Catherine Crofts went ‘off-grid’ to explore if and how her sleep changed when disconnected from the modern world. Find out what happened in her N=1 sleep study.

We often hear the phrase “I’ll sleep when I am dead”, especially from people who are considered leaders. Not sleeping can then be associated with ways to success. But sleep is critical for health. All animals sleep, even those that appear to be awake all the time, such as dolphins. When dolphins sleep, only one half of their brain sleeps at a time, giving them the impression of always being awake.

The bad effects of sleep deprivation

People deprived of all sleep will die within a few weeks, with symptoms like dementia.   Total hours of sleep are important for physical performance. Less than eight hours of sleep per night – and especially less than six hours – is associated with a decrease in aerobic output and muscle strength; a shorter time to physical exhaustion; and an increased risk of injuries. Appetite is also increased, with a raised desire for energy-dense foods.


“Less than eight hours of sleep per night – and especially less than six hours – is associated with a decrease in aerobic output and muscle strength.”


Sleep can be divided into two main phases – rapid eye movement (REM) and non-rapid eye movement (NREM) with the latter phase being divided into light and deep stages. Each of these three different stages of sleep are associated with different benefits to the brain and can occur at different times of the night. All stages are critical for good health.

NREM sleep heals and cleans

NREM is associated with memory, especially with the transfer of short-term memory into longer-term storage. Adequate amounts of NREM sleep are believed to both increase ‘learning capacity’ (ability to learn new material) and decrease the risk of information being lost or forgotten. NREM helps with curbing inflammation and boosting muscle repair. Deep NREM is also important for metabolic health as it lowers central sympathetic tone, thus decreasing blood pressure and insulin resistance. Physiological changes occur within the brain during deep sleep. Glial cells shrink, which allows the cerebrospinal fluid to ‘clean’ the brain of debris, especially the amyloid proteins associated with Alzheimer’s disease.

REM sleep boosts resilience

REM sleep, especially that acquired in the later hours of sleep in the early morning, is important for emotional regulation and may help us develop resilience for the stressors of the day, as well as being a critical life skill. REM sleep may also be important for motor memory development, such as athletic or artistic performance (e.g. playing a musical instrument).


“So overall, sleep is critical for memory, resilience, metabolic health, the prevention of degenerative brain diseases, amongst other health benefits of which we may not yet be aware.

We sleep in cycles, going through periods of light NREM into deep NREM, then into REM sleep. Each cycle takes about 90-120 minutes, but the amount of time we spend in deep NREM or REM sleep changes throughout the night. Earlier in sleep we have higher amounts of deep NREM sleep and less REM, with the reverse as sleep progresses. But this is also time-critical because if we go to bed late, we lose deep NREM, but if we wake up too early, we lose REM sleep. This suggests that daylight is important for helping to set the sleep cycles.

How well do I sleep and what is the impact of a holiday, while connected to the modern world (e.g. electricity and wifi) versus a holiday mostly ‘off grid’? I decided to test this using the FitBit Charge 3 to record my sleep patterns.


A Fitbit Charge 3 was worn every night from 26 November 2019 to 2 January 2020 inclusive, except for December 13th.

Time periods included normal work/home balance (26 Nov–15 Dec); a holiday period with no normal home routines but including access to electricity and internet (16 Dec-26 Dec); a holiday period with minimal access to electricity and wifi (27 Dec–2 Jan).

Sleep data collected included:

  • Sleep opportunity (time in bed)
  • Number of awakenings
  • Minutes of overall sleep time
  • Overall awake time
  • REM sleep, light sleep, and deep sleep
  • Fitbit sleep score – a proprietary calculation from Fitbit based on duration, quality (time spent in REM and deep sleep) and sleep heart rate, for a total score of up to 100. The higher the score, the better the sleep.


Descriptive statistics and ANOVA compared the three different groups.


Work Holiday “on grid” Holiday “off grid”
Number of nights 19 11 7
Time in bed 446 (67.3) 467 (55.4) 556.4 (59.0)**


400 (57.2) 423 (48.5) 492 (46.1)**
Awake 45.4 (12.0) 43.6 (13.6) 64.7 (17.0)**
Number of awakenings 24.7 (4.3) 28.6 (4.8) 25.9 (5.0)
REM sleep 96.2 (25.3) 109.9 (27.2) 101.3 (14.7)
Deep sleep 73.7 (28.4) 79.6 (17.4) 85.0 (24.5)
Light sleep 230.6 (40.0) 233.6 (28.5) 305.4 (40.5)**
Fitbit score 78.5 (4.6) 78.7 (4.1) 83.6 (3.6)*

All values in minutes unless otherwise specified

* p < 0.5, ** p < 0.01  compared to “work”


This study showed that overall, I probably sleep pretty well. I average about 7.5 hours in bed under ‘normal’ conditions with 6.7 hours of actual sleep time. Mostly I do not remember being awake. When I went on holiday and had normal access to electricity and mobile data (Rotorua) there was minimal difference to sleep. However, when I went ‘off-grid’ (Great Barrier Island) where I went for days without access to electricity and minimal wifi and cellphone coverage, there was a significance difference – in time spent in bed, asleep, awake, and light sleep. Some of these differences would be reflected in that we mostly slept in a light tent on thin mattresses and it was windy so it was very noisy in a flapping tent! This meant that I woke up more, was aware of being awake, and took longer to get back to sleep. But despite subjectively feeling that I had slept poorly – the FitBit score was higher than usual and I never felt sleepy the next day.

Earlier bedtime

The lack of electricity meant that we were more likely to eat dinner earlier than normal and retire to bed not long after it got dark (sunset about 8.30pm). As the tent was low, with minimal space, we had to lay down/recline. Although we normally read books before sleeping (paper or e-book with no blue light), we were asleep earlier than when we have access to electricity. I don’t know how much the lack of access to entertainment (TV, computer, smartphone), or normal household and work routines, or greater energy expenditure (tramping with a heavy pack) contributed to the early sleep, or whether it was more reflective of the lack of blue light and a return to normal circadian/ultradian rhythms.

Biggest increase in light sleep

It was interesting to note the lack of change between deep sleep and REM sleep between the three different times. Although we know that the quality of sleep is important to health, we do not know objectively how much time we should spend in these sleep zones. Fitbit suggests benchmark ranges for women my age of 10-20% time spent awake, 15-25% in REM sleep, 40-60% in light sleep, and 12-18% in deep sleep. However, while I normally reach these benchmarks, there is a significant difference in minutes spent in these sleep stages between nights when I get 5 hours of actual sleep time, compared to night when I get 7+ hours of sleep.

It is unclear why the biggest increase was in light sleep. Either this means that there is an optimal time in minutes for REM and deep sleep, then any additional sleep time could be spent in light sleep, or the increase in time spent awake, meant that additional time was spent in light sleep to restore proper sleep staging patterns. Looks like I need to spend more time off-grid by with access to a more comfortable and quiet bed to be able to better assess this situation!

A possible confounder includes the change in physical activity status while we were off-grid. We tramped and backpacked on Great Barrier Island. With minimal access to stores and cooking facilities, we had to carry all our food and a tramping stove. I carried between 13-18kg equipment in my pack for at least 2 hours (normally 4-6 hours) each day for six of the seven days on mostly hilly terrain, walking about 8-10 km per day.

So what does this mean? 

As always, more research is needed, because this was a small, limited and uncontrolled study. But as we are all individuals, maybe some self-experimentation with increasing sleep or sleep quality is a good idea.

  • For starters, we need to prioritise sleep.

We need to make sure we have access to a reasonable amount of ‘sleep opportunity’ or time spent in bed being able to sleep – at least 7-8 hours. Ideally, this time should be both early enough in the evening and late enough in the morning to make sure we get adequate amounts of both deep NREM and REM sleep.   This is not always possible, (especially for shift workers or parents with small children), but perhaps we need to deprioritise our screens and move them out of the bedroom and putting a timer on how late we watch Netflix.


“…perhaps we need to deprioritise our screens and move them out of the bedroom and putting a timer on how late we watch Netflix.”


  • Keep a record of how much sleep you get, what time you go to bed and when you wake up.

Wearables like FitBit can do this automatically, but a pen and notebook also work well.  What combination of bedtime and awake time makes you feel refreshed and energised? If you have a bad night’s sleep, do you get the 3pm cravings, or does your coffee, or worse, your energy drink need increase? What does sleep do to your physical activity?

  • Having regular digital detoxes can also be a good idea.

I try and have at least a continuous ‘off grid’ week every year. “But what if you miss the [momentous thing]?” is always asked. After about 20 years of doing this, I can recall missing one momentous occasion: The Boxing Day tsunami in 2004. (We learned about it approximately 8 days later.)