0109 A Gender Disparity in the Experience of Fear in Dreams
Mallory Wells,Garrett Baber,Nancy Hamilton
DOI: https://doi.org/10.1093/sleep/zsae067.0109
IF: 6.313
2024-04-20
SLEEP
Abstract:Abstract Introduction Determining whether gender differences exist in dream experiences is crucial for understanding potential disparities in offline emotional processing. Women tend to experience more nightmares (i.e., dreams characterized by fear) than men, which could indicate a failure of overnight fear extinction. Yet, there is mixed evidence as to whether gender differences in dream emotions exist. To resolve this ambiguity, our study employed a novel natural language processing (NLP) tool on a large dataset to explore potential gender disparities in the experience of fear in dreams. Methods Data utilized in this study were sourced from Cunningham et al.'s (2020) sleep and mental health daily survey conducted at Boston College, involving 667 participants (Female = 569, Male = 98). Dream reports and sleep efficiency calculations were surveyed across several days and self-reported gender, age, and neuroticism scores were surveyed once. We applied a deep learning NLP model to dream reports to estimate a regression value representing fear intensity from the text. Mean values per participant were computed for dream fear, dream report word count, and sleep efficiency. We fitted a general linear regression model to test whether mean dream fear could be predicted using binary self-reported gender while controlling for mean dream report word count, mean sleep efficiency, neuroticism, and age. Results Female identity was significantly associated with higher mean dream fear independent of mean sleep efficiency, mean dream report word count, trait neuroticism, and age, β = -0.03, SE = 0.01, 95% CI: [-0.05, -0.01], p < .001. All variables except for age were significantly related to mean dream fear. Conclusion We found that those who experienced greater fear in dreams were more likely to be female identifying, after controlling for other variables. Mean sleep efficiency, mean dream report word count, and neuroticism were significantly related to higher mean dream fear. Word count may be related to dream fear due to a bias in the NLP model. Future research should investigate differences in dream fear among other gender identities, collect a more representative sample, and include measures of stress and trauma as possible explanatory variables of gender disparities in dream fear. Support (if any)
neurosciences,clinical neurology