The Effects of Cyberbullying and Non-consensual Sexting on Gender Minority Stress and Psychological Functioning in Transgender Adults

Abstract

Despite the comparative safety of online environments, transgender and gender diverse (TGD) people can be exposed to cyberbullying and non-consensual sexting. An online survey of 146 TGD adults examined whether this contributes to gender minority stress. Multiple mediation analyses revealed links between cyberbullying and gender victimization, and between non-consensual sexting and victimization, discrimination, and gender rejection. These forms of gender minority stress fully mediated paths from cyberbullying and sexting to psychological distress in both gender binary and non-binary TGD people. Results are discussed from the perspective of trans-affirming counselors seeking to understand emerging online threats to their clients’ mental health.

Keywords:

• Anxiety
• cyberbullying
• depression
• minority stress
• sexting
• stress

Gender minority stress and the use of online environments as safe havens

Transgender and gender diverse (TGD) people are a minority group who are at higher risk of being subjected to various forms of psychosocial stress (Testa et al., 2015). “Gender minority stress” can take the form of discrimination and disadvantage in relation to access to health, housing, and employment (Ho & Mussap, 2017), victimization in the form of harassment, threatened or actual physical harm, or unwanted sexual contact (Jenkins et al., 2020), rejection, exclusion, or non-affirmation of gender identity by family, friends, colleagues, religious groups, and romantic partners (Bry et al., 2018), and/or internalized transphobia in the form of guilt, shame, and self-doubt (Kline & Randall, 2021; Tebbe & Moradi, 2016). Gender minority stress can cause mental ill-health and interfere with psychosocial functioning (Lefevor et al., 2019). It can also interfere with the ability to express one’s gender identity openly (Fuller & Riggs, 2018), with around half of TGD people who come out to family and friends (James et al., 2015) or in the workplace (Budge et al., 2010) experiencing harassment, discrimination, or rejection, and/or reduced financial support (Pollitt et al., 2017).

The desire to express one’s gender and experience the psychosocial benefits of coming out while minimizing exposure to sources of gender minority stress has contributed to the rapid adoption by minority groups of online interactions (Austin et al., 2020). Numerous studies have documented this shift both in sexual and gender minorities, particularly among youth (Craig et al., 2021). Motivations include social interactions with likeminded people for the purposes of obtaining emotional support (Cannon et al., 2017; Green et al., 2015), exploring gender identity (Hanckel et al., 2019), and expressing this identity authentically (Cannon et al., 2017), connecting with TGD communities (Cavalcante, 2016), obtaining legal and medical information of relevance (Craig et al., 2021), and seeking intimate relationships with others, including “hooking up” (Lamont et al., 2018; Mazur, 2019). Central to these motivations is that the online environment facilitates the two key features of coming out—gender expression and access to social supports—in a relatively safe environment (Cannon et al., 2017), particularly when the off-line world is perceived as being hostile (Craig et al., 2020).

Online victimization in the form of cyberbullying

There is growing recognition that online interactions also pose new risks and threats to gender and sexual minorities in the form of cyberbullying (Cavalcante, 2016; Hanckel et al., 2019). These risks are typically less physical than experienced by minorities in the offline world but can nonetheless contribute to gender or sexual minority stress (Ingram et al., 2017; Mkhize et al., 2020).

Cyberbullying is defined as “willful and repeated harm inflicted through the use of computers, cell phone, or other electronic devices” (Hinduja & Patchin, 2015, p. 5). Cyberbullying behaviors include making and distributing offensive or threatening emails, texts, comments, or images (Ovejero et al., 2016). Electronic technologies facilitate bullying by allowing perpetrators rapid and widespread access to a large audience to whom offending material may be disseminated, and helping them remain anonymous and thus less likely to suffer repercussions for their actions (Barlett et al., 2016, 2019; Englander et al., 2009; Kowalski & Limber, 2007). Although cyberbullying in children and adolescents is less common than traditional offline bullying (Modecki et al., 2014), its prevalence is growing, with the prevalence of victimization experienced by children and adolescents ranging between 10 and 40%, and perpetration by children and adolescents ranging between 3 and 24% (Jadambaa et al., 2019). The prevalence of both cyberbullying perpetration and victimization tends to increase into young adulthood (<25 years of age) but then declines with age (Wang et al., 2019).

Research into the cyberbullying experiences of sexual and gender minorities has confirmed that cyberbullying in the form of verbal harassment and abuse is highly prevalent relatively and in absolute terms, with rates as high as 70% reported (Abreu & Kenny, 2018). Research also shows that cyberbullying can contribute to suicidal behaviors in minority youth and also adults (Schwickrath, 2012), depression, low self-esteem, and poorer academic performance (Abreu & Kenny, 2018). Moreover, these impacts are more severe and more commonly experienced in these minorities than in heterosexual and cisgender people of similar age (Powell et al., 2020). These early findings confirm that cyberbullying may contribute to gender minority stress in the online environment.

Online victimization in the form of non-consensual sexting

Sexting describes the sharing of sexually explicit content online, often in the form of images (Klettke et al., 2014). It has rapidly become a common form of expressing sexual interest by adolescent and young cisgender adolescents and adults (Dir & Cyders, 2015), with most young people reporting receiving sexts (e.g., Clancy et al., 2019; Klettke et al., 2014), and up to 20% admitting to sending or disseminating sexts (Walker & Sleath, 2017). However, there is evidence that the use of deceptive or coercive methods to obtain sexts, and/or the sending or sharing of these sexts without consent, have also been increasing (Walker & Sleath, 2017). These non-consensual sexting behaviors can contribute to negative mood, depression, and decreased self-esteem (Klettke et al., 2019), humiliation, shame, and reputational damage in cisgender people (Gassó et al., 2019; Orue & Calvete, 2019).

There is growing evidence that consensual sexting behaviors used by gender minority people to initiate and maintain relationships online may also expose them to victimization in the form of receiving unwanted sexts from others, being pressured to provide others with sexts, and having someone use or threaten to use sexts to “out” them without their consent (Ellyson et al., 2021; Lamont et al., 2018; Van Ouytsel et al., 2020). Sexual and gender minority people appear to be very aware of these risks, explaining to researchers how they are particularly vigilant during intimate online interactions, focusing on their personal safety and being wary of placing themselves at risk of emotional harm (Lamont et al., 2018; Ma et al., 2022). However, further quantitative analysis of non-consensual sexting experiences in this population is warranted (Ma et al., 2022), particularly research that focuses on the psychological risks posed by sexting, whether these risks constitute a distinct form of minority stress, or whether they are subset of cyberbullying stressors more generally.

Gender differences in cyber-victimization attitudes and behaviors

Ciswomen are more likely to be victimized online than cismen, including being more likely to receive unwelcome image-based sexts (Klettke et al., 2019), and having sexualized images of themselves distributed to others with harmful intent (Morelli et al., 2016; Stanley et al., 2018). These gender differences are thought to be an extension of gender roles and power imbalances that normalize interpersonal aggression, sexual violence, and coercion by cismen toward ciswomen in offline environments (Barlett et al., 2014; Bianchi et al., 2021; Lee et al., 2017; Van Ouytsel et al., 2018, 2019). Therefore, it might be expected that TGD people whose identity and presentation conform more closely to cis-normative assumptions about gender and appearance (Johnson, 2016) will be more likely to hold and be subject to the same gendered attitudes that underpin bullying in offline and online environments. This includes TGD people assigned female at birth but who identify as men (transmen) and those assigned male at birth but who identify as women (transwomen). Indeed, gender-binary TGD people are more likely to suffer harassment and violence offline compared to other gender minorities (Miller & Grollman, 2015; Scandurra et al., 2017; Sterzing et al., 2017), and also compared to those whose gender identity is outside the binary, such as non-binary, gender diverse, or genderqueer people) (Beemyn & Rankin, 2011; Dispenza et al., 2012; Sterzing et al., 2017; Thorne et al., 2019). They are also more likely to report social stigma and social anxiety in relation to their gender than their non-binary peers (Ho & Mussap, 2020). However, quantitative research in the area is limited, and further research needs to be conducted to examine differences in experiencing and responding to online victimization in these groups.

Present study

The literature review noted the increasing utilization of online environments by TGD people and the potential risks of exposure to cyberbullying associated with this. We also considered ways in which non-consensual sharing of image-based sexts might constitute an additional risk for this population. We noted both the limited research that exists into cyberbullying and non-consensual sexting in TGD people, and the dearth of research into the relevance of these victimization experiences to gender minority stress and psychological functioning more generally. Therefore, in the present study, we measured the separate contributions of cyberbullying and non-consensual sexting to gender minority stress in TGD people, we compared these victimization experiences as a function of the alignment of their gender identity to the traditional female/male binary, and we examined the extent to which victimization is associated with psychological distress. Our review also noted the (understandable) focus of previous research in the area of adolescents and young adults, and because of this the present study focused on adults (cf. Powell et al., 2020).

A convenience sample of people who identity as transgender, agender, or gender diverse was thus asked to complete a survey containing measures of cyberbullying victimization and non-consensual image-based sexting, gender minority stress, and psychological distress. We hypothesized:

H1: That online victimization experiences will be experienced as gender minority stress;

H2: that cyberbullying and sexting will make different/separable contributions to minority stress owing to the different contexts within which they occur;

H3: that the effects of online victimization on gender minority stress will result in psychological distress in the form of depression, anxiety, and/or stress; and

H4: that hypothesized relationships between online victimization, gender minority stress, and psychological distress will be stronger for gender-binary compared to non-binary TGD people.

Methods

Participants

Participants were 146 adults (18–58 years of age; M = 23, SD = 4.9) whose gender does not correspond to their sex assigned at birth; 34 (23.3%) of whom identified as transwomen, 28 (19.3%) identified as transmen, 69 (47.3%) identified as non-binary, and 15 (10.3%) who responded “other” or declined to respond to the question about gender. Self-reported ethnicities, in order of frequency, were: White (N = 102, 69.9%), Latinx (N = 10, 6.8%), Asian (N = 9, 6.2%), Jewish (N = 6, 4.1%), Indigenous (N = 5, 3.4%), Pacific Islander (N = 3, 2.1%), Black (N = 2, 6.2%), and mixed-race/multi-ethnic (N = 13, 0.9%). They resided in the USA (N = 44, 31.1%), Australia (N = 64, 45.7%), India (N = 22, 15.6%), Singapore (N = 3, 2.3%), UK (N = 3, 2.4%), and Canada (N = 2, 1.5%).

Materials

The study was conducted via a Qualtrics™ online survey. Demographic questions were followed by questions about sex assigned at birth and gender identity. Participants then completed the following measures:

Gender minority stress

The Gender Minority Stress and Resilience Measure was used to assess gender minority stress (Testa et al., 2015). Participants responded strongly disagree (0) to strongly agree (4) to statements describing examples of difficulties and challenges faced by TGD people. Note that only items relevant to minority stress (and not resilience) were utilized in the study. These items comprised the following subscales: gender-related discrimination, gender-related rejection, gender-related victimization, non-affirmation of gender identity, and internalized transphobia. Responses to items from each subscale were averaged. This measure has good reliability (Cronbach’s alphas between .80 and .95) and evidence of criterion validity in relation to measures of depression and anxiety (r values between .20 and .40) (Hidalgo et al., 2019). In the present study, Cronbach’s alphas were slightly lower (alphas ranging from .70 to .90) but correlations with measures of depression and anxiety (r = .20–.35) confirmed adequate validity in our sample (see Table 1).

Table 1. Bivariate Correlations and Descriptives (Gender Binary = Upper-Right Diagonal; Gender Non-binary = Lower-Left Diagonal)

														Gender binary
			1	2	3	4	5	6	7	8	9	10	11	Min	Max	M	SD	Skew/se	Kurtosis/se	α
1	Age		–	−.03	.10	−.15	−.15	.1	−.20	−.13	−.30*	−.23	−.27	18	58	24.76	6.79	7.34	13.87	–
2	Cyberbullying		.12	–	.30*	.32*	.19	.47**	.11	−.11	−.06	.15	.01	1	1.8	1.48	.26	−2.34	−1.28	–
3	Non-consensual sexting		−.06	.45**	–	.26	.28*	.46**	.12	.08	−.09	.29*	.13	0	5	1.08	1.27	4.04	1.66	–
4	Gender discrimination		−.13	.25*	.28*	–	.57**	.61**	.62**	.45**	.33*	.39**	.35*	0	2.67	.90	.66	1.82	.19	.65
5	Gender rejection		.01	.27*	.46**	.62**	–	.44**	.42**	.41**	.38**	.31*	.27*	0	2.17	.77	.58	2.16	−.14	.70
6	Gender victimization		.07	.41**	.52**	.47**	.53**	–	.47**	.29*	.02	.27*	.09	0	2	.48	.46	3.33	1.5	.68
7	Non-affirmation of gender		−.21	.17	.11	.47**	.49**	.30*	–	.44**	.26	.41**	.32*	0	3	1.44	.85	.00	−1.34	.87
8	Internalized transphobia		−.07	−.02	.10	.10	.49**	.12	.35**	–	.43**	.36**	.40**	0	3	1.25	.92	1.25	−1.7	.94
9	Depression		−.06	.15	.19	.19	.29*	.37**	.35**	.18	–	.57**	.64**	0	20	9.55	5.52	.72	−1.32	.92
10	Anxiety		−.08	.16	.29*	.16	.37**	.40**	.20	.10	.59**	–	.75**	0	19	6.15	5.07	1.96	−.95	.88
11	Stress		−.12	.03	.18	.18	.34**	.34**	.21	.21	.67**	.74**	–	1	20	8.78	4.66	.98	−.68	.83
	Gender non-binary	Min	18	1	0	0	0	0	0	0	0	0	0
		Max	39	1.81	4	3	2.33	2.11	3	2.75	21	21	21
		M	24.83	1.43	1	.77	.78	.4	1.7	.85	11.09	7.18	11.1
		SD	4.11	.25	.97	.67	.61	.5	.76	.81	6.43	5.02	5.64
		Skew/se	2.72	−1.48	2.75	3.65	1.99	5.79	−.91	2.47	−.21	2.28	.24
		Kurtosis/se	2.5	−1.78	.35	1.57	−.99	4.74	−.88	−1.21	−2.3	−.24	−1.32
		α	–	–	–	.74	.72	.81	.91	.91	.92	.83	.89

Notes: se: standard error; α: Cronbach’s alpha; *p < .05, **p < .01.

Cyberbullying

Cyberbullying victimization was assessed against three types of behavior experienced in the past year: (i) Being disparaged online (“Have you received rude or nasty comments from someone while online”), (ii) being the topic of rumor online (“Have you been the target of rumors spread online, whether they were true or not?”), and (iii) being the target of online attacks (“Have you received threatening or aggressive comments while online?”) (Ybarra et al., 2007). Participants responded never (1) to everyday/almost every day in the past year (6), and responses were averaged across the items.

Non-consensual image-based sexting

Non-consensual sexting was assessed against five types of non-consensual or coerced image-based behaviors experienced over the past year: (i) Receiving a sext (“Have you received an image-based sext that wasn’t asked for, or was unexpected?”), (ii) receiving an unwanted/unwelcome sext (“Have you received an image-based sext that was unwanted/unwelcome?”), (iii) sending a sext that one did not want to send (“Have you consented to send an image-based sext when you actually did not want to sext?”); (iv) experiencing negative consequences in relation to sending a sext (“Have you experienced negative consequences as a result of sending sexually explicit images of yourself to another person via text message or mobile app?”), and (v) having a sext forwarded without one’s consent (“Have you sent an image-based sext of yourself that was subsequently forwarded (to your knowledge)?”). These items were developed by Clancy et al. (2019), and participants responded No (0)/Yes (1) to each. Responses to the five types of behaviors that comprise the measure were tallied to give a total score for experiences of non-consensual sexting.

Psychological functioning

The Depression, Anxiety, Stress Scale (DASS-21; Antony et al., 1998) was used to measure generalized anxiety and stress. Participants responded (1) did not apply to me at all to (4) applies to me very much, or most of the time in response to symptoms of depression, anxiety, and stress. The DASS-21 has good psychometric properties even with non-clinical samples, including evidence of the reliability of each subscale (Cronbach’s alpha = .88, .90, and .93, for depression, anxiety, and stress, respectively) and evidence of criterion validity in relation to measures of state negative effect (r in the range of .60–.70) (Henry & Crawford, 2005) as well as demonstrated utility with TGD samples (Ho & Mussap, 2017). In the present study, the reliability of the three DASS-21 subscales was found to be adequate (Cronbach’s alpha = .92, .86, and .88, for depression, anxiety, and stress, respectively) (see Table 1).

Procedure

The research team complied with APA ethical standards in the treatment of participants, and the project received prior ethics approval from the Human Research Ethics Committee of our institution. Respondents were recruited via trans support groups online, through gender health clinics, and via Prolific. After reviewing an online information statement and indicating their consent, all participants completed the online survey with the measures presented in the order described in the Materials subsection.

Results

Data screening and variable creation were carried out using IBM SPSS, V29 (IBM Corp., Armonk, NY, USA). Missing items were randomly distributed and constituted fewer than 2% of responses (these items were omitted from variable creation). A single violation from normality—in the cyberbullying variable—was corrected using reciprocal transformation (with reverse-coding). No cases were identified as multivariate outliers (Mahalanobis’s distance p < .001). Descriptives and bivariate correlations are provided in Table 1.

Chi-square tests on victimization experiences revealed no significant differences between binary and non-binary respondents in terms of exposure to cyberbullying in the form of being disparaged online, being the topic of rumor online, or being the target of online attacks, χ²(5) = 5.69, 7.71, and 2.12, p = .338, .173, and .833, respectively. No significant group differences were observed in relation to exposure to non-consensual image-based sexting in the form of receiving a sext, receiving an unwanted/unwelcome sext, sending a sext that one did not want to send, experiencing negative consequences in relation to sending a sext or having a sext forwarded without one’s consent, χ²(1) = 2.77, .53, .47, .25, and 3.59, p = .096, .465, .491, .616, and .058, respectively. Independent-samples t-tests also confirmed no significant group differences in levels of gender minority stress in the form of gender discrimination, rejection, victimization, or non-affirmation reported, t(118) = −1.12, .05, −.89, and 1.78, p = .266, .959, .374, and .079 (equal variances assumed). The only exception was internalized transphobia which was significantly higher in binary (M = 1.25, SD = .92) compared to non-binary (M = .85, SD = .81) respondents, t(118) = −2.55, p < .05.

Table 1 shows that cyberbullying and non-consensual image-based sexting were moderately correlated with one another as well as with several of the gender minority stressors (gender discrimination, rejection, and victimization). This was evident in both binary and non-binary TGD respondents. The table also shows that gender minority stressors were broadly correlated with depression, anxiety, and stress (apart from gender discrimination in relation to stress, and internalized transphobia in relation to anxiety), again, for both groups.

To examine the nature of these relationships, a multiple mediation analysis was conducted in AMOS™ Version 29 (Amos Development Corporation, Meadville, PA, USA) via a bias-corrected bootstrap re-sampling method (Shrout & Bolger, 2002) using the path model shown in Figure 1. The model includes paths from the two online stressors (cyberbullying and non-consensual image-based sexting) to the multiple mediators in the form of gender minority stress variables (discrimination, rejection, victimization, non-affirmation, and internalized transphobia), and from these mediators to the three measures of psychological distress (depression, anxiety, and stress). Fit indices for the “optimized” version of this model fit on the combined data from binary and non-binary participants, with non-significant paths removed, confirm that it performed well against standard fit criteria (criteria are included in parentheses): χ²(23) = 29.46, p = .166, χ²/df = 1.28 (<5), root mean square error of approximation = .044, 90% CI [.000, .086] (RMSEA < .08), standardized root mean square residual = .066 (SRMR < .08), comparative fit index (CFI) = .99 (CFI > .95), and Tucker-Lewis Index = .98 (TLI > .95) (Byrne, 2010; Hu & Bentler, 1999).

Figure 1. Path Model Used to Conduct Multiple Mediation Analysis (Optimized by Removal of Non-significant Paths) on the Combined Data from Gender Binary and Non-binary Participants, along with a Table Insert Showing Correlations between the Gender Minority Stress Variables Included in the Model. Parameters Included in the Model Are Presented in Triads, Consisting of Values Obtained on the Combined Data (Leftmost Values), on Gender Binary Respondents Only (Middle Values), and Non-binary Respondents Only (Rightmost Values). Model Fit on the Combined Data: χ²(23) = 29.46, p = .166 ns; χ²/df = 1.28; RMSEA = .044, 90% CI [.000, .086]; SRMR = .066; CFI = .99, and TLI = .98, and on the Same Model but with Gender (Binary vs. Non-binary) Included as a Free Parameter: χ²(46) = 62.06, p = .057 ns; χ²/df = 1.35; RMSEA = .052, 90% CI [.000, .083]; SRMR = .074; CFI = .97, and TLI = .93. All Coefficients Significant at p < .05 Unless Indicated by “ns”

Figure 1 shows positive direct paths from cyberbullying to gender victimization, and from non-consensual sexting to gender discrimination, rejection, and victimization. R² values indicate that gender victimization was most closely related to the online stressors, with ∼25% of its variance determined by the combination of cyberbullying and sexting. Positive direct paths were also evident from gender rejection to depression, anxiety, and stress, and from gender victimization to anxiety. R² values show these relationships to be in the range of 7% (in relation to stress) to 14% (in relation to depression).

To test for group differences in these paths, multigroup path analysis was conducted in AMOS™ against a “baseline” model consisting of the original model depicted in Figure 1 but with gender group (binary vs. non-binary) included as a free parameter. Model fit for this baseline was confirmed to be adequate for subsequent comparisons: χ²(46) = 62.06, p = .057, χ²/df = 1.35, RMSEA = .052, 90% CI [.000, .083], SRMR = .074, CFI = .97, and TLI = .93. Imposing constraints on regression weights resulted in no significant change to model fit, Δχ²(Δdf) = 7.86(8), p = .448, indicating no statistically significant group differences in relation to direct paths. Additional constraints imposed on intercepts also resulted in no significant change to model fit, Δχ²(Δdf) = 9.08(6), p = .169, indicating no statistically significant group differences in terms of variance explained.

Significant positive indirect paths were observed from cyberbullying to anxiety by way of gender victimization, β = .04, p < .01, and from sexting to depression, anxiety, and stress, by way of a combination of discrimination, rejection, and victimization, β = .08, .13, and .07, respectively, p < .01. Inspection of results by gender group confirmed that the indirect effects from sexting were typically of greater magnitude in the non-binary group and that the single indirect effect from cyberbullying to anxiety was of greater magnitude in the binary group. The absence of direct paths from either cyberbullying or sexting to depression, anxiety, or stress, for both groups indicates that gender minority stress variables fully mediated the relationships between the online stressors and psychological distress.

Discussion

It has been suggested that the perceived safety of online environments has encouraged their use among sexual and gender minorities (Cannon et al., 2017) but also exposed them to new online stressors in the form of cyberbullying and non-consensual sexting (Cavalcante, 2016; Hanckel et al., 2019). The primary question addressed in our study was whether these online stressors have contributed to gender minority stress in TGD people and the extent to which it has affected their psychological well-being.

Our results suggest that cyberbullying and image-based non-consensual sexting were indeed associated with gender victimization, with ∼25% of victimization experiences linked to exposure to these online stressors. Non-consensual sexting was additionally associated with both gender discrimination and rejection, although neither online stressor was relevant to non-affirmation of gender or internalized transphobia. Together, these results suggest that the gender minority stress experienced by TGD people is shaped by a combination of online and offline experiences (cf. Barlett et al., 2014; Klettke et al., 2014; Pingault & Schoeler, 2017).

Neither cyberbullying nor image-based sexting were directly associated with depression, anxiety, or stress. At face value, this was unexpected given previous research showing that online hostility can cause severe psychological distress in TGD people (Cavalcante, 2016; Hanckel et al., 2019; Ingram et al., 2017; Mkhize et al., 2020). However, subsequent multiple mediation analysis revealed that the online stressors were related to anxiety, and to a lesser extent, depression and stress, indirectly by way of gender minority stress. The most important mediator was gender rejection—it was broadly relevant to all three outcome variables and entirely responsible for mediating the effects of sexting. The second mediator was victimization, however, only in relation to anxiety. The superiority of paths to anxiety is presumably because anxiety was the only outcome variable related both to gender rejection and victimization, thus providing it with two mediation paths.

No gender group differences were observed in terms of relationships between these variables. Based on previous evidence that gender binary TGD people are exposed to more social stigma, harassment, and violence in offline environments (Ho & Mussap, 2020; Miller & Grollman, 2015; Scandurra et al., 2017), our expectation was that gender binary TGD respondents would report more gender-based victimization and experience more gender minority stress. However, this prediction was not supported in our study, with both gender binary and non-binary groups reporting similar rates of online victimization, similar levels of gender minority stress (except for internalized transphobia which was more common in gender binary respondents), and similar relationships between online victimization, gender minority stress, and psychological functioning. Our results underscore the need to support TGD people and address the interpersonal, social, and structural factors that contribute to their victimization online, regardless of their specific gender identity or presentation.

Limitations and future research

Our reliance on self-report measures limited our ability to draw causal inferences from the results. Furthermore, because it was impossible to conceal from participants the general purpose of the study, or our focus on TGD people as a potentially vulnerable group in the cyber context, this may have induced various response biases in participants. One possibility was demand characteristics leading to over-reporting of negative psychological responses to online victimization. An alternative possibility was under-reporting due to social desirability (e.g., to protect self-esteem). Participant reactance was also possible, with respondents deliberately seeking to undermine what they might perceive as the vulnerable/victim framework underpinning our research. These problems cannot be addressed solely through improvements to survey methods, but call for integration of diverse methods, such as the use of mixed-methods research where survey responses can be validated against qualitative interview data from the same participants. Selection biases may also have occurred if, for example, our study tended to attract participants who had experienced negative psychological responses to online victimization in the past.

There was also the problem of time-order ambiguity inherent in our cross-sectional research design. For example, it is possible that respondents who had experienced online victimization in the past had become more aware of and/or more sensitive to subsequent actual or anticipated recurrences of victimization. Although this issue has not been studied in the trans context, sensitization effects of this sort have been confirmed in the context of cyberattacks perpetrated on cisgender people and this can influence survey responses (Debb & McClellan, 2021). This problem calls for more experimental research (while acknowledging the ethical challenges of this approach), and the need for longitudinal research to better determine cause-effect relationships.

The generalizability of our results was also limited by the ethnic diversity of our sample, with most of our respondents (70%) self-describing as white and residing in Western developed nations. This limitation also prevented us from conducting group comparisons by ethnicity or exploring intersectionalities involving ethnicity. Therefore, we encourage future researchers to apply an intersectional lens to the study of online experiences in gender minorities, particularly as a function of their ethnicity, gender group, and privilege/disadvantage (Simons et al., 2021; Taube & Mussap, 2021).

We also recognize that adversities in the form of aggression (experienced online or offline) can serve as an impetus for psychosocial growth (Counselman-Carpenter & Redcay, 2022). This is a potentially important consideration given the therapeutic opportunities offered by positive-psychology perspectives and growth-based approaches, particularly when dealing with adversity that is essentially unavoidable and/or irreversible (Taube & Mussap, 2022).

Finally, we note that our measure of non-consensual sexting was potentially limited in terms of its focus on image-based sexting and that our survey did not measure online victimization experiences as a proportion of overall time spent online.

Implications for trans-affirming counselors

Several aspects of our findings are potentially relevant to counselors working with TGD clients. Our results suggest that online aggression perpetrated against TGD people is not experienced merely as generic stress that happens to take place online. Rather, this aggression appears to target gender identity in a manner that causes gender minority stress. Concerningly, the dimension of gender minority stress that is most closely related to the online stressors—gender victimization—is arguably the most serious. It describes stress due to verbal harassment and threatened or actual physical violence and sexual assault directed to someone because of gender identity (Testa et al., 2015), with clear links to physical and mental ill-health, harm, and even mortality (Grant et al., 2011; Jones et al., 2015; Marshall et al., 2016).

Our results also show that the effects of online stressors on gender minority stress are relevant to mental health outcomes, particularly in relation to anxiety. We note that previous research with cisgender groups has found that online threats often lead to anxious anticipation of potential harm in the form of feared reputational damage, shame, and threatened disclosure of information (Gassó et al., 2019; Orue & Calvete, 2019). Similarly, offline threats directed at TGD people, such as being outed, are typically experienced as anticipatory anxiety/rumination (Brooks et al., 2020; Hunter et al., 2021; Timmins et al., 2017; Weiss & Raymond, 2020). One can draw parallels with research into social anxiety in TGD people, where there is evidence that social anxiety is due both to generic social stressors (e.g., interacting with authority figures, talking to strangers, speaking in public, etc.), as well as social stressors specific to anti-trans stigma (e.g., using public bathrooms, applying for and providing proof of ID, navigating health professionals and services, and interacting with conservatives and religious authorities, etc.) (Ho & Mussap, 2020).

The implication for trans-affirming counselors is that responses to online victimization of TGD people should not rely on general public safety or cyber safety prevention and treatment programs. As advocated by Price and Hollinsaid (2022), a more tailored approach is required, one that directly addresses the effects of stigma on the mental health of gender (and sexual) minorities, and better prepares therapists to identify and respond to these mental health challenges within the social contexts in which they typically occur (including in online environments). Such approaches need to prioritize the development of structural-level interventions that challenge the societal inequalities underpinning perpetration behaviors targeting TGD people. Our results, obtained with an adult sample, also caution against an exclusive focus on TGD adolescents (Wang et al., 2019) and an over-reliance on school-based interventions for cyber-victimization and perpetration (Cross et al., 2016). However, given the similar results obtained with binary and non-binary TGD participants, we suggest that there is probably no need to differentiate between TGD subgroups in relation to prevention or treatment.

Our results also confirm that cyberbullying and non-consensual sexting experiences are only moderately related to each other and are different in their relevance to gender minority stress. This challenges the assumption that non-consensual sexting is merely cyberbullying in the dating/hook-up context. There is growing recognition that non-consensual sexting is a distinct form of aggressive behavior, separate from bullying or harassment more generally (Bianchi et al., 2021; Morelli et al., 2016; Van Ouytsel et al., 2018, 2019). The present results support this distinction in the context of online acts of aggression against TGD people.

On this basis, we suggest that future research in this area ensures that separate assessments are made of cyberbullying and non-consensual sexting and that it is not simply assumed that a generic measure of online bullying/harassment will suffice. Similarly, for counselors working in the area, more effort should be made to distinguish between sources of aggression, with particular attention paid to the impacts of sexting on TGD mental health because of its broader associations with gender minority stress.

Conclusions

Our study confirmed that TGD people face online stressors in the form of cyberbullying and non-consensual image-based sexting, that these stressors are experienced as gender minority stress, and that they are associated with anxiety and, to a lesser extent, depression, and stress. We found that the online stressors were most closely related to gender victimization, that is, the stress in the form of harassment, coercion, and threatened or actual harm. Our study also highlighted the separability of non-consensual image-based sexting from cyberbullying in the TGD context. Specifically, it should not be assumed that sexting is merely an expression of cyberbullying in the dating/hook-up context. These results add to our understanding of the specific challenges facing gender minorities in online environments and call into question the belief (or perhaps hope) that online environments can serve as safe havens for TGD people.

Disclosure statement

No potential conflict of interest was reported by the author(s).