The Effect of Intranasal Oxytocin on the Association Between Couple Interaction and Sleep: A Placebo-Controlled Study.

Objective: Although most people in romantic relationships cosleep, biosocial modulators of sleep quality have only recently come into focus. Oxytocin (OT) might be one such modulator, as it had been shown to increase social attachment and safety. We investigated the association between everyday life couple interaction and sleep quality, as well as the effects of OT on this association.

Methods: Eighty heterosexual couples ( N = 160 individuals, mean [standard deviation] age = 28 [5] years) were randomized to self-administer a) 32 international units of intranasal OT or b) placebo during 5 consecutive days. Each morning, they reported on sleep quality, and on subjective feelings of closeness and valence of couple interaction at a maximum of four times a day. Data were analyzed using hierarchical linear models.

Results: Subjective closeness ( B = 0.43, t (73) = 3.80, p < .001) and valence (negative - positive) of couple interaction ( B = 0.50, t (73) = 3.91, p < .001) were positively associated with sleep quality. Persons with OT reported higher levels of sleep quality than those without ( B = 0.47, t (74) = 2.32, p = .023). The association between closeness and sleep quality was stronger with OT than without (OT by closeness: B = 0.31, t (72) = 2.29, p = .025; OT by valence of interaction: B = 0.27, t (72) = 1.77, p = .081). Whereas the effect of couple interaction on sleep quality was strong in men, the OT effects were especially pronounced in women.

Conclusions: Our results suggest that enhancing closeness and positive couple interaction in cosleeping partners might be a way to improve sleep quality. The moderating effects of OT and sex on the association between couple interaction and sleep quality can have important implications for sleep therapy.Trial Registration: The study was preregistered at ClinicalTrials.gov ("Oxytocin, Couple Interaction, and Wound Healing" study, identifier NCT01594775). The present analyses were not preregistered.