Sexual Fantasies and Gender/Sex: A Multimethod Approach with Quantitative Content Analysis and Hormonal Responses

Abstract

Research links explicit sexuality (e.g., physical attraction and pleasure) to high testosterone (T) and nurturance (loving contact) to low T. Engaging in sexual fantasy, which can include explicit sexual and nurturant elements, increases T in women but not in men. We examined whether individual differences in the explicit sexual and nurturant content of fantasy were linked with T or with estradiol (E₂). In addition, we explored whether fantasy content differed or overlapped by gender/sex. Participants (26 women, 23 men) provided saliva samples for hormones before and after imagining a self-defined positive sexual encounter and responding to open-ended questions about the situation they imagined. We systematically content-coded responses for explicit sexual and nurturant content. In men, lower inclusion of nurturant content predicted larger T responses to fantasy. Fantasy content was not linked with T in women or with E₂ in women or men. Women and men did not differ significantly in explicit sexual and nurturant content. Our findings suggest that individual experiences of fantasy as more or less nurturant affect T in men, provide support for the Steroid/Peptide Theory of Social Bonds, and highlight the value of integrating hormones and content analysis to investigate research questions relevant to sexuality and gender/sex.

Acknowledgments

The authors would like to thank Elizabeth Cole, Sara McClelland, and David Winter for helpful discussions. K.L.G. was supported by a National Science Foundation Graduate Research Fellowship (grant no. DGE0718128).

Notes

¹We use the term gender/sex because it is impossible to definitively know whether any differences between women and men reflect biological sex differences, gender socialization, or a combination of these; this term has been used in previous publications (van Anders, 2010; van Anders, Goldey, & Kuo, 2011; van Anders, 2012b).

^a Self-identified in response to an open-ended item.

^b Participants indicated their relationship status based on definitions we provided (van Anders & Goldey, 2010).

^c Defined as contact with another person involving the participant's or the other person's genitals.

Note. Nicotine use was coded as a dummy variable (yes/no). Relationship status was a categorical variable (single, casually partnered, committed partnered) represented as two contrast codes in the regression.

²Participants are designated by an identification number preceded by W (for woman) or M (for man).