|Year : 2021 | Volume
| Issue : 1 | Page : 9-14
The associations of genital-normalizing surgery and assigned gender in predicting gender outcomes: A pooled nested case study analysis of 282 adults with differences of sex development in 58 academic articles
Jaimie F Veale
School of Psychology, University of Waikato, Hamilton, New Zealand
|Date of Submission||21-May-2020|
|Date of Decision||13-Aug-2020|
|Date of Acceptance||17-Sep-2020|
|Date of Web Publication||27-Mar-2021|
Jaimie F Veale
School of Psychology, Private Bag 3105, Hamilton 3240
Source of Support: None, Conflict of Interest: None
Purpose: The purpose was to systematically analyze the associations of childhood gender assignment and genital normalizing surgeries with adult gender in case reports of adults with differences of sex development (DSD). We did this using a pooled nested analysis of DSD adult cases reported in the academic literature. Materials and Methods: A search of academic databases uncovered 282 adult DSD cases reported in 58 academic articles that met the criteria for inclusion in the analysis. These were cases with 46, XX congenital adrenal hyperplasia, partial androgen insensitivity, mixed or partial gonadal dysgenesis, and micropenis where data about the assigned gender, childhood genital-normalizing surgery, and adult gender could be extracted. We conducted a generalized logistic mixed-model regression analysis, with multiple predictors of adult assigned gender incongruence. Results: Controlling for assigned gender, age, year of article publication, and DSD type, childhood genital surgery was not significantly associated with adult assigned gender incongruence. Cases assigned female had more than five times greater likelihood of assigned gender incongruence (95% confidence interval = 1.96–14.92). Conclusion: This study did not find evidence that childhood genital surgery is related to assigned gender incongruence; it also found that assigning a child as a female increased this possibility. These findings may inform decisions taken by clinicians and family members about the early management of DSD cases. The study was limited by publication bias in the types of cases that get published. Ethical considerations should always take precedence in decision-making regarding these surgeries.
Keywords: 46, XX differences of sex development, 46, XY differences of sex development, atypical sex characteristics, sex reassignment surgery
|How to cite this article:|
Veale JF. The associations of genital-normalizing surgery and assigned gender in predicting gender outcomes: A pooled nested case study analysis of 282 adults with differences of sex development in 58 academic articles. Urol Sci 2021;32:9-14
|How to cite this URL:|
Veale JF. The associations of genital-normalizing surgery and assigned gender in predicting gender outcomes: A pooled nested case study analysis of 282 adults with differences of sex development in 58 academic articles. Urol Sci [serial online] 2021 [cited 2021 Apr 19];32:9-14. Available from: https://www.e-urol-sci.com/text.asp?2021/32/1/9/312434
| Introduction|| |
People born with sex characteristics that are not clearly of male or female are known as having differences of sex development (DSDs) within the medical community. Intersex is the more commonly used term in the general community. DSDs can be associated with chromosomal differences (e.g., 45, X/46, XY mixed gonadal dysgenesis) and differences in hormone production (e.g., 46, XX congenital adrenal hyperplasia [CAH]) or synthesis (e.g., partial androgen insensitivity syndrome [PAIS]).
The optimal gender of rearing policy developed at Johns Hopkins University in the 1950s and 1960s required that DSD infants and children be assigned as a male or a female as early in life as possible, along with genital surgery to alter the appearance of their genitals to be more congruent with what was expected of this assigned gender. Money et al., who developed the policy, believed that having genitals that conform to the society's norms was necessary for gender identification and to help parents be committed to raising the child in that assigned gender.,,,, Conventionally, there had been a preference toward assigning DSD children as girls because feminizing genitoplasty is generally more technically straightforward than surgical phalloplasty.,,
Although the stability of gender identity is no longer explicitly given as a reason for these surgeries, there is evidence that decisions about surgery and gender assignment based on the optimal gender of rearing policy continue.,
Genital normalizing surgeries are controversial. Many DSD people are extremely unhappy that they have received surgeries without their consent, noting that receiving these surgeries has impacted on their rights to bodily autonomy, fertility, and sexual pleasure., United Nations experts and human rights committees have called for an end to them, and there have been moves in some global jurisdictions to legally restrict these surgeries. Despite this, these genital normalizing surgeries continue.,,
In 2014, Diamond and Garland noted that there had not been any research to support the belief that DSD children require early genital surgery to avoid assigned gender incongruence, which occurs when a person's gender does not align with the gender that they were assigned. We are not aware of any research that has quantitatively examined the predictors of assigned gender congruence or incongruence among people with DSDs. The purpose of this article is to explore whether genital-normalizing surgery, assigned gender, and DSD type predict gender outcomes through a pooled nested case study analysis of the available published case studies.
| Methods|| |
Literature search strategy
We conducted searches of PubMed and Google Scholar using keywords such as “intersex,” “disorders of sex development,” “congenital adrenal hyperplasia,” “androgen insensitivity,” and “gonadal dysgenesis” to identify DSD case reports. We also identified cases reported in two seminal review articles, and searched these databases for studies citing these articles. We relied only on sources from 1960 onward because we could not say with certainty the reasons for the atypical sex characteristics (DSD type) for many of the cases reported before this time.
We extracted data about cases that were born with externally visible sex characteristics, which were not typical of males or females, that is, we included cases of CAH, PAIS, mixed or partial gonadal dysgenesis (MGD), or micropenis. We limited our inclusion to these types of DSD because these are the cases where surgery is usually performed on specifically to normalize genital appearance. We also confined our criteria to adult cases aged 18 years or older to allow time for gender development. Moreover, articles had to state whether or not genital-normalizing surgery had been conducted on the case; we coded cases as having undergone childhood genital-normalizing surgery if this surgery was conducted at age 5 or earlier. We chose this age because it would be less likely that the child's gendered behavior would influence the decision to undergo surgery at this young age and because the optimal gender policy required surgical intervention as early as possible.
We extracted data from articles about whether each case had received childhood genital-normalizing surgery and what their assigned gender was. Cases were also coded as assigned gender incongruent if the article reported that the case's gender identity or the gender they were living in was not their assigned gender. In all cases, only binary (male or female) adult genders and assigned genders were reported in these articles. We also extracted data about the age of the case and the year of the article's publication because we believed these may also be associated with assigned gender incongruence.
A total of 282 cases were identified from 58 academic articles. These sources and summary details of the cases are outlined in [Supplementary Table 1], and the full dataset is available at https://doi.org/10.6084/m9.figshare.11390277.v1.
We analyzed data using IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp generalized mixed model function. Data from cases extracted were pooled and this meant that we had data that were structured with two levels: cases nested within articles. We calculated an intraclass correlation coefficient to assess the extent that assigned gender incongruence varied within and between clusters. Details of these results are given in the supplementary online materials, but these analyses suggested that there were substantial differences between articles in the chances of assigned gender incongruence, meaning that the use of mixed (also called hierarchical or multilevel) analysis was warranted to control for the effects of any shared variance among cases within each article.
For the main analyses in this article, we conducted a binary logistic mixed regression model with childhood genital surgery, age, assigned gender, DSD type, and year of publication predicting assigned gender incongruence. Random-effect intercept variance was estimated in the model. To reduce the effect of collinearity of predictors, continuous predictors (age and year of publication) were mean centered.
| Results|| |
Characteristics of cases identified
Cases included CAH (n = 91, 32%), micropenis (n = 78, 28%), PAIS (n = 70, 25%), and MGD (n = 43, 15%) DSD types. Age data could be extracted for 244 cases (87%). Ages ranged from 18 to 69 years, with a mean of 28.07, a standard deviation of 9.74, and a median of 25.00. Individual age data were not given for 38 cases because the article reported ages for a group instead of individuals (and all cases in the group were aged 18 or older). For these cases, we imputed ages with the mean age that was reported by the article for that group of cases.
A total of 74 out of 282 cases (26%) had undergone childhood genital-normalizing surgery; 23 out of 153 assigned male cases (15%) received masculinizing genital surgery and 51 out of 129 assigned female cases (40%) received feminizing genital surgery. In total, there were 49 cases (17%) that had assigned gender incongruence. There were similar proportions of assigned gender incongruence among those who had received childhood genital surgery (14 out of 74, 19%) and those who had not received childhood genital surgery (35 out of 208, 17%). A higher proportion of cases that were assigned female (38 out of 129, 30%) had assigned gender incongruence than those that were assigned male (11 out of 153, 7%). These descriptive findings, however, do not account for the fact that many articles reported more than one case, and cases reported within each article may be more similar to cases reported across different articles. We, therefore, conducted a generalized mixed regression model to statistically account for cases being nested within each article from which they were sourced.
Regression analysis of predictors of assigned gender incongruence
The results of the generalized mixed regression model are outlined in [Table 1]. This model shows that after accounting for other predictors in the model, childhood genital surgery, DSD type, and year of publication were not statistically significantly associated with adult assigned gender incongruence. Being assigned female was the strongest predictor of assigned gender incongruence, with the model's odds ratio indicating that cases assigned female had more than five times greater odds of assigned gender incongruence. The model also had a statistically significant effect for age, where the odds ratio of 1.04 meant that the model predicted that for every year older that a case was, it had a 4% increased odds of assigned gender incongruence.
|Table 1: A mixed-effects logistic model with predictors of adult assigned gender incongruence|
Click here to view
The results of this generalized mixed-model are further illustrated in [Table 2], which outlines the predicted probabilities for assigned gender incongruence broken down by assigned gender, DSD type, and surgery status. The statistically significant assigned female predictor in [Table 1] is reflected in the predicted probabilities of assigned gender incongruence for assigned female cases, being generally higher than for assigned male cases in [Table 2]. The overall predicted probability (PP) of assigned gender incongruence for cases assigned female (0.24 [95% confidence interval = 0.13–0.40]) represents a 4.36 times increased risk compared to the probability of assigned gender incongruence for cases assigned male (0.06 [0.02–0.14]).
|Table 2: Predicted probability of adult assigned gender incongruence by differences of sex development type, assigned gender, and childhood genital surgery|
Click here to view
The nonsignificant result for childhood genital-normalizing surgery is illustrated in [Table 2], with the column showing the PP for cases that had received this surgery not being notably different from their corresponding column of PP for those with no surgery. Finally, the nonsignificant result for DSD type is illustrated in [Table 2] by each cluster of four DSD types' predicted probabilities not being significantly different from each other.
Two supplementary analyses were conducted to further test the robustness of our conclusions. In the first supplementary analysis, we tested whether including sex chromosomes in the model, rather than DSD type, would affect the results. In the second supplementary analysis, we tested whether altering the childhood genital surgery variable from a binary (yes or no) variable to become a nominal variable that delineated masculinizing and feminizing surgeries would affect the results. As outlined in [Supplementary Table 2] and [Supplementary Table 3], both sex chromosomes (XX, XY, or mosaic karyotype) and childhood gender surgery type (masculinizing, feminizing, or no surgery) were not statistically significant predictors of assigned gender incongruence, and these alternative supplementary models produced minimal changes to the overall results, as shown in [Table 1].
| Discussion|| |
This study found that assigned gender, but not childhood genital surgery or DSD type, was related to assigned gender incongruence, in a combined analysis of 282 cases reported in 58 articles. The following subsection will discuss each of the predictors of assigned gender incongruence in detail.
Childhood genital-normalizing surgeries
The results of this study are consistent with what we have expected to find if childhood genital-normalizing surgeries had no impact on gender identity development, or that they had a small effect that this study did not have the power to detect. While it is always important to be cautious with correlational data, a strength of this study is that it controlled for potentially confounding third variables: assigned gender, DSD type, age of the case, and year in which that case was published. Further testing with a sex chromosome variable and a surgery variable that delineated feminizing and masculinizing surgeries did not notably change these results (see supplementary electronic materials). Our data were limited in that we were not able to examine specific surgical methods.
While the past justification for these surgeries has been that they help to create stable gender identities, this does not appear to be founded based on this research. While this justification may not have been explicitly part of the current surgical practice for some time, this finding still adds a useful contribution to scientific knowledge about the relative effects of different factors in gender development.
It is also possible that there might be implicit unconscious biases continuing in decisions about these surgeries as well as decisions about assigning DSD children female based on surgical considerations. Unconscious biases occur when prejudice or stereotyping toward a marginalized group, such as people with DSDs, influences medical decision-making, even if those decision makers are not consciously aware of it. Evidence for this is reflected in writing that gives more subtle indications that surgery is performed to influence gender development. The 2006 “Consensus statement on management of intersex disorders” stated that “successful (male) gender assignment depends on (phalloplasty)” (p. e492) and a 2014 commentary by the two lead authors of the consensus statement argued that leaving “children with genitalia that are neither primarily male nor female, to later decide on gender and sexuality seems to represent an inappropriate form of social experimentation” (p7, emphasis added). In 2014, the European Society for Paediatric Urology, and the Society of Pediatric Urology in the United States, stated that therapeutic management, which includes but is not limited to genital-normalizing surgery, directly impacts the “development of an individual's gender identity and gender role, a multifactorial process in which anatomical appearance likely plays a role” (p9).
There is also evidence that parents and family members of DSD children consider a stable gender identity a potential benefit of undertaking these surgeries. This means it is important that family members receive full information about the lack of evidence that genital-normalizing surgeries will have any impact on gender development. It is pleasing to see observations that more clinics are providing families with education and support to delay these normalizing surgeries until the person receiving this surgery is able to consent to it.
This study's nonsignificant findings should be considered in the context of the sample size and statistical power. We analyzed data from 282 cases (level 1) from 58 articles (level 2), which is greater than the 50 level 2 units suggested to avoid estimation bias and gave us the power to detect medium and large effects. It is possible that there was a weak effect for genital-normalizing surgery that we were not able to detect.
This research covered the substantial influence of assigned gender: assigned female cases had more than four times greater likelihood of assigned gender incongruence. This finding suggests that there has been a bias toward female assignment for many DSD cases, perhaps due to surgical constraints and assumptions about genitalia requirements for men.,,,
Age and year of publication
Older cases were more likely to have assigned gender incongruence. This finding was expected because older cases would have had more time to consider their gender and become aware of any incongruence with their assigned gender. We did not find any statistically significant association between the year of case publication and assigned gender incongruence. This finding suggests that changes in social environments, surgical techniques, and clinical care over time have not impacted the likelihood of assigned gender incongruence.
Differences of sex development type
There was also no statistically significant difference in the likelihood of assigned gender incongruence for the different DSD types: CAH, micropenis, PAIS, and MGD. This may be due to these different DSD types having similar effects from either androgen exposure or sensitivity to androgen exposure. We did not have the statistical power to examine differences in the manifestation or genotypes within these different DSD types, which might have given us a more sensitive measure of developmental androgen exposure.
Strengths and limitations
The type of data that we used was both a strength and a limitation of this research. While it was a strength that we could analyze cases reported from a diverse range of sources and statistically account for within-article variance, it is possible that there was a publication bias effect. Those cases with assigned gender incongruence may have been more likely to be published because they were more newsworthy. Alternatively, the authors might have been inclined toward wanting to present cases that they saw as successful (i.e., no assigned gender incongruence). We were also only able to include studies that reported individual case details; this limited us because we were not able to include more recent researches that tended to use larger samples without individualized case details. Nevertheless, it appears common for DSD research to continue to report individual case studies, and other researches have combined cases to draw conclusions about the rates of assigned gender incongruence;, this study takes a step further to do this with a quantitative analysis. While this research could examine cases with a range of different DSD types, the definition of these types of DSD might not always be consistent among article authors. For example, we included cases that were labeled as having micropenis and did not assess whether they met a certain criterion for penile length.
Most of the cases analyzed in this research had not received childhood genital surgery. There may be publication bias, in which teams that undertake childhood genital surgeries may be less likely to publish their cases. Our sense from this review of the academic literature is that a large proportion of the cases that had received this surgery were not followed up long enough to reach adulthood. Another study has found evidence that there is a discrepancy between those who are publishing in this subject area and what is happening in real-life practices.
It is important to note that assigned gender incongruence is not a negative outcome in itself. Transgender community advocacy has led much progress to depathologize transgender people. Despite this, if there has been a bias toward the female assignment of DSD children, then depriving children of childhoods in a gender that they are comfortable with, and possibly creating gender dysphoria, are all negative outcomes that should be avoided.
Our review of the literature has also led us to recommend more research into the reported life satisfaction of people with DSDs. Such research may alleviate concerns that having delayed gender assignment or atypical genitals will impact a child's gender identity or general well-being.
With the present state of controversy regarding these issues among clinicians, researchers, and advocates, it is imperative to note that the ethical considerations (e.g., bodily autonomy, informed consent, beneficence, and justice) should always be at the forefront of the minds of those making decisions about early gender assignment and undertaking surgeries to normalize a child's genitals. While this research has no bearing on these paramount ethical issues, it adds to our body of knowledge of gender development and it may help to encourage health professionals to explore if they consciously or unconsciously hold biased beliefs that early gender assignment and typical genitals are required for stable gender development.
Many thanks to Kaison Montieth for assisting with the literature search for this article.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| Supplementary Materials|| |
Calculation of intraclass correlation coefficient: To assess the extent that assigned gender incongruence varied within and between clusters, we first tested an intercept only binary logistic mixed model (binomial distribution with logit link function) and extracted the random intercept variance, var (u0j) = 5.19. We then calculated the intraclass correlation coefficient (ICC) by hand, ICC = var (u0j) + [var (u0j) + (π2/3)] = 5.19 + (5.19 + 3.29) = 0.61
| References|| |
Hiort O, Birnbaum W, Marshall L, Wünsch L, Werner R, Schröder T, et al
. Management of disorders of sex development. Nat Rev Endocrinol 2014;10:520-9.
Lee PA, Houk CP, Ahmed SF, Hughes IA. Consensus statement on management of intersex disorders. Pediatrics. 2006;118:e488-500.
Creighton SM, Liao LM. Changing attitudes to sex assignment in intersex. BJU Int 2004;93:659-64.
Creighton SM. Long-term outcome of feminization surgery: the London experience. BJU Int 2004;93 Suppl 3:44-6.
Creighton SM, Chernausek SD, Romao R, Ransley P, Salle JP. Timing and nature of reconstructive surgery for disorders of sex development – Introduction. J Pediatr Urol 2012;8:602-10.
Wolffenbuttel KP, Crouch NS. Timing of feminising surgery in disorders of sex development. In: Hiort O, Ahmed SF, editors. Understanding Differences and Disorders of Sex Development (DSD). Vol. 27. Karger: Endocrine Development; 2014. p. 210-21. Available from: https://www.karger.com/Article/FullText/363665
. [Last accessed on 2019 Jan 10].
Lee PA, Houk CP. Commentary to 'Evidence regarding cosmetic and medically unnecessary surgery on infants.' J Pediatr Urol 2014;10:7.
Mouriquand PD, Gorduza DB, Gay CL, Meyer-Bahlburg HFL, Baker L, Baskin LS et al
. Surgery in disorders of sex development (DSD) with a gender issue: If (why), when, and how? J Pediatr Urol. 2016;12:139-49.
Clune-Taylor C. Securing cisgendered futures: Intersex management under the “disorders of sex development” treatment model. Hypatia 2019;34:690-712.
Schweizer K, Brunner F, Handford C, Richter-Appelt H. Gender experience and satisfaction with gender allocation in adults with diverse intersex conditions (divergences of sex development, DSD). Psychol Sex 2014;5:56-82.
Mouriquand P, Caldamone A, Malone P, Frank JD, Hoebeke P. The ESPU/SPU standpoint on the surgical management of Disorders of Sex Development (DSD). J Pediatr Urol 2014;10:8-10.
Creighton SM, Michala L, Mushtaq I, Yaron M. Childhood surgery for ambiguous genitalia: Glimpses of practice changes or more of the same? Psychol Sex 2014;5:34-43.
Kreukels BP, Cohen-Kettenis PT, Roehle R, van de Grift TC, Slowikowska-Hilczer J, Claahsen-van der Grinten H, et al
. Sexuality in adults with differences/disorders of sex development (DSD): Findings from the dsd-LIFE study. J Sex Marital Ther 2019;45:688-705.
Garland F, Travis M. Legislating intersex equality: Building the resilience of intersex people through law. Leg Stud 2018;38:587-606.
Schoer MB, Nguyen PN, Merritt DF, Wesevich VG, Hollander AS. The role of patient advocacy and the declining rate of clitoroplasty in 46, XX patients with congenital adrenal hyperplasia. Clin Pediatr (Phila) 2018;57:1664-71.
Yankovic F, Cherian A, Steven L, Mathur A, Cuckow P. Current practice in feminizing surgery for congenital adrenal hyperplasia; a specialist survey. J Pediatr Urol 2013;9:1103-7.
Sturm RM, Durbin-Johnson B, Kurzrock EA. Congenital adrenal hyperplasia: current surgical management at academic medical centers in the United States. J Urol 2015;193:1796-801.
Diamond M, Garland J. Evidence regarding cosmetic and medically unnecessary surgery on infants. J Pediatr Urol. 2014;10:2-6.
Dessens AB, Slijper FM, Drop SL. Gender dysphoria and gender change in chromosomal females with congenital adrenal hyperplasia. Arch Sex Behav 2005;34:389-97.
Mazur T. Gender dysphoria and gender change in androgen insensitivity or micropenis. Arch Sex Behav 2005;34:411-21.
Sommet N, Morselli D. Keep calm and learn multilevel logistic modeling: A simplified three-step procedure using Stata, R, Mplus, and SPSS. Int Rev Soc Psychol 2017;30:203-18.
Kapur N. Unconscious bias harms patients and staff. BMJ 2015;351:h6347.
Crissman HP, Warner L, Gardner M, Carr M, Schast A, Quittner AL, et al
. Children with disorders of sex development: A qualitative study of early parental experience. Int J Pediatr Endocrinol 2011;2011:10.
Maas CJ, Hox JJ. Sufficient sample sizes for multilevel modeling. Methodol Eur J Res Methods Behav Soc Sci 2005;1:86-92.
Chen H, Cohen P, Chen S. How big is a big odds ratio? Interpreting the magnitudes of odds ratios in epidemiological studies. Commun Stat Simul Comput 2010;39:860-4.
Kessler SJ. The medical construction of gender: Case management of intersexed infants. Signs 1990;16:3-26.
Michala L, Liao LM, Wood D, Conway GS, Creighton SM. Practice changes in childhood surgery for ambiguous genitalia? J Pediatr Urol. 2014;10:934-9.
Inch E. Changing minds: The psycho-pathologization of trans people. Int J Ment Health 2016;45:193-204.
[Table 1], [Table 2]