Effects of all-night exposure to ambient odour on dreams and affective state upon waking

Previous laboratory research has shown that exposure to odours of contrasting pleasantness during sleep differentially affects the emotional tone of dreams. In the present study, we sought to investigate how a generally pleasant (vanillin) and unpleasant (thioglycolic acid [TGA]) smell influenced various dream characteristics, dream emotions, and post-sleep core affect during all-night exposure, controlling for appraisal of the olfactory environment during the assessments and sleep stage from which the participants woke up. We expected that exposure to vanillin would result in more pleasant dreams, more positive and less negative dream emotions, and a more positive post-sleep core affect compared to the control condition, whereas exposure to TGA would have the opposite effect. Sixty healthy volunteers (36 males, mean age 24 ± 4 years) were invited to visit the sleep laboratory three times in weekly intervals. The first visit served to adapt the participants to the laboratory environment. On the second visit, half the participants were exposed to an odour (vanillin or TGA, 1:1) and the other half to the odourless control condition. On the third visit, they received control or exposure in a balanced order. On each visit, the participants woke up twice, first from the rapid eye movement (REM) sleep stage and then in the morning, usually from a non-REM sleep stage. Repeated measures were taken upon each awakening. Dream pleasantness, emotional charge of the dream, positive and negative emotions experienced in the dream, and four dimensions of post-sleep core affect (valence, activation, pleasant activation - unpleasant deactivation, and unpleasant activation - pleasant deactivation) were assessed. We found a small effect of condition (exposure vs. control) in interaction with appraisal of the ambient olfactory environment on dream pleasantness. Specifically, false alarms (i.e., perceiving odour in the absence of the target stimulus) were associated with lower dream pleasantness than correct rejections. Although exposure had a statistically significant positive influence on post-sleep core affect (namely, valence, activation, and pleasant activation - unpleasant deactivation), the size of the effect was small and lacked practical significance. The hypothesised differential effects of vanillin and TGA were only modelled for dream ratings because they decreased the fit of the other models. Neither dream pleasantness nor emotionality differed according to the odour used for stimulation. The results of the present study suggest that all-night exposure to odours is unlikely to produce practically significant positive effects on dreams and post-sleep core affect.