Play Live Radio
Next Up:
Available On Air Stations

Getting to the Bottom of REM

Fifty years ago, two researchers at the University of Chicago reported in the journal Science that people moved their eyes even when they were fast asleep. They also showed that these periods of rapid eye movements during the night were associated with dreaming.

In the ensuing decades, thousands of scientific studies have shown there's more to REM sleep, as these periods came to be known, than just dreaming. Before becoming a journalist, NPR's Joe Palca conducted some of those studies. Palca, who received his doctorate from the University of California at Santa Cruz, wrote his dissertation on how people maintain their body temperature during REM.

Marking the 50th anniversary of the discovery of REM sleep, Palca reports how the finding changed sleep research.

Read excerpts from Palca's 1982 dissertation, Effects of Cold Ambient Temperature on Human Sleep, Metabolism and Thermoregulation.


Research to date on humans has tended to substantiate the finding in various animal species that thermoregulatory processes are curtailed or abolished during rapid eye movement (REM) sleep. Cold ambient temperatures reduce REM sleep throughout the entire duration of several days continuous exposure, whereas in most stressful situations REM sleep is initially depressed, but subsequently returns to baseline levels. The current study examines the effect of several nights exposure to cold on sleep and thermoregulatory variables.

Four male subjects, aged 20-24 years, slept twelve consecutive nights in the sleep laboratory. The first two nights were considered adaptation, and were not analyzed. There followed three nights at 29 degrees C (Baseline condition, thermoneutral for humans), five nights at 21 degrees C (Cold condition), and two nights at 29 degrees C (Recovery condition). A standard somnograph consisting of electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) was recorded, together with skin, rectal, and tympanic temperature and oxygen consumption (VO2).

Cold disrupted the sleep of all subjects. Stage 2 decreased significantly in the cold, whereas wakefulness increased significantly. Increases in stage 1 and decreases in REM sleep approached significance. The distribution of sleep stages was not significantly altered by the cold, and there were no significant rebounds in the recovery condition.

Individual differences in sleep patterns in the cold allowed the separation of the subjects into two groups: good sleepers and poor sleepers. Good sleepers exhibited little or no decline in REM sleep in the cold and generally exhibited lower skin and rectal temperatures than poor sleepers. They also decreased their VO2 in the cold compared with thermoneutrality, whereas poor sleepers increased their VO2. Tympanic temperature did not exhibit differences among thermal conditions or among subjects.

At REM onset in the cold tympanic temperature rose significantly in subjects. Skin temperature changes at REM onset in the cold varied by body location. The temperatures of trunk skin sites tended to remain fairly stable following REM onset, whereas those of peripheral sites tended to decline. VO2 increased during REM sleep compared to continuous NREM sleep both in the cold and at thermoneutrality.

The pattern of results did not point to an impairment of thermoregulatory functions equivalent to that seen in animals.



Four male subjects between the ages of 20 and 24 were chosen by their responses to a sleep questionnaire concerning the regularity of their sleep habits, abstinence from drugs, and willingness to participate in a time consuming and occasionally uncomfortable experiment. All subjects except the first were asked to spend one night at a cold temperature of 21 degrees C before beginning the experiment to assess their ability to tolerate this temperature.* This single cold night occurred at least one week before the start of the experiment, and during it electrodes were not attached to the subjects.


Subjects were required to abide by a rigid protocol throughout experiment, which included abstinence from alcohol and drugs, no strenuous exercise during the day, and no food for at least four hours their usual bed-times.

Subjects arrived at the lab about one hour before their chosen bed-time, which remained constant throughout the experiment. Electrodes for electrophysiological measures were applied in the recording room adjacent to the bedroom (Ta approximately 22 degrees C). Subjects then entered the 3.0 X 3.0 X 2.7 meter shielded bedroom where thermocouples for temperature measurements were attached. The latter procedure took about twenty minutes, and allowed the subjects to adapt to the bedroom's temperature while awake.

Subjects slept nude except for undershorts on a bed made from nylon webbing stretched around a metal frame. This material was used to maximize skin exposure to the ambient air, and did not provide the high levels of insulation of normal bedding. Ta was maintained to within 0.5 degrees of set values by proportionally controlled heating and cooling units in the room, and was monitored by an electronic thermometer located near the subject's head. An awning made from bed-sheets and mosquito netting was hung around the bed to reduce air flow around the subject.

Subjects spent twelve consecutive nights in the sleep lab. The first two nights (1, 2) of the twelve nights were considered as adaptation nights and were not analyzed. The next three nights (3, 4, 5) at 29 degrees C constituted the baseline condition, the next five nights (6, 7, 8, 9, 10) at 21 degrees C were the cold condition, and the last two nights were recovery nights (11, 12) at 29 degrees C. Subject LM suffered a deep cut on his finger on the day following his fifth night in the lab. The experiment was suspended for three days while he recovered. The two nights following the interruption were spent at 29 degrees C, and these, together with night 5, were considered his baseline condition.

* This procedure was instituted when a prospective subject left the experiment after his first night of cold exposure.

Copyright 2022 NPR. To see more, visit

Joe Palca is a science correspondent for NPR. Since joining NPR in 1992, Palca has covered a range of science topics — everything from biomedical research to astronomy. He is currently focused on the eponymous series, "Joe's Big Idea." Stories in the series explore the minds and motivations of scientists and inventors. Palca is also the founder of NPR Scicommers – A science communication collective.
More from Hawai‘i Public Radio