Introduction

Obsessive–compulsive disorder (OCD) is characterized by persistent, recurring, and unwanted obsessions, as well as repetitive compulsions. Obsessions are intrusive thoughts, images, or impulses that cause severe distress in the individual; compulsions are repetitive mental or physical actions designed to alleviate distress caused by the obsession (American Psychiatric Association, 2013). It is estimated that the lifetime prevalence rate for OCD is approximately 2.0 %, and the average age of onset is 19.5 years (Kessler, Berglund, & Demler, 2005; Ruscio, Stein, Chiu, & Kessler, 2010).

Obsessions and compulsions are expressed in a wide variety of ways depending on the individual. Several studies have attempted to classify OCD symptom dimensions based on specific symptoms. Across the literature, it is commonly agreed that there are approximately four OCD symptom dimensions: (1) contamination, (2) doubt/checking, (3) symmetry, and (4) unacceptable thoughts (Abramowitz, Franklin, Schwartz, & Furr, 2003; Williams, Mugno, Franklin, & Faber, 2013). The unacceptable thoughts subtype has also been described as obsessions without overt compulsions (McKay et al., 2004; Williams et al., 2011b). There is some evidence that suggests that unacceptable thoughts are associated with more fear and shame and less social acceptance than contamination or doubt/checking obsessions (Simonds & Thorpe, 2003). Three main areas of unacceptable thoughts are: sexual, scrupulous/religious, and violence (Gordon, 2002). Sexual obsessions are unwanted sexual thoughts, often involving children, family members, animals, violence, or same-sex sexual activities (e.g., Williams, Crozier, & Powers, 2011a). While much research has been conducted on certain symptom dimensions, relatively less research has been conducted examining sexual obsessions (Williams et al., 2013). This may be because the corresponding compulsions are often covert and the taboo nature of such thoughts (Williams et al., 2011b, 2013).

Unwanted sexual obsessions are common in both non-clinical samples and individuals with OCD. Research on student samples has found that 84 % of individuals have experienced unwanted sexual thoughts during their lifetime (Byers, Purdon, & Clark, 1998; Wetterneck, Smith, Hart, & Burgess, 2011). Grant et al. (2006) assessed individuals with OCD and found that 24.9 % had a history of sexual obsessions and 13.3 % were experiencing sexual obsessions during the time of the study. When looking specifically at sexual orientation obsessions, an analysis from a large DSM-IV Field Trial (n = 409; Foa et al., 1995) found that 8 % reported current sexual orientation obsessions, and 11.9 % endorsed lifetime symptoms (Williams & Farris, 2011). Using data from the OCD Collaborative Genetics Study (Samuels et al., 2006), Pinto et al. (2008) reported that 9.9 % of their sample (n = 485) endorsed past or present obsessions related to homosexuality.

While sexual obsessions in OCD may appear to be less common than other OCD symptom dimensions, these numbers may be an underestimate of the actual number of people suffering from sexual obsessions because the stigma associated with sexual thoughts may cause individuals to avoid reporting their obsessions (Grant et al., 2006). Given that the most common topics of sexual obsessions include incest, religion, AIDS, pedophilia, unfaithfulness, and sexual orientation (Gordon, 2002; Williams, 2008), the perceived stigma may be great for many with this type of OCD. One study found that there was more stigmatization toward those with sexual obsessions than contamination obsessions (Cathey & Wetterneck, 2013).

When conceptualizing sexual obsessions within OCD, it is important to recognize that people with sexual obsessions find their thoughts ego-dystonic and do not wish to act them out (American Psychiatric Association, 2013). These are different from fantasies because they are unpleasant and provoke distress, rather than being pleasurable (Gordon, 2002; Williams et al., 2011b). As a result, the thoughts may result in anxiety related to the threat the individual perceives from certain emotions, such as sexual desire, disgust, anger, and guilt (Wetterneck et al., 2011). In non-clinical samples, this distress can be predicted by the frequency of the sexual obsessions, with more distress experienced by those who are avoidant and feel threatened by feelings of sexual desire.

Less is known about variables associated with distress in OCD populations. Sexual orientation obsessions typically take the form of worry about becoming or being a lesbian/gay/bisexual/transgendered (LGBT) individual or worries that others may perceive one as LGBT (Williams, 2008). People with sexual orientation themed OCD (SO-OCD) differ from those who are actually conflicted about their sexuality because a heterosexual person with SO-OCD does not typically feel a strong sexual attraction or sexual arousal to members of the same sex. However, it is also possible for LGBT people to have unwanted anxiety about heterosexual thoughts (e.g., Goldberg, 1984), or SO-OCD about other-sex attraction.

Distressing obsessions are thought to be a result of flawed cognitive appraisals about an unwanted sexual thought, image, or impulse (e.g., Rachman, 1998), resulting in fears about the future of the individual’s sexual orientation. For example, a heterosexual individual may first experience these symptoms when noticing a member of the same sex and finding that person attractive. The individual then questions the meaning of the thought. These obsessions then lead to compulsions, which may present as the person checking for sexual arousal when around others or mental reminders about being heterosexual. Symptoms may also include avoidance, such as not watching television shows in which there is an LGBT character or keeping physical distance from others of the same sex (i.e., Williams et al., 2011a). The individual may watch pornography with same-sex themes to determine if it produces sexual arousal and compare that reaction to heterosexual pornography (Williams, Slimowicz, Tellawi, & Wetterneck, 2014). Another compulsion is to increase sexual intercourse with an other-sex partner to demonstrate that the person’s sexual preference has not changed.

It is important to understand that SO-OCD is not caused by homophobia/heterosexism, as the individual may or may not have negative feelings toward LGBT individuals (Williams, 2008). This is also different from internalized homophobia/heterosexism (IH), which occurs when an LGBT person has negative feelings about him/herself due to sexual orientation (Szymanski, Kashubeck-West, & Meyer, 2008), although LGBT people with OCD can also suffer from IH. Because it can be difficult for clinicians who do not have experience with OCD to distinguish between “sexual identity confusion” and SO-OCD, SO-OCD may be misdiagnosed (Glazier, Calixte, Rothschild, & Pinto, 2013).

Williams and Farris (2011) found that more males reported SO-OCD than females in a clinical sample. OCD severity was reported to be moderate, and these individuals appeared to have similar insight to people with other types of OCD. When assessed with the Yale-Brown Obsessive Compulsive Severity Scale (Goodman et al., 1989), of the 10 items, three items on the obsession subscale (time, interference, and distress) were significantly greater among those with sexual orientation obsessions than those without. The results indicated that those with SO-OCD spent more time occupied by their thoughts than people without these obsessions. The patients also reported more interference and more distress from their obsessions. Thus, SO-OCD may be more distressing than other forms of OCD, indicating a clinical need for more research on assessing and treating SO-OCD.

Sexual orientation obsessions in OCD are not uncommon, yet they have not been the subject of much research. To date, there are few published articles that focus on SO-OCD: a chapter, a case study, and a single empirical article (Williams, 2008; Williams & Farris, 2011; Williams et al., 2011a). The purpose of the current study was to gain a greater understanding of SO-OCD to facilitate future clinical investigations. To this end, we developed and deployed a survey of items about sexual orientation concerns to explore relationships between symptoms of SO-OCD, distress, and demographic characteristics, to better understand factors uniquely related to distress in this population.

Method

Participants

Participants included 1,176 adults who completed an online survey about sexual anxieties, behaviors, and cognitions. Data were collected from October 2009 to August 2010. Of these, 74.6 % were male and 25.4 % were female. The mean age was 25.7 years (SD = 8.76). Of the 85.3 % who provided information about nationality, the majority was from the United States (51.3 %) and other English-speaking countries (13.3 % UK/Ireland, 6.4 % Other Europe, 5.7 % Canada, 3.3 % Australia/New Zealand, 5.4 % Other, 14.7 % Did Not Report). In terms of sexual orientation, 72.0 % identified as heterosexual, 3.1 % as homosexual, 3.0 % as bisexual, and 21.9 % as unsure. In terms of mental health history, 25.9 % had a previous OCD diagnosis, 71.9 % did not, and 2.2 % did not disclose.

Procedure

Individuals were recruited from informational websites about OCD (13.4 %), an online forum about OCD (81.5 %), and email advertisements to LGBT groups (5.0 %). Participants were told this was a study about “sexuality concerns in OCD.” Informed consent was obtained by use of an information form provided online that the participant was required to read and acknowledge before proceeding to the next screen. The study was approved by University of Pennsylvania’s IRB, where the data were collected by the first author at the Center for the Treatment and Study of Anxiety in the Department of Psychiatry.

Measures

The survey consisted of demographic questions, OCD status, distress, and a sexuality questionnaire that was developed for the purposes of this study by the authors. Participants were asked to provide their age, gender, nationality, and if they had ever been diagnosed with OCD and, if so, when and by whom. Distress surrounding sexual orientation concerns (“On the scale below, how much distress have your thoughts about sexual orientation caused you?”) was rated on a scale of 0–6, with 0 representing “none” and 6 indicating “suicidal” levels of distress. Participants were asked to choose their sexual orientation on a nominal scale, with choices being homosexual, heterosexual, bisexual, and not sure.

The sexuality questionnaire consisted of 70 items generated by psychologists with extensive experience treating OCD. Items were based on the clinical observation that people with SO-OCD tend to have worries in three main areas: fears of becoming or being LGBT, worries that others may think one is LGBT, and experiencing unwanted same-sex thoughts (Williams, 2008). Items were developed to assess these worries and also to separate people with SO-OCD from those with other forms of OCD and people without OCD. Each item was rated on a scale of 1–5, with greater numbers corresponding to greater agreement by the participant. Lower ratings corresponded to responses indicative of greater distress.

Results

A principal components factor analysis (PCA) was employed with a Promax (oblique) rotation, given that the components were expected to correlate with one another. Although 23 possible components had an eigenvalue over 1, a visual inspection of a scree plot suggested 4–6 components, and a Horn’s (1965) parallel analyses suggested 6 components. Items were included on individual components if they demonstrated a loading of .50 or higher [considered strong loadings as suggested by Costello and Osborne(2005)]. In order to ensure the unique contribution to only one component, items were kept if they were at least .20 higher than their next highest loading.

The PCA with six components explained 48.4 % variance of the overall model with individual components accounting for 22.2, 13.1, 4.9, 3.6, 2.4, and 2.2 %, respectively. Table 1 shows the item loadings for each component. The main themes of the components centered on worries about one’s sexual orientation changing (Factor 1: Worry about Sexuality; 12 items), experience and desires for same sex partners (Factor 2: Desire Same Sex; 11 items), experience and desires for other sex partners (Factor 3: Desire Other Sex; 5 items), beliefs that a same-sex preference is wrong or immoral (Factor 4: Sexual Immorality; 4 items), beliefs that one needs to avoid other’s judgments of sexual orientation (Factor 5: Avoidance Judgment), and sexual orientation shame or dissatisfaction (Factor 6: Orientation Shame; 3 items). Cronbach’s alphas ranged from excellent to acceptable (.93, .91, .81, .78, .79, and .73, respectively).

Table 1 PCA of group reporting OCD diagnosis with sexual orientation symptoms
Full size table

In order to explore relationships between the components and levels of distress in those with OCD, we performed a set of correlations with those who reported having a prior diagnosis of OCD, were heterosexual, and who endorsed any distress from sexual orientation thoughts, that is, people with SO-OCD (n = 237). There were not enough individuals in the homosexual, bisexual, or unsure categories to run similar comparisons. These results are shown in Table 2. Level of distress was most strongly correlated with Worry about Sexuality, and moderately correlated with Avoidance of Other’s Judgment and Sexual Immorality. Although this group endorsed being heterosexual and there was a significant relationship between Same Sex Desires and Opposite Sex Desires, the magnitude was small, perhaps reflecting ambivalence of uncertainty surrounding sexual orientation obsessions. Additionally, the majority of those with SO-OCD (91 %) reported high levels of distress related to same sex thoughts, with 2 % reporting little distress, 2 % reporting some distress, 5 % reporting moderate distress, 19 % reporting much distress, 51 % reporting extreme distress, and 21 % reporting a “suicidal” level of distress.

Table 2 Component correlation matrix of SO-OCD factors and level of distress in participants with sexual orientation symptoms
Full size table

Individual t tests were used to explore potential sex differences on the factors and level of distress in the SO-OCD group (n = 237; 172 males and 65 females). Only the first component, Worry About Sexuality (becoming LGBT), demonstrated significant differences between males and females, with males (M = 27.31, SD = 10.98) scoring significantly higher than females (M = 22.74, SD = 9.84) [t (1, 235) = 2.70, p = .004]; the lower scores for females indicate that they exhibit more worries concerning changing sexual orientation than males. See Table 3 for means by factors and sex.

Table 3 Means for SO-OCD subscales in participants with sexual orientation symptoms
Full size table

Discussion

The purpose of this study was to better understand sexual orientation worries in OCD and the distress associated with these thoughts. The SO-OCD survey questions resulted in six subscales that represent factors related to this obsession, including: worry about a change in sexual orientation, desire for same sex partners, desire for other sex partners, beliefs that a same-sex preference is wrong or immoral, avoidance of other’s judgment, and sexual orientation shame. These components demonstrated good internal consistency.

The findings indicated that with increasing worry about sexual orientation changing, increased belief that same sex thoughts are immoral, and increased attempts to avoid other’s judgment about orientation, the level of distress also increased. These findings were consistent with expectations regarding SO-OCD and clinical observation (Williams, 2008).

Perhaps the most pronounced finding was that, among those with SO-OCD, 91 % had levels of distress ranging from “much” to “suicidal” as a result of their sexual orientation obsessions. This is an alarming percentage that underscores the importance of recognizing and assessing these symptoms. People with OCD are highly upset by such obsessions, many to the point of considering ending their lives. However, it should be noted that this sample may not be representative of all people with SO-OCD.

The findings of this study raise important questions about SO-OCD and risk for suicide. Sexual minorities may be more likely to make suicide attempts than their heterosexual counterparts, due to social stigma, rejection, discrimination, and marginalization (Haas et al., 2011; King et al., 2008). Poteat, Aragon, Espelage, and Koenig (2009) examined a large adolescent sample (n = 4,439) for significant differences on psychosocial concerns based on sexual orientation including heterosexual, LGBT, and a category termed “questioning/less certain.” Those in the questioning/less certain category were much more likely to have depressed and suicidal thoughts than either heterosexual or LGBT youth, perhaps because they did not feel LGBT or heterosexual labels adequately described them. People with SO-OCD will frequently express uncertainty about their sexual orientation and may wonder if they belong in this “questioning” category. Thus, they may experience increased marginalization, feeling unable to relate to LGBT or heterosexual peers. People with SO-OCD worry about negative repercussions from being LGBT and anxieties may be heightened due to the nature of their disorder. They may feel shame for being unable to control their thoughts, much in the way LGBT individuals with IH experience shame about their unwanted same-sex feelings.

There is little research on suicide risk in OCD, perhaps because people with OCD have generally been assumed to be less likely to harm themselves than people with other mental disorders. Comorbid major depressive disorder is common in people with OCD (40.7 %) (Ruscio et al., 2010), and this appears to be the cause of most of the suicide risk in this population (Maina, Salvi, Tiezzi, Albert, & Bogetto, 2007). Torres et al. (2011) found that, in addition to comorbidity, unacceptable thoughts (sexual/religious) were also associated with suicidal thoughts and plans in people with OCD. Given that people with sexual obsessions are more depressed (Dell’Osso et al., 2012; Grant et al., 2006), and those with SO-OCD experience greater distress (Williams & Farris, 2011), it is reasonable to conclude that increased depression may be part of the clinical picture.

Finally, sex was also a significant predictor of distress related to sexual orientation obsessions, with females experiencing more distress over same-sex thoughts than males. Consistent with earlier work, males were more prevalent in our SO-OCD sample (Williams & Farris, 2011), but it could be that although more males are more likely to be afflicted by SO-OCD, females are more upset by their symptoms.

This study had a number of limitations. The clinical information was based on self-report of a prior diagnosis, and the sample was comprised of online participants and therefore may only be representative of those interested in the particular topic. Even though this population is of interest to this study, assessing additional populations, including well-characterized clinical samples, would be beneficial. Validating these findings with a larger sample of people with and without SO-OCD and of different sexual orientations is an important next step.

Further research should be done to determine if the presence of certain symptoms, including depression, is a factor in the level of OCD-related distress, as well as factors such as religiosity and culture. Studies should also be conducted on SO-OCD in adolescents, as young people, who generally have less sexual experience, may have particular difficulty managing confusing and stigmatizing obsessions of this nature. Future directions include the development of a new measure based on these findings that can be used to reliably differentiate those with SO-OCD from those with sexual orientation confusion from others who may be secure in their LGBT orientation. Such an instrument could provide clinicians with a reliable diagnostic tool to better understand clients with SO-OCD. Moreover, it may provide a way of quantifying the presence of distress resulting from SO-OCD in research on this OCD symptom dimension and give insight into variables that affect the development, maintenance, and treatment of these symptoms. More work is urgently needed, as the level of distress experienced by those with SO-OCD demands a response from the research and clinical community.