Introduction

Several groups have been identified as high risk for HIV/sexually transmitted diseases (STD) acquisition. The risk of HIV/STD acquisition is related to the low frequency of condom use in vaginal or anal contact, earlier vaginal or anal sexual debut, greater number of sex partners, HIV/STD burden in the sexual network and concomitant sex encounters, as well as the concurrent use of substances before or during sex. The prevalence of HIV/AIDS in Israel in 2010 was 5.6 cases per 100,000 population, and included mainly heterosexual migrants from countries of generalized HIV epidemic, intra-venous drug users, and men who have sex with men (MSM), who accounted for 33.4 % of all male cases reported between 1981 and 2010. A predominant increase of HIV/AIDS burden in Israel was reported among MSM in the last 15 years, who have reached 60 % of all males newly notified with HIV/AIDS in recent years (Mor, Weinstein, Grotto, Levin, & Chemtob, 2013).

Sexual orientation can be described by sexual attraction, self-definition, and practices. These three dimensions of sexual orientation are correlated, yet no single dimension captures the full complexity of sexual orientation. Self-identification categories mask a diversity of sexual behaviors exposing them to variable risks (Lhomond, Saurel-Cubizolle, Michaels, & CSF Group, 2014). Heterosexuals do not constitute a single, homogeneous group. The definition of “heterosexual” differs based on what measure of sexual orientation is being captured, as some persons define themselves as heterosexual, but who do not report purely heterosexual attraction or behavior, as they may have engaged in same-sex encounters and/or report some same-sex attraction. Given that discordant heterosexuals do not identify themselves as gays or bisexuals, they are distant from the gay community, where most activities that focus on HIV prevention are performed. Therefore, discordant heterosexuals may not be exposed to these interventions and can be at increased risk for HIV/STD (Gattis, Sacco, & Cunningham-Williams, 2012). Females who are attracted to the same sex were also found at higher risk for substances use and sexual risk behavior than concordant heterosexuals (Austin, Roberts, Corliss, & Molnar, 2008). Additionally, current research findings suggest that sexual minorities have a higher risk of psychiatric and substance use disorders compared with heterosexuals. These include higher risk of mood disorders, anxiety disorders, and suicidality among homosexual individuals compared to heterosexuals (Gattis et al., 2012). There may also be important sociodemographic and behavior differences between the various groups of heterosexuals which should be better understood to prepare and implement effective HIV prevention strategies and to further understand the impact of discordant heterosexuals on HIV/STD transmission to the heterosexual community (Ellen et al., 2015).

Along with the increased reports of HIV and STD globally (World Health Organization, 2013a, 2013b) and in Israel (Mor et al., 2013), there is a pressing need for a reliable estimation of the size of the populations at risk for HIV/STD acquisition, defined by their sexual orientation. Understanding the link between sexual orientation and risk behaviors is important to assess the extent of populations at risk in order to design interventions to reduce these disparities. These data will facilitate the understanding of unmet needs for this population, appropriate service development, monitoring of discrimination, monitoring of temporal changes in sexual identity, and benchmarking of other data sources. Key users of these data include HIV/AIDS program planners, researchers, prevention interventionists, policymakers, local government, and other public service providers, social marketers, as well as communities of lesbians, gays, and bisexuals.

Most of the current data regarding sexual behavior and HIV/STD burden in Israel have been obtained through periodical studies among specific risk groups (Mor & Dan, 2012; Mor, Shohat, Goor, & Dan, 2012) and through passive surveillance at the central level (Brosh-Nissimov et al., 2012; Mor, Grayeb, Beany, & Grotto, 2012, 2013). However, the number of individuals in the general population in Israel who report same-sex attraction, same-sex practices, define themselves as gay/lesbian, have sex with CSW, and other sexual practices associated with high levels of risk is yet unknown. This study aimed, for the first time in Israel, to provide a more comprehensive view of contemporary dimensions of sexual orientation among adults Jews in Israel, and its association with risky sexual behaviors.

Method

Participants

This cross-sectional study was performed in December 2012 and included a representative sample of Jewish males and females aged 18–44. Study participants were asked to complete an anonymous electronic questionnaire regarding their sexual attraction and practices.

Individuals in this study were chosen from an existing Israeli panel of 77,000 persons aged 15–85. This online omnibus opinion poll panel is maintained to respond to a variety of topics and its members are identified using an interactive selection algorithm. A random sample was selected from the fraction of Jewish panel members aged 18–44, as they are more likely to be exposed to HIV and in accordance with the age group used for publication by the Statistical Abstract of Israel. The sample reflected sex, age, geographic district, and religiosity of the general Jewish population in Israel, based on figures from the Central Bureau of Statistics 2008 (Table 6 in Appendix). Those who have not performed vaginal or anal sex were excluded from the study. Arabs, orthodox Jews, and other minority panelists were also excluded, as they comprised less than 20 % of the Israeli population; thus, their participation required a much larger study sample.

Measures

Participants who were selected for this sample were requested to respond to an anonymous online tailored questionnaire, which included a maximum of 46 questions, depending on their gender and sexual practices. Participants were requested to provide their demographic characteristics, report the gender of their partners, define their own sexual orientation, past and present sexual behavior with their steady and casual partners, HIV testing, and condom or substance use.

Definitions

Sexual orientation was measured in three dimensions. First, same-sex sexual attraction was measured through a 6-point Likert Kinsey scale (Kinsey, Pomeroy, & Martin, 1948), ranging from attraction to opposite sex only to same sex only. Any same-sex attraction was considered when participants reported they were attracted exclusively or mostly to the same sex, equally to both sexes, or mostly to the opposite sex (Hayes et al., 2012). Second, measurement looked at their sexual practices: having ever had lifetime history of sexual contact either with the same or opposite sex or both (Corliss, Austin, Roberts, & Molner, 2009). Sexual contact in this study was defined as oral (oral-genital, yet excluding oral-anal), and vaginal, or anal intercourse. Third, through measurement of self-identified sexual orientation at the time of filling out the questionnaire being heterosexual or gay, lesbian and bisexual men and women (Ward, Dahlhamer, Galinsky, & Joestl, 2014). Based on these responses, heterosexual participants were categorized as concordant if they self-reported heterosexual identity, were attracted and had sex only with the opposite sex. Discordant heterosexuals included participants who self-reported heterosexual identity, yet were attracted or performed same-sex intercourse (Nield, Magnusson, Brooks, Chapman, & Lapane, 2015). Risky behavior was defined as unprotected last vaginal or anal coitus with a partner whose HIV-status was discordant or unknown, or participant who was previously diagnosed with STD/HIV (Corliss et al., 2009; Mor, Parfionov, Davidovitch, & Grotto, 2014). CSW was considered money or any other valuables received in exchange for sex (Baral et al., 2015). Steady sexual partnership was determined subjectively by the participants as boy/girlfriend, spouse, or de-facto spouse, while casual partner included all other types of sexual encounters, even anonymous (Mor et al., 2011; Richter et al., 2014).

Calculations

The estimated number of gays, lesbians, and bisexuals Jews aged 18–44 who were living in Israel and Tel Aviv was projected on the number of the respective population from the 2012 Statistical Abstract, stratified by age groups.

Analysis

In order to assess demographic and behavioral characteristics by sexual orientation, concordant heterosexuals, discordant heterosexuals and gays, lesbians or bisexuals were compared separately for males and females. Independent variables included demography, age for first sexual debut, number of sex partners, type of partnership, sex for pay, HIV testing, and sexual risk. Continuous variables were compared by the ANOVA test, and categorical variables were compared by the chi-test. p values lesser than 5 % were considered statistically significant.

In order to identify demographic, sexual orientation, and behavioral characteristics by risk, participants were later stratified by their sexual risk and were stratified to risky and non-risky sexual behavior. Comparisons of continuous variables between risky and non-risky males and females were performed by two-sided Student’s t test if normal distribution was demonstrated, and by the Mann–Whitney in other cases. Categorical variables were compared by the chi square test. Variables which were statistically significant in the univariate analyses were included in the multiple logistic regression in the backward technique, adjusted for age and after assessing for collinearity and normal distribution, generating odds ratio (OR) and 95 % confidence interval (95 % CI) to identify variables which predict risky behavior. Analyses were conducted by SPSS® 20.0 package for Windows software (SPSS® Inc., Chicago, IL, USA). The study was approved by the E. Wolfson Medical Center Review Board (WOMC-11-0082).

Results

This representative sample of Israeli Jews aged 18–44 included 995 males and 1005 females, and the response rate from the panelists was 73.2 and 79.5 %, respectively. Of all study participants, 1734 (86.7 %) were attracted only to the opposite sex, while 266 (13.3 %) were attracted at least to some extent to the same sex (Table 1). Ever having had same-sex sexual contact was reported by 189 (9.4 %) participants (10.2 % of males and 8.7 % of females, p = .82). While 82 (8.2 %) of all males identified themselves as gays or bisexuals (4.5 and 3.7 %, respectively), only 48 (4.8 %) of females self-identified as bisexuals or lesbians (3.7 and 1.1 %, respectively, p = .001). Discordant heterosexuals who reported same-sex attraction included 3.4 % of the males and 11.0 % of the females, while discordant heterosexuals who were engaged in same-sex practices comprised 4.1 and 6.1 %, respectively.

Table 1 Same-sex attraction and practices, by gender and sexual identity
Full size table

Discordant heterosexual males and gays and bisexual males were more likely to live in Tel Aviv, experienced earlier anal sex debut, reported greater number of lifetime sex partners, were more likely to be single and if in steady relationship—more commonly had a concomitant casual sex partner/s, ever been paid or coerced to have sex, used drug/alcohol during sex and were engaged in riskier sexual behavior than concordant heterosexuals (Table 2). Discordant heterosexual females and lesbians or bisexual women were more likely to experience earlier vaginal sex debut, reported greater number of lifetime sex partners, were more likely to have a concomitant casual sex partner, while in steady relationships, been paid or coerced to have sex, used drug/alcohol during sex and were engaged in riskier sexual behavior than concordant heterosexuals (Table 3).

Table 2 Demographic and behavioral characteristics of male participants, by self-identified sexual orientation
Full size table
Table 3 Demographic and behavioral characteristics of female participants, by self-identified sexual orientation
Full size table

Males whose sexual behavior was classified as risky were more likely to live in Tel Aviv, to have lower education and income, had an earlier same-sex encounter, reported greater number of lifetime sex partners, being single, and if in steady relationship—more likely to have a concomitant casual sex partner/s, paid for sex, used drugs/alcohol during sex, self-identified as gays or bisexual males or were discordant heterosexuals who experienced same-sex encounter than non-risky males (Table 4). In multivariate analysis, lower income, greater number of lifetime sex partners, been single and if in steady relationship having concomitant casual sex partner/s, paying for sex, self-identification as gay or bisexual males and being discordant heterosexual who experienced same-sex encounter were all variables predicting risky sexual behavior.

Table 4 Demographic and behavioral characteristics of male participants, by level of sexual risk
Full size table

Females whose sexual behavior was classified as risky were more likely to have lower income, reported earlier vaginal or same-sex debut and greater number of lifetime sex partners, more likely to be single, and if in steady relationship—more likely to have a concomitant casual sex partner/s, been coerced to have sex, used drugs/alcohol during sex or were discordant heterosexuals who experienced same-sex encounter than non-risky females (Table 5). In multivariate analysis, lower income, greater number of lifetime sex partners, using drugs/alcohol and being discordant heterosexual who experienced same-sex encounter were variables predicting risky sexual behavior.

Table 5 Demographic and behavioral characteristics of female participants, by level of sexual risk
Full size table

Of all study participants, those who lived in Tel Aviv had the highest rate of males and females who have ever experienced same-sex attraction (17.4 and 15.3 %, respectively), the highest rate of same-sex practice ever (17.0 and 11.1 %, respectively), and the highest rate of gay/lesbian or bisexual self-identification (13.6 and 7.4 %, respectively). It is, therefore, estimated that there are 94,176 (95 % CI 93,608–94,778) self-identified Jewish gays or bisexual males and 57,671 (95 % CI 57,251–58,183) lesbians or bisexual females aged 18–44 who are living in Israel, and about 33,839 (95 % CI 33,492–34,168) and 18,493 (95 % CI 18,234–18,752) gays or bisexual males and lesbians or bisexual females, respectively, living in Tel Aviv (Table 7 in Appendix).

Of all 753 men and 785 women who were in a steady relationship at the time of the study, 105 (13.9 %) and 76 (9.6 %), respectively, had ever had a casual sex partner during their relationships. Of all these 181 men and women who were in steady relationships and had concurrent casual partners, 16.2 and 1.4 %, respectively, had same-sex casual partners.

Discussion

This first study was performed among a representative sample of Jewish participants aged 18–44 in Israel and found that 10.2 % of males and 8.7 % of females had at least a single lifetime same-sex sexual encounter, while 4.5 and 3.7 % self-identified as gays or bisexual men, and lesbians or bisexual women, respectively. Gays and bisexual males, as well as discordant males and females heterosexuals who performed same-sex practice, were more likely to engage in risky sexual behavior than concordant heterosexuals or discordant heterosexuals who were attracted to some extent to the same sex.

The rates of lifetime same-sex practices among men in this study were higher than reported in the U.S. (Purcell et al., 2012; Ward et al., 2014), the United Kingdom (Hayes et al., 2012), and France (Lhomond et al., 2014) (6.9, 5.9, and 3.5 %, respectively). Rates of lifetime same-sex practices among women were lower than those reported from the U.S. (Nield et al., 2015) and United Kingdom (Hayes et al., 2012), yet higher than reported from France (Purcell et al., 2012) (11.2, 7.1, and 3.5 %, respectively). The heterogeneity in the results between the different studies might be associated with the distinguishable wording used in the relevant questions presented to the participants across the various studies, the different periods in which these studies were performed, and the different age groups which were captured in each of the publications. It might also be that the different timeframes pertaining to the questions regarding same-sex practices distorted the results, as participants were requested to respond to questions concerning same-sex practices over their lifetime, while same-sex attraction was referred to at the time they completed the questionnaires.

The differentiation between the three dimensions of sexual orientation (same-sex sexual attraction, sexual practices, and self-reported sexual definition) once again underlines the importance of using the precise study instruments in sex research in order to define populations at risk (Hayes et al., 2012). For example, asking research participants if they have ever had same-sex encounters in a sensitive and direct approach may encourage participants to respond more accurately regarding their sexual orientation than asking if they are gay or straight. This may also be applied to in clinical anamnesis in medical settings. Practitioners should have better understanding of the sexuality of their clients and be aware that self-reported sexual definition is not always concordant with sexual practices.

Our findings indicate that men and women who reported same-sex practices were generally engaged in riskier sexual behavior than those who only had opposite-sex encounters, which included a greater number of partners, earlier sexual debut, and more frequent substance use, as found in other studies (Hayes et al., 2012; Nield et al., 2015). Detailed medical anamnesis, which includes relevant questions about sexual behaviors and risks, allows providers to gain deeper insight into the health needs of their patients and tailor screening and prevention strategies. Prevention advice should also include additional lifestyle aspects, such as substance use and smoking.

Sexual risk among discordant heterosexuals were associated with same-sex practices, but not with same-sex attraction, in line with other studies (Nield et al., 2015). Discordant individuals, in a similar way to bisexuals, may be confronted with discomfort with their sexuality, fear of social isolation, expectations of rejection, express guilt or shame, concealment, and internalized homophobia (Mor et al., 2015; Rosario, Schrimshaw, & Hunter, 2006). It was also demonstrated that discordant heterosexuals have poorer social family support and limited social network (Corliss et al., 2009). They also experience a greater density of intrapersonal, interpersonal, and environmental difficulties that increase their risk behavior and vulnerability to HIV/STD. Finally, it was also speculated that discordant heterosexuals could have higher level of sex drive, while exploring their sexuality in a greater depth despite the stigma associated with same-sex practices, which increases their sexual repertoire and engagement in greater number of sexual partners and practices (Vrangalova & Savin-Williams, 2014). Discordant heterosexuals may therefore not only experience more stress and fewer coping resources than concordant heterosexuals (Ellen et al., 2015), but also less connected with the homosexual culture and circumvent venues that are perceived to be gay/lesbian, thereby avoiding the support that these establishments can provide (Reback & Larkins, 2010). In addition, discordant heterosexuals were found to perform higher risk practices, and may form a “bridge” population for HIV/STD to their steady partners and who are not members of a high-risk group.

Although males and females who only had opposite-sex partners were more likely to be in steady partnership than those who reported having same-sex encounters, almost 14 % of male and 10 % of females in steady relationships reported concomitant casual partners. In addition, nearly 20 % of the males had ever paid for sex. Individuals who had sexual encounters within an active (and probably riskier) sexual network such as CSW, are at a higher risk for HIV/STD acquisition and may also “bridge” the infections to their steady partners, with whom they probably do not use condoms regularly. Prevention efforts should incorporate education regarding safe-sex, while in steady relationships, establishing HIV/STD tests prior to engagement in unprotected sex with steady partners and suggest strategies to avoid “importing” HIV/STD into the dyadic relationships (Mor et al., 2011).

Tel Aviv, being the biggest metropolitan center in Israel, showed the highest rate of participants who reported same-sex practices, as also found in other studies exploring same-sex behavior in major cities (Brown, 2012). Big cities are usually characterized with a greater social liberalism and greater degree of independence and freedom from social pressure and family supervision (Michaels & Lhomond, 2006) and therefore it is recommended that safe-sex interventions are portrayed in big metropolitan areas.

As expected, gays and bisexual men reported the riskiest sexual risk behaviors among all study participants. In addition, their substance use rates were higher than those who only had encounters with the opposite sex, as found in other publications (Lhomond et al., 2014; Mravcak, 2006). Substance use may be a response to lower perceived or actual well-being or minority stress (Wong, Kipke, Weiss, & McDavitt, 2010). Additionally, gays and lesbians may also have been exposed to drugs and smoking in the gay/lesbian community and are probably under higher social pressure to use drugs or smoke.

This is the first study performed in Israel to describe sexual behavior on a national scale in a representative sample, yet it is subject to several limitations. First, we did not have enough statistical power to differentiate bisexual men and women from gays and lesbians, and therefore we collapsed the two sexual dimensions into one non-heterosexual group. Previous studies have outlined the physical, mental, and sexual differences between bisexual men and women with those of gays and lesbians. Bisexuals tend to display more psychological distress and report riskier sexual behavior than exclusively heterosexuals or homosexuals (Friedman et al., 2014). Additionally, bisexuals’ community support is limited and their social resilience may be inadequate. Moreover, their unique health needs are often not appropriately responded to the health system, as bisexuals express discomfort in disclosing their sexual identity in the health settings (Dodge et al., 2012). Similar findings were also reported among Israeli bisexual women (Mor et al., 2015). They were scored lower on socioeconomic status, engaged in higher risk behaviors, and had inferior health-related outcomes in comparison to lesbians. Accordingly, bisexual women reported more frequent visits to the emergency rooms, were less likely to visit their primary care clinicians, and reported poorer subjective health status. They also were less comfortable sharing their sexual orientation with the healthcare provider. As a result of the limited number of bisexuals in our study (especially females), we were unable to replicate previous findings and examine the possible differences health outcomes among bisexual men and women. Second, the cross-sectional nature of this study limits our ability to define causality. Third, although explicitly defined in the questionnaire, participants may have erroneously classified only penile-vaginal/anal penetration as the difference between being heterosexual versus gay/lesbian, rather than, for example, oral sex, resulting in a possible reporting bias and underestimation of same-sex behaviors. Fourth, participants were offered the opportunity to respond only to 3 options of sexual identities (i.e., heterosexual, gay/lesbian, and bisexual), limiting the heterogeneity of additional alternatives. Fifth, sexual contact included penetration practices only, while other practices, such as body contact/kissing or digital sex were excluded. Yet, this limitation is conservative, as we used only those practices which may expose the participants to HIV/STD. Sixth, this study is subject to recall bias regarding same-sex practices. However, we believe that even incidental same-sex practices, especially encounters which involved penetrative practices are unforgettable experiences throughout life. This may be even more noticeable among discordant heterosexuals. Additionally, in order to increase the reliability of self-reported sexual behavior, the question regarding condom use was limited to the last sexual encounter (Richter et al., 2014). Last, the age group of 18–44 was chosen to reflect the Statistical Abstract’s publications, while excluding members from other age groups and religions.

In summary, this first representative study demonstrated that 9.5 % of participants have ever had same-sex practices, while 6.5 % identified themselves as gay/lesbian or bisexual. Those who had same-sex practices reported greater risk for acquiring HIV/STD than those who only had opposite-sex partners. Risky sexual behavior among discordant heterosexual was associated with same-sex practices rather than same-sex attraction. Discordant heterosexual and gays or bisexual males should be aware of a potential for increasing the risk of acquiring or transmitting HIV/STD. It is also recommended that medical practitioners and health educators should not necessarily assume a low-risk profile for those who self-identify as heterosexuals. Practitioners should be able to provide appropriate prevention strategies and recommend screening tests to the populations they care for, such as frequent HIV/STD tests and psychological support if needed. Sexual health professionals should improve their communication skills with heterosexual clients to identify discordant heterosexuals. Furthermore, effective sex education should illuminate the syndemic effects of drugs/alcohol, stigma against minorities, and social norms have on sexual health risks.