Dr Neil Stanley Independent Sleep Expert
Thomas Wehr and segmented sleep It is clear that the research work of Thomas Wehr does not in any way lend credence to, or support for, Ekirch’s conception of segmented slumber. Ekirch seems to put a great deal of emphasis on the work of Thomas Wehr to support his concept of segmented sleep. He writes “In attempting to recreate conditions of "prehistoric" sleep, Dr. Thomas Wehr and his colleagues at NIMH found that human subjects, deprived at night of artificial light over a span of several weeks, eventually exhibited a pattern of broken slumber—astonishingly, one practically identical to that of pre-industrial households.” However Wehr explicitly was not trying to “recreate conditions of "prehistoric" sleep”, the purpose of his study was to “In animals, circadian pacemakers respond to seasonal changes in day length by making corresponding adjustments in the durations of diurnal and nocturnal periods of circadian rhythms; these adjustments mediate effects of photoperiod on breeding and other seasonally recurring phenomena. Little is known about photoperiod responses of human circadian pacemakers. To investigate this question, we recorded and compared circadian rhythm profiles of 15 individuals after chronic exposures to short (8 h) and long (14 h) nights.” (Conservation of Photoperiod-Responsive Mechanisms in Humans R846) Ekirch claims that In the experiment the subjects “ exhibited a pattern of broken slumber—astonishingly, one practically identical to that of pre-industrial households”. It would appear that Ekirch fails to understand the phrase ‘practically identical’ as Wehr's subjects remained “awake for an average of 2hrs before” they fell asleep. (‘The Impact of Changes in Nightlength’ p267). This does not fit with Ekirch’s conception of segmented sleep; nowhere does he mention such a long sleep onset, nor do any of sources that he quotes give any indication of long sleep onsets. Wehr states that “when individuals eventually fall asleep in long nights, they usually sleep in two major bouts, on in the evening and one in the morning, with a period of quiet wakefulness between the bouts . Each bout last 2-5 hours, and the period of intervening wakefulness last about 1-3hr. (‘The Impact of Changes in Nightlength’ p269). But this is a simplistic presentation of the findings of the experiment, in Wehr’s original paper he actually states that (Conservation of Photoperiod-Responsive Mechanisms in Humans R852) “ In long nights sleep generally separated into ≥2 fragments and often exhibited a symmetrically bimodal pattern of distribution” (This also perhaps gives support for the reports of “third sleep” that I have uncovered). This is not ‘practically identical’ to Ekirch’s conception that “Until the close of the early modern era, Western Europeans on most evenings experienced two major intervals of sleep bridged by up to an hour or more of quiet wakefulness”.  Only by averaging the data across all nights and the 12 ‘responders’ does the ‘bimodal pattern emerge. “The bouts of sleep are distributed in such a way that when nightly profiles of sleep are averaged together, a remarkably symmetrical bimodal pattern emerges” ('Clock for All Seasons' p334). But surely if “Until the close of the early modern era, Western Europeans on most evenings experienced two major intervals of sleep bridged by up to an hour or more of quiet wakefulness” as Ekirch claims such mathematical ‘manipulation’ would not be necessary, the subjects would all have slept in two periods, on most nights, (given that the conditions of the experiment remained constant, one might speculate that it would be more often than not. Ekirch writes that “Significantly, the intervening period of "non-anxious wakefulness" possessed "an endocrinology all its own," with visibly heightened levels of prolactin, a pituitary hormone best known for permitting chickens to brood contentedly atop eggs for long stretches of time”. Why Ekirch believes that broodiness in some breeds of chicken, (many chicken breeds rarely exhibit broodiness), has anything to do with nocturnal “non-anxious wakefulness” in humans is anyone’s guess.  This must have been from the personal communications as it does not appear in the references and is actually flatly contradicted by the following (’Clock for All Seasons' p281) “In 14-hr scotoperiods, prolactin reaches its high nighttime levels after the first 30 minutes, even though sleep onset has not yet occurred  (the average sleep latency in 14-hr scotoperiods is 2 hr) moreover there is no further increase in prolactin level after sleep begins. In addition, there is no decline in prolactin secretion during the periods of quiet wakefulness that occur in the middle of the scotoperiod between the evening and morning bouts of sleep.”   So if prolactin does not go up or down how can “the intervening period of "non-anxious wakefulness" possessed "an endocrinology all its own," with visibly heightened levels of prolactin.” (Interestingly Wehr states R851 “Although a parallel to the human PRL response to photoperiod can be found in the domestic cat, the response of other animals is opposite, in that they secrete less PRL in long nights” would suggest that the human response to long nights is not as, shall we say, natural as it would appear) Ekirch writes that “In fact, Wehr has likened this period of wakefulness to something approaching an altered state of consciousness not unlike meditation”. It is possible that this statement was made in a personal communication with Ekirch because what Wehr actually writes is “In humans, the nocturnal rise in PRL secretion is thought to be sleep dependent. If so, then it is not surprising that the cumulative duration of high PRL secretion increased when the cumulative duration of sleep increased in long nights. However, the pattern of PRL secretion in long nights suggests that PRL secretion is not strictly sleep dependent. In long nights, the onset of the nightly surge in PRL secretion was detected in samples obtained 30 min after the beginning of the dark period even though none of the subjects had fallen asleep at this time (average sleep latency ~2 h). An alternative possibility is that darkness stimulates human PRL secretion. Results of a subsequent experiment, however, failed to support this hypothesis (unpublished data). Another possibility is that quiet rest stimulates human PRL secretion. A previously published report that meditation stimulates human PRL secretion seems consistent with this hypothesis. Indeed, our subjects, when lying awake in long nights, resembled meditating individuals in displaying prominent and sustained alpha rhythms in their EEGs”. However you get alpha simply when you close your eyes and before sleep onset and given that spectral analysis was not done on the PSG we have no idea if the levels seen were “prominent” or not compared to ‘normal’ quiet wakefulness or pre-sleep levels, as Wehr says this is merely based on “unpublished observations”, doubtless post hoc. The paper that Wehr referes to ‘Jevning, R., A. F. Wilson, and E. F. VanderLaan. Plasma prolactin and growth hormone during meditation. Psychosom. Med. 40: 329-333, 1978’ found that  “Individuals of the short-term practitioner group were studied as their own controls before, during, and after a 40-min eyes-closed rest period. Prolactin concentration began to increase toward the end or after meditation in both groups of practitioners with levels continuing to increase in the post-TM period. The increases were not correlated with sleep occurrence. Prolactin levels were stable in controls throughout the experiment.” Wehr clear states that prolactin levels did not change during the long nights including during the period of wakefulness so what the subjects exhibited, with regards to prolactin, does not actually resemble mediation Ekich says that Thomas Wehr has found, "transitions to wakefulness are most likely to occur from REM periods that are especially intense," ('Clock for All Seasons' page 340) “In long nights, periods of quiet rest and contemplation often begin after transitions to wakefulness from periods of REM sleep that are particularly intense” Notice the use of the plural in this statement implying more than one period a night if wakefulness follow from REM periods again not fitting Ekirch’s conception that “Until the close of the early modern era, Western Europeans on most evenings experienced two major intervals of sleep bridged by up to an hour or more of quiet wakefulness”   Ekirch claims that there periods “typically accompanied by "particularly vivid dreams" distinguished by their "narrative quality," which many of the subjects in the experiment contemplated in the darkness”. However this is not a report of data obtain by Wehr as no dream recall were elicited from the subjects. (The phrase “which many of the subjects in the experiment contemplated in the darkness” seems to be an embellishment by Ekirch to support his hypothesis). What Wehr does say is based on an, untested, hypothesis ('Clock for All Seasons'  273) “If, as some research suggests, dreams that occur during REM sleep are distinguished by their vivid and emotional nature and their narrative quality, one wonders whether dreaming associated with REM sleep might have had a greater impact on ancestors who may have slept long nights than on their decedents who sleep in short nights today. At some time each night, they would have awakened from a vivid REM-type dream and entered an extended period of quiet wakefulness in which the effects of the dreams might reverberate in conscious awareness. Perhaps these aspects of the physiology of sleep in long nights will help explain why individuals who live in traditional societies sometimes seemed to attribute more importance to drams than we do” Ekirch, unwitting it seems, actually undermines the support he believes that Wehr provides for his conception of segmented sleep by admitting that “Wehr, to his credit, has speculated that other conditions in his experiments, apart from darkness, might have produced a bimodal pattern of sleep—such as boredom or the enforced rest of his subjects.”  "Further research will be necessary," he has written, "to determine whether, and to what extent, darkness per se or factors associated with the dark condition" were "responsible for the differences that we observed in the subjects' sleep." It is interesting perhaps to note that to date no such research has been done. The actual passage for Wehr is as follows “Confounding of darkness, enforced bed rest, and boredom. Further research will be necessary to determine whether, and to what extent, darkness per se or factors associated with the dark condition, such as instructions to rest or confinement in a boring environment, were responsible for the differences that we observed in the subjects’ sleep and circadian rhythm profiles in short and long nights. Whatever the answer to this question, the conditions and results of the experiment are still likely to be relevant to the behavior of photoperiod-responsive mechanisms in natural environments. In natural environments, darkness, enforced rest, and confinement tend to occur together and act in concert, because darkness inhibits arousal and restricts movement of day-active animals, including humans. In fact, in animals, changes in level of arousal evoked by the immediate presence of light and darkness have been shown to help mediate the phase-resetting effects of light and darkness on circadian oscillators. Until recently, this possibility was ignored in the interpretation of the responses of animals to findings in animals, it would be surprising if the de-arousing properties of enforced bed rest did not contribute significantly to the results of the present experiment.”  So the conditions in the experiment were in fact just as artificial as “today’s “artificial world”” as Wehr himself admits (‘The Impact of Changes in Nightlength’ p267).  A night length of 14 hours rarely occur for example in London only between 26th October and 15th February; in Rome between 10th November and 1st February; Washington DC between November 16th- January 25th, Tromso between 9th October - 4th March and never around the equator, so such a night length occurs for a majority of days in the most  northernmost, (and southernmost), fringes of civilisation, indeed in areas where humans are on relative, that is post Mesolithic era, newcomers. Even more damaging to Ekirch’s contention that “Until the close of the early modern era, Western Europeans on most evenings experienced two major intervals of sleep bridged by up to an hour or more of quiet wakefulness”, and for some reason conveniently ignored Ekirch, are the nonresponsive subjects in Wehr’s experiment   “A few individuals failed to exhibit photoperiod responses that were characteristic of the group as a whole. For example, 12 individuals’ sleep periods expanded and broke up into two or more fragments in long nights, but 3 individuals’ sleep periods remained compressed and consolidated, as though no change in photoperiod had occurred. In other photoperiod-responsive species, such as the Djungarian hamster (Phodopus sungorus), individuals of certain strains are similarly unresponsive to shortening of the photoperiod. Their nocturnal activity periods remain compressed and consolidated after transfer to long nights” Notice it is not “a few individuals” but 3 of the 15 completers i.e 20% and Wehr explains this by reference to strains of hamster,  bizarrely seemingly implying that the nonresponders are somehow a particular “strain” of human” And even more worryingly the “16th individual progressively became severely depressed and suicidal during the first 5 days of exposure to long nights and was removed from the experiment. He gradually recovered during the first week of his return to a normal schedule. Although he had no depressed and suicidal during the first 5 days of exposure to long nights and was removed from the experiment. He gradually recovered during the first week of his return to a normal schedule. Although he had no prior history of affective disorder, his second-degree relatives did. Since terminating the experiment 2 years ago, he has remained well” (Conservation of Photoperiod-Responsive Mechanisms in Humans R847). The fact that Wehr suggest that this subject “may have been genetically predisposed to affective illness” is disingenuous in the extreme (P856) and the fact that the experiment has such disastrous effect on one individual shows that it was not “practically identical to that of pre-industrial households” as Ekirch would have us believe. Ekirch writes “Of greater historical relevance is whether, as Thomas Wehr speculates, "It is tempting to speculate that in prehistoric times this arrangement provided a channel of communication between dreams and waking life that has gradually been closed off as humans have compressed and consolidated their sleep." Wehr goes on to say “if so, then this alteration might provide a physiological explanation for the observation that modern humans seem to have lost touch with the wellspring of myths and fantasies” This of course has nothing to do with 3000 years of rationalist or empiricist thought.  Ekirch writes “One remarkable implication of segmented sleep is that our pattern of seamless slumber for the past two hundred years has been a surprisingly recent phenomenon, the product of modern culture, not the primeval past. From the standpoint of scientists on the front lines of research, this discovery may lead to a better understanding of common "sleep disorders," a misnomer, perhaps, arising from the "natural pattern of human sleep . . . breaking through" into today's "artificial world." But Wehr actually writes “When modern Humans use artificial light to extend their waking activities into the nighttime hours, they compress, reduce and consolidate their sleep, and the delay its timing to the day-night cycle. When human sleep is decompressed in long scotoperiods, it can appear to modern eyes to be disordered. Sleep latency is long, and sleep is fragmented and interspersed with period of wakefulness. Because this type of sleep is similar to the sleep of many other animals, and because it occurs in health young individuals who sleep in long nights, we interpret it to be a natural pattern of human sleep, as it may have once occurred when humans lived from sun to sun. When modern humans find that their sleep is fragmented and interrupted by period of wakefulness during the night they regard it as being disordered. For some, an alternative interpretation could be a natural pattern of human sleep is breaking through into an artificial world in which it seems unfamiliar and unwelcome.”  (‘The Impact of Changes in Nightlength’ p 283) Note that Wehr uses the phrase ‘a natural pattern’ whereas Ekirch states that it is “the natural pattern”. Note Wehr makes no speculation about “common “sleep disorders”.
© Dr. Neil Stanley 2013-2018