The eyes have it: sex and sexual orientation differences in pupil dilation patterns

Abstract

Recent research suggests profound sex and sexual orientation differences in sexual response. These results, however, are based on measures of genital arousal, which have potential limitations such as volunteer bias and differential measures for the sexes. The present study introduces a measure less affected by these limitations. We assessed the pupil dilation of 325 men and women of various sexual orientations to male and female erotic stimuli. Results supported hypotheses. In general, self-reported sexual orientation corresponded with pupil dilation to men and women. Among men, substantial dilation to both sexes was most common in bisexual-identified men. In contrast, among women, substantial dilation to both sexes was most common in heterosexual-identified women. Possible reasons for these differences are discussed. Because the measure of pupil dilation is less invasive than previous measures of sexual response, it allows for studying diverse age and cultural populations, usually not included in sexuality research.

Figure 1. Pupil dilation to the same sex and the other sex.

Panel A shows men’s responses and Panel B shows women’s’ responses. Y Axes reflect z-scores within participants: positive numbers indicate dilation to the same sex, and negative numbers indicate dilation to the other sex. X Axes reflect self-reported sexual orientation: 0 represents an exclusive heterosexual orientation, 3 an even bisexual orientation, and 6 an exclusive homosexual orientation. Triple lines represent regression coefficients with 95% confidence intervals. Dots represent participants’ average scores. β’s are standardized coefficients for linear effects.

Figure 2. Pupil dilation to the more arousing sex and less arousing sex.

Panel A shows men’s responses and Panel B shows women’s’ responses to the more arousing sex (i.e., more dilation-eliciting sex; upper lines) and the less arousing sex (lower lines). Y Axes reflect, within participants, z-scores compared to a neutral stimulus. X Axes reflect self-reported sexual orientation: 0 represents an exclusive heterosexual orientation, 3 an even bisexual orientation, and 6 an exclusive homosexual orientation. Triple lines represent regression coefficients with 95% confidence intervals. β’s are, for men, standardized coefficients for curvilinear effects, and, for women, standardized coefficients for linear effects.