A team of researchers from Austria and Switzerland have produced evidence suggesting that the brain can distinguish between familiar and unfamiliar voices during sleep.

The study, published this month in the Journal of Neuroscience, exposed sleeping participants to recordings of a voice familiar to the subject, such as a friend or relative, and an unfamiliar voice. The recordings consisted of subject’s own name as well as two other names. Participant responses to this stimuli were measured by polysomnography data collected over the course of a full night.

Measuring brain activity during sleep using electroencephalography

Polysomnography is the process by which a suite of physiological data is collected during sleep. Some of this data is collected by electroencephalogram (EEG), a test that quantifies electrical activity in the brain via electrodes attached to the scalp. The researchers focused on three EEG events related to information processing or sensory inhibition: K-complexes, sleep spindles, and micro-arousals.

K-complexes and sleep spindles are hallmarks of non-REM stage 2 (N2) sleep. Described as “the beginning of proper sleep”, adults spend half their sleeping time in the N2 sleep stage. K-complexes are patterns that can occur spontaneously but are more often elicited by sensory stimulation. Sleep spindles are also evoked by sensory input and involve communication between the thalamus and the cortex. The function of both EEG events is disputed. K-complexes have been linked to both arousal and sleep preservation. Sleep spindles may have a role in sensory inhibition or memory-related auditory processing.

Micro-arousals can occur in any phase of sleep and consist of brief shifts towards wave patterns associated with arousal. These events are thought to be involved in information processing.

Evidence of sensory, but not linguistic, processing

The researchers observed change in sleep stage evoked by the voices, but unfamiliar voices did result in a higher incidence of EEG events within a sleep stage. This increased EEG activity had no relationship to whether participants were exposed to their own name or another name. While voice familiarity had no significant effect on sleep spindles and only an insignificant effect on micro-arousals, K-complexes occurred significantly more often in response to an unfamiliar voice. K-complexes resulting from an unfamiliar voice also had a stronger peak than those induced by a familiar voice.

This increase in K-complexes in response to the unfamiliar voice decreased in the second half of the night while there was no change in K-complexes evoked by familiar voices. These results suggest that as participants gained exposure to the unfamiliar voice, the increased familiarity led to decreased unfamiliarity-evoked activity.

When the researchers analyzed the brain activity surrounding the K-complexes, the responses following the unfamiliar voices appeared significantly more synchronized than those following familiar voices, implying coordinated sensory processing in response to unfamiliar voices.

The authors propose that unfamiliar voices are more “relevant” than familiar voices because they represent a greater threat from an evolutionary perspective. The longer a person is exposed to these voices, they are perceived as less of a threat and evoke less of a response so as to not disturb sleep. They argue that the lack of difference in response elicited by a non-subject name compared to a subject’s own name demonstrates an inhibition of linguistic processing during sleep while the sensory processing involved in distinguishing voices is somewhat preserved.

What does this research tell us about sleep?

This study only involved 22 participants, all female, so it is too early to make sweeping claims about how the brain interprets unfamiliar voices during sleep based on this data alone. If this study were replicated with a larger sample size, the greater statistical power may also uncover a significant change in micro-arousals.

It is still unclear if what benefits these EEG events confer to a sleeping person as they do not result in any change to sleep stage. The authors caution that while this data provides evidence for active sensory processing during sleep, the altered activity in response to unfamiliar voices should not be interpreted as an actual disturbance to sleep.

This study enhances scientific understanding of auditory processing during sleep. Future research may contextualize these findings within a more robust body of knowledge regarding sensory activity in the sleeping brain.