio and electrical safety

Aryal’s contactless sensor is a low-power, short-range radio-frequency sensor. Like any consumer product with a radio component, Aryal must meet electrical safety, electromagnetic compatibility and radio equipment requirements in each market where it is sold, including the relevant UK and EU directives and FCC rules in the United States.

Data

Aryal collects sleep data and data about the room. It processes data on the device where possible and uses cloud processing for the personal circadian model. Data processing is designed to comply with UK and EU GDPR, and users retain ownership of their data.

Section 12

Myths, limits and open questions

A credible scientific argument is open about what it does not know. This section corrects some common misconceptions about circadian science and light, and sets out the main gaps in the evidence.

Myth: screens are the main problem

Screens do emit light in the range that melanopsin responds to, and evening use of light-emitting devices can delay the clock⁴⁹. The intensity at the eye is modest, though. One laboratory measurement of a large LED-backlit monitor at normal viewing distance found about 35 to 50 lux at the eye⁷⁶. Ordinary room lighting before bed is enough by itself to suppress melatonin¹⁶, so the screen is seldom the only source worth considering. Timing matters more than wavelength alone. Blue-enriched light at 7am supports the clock, and the same light at 11pm works against it.

Myth: everyone needs 10,000 lux

The 10,000 lux figure comes from the clinical protocol for seasonal affective disorder, a clinical dose for a clinical population⁴⁸. The consensus minimum for healthy adults is 250 melanopic EDI by day², and a source weighted toward the blue-cyan region can reach that level at far lower total output than a broad-spectrum light box needs to reach 10,000 photopic lux.

Two numbers worth checking

The claim that people spend 90 per cent of their time indoors comes from a US survey of daily activity in 9,386 people¹. It is robust as a US average for the period studied but has not been replicated at the same scale elsewhere, and the exact share varies with latitude, season, occupation and age. Popular accounts also sometimes say the human clock runs at 25 hours. Those figures come from early isolation studies with incomplete control of light exposure, and under careful control the average period in healthy adults is 24.18 hours³⁴.

What the science does not yet know

The consensus gives thresholds, but the dose-response relationship above them is not well characterised. It is not known whether 500 melanopic EDI gives twice the benefit of 250, or whether the benefit levels off². Individual variation in the response to light has not been measured in samples large enough to build reliable personalisation from published data alone, which is why Aryal learns from each person’s own data, an approach still to be validated in a clinical trial.

The longest controlled light trial ran for three and a half years⁴⁴. There are no trials of five years or more, and the long-term effects of sustained, optimised light in healthy people are unknown. Some well-designed studies have found no benefit. A 24-week cluster randomised trial in 61 people with severe dementia found that ambient bright light did not improve rest-activity rhythms overall⁷⁷, and pooled analyses differ on whether light improves mood and agitation in dementia^52,53. Several influential trials were supported in part by lighting manufacturers. None of this reverses the overall direction of the evidence, but it argues for care in how far that evidence is extended.

The mechanistic evidence for red and near-infrared photobiomodulation is promising, but clinical evidence in healthy people is at an early stage^50,51, and Aryal makes no claims for red or near-infrared light. The benefit of personalised, adaptive light at scale also remains to be shown.