The quality and complexity of dreams appear to change with our stages of sleep, according to a new analysis.
Before the twenty-first century, we used to think dreams only occurred during rapid eye movement (REM) sleep, but more recent research shows people sometimes recall dreams even when they are woken from non-REM stages of sleep.
Whether these two types of dreaming are inherently different is something neuroscientists are still trying to figure out.
When patients are woken during REM sleep, research shows they can usually recall elaborate, vivid, and emotional story-like dreams. In contrast, those woken during non-REM stages remember their dreams less, and the dreams themselves tend to be more thought-like.
These are important findings, but they are also based on subjective reports. REM dreams are often described in more words, for instance, but when the length of the description is controlled for, differences in elaboration disappear or are highly diminished.
Researchers in Brazil have now developed a high-speed analysing tool that can take these qualitative reports and display them in a more objective graph form, taking into account biases for both length and language.
“We know REM dreams are longer and more like movies,” says neuroscientist Sidarta Ribeiro from the University of São Paulo in Brazil.
“Automating the process of analysis, as we did in the study, made possible the first-ever quantitative measurement of this structural difference.”
Compared to traditional methods, which rely on parsing out the meaning of words, this non-semantic graph analysis was able to instead focus on the overall tone of what was said.
Focusing on 133 previously collected dream reports from 20 participants, who were woken at different stages of dreaming, researchers graphed out the words, replacing them with nodes on a graph.
Analysing their structural organisation, the new tool found REM dream reports were far more complex and full of connected information compared to dreams during non-REM sleep.
And this was true regardless of the report’s length.
“This is the first study to use graph theory to show that REM dream reports have more structural connectedness than non-REM dream reports,” says neuroscientist Joshua Martin from Humboldt University in Berlin.
“Not to depreciate the relevance of traditional methods, but these results are important because they show that computational methods can be applied to studies of dreaming.”
While non-REM sleep is suspected of having some restorative function, we’re still not really sure why REM sleep exists. If dreaming during this stage is truly of a different quality, as this new research suggests, then REM and non-REM dreaming might be driven by distinct underlying mechanisms that could play differing roles in our biology.
Compared to REM dreams, dreams from the N2 stage – a deep, non-REM, slow-wave sleep – were shorter, less frequently recalled, less intense, and more thought-like.
Of course, sleep studies come with lots of limitations beyond mere subjectivity. Being woken up continuously throughout the night could itself be impacting the quality of sleep among volunteers.
Recall of dreams might also be warped by sleep inertia – that weird stage between waking and sleeping – although dreams’ narrative complexity appears to stay the same even once participants have woken up properly.
While complex dream narratives can still occur in non-REM sleep, the authors suspect the very physiology of REM sleep, which shows great cortical activity and muscle atonia, is a better time for interactive narratives to unfold uninterrupted.
“In this sense, dream experiences that are coherent, immersive, and story-like may be more easily organized into a report with larger connectedness, while dream experiences that are fragmented and isolated are relatively more difficult to organize mentally and thus are structurally less connected,” the authors explain.
Not only do the results of the study complement existing literature on dream reports and REM sleep, but they also support recent and more objective measurements of dream bank databases.
A study published in 2020, for instance, used an algorithm to sift through 24,000 dreams and found various “statistical markers” that support the hypothesis that our dreams are a continuation of everyday life.
One algorithm isn’t enough to put this mystery to bed, but mathematical tools like this one could be useful when it comes to assessing our sleep and our dreams with as little bias and with as many considered factors as possible.
The current study was conducted at a much smaller scale, but it offers some of the first really objective measurements on dreams that we’ve got.
The study was published in PLOS ONE.
