Introduction

In creating his Heterosexual–Homosexual rating scale, Kinsey conceptualized sexual orientation on a Likert-type scale from exclusive heterosexual attraction (score of 0) to exclusive homosexual attraction (score of 6), with bisexuality falling in the middle (Kinsey, Pomeroy, & Martin, 1948/1998; Kinsey, Pomeroy, Martin, & Gebhard, 1953/1998). Kinsey also recognized that some persons are asexual, which he described as an absence of “social-sexual contact or reactions” (Kinsey et al., 1948/1998). Indeed, asexual persons are thought to exhibit little or no sexual attraction (Bogaert, 2004; Hinderliter, 2009). However, the orientation of asexuality exists outside of the Kinsey 6-point rating scale, because of the inherent difficulty incorporating persons who report a complete lack of interest in sexual behavior within Kinsey’s one-dimensional conceptualization of sexual orientation. In addition, the Kinsey scale does not consider diverse descriptions of sexual identity, such as pansexual or non-binary gender, which in recent years anecdotally appear to have gained limited acceptance among some small and diverse communities. Therefore, the Kinsey scale may not adequately reflect current expressions of sexual identity, particularly in Western society.

Storms (1980) suggested that other conceptualizations of sexual orientation relate to either sex-role orientation or to erotic orientation. Specifically, sex-role orientation suggests that sexual attraction occurs because individuals identify with various presentations of masculine–feminine sex roles. In contrast, erotic orientation suggests that sexual orientation arises because individuals develop an erotic responsiveness toward stimuli associated with either one sex or the other. Storms found that sexual orientation is primarily related to erotic fantasy, which lies along two independent dimensions of heterosexuality and homosexuality (see Fig. 1).

Fig. 1
figure 1

Storms’ (1980) model of sexual orientation which quantifies four categories of orientation along two-dimensional axes

Full size image

A primary criticism of Storms’ model of sexual attraction is that it does not account for non-binary expressions of gender outside of the “normative” male–female representation of gender. For example, Storms’ model of sexual attraction does not reasonably account for composite expressions of sexuality, such as asexual persons who also identify as aromantic, heteroromantic, biromantic, or homoromantic. In addition, would aromantic persons be reasonably able to answer some of the scale items contained in Storms’ Erotic Response and Orientation Scale (EROS) (e.g., “How often have you masturbated while fantasizing a sexual experience with a woman?”).

Diverse Expressions of Sexual Orientation

We recently surveyed a large international sample of male, female, transgender, and intersex persons 18 years of age and older. The sample was recruited primarily in Western countries (e.g., United States, Canada, Australia, and United Kingdom). An anonymous online survey was used to evaluate whether clinical profiles exist for various presentations of hypersexual behavior. Preliminary examination of responses suggested that participants diversely described their sexual orientation well beyond the clear-cut categorical orientations of heterosexual, bisexual, and homosexual. The range of expressions of sexual orientation found in our diversely recruited sample is worthy of comment, and may be useful for researchers and clinicians to consider when asking persons about their sexual orientation in either research projects or in clinical assessment.

As background, at the beginning of our questionnaire battery, we asked participants “How would you describe your sexual orientation?” for which the response choices provided were heterosexual/straight, bisexual, gay, lesbian, and other. Participants in the study were required to select one out of the five listed response options to describe their sexual orientation. Participants who selected the category of “other” were asked to specifically describe the sexual orientation to which they identified.

A total of 2295 persons completed the sexual orientation survey item described, which comprised male (n = 1176), female (n = 871), transgender male-to-female (n = 129), transgender female-to-male (n = 100), and intersex (n = 19). The mean age of participants by sexual orientation was heterosexual (M = 36.17 years, SD = 14.30), bisexual (M = 30.89 years, SD = 13.21), gay (M = 36.49 years, SD = 15.04), lesbian (M = 28.91 years, SD = 11.20), and other (M = 26.66 years, SD = 10.08). In total, 310 (13.5 %) participants identified their sexual orientation as “other” (see Table 1).

Table 1 Breakdown of sexual orientation by gender
Full size table

Participants who were female (18.3 %), transgender male-to-female (27.1 %), transgender female-to-male (26.0 %), and intersex (47.4 %) variously described their sexual orientation within the subcategory of “other,” in comparison to only 6.9 % of male participants. Subsequent analysis of the subcategory of “other” found that the most commonly identified sexual orientation by gender was “exclusive” asexual for male and female participants (male: n = 38, 3.2 % of the sample; female: n = 76, 8.7 % of the sample). Pansexual was the most commonly identified sexual orientation for transgender male-to-female and transgender female-to-male participants, respectively (n = 18, 14.0 %; n = 9, 9.0 %), as well as pansexual and unsure–unknown sexual identities (in equal proportions) for intersex participants (n = 2, 10.5 %). A minority of participants also identified their sexual orientations as composite in nature. For example, some participants further categorized their sexual orientation of asexual to include various sexual identities across the romantic spectrum.

Asexuality and Sexual Activity

Following a preliminary analysis of the survey data, we noticed that within the sexual orientation category of “other,” approximately 41.6 % of participants chose “exclusive” asexual as their sexual identity. Indeed, asexual was the sexual orientation identified by participants next most often outside of the three more commonly recognized sexual orientations of heterosexual, bisexual, and homosexual. We performed a series of independent samples t-tests to explore whether asexuality may simply indicate very low levels of sexual arousal or an absence of sexual arousal for these persons. The t-tests were also conducted because asexuality may be considered a distinct sexual orientation for those persons who exhibit no interest in sexual activity. When conducting our analysis, participants were categorized into two groups: persons who identified their sexual orientation as asexual (including asexual spectrum) (n = 155, 6.8 % of the recruited sample, M age = 23.99 years, SD = 7.0), and persons who identified their sexual orientation as non-asexual (n = 2140, 93.2 %, M age = 33.80 years, SD = 14.1).

Four independent samples t-tests were performed to explore differences between asexual and non-sexual groups regarding their frequency of sexual activity for the sexual behaviors of intercourse, masturbation, kissing and petting, and sexual fantasy. A further four independent samples t-tests were performed to evaluate whether significant differences existed between asexual and non-asexual groups regarding their chronological age, level of sexual arousal (propensity for sexual excitation), and levels of anxiety and depressed mood. Because a total of eight independent samples t-tests were performed, we also performed a Bonferroni correction which changed the criterion for significance to p ≤ .006.

Sexual activity frequency was measured with an adapted version of the Derogatis Sexual Functioning Inventory (DSFI): Frequency of Sexual Activities subscale (Derogatis & Melisaratos, 1979). Participants indicated the frequency with which they typically engaged in sexual behaviors of intercourse, masturbation, kissing and petting, and sexual fantasy. Items were rated on a 9-point Likert scale that ranged from 1 (not at all) to 9 (four or more times per day). Research has previously found that the DSFI is a reliable and valid measure of sexual activity/behavior (Derogatis, 2011; Derogatis & Melisaratos, 1979).

Propensity for sexual arousal was measured with the Sexual Excitation subscale of the Sexual Inhibition/Sexual Excitation Scale (SIS/SES) (Janssen, Vorst, Finn, & Bancroft, 2002). The SES subscale assesses an individual’s propensity for sexual excitation/arousal, with higher scores indicating an increased propensity for these characteristics. For example, “When I look at erotic pictures, I easily become sexually aroused.” Items were rated on a 4-point Likert scale from 1 (strongly disagree) to 4 (strongly agree) and summed to produce a SES subscale score of 20–80. Psychometrically, the SES subscale has been found to exhibit high internal consistency, with Cronbach’s α ranging from .88 to .89 (Janssen et al., 2002).

Depressed mood and anxiety were measured using the Depression Anxiety Stress Scales (DASS-21) (Antony, Bieling, Cox, Enns, & Swinson, 1998; Lovibond & Lovibond, 1995). The DASS-21 depression and anxiety subscales each comprise a 7-item self-report measure of participants’ depressed mood and anxiety over the past week. Subscale items for depression included “I couldn’t seem to experience any positive feeling at all” and anxiety “I felt scared without any good reason.” Items were rated on a 4-point Likert scale from 0 (did not apply to me at all) to 3 (applied to me very much, or most of the time). Items were summed to provide total subscale scores of between 0 and 21, with higher scores representing greater levels of depressed mood and anxiety. Subscale scores for depression and anxiety were also multiplied by a factor of 2 to interpret the results against the full version of the DASS (Lovibond & Lovibond, 1995). Previous research suggests high concurrent validity for the subscales of depression (r = .87) and anxiety (r = .84), and high internal consistencies for the subscales of depression (α = .96) and anxiety (α = .89) (Akin & Cetn, 2007).

As frequently occurs with anonymous research questionnaires conducted online, we observed a sizable attrition of participants as they worked their way through the lengthy questionnaire battery. The completion rate for the recruited sample was approximately 68 %. A total of 155 asexual (spectrum) persons answered the chronological age item which was located at the beginning of the questionnaire, and 132 asexuals completed the DSFI-Frequency of Sexual Activities subscale items. In total, 121 asexuals completed the sexual excitation (sexual arousal) subscale. Furthermore, 106 asexuals and 1529 non-asexuals completed the DASS-21 anxiety and depression subscales located in the second half of the online survey. Notwithstanding that some asexuals elected not to fully complete the survey, the proportion of asexual participants by gender remained at similar levels throughout the aforementioned points in the survey.

Results indicated that asexuals, on average, engaged in intercourse significantly less often (M = 1.32 [equated to a DSFI rating scale of approximately not at all], SD = .98) compared to non-asexuals (M = 3.22 [approximately 1–2 times per month], SD = 1.83), t(208.09) = −19.76, p < .006. The mean difference in intercourse scores between the two groups was −1.90 (95 % CI −2.09 to −1.71), and Cohen’s d = 1.29, which indicated a large effect size. Second, asexuals reported significantly less masturbation (M = 3.50 [between 1 and 2 times per month to once per week], SD = 2.13) compared to non-asexuals (M = 5.04 [approximately 2–3 times per week], SD = 2.05), t(1873) = −8.26, p < .006. The mean difference in masturbation scores between the two groups was −1.54 (95 % CI −1.90 to −1.17), and Cohen’s d = 0.74, which approximated a large effect size. Third, asexuals reported significantly less kissing and petting (M = 1.84 [approximately less than once per month], SD = 1.84) compared to non-asexuals (M = 4.19 [approximately once per week], SD = 2.66), t(175.21) = −13.62, p < .006. The mean difference in kissing and petting scores between the two groups was −2.35 (95 % CI −2.69 to −2.01), and Cohen’s d = 1.03, which also indicated a large effect size. Fourth, on average, asexuals reported engaging in significantly less sexual fantasy (M = 2.33 [approximately less than once per month], SD = 2.03) compared to non-asexuals (M = 5.20 [approximately 2–3 times per week], SD = 2.63), t(166.29) = −15.31, p < .006. The mean difference in sexual fantasy scores between the two groups was −2.87 (95 % CI −3.24 to −2.50), and Cohen’s d = 1.22, which once again indicated a large effect size.

Regarding propensity for sexual arousal, asexuals reported, on average, significantly less sexual excitation (M = 27.44, SD = 6.66) compared to non-asexuals (M = 54.02, SD = 10.19), t(165.19) = −40.49, p < .006. The mean difference in sexual arousal scores between the two groups was −26.58 (95 % CI −27.88 to −25.29), and Cohen’s d = 3.09, which indicated a large effect size. When analyzing participants’ chronological age, asexuals were significantly younger (M = 23.99, SD = 6.96) than non-asexuals (M = 33.80, SD = 14.10), t(257.42) = −15.40, p < .006. The mean difference in age scores between the two groups was −9.81 years (95 % CI −11.06 to −8.55), and Cohen’s d = 0.88, which also indicated a large effect size.

We also performed two independent samples t-tests to determine whether lower levels of sexual behavior for asexuals may be attributable to dysphoric mood compared to non-asexuals. Asexuals reported, on average, similar levels of anxiety (M = 7.47, SD = 7.57) as non-asexuals (M = 6.46, SD = 7.10), t(1633) = 1.42, p = .16. The mean difference in anxiety between the two groups was 1.01 (95 % CI −.39 to 2.42), and Cohen’s d = 0.14, which indicated a small effect size. A clinical cut-off score of 0–7 on the DASS anxiety subscale indicates normal levels of anxiety, whereas a score of 8–9 indicates mild levels of anxiety (Lovibond & Lovibond, 1995). Furthermore, asexuals reported similar levels of depressed mood (M = 12.55, SD = 11.95) as non-asexuals (M = 9.31, SD = 10.17), t(115.80) = 2.72, p = .007. The mean difference in depressed mood scores between the two groups was 3.24 (95 % CI 0.88 to 5.59), and Cohen’s d = .29, which indicated a small-medium effect size. A score of 0–9 on the DASS depression subscale suggests normal mood levels, whereas a score of 10–13 suggests mild levels of depressed mood (Lovibond & Lovibond, 1995).

Given the 10-year difference in average age between the asexual and non-asexual groups, analyses of covariance (ANCOVAs) were performed with age entered as a covariate. Similar significant differences were obtained between the two groups for all dependent variables measured in the independent samples t-tests.

Participants also responded to the survey item “I am interested in sexual activity” for which the response choices were either “yes” or “no.” The percentage of asexuals and non-asexuals who selected a “no” response to this item was 90.15 % (n = 119) and 4.36 % (n = 76), respectively. A chi-square test indicated a significant association between asexual/non-asexual groups and their interest in sexual activity, χ2(1, n = 1874) = 959.47, p < .001, phi = −.72; considered a large effect size using Cohen’s (1988) criteria.

Asexuality: Low Sexual Arousal and Biology

Our results indicated that self-identified asexuals engaged in significantly less sexual behavior in terms of intercourse, masturbation, kissing and petting, and sexual fantasy than non-asexuals. In addition, differences in sexual behavior between asexual and non-asexual groups corresponded to large Cohen’s d effect sizes. Low sexual behavior by asexuals may be partially explained by their significantly lower sexual arousal compared to non-asexuals. The results also suggested that no significant differences existed between asexuals and non-asexuals regarding their levels of anxiety or depressed mood, which could reasonably explain low sexual arousal among the asexuals.

Although asexuals exhibited significantly lower sexual arousal compared to non-asexuals, for asexuals, Spearman correlations (rho [ρ]) between SES and sexual activity frequency were positively related for intercourse (ρ = .15, p = .053), masturbation (ρ = .40, p < .01), kissing and petting (ρ = .18, p < .05), and sexual fantasies (ρ = .53, p < .01). For non-asexuals, corresponding Spearman correlations between SES and sexual activity frequency were positively related for intercourse (ρ = .20, p < .01), masturbation (ρ = .34, p < .01), kissing and petting (ρ = .11, p < .01), and sexual fantasies (ρ = .36, p < .01). Results suggested that although asexuals exhibited low sexual arousal, their frequency of masturbation, kissing and petting, and sexual fantasies were positively related to sexual arousal, even though asexuals engaged in sexual behavior at much lower levels than non-asexuals.

Asexuals (90.15 %) overwhelmingly reported they had no interest in sexual behavior, notwithstanding the fact that these persons consistently experienced low sexual arousal and engaged in low levels of sexual activity. Conversely, 95.6 % of non-asexuals reported they were interested in sexual behavior. Taken together, the results suggested that most asexuals were not psychologically interested in sexual behavior, although they reported low sexual arousal and sexual activity. In contrast, our results suggested that most non-asexuals were psychologically interested in sexual activity, and experienced significantly higher sexual arousal and sexual activity compared to asexuals.

We acknowledge that for some asexuals their sexual activity occurs because they live in committed relationships and may engage in sexual behavior principally to meet their partner’s wishes rather than because of either sexual interest or arousal (Van Houdenhove, Gijs, T’Sjoen, & Enzlin, 2015). However, the results also suggested that many asexuals engaged in low levels of solo masturbation and sexual fantasy irrespective of their relationship status. Therefore, when taken together, the results beg the question—why did our asexual sample generally report “no” psychological interest in sexual behavior, yet also reported some (albeit low) levels of sexual arousal and sexual behavior?

We hypothesize that perhaps asexuals’ low sexual arousal/behavior may in part be biologically explained. In addition, perhaps asexuals have a biologically innate need to engage in low levels of sexual behavior which overrides their reported psychological disinterest in sexual activity. Indeed, sexual behavior is reported to have evolved since prehistoric times and most humans appear primordially predisposed to engage in sexual behavior at varying levels (Gray & Garcia, 2013; Long, 2014). Diamond (1965) suggested that sexual behavior is a composite mix of prenatal and postnatal influences, with postnatal factors superimposed on various biologically inherent forms of human sexuality. In addition, Epstein, McKinney, and Fox (2012) suggested that sexual orientation may exist on a fluid continuum, and the range and point along the continuum differs between individuals, and is determined by a combination of various genetic and environmental factors.

Social Media and Education

Asexual participants in this study were principally recruited by targeting diverse sexual communities listed online. Therefore, given results indicated that asexuals were significantly younger in age than for non-asexuals (M = 23.99 years vs. M = 33.80 years), it is worth considering the role of social media in facilitating the diversity of sexual expression, particularly in younger persons who are also more likely to be female than male. Since the mid-1990s, commercialization of the internet world-wide has had a profound impact upon social culture, with persons electronically communicating via email, instant messaging, discussion forums, and blogs (Boyd & Ellison, 2008; Ryan, 2010). In addition, social networking has facilitated like-minded persons to form diverse and evolving social communities which transcend nationality and culture. Social networking has assisted persons en masse to discuss (often anonymously) and freely explore with each other their sexual orientation. Prior to commercialization of the internet, it is likely that human sexual expression was more confined to the “big three” sexual orientations, because significant barriers inhibited persons of diverse sexual orientations from connecting with like-minded individuals.

Today, school-aged children are more educated about diversity of sexual orientation. Australian primary school children as young as age 11 are currently taught about gender diversity as part of the Safe Schools program (McNally, 2015). These educational messages include teaching children about diverse sexual orientations, such as pansexual and gender queer, as well as gender neutrality. Given the roll-out of educational programs such as Safe Schools within Australia and most likely similar programs which operate internationally, it is not surprising that young people are diversely expressing either their gender or sexual orientation. In the future, a diverse range of genderless categories and new meanings which redefine “sexual citizenship” (Gressgard, 2013) may become the norm, although these new personal identities will most likely continue to be minority expressions of sexual identity in comparison to the “big three” recognized sexual categories.

Conclusion

We recognize that our sample may not accurately reflect either the prevalence or diversity of sexual orientation in the general population, in part because asexual and transgendered persons were specifically targeted during recruitment, and as such are likely to be disproportionately represented in the sample. In addition, 19 intersex participants were recruited for the survey, which is too small of a population sample to reliably reflect the likely diversity of sexual orientation within this subgroup of gender. Furthermore, there existed differences in the proportion of males and females between the asexual and non-asexual groups. Notwithstanding these research limitations, the study provides valuable insight into the current diversity with which many individuals uniquely express their sexuality beyond the commonly accepted “big three” expressions of sexual orientation.

Results suggest that a significant minority of individuals diversely express their sexual orientations, and these individuals are likely to view their sexual orientation as integral to their personal identity and about which they hold deep personal convictions. For example, although same-sex attraction between females is clinically recognized as homosexual, females of same-sex attraction who present for treatment will often report their sexual orientation as lesbian, and variously describe their sexual orientation as gay, dyke, queer, gender queer, asexual lesbian, or lesbian demisexual.

Our results suggest that further research of various sexual orientations may be useful to provide validity to some of the more commonly identified sexualities within the category of “other,” such as asexual, pansexual, and romantic spectrum attractions. To date, asexuality is the most investigated sexuality outside of the “big three” sexual orientations, although clinical research of asexual orientation is somewhat sparse. Research suggests that asexuals have little or no sexual attraction, low sexual excitation, low sexual desire, and arousal, as well as low frequency of sexual behavior (Bogaert, 2004; Brotto, Knudson, Inskip, Rhodes, & Erskine, 2010; Prause & Graham, 2007). Indeed, some research suggests that asexuality may be more usefully categorized as a distinct sexual orientation (Brotto, Yule, & Gorzalka, 2015).

We suggest that further research is required to reliably determine whether these different sexual orientations are simply variations of the “big three” expressions of sexual identity or orthogonal, distinct, and independent sexual identities to these main categories of sexual expression. Furthermore, our results raise the question—do some of these diverse and less common expressions of sexual orientation more accurately reflect variations in sexual arousal (as opposed to sexual attraction) or, alternatively, reflect distinct and independent sexual orientations?

Taken together, our results suggest that sexual orientation represents a rich and diverse tapestry of human sexual and romantic expression that is beyond current typical recognition and understanding. We also theorize that various levels of sexual arousal could include biological factors that to date are unexplained by sex research.