Biological rhythms

Human sleep before the industrial era

A quantitative study of sleep patterns in three pre-industrial societies implies that our natural sleep duration is close to seven hours, and that sleep cycles are determined by environmental temperature as well as the light–dark cycle.

How much sleep do we need? What time should we start school or work so that we are in tune with our body clocks? These are topical questions, and hardly a week goes by without media reports about the importance of sleep and daily rhythms for physical and mental health. The suggestion is that our modern sleep patterns are muddled. But what are our natural sleep patterns, and how are they influenced by environmental factors? What is the benchmark for 'good' sleep? Writing in Current Biology, a team of sleep neuroscientists and anthropologists (Yetish et al.1) reports a study of sleep patterns in three hunter-gatherer and hunter-horticulturist groups. During the study, and as is usual for these communities, none of the participants had access to electricity, and the only sources of light, apart from the Sun and Moon, were small fires.

The study participants came from the Hadza group in Tanzania, the Ju/'hoansi San in Namibia and the Tsimane' in Bolivia. The authors monitored the participants' sleep over several days or weeks using activity and light recorders, and found that the average sleep duration across the three groups — the interval between sleep onset and sleep end — was 7.7 hours (Fig. 1). Subtracting the awakenings that occur during the night left an average total sleep time of only 6.4 hours. How does this compare with sleep in modern industrialized societies that use electricity daily? The embarrassing answer is: we don't really know, because most estimates of sleep duration are based on self reports rather than the objective measures used by Yetish and colleagues. However, the self-reported average sleep duration in modern industrialized groups is around 7–7.5 hours, and although it depends on the day of the week and the age of the individual2,3, it differs little from that found in this study.

Figure 1: Daily patterns.

Yetish and colleagues' study1 of sleep patterns in hunter-gatherer and hunter-horticulturalist groups allows a comparison between typical sleep patterns and daily activity schedules of individuals in modern but pre-industrial societies and those in modern, industrial societies4. In the latter societies, regular access to electricity divorces people from the natural environmental light–dark cycle, with light-exposure levels being largely self-determined and fairly uniform across waking hours. The authors' findings show that average sleep duration in the two societies is similar, but that sleep timing is more in sync with environmental cycles in pre-industrial than in industrial societies. (Schematic of light intensity derived from refs 1 and 4.)

When do the 7.7 hours of sleep recorded by Yetish et al. occur? No watches or electronic clocks were available to the participants, so sleep timing was determined by environmental cues and internal biological clocks. Maybe surprisingly for those of us who think that napping is natural, the authors found that few daytime naps were taken. Furthermore, sleep did not begin at the onset of darkness, but on average 3.3 hours after sunset. This average conceals a remarkable night-to-night variation in sleep onset — it seems that a regular bedtime is not a hallmark of natural sleep. By contrast, individuals in each group tended to wake at similar times, generally before sunrise. How does this compare with modern, industrialized sleep? Many people do go to sleep well after sunset, but few of us consistently rise before sunrise.

The timing of natural sleep can be easily linked to what we know about our biological daily rhythms. Earth's rotation results in environmental cycles of light and dark and hot and cold, and evolution has favoured the survival of biological mechanisms that predict these daily geophysical patterns. This rhythmicity has been observed right down to the cellular and molecular level, with nearly every cell in the human body showing approximately 24-hour (circadian) oscillations in gene expression. It is thought that synchrony of these billions of individual cellular rhythms is orchestrated by the brain's master clock, and it is well established that the light–dark cycle is the most prominent environmental stimulus for synchronizing this clock with the external world. The timing of our internal clock determines when we feel sleepy.

However, since the discovery of fire, we have learnt to manipulate our exposure to light. We can extend the light period by lighting a fire, a candle, an oil lamp, an incandescent light bulb, a fluorescent lamp or a light-emitting diode, giving us behavioural control over the stimulus that synchronizes our clock. The participants in Yetish and colleagues' study have this control to only a limited extent — the dim red light emitted by their fires has less biological effect than electric light with a strong blue component, such as is emitted by electronic devices and many low-energy light bulbs4,5.

Recent studies exploring the impact of introducing electric light to the Toba Qom people in the Argentinean Chaco region6 or to rubber tappers in the Amazon7 indeed demonstrate delayed bedtimes and reduced sleep duration. And evidence from a study comparing US residents in daily life and on a camping trip away from artificial light suggests that access to electric light not only shifts the circadian clock but also exaggerates the natural individual variability of sleep timing8. At an extreme, perhaps the owl-like tendencies of adolescents, and the resulting debate on shifting school timing to later starts, are to a considerable extent a direct result of our manipulation of our light environment.

Yetish and colleagues also point to the influence of a second environmental cycle on sleep timing: temperature. Core body temperature rises and falls over each 24 hours, with sleep generally occurring as the body temperature falls. Because we are a diurnal species, awake during the day and asleep during the night, and because environmental temperature tends to drop during the hours of darkness, the rhythm of our core body temperature tends to align with that of the environmental temperature. From an energetic perspective, this makes sense — keep the difference between environment and body temperature as small as possible and you reduce the amount of energy needed to stay warm. This link between environmental temperature, metabolic demands and the timing of the sleep–wake cycle has received attention in the animal and human sleep literature9,10. Yetish and colleagues found that the timing of waking is closely associated with the environmental minimum temperature; in the one group that woke after sunrise, the San, this pattern occurred only during summer, when the environmental-temperature minimum also occurred after sunrise.

We are beginning to understand the impact of our artificial world on our sleep–wake rhythms. Although we have tantalizing hints about how the light environment of our modern world affects our sleep patterns, there are few data on how we manipulate our temperature environment and the effects of such manipulation on sleep. Yetish et al. have provided findings that alter our assumptions about sleep in our ancestors, and have opened the door to further studies of the effects of light and temperature on sleep today.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Yetish, G. et al. Curr. Biol. (2015).

  2. 2

    Roenneberg, T. Nature 498, 427–428 (2013).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Skeldon, A. C., Derks, G. & Dijk, D.-J. Sleep Med. Rev. (2015).

  4. 4

    Santhi, N. et al. J. Pineal Res. 53, 47–59 (2012).

    CAS  Article  Google Scholar 

  5. 5

    Czeisler, C. A. Nature 497, S13 (2013).

    ADS  CAS  Article  Google Scholar 

  6. 6

    de la Iglesia, H. O. et al. J. Biol. Rhythms 30, 342–350 (2015).

    CAS  Article  Google Scholar 

  7. 7

    Moreno, C. R. C. et al. Sci. Rep. 5, 14074 (2015).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Wright, K. P. Jr et al. Curr. Biol. 23, 1554–1558 (2013).

    CAS  Article  Google Scholar 

  9. 9

    van der Vinne, V. et al. Proc. Natl Acad. Sci. USA 111, 15256–15260 (2014).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Daan, S., Honma, S. & Honma, K. J. Biol. Rhythms 28, 403–411 (2013).

    Article  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Derk-Jan Dijk or Anne C. Skeldon.

Related links

Related links

Related links in Nature Research

Circadian rhythms: No lazing on sunny afternoons

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Dijk, D., Skeldon, A. Human sleep before the industrial era. Nature 527, 176–177 (2015).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Sign up for the Nature Briefing newsletter for a daily update on COVID-19 science.
Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing