From ‘Shark-Week’ to ‘Mangina’: An Analysis of Words Used by People of Marginalized Sexual Orientations and/or Gender Identities to Replace Common Sexual and Reproductive Health Terms

Purpose: To explore sexual and reproductive health (SRH)-related word-use among sexual and gender minority (SGM) individuals in the United States. Methods: In 2019, we fielded an online quantitative survey on the SRH experiences of SGM adults. Eligible participants included transgender, nonbinary, and gender-expansive (TGE) people assigned female or intersex at birth, and cisgender sexual minority women (CSMW) in the United States. The survey asked participants to indicate if they used each of nine SRH terms, and if not, to provide the word(s) they used. We analyzed patterns in replacement words provided by respondents and tested for differences by gender category with tests of proportions. Results: Among 1704 TGE and 1370 CSMW respondents, 613 (36%) TGE respondents and 92 (7%) CSMW respondents replaced at least 1 SRH term (p-for-difference <0.001). Many (23%) replacement words/phrases were entirely unique. For six out of the nine terms, TGE respondents indicated that use of the provided term would depend on the context, the term did not apply to them, or they did not have a replacement word/phrase that worked for them. Conclusions: SRH terms commonly used in clinical and research settings cause discomfort and dysphoria among some SGM individuals. To address inequities in access to and quality of SRH care among SGM individuals, and to overcome long standing fear of mistreatment in clinical settings, more intentional word-use and elicitation from providers and researchers could increase the quality and affirming nature of clinical and research experiences for SGM people.


Introduction
Language powerfully influences the ways individuals perceive and experience the world. 1 Word choice can erase or simplify certain experiences, or conversely, validate and amplify experiences. 2,3 Unfortunately, in the sexual and reproductive health (SRH) field, there is a long history of cisnormative and heteronormative language in both clinical and research spaces that can alienate sexual and gender minority (SGM) individuals. 4 ''Cisnormative'' describes the assumption that all people are cisgender, that is, they identify with the gender assumed of their sex assigned at birth. ''Heteronormative'' refers to the assumption that all people are heterosexual or straight. SGM is an umbrella term for individuals whose experience falls outside of cisgender and heterosexual experiences. An example of these assumptions in clinical and research spaces is the National Survey of Family Growth (NSFG) from 2015 to 2017, which divided participants into ''female'' and ''male,'' and referred to ''female'' participants' partners exclusively with he/him pronouns. 5 Beyond alienating SGM people, these assumptions undermine the rigor of research and the quality of clinical care. In research spaces, these assumptions can induce selection and measurement biases through inadequate definition of study eligibility criteria and insufficient specificity in definition of certain health behaviors or experiences. 3,[6][7][8] In clinical care, these assumptions inform the accuracy of language used by providers or health facilities and may contribute to gender dysphoria (distress related to one's gender identity as it relates to their sex assigned at birth and/or how their gender is perceived), reduce people's comfort in seeking care and/or disclosing health behaviors, and impact the appropriateness and quality of care provided. 3,6,9,10 Prior research indicates that transgender people evaluate health care providers on use of language and will seek out or avoid providers according to the language that affirms their experiences. 7 In response to cisnormative and heteronormative language, transgender and nonbinary individuals have adopted language to affirm their identities. 2 Yet, one formative study found that most of the 1788 transgender and nonbinary respondents had not been asked about language preferences for their bodies by their providers, despite wanting to be asked. 11 One approach to avoiding these issues is for researchers and clinicians to elicit personcentered language from individuals, potentially allowing for greater inclusivity and precision in research and clinical care. 12 To explore this understudied topic, we set out to measure SRH-related word-use among SGM individuals. Specifically, we aimed to ascertain how many study participants use standard SRH-related medical terms versus their own words, and what those words are through a novel, online questionnaire. Better understanding of word-use among SGM individuals may contribute to improved health equity by creating more affirming research and clinical environments where SGM individuals may feel more comfortable seeking care and disclosing relevant health information, and where providers and researchers will better understand the unique needs of these populations. 3,11 Methods

Study population
Eligible participants included SGM individuals assigned female or intersex at birth, transgender, nonbinary, and gender-expansive (TGE) individuals or cisgender sexual minority women (CSMW), who resided in the United States or its territories, were 18 years or older, and were able to read and understand English. Participants were recruited through social media, email lists, and announcements at community events, and through The Population Research in Identity and Disparities for Equality (PRIDE) Study, 13 an online national prospective cohort study of SGM adults. Participants in The PRIDE Study had consented to be part of the longitudinal cohort, whereas participants recruited outside of The PRIDE Study agreed to participate in this single survey. Recruitment materials invited interested individuals to participate in an online quantitative survey covering a broad range of SRH topics. 12

Measures
Participants completed an online survey, fielded between May and September 2019 through Qualtrics (Qualtrics LLC; Provo, UT), which included several previously described SRH domains. 12 To measure the primary outcome, we designed the survey to allow use of respondent-provided words and phrases in place of standard SRH terms. To calibrate term meaning, we provided gender-neutral definitions for nine medical terms: ''abortion,'' ''birth control,'' ''breasts,'' ''penis,'' ''period,'' ''pregnant,'' ''sperm,'' ''uterus,'' and ''vagina.'' After reading each definition, the survey prompted all respondents to indicate whether they used that word, did not use that word, or preferred not to say. Respondents who indicated use of another word had to provide their replacement word/phrase to continue. After data were collected separately for each term, these replacement words were substituted for the medical terms throughout the remaining questions and answer choices. 12 These replacement words/phrases comprise the primary data for these analyses.
We also collected data on sociodemographic characteristics, including race/ethnicity, gender identity, sex assigned at birth, and sexual orientation. We measured gender identity with two questions: (1) an open-text field wherein participants could describe their gender identity in their own words and (2) a multiple-choice question asking participants to select all that apply from 12 gender-identity options. 14 We asked sex assigned at birth with a multiple-choice question with four response options: ''female,'' ''male,'' ''not listed (write-in),'' and ''prefer not to say.'' 14 Separately, we asked participants if they had ever received an intersex or difference in sex development diagnosis as well as whether they personally identified as intersex.

Analysis
We conducted descriptive quantitative analyses to summarize data on respondent sociodemographic characteristics with Stata 15.1 (StataCorp, College Station, TX). We classified respondents into one of two gender categories based on analyses of responses to the gender identity and sex assigned at birth questions: (1) TGE and (2) CSMW. 14, 15 We excluded data from those whose gender identity was not classifiable (n = 4).
Data cleaning involved eliminating explanations of term usage, separating multiple responses from a single participant so each registered as an individual response, and conversely, connecting multi-word responses with hyphens (e.g., ''I use both shark week and bleeding when talking with friends'' would become ''sharkweek'' and ''bleeding''). We did not alter the spelling of replacement words, as some misspellings may have been intentional. We included mentions of the medical term to capture respondents who stated they used multiple words, including the medical term, depending on context. We then imported cleaned responses into the qualitative analysis software MAXQDA (2020). 16 In the first phase of analysis, we quantified the frequency with which respondents provided replacement words, and the most common words/phrases per medical term utilizing the Word Frequencies tool in MAXQDA. In the second phase of analysis, we employed a phenomenological qualitative content analysis approach 17 to explore patterns and themes across re-placement words. Specifically, we utilized the Word Cloud generator in MAXQDA to generate visual representations of replacement words such that larger terms reflect a higher number of mentions. We compared the word clouds and word frequency charts to identify patterns in the data, particularly around the use of ''masculinized'' words for those terms often associated with bodies assigned female at birth (e.g., ''mancave'' for ''vagina'') and for words that expressed aversion to the medical term (e.g., ''unwanted-chest'' for ''breasts''). To evaluate patterns in word-use by broad gender category (TGE and CSMW), we conducted several tests of proportion in Stata.

Ethical review
This study was reviewed and approved by the Institutional Review Boards of Stanford University School of Medicine and the University of California, San Francisco. The PRIDE Study Research Advisory Committee and The PRIDE Study Participant Advisory Committee (PAC) (pridestudy.org) also reviewed and approved this study. All participants consented to participation before questionnaire initiation.

Respondent characteristics
Overall, 5005 people initiated the questionnaire. Of these, 3074 people were classified as TGE (n = 1704) or CSMW (n = 1370) and responded to at least one of the nine two-part word-use questions. Respondents had a median age of 28 years (interquartile range: 23-35) and resided across all four U.S. Census regions ( Table 1). Most respondents were white (87%) and college graduates (67%), and had health insurance (90%). A majority of respondents selected more than one gender identity (52%) and sexual orientation (59%). Most (99.7%) respondents reported their sex assigned at birth as female, and 0.3% (n = 10) reported their assigned sex at birth as ''not listed.'' In a separate question, 3.2% (n = 99) reported identifying as intersex.
Although there is overlap in the words most and least replaced by CSMW and TGE respondents, the greatest difference in replacement word use occurred for ''breasts'' and ''vagina,'' anatomical words often associated with bodies assigned female at birth. TGE respondents were over 23 times more likely to replace ''vagina'' than CSMW respondents (Table 2; p-for-difference < 0.001). An example word cloud for TGE replacement terms for ''vagina'' can be found in Figure 1. TGE respondents most commonly replaced ''vagina'' with the gender-neutral ''fronthole,'' comprising over 33% of the replacement words/phrases. Many TGE respondents also put some variation of ''front-hole'' (e.g., ''frontal-hole'') (n = 22, 8%) and ''genitals,'' (n = 16, 6%). Six TGE respondents provided the word ''dick,'' and 13 provided a masculinized term (e.g., ''mangina,'' ''mancave,'' and ''boypussy''); no CSMW respondents provided a neutral or masculinized replacement. Similarly, TGE respondents replaced the word ''breasts'' more than eight times as often as CSMW respondents ( p-for-difference < 0.001). Contextual replacement of and aversion to medical terms Respondents in both gender categories provided contextual responses for replacement words, delineating when or with whom they might use a given word (Table 3). Specifically, five TGE respondents noted that the term ''vagina'' was either ''too clinical'' or that they would only use ''vagina'' in a medical context. Two participants reported using ''vagina'' either in a medical setting or to discuss their period, but avoiding that word during sex. For ''breasts,'' six respondents indicated that word-use depends on whose body is being discussed. For ''period,'' another wrote, ''I use both; 'period' to doctors that don't understand transgender, or use 'shark week'.'' In response to six of the nine medical terms (''breasts,'' ''penis,'' ''period,'' ''sperm,'' ''uterus,'' and ''vagina''), 34 TGE respondents (2%) stated that the term did not apply to them, they did not have a word, or they avoid talking about the subject. For ''vagina,'' nine respondents reported they do not have a word or never reference a ''vagina.'' Notably, one respondent wrote ''I NEVER use this word, it causes me to dissociate to the point I can't walk. I don't use any terms for it.'' Five respondents reported not having breasts, having had top surgery, or associating breasts with their removal (e.g., ''bullshit flesh bags removed from ur [sic] body''). Similarly, for ''uterus,'' one respondent explained, ''the only time I ever bring up that area is when talking about a hysterectomy.'' CSMW respondents provided contextual responses and responses that expressed aversion with similar frequency to TGE respondents ( p = 0.66 and p = 0.38). For ''breasts,'' one CSMW respondent stated they would use ''tits or boobs, unless I'm in a medical setting''; another explained, '''breasts' makes me anxious and uncomfortable for whatever reason.''

Key findings
These results from a national survey of TGE people and CSMW assigned female or intersex at birth highlight several important findings. First, more than one in three respondents (36%) reported not using at least one medical term. Second, TGE respondents were more likely to replace words, with the greatest difference occurring for body parts often associated with bodies assigned female at birth: ''vagina'' and ''breasts.'' Third, the replacement words varied widely; almost a quarter were unique. Fourth, many respondents delineated word-use by context, indicating (1) that word-use is dependent on setting and (2) an assumption that health care providers may not understand words that reflect the experiences of SGM people. Finally, many respondents expressed discomfort with certain terms and explained that some terms did not apply to them.

Limitations and strengths
As with all research, this study had limitations. As no comprehensive sampling frame of TGE people exists, nor one for CSMW, we relied on convenience sampling; thus, our sample may not be representative of these populations. Furthermore, while 22% of respondents identified with a non-white racial/ethnic identity, our low number of respondents who identify as Black/ African American, Hispanic/Latinx, or another racial minority does not adequately reflect the racial diversity that exists among SGM people. In addition, we recognize the limitations of comparing TGE individuals with CSMW. This is not to imply that CSMW serve as an adequate comparison population nor representative of the general population, rather to reflect on potentially different experiences (e.g., transphobia) in clinical and research settings.
Specific to word-use data, while many participants explained why or when they use a specific word, we did not systematically measure word-use by context. Consequently, we cannot say whether these replacement words would or would not be acceptable for providers to use in clinical settings (as respondents may have provided them only as words that they feel comfortable using for their own bodies in private settings). Relatedly, in one study on provider communication preferences of transgender men and trans-masculine nonbinary individuals, most respondents reported preferring that providers use medical terms rather than slang when referring to their bodies. 11 Future research should ask participants what words they use for a given range of contexts and why.
Separately, our gender-neutral definitions for each SRH medical term, intentionally created with feedback from community stakeholders, may have influenced the desire to provide replacement words. We cannot know if or how replacement word-use might vary with standard medical definitions or those derived in another context, compared to the study-specific, gender-neutral definitions we provided in this survey. Because we decided to offer all respondents the opportunity to provide their own words, we cannot assess whether this feature influenced survey retention.
Specific to data cleaning, because we did not alter spelling to capture potential alternate spellings of words, we may have missed patterns across words that were unintentionally misspelled. As MAXQDA tools were not case sensitive, differences in use of caps and hyphens were not captured.
These limitations are balanced by strengths, namely the innovative survey design that allowed respondents to customize the language of the survey, and the large sample size spanning multiple gender identity categories, ages, and regions.

Conclusions
The diversity of replacement words provided by respondents emphasizes the importance of avoiding assumptions about the SRH words that people use. 11 The many contextual responses suggest that some individuals may not always use their preferred word for safety, to avoid having to explain themselves, or because of assumed nonacceptability in certain (primarily clinical) contexts. These findings add context to prior research findings that the communication skills and language used by health care providers can affect if and to what extent patients disclose relevant details about their sexual orientations and gender identities. 3,11,18,19 These findings emphasize the importance of personcentered language, mirroring an individual's own language, specifically inviting each individual to provide the word(s) they use to refer to a particular body part or process. Using person-centered language addresses issues of discrimination as individuals may associate inaccurate language, including the use of incorrect pronouns, with negative attitudes toward them on the part of providers. [20][21][22] In addition, defining words and then inviting individuals to provide their own may allow for the provision of more accurate and responsive clinical care and research, as we saw many respondents provided replacement words that applied to different body parts, expressed strong aversion to certain terms, or expressed that a given term did not apply to them.
To guide interactions between health care providers and patients, and between researchers and participants, it is important to clearly state the rationale for asking certain questions (and to ensure each question is necessary before asking it). 23 For example, ''In order to assess your risk of X, I am going to ask you a few questions about Y.'' To get on the same page about the words being used, a provider or researcher could say ''This is the word I use to talk about Z. What word do you use?'' In the case of written research materials, investigators could adopt a similar piped-in language approach to the customizable survey used in this study.
The full text of this survey and associated methodology has been published elsewhere. 12 Cisnormative and heteronormative language have been a source of great discrimination, erasure, and oppression for SGM people, contributing to health disparities in SRH. We offer these findings in the hope that others will adopt and adapt these methods to create more inclusive and individualized clinical and research experiences, which place greater autonomy in the hands of patients and research participants. These findings offer insight into the diversity of ways people talk about their bodies, and the impact on patients and research participants of imposing language that may carry cisnormative assumptions. Our work highlights the importance of avoiding assumptions, creating as much opportunity as possible to use person-centered language to eliminate potential gaps in knowledge, care, and measurement, and, most importantly, to ensure that we use affirming language that promotes health rather than harm.

Disclaimer
The statements in this article are solely the responsibility of the authors and do not necessarily represent the views of Patient-Centered Outcomes Research Institute, its Board of Governors or Methodology Committee, or the National Institutes of Health.