Content uploaded by Caroline Sten Hartnett
Author content
All content in this area was uploaded by Caroline Sten Hartnett on Jan 09, 2019
Content may be subject to copyright.
Family Planning
Congruence across Sexual Orientation Dimensions and Risk
for Unintended Pregnancy among Adult U.S. Women
Caroline Sten Hartnett, PhD
a
,
*
, Lisa L. Lindley, DrPH, MPH, CHES
b
,
Katrina M. Walsemann, PhD, MPH
c
a
Department of Sociology, University of South Carolina, Columbia, South Carolina
b
Department of Global & Community Health, College of Health & Human Services, George Mason University, Fairfax, Virginia
c
Department of Health Promotion, Education, and Behavior, Arnold School of Public Health, University of South Carolina, Columbia,
South Carolina
Article history: Received 23 April 2016; Received in revised form 25 October 2016; Accepted 31 October 2016
abstract
Objectives: To date, no studies have investigated whether sexual minority women (SMW) are more likely to experience
unintended pregnancies compared with their heterosexual peers. The aim of this study was to explore whether adult
SMW were more likely to have unintended pregnancies compared with heterosexual women, to examine the role of
identity–attraction congruence in unintended pregnancy risk, and to evaluate possible mediators.
Methods: Data on pregnancies to women ages 18 to 44 were drawn from the 2006 through 2013 National Survey of
Family Growth (n¼25,403). Weighted logistic regression models estimated the likelihood of reporting an unintended
(rather than intended) pregnancy by identity–attraction congruence and the extent to which this association was
mediated by sexual experiences with men, including age at first sex and number of sexual partners. Supplementary
analyses addressed the issue of abortion underreporting.
Results: Pregnancies to SMW were more likely to be unintended compared with pregnancies to heterosexual women
(adjusted odds ratio, 1.26; 95% confidence interval, 1.08–1.46). This was driven by the elevated risk experienced by
heterosexual-identified women with same-sex attractions, specifically (adjusted odds ratio, 1.28; 95% confidence
interval, 1.08–1.51). Greater unintended pregnancy risk among these women was mediated by a greater number of male
sex partners compared with heterosexual women.
Conclusions: Unintended pregnancy risk among SMW has historically received little attention from scholars and
clinicians. Future research should explore the specific conditions that put heterosexual-identified women with same sex
attractions at increased risk for unintended pregnancy. Clinicians should consider these dynamics when screening
patients for contraceptive counseling.
Ó2016 Jacobs Institute of Women's Health. Published by Elsevier Inc.
Unintended pregnancy rates in the United States are higher
than in many other industrialized countries; nearly one-half of
U.S. pregnancies are unintended (Finer & Zolna, 2016; Singh,
Sedgh, & Hussain, 2010). These pregnancies are associated with
lower levels of prenatal care and breastfeeding and higher levels
of premature delivery, low birth weight, child abuse, intimate
partner violence, and maternal depression and anxiety (Barber,
Axinn, & Thornton, 1999; Gipson, Koenig, & Hindin, 2008;
Kane, Morgan, Harris, & Guilkey, 2013; Korenman, Kaestner, &
Joyce, 2002; Kost, Landry, & Darroch, 1998a, 1998b; Myhrman,
1988; Najman, Morrison, Williams, Andersen, & Keeping, 1991).
Because of these negative consequences, Healthy People 2020
names as a national priority the reduction of unintended preg-
nancies as well as the elimination of disparities in unintended
pregnancy rates. This paper contributes to the burgeoning liter-
ature on unintended pregnancy among sexual minority women
(“SMW,”an umbrella term for women who identify as lesbian,
gay, or bisexual, have same-sex attractions, or are involved in
same-sex sexual behaviors). Although a few studies have
Funding Statement: Supported by grant R40MC28314-01-00, MCH Research
Program, from the Maternal and Child Health Bureau (Title V, Social Security
Act), Health Resources and Services Administration (HRSA), Department of
Health and Human Services. A version of this paper was presented at the 2015
Population Association of America Annual Meeting.
*Correspondence to: Caroline Sten Hartnett, PhD, Department of Sociology,
University of South Carolina, Sloan College #321, Columbia, SC 29205. Phone:
803-777-3123; fax: 803-777-5251.
E-mail address: hartnetc@mailbox.sc.edu (C.S. Hartnett).
www.whijournal.com
1049-3867/$ - see front matter Ó2016 Jacobs Institute of Women's Health. Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.whi.2016.10.010
Women's Health Issues 27-2 (2017) 145–151
investigated disparities in pregnancy risk between sexual mi-
nority and heterosexual individuals, these studies were limited
to adolescents and most did not examine pregnancy intentions
or the role of congruence across dimensions of sexual orientation
(Lindley & Walsemann, 2015; Saewyc, Bearinger, Blum, &
Resnick, 1999; Tornello, Riskind, & Patterson, 2014). We explore
whether adult SMW (aged 18–44) experience an increased risk
for unintended pregnancy compared with heterosexual women,
examine whether the risk differs depending on the congruence
of women’s sexual identities and attractions, and test possible
mediators.
Stress, Stigma, and Congruence across Sexual Orientation
Dimensions
Theoretical and empirical work on sexual minority experi-
ences provide a framework for understanding why unintended
pregnancy risk may differ between sexual minority and het-
erosexual women. According to the stress process framework,
individuals experience stress not only owing to acute events
but also owing to everyday social conditions, which can, in
turn, contribute to poor health (Mirowsky & Ross, 1989;
Pearlin, 1999). The effect can be particularly pernicious for
stigmatized minorities, including sexual minorities, as outlined
in the minority stress framework (Meyer, 2003). Those who
experience stigma-related stress may engage in risky behaviors
as a means of coping (Hatzenbuehler, Phelan, & Link, 2013).
This basic framework leads us to expect higher rates of
unintended pregnancy among SMW. This is supported by
empirical studies which find that sexual minority women are
more likely to become pregnant as adolescents than their
heterosexual peers (Charlton et al., 2013; Lindley &
Walsemann, 2015).
Not all SMW, however, experience stigma-related stress to
the same degree, so unintended pregnancy risk may vary as well.
Sexual orientation is typically defined in terms of sexual attrac-
tion, behavior, and identity, and women differ in their congru-
ence or incongruence across these dimensions. The extent to
which minority stigma translates into stress and other tangible
outcomes depends, in part, on whether one identifies as a sexual
minority and whether that identity is important to the individual
(“prominence”), whether it is integrated with other identities
(“synthesis”), and the degree of self-acceptance (“valence”;
Meyer, 2003). These identity characteristics can buffer the effects
of stigma and stress, because they may allow for greater social
support, particularly from others with the same identity. How-
ever, some individuals will not identify as lesbian, gay, or
bisexual (LGB), despite reporting some same-sex attraction,
owing to internalized homo/biphobia, fear of rejection, or
cultural differences. This incongruence across sexual orientation
dimensions (particularly individuals who identify as heterosex-
ual but report same-sex attractions or behaviors) is associated
with poorer mental health and greater engagement in health risk
behaviors (such as having intercourse under the influence of
drugs or alcohol), compared with individuals who express
congruence (Gattis, Sacco, & Cunningham-Williams, 2012;
Reback & Larkins, 2013; Trocki, Drabble, & Midanik, 2009).
Following this theoretical and empirical work, we anticipate
that, among SMW, the risk of unintended pregnancy will be
concentrated among those whose sexual identity and sexual
attractions are incongruent (i.e., heterosexual-identified women
with same-sex attractions).
We also explore possible mediators of the association
between sexual orientation congruence and unintended preg-
nancy. Scholars have theorized that minority individuals may
have poorer health-related outcomes in part because they
engage in riskier behaviors to cope with stigma-related stress,
and this assertion has been supported by empirical research
(Hatzenbuehler, McLaughlin, Keyes, & Hasin, 2010;
Hatzenbuehler et al., 2013; Ryan, Huebner, Diaz, & Sanchez,
2009). In the case of unintended pregnancy, heterosexual-
identified women with same-sex attractions (i.e., incongruent
SMW) may be more likely to have risky sex with men
(e.g., having multiple male sex partners) in response to stress and
stigma, compared with LGB-identified women with same-sex
attractions (i.e., congruent SMW), who may use different
means of handling the stress they experience. Our second
hypothesis, therefore, is that an higher risk for unintended
pregnancy among incongruent SMW would be explained
partially by a greater number of and a greater turnover of male
sexual partners and higher likelihood of early sex. These specific
factors have been linked to unintended pregnancy and may be
more common among sexual minority individuals (Heywood,
Patrick, Smith, & Pitts, 2014; Kusunoki, 2014; Magnusson,
Masho, & Lapane, 2011; Miller et al., 2010; Moore, Frohwirth, &
Miller, 2010; Tornello et al., 2014).
Material and Methods
Data
Data came from the National Survey of Family Growth (NSFG),
a nationally representative sample of the civilian, noninstitu-
tionalized, reproductive-age female population. The data were
cross-sectional, with new samples taken every year.
Respondents were interviewed face-to-face and also used self-
interview software to answer sensitive questions.
We pooled data from 2006 to 2013, which is the most recent
survey year available. The pooled sample included 11,041
women, who reported on 30,035 pregnancies. We excluded
pregnancies that occurred before the woman was 18 years old
(n¼3,872 pregnancies; 13%) because we expected pregnancy
dynamics would be quite different for younger adolescents.
In addition, 337 pregnancies (1%) were excluded because preg-
nancy intentions were either missing or reported as “indifferent”
or “don’t know.”We omitted 419 pregnancies (1%) owing to
missing information on the respondent’s sexual orientation and
4 pregnancies to women who had not had sex with a male
partner, because these pregnancies could not have been unin-
tended. The final analytic sample included 25,403 pregnancies,
of which 11,827 were unintended. The response rate was 78% for
the 2006 through 2010 surveys and 73% for the 2011 through
2013 surveys (National Center for Health Statistics, 2011, 2014).
Measures
Unintended pregnancy
Questions in the NSFG for ascertaining intention status were
coded using the standard approach (Chandra, Martinez, Mosher,
Abma, & Jones, 2005; National Center for Health Statistics, 2016).
Pregnancies were classified as “intended”if they came at the
“right time”or later than the woman wanted; they are classified
as “unintended”if the pregnancy came “too soon”or if the
woman wanted no future pregnancy. Pregnancies were the unit
of observation in the analysis and the outcome of interest was
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151146
whether each pregnancy was intended or unintended. Although
unintended pregnancy is a complex construct and subject to
misclassification, it has fairly high external validity and is widely
used (Joyce, Kaestner, & Korenman, 2002).
Sexual orientation congruence
Regarding identity, respondents were asked, “Do you think of
yourself as.Heterosexual or straight; Homosexual, gay, or
lesbian; Bisexual; Something else.”Those who answered
“Heterosexual or straight”were considered to have heterosexual
identity; all others were considered to have non-heterosexual
identities. Attraction was evaluated using the question, “People
are different in their sexual attraction to other people. Which
best describes your feelings? Are you.Only attracted to males;
Mostly attracted to males; Equally attracted to males and
females; Mostly attracted to females; Only attracted to females?”
Those who answered “Only attracted to males”were considered
to have heterosexual attraction; all others were considered to
have non-heterosexual (or “same-sex”) attraction.
Based on these two questions, respondents were classified into
3 mutually exclusive categories: heterosexual women (hetero-
sexual in both identity and attraction; n¼21,305 pregnancies),
congruent SMW (non-heterosexual in both identity and attrac-
tion; n¼1,441 pregnancies), and incongruent SMW (heterosexual
identity and non-heterosexual attraction; n¼2,657 pregnancies).
Sexual behavior is the third dimension of sexual orientation
and was dealt with in two ways. First, sexual experiences with
men were treated as mediators because these encounters are
proximate determinants of unintended pregnancy. Second, we
conducted additional analyses examining identity–behavior
congruence. These findings were nearly identical to those for
identity–attraction congruence and are discussed in the
supplementary analyses section.
Mediators
We hypothesized that an increased risk of unintended preg-
nancy among SMW with identity–attraction incongruence
would be partially explained by sexual experiences. Early first
sex with a man was categorized as first sex at age 13 or younger,
ages 14 to 15, ages 16 to 17, or no sex with a man before age 18.
We also included an indicator of the number of lifetime male
sexual partners respondents reported as well as the number of
male sexual partners respondents had in the last 12 months.
Covariates
A number of factors may confound the relationship between
sexual orientation and pregnancy intentions. Self-reported race/
ethnicity was categorized as non-Hispanic White, non-Hispanic
Black, Hispanic (any race), or non-Hispanic other. We also exam-
ined age at conception and pregnancy order (specifically whether
this was the first, second, third, or a higher order pregnancy).
Respondents’mother’s education level was categorized as less
than high school, high school or equivalent, some college, or a
bachelor’s degree or higher. We used mothers’education level
rather thanrespondents’because the two arehighly correlated and
younger women’s education had not yet been completed. House-
hold income was categorized as below the poverty line, 100% to
199% of poverty, 200% to 299% of poverty, or 300% and above.
Analysis
For all analyses, pregnancies were the unit of analysis and
probability weights were applied to account for the complex
sampling design and the fact that some women contributed
multiple pregnancies (personal communication, Daniels, K.,
March 11, 2015; National Center for Health Statistics, 2011, 2014).
To address our first hypothesis, we used multivariable logistic
regression to test whether sexual orientation was associated
with having an unintended pregnancy rather than an intended
pregnancy. First, we compared heterosexual women with all
SMW (i.e. anyone with non-heterosexual identity and/or
same-sex attraction). Then we examined the role of identity–
attraction congruence. Both models controlled for possible
confounders, including age at conception, race/ethnicity,
pregnancy order, mothers’education, and household income.
To address the second hypothesis, we examined whether
sexual experiences mediated the relationship between sexual
orientation congruence and the odds of having an unintended
pregnancy. These mediators included early first sex and the
number of male partners (recent and lifetime). These variables
were added sequentially and the models used the same set of
control variables listed previously.
Results
Descriptive statistics for the weighted sample are presented
by sexual orientation congruence in Table 1. The mean age at
conception was 25.3 and most pregnancies were first (31%) or
second pregnancies (29%). Fifty-seven percent of reported
pregnancies were to non-Hispanic White women. Most
pregnancies were to women whose mothers had a high school-
level education (33%) or below (29%). It was most common for
household incomes to be below poverty (29%) or to be in the
highest income category (29% had incomes above 300% of
poverty).
Table 2 presents coefficients from weighted multivariable
logistic regression models predicting whether a pregnancy was
unintended (rather than intended). We tested whether SMW
were at elevated risk for experiencing an unintended pregnancy
(model 1) and the role of identity–attraction congruence (model
2). SMW overall were at higher risk for unintended pregnancy
than heterosexual women (odds ratio [OR], 1.26; p<.01; model
1), and SMW with identity–attraction incongruence, in partic-
ular, were at increased risk for unintended pregnancy (OR, 1.28;
p<.01; model 2).
Table 3 examined the role of sexual experiences in mediating
the association between sexual orientation congruence and un-
intended pregnancy. We found that the coefficient for SMW with
identity–attraction incongruence was attenuated after adjusting
for early first sex (OR, 1.22; p<.01; model 2). When lifetime and
recent male sexual partners were entered into the model, the
coefficient for incongruent SMW was reduced to nonsignificance
(OR, 1.08; p>.05; model 3).
Supplemental Analyses
As with other U.S. data sources that rely on women’s
self-reports of pregnancies, abortions in the NSFG are under-
reported (Jones & Kost, 2007). Because we were interested in
whether this underreporting biased our results, we examined if
the association between sexual orientation and unintended
pregnancy risk differed for abortions versus other types of
unintended pregnancy outcomes (see Appendix A). When
examined separately, both unintended pregnancies ending in
abortion and unintended pregnancies with other outcomes
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151 147
(e.g., birth) were more common for SMW than heterosexual
women, which supports our findings.
We also wanted to test whether our results would differ if we
examined identity–behavior congruence (rather than identity–
attraction congruence). Following the analytic approach used by
Gattis et al. (2012), we repeated our analyses using a set of
dummy variables for identity–behavior congruence. In these
analyses, respondents were considered “SMW congruent”if they
identified as something other than heterosexual and had at
least some experience with a female sex partner, whereas
respondents were considered “SMW incongruent”if they iden-
tified as heterosexual but had a female sex partner. The results
for identity–behavior congruence (presented in Appendix B)
were fully consistent with the pattern of results for identity–
attraction congruence (presented in Tables 2 and 3) with one
exception: the odds ratio for identity–behavior incongruence
remained significant when the number of male sex partners was
included in the model.
Discussion
Unintended pregnancy is an important indicator of popula-
tion health because it increases the risk of poor physical and
mental health for mothers and children. In this paper, we
showed that adult SMWdparticularly SMW who were incon-
gruent, with heterosexual identities but some same-sex attrac-
tiondwere more likely to have unintended pregnancies,
compared with heterosexual women. These differences in
unintended pregnancy risk were tied to the fact that SMW with
identity–attraction incongruence had more male sex partners
than their heterosexual counterparts.
Our findings fit with prior research on adolescent pregnancy
among SMW and extend it in important ways. Using multiple
dimensions of sexual orientation, Charlton et al., 2013 found that
adolescent sexual minorities were more likely to experience
early pregnancies than their heterosexual peers. Tornello et al.
(2014) showed that adolescents who identified as bisexual
were more likely to report terminating a pregnancy, and a third
study demonstrated higher rates of pregnancy among New York
City high school students who identified as sexual minorities or
had same-sex partners (Lindley & Walsemann, 2015). Our work
builds on this prior research by expanding the age range under
consideration to all adult reproductive-age women (18–44 years)
and by showing disparities in unintended pregnancy, rather than
all pregnancies or terminated pregnancies. Further, we examined
the role of congruence across dimensions of sexual orientation,
which most studies have not (Charlton et al., 2013 is one
exception). We found that adult SMW experienced disparities in
unintended pregnancy risk, and these increased risks were
concentrated, specifically, among women who identified as
heterosexual but had some same-sex attraction.
We also found support for our hypothesis that unintended
pregnancy risk would be mediated by sexual experiences.
Specifically, we found that the elevated risk of unintended
pregnancy among SMW with identity–attraction incongruence
was explained by their greater average number of male sexual
partners (and, to a lesser extent, earlier first sex), compared with
heterosexual women. Previous research by Tornello et al. (2014)
had found that lesbian- and bisexual-identified teenagers had
more male and female sexual partners than their heterosexual-
identified peers, and other research indicated that the same
pattern existed among young women ages 25 to 34 (Lindley,
Walsemann, & Carter, 2013). In addition, research by Saewyc
et al. (1999) found that adolescents in Minnesota who identi-
fied as lesbian or bisexual were more likely to have had early first
sex than those who identified as heterosexual. Our study
Table 1
Background Characteristics (Pregnancies), National Survey of Family Growth 2006–2013 (Weighted)
All Pregnancies Pregnancies to Women Who Are
Heterosexual
*
SMW Congruent
y
SMW Incongruent
z
Age (yrs) of respondent at conception (mean) 25.3 25.5 23.5 24.8
Race/ethnicity of respondent (proportion)
Non-Hispanic White 0.57 0.56 0.66 0.65
Non-Hispanic Black 0.16 0.17 0.18 0.12
Hispanic 0.20 0.21 0.11 0.15
Non-Hispanic other 0.06 0.06 0.05 0.08
Pregnancy order (proportion)
1st pregnancy 0.31 0.31 0.28 0.33
2nd pregnancy 0.29 0.29 0.26 0.27
3rd pregnancy 0.19 0.20 0.18 0.18
4th or higher 0.21 0.20 0.27 0.21
Respondent’s mother’s education (proportion)
Less than high school 0.29 0.30 0.23 0.18
HS/GED 0.33 0.33 0.36 0.33
Some college 0.21 0.20 0.24 0.30
BA or higher 0.16 0.15 0.15 0.18
No mother figure 0.01 0.01 0.02 0.01
Respondent’s household income (proportion)
Below 100% poverty level 0.29 0.28 0.40 0.29
100–199% of poverty 0.24 0.25 0.21 0.23
200–299% of poverty 0.18 0.18 0.18 0.19
300% of poverty 0.29 0.30 0.22 0.29
n25,403 21,305 1,441 2,657
Abbreviations: BA, bachelor’s degree; HS, high school; SMW, sexual minority women.
*
Heterosexual (identity and attraction).
y
Non-heterosexual (identity and attraction).
z
Heterosexual identity, non-heterosexual attraction.
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151148
suggests that these differences in sexual experiences may be one
of the key mechanisms shaping the higher risk of unintended
pregnancy among adult SMW with identity–attraction incon-
gruence. Although incongruent SMW engaged in these behaviors
at higher rates than other women, we were unable to examine
the reasons for this directly. Qualitative work may be useful for
addressing this question in the future.
There are several limitations in the analysis. First, pregnancies
occurred in the past, whereas sexual orientation was measured
at the time of the survey. Unfortunately, there is no nationally
representative dataset that includes both the intention status of
pregnancies and women’s sexual orientation at the time of those
pregnancies. Assuming that we have misattribution in both
directions (i.e., women who had non-heterosexual identity at the
time of the survey but heterosexual identity at the time of the
pregnancy, and vice versa), it would have the effect of adding
noise to the data, such that our reported results would be con-
servative. Ideally, future studies would collect data on sexual
Table 3
Pregnancy Intention Status Regressed on Identity–Behavior Congruence with Sexual Experience Mediators. Odds Ratios from Logistic Regressions, National Survey of
Family Growth 2006–2013 (Weighted) (n¼25,403)
Baseline Early First Sex Male Partners Full Model
Model 1 Model 2 Model 3 Model 4
OR 95% CI OR 95% CI OR 95% CI OR 95% CI
SMW, by congruence (ref ¼heterosexual
*
)
SMW Congruent
y
1.21 0.94–1.56 1.14 0.88–1.47 1.01 0.78–1.30 1.00 0.77–1.29
SMW Incongruent
z
1.28
**
1.08–1.51 1.22
*
1.03–1.43 1.08 0.93–1.26 1.07 0.92–1.25
Age at first sex with a male (ref ¼not before age 18)
Age 13 or younger 1.51
**
1.19–1.92 1.14 0.88–1.48
Age 14–15 1.55
**
1.35–1.77 1.26
**
1.10–1.46
Age 16–17 1.27
**
1.12–1.43 1.15
*
1.02–1.30
No. of male sex partners in past year 1.04 0.99–1.08 1.04 1.00–1.08
No. of male sex partners in lifetime 1.04
**
1.03–1.05 1.04
**
1.03–1.04
Abbreviation: SMW, sexual minority women.
*
p<.05,
**
p<.01
Note: The dependent variable ¼1 if the pregnancy is unintended (rather than intended). Models include controls for age at conception, race-ethnicity, pregnancy order,
mother’s education, and household income.
*
Heterosexual (identity and attraction).
y
Non-heterosexual (identity and attraction).
z
Heterosexual identity, non-heterosexual attraction.
Table 2
Pregnancy Intention Status Regressed on Identity–Attraction Concordance. Odds Ratios from Logistic Regressions, National Survey of Family Growth 2006–2013
(Weighted) (n¼25,403)
Model 1 Model 2
OR 95% CI OR 95% CI
SMW
*,y
(ref ¼heterosexual
z
) 1.26
**
1.08–1.46
SMW, by congruence (ref ¼heterosexual
z
)
SMW congruent
*
1.21 0.94–1.56
SMW incongruent
y
1.28
**
1.08–1.51
Controls
Age at conception 0.87
**
0.86–0.89 0.87
**
0.86–0.89
Non-Hispanic White (ref)
Non-Hispanic Black 1.88
**
1.65–2.16 1.88
**
1.65–2.16
Hispanic 1.11 0.94–1.31 1.11 0.94–1.31
Non-Hispanic Other 0.88 0.70–1.11 0.88 0.70–1.11
1st pregnancy (ref)
2nd pregnancy 0.80
**
0.71–0.89 0.80
**
0.71–0.89
3rd pregnancy 1.13 0.99–1.29 1.13 0.99–1.29
4th or higher pregnancy 1.80
**
1.49–2.17 1.80
**
1.49–2.18
Mom less than HS (ref)
Mom HS/GED 1.08 0.94–1.25 1.08 0.94–1.25
Mom some college 1.24
**
1.06–1.46 1.24
**
1.06–1.45
Mom BA or higher 1.13 0.93–1.36 1.12 0.93–1.36
No mother figure 1.48 0.97–2.25 1.48 0.97–2.25
Below 100% of poverty (ref)
100–199% of poverty 0.99 0.86–1.13 0.99 0.86–1.13
200–299% of poverty 0.84
*
0.72–0.98 0.84
*
0.72–0.98
300% of poverty þ0.81
**
0.70–0.94 0.81
**
0.70–0.94
Abbreviations: BA, bachelor’s degree; CI, confidence interval; HS, high school; OR, odds ratio; SMW, sexual minority women.
Note: The dependent variable ¼1 if the pregnancy is unintended (rather than intended).
*
p<.05;
**
p<.01
*
Non-heterosexual (identity and attraction).
y
Heterosexual identity, non-heterosexual attraction.
z
Heterosexual (identity and attraction).
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151 149
orientation, pregnancy intentions, and actual pregnancies
prospectively.
Second, approximately 11% of pregnancies experienced by
NSFG respondents were unreported because of respondents’
reluctance to discuss prior abortions.
1
In our supplementary
analyses, we found that both unintended pregnancies ending in
abortion and unintended pregnancies with other outcomes
(e.g., birth, miscarriage) were more common for SMW than
heterosexual women, which supports our findings. For the
NSFG’s“missing abortions”to invalidate the results in Tables 2
and 3, heterosexual women would have to have been more
likely to underreport their abortions than SMW. This is unlikely,
because economically or socially disadvantaged women are most
likely to underreport (Jones & Kost, 2007). Thus, if any difference
in abortion reporting occurred, it is likely that SMW, a socially
disadvantaged group, would have been more prone to under-
reporting than heterosexual women, which would make our
findings conservative. For the future, we suggest that questions
related to sexual identity and attraction be added to clinic-based
surveys of abortion patients to empirically identify the extent of
abortion underreporting for SMW.
Finally, we were not able to identify women who became
pregnant via artificial insemination; instead, we were only able
to exclude pregnancies to women who had never had sex with a
man. Ideally, we would have excluded pregnancies from artificial
insemination to limit our analysis to a more homogeneous group
of pregnancies. Based on the available data, however, there
seems to be little difference between heterosexual and SMW in
receipt of reproductive assistance. Specifically, the proportion of
pregnancies to women who had ever received medical help to
get pregnant was very similar for heterosexual women and SMW
in the NSFG (results available upon request).
Implications for Practice and/or Policy
Although SMW have been largely ignored in previous preg-
nancy prevention efforts, our findings indicate that they are in
need of comprehensive reproductive health care to reduce their
risk of unintended pregnancy. Reproductive health providers
should be aware that those at highest riskdthat is, women with
heterosexual identity and same-sex attractiondwill be difficult
to identify. Thus, providers should routinely ask women,
regardless of their sexual identity, about their sexual attractions
and behaviors with male and female partners, as well as about
their pregnancy intentions, and provide information and
services accordingly. Many SMW will likely need contraceptive
counseling and other preventive services, including human
papillomavirus vaccinations, sexually transmitted infection
testing, and screenings for interpersonal violence, as well as
referrals for LGBT support services. Moreover, many providers
may want or need training in how to provide culturally compe-
tent care to these women, particularly in how to raise and discuss
issues regarding sexual orientation. Last, additional research is
needed to explore the diverse experiences of SMW regarding
their sexual, reproductive, and contraceptive needs, as well as
their preferred ways to engage providers on these issues. Such
information will be vital to developing pregnancy prevention
interventions that are effective in helping SMW.
Acknowledgments
The authors thank Debra Umberson for feedback on an earlier
draft and Brittany Pittelli for editing assistance.
References
Barber, J. S., Axinn, W. G., & Thornton, A. (1999). Unwanted childbearing, health,
and mother-child relationships. Journal of Health and Social Behavior, 40(3),
231–257.
Chandra, A., Martinez, G. M., Mosher, W. D., Abma, J. C., & Jones, J. (2005).
Fertility, family planning, and reproductive health of US women: Data from the
2002 National Survey of Family Growth. Hyattsville, MD: National Center for
Health Statistics.
Charlton, B. M., Corliss, H. L., Missmer, S. A., Rosario, M., Spiegelman, D., &
Austin, S. B. (2013). Sexual orientation differences in teen pregnancy and
hormonal contraceptive use: An examination across 2 generations. American
Journal of Obstetrics and Gynecology, 209(3), 204.e1–204.e8.
Finer, L. B., & Zolna, M. R. (2016). Declines in unintended pregnancy in the United
States, 2008–2011. New England Journal of Medicine, 374(9), 843–852.
Gattis, M. N., Sacco, P., & Cunningham-Williams, R. M. (2012). Substance use and
mental health disorders among heterosexual identified men and women
who have same-sex partners or same-sex attraction: Results from the
National Epidemiological Survey on Alcohol and Related Conditions. Archives
of Sexual Behavior, 41(5), 1185–1197.
Gipson, J. D., Koenig, M. A., & Hindin, M. J. (2008). The effects of unintended
pregnancy on infant, child, and parental health: A review of the literature.
Studies in Family Planning, 39(1), 18–38.
Hatzenbuehler, M. L., McLaughlin, K. A., Keyes, K. M., & Hasin, D. S. (2010). The
impact of institutional discrimination on psychiatric disorders in lesbian,
gay, and bisexual populations: A prospective study. American Journal of
Public Health, 100(3), 452–459.
Hatzenbuehler, M. L., Phelan, J. C., & Link, B. G. (2013). Stigma as a fundamental
cause of population health inequalities. American Journal of Public Health,
103(5), 813–821.
Heywood, W., Patrick, K., Smith, A. M. A., & Pitts, M. K. (2014). Associations
between early first sexual intercourse and later sexual and reproductive
outcomes: A systematic review of population-based data. Archives of Sexual
Behavior, 44(3), 531–569.
Jones, R. K., & Kost, K. (2007). Underreporting of induced and spontaneous
abortion in the United States: An analysis of the 2002 National Survey of
Family Growth. Studies in Family Planning, 38(3), 187–197.
Joyce, T., Kaestner, R., & Korenman, S. (2002). On the validity of retrospective
assessments of pregnancy intention. Demography, 39(1), 199–213.
Kane, J. B., Morgan, S. P., Harris, K. M., & Guilkey, D. K. (2013). The Educational
consequences of teen childbearing. Demography, 50(6), 2129–2150.
Korenman, S., Kaestner, R., & Joyce, T. (2002). Consequences for infants of
parental disagreement in pregnancy intention. Perspectives on Sexual and
Reproductive Health, 34(4), 198–205.
Kost, K., Landry, D. J., & Darroch, J. E. (1998a). Predicting maternal behaviors
during pregnancy: Does intention status matter? Family Planning Perspec-
tives, 30(2), 79–88.
Kost, K., Landry, D. J., & Darroch, J. E. (1998b). The effects of pregnancy planning
status on birth outcomes and infant care. Family Planning Perspectives, 30(5),
223–230.
Kusunoki, Y. (2014). Relationship dynamics and contraception: The role of
seriousness and instability. Presented at the Population Association of
America 2014 Annual Meeting, Boston, MA. Available: http://paa2014.
princeton.edu/papers/141382. Accessed: September 1, 2016.
Lindley, L. L., & Walsemann, K. M. (2015). Sexual orientation and risk of preg-
nancy among New York City high-school students. American Journal of Public
Health, 105(7), 1379–1386.
Lindley, L. L., Walsemann, K. M., & Carter, J. W. (2013). Invisible and at risk: STDs
among young adult sexual minority women in the United States. Perspectives
on Sexual and Reproductive Health, 45(2), 66–73.
Magnusson, B. M., Masho, S. W., & Lapane, K. L. (2011). Early age at first inter-
course and subsequent gaps in contraceptive use. Journal of Women’s Health,
21(1), 73–79.
Meyer, I. H. (2003). Prejudice, social stress, and mental health in lesbian, gay, and
bisexual populations: Conceptual issues and research evidence. Psychological
Bulletin, 129(5), 674.
1
One of the few studies with more complete data (relying on the NSFG,
abortion provider surveys, and other data) found that approximately 45%
of U.S. pregnancies are unintended, and that 42% of these unintended pregnan-
cies result in abortion (Finer & Zolna, 2016). Comparing these more complete
numbers to the NSFG, researchers have determined that approximately 60% of
abortions are “missing”from the NSFG (Jones & Kost, 2007). Therefore, the
NSFG lacks information on approximately 11% of the pregnancies actually
experienced by respondents (0.45 0.42 0.60 ¼.11).
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151150
Miller, E., Decker, M. R., McCauley, H. L., Tancredi, D. J., Levenson, R. R.,
Waldman, J., .Silverman, J. G. (2010). Pregnancy coercion, intimate partner
violence and unintended pregnancy. Contraception, 81(4), 316–322.
Mirowsky, J., & Ross, C. E. (1989). Social causes of psychological distress.
Hawthorne, NY: Aldine de Gruyter.
Moore, A. M., Frohwirth, L., & Miller, E. (2010). Male reproductive control of
women who have experienced intimate partner violence in the United
States. Social Science & Medicine, 70(11), 1737–1744 .
Myhrman, A. (1988). The Northern Finland Cohort, 1966–82: A follow-up study
of children unwanted at birth. In David, H. P., Dytrych, Z., Matjcek, Z., &
Schiiller, V. (Eds.), Born unwanted: Developmental effects of denied abortion
(pp. 103–110). New York: Springer, .
Najman, J. M., Morrison, J., Williams, G., Andersen, M., & Keeping, J. D. (1991). The
mental health of women 6 months after they give birth to an unwanted
baby: A longitudinal study. Social Science & Medicine, 32(3), 241–247.
National Center for Health Statistics. (2011). User’s guide. Public use data file
documentation 2006–2010 National Survey of Family Growth. Hyattsville, MD:
Centers for Disease Control and Prevention, National Center for Health
Statistics. Available: http://www.cdc.gov/nchs/data/nsfg/NSFG_2006-2010_
UserGuide_MainText.pdf. Accessed: October 14, 2014.
National Center for Health Statistics. (2014). User’s guide. Public use data file
documentation 2011–2013 National Survey of Family Growth. Hyattsville, MD:
Centers for Disease Control and Prevention, National Center for Health
Statistics. Available: http://www.cdc.gov/nchs/data/nsfg/NSFG_2011-2013_
UserGuide_MainText.pdf. Accessed: March 24, 2015.
National Center for Health Statistics. (2016). NSFG - 2011-2013 NSFG - Public use
data files, codebooks and documentation. Available: http://www.cdc.gov/
nchs/nsfg/nsfg_2011_2013_puf.htm. Accessed: September 1, 2016.
Pearlin, L. I. (1999). The stress process revisited. In Aneshensel, C. S., &
Phelan, J. C. (Eds.), Handbook of the sociology of mental health (pp. 395–415).
New York: Springer, .
Reback, C. J., & Larkins, S. (2013). HIV risk behaviors among a sample of het-
erosexually identified men who occasionally have sex with another male
and/or a transwoman. Journal of Sex Research, 50(2), 151–163.
Ryan, C., Huebner, D., Diaz, R. M., & Sanchez, J. (2009). Family rejection as a
predictor of negative health outcomes in White and Latino lesbian, gay, and
bisexual young adults. Pediatrics, 123(1), 346–352.
Saewyc, E. M., Bearinger, L. H., Blum, R. W., & Resnick, M. D. (1999). Sexual in-
tercourse, abuse and pregnancy among adolescent women: Does sexual
orientation make a difference? Family Planning Perspectives, 31(3), 127–131.
Singh, S., Sedgh, G., & Hussain, R. (2010). Unintended pregnancy: Worldwide
levels, trends, and outcomes. Studies in Family Planning, 41(4), 241–250.
Tornello, S. L., Riskind, R. G., & Patterson, C. J. (2014). Sexual orientation and
sexual and reproductive health among adolescent young women in the
United States. Journal of Adolescent Health, 54(2), 160–168.
Trocki, K. F., Drabble, L. A., & Midanik, L. T. (2009). Tobacco, marijuana, and
sensation seeking: Comparisons across gay, lesbian, bisexual, and hetero-
sexual groups. Psychology of Addictive Behaviors, 23(4), 620–631.
Author Descriptions
Caroline Sten Hartnett, PhD, is an Assistant Professor in the Sociology Department
at the University of South Carolina. Her research examines how social factors
influence decision making around childbearing, including attitudes toward
children, parity desires, and unintended pregnancy.
Lisa L. Lindley, DrPH, MPH, CHES, is Associate Professor, Department Global and
Community Health, George Mason University. Her research interests are in sexual
health promotion and the prevention of HIV, STIs, and unintended pregnancies
among adolescents/young adults and sexual minority populations.
Katrina M. Walsemann, PhD, MPH, is Associate Professor, Department of Health
Promotion, Education, and Behavior, University of South Carolina. Her research
focuses on understanding how social inequalities (e.g., race, class, gender, sexual
minority status) impact health across the life course.
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151 151
Appendix A
Appendix Table A.1
Proportion of Pregnancies Unintended, by Sexual Orientation Congruence and Pregnancy Outcome: National Survey of Family Growth 2006–2013 (Weighted;
n¼25,403)
Proportion Unintended, When Unintended Pregnancies Are Limited to Those Ending in.
Abortions Births and Other Outcomes
x
Heterosexual
*
0.10 0.36
All SMW
y,z
0.19
k
0.42
k
SMW congruent
y
0.21
k
0.46
k
SMW incongruent
z
0.19
k
0.41
k
Total 0.12 0.37
Abbreviation: SMW, sexual minority women.
*
Heterosexual (identity and attraction).
y
Non-heterosexual (identity and attraction).
z
Heterosexual identity, non-heterosexual attraction.
x
Includes all pregnancies besides those ending in abortion, including miscarriages, stillbirths, ectopic pregnancies and current pregnancies.
k
The proportion differs from the heterosexual category at the p<.05 level.
Appendix B: Supplemental Analyses of Identity–Behavior Congruence
Appendix Table B.1
Pregnancy Intention Status Regressed on Behavior Congruence: Odds Ratios From Logistic Regressions, National Survey of Family Growth 2006–2013 (Weighted)
(n¼24,983)
Model 1 Model 2
OR 95% CI OR 95% CI
All SMW
y,z
(ref ¼heterosexual
*
) 1.57
**
1.31–1.87
SMW, by behavior congruence (ref ¼heterosexual
*
)
SMW congruent
y
1.31 0.99–1.73
SMW incongruent
z
1.69
**
1.39–2.06
Controls
Age at conception 0.88
**
0.86–0.89 0.88
**
0.86–0.89
Non-Hispanic White (ref)
Non-Hispanic Black 1.88
**
1.64–2.16 1.88
**
1.64–2.15
Hispanic 1.16 0.98–1.38 1.16 0.98–1.37
Non-Hispanic other 0.90 0.70–1.16 0.90 0.70–1.16
1st pregnancy (ref)
2nd pregnancy 0.79
**
0.70–0.88 0.79
**
0.70–0.88
3rd pregnancy 1.11 0.97–1.27 1.11 0.97–1.27
4th or higher pregnancy 1.76
**
1.46–2.13 1.76
**
1.46–2.13
Mom less than HS (ref)
Mom HS/GED 1.08 0.93–1.24 1.08 0.93–1.24
Mom some college 1.25
**
1.07–1.45 1.24
**
1.07–1.45
Mom BA or higher 1.12 0.93–1.35 1.12 0.94–1.35
No mother figure 1.41 0.92–2.17 1.42 0.92–2.17
Below 100% of poverty (ref)
100%–199% of poverty 0.99 0.87–1.13 0.99 0.86–1.12
200%–299% of poverty 0.84
*
0.72–0.98 0.83
*
0.71–0.98
300% of poverty þ0.82
**
0.70–0.95 0.82
**
0.70–0.95
Abbreviations: BA, bachelor’s degree; CI, confidence interval; HS, high school; OR, odds ratio; SMW, sexual minority women.
*
p<.05;
**
p<.01
Note: The dependent variable ¼1 if the pregnancy is unintended (rather than intended).
*
Heterosexual (identity and behavior).
y
Non-heterosexual (identity and behavior).
z
Heterosexual identity, non-heterosexual behavior.
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151151.e1
Appendix Table B.2
Pregnancy Intention Status Regressed on Behavior Congruence, with Sexual Experience Mediators: Odds Ratios from Logistic Regressions, National Survey of Family
Growth 2006–2013 (Weighted) (n¼24,983)
Baseline Early First Sex Male Partners Full Model
Model 1 Model 2 Model 3 Model 4
OR 95% CI OR 95% CI OR 95% CI OR 95% CI
SMW, by behavior congruence (ref ¼heterosexual
*
)
SMW congruent
y
1.31 0.99–1.73 1.22 0.93–1.60 1.03 0.78–1.36 1.02 0.77–1.34
SMW incongruent
z
1.69
**
1.39–2.06 1.57
**
1.29–1.91 1.29
**
1.07–1.56 1.27
*
1.05–1.53
Age at first sex with a male (ref ¼Not before age 18)
Age 13 or younger 1.43
**
1.13–1.81 1.11 0.86–1.43
Age 14–15 1.51
**
1.31–1.73 1.26
**
1.09–1.45
Age 16–17 1.25
**
1.10–1.41 1.14
*
1.01–1.29
No. of male sex partners in past year 1.04 0.99–1.08 1.04 1.00–1.08
No. of male sex partners in lifetime 1.04
**
1.03–1.05 1.03
**
1.03–1.04
Abbreviations: CI, confidence interval; HS, high school; OR, odds ratio; SMW, sexual minority women.
Note: The dependent variable ¼1 if the pregnancy is unintended (rather than intended). Includes controls for age at conception, race-ethnicity, pregnancy order,
mother’s education, and household income.
*
p<.05,
**
p<.01
*
Heterosexual (identity and behavior).
y
Non-heterosexual (identity and behavior).
z
Heterosexual identity, non-heterosexual behavior.
C.S. Hartnett et al. / Women's Health Issues 27-2 (2017) 145–151 151.e2
