The impact of romantic relationships on mental health is well-established, with sleep potentially mediating this relationship. Sharing a bed with a partner is common, yet its effects on sleep are inconsistent across studies. Previous research using actigraphy has yielded conflicting results, showing either disrupted sleep, no effect, or improved sleep depending on the study. Actigraphy, however, lacks the neurophysiological detail provided by polysomnography, which measures sleep stages directly. Only a few polysomnographic studies exist, with conflicting findings regarding the impact of co-sleeping on various sleep parameters. These studies have also lacked consideration of important covariates such as relationship characteristics and chronotype. Interpersonal synchronization during wakefulness is associated with positive social outcomes, suggesting that sleep-stage synchrony during co-sleep might be an important factor. This study used dual simultaneous polysomnography and cross-recurrence quantification analysis to investigate the impact of bed partner presence on objective and subjective sleep parameters, including sleep-stage synchronization, in a larger sample of young heterosexual couples. The study included a confirmatory part to reassess previous pilot study findings, and an exploratory part to investigate the influence of relationship factors and complex forms of interpersonal coupling.

Prior research on the effects of bed sharing on sleep has yielded mixed results. Actigraphic studies comparing co-sleep to individual sleep have reported either more disrupted sleep patterns (in both or only women), or increased sleep time in men. Studies comparing married couples to single individuals have shown longer total sleep time and less wakefulness after sleep onset in married couples. Furthermore, some research indicates synchronization of movements and increased sleep-wake concordance during co-sleep, which are influenced by relationship characteristics. However, actigraphy's limitations in assessing sleep stages prevent a complete understanding of the neurophysiological impact of bed sharing. Existing polysomnographic studies are limited in number and sample size and show inconsistencies in their findings beyond a consistent increase in REM sleep during co-sleep. These studies have also neglected crucial factors like relationship characteristics and chronotype.

Twenty-four childless, healthy young adults (12 heterosexual couples) participated in a within-subjects study involving four nights in a sleep laboratory. Couples spent two consecutive weekends in the lab, with one weekend dedicated to individual sleep in separate rooms and the other to co-sleep in adjacent beds. The first night of each weekend served as an adaptation night and was excluded from analysis. Participants completed questionnaires assessing inclusion criteria (absence of shift work, pregnancy, sleep-affecting medications or disorders, depression, addictions, etc.), chronotype, relationship characteristics (duration, passionate love, conflict, social support, and relationship depth), and subjective sleep parameters. Polysomnography data were collected using EEG, EOG, EMG, ECG, pulse oximetry, and respiratory parameters. Sleep stages were manually scored according to AASM criteria, and leg movements were calculated automatically. Cross-recurrence quantification analysis was used to assess sleep-stage synchronization between partners, considering both simultaneous and lagged synchronization. Paired t-tests and Wilcoxon signed-rank tests were used to compare co-sleep and individual sleep conditions. ANOVAs and ANCOVAs were conducted to examine the effects of gender, relationship characteristics, chronotype, snoring, and leg movements on sleep parameters and synchronization.

Compared to individual sleep, co-sleeping was associated with approximately 10% more REM sleep (p = 0.005), less fragmented REM sleep (p = 0.008), longer undisturbed REM fragments (p = 0.0006), and more limb movements (p = 0.007). No significant differences were found for other sleep stages. Social support interacted with sleeping arrangement, with individuals having lower social support showing a more pronounced increase in REM sleep during co-sleep. Sleep architectures were significantly more synchronized between partners during co-sleep (p = 0.005), even when wake phases were excluded (p = 0.022). Significant coupling of sleep stages was observed across a lag of ±5 min during co-sleep. Relationship depth was positively associated with synchronization. Gender, chronotype, and other relationship characteristics did not significantly influence REM sleep or synchronization.

The findings support previous reports of increased REM sleep during co-sleep, but contradict some findings regarding other sleep parameters and the absence of interaction effects of gender and sleep arrangement. The increased and stabilized REM sleep during co-sleep, coupled with increased sleep-stage synchronization, suggests a possible positive feedback loop between social interaction and REM sleep. The study proposes that REM sleep, with its potential benefits for memory consolidation and emotional processing, enhances social interaction, and this enhancement stabilizes REM sleep. The REM sleep-related effects of co-sleeping might play a role in the protective effect of partnerships against mental illness, given the link between REM sleep fragmentation, insomnia, and mental health problems. The increased sleep-stage synchronization is also significant, suggesting that social interactions extend into sleep and further support the idea of co-sleeping as a mechanism affecting mental well-being. The lack of other changes in sleep parameters besides REM sleep and increased movements may be because co-sleeping induces alterations distinct from those related to typical good or poor sleep quality.

This study provides evidence for increased and stabilized REM sleep and enhanced sleep-stage synchronization in co-sleeping couples. Social support and relationship depth appear to be important factors influencing these effects. The findings suggest a potential positive feedback loop between sociality and REM sleep, highlighting the possibility that co-sleeping may contribute to mental well-being through REM sleep stabilization and interpersonal synchronization. Future research should explore these relationships further, with larger samples and incorporating a dyadic approach to analyze couple dynamics.

The study's relatively small sample size, particularly for the ANOVAs and ANCOVAs, limits the generalizability of some exploratory findings. The laboratory setting might not fully reflect real-world co-sleeping conditions. The lack of actigraphy data prevents a direct comparison with studies using this method. While the data support an analysis at the individual level, future studies could benefit from a dyadic approach. Furthermore, the study focused on young, healthy heterosexual couples, limiting the generalizability to other populations and relationship types.