Introduction

Growing public attention has been paid to children whose gender identity differs from that normatively expected of their natally assigned sex (referred to within this paper as gender variant children), such as in conversations in the mainstream media and social media outlets (see for example, Parry 2013 for a discussion of the DSM5 and blogs such as Raising My Rainbow). With this attention comes a demand upon health care professionals, educational services and the child’s parents to develop appropriate responses. With regard to the latter group (i.e., parents), whilst they may likely experience conflicting advice over how best to respond to their child’s gender variance, there is currently little empirical research which examines how parents experience their child’s gender variance and what requirements they have in terms of support for themselves and their children.

One reason for examining the experiences of parents is the diversity within academic thinking on the topic of gender variance amongst children. Indeed, within the academic literature, advice to both parents and health care practitioners concerning how to respond to gender variant children primarily advocates a highly cautious approach. This is best exemplified in the work of Zucker and Bradley (1995), which emphasized a reparative approach to gender variance that situates it as a disorder that may correct itself over time. Reparative approaches typically advocate that children should ‘return’ to their birth gender, and be actively taught and presented with gender norms by their parents in order to facilitate this return (Zucker and Bradley 1995). Arguably, such an approach may be useful for parents struggling to accept their child’s gender variance, for whom it may be easiest or most in line with their values to teach their child gender normative behaviours. This approach is also more in line with ongoing societal stigma related to gender variance, in which gender norms dictate particular behaviours for boys as compared with girls, with any difference from these norms typically considered to be problematic (Carroll et al. 2002; Drescher 2002; Pearlman 2006). However, whilst congruent with both some parent’s and broader society’s attitudes towards gender conformity, reparative approaches are likely to be experienced as unsupportive by both parents who are accepting of their child’s gender variance, and by gender variant children themselves (see Ehrensaft 2012).

By contrast, an affirmative approach, such as that advocated by Hill et al. (2010), encourages practitioners to support gender variant children, and to do so by working with parents to accept and support their child’s self-identified gender (Malpas 2011; Menvielle and Hill 2010). As such, affirmative approaches encourage parents to acknowledge their child’s gender identity as they experience it, and act as an advocate for their child. In addition to taking this view of gender variance, advocates of affirmative approaches such as Brill and Pepper (2008) have suggested that interventions need to provide parents with the skills and resources to support their child, while also assisting parents to come to terms with any grief they may feel themselves (Brill and Pepper 2008; Menvielle and Hill 2010; Wyss 2013).

As evident from the differences between these two approaches, there is currently little consistent information provided to parents in relation to both the best medical or psychological approach to take with their child, or how to view their child’s gender variance (Dreger 2009). This is problematic given that support for parents of gender variant children is crucial to ensure best outcomes for gender variant children. This is particularly so given the ongoing stigma attached to gender variance in society more widely (Carroll et al. 2002; Drescher 2002; Pearlman 2006). In terms of parent’s responses to such stigma, Saeger’s (2006) case study with a gender variant child and his lesbian mothers indicated that it is possible that parents experience particular anxiety in relation to the inference that it is their own poor parenting which ‘caused’ their child to behave in a non-normative manner. Such stigma may lead to a range of conflicting emotions in parents, and raises questions of how best to communicate and support children. Correspondingly, Wren (2002) suggested that supporting parents in this area is crucial, as promoting a “reflective self-awareness” within parents can lead to a similar awareness in children, and correspondingly foster secure attachments (p. 392).

In terms of the specific experiences of parents in relation to a gender variant child, Saeger’s (2006) case study indicated that parents and extended family can be conceptualized as going through stages of acceptance, including: discovery, turmoil, decision-making, and finding balance. Saeger suggested that if families are able to reach the final stages of decision-making and finding balance, then they are able to not only find satisfaction in supporting their child to live as their self-identified gender, but also advocate successfully for their gender variant child. This stage-based conceptualization indicates the importance of support from extended family for families of gender variant children, and correspondingly Saeger suggested that support from extended family can both assist families in relation to advocacy and promote resilience amongst gender variant children.

In addition to the usefulness of support from extended family, it is likely that parents will also require a degree of support from healthcare professionals (for both themselves and their children). For example, parents are frequently required to make a number of medical choices in relation to their children, particularly in the case of pre-pubescent children (such as whether or not to take a medical approach to working with their children, including whether or not to give their children hormone blockers; see Riley et al. 2011), and some healthcare routes are only possible if children have formal diagnoses related to gender variance. Yet despite the importance of engaging with professionals, there is only a small body of empirical literature that examines parent’s interactions with health care or support professionals. What literature there is suggests that interactions with healthcare professionals can be difficult to negotiate, and that parents may find it hard not only to seek professional assistance that they perceive as supportive, but also to find professional assistance in the first place. For example, Meadow’s (2011) interview study with parents of gender variant children indicated that the participants often struggled to identify appropriate resources to support themselves and their children, and further that they were unsure of whether to seek medical or psychological support for their children in the first place. Where professionals have been located and used by parents, survey research by Riley et al. (2011) found that both schools and health care professionals could be a source of stress for parents, in terms of a lack of both the availability of support and an understanding of the child’s gender identity.

Further in relation to understanding gender variant children’s identity, research by Ansara (2010) demonstrated that pronoun use can be particularly difficult for parents (especially male parents), and can therefore pose issues in relation to support for children (that is, through referring to children with the pronoun associated with their natally assigned sex rather than their current gender identity). Research by Kane (2006) found that male parents find issues related to pronoun use more challenging than do female parents, which suggests that heterosexual male parents in particular struggle with gender non-conforming children (in Kane’s research this term referred to children who may or may not be gender variant but who exhibit preferences for aspects of behaviours and clothing typically associated with a gender different to that normatively expected of their natally assigned sex). Kane’s research found that this struggle is particularly salient in relation to young male assigned children, where there is a smaller range of societally acceptable gender non-conforming behaviours available. These difficulties surrounding pronoun use and acceptance of gender non-conforming behaviours can be exacerbated by health care professionals if they also insist upon using pronouns applicable to the gender variant child’s assigned sex, rather than their self-identified gender.

As this overview suggests, whilst previous research on gender variant children and their parents is divided in terms of the approach to take to gender variance in the first place (see Dreger 2009), research that emphasizes an affirming approach highlights the need for adequate support of both parents and children from professionals and support networks. More research is therefore required in order to understand the experiences that parents have of these forms of support. The aim of the current research was therefore to conduct a scoping survey in the Australian context to identify the attitudes and experiences of parents of gender variant children in relation to support, including the support that parents felt they had in relation to the healthcare needs of their children.

Method

Participants

Participants were parents who self-identified as raising at least one gender variant child, and were sourced via the first author’s existing networks and snowball sampling (such as through email lists and social media sites). Sixty-one heterosexual participants who felt their child met the inclusion criteria outlined below completed the survey. The majority (90.5 %) of participants identified as cisgender females (that is, people whose gender identity normatively accords with that expected of their natally assigned sex), with the remainder identifying as cisgender males (9.5 %). The majority of participants were in heterosexual relationships (90.5 %), with the remainder stating that they were not currently in a relationship (9.5 %). In terms of state of residence within Australia, 21 participants lived in Queensland, 15 lived in Victoria, 15 lived in New South Wales, nine lived in South Australia, and three lived in Tasmania. The average number of children within each family was 2.5 (SD = 1.05). Each family only had one gender variant child.

Of the gender variant children about whom parent participants completed the survey, 52.4 % were identified by their parent as being male assigned at birth but now identifying as female, and 47.6 % were identified by their parent as being female assigned at birth but now identifying as male. The average age of these children was 10.33 years (SD = 4.08), with the range being 4–18 years. In terms of a diagnosis, 71.4 % indicated that their child had received a formal diagnosis of gender identity disorder, and 28.6 % indicated that their child had not received a formal diagnosis. (It should be noted that whilst the inclusion criteria for participation in the survey utilized the DSM5 diagnosis of gender dysphoria, all participants had received a diagnosis for their child when the DSM IV-TR was still in use, in which gender identity disorder was the diagnosis). There was no significant difference between having had a diagnosis or not in regards to age of the child.

Procedure

This research was granted ethics approval by the Social and Behavioural Research Ethics Committee at the first author’s institution. Data were collected via an online survey administered through Survey Monkey. For the purposes of the survey, a modified version of the DSM5 diagnostic criteria for gender dysphoria in children was included on the opening screen, to indicate how gender variance was defined for the purposes of the survey. Emphasis was placed upon the gender variant child—about whom the parent participant completed the survey—meeting criterion A1 (‘A strong desire to be of the other gender or an insistence that he or she is the other gender’), as well as at least one of the other criteria (that is; ‘in children who are assigned as male at birth, preference for wearing stereotypically female attire; in children who are assigned as female at birth, an insistence on wearing only stereotypically masculine clothing’, ‘a strong and persistent preference for cross-sex roles in make-believe play or persistent fantasies of being the opposite sex to that assigned at birth’, ‘an intense desire to participate in the stereotypical games and pastimes of the opposite sex to that assigned at birth’, and ‘a strong dislike of their sexual anatomy or a desire for anatomy that accords with their own sense of gender identity’). This modified version was utilized given the purpose was not to diagnose gender dysphoria, but rather to identify the experiences of parents and their children who fall within a spectrum of gender variance. Once participants consented to participate in the study and indicated that they had at least one child who met these criteria, they were taken to the remainder of the survey, which they completed online.

Measures

The survey was designed by the authors. In addition to asking general demographic questions as outlined above, the survey utilized forced choice, Likert scale, and open-ended response options. Forced choice options involved questions about formal diagnosis, and access to legal and health care professionals. For example, ‘have you accessed a health care professional to help you understand your child’s gender variant identity?’ These forced choice questions were then followed up with open-ended questions to give participants an opportunity to provide further details about their experiences. For example, if participants answered ‘yes’ to the answer above, they were asked ‘please tell us about any positive experiences you had with these health professionals, or any benefits you or your child experienced as a result of this contact?’, and ‘Please tell us about any negative experiences you had with these health professionals, or any challenges you or your child experienced as a result of this contact?’

Likert scale questions required participants to rate the degree of support they felt towards their gender variant child both in the present and 3 years ago, similarly in regards to their perceptions of the support offered by their partner and other children in the family. The point of 3 years was chosen as the authors were keen to include parents of young children (for example, including children aged 4 or 5 years’ old) in the survey, and it was necessary that parents were able to respond to this question on the basis of children who displayed gender variant behaviours 3 years earlier (for example, by expressing a preference for dressing in their preferred gender from 1 year old). Likert scales were also used to ask about support provided by extended family members, schools, and health care professionals. All such scales utilized the options 1 = not supportive, 2 = somewhat supportive, 3 = quite supportive, and 4 = very supportive. Participants were also asked to indicate their beliefs about what determines gender expression in response to two statements: ‘Gender is biologically-determined’ and ‘Gender is a product of upbringing’. Each of these utilized a seven-point scale, where 1 = strongly disagree, 4 = neutral, and 7 = strongly agree. Open-ended responses included questions asking participants to explain behaviours they considered to exemplify their child’s gender variance, their experiences with health care professionals, their own feelings about their child’s gender variance, and their partner’s responses to their child’s gender variance.

Data Analysis

Likert and forced choice responses were entered into SPSS 17.0, which was used to conduct all descriptive and inferential analyses. Open-ended responses were entered into NVIVO for identification of trends within each response group. A sample of each set of open-ended responses is provided in the analysis below, differentiated according to the variables identified as salient within the NVIVO analysis. Non-parametric tests were conducted on any categorical differences in open-ended responses. The examples included from each set were selected on the basis of their representativeness of the broader trends across responses within the set.

Results

Support Within the Immediate Family

Participants whose child had a formal diagnosis of gender identity disorder reported they were more supportive of their child’s gender variance (M = 3.80, SD = .40) than those whose child did not have a formal diagnosis (M = 3.10, SD = .62), t = 2.456, p < .01. Similarly, participants whose child had a formal diagnosis reported that their partner was more supportive of their child’s gender variance (M = 3.79, SD = .42) than those whose child did not have a formal diagnosis (M = 1.22, SD = .50), t = 4.537, p < .001. There was no significant relationship between having had a formal diagnosis of gender identity disorder and support from cisgender children toward their gender variant sibling.

Participants rated the amount of support they felt at present towards their gender variant child as higher (M = 3.88, SD = .05) than it was 3 years ago (M = 3.11, SD = .14), t = 5.503, p < .001. Participants rated the amount of support their partner presently showed towards their gender variant child as higher (M = 3.71, SD = .07) than it was 3 years ago (M = 2.57, SD = .18), t = 7.394, p < .001. Participants rated the amount of support their cisgender children presently showed towards their gender variant sibling as higher (M = 3.40, SD = .20) than it was 3 years ago (M = 2.80, SD = 1.09), t = 3.525, p < .001.

Male participants rated their female partners as more supportive of their child’s gender variance (M = 4.00, SD = .00) than did female participants rate their male partner’s degree of support (M = 3.23, SD = .63), t = 2.181, p < .05.

Using the mean age of gender variant children as a mid way point, children were divided into two groups (those aged nine or under and those aged 10 or above). The siblings of children in the older age category were reported to be more supportive of their gender variant sibling (M = 3.71, SD = .71) than were siblings of children in the younger age category (M = 3.12, SD = .77), t = 2.050, p < .05. No significant relationship was found between the two age categories and support from either participants or their accounts of the supportiveness of their partner.

In terms of participant’s accounts of what they believed constituted examples of their child’s gender variance, Table 1 outlines a sample of these accounts, differentiated by the natally assigned sex of the child and sex of the participant.

Table 1 Children’s behaviours reported by participants as indicating their child’s gender variance
Full size table

As outlined in Table 1, most participants identified that their child wished to dress as their preferred gender identity. In addition, a number of participants also identified that their child expressed a desire to be their preferred gender, and that they spoke about concerns related to developmental changes to their body.

Interestingly, a trend was apparent such that male participants appeared more frequently to use pronouns applicable to assigned sex rather than the child’s own expressed gender. The data were re-coded using two binary categorical variables; male or female participant and pronoun relevant to assigned sex or pronoun relevant to child’s expressed gender identity. A log-likelihood ratio test indicated significant differences: Λ (2, N = 61) = 12.40, p < .05. Participants who were male were significantly more likely to use the pronoun applicable to their child’s natally assigned sex than would be expected in an even distribution. This pattern of male participants being more likely to use pronouns applicable to their child’s natally assigned sex appeared also in participant’s accounts of how supportive they felt towards their child’s gender variance, Λ (2, N = 61) = 15.23, p < .05, as highlighted in Table 2.

Table 2 Participant’s accounts of how supportive they felt towards their gender variant child
Full size table

Evident in the quotes included in this table is also a difference in relation to how participants spoke about their acceptance of their child’s gender identity. For example, some participants spoke of their child’s gender variance as “painful” and as something which required “acceptance”, while others used language that reflected a belief that their child simply “was” their preferred gender, and accounted for the child’s gender variance as being a defect that simply needed to be corrected. The differences seen in these responses are reflected in the section below concerning beliefs about gender.

The afore-mentioned relationship between pronoun use and sex of parent was further highlighted in open-ended responses in which female participants spoke about their male partner’s responses to their child’s gender variance, as outlined in Table 3.

Table 3 Participant’s accounts of their partner’s responses to their gender variant child
Full size table

The statements included in Table 3 indicate that female participants felt that their male partners were slower to accept their child’s gender variance than they were themselves (reflecting the quantitative findings reported above). However, most female participants indicated that despite this, their male partners were currently accepting of their child’s gender variance.

Looking across all open-ended responses, a trend was identified where male participants were more likely to use pronouns related to their child’s assigned sex where the child was male-assigned at birth. To examine this, the data were re-coded using three binary categorical variables: male or female participant, male assigned child female pronoun or male assigned child male pronoun, and female assigned child female pronoun or female assigned child male pronoun. Log-likelihood ratio tests only identified a significant relationship between the first two variables, where male participants were more likely to use a masculine pronoun to refer to their male assigned child, Λ (2, N = 61) = 9.76, p < .05.

Support from Outside the Immediate Family

Participants whose child had a formal diagnosis of gender identity disorder reported higher levels of support from schools (M = 3.22, SD = .45) than those who child did not have a formal diagnosis (M = 1.77, SD = .56), t = 3.712, p < .01. Similarly, participants whose child had a formal diagnosis of gender identity disorder reported higher levels of support from extended family members (M = 3.28, SD = .68) than those who child did not have a formal diagnosis (M = 1.80, SD = .42), t = 4.661, p < .001.

Parents whose child was in the younger age category reported lower levels of support from extended family members (M = 2.71, SD = .95) than did parents whose child was in the older age category (M = 3.52, SD = .65), t = 3.275, p < .01.

In regards to non-familial support, 47 % of participants had accessed legal support in regards to their gender variant child and 52.4 % had not. Children of parents who had accessed legal support were older (M = 12.20, SD = 2.94) than were children of parents who had not accessed legal support (M = 8.63, SD = 4.20), t = 3.861, p < .001. 76.2 % of participants had accessed a health care professional in regards to their gender variant child, and 23.8 % had not. Again, children of parents who had accessed a health care professionals were older (M = 11.12, SD = 3.60) than were children of parents who had not accessed a health care professional (M = 7.80, SD = 4.45), t = 2.943, p < .01. Parents indicated a range of health care professionals they had engaged with, including counselors, psychologists, psychiatrists, general practitioners, endocrinologists, pediatricians, mental health nurses and specialist gender clinics. Table 4 outlines open-ended responses about health care professionals, differentiated by positive responses and negative responses.

Table 4 Experiences with healthcare professionals
Full size table

As indicated by statements included in Table 4, positive experiences of healthcare professionals included experiences where being taken seriously by a professional took the weight off the shoulders of participants and their children, making them feel more like someone believed them and “was listening”. As a direct corollary to this, negative experiences were those in which professionals did not offer support, but instead were judgmental and placed the ‘blame’ on the shoulders of parents [reflecting Saeger’s (2006) findings concerning the impact of this blame on parents].

Both positive and negative experiences appeared to be uniformly shared across both the sex of participants, and the assigned sex of the child, hence no differentiation is provided. However there was a relationship between level of support received from health care professionals and parents’ self-reported support for their gender variant child, r = .453, p < .01.

Beliefs About Gender

In terms of beliefs about gender, on average participants indicated that they were more likely to hold the belief that gender is biologically determined (M = 5.95, SD = 1.70) than they were to hold the belief that gender is determined by upbringing (M = 1.20, SD = 1.22), r = −.345, p < .05.

Parents of children female assigned at birth were more likely to believe that gender is biologically determined (M = 6.30, SD = 1.34) than were parents of children male assigned at birth (M = 4.34, SD = 1.75), t = 2.918, p < .01.

Parents of older children were more likely to believe that gender is biologically determined, r = .496, p < .001. Parents of older children were less likely to believe that gender is determined by nurture, r = −.569, p < .001.

Discussion

The findings presented in this paper both affirm and extend the small body of previous research undertaken with parents of gender variant children. In relation to the findings concerning acceptance, the results of the present study reinforce previous research suggesting that a parent’s acceptance of a child’s gender variance grows over time, potentially reflecting the stages to acceptance proposed by Saeger (2006). However, the level of acceptance indicated by participants in this study was to some degree dependent on the parent’s own gender (with male parents showing more resistance to adopting the child’s self-identified gender pronouns). This lower level of acceptance from male parents was also reflected in the fact that females rated their male partner’s acceptance levels as lower than male parents did their female partner’s acceptance levels, and also in the qualitative findings of the use of pronouns (with male parents more frequently using the pronoun associated with their assigned sex). In addition, qualitative data demonstrated that female parents more frequently used pronouns that reflected their child’s preferred gender identity, while male parents often used pronouns normatively associated with their child’s natally assigned sex, particularly for children assigned male at birth. These findings also support previous research by Kane (2006), who found that the male parents in her sample struggled more with pronoun use and gender non-conforming behviours than did female parents. As such, our research suggests that similar patterns are found to those seen in Kane’s research even where children have a formal diagnosis of gender variance. This finding also supports suggestions in the literature that heterosexual male parents frequently leave the emotional work of caring for gender variant children to female parents (Wren 2002), and this area warrants further research in terms of providing support to male parents of gender variant children.

Participants in this study also indicated that experiences with healthcare professionals were highly variable. Specifically, open-ended responses appeared to indicate that positive responses from health care professionals were experienced as affirming for both children and parents, whilst negative responses appeared to be based on gender normative judgments about child development. These findings also affirm previous research suggesting that health care professionals may play either a supportive or marginalizing role in the care of gender variant children (Mallon 2000). However, despite this variability the qualitative data also demonstrated that support from healthcare professionals was highly valued by participants. Given the emphasis in the DSM5 on gender dysphoria as a non-pathologising account of gender variance, these results suggest that at least some Australian health care professionals are in need of updating their knowledge about gender variance in order to provide adequate support to families.

While the results indicated that male parents in particular struggled with pronoun use, the quantitative data nevertheless found that most parents were supportive of their gender variant child, and that this was particularly the case where the child had received a formal diagnosis. This finding was also reflected in relation to schools, extended family, and siblings, with participants whose child had a formal diagnosis indicating higher levels of support from other people or institutions than parents of children who were not formally diagnosed. As such, it is likely that experiences of gender variance will remain mediated to at least some degree by the assessments of health care professionals (Riggs et al. 2014). It is not unreasonable to suggest that, given general public faith in health care professionals, diagnoses provided by such professionals may serve to encourage support for gender variant children. Correspondingly, whilst some authors have encouraged a move away from a medical or psychological model of gender variance in children (see (Ehrensaft 2012), who advocates for the use of diagnosis only in cases where children exhibit stress or distress), the findings of the current study would suggest that there may be considerable utility in applying a diagnosis with the aim of engendering support and understanding.

Finally, the research also found that the majority of parents in this sample felt that gender was biologically determined rather than determined by upbringing. This finding was particularly relevant in relation to children female assigned at birth and older children. These findings add to previous literature in relation to attitudes of parents concerning gender, and particularly in relation to the findings of Kane (2006) who similarly found that parents commonly expressed biological explanations for gender. However, Kane also found that despite attitudes concerning biological explanations for gender, parents of young male assigned children expressed feeling some responsibility for their child’s masculinity, indicating that they felt it was something they could “craft” (p. 172). Kane’s findings may therefore help explain the results of the current study, which suggest that parents with children female assigned at birth were more likely to indicate that they felt that gender was biologically determined. As such, it is possible that parents with children assigned male at birth may have felt more ‘responsible’ for shaping their child’s masculinity, consistent with attitudes relating to gender being determined by upbringing. These findings are interesting given that Kane suggests that if parents are consciously crafting their child’s gender identity in relation to masculinity, then that conscious effort could instead be channeled to support (rather than hinder) non-normatively gendered behaviours in children (and particularly those assigned male at birth), and this is an area which warrants further research.

In terms of limitations, the findings reported here are limited primarily by sample size. Importantly, however, of the two Australian social support groups facilitated by parents of gender variant children, one indicated its membership to the first author as approximately 70. Given there would likely be overlap between the two groups, a sample of 61 may be taken as broadly indicative of the population. The survey itself was limited by using single items to assess individual variables, and thus future research may focus on further developing or adapting scales to better capture the support experiences and needs of gender variant children and their parents. In addition, the study’s findings are limited to the experiences of parents of gender variant children. Future research could usefully explore the issue of support for parents from a variety of perspectives, including healthcare professionals and extended family members.

Given ongoing debates over the role of health care professionals in the lives of gender variant children and their families, the findings presented here would appear to clearly demonstrate the need for services that support parents to understand their child’s gender variance, and to affirm their child’s gender identity as they experience it.