**Sexual Dysfunctions and Their Association with the Dual Control Model of Sexual Response in Men and Women with High-Functioning Autism**

#### **Daniel Turner 1,2,\*, Peer Briken <sup>2</sup> and Daniel Schöttle 2,3**


Received: 20 February 2019; Accepted: 22 March 2019; Published: 28 March 2019

**Abstract:** Adults with an Autism Spectrum Disorder (ASD) are characterized by impairments in social interaction and communication, repetitive and stereotyped interests and behaviours as well as hyper- and/or hyposensitivities. These disorder specific symptoms could be associated with the development of sexual disorders. The Dual Control Model of Sexual Response presents one approach that is frequently used to explain the emergence of sexual dysfunctions. The aim of the present study was to assess the extent of symptoms of sexual dysfunctions in men and women with ASD and to evaluate their association with the individual propensity of sexual excitation and inhibition as defined by the Dual Control Model. Both men and women with ASD were more likely to report about sexual dysfunctions than individuals from the control group. In men with ASD, sexual inhibition was significantly correlated with the emergence of sexual dysfunctions, while there was no association between sexual functioning and sexual excitation. In women, the opposite pattern was found. Especially the peculiarities in sensitive perception could be responsible for the observed problems with sexual functioning in individuals with ASD. The present findings highlight the great need for specialized treatment programs addressing the frequently observed sexuality-related problems in individuals with ASD. However, up to now such treatment programs are lacking.

**Keywords:** sexual dysfunction; autism; erectile dysfunction; sexual satisfaction; Asperger syndrome; sexual desire; lubrication; sexual intercourse; sexual excitation; sexual inhibition

#### **1. Introduction**

Autism Spectrum Disorder (ASD) is characterized by impairments in social interaction and communication, as well as repetitive and stereotyped interests and behaviours [1]. It is estimated that up to 1.7% of the population are affected by ASD [2,3]. About 50% of individuals with ASD have average intellectual functioning and in the meantime more and more adults are being diagnosed in later life [4]. Just like in other neurodevelopmental disorders, there is a male preponderance in ASD and the male to female ratio is estimated to be around 3–4:1 [5,6]. However, these reported gender differences are currently subject of a controversial discussion and it is suggested that this effect might be largely attributable to the possible gender-biased artefact of a male-symptomatic based diagnostic system with later diagnosed females requiring heavier symptom loads for diagnosis [7].

Nevertheless, all individuals with ASD have in common that they have (in varying degrees) difficulties in interpreting non-verbal cues, such as decoding and interpreting facial expressions and have limited capabilities in theory of mind skills [1]. When throughout development social

interactions become more complex and romantic and sexual relationships become increasingly important, the learned social skills often cannot keep up with the social demands needed for the initiation and maintenance of romantic peer-relationships [8]. Thus, many stereotypes around individuals with ASD concerning sexuality related issues have arisen, such as, ASD individuals are seen as being only sparsely interested in sexual and romantic relationships or as being mainly asexual [9,10]. Contrary to these stereotypes, however, in recent years a growing body of research has accumulated showing that most individuals with ASD report a general interest in solitary and dyadic sexual behaviours and show the full range of sexual behaviours, just like their clinically non-affected counterparts [11–14]. Nevertheless, the deficits in intuitively understand social and nonverbal communication cues, difficulties in perspective-taking, inflexibility, affective dysregulation, repetitive and stereotyped interests and peculiarities in sensitive perception leading to either over- or underreactions to sensory stimuli can hamper the development of romantic and sexual relationships, can be associated with impaired sexual functioning and sometimes also with the development of sexual disorders [15–17].

In a first study of our working group, focusing on paraphilias and hypersexuality in high-functioning men and women with ASD, it was found that high-functioning ASD men reported more frequently about masochistic, sexually sadistic, voyeuristic, frotteuristic and paedophilic fantasies and more frequently about frotteuristic behaviours compared to men from a control group. Furthermore, more men with ASD reported about hypersexual fantasies and behaviours than their non-affected peers. High-functioning ASD women reported more frequently about masochistic behaviours than healthy women, while no other differences occurred [18]. Men with ASD usually show a more pronounced ASD symptomatology regarding for example, repetitive behaviours or hypo- and hypersensitivities, which could be one possible explanation for the higher prevalence of hypersexual and paraphilic fantasies and behaviours in ASD men compared to ASD women [18,19]. Thereby, the more frequently observed repetitive behaviours and obsessive interests could translate into sexualized interests and behaviours, which result in a faster habituation leading the individual to seek novel sexual activities, for example, paraphilic sexual activities.

Besides paraphilic and hypersexual behaviours, the disorder-inherent deficits and symptoms could also be accompanied by sexual dysfunctions in ASD men and women. Sexual dysfunctions are disorders characterized by a clinically significant disturbance in a person's ability to respond sexually or to experience sexual pleasure [1]. Sexual dysfunctions are usually classified in accordance with the four phases of the sexual reaction cycle: disorders of sexual appetence (e.g., female sexual interest/arousal disorder), disorders of sexual desire (e.g., erectile disorder, male hypoactive sexual desire disorder), orgasm disorders (e.g., premature (early) ejaculation and delayed ejaculation, female orgasmic disorder) and sexual pain disorders. (e.g., genito-pelvic pain/penetration disorder). In the general population it is estimated that about 40% to 50% of all women report at least one sexual dysfunction throughout their lifetime, while the life-time prevalence in men is estimated to be about 20% to 40% [20–22].

Bancroft and Janssen have developed a theoretical model, which could help to explain some aspects of the emergence of sexual dysfunctions in both men and women: the Dual Control Model of Sexual Response [23]. The Dual Control Model postulates that whether or not a sexual response occurs in a particular situation depends on the interaction between an excitatory and an inhibitory neuroanatomical and neuroendocrinological network [23,24]. Individuals high in sexual inhibition and low in sexual excitation are more likely to develop sexual dysfunctions [23,25,26].

Based on these findings we aimed at assessing symptoms of sexual dysfunctions in men and women with ASD using standardized assessment scales and at evaluating the association between the individual propensity of sexual excitation and inhibition and sexual dysfunctions in both ASD individuals and healthy controls. Due to the above-stated ASD specific symptoms we hypothesized that (1.) both men and women with ASD would show more signs of sexual dysfunctions than healthy controls and (2.) that in individuals with ASD as well as in healthy controls higher sexual excitation and lower sexual inhibition scores would be related with less signs of sexual dysfunctions.

#### **2. Materials and Methods**

#### *2.1. Participants*

The present study included *n* = 96 adults with high-functioning Autism or Asperger syndrome who were compared to *n* = 96 healthy controls. In order to control for the influence of age and education on the sexual outcome measures the participants were matched concerning these variables (Table 1). All patients with ASD self-reported that they had been diagnosed by an experienced psychiatrist or psychologist. However, due to data protection regularities we did not gather any more information from the diagnosing clinicians about the diagnostic procedures. In Germany, mental disorders are usually diagnosed based on the diagnostic criteria of the International Classification of Diseases, 10th version (ICD-10) of the World Health Organization (WHO) and thus it could be assumed that all diagnoses of our study participants were made according to the ICD-10. Mean age at which patients received their ASD diagnosis was 35.7 years (SD = 9.1 years; range = 17 to 55 years). To assess the extent of autism symptoms all participants rated the German version of the Autism Spectrum Quotient-Short Form (AQ-SF) [27]. ASD patients had significantly higher scores on the AQ-SF than the healthy controls (Table 1). While all of the ASD patients scored above the proposed cut-off value of 17 points in the AQ-SF, none of the healthy controls did so. Both, the ASD individuals as well as our control participants had on average 12 years of school education suggesting that all of them had at least average intellectual functioning.

**Table 1.** Sociodemographic and clinical characteristics of study participants.


ASD = Autism Spectrum Disorder; HCs = Healthy Controls. Regular intake is defined as at least three times a week. \*\* *p* ≤ 0.01.

Of the ASD patients 78.2% (*n* = 75) indicated being exclusively or predominantly heterosexual, 10.4% (*n* = 10) being exclusively or predominantly homosexual, 8.3% (*n* = 8) being equally hetero- and homosexual and 3.1% (*n* = 3) indicated having no sexual orientation. In contrast, all healthy controls (HCs) were exclusively or predominantly heterosexual. Sexual orientation was assessed using the Kinsey scale [28]. More HCs (*n* = 78, 81.3%) were currently in a relationship than individuals with ASD (*n* = 27, 28.1%; *p* < 0.001) and more HCs (*n* = 96, 100%) indicated that they had previously been in a relationship lasting more than three month than individuals with ASD (*n* = 60; 62.5%; *p* < 0.001).

The control participants consumed alcoholic beverages on a more regular basis than the ASD individuals. On the other side more individuals with an ASD reported about psychiatric comorbidities, about regular medication intake in general and intake of psychopharmacological drugs in specific.

#### *2.2. Procedure*

All information about study participants were gathered using self-report questionnaires. These could be answered at home. Individuals diagnosed with ASD were recruited via self-help groups throughout Germany and through the Autism outpatient centre at the University Medical Center Hamburg-Eppendorf, Germany. Healthy controls were recruited through advertisements at the University Medical Center Hamburg-Eppendorf, at the University Medical Center Mainz, at local shopping malls and through personal contacts of the principal investigators.

The ethical review board of the Hamburg Medical Council approved the study protocol of the present study (PV4380).

#### *2.3. Measures*

#### 2.3.1. International Index of Erectile Function (IIEF)

The IIEF consists of 15 items and assesses the extent of sexual problems in male respondents. Thereby, sexual functioning is measured on five subscales: erectile functioning, orgasmic functioning, sexual desire, satisfaction with sexual intercourse and overall sexual satisfaction. Lower scores on each subscale represent more problems. The guidelines on the clinical application of the IIEF recommend that patients with a score below 14 out of 30 points on the erectile functioning subscale should be considered for treatment with Sildenafil [29]. Internal consistency for the total score as well as for all subscales of the original version of the IIEF was between α = 0.73 and 0.91 [29]. In the validation study of the German version of the questionnaire internal validity of the total score was α = 0.95, however, in contrast to the English version only a four-factorial solution was found [30]. In a follow-up study with 261 German men the original five factor structure could be replicated by confirmatory factor analysis, although a four-factor model represented an acceptable fit as well [31]. Nevertheless, in the present study we followed the original five-factor model of the questionnaire.

#### 2.3.2. Female Sexual Function Index (FSFI)

The FSFI consists of 19 items and assesses the extent of sexual problems in women on six domains: sexual desire, sexual arousal, lubrication, orgasm, sexual satisfaction and sexual pain. Lower scores represent more problems. Internal consistency for the total score as well as for all subscales of the original version of the IIEF was above α = 0.82 and test-retest reliabilities were between r = 0.79 and 0.86 for the subscales [32]. The German validation study was performed using an online sample of 1243 German women and supported the six factorial design of the original version. Internal consistencies were between α = 0.75 and 0.95 for the total score and the scores of the subscales [33].

#### 2.3.3. Sexual Inhibition/Sexual Excitation Scales-Short Form (SIS/SES-SF)

Based on the Dual Control Model of Sexual Response the SIS/SES-SF is a 14-item questionnaire that assesses participants' reactions in sexual situations on three subscales: one sexual excitation subscale (SES) and two sexual inhibition subscales (SIS1 and SIS2) [34]. While SIS1 measures sexual inhibition due to a threat of performance failure, SIS2 assesses sexual inhibition due to a threat of performance consequences, for example, unwanted pregnancy or sexually transmitted diseases [35]. In a first validation study of the German version of the SIS/SES-SF, internal consistencies of *α* = 0.82 for SES, *α* = 0.60 for SIS1 and *α* = 0.70 for SIS 2 were reported [36].

#### **3. Results**

#### *3.1. Relationship and Sexual Satisfaction*

While more women from the control group viewed sexuality as an important part in their lives (HCs: 53.8% vs. ASD: 20%), no differences occurred concerning relationship and sexual satisfaction between women with ASD and women from the control group. Moreover, more female controls rated themselves as being sexually attractive (HCs: 53.8% vs. ASD: 20%).

When comparing ASD men to men from the control group, it was found that more male controls were satisfied with their current relationship (HCs: 63.8% vs. ASD: 11.1%) and sexual life (HCs: 59.6% vs. ASD: 10.7%). Furthermore, more male controls viewed themselves as being sexually attractive (HCs: 73.7% vs. ASD: 3.6%), while no differences occurred concerning the importance of sexuality.

Finally, when comparing ASD women with ASD men it was found that more ASD women were currently in a relationship (women: 46.2% vs. men: 16.1%), more ASD women were satisfied with their current relationship (women: 44.4% vs. men: 11.1%) and ASD women viewed themselves as more sexually attractive than ASD men (women: 20.0% vs. men: 3.6%). On the other side more ASD men viewed sexuality as an important part in their life (women: 20.0% vs. men: 50.0%). No differences occurred concerning sexual satisfaction.

#### *3.2. Sexual Dysfunctions*

The female controls scored significantly higher on all FSFI subscales indicating that they reported less problems with sexual desire, sexual arousal and sexual satisfaction, lubrication, orgasm quality and less sexual pain compared to the women with ASD (Table 2). The male controls also reported about significantly better overall sexual functioning than the ASD men. However, when assessing the IIEF subscales this accounted only for erectile functioning and sexual intercourse satisfaction, while no differences were found concerning orgasmic functioning and sexual desire. Furthermore, more ASD men than male controls were below the cut off for erectile functioning problems justifying the use of medication to treat these problems, however, this difference only closely approached the intended level of significance (Table 2).


**Table 2.** Average questionnaire sum and subscale scores compared between autism spectrum disorder (ASD) patients and healthy controls (HCs).

FSFI = Female Sexual Function Index, IIEF = International Index of Erectile Functioning.

In order to address the impact of the assessed clinical characteristics on sexual functioning in our ASD participants, we calculated two linear logistic regression analyses (one for the male ASD participants and one for the female ASD participants) with overall sexual functioning (IIEF sum score in men and FSFI sum score in women) as the outcome variable and regular alcohol or illegal drug intake, any psychiatric disorders, any endocrine disorders, genital abnormalities, regular intake of psychopharmacological agents and hormone replacement therapy as predictors. Table 3 gives an overview about the results of the logistic regression analyses showing that neither in ASD men nor in ASD women any of the additionally assessed clinical features had a significant influence on the overall sexual functioning scores.

**Table 3.** Linear logistic regression addressing the relationship between clinical factors and sexual dysfunctions in individuals with Autism Spectrum Disorder (ASD).


#### *3.3. Sexual Excitation and Sexual Inhibition*

Table 2 also provides an overview about the SIS/SES-SF scores in ASD women and men compared to the HCs. While ASD women had significantly lower scores in sexual excitation compared to their non-ASD counterparts, ASD men had significantly higher scores on the sexual excitation subscale. Furthermore, women with ASD also had higher scores in SIS1 and SIS2, while no differences occurred between ASD men and the HCs.

#### *3.4. Correlational Analyses*

Women with ASD scoring higher on SES reported fewer overall problems with sexual functioning (Table 4). More specifically, higher SES scores were correlated with fewer problems with sexual desire, sexual arousal, lubrication and orgasm. No significant correlations were found between SIS1 and SIS2 and any of the FSFI subscales in ASD women. Comparably, in healthy women SES was also positively correlated with overall sexual functioning. Furthermore, SIS2 was negatively correlated with sexual desire and sexual arousal, meaning that those with higher SIS2 scores reported about more problems with sexual desire and sexual arousal.

In the male controls higher scores in SES were correlated with higher scores with overall sexual functioning, erectile functioning, sexual desire and overall sexual satisfaction (Table 5). In contrast, no association was found between SES and any of the IIEF subscales in ASD men, however, SIS1 and SIS2 were negatively correlated with overall sexual functioning as well as most of the IIEF subscales.


**Table 4.** Correlational analysis between the sexual inhibition/sexual excitation scales short form (SIS/SES) and female sexual functioning assessed with the female sexual function index (FSFI).

\*\* *p* < 0.01.

**Table 5.** Correlational analysis between the SIS/SES and male sexual functioning assessed with the IIEF.


\* *p* < 0.05; \*\* *p* < 0.01.

#### **4. Discussion**

To our knowledge, this is the first study to explore symptoms of sexual dysfunctions using self-report scales in a cohort of women and men with high-functioning ASD in comparison with a matched control group. In line with previous research, significantly less ASD men and women were currently in a romantic relationship compared to the HCs [37,38]. As was suggested in the introduction the disorder-specific symptoms like deficits in intuitively understanding social and nonverbal communication cues, difficulties in perspective-taking, cognitive and behavioural inflexibility as well as affective dysregulation, might hamper the initiation of romantic relationships in ASD individuals. Furthermore, both men and women with ASD reported lower relationship and sexual satisfaction than the HCs [39]. Within the present study we did not evaluate whether or not the current spouse of our study participants was diagnosed with ASD as well, however, this seems to be quite important, because it was shown that having a relationship with another autistic individual leads to an improved relationship satisfaction [13]. Women with ASD often have better social learning abilities, share more common interests with their peer group, have more advanced coping strategies and show less overt restricted interests and repetitive behaviours [40,41]. Thus, their problems in initiating and maintaining a romantic relationship are often not as pronounced as in ASD men, explaining why more women with ASD than men within the present study were in a romantic relationship [42]. Although fewer men with ASD were in a relationship compared to female ASD individuals, more men reported that sexuality was an important part in their life. These unfulfilled sexual desires could indicate that overall ASD men experience more distress concerning their own sexuality than ASD women.

Concerning sexual functioning it was found that men with ASD reported more problems with erectile functioning than the HCs. However, despite the findings of previous research that men with erectile dysfunctions from the general population usually have lower SES scores than those without erectile dysfunctions, the ASD men had significantly higher scores in sexual excitation than their non-affected counterparts [23,25,26]. This quite unexpected result could be the consequence of the peculiarities in sensitive perception in ASD men. On the one side the hypersensitivities experienced

by many ASD men could cause that discrete (and even non-sexual) cues could be perceived as quite intense and sexually arousing, meaning that ASD men get sexually aroused more easily. In terms of the Dual Control Model this could be translated to a higher sexual excitation (e.g., Item 1 of the SIS/SES-SF: "When a sexually attractive stranger accidentally touches me, I easily become aroused"). On the other side, as quickly as sexual arousal might arise in ASD men, it could also decline again because due to the pronounced hypersensitivity a constant and increasingly strong stimulation is necessary in order to hold sexual arousal on an adequate level. This in many cases might not be possible and thus in the long run men with ASD experience more problems with erectile functioning because of the possibly more rapidly decreasing sexual arousal during (sexual) stimulation. Supporting this line of argument, the individual propensity of sexual excitation did not correlate neither with the IIEF sum score nor with any of the IIEF subscales, suggesting that sexual excitation might refer to a different kind of behaviour in ASD men compared to healthy men. A further possible explanation could be that ASD men have difficulties in recognizing and classifying signs of excitement and therefore answered the questions regarding excitement in a different manner than the male controls. Concerning the individual propensity of sexual inhibition, no differences were found between ASD men and the male controls. Furthermore, medium to large correlations in the expected direction were found between both sexual inhibition factors and the IIEF sum score and most of the IIEF subscales in the male ASD sample. These findings indicate that just like their non-affected counterparts, ASD men with a stronger propensity of sexual inhibition due to a threat of performance failure or due to a threat of performance consequences report about more sexual dysfunctions [23,26].

Comparably to the ASD men, the ASD women also reported significantly more sexual dysfunctions across all of the FSFI domains compared to the female controls. Just like in the ASD men this could be the consequence of the peculiarities in sensitive perception. However, the significantly lower sexual excitation and significantly higher sexual inhibition scores suggest that while in men hypersensitivities might be more important in the aetiology of sexual dysfunctions, in women it might rather be hyposensitivities. Women with ASD might need more intense sexual stimulation to become and stay sexually aroused during having sex and to reach an orgasm, explaining the lower sexual excitation scores. However, the ASD women in the present study also reported more frequently about sexual pain problems, suggesting that not only hyposensitivities but also hypersensitivities could be of relevance and it could be possible that normotypical sexual intercourse is perceived as painful by some ASD women. Both women with ASD and female controls scoring higher on sexual excitation reported better sexual functioning. Comparably, previous research found that in women from the general population higher SES scores were positively correlated with a more positive attitude towards sexuality, higher overall sexual functioning, higher sexual desire, higher sexual arousal, less problems with lubrication and higher orgasm quality [36,43,44]. Although women with ASD had significantly lower sexual inhibition scores than female HCs, no association was found between sexual inhibition and sexual functioning in the ASD women. ASD women have an up to three times increased risk to be sexually victimized than non-ASD women, which could explain the higher sexual inhibition scores. It could have been expected that those individuals with an especially pronounced propensity of sexual inhibition would also show more sexual dysfunctions, however, this was obviously not the case [45].

The findings of the present study are limited because diagnoses were assessed via self-report and one cannot be sure that all participants were diagnosed by a trained psychologist or psychiatrist. Due to data protection regulations we were not allowed to contact the diagnosing clinicians in order to verify the diagnoses of our study participants. We tried to reduce false positives by using the well-established cut-off of the German version of the AQ-SF, which proved in other studies to be sufficiently sensitive and specific to assess autistic symptomatology [27]. Nevertheless, future studies should choose a more standardized assessment approach concerning the verification of clinical diagnoses, for example by conducting a clinical interview. Furthermore, all participants were recruited through ASD self-help groups or ASD outpatient care centres, indicating that their contact with the medical system was due to their symptomatology. Although we assessed comorbid psychiatric disorders in general, we did

not evaluate specific disorders, such as depressive or anxiety disorders, which are highly prevalent in autistic individuals and could affect sexual well-being and functioning. Using diagnostic interviews in future studies could help to also prevent this shortcoming. Furthermore, we did not assess intellectual functioning of our participants (e.g., by assessing IQ scores), however, as our study participants had on average 12 years of school education it can be assumed that all participants possessed at least average intellectual abilities. It is possible that especially individuals with a higher interest in sexuality-related issues and perhaps also with more sexual problems, were more likely to volunteer to participate in the present study leading to a sampling bias and an overestimation of sexual problems. However, it is likely that this should have also accounted for the individuals in the control group, thereby equalizing a possible overestimation of the actual rate of sexual dysfunctions in the ASD group at least to some degree. Our results are further limited by the fact that we did not evaluate whether or not the spouses of our ASD individuals were diagnosed with ASD as well. As stated above previous research has suggested higher sexual and relationship satisfaction when both companions are diagnosed with ASD. Thus, future studies addressing sexual functioning of ASD individuals should definitely consider this point. Finally, we did not evaluate hormonal profiles of our study participants, although differences in hormone serum concentrations could have a great impact on sexual functioning as well. Future studies should therefore assess the hormonal profiles of ASD individuals in order to find out if the increased prevalence of sexual dysfunctions found in ASD individuals is due to somatic or psychiatric reasons or both. At least though we did not find any differences in the self-reported frequency of endocrine disorders, genital abnormalities or hormonal substitution treatment.

The present study has shown that a considerable number of individuals with ASD report about a general relationship and sexual dissatisfaction and about sexual dysfunctions. Furthermore, the sexual problems are probably to a large part attributable to the disorder-specific symptoms, such as impaired social and interpersonal skills, difficulties in perspective taking and theory of mind and the peculiarities in sensitive perception. This points out that there is a great need for specialized treatment programs teaching individuals with ASD how they can, despite their disorder, have a fulfilling and satisfying sexual life. Unfortunately, such treatment programs are almost non-existent up to now, at least for adults with high-functioning ASD.

**Author Contributions:** Conceptualization, D.T., P.B. and D.S.; methodology, D.T., P.B. and D.S.; validation, D.T., P.B. and D.S.; formal analysis, D.T.; investigation, D.T. and D.S.; resources, D.T., P.B. and D.S.; data curation, D.T. and D.S.; writing—original draft preparation, D.T.; writing—review and editing, P.B. and D.S.; supervision, P.B.; project administration, D.T. and D.S.

**Acknowledgments:** The present study was part of the doctoral thesis of Stefanie Schmidt, therefore we want to thank Stefanie Schmidt for her assistance in collecting the data.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Mental Health and Proximal Stressors in Transgender Men and Women**

#### **Noelia Fernández-Rouco 1, Rodrigo J. Carcedo 2,\*, Félix López <sup>2</sup> and M. Begoña Orgaz <sup>2</sup>**


Received: 1 March 2019; Accepted: 20 March 2019; Published: 25 March 2019

**Abstract:** This paper explores the subjective perception of some personal and interpersonal aspects of the lives of transgender people and the relationship they have with their mental health. One hundred and twenty transgender people (60 men and 60 women) participated in semi-structured interviews. Following quantitative methodology, analysis highlighted that social loneliness is the main predictor of lower levels of mental health (anxiety and depression) for both genders and recognized romantic loneliness as the strongest factor among transgender men. In both cases, higher levels of loneliness were associated with lower levels of mental health. The results have guided us to improve institutional and social responses and have provided an opportunity to promote the mental health of transgender people.

**Keywords:** transgender; anxiety; depression; social loneliness; romantic loneliness

#### **1. Introduction**

The mental health of transgender people is frequently disturbed in several spheres [1]. According to this, the Minority Stress Model asserts that mental health distress is often the result of a hostile or stressful social environment [2]. This model describes the processes by which sexual and gender minorities are subjected to minority stress: (a) distal or external stressors (environmental), such as exposure to discrimination and violence; (b) proximal interpersonal stressors such as feelings or expectations that external stressors will occur and the need to protect oneself from these external stressors; and (c) proximal personal stressors that reflect an internalization of negative attitudes and prejudice from society. Conversely, interactive and internalized proximal resilience is also possible, with internalization of positive self-image, use of adaptive coping skills and community attachments. Interactive and internalized proximal stressors are frequently described as distressing. The cumulative stressors can serve to overwhelm themselves and to lead to poor mental health outcomes [3].

Over the last decades, several studies have been focused on transgender people's mental health and other personal and interpersonal variables (stressors) including self-esteem and body image [4], coping skills [5], social and emotional loneliness [6], sexual satisfaction [7] or anxiety and depression [8], yet there are no studies in Spain analysing how all these topics are able to explain the state of transgender people's mental health.

This work focuses on internalized proximal stressors (self-esteem, body image and coping skills) and interpersonal ones (social and emotional loneliness and sexual satisfaction), as well as the associations occurring in transgender people's mental health (anxiety and depression). To improve the empowerment and mental health of transgender people, proximal stressors (more modifiable taking into account personal aspects) need to be identified, which would provide both transgender people and professionals the opportunity to intervene.

#### *1.1. Mental Health: Anxiety and Depression in Transgender People*

The concepts of mental health and the specific nature of the relationship between anxiety and depression have been much debated. Research from the past decades has been reviewed to assess whether there is a quantitative or qualitative difference between anxiety and depression. Anxiety and depression syndromes have been studied both separately and combined to determine whether a quantitative or qualitative difference exists between them [9]. In the end, although there are several studies supporting comorbidity between anxiety and depression [10], they are commonly perceived as different; depressed disorders are characterized by a devaluation of self and negative attitudes toward the past and future, whereas anxiety disorders are marked by themes of danger and anticipated harm [11].

Although mental health problems may be self-limiting or may respond to self-help or to lay-help [12], delaying or avoiding formal care can result in problematic consequences. Too, the duration of untreated illness is associated with worsened outcomes in mental health problems such as major depressive and anxiety disorders [13]. The stigma resulting from a context in which power is exercised to the detriment of members of a social group [14], in this case, transgender people, includes such behaviours as labelling, separation, stereotype awareness and prejudice and discrimination. This stigma, along with mental health problems, is an important factor which prevents people from seeking help [15]. Additionally, this stigma plays an important role in limiting the opportunities and access to resources of transgender people in a number of critical domains (e.g., employment, healthcare, etc.), while continuously having a detrimental effect on their mental health [16]. A large body of literature points out that transgender people experience greater mental health problems, such as depression and anxiety [17,18] than do cisgender individuals (cisgender refers to those who are not transgender). Concretely, transgender people experience greater quantities of stressors from childhood which result in an increase of mental health problems such as depression and anxiety [19]. Transgender individuals, too, face a host of minority stressors specific to their sexual and gender minority identities. Viewed from a broader perspective, stigmatized people may be more susceptible to mental health problems due to the accumulation of stressors experienced over the course of a lifetime, as opposed to simply experiencing those stressors in isolated, discrete moments [18].

In addition, many community-based surveys have found that women (with no differences between cisgender and transgender), on average, experience depressed moods more frequently than men, as measured by self-report scales [20]. Women also self-report higher levels of anxiety [21]. Too, though the range of anxiety being studied varies, findings show that transgender men experience anxiety more frequently than transgender women [8,22,23].

#### *1.2. Proximal Stressors in Transgender People*

Transgender people have been found to face multiple difficulties and interpersonal challenges [24]. Forms of rejection from family and loved ones [25], low levels of self-esteem [26] and body image problems resulting from an attempt to reject those body parts that they do not identify with [27], are all examples of such challenges. Furthermore, although the association between transgender status and sexuality is commonly taken for granted and though research exists regarding improvements in sexual functioning after transition [28] and the importance of sexual life for humans in general [29], there is no substantial evidence pointing to sexual satisfaction in this population but rather to an unsatisfactory sex life [30]. In any case, the importance of social relations is not unique to the transgender population; humans are social beings who form attachments from the moment they are born [31]. They have a fundamental, adaptive need to belong [32]. Additionally, coping skills are vital to living a successful life and to maintaining a healthy mental health state [33,34]. Coping mechanisms, therefore, have been theorized to buffer the effects of mental health problems which result from stigmatization [2].

The impact that stressors have on both physical and mental health have been summarized in previous studies [3]. This literature, however, does not take gender into account when studying transgender status, nor have previous studies looked at transgender men or women individually. Finally, although certain stressors were studied both separately and jointly, no comprehensive studies yet exist in which proximal stressors are examined, including that of self-esteem, body image, coping skills, loneliness (social, family and romantic) and sexual satisfaction.

#### *1.3. Associations Between Proximal Stressors and Mental Health for Transgender People*

Much research exists linking different stressors to anxiety and depression. A large body of literature exists in which the relationship between self-esteem and depression is discussed. Furthermore, there is a growing body of longitudinal studies which indicate low levels of self-esteem predetermine depression; and correspondingly, people with high levels of self-esteem appear to have a lesser risk of suffering from depression [35]. In the same line, several theories postulate that a higher level of self-esteem serves as a buffer against anxiety [36]. This association was found within the transgender population as well, in relation to both anxiety and depression [37,38].

Body image is yet another factor that plays an important role in mental health [39]. Dissatisfaction with body image has been associated with an increase in mental health problems [40], a fact which holds true in the case of the transgender population [41]. The reinforcement of coping strategies, on the other hand, has proven effective in the management of issues encountered in the day to day, specifically in the prevention of problems related to mental health [42]. In fact, problem-focused coping predicted positive mental health outcomes among transgender youth [43] and the application of avoidant coping strategies during transitioning to manage gender-related stress has been associated with both depression and anxiety [8].

Interpersonal context has shown to be a major theme in the prevention or reduction of mental health problems. General loneliness was found to be an important variable for mental health [44]. Some authors have demonstrated that the emotional loneliness resulting from being cut-off from one's family is the strongest variable related to issues in mental health [45]. A large percentage of transgender individuals experience family rejection, social isolation and loneliness, which can result in a number of negative issues including mental health problems [46].

Sexuality is also a central topic for human development [47]. Specifically, there is a reciprocal relationship between certain mental problems such as anxiety and depression and sex problems [48]. Some studies in which other excluded populations were subjects, discovered that sexual satisfaction predicts positive mental health [49–52]. In terms of the transgender population, most studies which investigate sexual function are focused only on post-surgical outcomes [53].

Previous work studies the relationship that exists between different stressors and mental health but does not take into account the role that gender may play in this relationship owing to the fact that men and women are usually studied together [54], nor are the ways in which gender could affect the associations between stressors and mental health yet determined.

In summary, existing research has demonstrated that self-esteem, body image, coping skills, loneliness and sexual satisfaction are predictors of depression and anxiety. However, existing studies have not yet examined the relationships that exist between each of these variables and how gender moderates these relationships. The purpose of this study, therefore, is to examine the pattern of connections among each of these variables as they relate to transgender individuals, both transgender men and women. The current study aims to investigate (1) whether higher levels of self-esteem, body image, proactive coping skills, sexual satisfaction and lower levels of loneliness will be associated to better mental health and (2) whether differences exist between men and women.

#### **2. Experimental Section**

#### *2.1. Participants*

The sample consisted of 120 transgender people residing within Spain (93.3% Spanish and 6.7% foreigners, all from South America), 60 men (female-to-male) and 60 women (male-to-female). Participants were recruited in different cities and villages by this article's authors. Contact was made

via phone call or emails sent to people in LGTB or Transgender non-profit organizations and internet forums on websites aimed at LGTB or transgender information. The age range of the sample was 18 to 63 years old (M = 33.8; S.D. = 10.1). Of the participants, 19.1% had primary studies, 15% finished secondary school, 38.3% finished professional training and 27.5% finished university. We selected participants while maintaining a balanced number of men and women in three different reassignment moments (i.e., persons who assumed gender without any hormonal or surgery treatment, persons in hormonal treatment, persons in surgery reassignment process and persons who fully reassigned their sex). After stratifying by gender and reassignment moments, they were selected under a "snowball" sampling scheme [55].

#### *2.2. Procedure*

The people who responded positively to the recruitment method were given a standard description of the study and were evaluated for their eligibility to participate which consisted of the following criteria: individuals had to identify themselves as exclusively transgender at the time of the interview, did not have any mental health problem diagnosis or current state that impede to answer accurately to an interview (e.g., schizophrenia or being under the influence of drugs, etc.), expressed a consistent desire to have reassignment surgery and were 18 years of age or older. Eligible participants who expressed an interest in participating in the study were interviewed in-person at a location of their choosing (e.g., home, cafeteria, etc.). Individuals participating in the study did so voluntarily and there were no incentives in exchange for participation. The study was conducted in Spanish.

Face to face interviews lasting about 90 min were conducted in which each participant was orally asked all the questions in order to assure that everything was fully understood, taking into account the modest educational level of a considerable percentage of participants. First author of this paper introduced herself as member of the university staff and expressed our interest in the experiences of transgender people. Only upon establishing rapport, informing participants that they were free to leave the study whenever they wished and that their participation was confidential and voluntary and explicitly obtaining informed consent, did interviews commence. Upholding these ethical standards is vital for the collection of good quality data. The Good Practice Manual for Research of CSIC (2011) was followed regarding ethical standards [56]. In addition, this study respected the norms of the Declaration of Helsinki.

#### *2.3. Measures*

#### 2.3.1. Predictor Variables

Self-esteem. The instrument used was the Tennessee Self-Concept Scale 2nd Edition (TSCS:2) developed by Fitts and Warren as a review of Tennessee Self-Concept Scale [57,58]. The complete scale consists of 82 statements. The items are classified into three dimensions: (1) identity and self-concept: how does the individual see him/herself (30 items); (2) self-satisfaction or self-esteem: how does the individual accept him/herself (30 items); (3) self-behaviour: how does the individual behave towards him/herself (30 items). The short form is used with the first 20 questions and gives an indication of whether a person tends to see him/herself as generally positive and consistent or negative and variable. Scores from 1 (always false) to 5 (always true) are used, with higher scores reflecting higher levels of self-esteem. This instrument was chosen because it is standardized, easy to administer and has presented a good validity showing high correlations with other self-esteem scales [57]. Cronbach's alpha in this study was 0.83.

Body Image. The Body Image Scale [59] was used. A higher score indicates higher levels of dissatisfaction. On the 14-item Appraisal of Appearance Inventory (AAI), three independent observers (the diagnostician, a nurse from the gender team and the researcher) rated their subjective appraisal of the appearance of the subject on a 5-point scale of femininity/masculinity. Higher scores indicate higher levels of incompatibility with the appearance of the new gender. Cronbach's alpha in this study was 0.93.

Coping Skills. The Coping Skills Scale summarizes the dimensions described by Lazarus and Folkman [60,61]. It is a multidimensional instrument that assesses active coping, social support coping, avoidant coping cognitive passivity and repression and avoidant coping behaviour or refusal.

Certain items were eliminated from the scale for use in this study as they were not considered to be adequate indicators of the coping strategy. The scale has been adapted according to the characteristics of the participating sample and by combining the two avoidance coping subscales into one. The response format, however, was not altered: a scale from 1 (I have never faced a situation like that) to 4 (I have come into contact with a situation like that many times) was applied. An exploratory factor analysis was conducted in order to pinpoint these modifications, yielding five factors (66.95% variance explained) that ultimately were grouped into three factors to rule out the items in our sample did not indicate the use of the strategy for coping in the original scale: active coping strategy, coping strategies and social support avoidant coping strategy, which account for 60.96% of the variance.

In our study, internal consistency for the subscale of social support coping corrected (7,8,13,17) was an alpha of 0.70 for active coping subscale corrected (1,6,9,12) alpha was of 0.75 and avoidant coping subscale corrected (2,3,4,5,14) was 0.77.

Social and emotional loneliness. The short version of the Social and Emotional Loneliness Scale for Adults (SELSA-S) was used to measure both types of loneliness [45]. In fact, SELSA-S consists of three subscales labelled (a) social loneliness, (b) family-emotional loneliness and (c) romantic-emotional loneliness. Participants rated 15 items, 5 of every subscale. Items were rated on a 7-point Likert-type scale that ranged from 1 (strongly disagree) to 7 (strongly agree). The total score of every subscale was obtained by summing up the items, with possible scores ranging from 7 to 35. There is no total score for loneliness because this measure comes from a multidimensional perspective of loneliness. In this study Cronbach's alpha was 0.83 for family-emotional, 0.77 for social and 0.74 for romantic-emotional.

Sexual satisfaction. The subscale of sexual satisfaction of the Multidimensional Sexual Self-Concept Questionnaire (MSSCQ) was used to measure this aspect [7]. A total 5 of 5 items were scored on a 7-point Likert-type scale (expanding upon the original 5-point Likert-type scale) ranging from 1 (not at all characteristic of me) to 7 (very characteristic of me) comparable to a SELSA scale. Alpha was 0.96 and 0.95 in this study.

#### 2.3.2. Moderator Variable

Gender was recorded as 0 for transgender women and 1 for transgender men.

#### 2.3.3. Outcome Variables

The Anxiety and Depression subscales of The Symptom Checklist of Derogatis (SCL-90-R) were used to assess anxiety and depression [62]. Twenty-three items were scored, ten items for anxiety and thirteen for depression. For each item the person was asked to rate severity of depression experienced over the past week. Responses were scored on a five-point scale ranging from (1) not at all to (5) extremely. Cronbach's alpha was 0.92 for anxiety and 0.94 for depression,

For all the scales and subscales, a total score was obtained by adding up the individual scores and dividing them by the number of items answered.

#### *2.4. Analysis Strategy*

As the method of obtaining data was the interview method, no missing data was obtained; all of the participants answered every question. After data curation, statistical techniques were used to process the data using descriptive, Pearson correlations and hierarchical regression analysis with the IBM SPSS 22 package (IBM, Armonk, NY, USA). Firstly, pertinent analyses were carried out to verify the reliability, normality, independence and homoscedasticity assumptions using Cronbach's alpha, Kolmogorov-Smirnov test and Q-Q plots, the collinearity statistics (tolerance index and variance inflated factor—VIF and the Breusch-Pagan test respectively. Secondly, independent samples *t*-test were used to assess the statistical significance of gender differences. Thirdly, Pearson bivariate correlations were used to explore the associations between men's and women's mental health and stressors. Fourthly, hierarchical multiple linear regression analysis was used to study the moderating effect of gender on the criterion variables (anxiety and depression). Before computing these, the assumptions of the presence of normality, linearity and homoscedasticity, along with the absence of multicollinearity were tested. Predictors were entered into the first step (main effects) and interactions between gender and predictors were entered in the second step (interactions between gender and those predictor variables that showed a different association seem to have responded in a different way in relation to the gender of the participants). When an interaction is significant, two separate regression models for each level of the moderator were conducted. Alpha level of 0.05 was used. Finally, power analysis was obtained using the G\*Power program [63] and heteroscedasticity between the predictor and the criterion variable was run through the macro Heteroskedasticity SPSS [64].

#### **3. Results**

The Cronbach's alpha showed a good reliability and the residual variance was constant with normality distribution.

All the predictors showed a linear relationship with anxiety and depression as it was observed in the scatterplot of the standardized residuals with the standardized predicted values. Q-Q plots and the level of significance obtained when applying the Kolmogorov-Smirnov test (up to 0.05) showed a good normality. When testing multicollinearity, the tolerance index values for the studied variables were up to 0.78 for anxiety and 0.69 for depression, which indicated the independence of the contributions of the predictor variables, producing variance inflated factor (VIF) scores lower than 10 for all the predictors. Finally, heteroscedasticity was an accomplished assumption because Breusch-Pagan (LM = 9.87; *p* = 0.20 for anxiety and LM = 3.05; *p* = 0.96 for depression) test was not found significant.

#### *3.1. Gender Differences in Proximal and Mental Health Variables*

Descriptive statistics of predictor and outcome variables are displayed in Table 1 for transgender men and women respectively. To examine whether there are mean differences based on the study variables, *t*-tests for independent samples were conducted. Differences in anxiety, body image, social loneliness and sexual satisfaction were found. Transgender women showed higher levels of anxiety, social loneliness and sexual satisfaction and a poorer body image.


**Table 1.** Descriptive statistics for men and women in predictor and outcome variables.

#### *3.2. Proximal Aspects and Mental Health for Transgender Men and Women*

Bivariate correlations of interpersonal variables with anxiety and depression for both men and women, are shown in Table 2. All the stressors were associated with anxiety and depression except for the case of active and social support coping strategies which were not significantly correlated with anxiety. Similarly, high correlations were observed between anxiety and depression.


**Table 2.** Bivariate correlations for all the sample (men and women together).

\* *p* < 0.05; \*\* *p* < 0.01.

To identify whether associations of stressors with anxiety and depression varied by gender, two separate bivariate correlational analyses were conducted for transgender men and women (see Table 3). Regarding the correlation between the stressors and anxiety, men showed higher associations for family loneliness, whereas women showed higher associations for body image and social loneliness. With respect to the correlations between stressors and depression, men showed higher correlations for family and social loneliness, whereas women presented higher correlations for body image and avoidant coping strategies, interestingly not showing significant correlations with loneliness. Some correlations differ for men and women: anxiety and self-esteem, body image, avoidant coping strategy and romantic loneliness, depression and active coping strategy, social support coping strategy and romantic loneliness. These variables would be entered in the regression models as interactions with gender (see Table 3).

**Table 3.** Bivariate correlations for men and women separately (transgender men above the diagonal and transgender women below the diagonal).


\* *p* < 0.05; \*\* *p* < 0.01.

#### *3.3. Proximal Stressors as Predictors of Mental Health (Anxiety and Depression)*

To test the effects of all predictor variables on symptoms of anxiety and depression, taking into account the role of gender variable in those relationships (moderating effect), two hierarchical multiple regression analysis were conducted using the two-step model with two steps of independent variables.

Taking into account anxiety as a criterion variable, the main effects model was significant (F (1, 118) = 36.87, *p* < 0.001). This model accounted for 23% of the variance of anxiety. In order to the study the moderating effect of gender, the interaction of each predictor with gender was also included in a second step (i.e., self-esteem, body image, avoidant coping strategy and romantic loneliness). The interactions model produced an increment of 4% variance and romantic loneliness × gender interaction was found to be significant. Hence, the interaction effects model was selected to explain anxiety (F (1, 117) = 7.79, *p* < 0.01) and its observed power was 0.85. In this sense, the predictor found to be significant for both genders was social loneliness and the predictor found only for men was romantic loneliness. Therefore, higher scores in social loneliness were associated with higher levels of anxiety for transgender men and women and higher scores in romantic loneliness were associated with higher levels of anxiety only for transgender men (F (1, 58) = 32.77, B = −0.78, *p* < 0.001) (see Table 4 and Figure 1).

Regarding the regression analysis conducted to explain depression, the main effects model was also significant (F (1, 116) = 12.15, *p* < 0.001). This model accounted for 53% of the variance of depression. The interaction of each predictor with gender was also included in a second step (i.e., active coping strategy, social support strategy and romantic loneliness) to study the moderating effect of gender. The interactions model produced an increment of 5% variance and romantic loneliness × gender interaction was found to be significant. Hence, the interaction effects model was selected to explain depression (F (1, 115) = 10.84, *p* < 0.001) and its observed power was 0.99. In this sense, the predictor found to be significant for both genders was social loneliness, avoidant coping strategy and body image and the predictor found only for men was romantic loneliness. Therefore, higher scores in social loneliness, avoidant coping strategy and a poor body image was associated with higher levels of depression for transgender men and women, and higher scores in romantic loneliness explained higher levels of depression for transgender men (F (1, 55) = 6.31, B = −0.49, *p* < 0.05) (see Table 4 and Figure 1).


*J. Clin. Med.* **2019**, *8*, 413

**Figure 1.** Romantic loneliness × gender interaction effect on mental health: (**a**) Depression; (**b**) Anxiety.

#### **4. Discussion**

The study aimed to investigate the situation and the relationship between proximal stressors and mental health capacity among transgender adults in Spain. Several significant differences were found in some stressors and in the mental health of both men and women. Specifically, transgender women were found to have higher levels of anxiety, poorer body image, higher social loneliness and higher sexual satisfaction, similar to the results found in previous literature [65,66]. Women have been particularly stigmatized because by transitioning from male to female and deviating from their expected gender role, prior social status is lost [1,67]. Perhaps women show greater sexual satisfaction due to the fact that their sexual life is a private sphere, one that is not publicly visible and, unlike other contexts such as social relationships, it is a realm in which they can experience more freedom. There is no literature on this subject but it has become a topic of special interest for future research.

On the subject of mental health, anxiety and depression typically occur simultaneously [68], a fact which holds true for transgender people as well [69], although they are commonly accepted as separate concepts [70]. Considering the association between proximal stressors and mental health, this study investigated the effects of proximal personal (self-esteem, body image and coping skills) and proximal interpersonal (social, family and romantic loneliness and sexual satisfaction) stressors on transgender men's and women's psychological health (anxiety and depression). Identifying which are the most important predictors and how to minimize them would be a useful tool for the design of future clinical and research interventions. The findings are consistent with previous research in that several proximal stressors were found to be associated with poor mental health among transgender people [8,46,71].

With respect to the ways in which these stressors are associated with poor mental health, social loneliness accounts for anxiety in both men and women, whereas romantic loneliness only accounts for it in men. Additionally, depression is accounted for by the level of social loneliness, body image and the use of avoidant coping skills in both men and women, though it is only accounted for by romantic loneliness, again, in in the case of men. Gender differences, therefore, are only significant in the case of romantic loneliness. These results substantiate previous findings regarding these variables and that of psychological health in different populations. In fact, there is empirical evidence regarding the fact that loneliness anticipates anxiety and depression [44,72–75]. There are no previous studies, however, concerning the role of gender in regard to romantic loneliness.

On the other hand, coping skills also play a role in one's mental health. The use of ineffective coping skills can either hinder or promote anxiety and depression [33]. The coping strategy of avoidance, which has shown to be ineffective in resolving complex life circumstances, causes people to experience significant levels of distress [76]. Discomfort with body and high desire of reassignment between the participants (at different levels) is common among members of our sample, related with a poor mental health [27].

For the results of our study to be accurately interpreted, certain limitations must be considered. First, the measures used in the study were all self-reported, a factor that may be associated to higher levels of response bias. Nevertheless, self-reported measures are an effective method in which to assess mental health [77]. Second, the study used a cross-sectional design which does not allow for understanding causal pathways. Nevertheless, the study contributes to our understanding of the significant association between proximal stressors and mental health, taking into account the moderating effect of gender. Third, the use of convenience sampling limits the generalizability of the findings, although it allowed us to access people in different situations. Fourth, the bidirectionality of the relationship between some of the stressors and the mental health can be considered a limitation. This issue is partially ameliorated by the fact that the outcome variables had a timeframe that was more proximal to the reporting period, whereas the independent variables had more distal timeframes, meaning that a larger body of literature exists in which proximal stressors foresee issues related to mental health. Finally, no other situations of disadvantage linked to mental health (socioeconomic status, culture, ethnicity, etc.), that could potentially have affected what was found in the analysis have been studied. Thus, in addition to transgender experience, gender was included as an important variable to be considered.

Future research should delve deeper, including looking into distal stressors and other mental health indicators. It could be interesting as well to separate those who have long-term mental health issues and those who do not. In this way, data collection and research projects are possible not only in the short-term but long-term as well, which would then allow for a more developmental perspective. Finally, qualitative research would allow for better understanding of subjective experiences in relation to stressors and mental health. All these research suggestions could be useful for a better understanding of the transgender experience.

Finally, notwithstanding the mentioned limitations and future research suggestions, the current study contributes to the literature on the subject by (1) exploring proximal personal and proximal interpersonal stressors and mental health in the transgender population, as well as differentiating between men and women; (2) highlighting the relationship between proximal stressors and mental health in this population; and (3) emphasizing the role of gender as a moderator of the relationship between stressors, specifically romantic loneliness and mental health. These contributions could lead to professional intervention which would promote the mental health of transgender people. Based on our results, interventions looking to reduce social loneliness, avoidant coping strategies, poor body image and romantic loneliness (among men) would be compelling, as would the impact each has on transgender individuals' mental health. Practitioners should to be aware the importance of relationships and the impact of loneliness on in transgender's mental health. Promoting a good relational network, both friendships and romantic partners, has always to be considered in any intervention with this population. In fact, working on social meaningful connections would buffer other feelings of loneliness, such as romantic loneliness. In this case, this seems to be especially important for transgender men in the context of romantic relationships.

Additionally, it is known that this population lives in stressful an environment due to different situations such as stigma, transphobia, and/or violence [1]. All these circumstances may promote the utilization of avoidant coping strategies in order to protect themselves from distress. However, as we have observed in this study, the use of these strategies may individuals be more prone to depression.

Finally, practitioners should focus on individuals' body image. This is an important aspect in order to prevent depression. Developing an accurate evaluation and intervention and also working on individuals' context to prevent from discrimination due to body image are important elements to be considered.

**Author Contributions:** Conceptualization, N.F.-R., F.L. and R.J.C.; Methodology, N.F.-R., R.J.C. and B.O.; Software, N.F.-R. and R.J.C.; Validation, N.F.-R. and R.J.C.; Formal analysis, N.F.-R., R.J.C. and B.O.; Investigation, N.F.-R.; Resources, N.F.-R. and R.J.C.; Data curation, N.F.-R.; Writing—original draft preparation, N.F.-R., R.J.C., F.L. and B.O.; writing—review and editing, N.F.-R., R.J.C., F.L. and B.O.; visualization, N.F.-R. and R.J.C.; supervision, N.F.-R.; Project administration, N.F.-R.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
