Sexual hallucinations are intrusive phenomena that can be experienced in various sensory modalities (1, 2). For example, people may experience orgastic sensations, touch of erogenous zones, or images with an explicit sexual content in the absence of corresponding stimuli from the outside world. Although valued positively by a small minority of people (notably those experiencing auras in the context of epilepsy) they are mostly unwanted and burdening (2). They are not as rare as previously thought, with estimated lifetime prevalence rates of 0.4% in the general population and up to 44% in chronically hospitalized patients with schizophrenia spectrum disorders (2, 3). In addition to their association with the latter group of disorders, they are chiefly reported on in the context of sleep paralysis, epilepsy, narcolepsy, sedative use and pelvic pathology (e.g. Tarlov cysts, clitoral engorgement (2), whether or not in combination with sexual delusions. Studies have established an association between sexual hallucinations, sexual delusions and childhood trauma with odds ratios of up to 8.7 in patients with schizophrenia spectrum disorders (1, 3). As far as we know, there are no published reports on these phenomena in people diagnosed with borderline personality disorder (BPD), a condition that is also strongly associated with childhood trauma. One reason for this omission may be that hallucinations in BPD have long been considered infrequent, mild and transient in nature (4–6). However, several studies report a lifetime prevalence of hallucinations in general in up to 60% of this group (4, 5) and moreover indicate that hallucinations in this group are often indistinguishable from those in schizophrenia spectrum disorders. For example, verbal auditory hallucinations in both groups share the same phenomenological characteristics (7, 8). In both groups, these hallucinations have also tended to be experienced for years, albeit with longer-lasting symptom-free intervals in BPD, and yet with ensuing levels of distress that are often markedly higher than in people with schizophrenia spectrum disorder (5, 7, 9, 10). Despite these empirical findings, major classifications such as the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and the International Classification of Diseases (ICD-11) do not mention hallucinations as diagnostic criteria for BPD (11, 12). This is probably a second reason why hallucinations in general, but sexual hallucinations too, have long been overlooked in this group. This could be another reason why not only people experiencing these phenomena may be reluctant to discuss them, but also the health professionals to whom they turn for help.

In line with previous findings for people with schizophrenia spectrum disorders, hallucinations in people with BPD are often associated with childhood trauma (5). Prevalence rates reported for childhood trauma in BPD are up to 93%, with rates for sexual abuse varying from 40% to 71% (13–15). Given these figures, we also expect high rates of sexual hallucinations and delusions in the group diagnosed with BPD. Therefore, this study aims to answer the following research questions:

We included 81 people with a DSM diagnosis of BPD, 88% of whom identified as female, 11% as male, and 1% as ‘other’. The majority of the participants had finished secondary school or a higher type of education, and half of them were currently employed. Depressive disorder and PTSD were the most commonly diagnosed comorbid disorders. Of all participants 84% used psychotropic medication, the largest group of which (used by almost 50%) was antidepressants. This, along with further demographic information, is reflected in Table 1.

Within this group, the lifetime prevalence rate for sexual hallucinations was 59.3%; for the present state, this was 35.8% (Table 2). For sexual delusions, the lifetime and present-state prevalence rates were respectively 49.4% and 30.9%. Most prevalent were delusions of reference. All participants reported on having experienced some type of childhood trauma. Table 3 provides an overview of total and mean scores on the CTQ per type of childhood trauma, as well as scores for the groups that had reported on hallucinations in general, sexual hallucinations, sexual delusions, and having had no (prior experience with) hallucinations.

Common examples of auditory sexual hallucinations were the voice of an abuser, sexually themed commands or comments, and hearing footsteps. Visual sexual hallucinations typically involved seeing the face or shape of an abuser, seeing people having sexual intercourse or seeing a vibrator. As for tactile sexual hallucinations, participants often reported feeling touch to erogenous zones or having the sensation of being sexually abused. Somatic sexual hallucinations included unwanted sexual arousal, random orgasmic sensations, sexual arousal in unexpected body parts (e.g. tips of fingers, or the back), perceived abnormal changes to the size or shape of genital organs and the sensation of being penetrated. Olfactory sexual hallucinations often involved the smell of genital odors or condoms, while gustatory ones typically involved the taste of sperm.

Sexual delusions frequently involved firmly believing in being watched in the shower or while undressing, third parties having sexual intentions, being followed for sexual reasons, feelings of jealousy or distrust about the faithfulness of one’s partner, disproportionate changes to the size or shape of genitalia and transgenderism.

As for all types of hallucination in the six main sensory modalities (whether of a sexual nature or not), the lifetime prevalence rate in the study group was 84%, and for the present state 57% (Table 4). Somatic, visual, tactile and auditory sexual hallucinations were more common than olfactory and gustatory ones. Most frequent were auditory hallucinations, experienced at a mean rate of once a week to daily. The highest reported levels of distress were associated with verbal auditory and tactile hallucinations with a median score of 3 indicating ‘considerable distress’ (with ensuing feelings of anxiety, nervousness, and depression), followed by visual hallucinations (some to considerable distress). Among these phenomena, verbal auditory hallucinations had the highest impact on people’s daily functioning, with an average score between 1 ‘minimal disruption to daily life’ (disturbances to some specific activities) and 2 ‘moderate disruptions to life’ (disturbances to multiple activities). People rated verbal auditory, visual and tactile hallucinations more often as negative, whereas they rated nonverbal auditory, gustatory and olfactory hallucinations more often as neutral or positive.

To explore the effects for the five subscales of childhood trauma on sexual hallucinations and delusions we conducted a logistic regression analysis. The results of the model with the childhood trauma subscales were statistically significant for sexual hallucinations and delusions (X²(5) = 16.850, p = 0.005 and X²(5) = 25.114, p = 0.000) and explained respectively 31.8% and 43.1% (Nagelkerke R²) of the variance (see Table 5 for the odds ratios). Participants who had high scores on the subscales of childhood emotional abuse and sexual abuse were more likely to experience sexual hallucinations and delusions. In the case of sexual delusions, this result also held for the total CTQ score. Interestingly, we found a significant result in the opposite direction for emotional neglect and sexual hallucinations. Upon further examination, there seemed to be a spurious effect for emotional abuse, in the sense that emotional neglect only showed an effect when emotional abuse was also included in the analysis. In analyses that did not consider emotional neglect, the effect of emotional abuse still stands. This phenomenon may be attributable to the correlation between emotional neglect and emotional abuse (r = 0.640). Upon visual examination of the mean scores of emotional abuse and emotional neglect, emotional abuse had a higher score in the group of participants who had sexual hallucinations than in the group who had not, with emotional neglect showing similar scores for both groups. Likewise, upon examination of correlations, we found a positive correlation for emotional abuse and sexual hallucinations (r = 0.336), whereas this was not the case for emotional neglect and sexual hallucinations (r = 0.022).

Our study of 81 outpatients diagnosed with BPD yields prevalence rates for hallucinations in general that are comparable to those reported on in previous studies (4, 5, 8, 18, 19), with the exception of visual, olfactory, and tactile hallucinations, which were slightly more prevalent in our sample than in other studies (20–22). The reported levels of distress due to hallucinations were comparable to the relatively high rates reported earlier, notably for auditory hallucinations in BPD (7, 8, 18, 24–27). More importantly though, more than a third of the participants in our study experienced sexual hallucinations during the last month, and almost 60% at some point throughout their lives, with somatic sexual hallucinations being the most prevalent type. Sexual delusions were reported on by almost a third of the participants in the last month, compared to roughly half the group during their lifetime. These rates are substantially higher than those reported on in the context of psychosis, with the lifetime prevalence rate for sexual hallucinations in our group being 1.4 times higher than in chronically hospitalized people with schizophrenia spectrum disorder (3), and 150 times higher than the rates known for the general population. For sexual delusions, such a comparison was harder to make due to a lack of studies in this field. However, compared to the rate of 15.2% for ‘attenuated or subthreshold psychotic symptoms with sexual content’ in an outpatient group with an ultra-high risk of psychosis (23), sexual delusions in our BPD group were more than twice as prevalent.

Although all participants reported childhood trauma, particularly childhood emotional abuse and sexual abuse increased the chance of experiencing sexual hallucinations and delusions. These results are in line with the literature on BPD (13–15) and with similar studies on sexual hallucinations and delusions in schizophrenia spectrum disorders (1, 3, 28). The smaller odds ratios that we report here might be explained by the high prevalence of childhood trauma in this study’s BPD population.

In our cross-sectional study of 81 outpatients diagnosed with borderline personality disorder, 60% reported having experienced sexual hallucinations, and 36% during the past month. For sexual delusions these rates were 49% and 31%, respectively. These rates are substantially higher than in other patient groups, and those for sexual hallucinations are 150 times higher than in the general population. The burden caused by hallucinations in general was substantial for most of the participants. The prevalence rate for childhood trauma was a full 100% in our sample. Our findings add to the existing body of published studies which challenge the traditional notion that hallucinations are rare and self-limiting in BPD. They thereby urge us to raise awareness of these phenomena, improve recognition, and develop evidence-based treatment methods, including those for trauma.