Linda Geddes decided to live for weeks in only candlelight – no bulbs, no screens. Along the way, she discovered simple things that everyone can try to sleep and feel better.

We spend a third of our lives either sleeping… or trying to get to sleep. But in the world of 24/7 living and artificial light, our sleep is increasingly under threat. 

Many of us don’t get the recommended seven to nine hours we need each night and struggle to get up in the mornings – especially on work days. But it isn’t only our quantity of sleep that’s affected. Since the discovery that light (particularly blue light, emanating from devices like smartphones) can affect our biological clocks, evidence has been building that exposure to even low levels of light in the evening or at the night is disrupting our sleep quality, as well.

So what would happen if we turned off the lights? Would it improve our sleep or have other benefits? And how easy would this be to achieve in a modern city?

You might also like:
• How much sleep do we really need?
• The vital time you shouldn’t be on social media
• Why you shouldn’t try to be a morning person

One recent winter, I decided to find out. Working with sleep researchers Derk-Jan Dijk and Nayantara Santhi at the University of Surrey, I designed a programme to go cold-turkey on artificial light after dark, and to try to maximise exposure to natural light during the day – all while juggling an office job and busy family life in urban Bristol.

The discoveries I've made have revolutionised my attitude to light ­­­– and how I live my life during the night and day. I now make simple, daily choices that can transform how I sleep, how I feel and perhaps even my cognitive abilities. Could you be doing the same?   

For millennia, humans lived in synchrony with the natural cycle of light and dark. This doesn’t mean that everyone went to sleep as soon as the Sun set. Studies of pre-industrial societies, such as tribes living in Tanzania or Bolivia today, suggest that people stay up for several hours after dark, often socialising by firelight. In fact, the amount of sleep they get is quite like people in industrialised countries, but the timing is more in line with the natural cycle of day and night: they tend to go to bed earlier and get up just before dawn. 

“In modern societies, at least on weekdays, we don’t sleep in tune with our body clock,” says Dijk. Exposure to artificial light at night is shifting our body clocks later. But we still need to go to work in the morning, so we set an alarm clock – even though the biological clock says we should still be asleep.

Pre-industrial societies such as the Hadza tribe in Tanzania also seem to have a far lower prevalence of sleep problems, like insomnia. “When we asked members of the Hadza whether they thought their sleep was good, they almost universally said ‘yes, it’s totally fine’. That statistically doesn't match up with what we see in the West,” says David Samson, an anthropologist at the University of Toronto in Mississauga, who has studied them.

Light has a powerful non-visual effect on our body and mind, something to remember when we stay indoors all day and have lights on late into the night – Nayantara Santhi

Why is this? Light enables us to see, but it affects many other body systems as well. Light in the morning advances our internal clock, making us more lark-like, while light at night delays the clock, making us more owlish. Light also suppresses a hormone called melatonin, which signals to the rest of the body that it’s night-time – including the parts that regulate sleep. “Apart from vision, light has a powerful non-visual effect on our body and mind, something to remember when we stay indoors all day and have lights on late into the night,” says Santhi, who previously demonstrated that the evening light in our homes suppresses melatonin and delays the timing of our sleep.

However, light also boosts alertness in its own right. It’s like drinking a double espresso. Although these stimulant effects are bad news if you’re trying to sleep, being exposed to more bright light during the daytime could make us more alert. Light also stimulates brain regions that regulate mood.

“The important thing is that we create a light exposure pattern with sufficient light during the day, and not too much light in the evening,” says Dijk.

Despite this logic, persuading my family to let me switch to this way of living took some effort. When I suggested to my husband that living by candlelight might be romantic, he rolled his eyes. But convincing him was a doddle compared to my six-year-old daughter and four-year-old son. Here’s how that conversation went:

Me: Children, we’re going to try living in the dark for a few weeks

Daughter: But that will be spooky.

Me: No, I think it could be lots of fun. We’ll have candles.

Daughter: *bursts into tears*

Me: Please don’t cry. It will be like going camping.

Son: Can we have marshmallows?

Several packets of marshmallows later, we were set – although I agreed that my husband could occasionally use electric lighting and the kids could watch television, provided I wasn’t nearby. Because I needed to maintain a normal work schedule, I also decided to keep the lights on until 18:00, although I switched my laptop to ‘night mode’ after sunset. 

The protocol looked like this: during the first week, I would try to maximise my exposure to daylight by moving my desk next to the window, lingering in the park on the way back from school drop-off each morning, getting outside at lunchtime and trying to exercise outdoors. The second experimental week would be spent minimising my exposure to artificial light after 18:00, relying on candle-light or dim red lighting instead. Then I would combine the two.

In between each of these intervention weeks, I would lead a normal life. These weeks would function as a baseline.

On the last evening of each week, I would spend the evening in darkness, taking hourly samples of my melatonin, our 'hormone of darkness'

To track my responses, I’d wear an ‘actiwatch’ to measure light exposure, activity and sleep. I'd also complete sleep diaries and questionnaires to assess my sleepiness and mood and undertake a battery of cognitive tests to assess my short-term memory, attention and reaction speeds. On the last evening of each week, I would spend the evening in darkness, taking hourly samples of my melatonin, which is released in response to a signal from the biological clock and therefore provides a marker of our internal time. “Melatonin is our hormone of darkness; it creates the biological night,” says Marijke Gordijn, a chronobiologist at the University of Groningen in the Netherlands, who measured my melatonin levels.

The idea was to see if these changes to my light exposure altered the timing of my biological clock. We were curious to see if any of the benefits predicted by larger, well-controlled laboratory studies would translate into real life.

“We’ve done a lot of experiments where we’ve given a dose of light and seen that it shifts the clock,” says Gordijn. “But if we want to apply those findings to help people, we need to know that it will have the same effect when the environment is more variable.”

Switched on

And so, on a bright and sunny December morning, I found myself in the local park, inconspicuously trying to work out on the monkey-bars and swings instead of going to a body pump session in the gym. "Mummy, what's that lady doing?" asked a small boy.

Because it was winter and most people were inside keeping warm, the park was largely deserted. Motivation had been a struggle for me also. It’s hard to overcome the belief that because it’s winter, it will be cold and miserable outside. However, I was reminded of something a Swedish friend used to say: there’s no such thing as bad weather, only inappropriate clothing. And I soon realised that its rarely as bad outside as it may look. Indeed, the more I did it, the more I came to regard getting outdoors in winter as a treat, rather than a chore.  

On another morning, I sat in the park with a cup of tea on my way back from dropping the kids at school and got out my light meter. The illuminance of light is measured in lux. On a cloudless day in summer, the light outdoors can reach as high as 100,000 lux; on an overcast day, it can be as low as 1,000 lux. Today, the reading was 73,000 lux.

Back indoors, I took a reading in the centre of my shared office: 120 lux – lower even than the 500 lux you’d expect outdoors immediately after sunset. Horrified, I returned to my temporary desk by the window, where it was colder, but a sunnier 720 lux.

I took a reading in the centre of my shared office: 120 lux – lower even than the 500 lux you’d expect outdoors after sunset

Despite my best efforts to get more daylight during the intervention weeks, my average light exposure between 7:30 and 18:00 was 397 lux during the first week and just 180 lux during the second. This was presumably because I still spent most of the day indoors, working at my computer, and because the Sun set at around 16:00. The likely reason for this variation was the weather. During the first week, there were 4.5 hours of bright sunshine per day on average, whereas in the second week it was just 0.9 hours. This was still an improvement on baseline weeks, though, when my average daytime exposure was just 128 lux.

It wasn’t only the weather that proved challenging. For the first few nights of the experiment, we slept with the curtains open to maximise exposure to the dawn light. Light at this time is thought to shift the body clock earlier.

But at night-time, the light from the streetlights made it difficult to sleep.

Urban dwellers are exposed to night-time light levels three to six times brighter than people in small towns and rural areas

I’m not alone in experiencing this problem. In 2016, researchers reported that people living in urban areas of more than 500,000 people are exposed to night-time light levels that are three to six times brighter than people in small towns and rural areas. Those living in areas of more intense light sleep less, are more tired during the daytime, and report feeling more dissatisfied with their sleep. They also go to bed and wake up later than people in darker areas.  

After a few days of this, I began closing the curtains and using a dawn simulation clock instead. It was an imperfect solution, as the light from these devices isn’t as bright as daylight. But it was better than nothing.

If exposing myself to more daylight wasn’t hard enough, I also had to eliminate evening light during December, the darkest month of the year. Doing so brought home to me just how useful artificial light is. Cooking by candle-light was a daily challenge, chopping vegetables an outright hazard. I began to prepare meals earlier in the day, which ate into my work time and meant I got less done.

Eventually, I employed another work-around. I installed some smart lightbulbs in my kitchen, that could be dimmed and colour-adjusted using an app on my smartphone. This, of course, created a paradox: to remove the alerting blue light from the bulbs, I had to expose myself to blue light from my phone, so I did it in the daytime to avoid invalidating the experiment. Now our kitchen glowed an eerie red-orange by night. But at least we could cook again.

During my ‘dark weeks’, I was exposed to an average illuminance of 0.5 lux between 18:00 and midnight and a maximum of 59 lux. That was compared to an average 26 lux (and a maximum of 9640 lux – I have no idea what this super-bright, artificial light source was) when I lived normally – although the actiwatch on my wrist wouldn’t have detected any light emitting from my smartphone or laptop during baseline weeks. That’s important, because there’s mounting evidence to suggest that these devices can disrupt sleep.

One 2015 study suggested that using an e-reader before bed prolonged the amount of time it took for people to fall asleep, delayed the circadian clock, suppressed REM sleep, and left participants feeling more tired the next morning, compared to people who read a print book for the same amount of time.

One 2015 study suggested that using an e-reader before bed prolonged the amount of time it took for people to fall asleep, suppressed REM sleep, and left participants feeling more tired the next morning

Another recent study compared people’s responses when they played computer games during the evening on a normal smartphone screen versus one that suppressed blue light. Players felt more alert after using the conventional smartphones and performed worse in cognitive tests the next day, suggesting that their sleep may have suffered.   

My pledge to avoid artificial light also made socialising difficult. A few days before my experiment started, a friend invited us over to her house for pre-Christmas drinks in the middle of a “dark week”. When I explained my predicament, she generously offered to let me sit upstairs in a candle-lit room and receive visitors. I politely declined, feeling how I imagine vegans must feel when they’re invited out to a meal at a steakhouse.

Instead, we encouraged friends to come to our house, and so they came: amused, curious, and occasionally concerned about what they might find. One family initially declined our invitation to stay for New Year because they were worried their son might knock over the candles. They changed their mind when I told them they could use the lights in their bedroom. (We kept all candles out of kids’ reach, just in case.)

Once we’d adapted to the challenges, living without artificial light was very pleasant. Conversation seemed to flow more easily, and visitors also commented on how mellow and relaxed they felt in the dim light. Another bonus was that our children seemed to settle down more easily in the evenings, although we didn’t collect any data on this.

But did any of this make any difference to my sleep or mental performance? There was a general trend towards earlier bedtimes during the intervention weeks – particularly during the week when increased daylight was combined with low evening light. On this week, my average bedtime was 23:00, compared to 23:35 on baseline weeks.

As it was December, I had a lot of social commitments, so I occasionally ignored my body’s sleep signals and stayed up later. It’s an issue that researchers often encounter in their studies. “People have social obligations, and it’s very hard for them to follow what their clock is telling them to do,” says Mariana Figueiro, director of the Lighting Research Center in Troy, New York. “We are constantly fighting our physiology.” 

I was significantly sleepier in the evenings during the increased daylight and low evening light weeks

Even so, I was significantly sleepier in the evenings during the increased daylight and low evening light intervention weeks. My body also started releasing the darkness hormone, melatonin, some 1.5 hours earlier during the daylight intervention week – and two hours earlier when I avoided evening light.

It is a pattern that’s been seen in other studies. Like me, Kenneth Wright at the University of Boulder in Colorado has long been fascinated by how our modern lighting environment might be affecting our internal timing. So in 2013, he sent eight people camping in the Rocky Mountains of Colorado for a week during the summer and measured how this affected their sleep. “Camping is an obvious way of removing ourselves from this modern lighting environment and just getting access to natural light,” says Wright.

Before the trip, the participants’ average bedtime was 00:30 and their wake time 8:00. Both had shifted approximately 1.2 hours earlier by the end of the trip. They also started releasing melatonin some two hours earlier once they were removed from artificial light – although they didn’t sleep for any longer.

In winter, people sent to camp outdoors for a week slept for around 2.3 hours longer than usual 

Wright recently repeated the study in winter. This time, he found that participants went to sleep some 2.5 hours earlier under natural lighting conditions but woke up at roughly the same time as when they were living indoors. This meant they slept for around 2.3 hours longer. “We think it’s because people were going back to their tents earlier to get warm, so they were giving themselves a longer opportunity to sleep,” says Wright.

Unlike his participants, I didn’t experience a sizeable increase in the amount of sleep I got during the intervention weeks – although there was a slight increase in sleep time and efficiency (the ratio of the total time you spend asleep versus the amount of time bed). However, this didn’t reach statistical significance, meaning it could have been the result of chance. Perhaps it was because I was living in a relatively warm house, which made defying my body clock easier. I also was forced to get up at the same time each morning by my children – who occasionally woke me at night, too.  

For every 100 lux increase in my average daylight exposure, I experienced an increase in sleep efficiency of almost 1% and got an extra 10 minutes of sleep

But when I correlated my sleep with the amount of light I was exposed to during the daytime, an interesting pattern emerged. On the brightest days, I went to bed earlier. And for every 100 lux increase in my average daylight exposure, I experienced an increase in sleep efficiency of almost 1% and got an extra 10 minutes of sleep.

I also felt more alert upon waking during all three intervention weeks – but particularly during the two weeks when I was exposed to more daylight.

This pattern has been seen in other studies too. The General Services Administration is the largest landlord in the United States. Many of the public buildings it manages either were designed to try and boost indoor daylight levels or have been remodelled, so its leaders were keen to find out if this had made any difference to the health of those working inside them. Working with the Lighting Research Center’s Figueiro, they picked four such office buildings, plus the GSA Regional Office Building in Washington DC – a converted warehouse which had little access to natural daylight at that time. Workers were asked to wear a device that collected light data around their necks, as well as to complete daily mood and sleep questionnaires for a week during summer and again in winter.

Our study revealed that if you are three, four, five feet from the window, you lose the daylight – Mariana Figueiro

When the light data started to come in, it was initially disheartening. Despite efforts to boost daylight in the workplace, many GSA workers weren’t receiving it. “Our study revealed that if you are three, four, five feet from the window, you lose the daylight,” Figueiro says. “It’s not just your distance from the window that matters. You have partitions, people pulling the shades. Having a window doesn’t necessarily mean you’re going to get good daylight.”

Probing further, Figueiro’s team divided the office workers into those who were receiving a high circadian stimulus – light that was bright or blue enough to activate the circadian system – and those who were receiving a low stimulus. 

Those exposed to a high amount of morning light took an average of 18 minutes to fall asleep at night, compared to 45 minutes in the other group

Those who received a high stimulus took less time to fall asleep at night and slept for longer. Morning light seemed to be particularly powerful: those exposed to a high stimulus between 8:00 and 12:00 took an average 18 minutes to fall asleep at night, compared to 45 minutes in the low stimulus group. They slept for an extra 20 minutes. Their sleep efficiency was 2.8% higher. And they reported significantly fewer sleep disturbances. These associations were stronger during winter, when people may have had less opportunity to receive natural light during their journey to work.

Gordijn also recently published a study which found that people slept better following more exposure to daylight. Here, the participants were wired up to polysomnography monitors to record their sleep in detail. “People had more deep sleep, and it was less fragmented after more exposure to daylight,” Gordijn says.

Light heart

Until recently, scientists had assumed that our urge to sleep was driven by two independent systems: the circadian system, which affects sleep timing, and a ‘homeostatic system’ which keeps tabs on how long you’ve been awake and ratchets up the pressure to sleep.   

Light was known to alter the timing of sleep via the circadian system. But recent work by Samer Hattar at the University of Maryland has suggested that the light-sensitive cells in the eye, which control the circadian system, also connect to the homeostatic system. “We propose that the timing and intensity of light exposure doesn’t only modulate circadian-driven aspects of sleep, but also homeostatic sleep pressure,” Gordijn says.

Daylight also affects mood. Those GSA office workers who were exposed to brighter morning light scored lower on a self-rated scale of depression. Other research has shown that morning light, as well as light during the day, can improve symptoms of non-seasonal depression.

“It probably has to do with being more entrained to the light/dark cycle and sleeping better,” says Figueiro. In her study, those who recorded a high circadian stimulus in the daytime tended to be more active during daylight hours and less active once it got dark, suggesting their sleep was more aligned with their internal clock.

These data are in accordance with office studies in the UK. In March 2007, Dijk and his colleagues replaced the light bulbs on two floors of an office block in northern England, housing an electronic parts distribution company. Workers on one floor of the building were exposed to blue-enriched lighting for four weeks; those on the other floor were exposed to white light. Then the bulbs were switched, meaning both groups were ultimately exposed to both types of light. They found that exposure to the blue-enriched white light during daytime hours improved the workers’ subjective alertness, performance, and evening fatigue. They also reported better quality and longer sleep.

My early-morning mood was significantly more positive during the intervention weeks compared to when I was living normally

This also fits with my own findings. Immediately after waking and before going to bed each night, I filled out a questionnaire to assess how positive and negative I was feeling. The results suggest that my early-morning mood was significantly more positive during the intervention weeks compared to when I was living normally. There was also a trend towards less negative feelings in the evening.  

And although I didn’t officially assess my mood at any other time of day, I felt more energetic and uplifted on those weeks when I spent more time outside. Because of my experience, I’m a convert to outdoor exercise. I’m also learning to embrace the long winter nights: seeing the season as an opportunity to make the house cosy with candles rather than bemoaning the darkness.

Even my daughter is a convert. Towards the end of the experiment, I asked her if she was looking forward to switching the lights back on. "No," she said. "It has been wonderful, because the candles are really relaxing." Instead, it was my four-year-old son who insisted: he wanted to see what he was eating at dinnertime. 

Although none of my cognitive test results achieved statistical significance, there was a trend towards faster reaction speeds during the intervention weeks, as well as slightly better performance in a test that involved remembering where a token was hidden in a series of boxes.    

Studies by Gilles Vanderwalle at the University of Liège in Belgium and Dijk have shown that exposure to bright light activates brain areas involved in alertness – although in these studies, the effects weren’t long-lasting.

However, in a separate study, researchers at Charité Universitätsmedizin in Berlin discovered that the energising effects of light continued for the rest of the day. When participants were exposed to bright, blue-enriched light in the morning, they reported feeling less sleepy during the evenings, and their reaction speeds were maintained, rather than declining as time wore on. Also, the bright morning light seemed to buffer their body clocks against the effects of blue evening light – a finding which is in accordance with current mathematical models of how light affects the human biological clock and sleep.

It supports the idea that brighter and blue-enriched morning light could be a useful countermeasure to artificial light in the evenings especially during the darker seasons, when less daylight is available. It means we don't necessarily need to spend our evenings in darkness, or stop using our computers and gadgets.

When we’re talking about kids looking at iPads in the evening, it’s having detrimental effects if they’re spending their daytimes in biological darkness. But if they’re in bright light during the day it may not matter. – Dieter Kunz

“The effects of light in the evening highly depend on the light you were exposed to in the morning,” says Dieter Kunz, who was involved in the study. “When we’re talking about kids looking at iPads in the evening, it’s having detrimental effects if they’re spending their daytimes in biological darkness. But if they’re in bright light during the day it may not matter.” 

It’s ridiculously simple. But spending more time outdoors during the daytime and dimming the lights in the evening really could be a recipe for better sleep and health. For millennia, humans have lived in synchrony with the Sun. Perhaps it's time we got reacquainted. 

Linda Geddes is the author of the upcoming book Chasing the Sun: The astonishing science of sunlight and how to survive in a 24/7 world. It will be published by Wellcome Collection in January 2019.

Join 800,000+ Future fans by liking us on Facebook, or follow us on Twitter.

If you liked this story, sign up for the weekly bbc.com features newsletter, called “If You Only Read 6 Things This Week”. A handpicked selection of stories from BBC Future, Earth, Culture, Capital, and Travel, delivered to your inbox every Friday.