Teen Pregnancy Risk Factors Among Young Women of Diverse Sexual Orientations

Home , Sexual attraction

Teen Pregnancy Risk Factors Brittany M. Charlton, ScD,a, b,c,d Andrea L. Roberts, PhD,e Margaret Rosario, PhD,f Sabra L. Katz- AmongWise, PhD,a, b Jerel P. Calzo, Young PhD,g Donna Spiegelman, Women ScD,c, d,h S. Bryn Austin,of ScDa,b,d,e Diverse Sexual Orientations OBJECTIVES: abstract

Young women who are sexual minorities (eg, bisexual and lesbian) are approximately twice as likely as those who are heterosexual to have a teen pregnancy. Therefore, we hypothesized that risk factors for teen pregnancy would vary across sexual orientation groups and that other potential risk factors exist that are unique to sexual METHODS: minorities. We used multivariable log-binomial models gathered from 7120 young women in the longitudinal cohort known as the Growing Up Today Study to examine the following potential teen pregnancy risk factors: childhood maltreatment, bullying victimization and – perpetration, and gender nonconformity. Among sexual minorities, we also examined the following: sexual minority developmental milestones, sexual orientation related stress, RESULTS: sexual minority outness, and lesbian, gay, and bisexual social activity involvement. Childhood maltreatment and bullying were significant teen pregnancy risk – factors among all participants. After adjusting for childhood maltreatment and bullying, the sexual orientation related teen pregnancy disparities were attenuated; these risk factors explained 45% of the disparity. Among sexual minorities, reaching sexual minority developmental milestones earlier was also associated with an increased teen pregnancy CONCLUSIONS: risk. The higher teen pregnancy prevalence among sexual minorities compared – with heterosexuals in this cohort was partially explained by childhood maltreatment and bullying, which may, in part, stem from sexual orientation related discrimination. Teen pregnancy prevention efforts that are focused on risk factors more common among young – women who are sexual minorities (eg, childhood maltreatment, bullying) can help to reduce NIH the existing sexual orientation related teen pregnancy disparity.

a WHAT’S KNOWN ON THIS SUBJECT: Compared with their Division of Adolescent and Young Adult Medicine, Boston Children’s Hospital, Boston, Massachusetts; bDepartment of Pediatrics, Harvard Medical School, Departments of cEpidemiology, eSocial and Behavioral heterosexual peers, young women who are sexual minorities Sciences, and hBiostatistics, Nutrition, and Global Health and Population, Harvard T.H. Chan School of Public (eg, bisexuals and lesbians) are at heightened teen pregnancy Health, Harvard University, Boston, Massachusetts; dChanning Division of Network Medicine, Department of risk. Young sexual minority women report established risk Medicine, Brigham and Women’s Hospital, Boston, Massachusetts; fDepartment of Psychology, The Graduate factors for teen pregnancy, such as childhood maltreatment, Center, City University of New York, New York City, New York; and gDivision of Health Promotion and Behavioral more frequently than heterosexuals. Science, Graduate School of Public Health, San Diego State University, San Diego, California WHAT THIS STUDY ADDS: There is limited research in which Dr Charlton conceptualized the project, conducted the analyses, and led the development and researchers empirically examine whether teen pregnancy writing of the article; Dr Austin supervised data collection; Drs Austin, Roberts, Rosario, Katz-Wise, risk factors vary across sexual orientation groups. To our Calzo, and Spiegelman aided in the interpretation of data and critically reviewed the manuscript knowledge, no authors of US-based studies have explored for important intellectual content; and all authors approved the final manuscript as submitted whether there are also additional risk factors unique to and agree to be accountable for all aspects of the work. sexual minorities. This work was presented in part at the North American Society for Pediatric and Adolescent Gynecology Annual Clinical and Research Meeting; April 7–9, 2016; Toronto, Ontario; the Society for To cite: Charlton BM, Roberts AL, Rosario M, et al. Teen Pregnancy Risk Factors Among Young Women of Diverse Sexual Orientations. Pediatrics. 2018;141(4):e20172278

Downloaded from www.aappublications.org/news by guest on October 1, 2021 PEDIATRICS Volume 141, number 4, April 2018:e20172278 ARTICLE Compared with heterosexuals, young minority developmental milestones gay) at younger ages may be exposed women who are sexual minorities (eg, age at which they first identified to earlier, more persistent stigma and – (eg, bisexual and lesbian) are at a as a sexual minority), (5) sexual have less well-developed coping– skills heightened– risk for teen pregnancy orientation related stress (eg, compared with sexual minorities26 28 (defined as1 a10 pregnancy before age worry that others might think they who reach milestones later. ‍ ‍ 20 years). ‍ This population reports are lesbian or bisexual), (6) sexual Young women who experience established risk factors for teen minority outness, and (7) lesbian, sexual minority stress may seek to pregnancy,11 such as earlier sexual gay, and bisexual (LGB) social activity strengthen the perception that they initiation, more frequently than involvement. are heterosexual by having sexual “ ” heterosexuals. It has been suggested Childhood maltreatment is an intercourse with men. Individuals that young women who are sexual who are not out to others about established14, risk15 factor for teen minorities may be exposed to these their sexual orientation report29 lower pregnancy ‍ and is more16, 17common established risk factors more often among sexual minorities. ‍ self-esteem and more anxiety because of their sexual orientation Both bullying victimization and and may engage in behaviors or may even have additional risk that put them at risk for a teen perpetration have been associated 18 factors for teen pregnancy that10,12 are with an increase in teen pregnancy. pregnancy. Young women who are unique to sexual minorities. ‍ sexual minorities may view sexual Victims and perpetrators are more “ ” For example, in this research, we intercourse with men and pregnancy likely to19, experience20 depression, 21 suggest that young women who are anxiety, ‍ and low self-regard, as a way to stay closeted. Being sexual minorities may try to avoid which may elevate their teen involved with the LGB community or cope with the stigma related to pregnancy risk. Sexual minorities through activities like25,30 LGB social or their sexual orientation by having educational events ‍ confers social report being22, 23victims and perpetrators sexual intercourse with men, putting of bullying more often than support and15 may act as a buffer to themselves at risk for pregnancy. their heterosexual peers. Although risk factors. In previous research, data has been it has yet to be studied, sexual We hypothesize that the following leveraged from Canadian adolescents, minorities may engage in sexual proposed risk factors would be but there is limited research from the behaviors that put them at risk for positively associated with teen “ ” United States in which researchers a teen pregnancy for many reasons, pregnancy among young women of have empirically examined whether including wanting to prove to bully all sexual orientations: childhood teen pregnancy risk factors vary 13 victims or perpetrators that they are maltreatment, bullying victimization across sexual orientation groups. heterosexual. and perpetration, and gender To our knowledge, no authors of conformity. Among sexual minorities, US-based studies have explored Regarding gender nonconformity, young women who are the most we hypothesize that risk factors whether there are additional risk unique to those participants would factors that are unique to sexual feminine may be at greater risk for having a teen pregnancy. One be positively associated with teen minorities. Therefore, we examined pregnancy, namely earlier timing potential pathway follows the gender – teen pregnancy risk factors, including 24 of sexual minority developmental a number of potential risk factors intensification hypothesis, which posits that young people experience milestones, more sexual orientation that have yet to be explored in related stress, less outness, and less previous research, using longitudinal increased pressure to conform to culturally sanctioned gender roles LGB social activity involvement. data from young women of diverse METHODS sexual orientations in the United during adolescence. Therefore, young States. women who are the most gender Study Population conforming (ie, most feminine) may The following three proposed risk be the least prepared to negotiate factors for teen pregnancy were contraception with a male partner, The Growing Up Today Study (GUTS) ’ examined among participants elevating their risk of pregnancy. began in 1996 when women from of all sexual orientations: (1) It is suggested in the Sexual the Nurses Health Study 2 enrolled 25 N childhood maltreatment, (2) bullying Minority Stress Theory that sexual their female and male children victimization and perpetration, minorities experience stress related ( = 27793), aged 9 to 14 years, and (3) gender nonconformity. to the stigma associated with their into a longitudinal cohort study. The following four risk factors sexual orientation. Individuals who The current analysis was limited to relevant to participants who are reach sexual minority developmental female (based on their sex assigned sexual minorities were examined milestones (eg, identifying as mostly at birth) GUTS participants who among those participants: (4) sexual heterosexual, bisexual, lesbian, or reported their sexual orientation Downloaded from www.aappublications.org/news by guest on October 1, 2021 2 CHARLTON et al TABLE 1 Characteristics by Sexual Orientation in a US-Based Cohort of Young Women (N = 7120) Completely Completely Mostly Bisexual Lesbian Pa Heterosexual Heterosexual Heterosexual With No Same-Sex With Same-Sex Partners Partners 84%, n = 6003 2%, n = 131 11%, n = 774 2%, n = 150 1%, n = 62 Demographics, % (n) Mean age in 1996, y, SD, range: 9–15 y 11.6, 1.6 11.3, 1.4 11.3, 1.7 11.6, 1.7 11.4, 1.6 <.0001 White race and/or ethnicity 94 (5651) 92 (120) 91 (702) 91 (136) 90 (56) .001 Potential risk factors in all participants, % (n) Childhood gender nonconformityb Below median 57 (3400) 54 (71) 44 (340) 32 (48) 32 (20) — Above median, below top decile 31 (1847) 34 (45) 34 (266) 38 (57) 18 (1511 — Top decile 13 (756) 12 (15) 22 (168) 30 (45) 50 (31) <.0001 Childhood physical and/or emotional abuse, Childhood Trauma Questionnaire, % (n) Quintile 1, least 21 (945) 15 (17) 13 (78) 6 (7) 10 (5) — Quintile 2 23 (1080) 21 (24) 20 (123) 13 (15) 17 (8) — Quintile 3 22 (989) 26 (30) 19 (116) 16 (19) 8 (4) — Quintile 4 20 (898) 18 (21) 23 (142) 24 (28) 31 (15) — Quintile 5, most 15 (684) 21 (24) 26 (162) 41 (47) 33 (16) <.0001 Childhood threatened and/or actual physical assault, Conflict Tactic Scales, % (n) Never 9 (408) 10 (11) 8 (51) 6 (7) 13 (6) — Rarely or sometimes 77 (3535) 72 (84) 70 (435) 62 (72) 71 (34) — Often or very often 14 (651) 18 (21) 22 (135) 32 (37) 17 (8) <.0001 Childhood sexual abuse 2 (106) 7 (8) 5 (33) 11 (12) 4 (2) <.0001 Bullying, % (n) None 77 (3113) 63 (64) 69 (373) 59 (62) 67 (28) — Victim only 9 (358) 6 (6) 13 (69) 15 (16) 17 (7) — Perpetrator only 7 (285) 18 (18) 8 (45) 14 (15) 7 (3) — Victim and perpetrator 7 (275) 14 (14) 10 (53) 11 (12) 10 (4) <.0001 Potential risk factors among participants who are sexual minorities, % (n) Sexual minority developmental milestonesc Late adolescence and/or young N/Ad N/Ad 38 (168) 20 (20) 11 (5) — adulthood Middle adolescence N/Ad N/Ad 40 (176) 42 (43) 44 (20) — Early adolescence N/Ad N/Ad 22 (96) 38 (39) 44 (20) <.0001 Sexual orientation–related stress,e N/Ad N/Ad 2.3, 1.0 1.9, 0.9 1.7, 0.8 <.0001 mean, SD Sexual minority outnessf Everyone knows N/Ad N/Ad 9 (50) 38 (52) 69 (36) — Most or a few people know N/Ad N/Ad 70 (409) 58 (79) 29 (15) — No one knows N/Ad N/Ad 22 (129) 4 (5) 2 (1) <.0001 LGB social activity involvementg ≥6 times per y N/Ad N/Ad 3 (23) 17 (24) 56 (30) — <6 times per y N/Ad N/Ad 66 (462) 72 (100) 43 (23) — Never N/Ad N/Ad 31 (219) 10 (14) 2 (1) <.0001 Outcome, % (n) Any pregnancy in ages <20 y 2 (117) 3 (4) 4 (28) 9 (14) 5 (3) <.0001 N/A, not applicable; —, P value was calculated for categorical variable. a P value was calculated by using an analysis of variance for continuous variables and χ2 tests for categorical variables (except for teen pregnancy in which it was calculated by using Fisher’s exact test). b Median was based on 4 items from the Recalled Childhood Gender Identity and Gender Role Questionnaire. c Age at which sexual minority identification, same-sex attraction, or same-sex sexual contact first occurred, categorized as late adolescence and/or young adulthood (ages –18 27 y), middle adolescence (ages 14–17 y), or early adolescence (ages ≤13 y). d Data collection among mostly heterosexual, bisexual, and lesbian participants only. e Sexual orientation–related stress was assessed via questions about the extent to which participants wished they were not attracted to the same sex, with response options 1 (strongly disagree) to 5 (strongly agree); higher scores are used to indicate more stress. f Mean of 3 items on outness to family, friends, and people at school or work. g Mean of 3 items on frequency of LGB social activity involvement.

Downloaded from www.aappublications.org/news by guest on October 1, 2021 PEDIATRICS Volume 141, number 4, April 2018 3 and information Non the proposed (made up of the mostly homosexual categorized any such reports into: teen pregnancy risk factors between and completely homosexual groups). never (ref), rarely or sometimes, and 1996 and 2010 ( = 7970). This We ran sensitivity analyses using often or very often. Sexual abuse ’ study was approved by the Brigham the sexual orientation report at age was measured35 with 2 additional and Women s Hospital Institutional 16 years (preceding most of the questions. The first question was Review Board. teen pregnancies); the results did about being touched by or forced Sexual Orientation not reveal meaningful change. For to touch an adult or older child in a the mediation analyses, the sexual sexual way when the participant did minority groups were collapsed into not want to, and a second question a single group for statistical power asked about an adult or older child Detailed information about sexual and interpretation ease. forcing or attempting to force sexual orientation has been collected on Potential Risk Factors for Teen activity by threatening, holding 36 every GUTS questionnaire since Pregnancy down, or hurting the participant. 1999 by using an item adapted from We considered any participant Childhood Maltreatment the Minnesota31 Adolescent Health with such reports as being sexually Survey that asks about feelings “ abused (yes or no). The 3 types of of attraction and identity. The item ’’ childhood maltreatment (physical read, Which of the following best In 2007, participants reported and/or emotional abuse, threatened describes your feelings? and childhood maltreatment before or actual physical assault, and sexual included the following response age 11 years. Physical and/or Bullyingabuse) were Victimization all examined and Perpetration separately. options: completely heterosexual emotional abuse was measured by (attracted to persons of the using 5 items from the Childhood32 opposite sex), mostly heterosexual, Trauma Questionnaire, capturing Using 2 items adapted from the World bisexual (equally attracted to men the frequency with which adults in Health Organization Health Behavior and women), mostly homosexual, and outside of the family as well as 37 of School-aged Children Survey, completely homosexual (gay or siblings did the following: yelled or participants reported in 2001 the lesbian, attracted to persons of the screamed, said hurtful or insulting frequency of past-year bullying same sex), and not sure. This item things, punished the child in a way victimization and perpetration. Items was combined with a question about that seemed cruel, hit so hard it “ ’ “ were measured on a 5-point Likert the sex of sexual partners that read, left bruises or marks, and made ’ ” During your life, the person(s) the child feel important (reverse scale ranging from I haven t been ” “ ” with whom you have had sexual coded). To make comparisons bullied or I haven t bullied anyone “ contact is (are): and included the across the literature, we created a to several times a week. We following response options: I have score summing across all potential categorized bullying experiences as not had sexual contact with anyone, perpetrators and all items, as has ” 33 none, victim only, perpetrator only, or female(s), male(s), and female(s) been done previously, to categorize bothGender victim Nonconformity and perpetrator. and male(s). These data allowed the amount of physical and/or participants to be categorized by all 3 emotional abuse into quintiles. dimensions of sexual orientation (ie, The lowest quintile was used to identity, behavior, and attraction), indicate the least abuse (including Participants responded in 2005 and thereby allowing participants none) and the highest quintile was 2007 to 4 items from the Recalled to endorse an identity such as used to indicate the most abuse. Childhood Gender Identity and Gender 38 completely heterosexual while also Threatened or actual physical Role Questionnaire about behaviors assault was measured with 4 items reporting behaviors such as having 34 during childhood up to age 11 years same-sex partners. from the Conflict Tactics Scales, regarding media characters imitated capturing the frequency with which Sexual orientation groups were adults in and outside of the family or admired, roles taken in pretend modeled by using the report at as well as siblings did the following: play, favorite toys and games, and feelings of femininity or masculinity. age 19 years (closest to most of pushed, grabbed, or shoved; spanked “ Response options were on a 5-point the teen pregnancies) as follows: for discipline; and threatened or ‘ ’” “ scale, ranging from always women completely heterosexual with no actually did kick, punch, or hit ‘ ’” same-sex partners (reference group with something that could hurt or or girls or very feminine to always [ref]), completely heterosexual physically attacked in another way. boys or men or very masculine. with same-sex partners, mostly As has been done previously, we We calculated a mean of the 4 heterosexual, bisexual, and lesbian combined perpetrator groups and questions and modeled this into Downloaded from www.aappublications.org/news by guest on October 1, 2021 4 CHARLTON et al TABLE 2 Multivariable Relative Risk of Teen Pregnancy in a US-Based Cohort of Young Women (N = 7120) Teen Pregnancy RR (95% CI) Individual Variable Modelsa Models Simultaneously Adjusted for Other Significant Variablesb Sexual orientation Completely heterosexual with no same-sex partners Ref Ref Completely heterosexual with same-sex partners 1.51 (0.56–4.03) 1.71 (0.62–4.74) Mostly heterosexual 1.78 (1.18–2.69) 1.32 (0.73–2.37) Bisexual 4.67 (2.74–7.97) 2.34 (1.14–4.81) Lesbian 2.37 (0.76–7.37) 1.14 (0.16–8.13) Potential risk factors in all participants Childhood gender nonconformityc Below median Ref N/Ab Above median, below decile 0.90 (0.64–1.26) N/Ab Top decile 1.00 (0.65–1.55) N/Ab Childhood physical and/or emotional abuse Quintile 1, least Ref. Ref Quintile 2 0.58 (0.29–1.16) 0.52 (0.24–1.09) Quintile 3 0.83 (0.43–1.58) 0.62 (0.30–1.27) Quintile 4 1.04 (0.56–1.92) 0.62 (0.30–1.27) Quintile 5, most 2.30 (1.34–3.95) 1.65 (0.90–3.03) Childhood threatened and/or actual physical assault Never Ref N/Ab Rarely or sometimes 0.93 (0.48–1.77) N/Ab Often or very often 1.33 (0.64–2.76) N/Ab Childhood sexual abuse 4.81 (2.78–8.32) 2.97 (1.52–5.79) Bullying No bullying involvement Ref Ref Victim only 0.93 (0.47–1.85) 0.56 (0.22–1.43) Perpetrator only 2.95 (1.88–4.63) 2.27 (1.30–3.97) Victim and perpetrator 2.40 (1.44–4.00) 1.98 (1.09–3.62) Potential risk factors among participants who are sexual minorities Sexual minority developmental milestonesd Late adolescence and/or young adulthood Ref N/Ab Middle adolescence 5.67 (1.30–24.82) N/Ab Early adolescence 3.15 (0.61–16.18) N/Ab Sexual orientation–related stresse 0.97 (0.71–1.31) N/Ab Sexual minority outnessf Everyone knows Ref. N/Ab Most or a few people know 0.86 (0.40–1.84) N/Ab No one knows 0.48 (0.15–1.54) N/Ab LGB social activity involvementg ≥6 times per y Ref N/Ab <6 times per y 1.51 (0.37–6.22) N/Ab Never 2.14 (0.49–9.28) N/Ab N/A, not applicable. a Each model used to assess the association with teen pregnancy is adjusted for age and race and/or ethnicity and is not adjusted for other variables. b The model used to assess the association with teen pregnancy is adjusted for age and race and/or ethnicity and is simultaneously adjusted for the other statistically significant variables (sexual orientation, childhood physical and/or emotional abuse, childhood sexual abuse, and bullying); therefore, the other risk factors do not have any estimate in this model. c Median was based on 4 items from the Recalled Childhood Gender Identity and Gender Role Questionnaire. d The age at which sexual minority identification, same-sex attraction, or same-sex sexual contact first occurred, categorized as late adolescence and/or young adulthood (ages 18–27 y), middle adolescence (ages 14–17 y), or early adolescence (ages ≤13 y). e Sexual orientation–related stress was assessed via questions about the extent to which participants wished they were not attracted to the same sex, with response options 1 (strongly disagree) to 5 (strongly agree); higher scores are used to indicate more stress and RR of teen pregnancy for a 1 U change in sexual orientation–related stress. f Mean of 3 items on outness to family, friends, and people at school or work. g Mean of 3 items on frequency of LGB social activity involvement.

Sexual Minority Developmental Milestones tertiles of nonconformity (below conforming were categorized below median, above median but below top the median and the participants who decile, and top decile), so as to not are the least gender conforming were assume linearity, as33 has been done in categorized in the top decile. Data In 2010, 3 items were adapted from previous literature. Therefore, the were used from 2005, and we imputed the Sexual Risk 39Behavior Assessment participants who are the most gender any missing data from 2007. Schedule-Youth to assess the age at Downloaded from www.aappublications.org/news by guest on October 1, 2021 PEDIATRICS Volume 141, number 4, April 2018 5 TABLE 3 Proportion of Sexual Orientation and Teen Pregnancy Association Mediated by Risk Factors in a US-Based Cohort of Young Women (N = 7120) Sexual Orientation Proportion Mediated (P) by Risk Factors Childhood Physical and/ Childhood Sexual Abuse Bullying Childhood Physical and/or or Emotional Abuse Emotional Abuse and Sexual Abuse and Bullying Completely heterosexual with no same-sex partners Ref Ref. Ref Ref Sexual minoritya 0.38 (0.01) 0.32 (0.03) N/Ab 0.45 (0.02) N/A, not applicable. a Sexual minority includes completely heterosexual with same-sex partners, mostly heterosexual, bisexual, and lesbian. b No mediation.

LGB Social Activity Involvement which the following milestones first 20 years were categorized as ever occurred among participants who havingConfounders a teen pregnancy. In 2003 and 2005,42 3 items adopted are sexual minorities: identifying from Rosario et al were used to as mostly heterosexual, bisexual, assess the participation in LGB lesbian, or gay; sexual attraction social activities of a participant who Potential confounders included to girls; and sexual contact with is a sexual minority. Participants baseline age in years (continuous) girls. On the basis of adolescent 27 indicated how often they went to LGB and race and/or ethnicity (white, development and previous literature, social events, cafes, dance clubs, bars, another race and/or ethnicity). In we categorized the youngest or hung around these places; went to sensitivity analyses, we explored age of reaching sexual minority LGB meetings or educational events whether estimates significantly identification, same-sex attraction, at a community center or other place; changed when adjusting for or same-sex sexual contact into late – and read or watched LGB magazines, other covariates, such as region adolescence and/or young adulthood newspapers, books, Web sites, – “ of residence at age 19 years (the (ages 18 27 years), middle videos, or movies. A 7-level response ≤ ” “ ” report closest to most of the teen adolescence (ages 14 17 years), and option ranged from 2 or more times pregnancies), coded as West, Sexualearly adolescence Orientation– (agesRelated 13Stress years). per week to never with 5 other Midwest, South, or Northeast; the middle-response options. We used results did not reveal meaningful a mean score of these items from change. Missing data on all potential ’ ≥ Answering an item adapted from both years and categorized this as cofounders were <0.01%, so we used occurring 6 times per year, <6 times Statisticala complete Analysiscase approach. Wright and Perry40 s Sexual Identity Distress Scale, participants who per year, or never. “ are sexual minorities indicated in ” “ It is possible that some of the 2003 and 2005 on a 1 to 5 scale ( 1 = First, we examined the association ” proposed risk factors, including strongly disagree to 5 = strongly – between sexual orientation, the bullying victimization and agree ) the extent to which they potential risk factors, and teen perpetration, sexual orientation wished they were not attracted to the pregnancy using bivariate models. related stress, and LGB social activity same sex; scores from 2003 to 2005 ’ Then, we used multivariate involvement were reported after were averaged, and higher scores regression from log-binomial models some participants pregnancies, so Sexualwere used Minority to indicate Outness more stress. adjusted for potential confounders. we ran sensitivity analyses excluding To account for sibling clusters, any individuals whose pregnancies we estimated the variance using ’ occurred after the report of these Answering items adapted from generalized estimating equations 41 factors; the results did not reveal D Augelli et al, participants who with a compound symmetry working meaningful change. are sexual minorities indicated in Teen Pregnancy correlation matrix. To examine if 2003 and 2005 their level of sexual each variable was a risk factor for orientation outness to family, friends, “ teen pregnancy, we fit models for and people at school or work using In 1999, lifetime histories of ” “ ” each individual potential risk factor, a 5-option scale from everyone pregnancy were collected from all resulting in risk ratios (RRs) and knows to no one knows. We participants; in each subsequent 95% confidence intervals (CIs). created a mean score of these 3 questionnaire, current pregnancies “ ” “ We also fit a model simultaneously items from both years, categorized have been queried. As is done in the ” “ ” 43 adjusting for the variables that were as everyone knows, most or a few literature, participants who had statistically significant risk factors. people know, and no one knows. any pregnancies occurring before age Finally, we calculated the percent Downloaded from www.aappublications.org/news by guest on October 1, 2021 6 CHARLTON et al 10 of the sexual orientation and teen each associated with teen pregnancy. Saewyc et al examined teen pregnancy association mediated After simultaneously adjusting for pregnancy risk factors across – by the statistically significant risk the significant risk factors, the sexual sexual orientation groups using factors using the 44publicly available orientation related teen pregnancy data from teens in British Columbia, %mediate macro . All analyses were disparities were attenuated (Table 2). Canada. In addition to established conducted by using SAS version 9.3 When testing for mediation, 38% teen pregnancy risk factors, the (SAS Institute, Inc, Cary, NC). of the disparity was explained by authors examined factors that may RESULTS physical and/or emotional abuse, and be unique to sexual minorities. 32% of the disparity was explained Compared with heterosexual peers, by childhood sexual abuse. Bullying female participants who are sexual n Of the 7120 participants in our was not a statistically significant minorities experienced sexual mediator. Together, childhood initiation at an earlier age and were sample, 84% ( = 6003)n identified as – completely heterosexual with no same- maltreatment and bullying explained more likely to have been sexually 45% of the sexual orientation related abused, both of which are teen sex partners, 2% ( n= 131) identified teen pregnancy disparity (Table 3). pregnancy risk factors. Our results as completely heterosexual withn same- from previous analyses support the sex partners, 11% ( = 774) identifiedn In individual models of the potential finding of an earlier age at sexual as mostly heterosexual, 2% ( = 150) risk factors unique to sexual initiation for young women who are identified as bisexual, and 1% ( = 62) minorities, reaching sexual minority 11 sexual minorities, and the current identified as lesbian. developmental milestones earlier study supports the sexual abuse was associated with an increase in Compared with completely finding. teen pregnancy (Table 2). Because heterosexuals with no same- 10 data were available only among sex partners, sexual minorities In that same study, Saewyc et al sexual minorities, no further experienced more childhood were also able to calculate an enacted mediation analyses were conducted maltreatment, more bullying stigma (defined as the exclusion, for those risk factors. victimization and perpetration, harassment, discrimination, and and greater gender nonconformity. None of the sensitivity analyses violence resulting from the disclosure Among participants who are sexual (modeling sexual orientation at age or even suspicion of a sexual minority minorities, lesbians reached sexual 16 years, excluding pregnancies that identity) composite score, including minority developmental milestones occurred before the proposed risk information on the following: at the youngest age, whereas factor was assessed, and adjusting experiencing discrimination on the bisexual and mostly heterosexuals basis of race, physical appearance, or – for geographic region) made any reported the highest amounts of meaningful differences in the results. sexual orientation; being excluded sexual orientation related stress, by others at school, being insulted DISCUSSION being the least out about their sexual or teased, or being physically orientation, and having the least LGB assaulted; or experiencing verbal social activity involvementn (Table 1). or physical sexual harassment. Young women who are sexual Compared with participants who A total of 2% ( = 166) of participants minorities in this cohort, particularly are sexual minorities without teen had a teen pregnancy. Compared bisexuals, were significantly more pregnancies, participants with teen – with completely heterosexuals with likely than their heterosexual peers pregnancies reported more sexual no same-sex partners, there was not to experience a teen pregnancy. The orientation based discrimination a statistically significant elevated risk higher teen pregnancy prevalence and a higher enacted stigma score. of a teen pregnancy for completely among sexual minorities was In the current study, we did not – – heterosexuals with same-sex partially explained by childhood have available data on sexual partners (RR: 1.51; 95% CI: 0.58 – maltreatment and bullying. One orientation related discrimination to 4.03), but mostly heterosexuals (RR: – additional variable, the earlier age make such a comparison. However, 1.78; 95% CI: 1.18 2.69) and lesbians ∼ of sexual minority developmental the higher prevalence of childhood (RR: 2.37; 95% CI: 0.76 7.37) had – milestones, was a significant risk maltreatment and bullying in 2 times the risk, and bisexuals (RR: factor for teen pregnancy among our participants who are sexual 4.67; 95% CI: 2.74 7.97) had nearly sexual minorities. minorities relative to heterosexuals – 5 times the risk (Table 2). With these findings, we build may, in part, stem from sexual – In individual models, childhood on the existing research about orientation related discrimination. maltreatment and bullying sexual orientation related teen In another study that was based on victimization and perpetration were pregnancy disparities. For example, data from Canadian teens and health Downloaded from www.aappublications.org/news by guest on October 1, 2021 PEDIATRICS Volume 141, number 4, April 2018 7 12 care providers, Travers et al factors that had yet to be explored, women only, so future researchers used qualitative data to examine including gender nonconformity, should explore teen pregnancy – teen pregnancy risk factors. The sexual minority developmental involvement among young men who focus groups discussed several milestones, sexual orientation are sexual minorities as well. potential risk factors, many of related stress, sexual minority CONCLUSIONS which stemmed from heterosexism. outness, and LGB social activity For example, male and female involvement. Whereas bullying has participants spoke about a lack of been a focus in previous research, Teen pregnancy prevention efforts sexual health information directed we were able to look at this with should focus on risk factors such as toward sexual minorities, including more nuance by examining not childhood maltreatment and bullying limited knowledge about their risk only bullying victimization but also among young people of all sexual of having or being involved in a bullying perpetration, which proved orientations. These risk factors are teen pregnancy. Participants also to be a strong teen pregnancy risk more common among young women mentioned teens who are sexual factor in this sample. Similarly, who are sexual minorities compared minorities feeling a desire to prove childhood maltreatment has been with heterosexual peers; therefore, – their heterosexuality or hide their the focus in previous research, but this focus can help to reduce the sexual minority status or even our data allowed us to examine existing sexual orientation related conform to heterosexual norms differences in the type of abuse. teen pregnancy disparity. Future through sexual activity with a ’ researchers should explore how The GUTS participants were children different gender or sex. these risk factors and others that 13 of Nurses Health Study 2 participants, are unique to sexual minorities Goldberg et al used the US-based so results from this sample may not (eg, reaching sexual minority National Longitudinal Study of generalize to other populations; this – developmental milestones earlier) Adolescent to Adult Health data to cohort is predominantly of white may be linked to sexual orientation examine how certain established race and/or ethnicity, their mothers ’ based discrimination and stigma; teen pregnancy risk factors attenuate are all nurses, and the majority of – such data can inform teen pregnancy sexual orientation disparities. As in participants annual household ≥ prevention efforts. the current study, sexual orientation incomes during their childhood related teen pregnancy disparities in was $75000. Teen pregnancy is a that cohort were partially attenuated relatively rare outcome in this sample, ABBREVIATIONS by factors such as a history of which limits the statistical power. It childhood sexual abuse. is possible that we lacked the data The current study does have some to detect small- to moderate-sized – CI: confidence interval limitations. GUTS data were not associations for some of the proposed GUTS: Growing Up Today Study available on sexual orientation based risk factors; future researchers should LGB: lesbian, gay, and bisexual discrimination or on the measures test these factors with an even larger Ref: reference group needed to create an enacted stigma sample size. Data on teen pregnancy RR: risk ratio score. However, we examined novel risk factors were available for young Reproductive Investigation Scientific Meeting; March 25–28, 2015; San Francisco, California; and the Society for Adolescent Health and Medicine Annual Meeting; March 18–21, 2015; Los Angeles, California. DOI: https://doi.org/10.1542/peds.2017-2278 Accepted for publication Dec 27, 2017 Address correspondence to Brittany M. Charlton, ScD, Division of Adolescent and Young Adult Medicine, Boston Children’s Hospital, 300 Longwood Ave, Boston, MA 02115. E-mail: [email protected] PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). Copyright © 2018 by the American Academy of Pediatrics FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose. FUNDING: Dr Charlton was supported by grant F32HD084000, Dr Katz-Wise by grant K99HD082340, Drs Rosario and Austin by grant R01HD057368, and Dr Austin by grant R01HD066963 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Dr Calzo was supported by grant K01DA034753 from the National Institute on Drug Abuse. Additional funds were provided to Dr Charlton by the Society of Family Planning through an Interdisciplinary Innovation grant and the American Cancer Society through grant MRSG CPHPS 130006. Funded by the National Institutes of Health (NIH). POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

Downloaded from www.aappublications.org/news by guest on October 1, 2021 8 CHARLTON et al REFERENCES 1. Flicker S, Flynn S, Larkin J, et al. teen pregnancy trends among gay, 20. Arseneault L, Walsh E, Trzesniewski Sexpress: The Toronto Teen Survey lesbian, and bisexual students in K, Newcombe R, Caspi A, Moffitt Report. Toronto, ON: Planned British Columbia. Can J Hum Sex. TE. Bullying victimization uniquely Parenthood Toronto; 2009 2008;17(3):123–139 contributes to adjustment problems in young children: a nationally 2. Har tnett CS, Lindley LL, Walsemann KM. 11. Charlton BM, Corliss HL, Missmer SA, representative cohort study. Congruence across sexual orientation et al. Reproductive health screening Pediatrics. 2006;118(1):130 138 dimensions and risk for unintended disparities and sexual orientation in – pregnancy among adult US women. a cohort study of US adolescent and 21. Egan SK, Perry DG. Does low self- Womens Health Issues. 2017;27(2): young adult females. J Adolesc Health. regard invite victimization? Dev 145–151.e2 2011;49(5):505–510 Psychol. 1998;34(2):299–309 3. Riskind RG, Tornello SL, Younger 12. T ravers R, Newton H, Munro L. “Because 22. Berlan ED, Corliss HL, Field AE, BC, Patterson CJ. Sexual identity, it was expected”: heterosexism as Goodman E, Austin SB. Sexual partner gender, and sexual health a determinant of pregnancy among orientation and bullying among among adolescent girls in the sexually diverse youth. Can J Commun adolescents in the growing up United States. Am J Public Health. Ment Health. 2011;30(2):65–79 today study. J Adolesc Health. 2014;104(10):1957–1963 13. Goldber g SK, Reese BM, Halpern CT. 2010;46(4):366–371 4. Blake SM, Ledsky R, Lehman T, Teen pregnancy among sexual minority 23. Schneider SK, O’Donnell L, Stueve Goodenow C, Sawyer R, Hack T. women: results from the National A, Coulter RW. Cyberbullying, school Preventing sexual risk behaviors Longitudinal Study of adolescent bullying, and psychological distress: among gay, lesbian, and bisexual to adult health. J Adolesc Health. a regional census of high school adolescents: the benefits of gay- 2016;59(4):429–437 students. Am J Public Health. sensitive HIV instruction in schools. Am 14. Noll JG, Shenk CE. Teen birth rates 2012;102(1):171–177 J Public Health. 2001;91(6):940–946 in sexually abused and neglected 24. Hill JP, Lynch ME. The intensification 5. Case P, Austin SB, Hunter DJ, et al. females. Pediatrics. 2013;131(4). of gender-related role expectations Sexual orientation, health risk factors, Available at: www.pediatrics.org/cgi/ during early adolescence. In: Brooks- and physical functioning in the Nurses’ content/full/131/4/e1181 Gunn J, Petersen AC, eds. Girls at Health Study II. J Womens Health 15. Saewyc EM, Taylor D, Homma Y, Puberty. Boston, MA: Springer; (Larchmt). 2004;13(9):1033–1047 Ogilvie G. Trends in sexual health and 1983:201–228 6. Charlton BM, Corliss HL, Missmer risk behaviours among adolescent 25. Meyer IH. Prejudice, social stress, SA, Rosario M, Spiegelman D, Austin students in British Columbia. Can and mental health in lesbian, gay, SB. Sexual orientation differences J Hum Sex. 2008;17(1–2):1–13 and bisexual populations: conceptual in teen pregnancy and hormonal 16. Saewyc EM, Skay CL, Pettingell SL, et al. issues and research evidence. Psychol contraceptive use: an examination Hazards of stigma: the sexual and Bull. 2003;129(5):674–697 across 2 generations. Am J Obstet physical abuse of gay, lesbian, and 26. Corliss HL, Cochran SD, Mays Gynecol. 2013;209(3):204.e1–e8 bisexual adolescents in the United VM, Greenland S, Seeman TE. Age 7. Goodenow C, Szalacha LA, Robin LE, States and Canada. Child Welfare. of minority sexual orientation Westheimer K. Dimensions of sexual 2006;85(2):195–213 development and risk of childhood orientation and HIV-related risk among 17. Corliss HL, Cochran SD, Mays VM. maltreatment and suicide attempts adolescent females: evidence from a Reports of parental maltreatment in women. Am J Orthopsychiatry. statewide survey. Am J Public Health. during childhood in a United 2009;79(4):511–521 2008;98(6):1051–1058 States population-based survey 27. Katz-Wise SL, Rosario M, Calzo JP, 8. Lindley LL, Walsemann KM. Sexual of homosexual, bisexual, and Scherer EA, Sarda V, Austin SB. orientation and risk of pregnancy heterosexual adults. Child Abuse Negl. Associations of timing of sexual among New York City high-school 2002;26(11):1165–1178 orientation developmental milestones students. Am J Public Health. 18. Lehti V, Sourander A, Klomek A, et al. and other sexual minority stressors 2015;105(7):1379–1386 Childhood bullying as a predictor with internalizing mental health symptoms among sexual minority 9. Saewyc EM, Bearinger LH, Blum RW, for becoming a teenage mother in young adults. Arch Sex Behav. Resnick MD. Sexual intercourse, abuse Finland. Eur Child Adolesc Psychiatry. 2017;46(5):1441 1452 and pregnancy among adolescent 2011;20(1):49–55 – women: does sexual orientation make 28. Parks CA, Hughes TL. Age differences 19. Hawker DS, Boulton MJ. Twenty years’ a difference? Fam Plann Perspect. research on peer victimization and in lesbian identity development 1999;31(3):127–131 psychosocial maladjustment: a meta- and drinking. Subst Use Misuse. 10. Saewyc EM, Poon CS, Homma Y, analytic review of cross-sectional 2007;42(2–3):361–380 Skay CL. Stigma management? The studies. J Child Psychol Psychiatry. 29. Jordan KM, Deluty RH. Coming out links between enacted stigma and 2000;41(4):441–455 for lesbian women: its relation to

Downloaded from www.aappublications.org/news by guest on October 1, 2021 PEDIATRICS Volume 141, number 4, April 2018 9 anxiety, positive affectivity, self-esteem, and preliminary psychometric data. 40. W right ER, Perry BL. Sexual identity and social support. J Homosex. J Fam Issues. 1996;17(3):283–316 distress, social support, and the health 1998;35(2):41–63 35. Moore D, Gallup G, Schussel R. of gay, lesbian, and bisexual youth. 30. Ker tzner RM, Meyer IH, Frost DM, Disciplining Children in America: A J Homosex. 2006;51(1):81–110 Stirratt MJ. Social and psychological Gallup Poll Report. Princeton, NJ: The 41. D ’Augelli AR, Hershberger SL, Pilkington well-being in lesbians, gay men, Gallup Organization; 1995 NW. Lesbian, gay, and bisexual youth and bisexuals: the effects of race, 36. Koss MP, Gidycz CA. Sexual and their families: disclosure of sexual gender, age, and sexual identity. Am J experiences survey: reliability and orientation and its consequences. Am Orthopsychiatry. 2009;79(4):500–510 validity. J Consult Clin Psychol. J Orthopsychiatry. 1998;68(3):361–371; discussion 372 375 31. Remafedi G, Resnick M, Blum R, Harris 1985;53(3):422–423 – L. Demography of sexual orientation in 37. Due P, Holstein BE, Lynch J, et al; Health 42. Rosario M, Hunter J, Maguen S, Gwadz adolescents. Pediatrics. 1992;89(4, pt Behaviour in School-Aged Children M, Smith R. The coming-out process 2):714–721 Bullying Working Group. Bullying and its adaptational and health-related 32. Bernstein DP, Stein JA, Newcomb MD, and symptoms among school-aged associations among gay, lesbian, et al. Development and validation children: international comparative and bisexual youths: stipulation and of a brief screening version of the cross sectional study in 28 countries. exploration of a model. Am J Community Childhood Trauma Questionnaire. Child Eur J Public Health. 2005;15(2):128–132 Psychol. 2001;29(1):133–160 Abuse Negl. 2003;27(2):169–190 38. Zucker KJ, Mitchell JN, Bradley SJ, 43. Romero L, Pazol K, Warner L, et al. 33. Rober ts AL, Rosario M, Corliss HL, Tkachuk J, Cantor JM, Allin SM. Reduced disparities in birth rates Koenen KC, Austin SB. Elevated The recalled childhood gender among teens aged 15–19 years–United risk of posttraumatic stress in identity/gender role questionnaire: States, 2006–2007 and 2013–2014. sexual minority youths: mediation psychometric properties. Sex Roles. MMWR Morb Mortal Wkly Rep. by childhood abuse and gender 2006;54(7–8):469–483 2016;65(16):409–414 nonconformity. Am J Public Health. 39. Meyer -Bahlburg HFL, Exner TA, Exner 44. Her tzmark E, Pazaris M, Spiegelman D. 2012;102(8):1587–1593 TM, Gruen RS. Sexual Risk Behavior %mediate [computer program]. 2009. 34. Straus MA, Hamby SL, Boney-McCoy Assessment Schedule-Youth. New Available at: https://www.hsph.harvard. S, Sugarman DB. The revised Conflict York, NY: New York State Psychiatric edu/donna-spiegelman/software/ Tactics Scales (CTS2): development Institute; 1994 mediate/

Downloaded from www.aappublications.org/news by guest on October 1, 2021 10 CHARLTON et al Teen Pregnancy Risk Factors Among Young Women of Diverse Sexual Orientations Brittany M. Charlton, Andrea L. Roberts, Margaret Rosario, Sabra L. Katz-Wise, Jerel P. Calzo, Donna Spiegelman and S. Bryn Austin Pediatrics 2018;141; DOI: 10.1542/peds.2017-2278 originally published online March 12, 2018;

Updated Information & including high resolution figures, can be found at: Services http://pediatrics.aappublications.org/content/141/4/e20172278 References This article cites 39 articles, 3 of which you can access for free at: http://pediatrics.aappublications.org/content/141/4/e20172278#BIBL Subspecialty Collections This article, along with others on similar topics, appears in the following collection(s): Adolescent Health/Medicine http://www.aappublications.org/cgi/collection/adolescent_health:med icine_sub Teen Pregnancy http://www.aappublications.org/cgi/collection/teen_pregnancy_sub Public Health http://www.aappublications.org/cgi/collection/public_health_sub LGBTQ+ http://www.aappublications.org/cgi/collection/lgbtq Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.aappublications.org/site/misc/Permissions.xhtml Reprints Information about ordering reprints can be found online: http://www.aappublications.org/site/misc/reprints.xhtml

Downloaded from www.aappublications.org/news by guest on October 1, 2021 Teen Pregnancy Risk Factors Among Young Women of Diverse Sexual Orientations Brittany M. Charlton, Andrea L. Roberts, Margaret Rosario, Sabra L. Katz-Wise, Jerel P. Calzo, Donna Spiegelman and S. Bryn Austin Pediatrics 2018;141; DOI: 10.1542/peds.2017-2278 originally published online March 12, 2018;

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://pediatrics.aappublications.org/content/141/4/e20172278

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. Pediatrics is owned, published, and trademarked by the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2018 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Downloaded from www.aappublications.org/news by guest on October 1, 2021