Results
Table
1 displays the proportion of Identify participants who had ever experienced SOGICE. In total, 124 participants (3.1%; 95% confidence interval [CI] 2.6–3.7) reported ever having experienced SOGICE. As presented in Table
2, a higher proportion of those exposed to SOGICE was aged 19–26 years (vs 14–18 years), were trans, non-binary, or another non-cisgender identity, were unsure/questioning gender, and reported severe deprivation. Table
3 outlines the associations between SOGICE exposure and NSSI and suicidality. Participants with SOGICE exposure were more likely than those without SOGICE experience to report NSSI, suicide planning, and suicide attempts. After adjusting for age, gender, and material deprivation, SOGICE exposure was significantly associated with increased NSSI frequency (OR 1.47; 95% CI 1.03–2.08), and more than two times the odds of planning suicide (OR 2.56; 95% CI 1.74–3.78) and attempting suicide (OR 2.73; 95% CI 1.70–4.39).
Table 2
Percentage of participants reporting SOGICE across demographic groups (N = 3948)
Total response ethnicity |
Māori | 16 (2.8) | 553 (97.2) | χ2 (1) = 0.11, p = 0.737 |
Samoan | 3 (4.1) | 70 (95.9) | χ2 (1) = 0.02, p = 0.883 |
Cook Island Māori | 0 | 41 (100) | χ2 (1) = 0.50, p = 0.481 |
Tongan | 2 (11.8) | 15 (88.2) | χ2 (1) = 1.83, p = 0.176 |
Chinese | 3 (1.9) | 157 (98.1) | χ2 (1) = 0.49, p = 0.486 |
Indian | 5 (4.7) | 101 (95.3) | χ2 (1) = 0.45, p = 0.502 |
Filipino | 1 (1.4) | 72 (98.6) | χ2 (1) = 0.28, p = 0.595 |
Dutch | 3 (5.3) | 54 (94.7) | χ2 (1) = 0.30, p = 0.582 |
British | 6 (6.2) | 91 (93.8) | χ2 (1) = 2.12, p = 0.145 |
Pākehā or New Zealand European | 101 (3.0) | 3212 (97.0) | χ2 (1) = 0.27, p = 0.601 |
Age groupa | | | χ2 (1) = 13.82, p < 0.001 |
14–18 | 40 (2.1) | 1898 (97.9) |
19–26 | 84 (4.2) | 1926 (95.8) |
Gender groups | | | χ2 (3) = 15.83, p = 0.001 |
Cisgender | 38 (2.0) | 1835 (98.0) |
Trans man and trans woman | 25 (4.5) | 534 (95.5) |
Nonbinary and another gender | 43 (4.4) | 943 (95.6) |
Unsure | 18 (3.5) | 503 (96.5) |
Total response Sexuality |
Heterosexual/straight | 2 (5.6) | 34 (94.4) | χ2 (1) = 0.13, p = 0.723 |
Mostly straight | 1 (0.9) | 112 (99.1) | χ2 (1) = 1.26, p = 0.262 |
Takatāpui | 12 (5.1) | 222 (94.9) | χ2 (1) = 2.57, p = 0.109 |
Queer | 66 (3.6) | 1749 (96.4) | χ2 (1) = 2.40, p = 0.121 |
Gay | 31 (3.3) | 895 (96.7) | χ2 (1) = 0.09, p = 0.763 |
Lesbian | 27 (3.5) | 745 (96.5) | χ2 (1) = 0.27, p = 0.606 |
Bisexual | 48 (2.8) | 1663 (97.2) | χ2 (1) = 0.94, p = 0.332 |
Pansexual | 30 (3.2) | 894 (96.8) | χ2 (1) = 0.01, p = 0.920 |
Asexual | 24 (4.2) | 547 (95.8) | χ2 (1) = 2.08, p = 0.150 |
Aromantic | 4 (2.4) | 164 (97.6) | χ2 (1) = 0.12, p = 0.725 |
Demisexual | 12 (3.2) | 364 (96.8) | χ2 (1) = 0.00, p = 1.00 |
Fluid/it changes | 17 (2.9) | 566 (97.1) | χ2 (1) = 0.04, p = 0.833 |
Unsure | 8 (2.6) | 301 (97.4) | χ2 (1) = 0.17, p = 0.681 |
Material Deprivation | | | χ2 (2) = 27.06, p < 0.001 |
No deprivation (0) | 59 (2.7) | 2158 (97.3) |
Mild deprivation (1–4) | 53 (3.3) | 1566 (96.7) |
Severe deprivation (5–9) | 12 (11.9) | 89 (88.1) |
Homelessness | 41 (11.0) | 331(89.0) | χ2 (1) = 82.29, p < 0.001 |
Staturory Care/Oranga Tamariki involvementa | 20 (5.1) | 369 (94.9) | χ2 (1) = 5.05, p = 0.025 |
Table 3
Association between SOGICE and NSSI and suicidality (n = 3948)
SOGICE | 1.58 (1.14–2.21)** | 1.47 (1.03–2.08)* | 1.48 (0.99–2.20) | 1.24 (0.81–1.89) | 2.60 (1.82–3.73)*** | 2.56 (1.74–3.78)*** | 3.12 (2.03–4.80)*** | 2.73 (1.70–4.39)*** |
Participants reporting SOGICE exposure had SOGICE suggested by family/whānau members (55.6%), religious or spiritual leaders (48.4%), myself (17.7%), another person (14.5%), and medical professionals (11.3%). Of the 22 participants who selected “myself,” some also selected religious leaders (
n = 11; 50.0%), family members (
n = 11; 50.0%), medical professionals (
n = 3; 13.6%), or another person (
n = 3; 13.6%), and nine selected “myself” only (41.0%). Table
4 presents the association between SOGICE exposure from each suggester type and NSSI and suicidality. In bivariate models, those who experienced SOGICE suggestion from religious or spiritual leaders had a higher likelihood of reporting a suicide plan and suicide attempt. Similar findings were observed for medical professionals and family/whānau members as SOGICE suggesters. However, significantly higher NSSI frequency and suicide ideation were found amongst participants whose family/whānau members suggested SOGICE to them. In multivariate models that adjusted for age, gender, and material deprivation, those exposed to SOGICE-suggestion from religious or spiritual leaders were significantly more likely to engage in NSSI, suicidal planning, and suicide attempts (OR 3.00; 95% CI 1.54–5.84). Having family/whānau members who suggested SOGICE was also a significant predictor for suicidal planning and at least one suicide attempt (OR 2.27; 95% CI 1.20–4.31).
Table 4
Association between SOGICE suggester and NSSI and suicidality (n = 3948)
Religious leaders | 1.62 (1.00–2.63) | 1.71 (1.03–2.83)* | 1.36 (0.78–2.38) | 1.22 (0.68–2.19) | 2.43 (1.46–4.05)*** | 2.70 (1.56–4.67)*** | 2.98 (1.62–5.48)*** | 3.00 (1.54–5.84)** |
Medical professionals | 2.56 (0.96–6.82) | 1.72 (0.62–4.82) | 3.50 (0.78–15.66) | 2.11 (0.44–10.10) | 3.42 (1.19–9.89)* | 2.47 (0.78–7.82) | 2.62 (0.73–9.43) | 1.21 (0.29–5.02) |
Family/whānau member | 1.63 (1.05–2.53)* | 1.44 (0.91–2.29) | 1.76 (1.01–3.06)* | 1.42 (0.79–2.54) | 2.78 (1.71–4.47)*** | 2.57 (1.53–4.32)*** | 3.03 (1.71–5.36)*** | 2.27 (1.20–4.31)* |
Myself (self-suggested SOGICE) | 1.53 (0.70–3.36) | 1.70 (0.76–3.82) | 1.98 (0.73–5.39) | 1.81 (0.64–5.12) | 2.57 (1.11–5.94)* | 2.92 (1.20–7.08)* | 3.63 (1.41–9.32)** | 3.86 (1.39–10.66)** |
Another person | 1.56 (0.66–3.65) | 1.16 (0.48–2.80) | 1.16 (0.44–3.11) | 0.82 (0.29–2.28) | 2.05 (0.81–5.21) | 1.57 (0.59–4.19) | 2.75 (0.90–8.40) | 2.20 (0.68–7.16) |
Number of categories of suggesters [0–2] - 0, 1, 2, or morec | 1.37 (1.08–1.74)** | 1.29 (1.01–1.66)* | 1.33 (1.00–1.78) | 1.17 (0.86–1.58) | 1.81 (1.40–2.34)*** | 1.77 (1.34–2.33)*** | 1.92 (1.42–2.60)*** | 1.71 (1.22–2.40)** |
Participants who reported self-suggesting SOGICE were separated from those who had been exposed to SOGICE-suggestion by other individuals. In the sample of those exposed to other-suggested SOGICE (n = 114), 61.4% (n = 70) of participants had reported one type of SOGICE-suggester, and 38.6% (n = 44) were exposed to two or more types of SOGICE suggesters. In bivariate models, participants exposed to increasing types of SOGICE suggesters had a higher likelihood of engaging in NSSI and all the suicidality variables that were examined. After adjusting for sociodemographic variables, an increment of exposure to one SOGICE-suggester (on a scale of 0 to 2), on average, was significantly associated with 29% increased odds of NSSI frequency, 77% for a suicidal plan, and 71% for a suicide attempt.
Discussion
Studies on SOGICE, predominantly with adult populations, have demonstrated associations with adverse mental health outcomes and highlight that particular groups may be at higher risk of harm. However, few of these studies have been explicitly designed to explore young people’s experiences. Young people may face additional challenges when exposed to SOGICE, as adolescence is the time when identity development and consolidation are central to positive youth development. The effects of SOGICE that seek to change a key part of young people’s identity at this developmental stage may therefore be particularly harmful. Relationships are fundamental to identity development; however, there is limited research on whether particular categories of people who suggest SOGICE to young people are differentially associated with harm. This study addressed these gaps by collecting mental health and demographic data, including data on who suggested SOGICE to participants. The results confirmed that SOGICE is occurring for an important minority of young people in New Zealand and that it is often very harmful. Mental health impacts did vary according to who suggested SOGICE to the young person, the frequency of categories of people who suggested it, and young people’s membership in particular demographic groups that were associated with increased stigma and distress.
The findings indicate that at least 3.0% of a contemporary youth sample have experienced SOGICE. Studies using alternative measures targeting particular SOGICE behaviors or phenomena have produced higher estimates among Canadian participants (Salway et al.,
2021) and transgender and gender-diverse participants in New Zealand (Veale et al.,
2021). One potential explanation for the relatively low reporting of SOGICE in the present study could be due to the wording of the question around SOGICE. The question in the current study was written to describe experiences and use language that was likely to be recognizable to the local younger age group of the study. However, the question wording may have emphasized direct SOGICE experiences rather than indirect processes or the withholding of services to enforce a change in identity. As a measure, this may have produced a more conservative estimate of prevalence than measures that explicitly include SOGICE through indirect means (e.g., the denial of appropriate health care or services based on gender identity, etc.) (Veale et al.,
2021). In addition, the question in the current study initially required participants to endorse that their experience was “conversion therapy.” Although the wording of the question described a range of “practices,” the initial use of the word therapy may have made it difficult for participants to reconcile a negative experience with something that was purported to be “therapeutic.”. As such, when including participants who reported they were too upset to answer the initial question or who said they preferred not to respond once asked the question (see Appendix Table
5), the prevalence rose to 7.1%, which may be a more accurate estimate of prevalence in a nonprobability youth sample.
The study shows that older youth were more likely to report SOGICE. In this way, the study reflects other research predominantly with adults showing increased age-related prevalence in Hong Kong (Chan et al.,
2022) and South Korea (Lee et al.,
2021) but differs from recent US (Green et al.,
2020) and NZ (Veale et al.,
2021) research showing greater prevalence in younger participants. The age differences in the current study may indicate that older youth may have had more time to understand and recognize their SOGICE experiences, and may therefore be more likely to report such experiences. In addition, older youth may be more willing to answer questions about experiences that may now be historic and potentially less distressing than younger participants who may currently have, or only recently had, SOGICE experiences. The findings show that younger youth were more likely to skip or decline to answer these questions (see Appendix Table
5), which might mean that SOGICE experiences for younger youth are underrepresented in the study. Internationally, more research with youth populations is required to understand these trends and acknowledge that many presently enduring such experiences may be too distressed to answer these items, leading to underreporting among younger youth.
Critically, the results align with findings from international literature that show an association between SOGICE and NSSI and suicidality among adults, as well as among young people aged ≤24 years (Green et al.,
2020). The odds for suicide attempts in the past year were nearly three times for participants who had experienced SOGICE compared with those who had not; this effect size was similar to the findings observed among gender- and sexuality-diverse youth exposed to SOGICE in the USA (Green et al.,
2020). The results are consistent with the hypothesis that the effects of SOGICE are particularly damaging to young people’s mental health. The higher risk of NSSI, suicide planning, and suicide attempts in the past year highlights that experiences of SOGICE that may have occurred earlier in life, even in the relatively short lives of young people, may produce long-term harm. SOGICE practices are associated with life-threatening implications, emphasizing the importance of prohibiting such practices and providing trauma-informed mental health support, including for historical experiences.
The findings show that the prevalence of SOGICE varies by some sociodemographic features and not others. For instance, SOGICE was relatively equally reported by participants across all ethnicities and sexualities but was more likely to be reported by young people who were transgender, non-binary or another gender, or unsure whether they were transgender, than by cisgender young people. The higher rates reported by transgender and gender-diverse young people could indicate that they may have experienced change efforts directed at their gender identity and their sexual orientation, and this may increase the potential for distress. Service provision and future research should recognize those young people who are both gender- and sexuality-diverse may face increased challenges compared to those who occupy only one of these identities.
Additionally, transgender and gender-diverse young people may affirm their non-cisgender identities at an early age (Fast & Olson,
2018) than sexuality-diverse young people do their non-heterosexual identities (Bishop et al.,
2022). In this way, they may face increased exposure to earlier SOGICE than sexuality-diverse young people, and this earlier age of SOGICE exposure, or the increased duration of SOGICE exposure, may also explain the increased associations of SOGICE-related harm for these young people. Future research is required to explore whether the age of recognizing one’s gender- and sexuality-diverse identity, alongside the age of first SOGICE exposure, are related, and potentially associated with mental health outcomes.
Transgender and gender-diverse young people may also be more likely to face SOGICE in healthcare settings than sexuality-diverse young people because accessing gender-affirming care requires the disclosure of a gender-diverse identity. Disclosing a gender-diverse identity may increase the chances that these young people face SOGICE by healthcare practitioners. The study shows that family members are often involved in suggesting SOGICE to transgender and gender-diverse young people. One way that they may do this is to delay, discourage and prevent access to gender-affirming health care. The study underscores that transgender and gender-diverse young people need stronger protections and must have access to unbiased gender-affirming healthcare without cisheteronormative gatekeeping by health care practitioners or parents and guardians. The study emphasizes that transgender and gender-diverse young people must be explicitly named and visible in SOGICE policy, intervention, and practice outcomes to ensure that their unique and increased needs are addressed.
Young people reporting severe deprivation were also more likely to report SOGICE than those reporting less deprivation. While this may be a byproduct of the increased rates of homelessness and statutory care experience reported by those who reported SOGICE experiences, it nonetheless emphasizes the need for accessible and affordable mental health supports for gender- and sexuality-diverse young people and adults. Accessible and affordable mental health care support is critical in this situation, as those reporting severe deprivation may not be able to afford treatment otherwise.
Equally, policies and processes for young people who are homeless or have experienced homelessness or involvement with statutory child services must recognize that SOGICE experiences are statistically more common for this group and respond accordingly. A range of research demonstrates the overrepresentation of gender- and sexuality-diverse young people experiencing homelessness or statutory care (Baams,
2019). SOGICE may be a direct or indirect cause of homelessness or statutory care experience, which means that secular statutory care services may be essential for young people who may have been traumatized by religious SOGICE experiences. The study could not determine whether family members who suggested SOGICE were foster parents or siblings or whether the young people were in statutory care because of such experiences from family members. Given the prevalence of young people in care reporting SOGICE experiences, statutory care workers must be educated on the damaging effects of SOGICE and screened to ensure such practices are avoided in future.
A sensitivity analysis explored whether a similar pattern of sociodemographic characteristics was associated with reluctance to answer the SOGICE items. Appendix Table
5 shows that other than the noted higher prevalence among younger young people to skip the SOGICE section, the other differences were higher rates of skipping by nonbinary young people, Māori, and Cook Island Māori participants. Sample sizes were small in this analysis, so caution is required when drawing comparisons; however, future research is needed to understand why participants from these groups were more likely to skip these questions. It may be that more members of these groups experienced current or recent SOGICE exposure compared to participants from other groups in the study. However, the higher rates of non-binary youth who declined to answer SOGICE questions, alongside the higher rates of non-binary young people who reported SOGICE when they did answer these items, highlights a concerning trend for this group. One possible reason for the overrepresentation of non-binary participants in both sets of these results may reflect the largely transnormative nature of gender-affirming care (Pasley et al.,
2022) that may place non-binary populations at greater risk of SOGICE compared to young people with other gender identities. More research on these populations, as well as continued validation and improved understandings of Māori, Cook Island Māori and non-binary identities, are important opportunities that may improve outcomes.
A wide range of people may suggest SOGICE to a young person, including family members. The study found a strong association between family member-suggested SOGICE and suicidality, which reflects USA research showing that participants whose parents had initiated sexual orientation change efforts had three times the odds of reporting a suicide attempt than those who had not experienced this (Ryan et al.,
2020). The significance of family members’ suggestions of SOGICE may reflect the fact that the majority of young people rely on family members for their survival and wellbeing, including meeting their housing, living and financial needs. Many family members will also meet many of the emotional needs of young people. Therefore, when such family members, who often have a lot of power and influence on young people, suggest SOGICE, it may be more distressing than when other, less powerful and intimately aquatinted adults and peers suggest SOGICE. In contrast, other research has highlighted that families’ positive responses to their child’s gender or sexual orientation can foster positive outcomes in health (Newcomb et al.,
2019), education (Fenaughty et al.,
2019), and self-identity (Katz-Wise et al.,
2016). These effects emphasize the need for family education and support, particularly around the negative impacts of cisheteronormative rejection (American Psychological Association,
2021a) and the reconciliation of religion and LGBTQIA+ identities (VanderWaal et al.,
2017).
SOGICE suggested by religious leaders was also significantly more likely to be associated with higher NSSI and suicide planning and attempts in the past year. The significance of religious leaders’ role in SOGICE-related harm may reflect that such leaders are often highly trusted by young people and their families. When such people, who ostensibly have young people’s best interests at heart, suggest SOGICE, a disproven therapeutic activity (American Psychological Association,
2021b), this results in significantly increased odds of suicidality and NSSI. Prohibiting people in such roles from suggesting or initiating SOGICE is an essential public health response to protect gender- and sexuality-diverse young people from harm. Alongside education for parents and families, policy and practice interventions are required to support the education of religious and spiritual communities about the harm associated with SOGICE to help prevent such experiences from continuing.
People who self-suggest SOGICE may be particularly negative about their diverse gender and sexuality. Some of the increased harm reported for self-suggestion of SOGICE may represent the increased anguish and shame at “failing” not only themselves but also others when the promises of SOGICE are not realized. Understanding young people’s self-suggestion within the context of the cisheteronormative beliefs and norms that influence youth development is critical. Although only a minority of gender- and sexuality-diverse young people self-suggest SOGICE, such decisions are hazardous and associated with the most harm of any group in the study. Prohibiting SOGICE experiences, even when self-suggested, is a critical public health recommendation from this study, and young people who seek out such experiences are an urgent priority for mental health and support services. The increased risk of NSSI and suicidality for young people who self-suggest SOGICE also strengthens and broadens the recommendations for education and support on this issue. A public health response requires that all young people, irrespective of their gender or sexuality, receive comprehensive education about gender and sexuality diversity to counter destructive norms that devalue and stigmatize their identities (Fenaughty,
2019).
Half of the young people who self-suggested SOGICE also reported that another person had suggested SOGICE to them, indicating that others’ attitudes and beliefs often inform self-suggestion decisions. Further, Table
4 shows that harm increased as the frequencies of categories of suggesters increased. The increased risk associated with multiple types of SOGICE-suggesters underscores the power of others to influence or mitigate SOGICE self-suggestion and initiation. Further studies are required to determine the nuances within the relationship between the number of SOGICE-suggester types and NSSI and suicidality risks; nonetheless, mental health services are advised to screen for the frequency of types of suggesters of SOGICE and to consider the increased risk of harm for young people who report more than one type of suggester.
Sample bias is one of the crucial limitations of a nonprobability sample. Recruitment for the study relied on the internet and social media, as well as regional libraries, mass media stories, and posters in schools and tertiary education providers. The call to participate in the research was also widely shared through rainbow community networks and media. Young people connected to rainbow communities and media may therefore have been more likely to see the call to participate. Such young people may differ from those not connected to rainbow communities and media, as they may have more rainbow-friendly social connections and supports, which may operate as protective factors for their mental health. The greater concentration of more-connected participants in the study means the data may underestimate the adverse effects of SOGICE experiences because it cannot account for those who have fewer connections and, therefore, less supports. Young people currently undergoing SOGICE may also have been less likely to be engaging with rainbow communities and media, and they may have been less likely to see the call to participate, resulting in a potential underestimate of the prevalence of SOGICE. Nonetheless, a key strength is the large sample size, which enabled comparisons within the sample to tease out effects while accounting for bias to the greatest degree possible.
The research literature and professional guidelines emphasize that SOGICE is ineffective and harmful (American Psychological Association,
2021a). It is doubtful that a large population of “converted” young people who are satisfied with their SOGICE experience exist or report positive mental health outcomes due to SOGICE. The literature demonstrates that SOGICE is associated with adverse short- and long-term mental health outcomes, including suicidality. Extrapolating these mental health findings means that many young people with SOGICE exposure may experience debilitating mental health outcomes and may not be able, or even alive, to participate in research on SOGICE exposure. As such, all cross-sectional SOGICE survey findings are more likely to underestimate the prevalence and adverse effects of SOGICE.
The question wording may have inadvertently created barriers to young people reporting a negative “therapy” experience. A question focusing on practices and withholding services may generate a higher prevalence of SOGICE, including by health practitioners, who were less implicated than in other studies. While the adverse effects of SOGICE in the study were comparable to international studies, the prevalence was lower. For these reasons, the prevalence data on young people who said they were too upset to answer the SOGICE item or said they preferred not to answer further questions on SOGICE after the initial question is included (see Appendix Table
5). The combined proportion of participants who requested to skip this question due to being upset, alongside those who did report SOGICE, may better represent those who experienced SOGICE.
As a cross-sectional study, the findings are correlational and do not prove causality; however, these findings are consistent with the literature detailing the harmful effects of SOGICE among mainly adult participants, as well as a large non-probability youth sample in the US (Green et al.,
2020), and qualitative findings (Kinitz et al.,
2021). Furthermore, the outcome measures of NSSI, suicide plans and suicide attempts were limited to the past 12 months. It is implausible that all of the SOGICE practices in this study happened in the past year alone, and these data suggest the potential for long-term harm that occurs after SOGICE exposure in young people. Further research is required to understand how SOGICE is associated with mental health problems over time, which might explore whether there are compounding effects of SOGICE over time.
The analysis presented focuses on gender- and sexuality-diversity but not on variations in sex characteristics. While intersex young people were included in the analysis, the sample size was too small to separately analyze young people with variations in sex characteristics. Further research is required to explore key factors for young people with variations in sex characteristics in relation to SOGICE.