Verified by 
Breathing during wakefulness is often involuntarily influenced by physical activities and emotions; our breathing becomes quick with fear or with physical exertion, or slows with relaxation. Although in deep sleep our breathing becomes more regular, it seems to be more erratic and marked by irregularities similar to wakefulness when we enter REM sleep, the stage associated with the most vivid dreams.
A recent study conducted by Oudiette and colleagues and published in Scientific Reports examined whether breathing irregularities during REM sleep may correspond with dream content, either physical activities or emotional arousal occurring within dreams.
In order to study this, the authors used the special case of lucid dreams, where the sleeper is aware that they are dreaming and can control the dream to some extent. Although lucid dreaming occurs rarely in the sleep laboratory, the authors decided to recruit patients with narcolepsy to participate in the study, as narcoleptics experience quick transitions between wakefulness and REM sleep, and have more frequent lucid dreams.
The patients came to the laboratory and first performed waking imaginative activities, prior to taking a nap. During the nap, they were told if they became lucid to signal to the researchers by moving their eyes left and right multiple times; these eye movements can be clearly seen by researchers on the recorded electrooculography.
Patients were instructed to, once lucid, modify their dream content towards a scenario involving irregular respiration, such as holding their breath while diving underwater, or yodeling or singing. They were supposed to signal with eye movements both before and after performing the desired dream activity.
Upon awakening, patients then reported whether they became lucid and gave the eye signal, and what their dream content was after the eye signal. The researchers could then compare the reported content with physiological signals that were recorded during sleep, including eye movements, flow of respiration, heart rate, and muscle tension in abdominal and thoracic muscles.
In all, 21 patients were recruited to participate, for a total of 98 naps; 74 of these naps contained REM sleep and were included in the analyses. 18 of the 21 patients succeeded at least once in signaling lucidity during REM sleep. Among the 74 REM naps, 32 included signaled lucid dreams.
There were several examples where the dream content and physiological measures seemed to match. For instance, one patient became lucid, gave the eye signal, and then had a dream of holding his breath in fear as someone threatened him with a gun. The researchers matched this content with a clear change in respiration: an 11-second long cessation or airflow and an absence of abdominal movements, even a mild heart rate decrease during the airflow cessation.
Overall, 50% of the lucid dreams incorporated clear eye signals followed by changes in respiratory behaviors, such as a large inspiration or cessation of airflow, which matched the reported dream content. The remaining 50% of lucid dreams either had poor ocular codes or unclear respiratory events, or mismatched dream content; it may be the case that lower levels of lucidity correspond with poor control of eye movements, respiration, and dream content.
For example, one patient dreamt she was swimming the breaststroke, and was repeatedly putting her head in and out of water with each stroke thus repeatedly holding her breath then inhaling deeply. This corresponded with recorded physiological measures, including repeated cessation of airflow followed by hyperventilation. However, there were no eye movement signals that preceded these respiratory behaviors, so it is not clear whether the dreamer was truly lucid at this time. Nevertheless, this is still strong evidence for the link between dream content and physical respiration during sleep.
Overall, the results support that lucid dreamers can actively influence respiration, abdominal and thoracic muscle movements, and eye movements through controlling dream content. This has broader implications for the relationship between the dreaming and sleeping body, and the authors suggest that even in nonlucid dreams, the content of mental imagery may physically manifest in the sleeping body, and may in part explain the erratic respiration, changes in heart rate, and muscular activity exhibited particularly during REM sleep. This research points to a stronger role for dreaming in sleeping physiology than has previously been acknowledged. For example, if habitual waking behaviors, such as physical activities or emotional dispositions, can influence dream content, then it is likely that this dream content also impacts the quality of REM sleep physiology. This has been supported to some extent by findings that heart-rate, muscle tension and muscle twitches, and eye movement activity correspond with emotional or sensory intensity of dreaming (e.g. nightmares are associated with increased heart rate and respiration).
Nevertheless, the authors express some limitations, namely that many participants reported that the physical act of holding their breath wasn’t truly a consequence of dream content, but rather a fully intentional act, as most participants found they did not need to alter their breathing in response to the dream. For example, some participants reported that they could quite easily breath underwater in their dream, but decided to force themselves to hold their breath for the experiment. Thus it’s possible the physical manifestation of dream content would be less potent in non-lucid dreams.
References
Oudiette, D., Dodet, P., Ledard, N., Artru, E., Rachidi, I., Similowski, T., & Arnulf, I. (2018). REM sleep respiratory behaviours match mental content in narcoleptic lucid dreamers. Scientific reports. https://doi.org/10.1038/s41598-018-21067-9
