Sexual Orientation Disparities in Preconception Health

Abstract

Objective: In the United States, there have been very few improvements in adverse birth outcomes, such as infant mortality, low birthweight, and preterm birth in recent years. Health promotion before pregnancy (e.g., preconception care) has been increasingly recognized as an important strategy by which to improve these reproductive outcomes. As of yet, no research has examined sexual orientation disparities in preconception health which has important implications for birth outcomes in the United States, since sexual minority women (SMW) are more likely to report stillbirths, low birthweight, and preterm infants than heterosexual women.

Methods: This study addresses this gap by utilizing data from the National Longitudinal Study of Adolescent to Adult Health (Add Health) to examine sexual orientation disparities in women's preconception health 1 and 3 years before a live birth (n = 3,133).

Results: Our findings suggest that, even after controlling for maternal characteristics, SMW are more likely to report adverse health conditions and behaviors before pregnancy relative to heterosexual women 1 year before the survey, including higher odds of binge drinking, other substance use, having a sexually transmitted infection diagnosis, and depression.

Conclusions: Despite new public health policies aimed at improved preconception health, our findings suggest that SMW are even more vulnerable to poor preconception health than their heterosexual counterparts, which has important implications for maternal and child health. This study provides important evidence for the need to invest in the reproductive health of SMW, particularly in the context of pregnancy.

Introduction

In the United States, adverse birth outcomes, such as infant mortality, low birthweight, and preterm birth, have stagnated in recent years.^1–4 In fact, recently the United States has actually seen increases in maternal mortality.⁵ Although research has emphasized the importance of prenatal care for improving maternal and child health,^6–8 it has been argued that intervening during pregnancy is not early enough in the reproductive process to ensure optimal maternal and child health outcomes.⁹ In 2006, the Centers for Disease Control established the select panel on preconception health to address the health of women before pregnancy.¹⁰ Since its establishment, multiple studies have documented that health status and behaviors before pregnancy, such as sexually transmitted infections (STIs),¹¹ elevated body mass index (BMI),^12,13 tobacco, alcohol, and drug use,¹⁴ and poor mental health,¹⁵ are associated with adverse birth outcomes.^16–18

Although some research has examined disparities related to preconception health based on race/ethnicity¹⁹ and disability,²⁰ to our knowledge, no research has examined sexual orientation disparities in preconception health. It is well documented that in the general population, sexual minority women (SMW) (e.g., women who do not identify as “exclusively heterosexual” or report same-sex attraction or same-sex sexual or romantic relationships) are more likely to report several health behaviors and conditions that may compromise maternal and infant health if they occur during the preconception or prenatal periods. These include increased risks of hazardous drinking,²¹ tobacco use,²² and marijuana as well as other drug use^23,24 among SMW compared with heterosexual women. SMW are also more likely to report an STI diagnosis,²⁵ as well as disordered eating and obesity,^26,27 than heterosexual women. In addition, mental health disparities have been observed based on sexual orientation, with lesbian and bisexual women exhibiting a higher prevalence of anxiety disorders, post-traumatic stress disorder, major depression, and higher rates of suicide attempts relative to heterosexual women.^28–30 The lack of research related to sexual orientation disparities in health during the preconception period is particularly troubling given that new research shows SMW are more likely to report stillbirths, low birthweight, and preterm births than heterosexual women.³¹

The Office of Disease Prevention and Health Promotion released its goals for Healthy People 2020, including specific but separate goals for improving maternal, infant, and child health on the one hand, and sexual and gender minority health on the other hand. Despite these goals being discrete, understanding the maternal and infant health of SMW during the preconception period is of critical importance for several reasons. First, as stated previously, SMW are more likely to report adverse birth outcomes than heterosexual women.³¹ Second, SMW report more unmet medical needs,^32–34 are less likely to have health insurance,³² and have lower rates of utilization of reproductive health services.^32,35 In addition, SMW who do utilize health care services are often given incomplete or inaccurate sexual health information.^36,37 Third, SMW have an increased risk of unintended pregnancy,^38,39 which research suggests is associated with delayed prenatal care and low birthweight infants.^40–43 Pregnancy planning allows women time to address health problems and health behaviors during the preconception period, including behavioral risk factors such as smoking or drinking, as well as mental and physical health conditions.⁴¹

Identifying disparities in preconception health is the first step to establishing whether SMW may benefit from targeted public health interventions to improve health before a pregnancy and birth. This study addresses this goal by using longitudinal data from the National Longitudinal Study of Adolescent to Adult Health (Add Health) to examine sexual orientation disparities in women's preconception health 1 and 3 years before a live birth. Because health behaviors are reported prospectively, this study eliminates recall-related biases, allowing us to measure women's health status before a pregnancy begins.

Methods

Data

Data came from the Add Health. The initial Add Health sample was drawn from 80 high schools and 52 middle schools throughout the United States with unequal probabilities of selection.^44,45 A subsample of students (n = 20,747) were asked to complete additional in-home interviews and were contacted for follow-up interviews between 2001 and 2002 (Wave III) as well as 2007 and 2008 (Wave IV). Between 2016 and 2018 (Wave V), respondents were contacted again for follow-up interviews. A subsample of Wave V responses was released in 2018; thus, women who reported pregnancies in Wave V were also eligible to be in the sample. Response rates for this survey were 77.4% for Wave III and 80.3% for Wave IV. Response rates for Wave V have not yet been released.

Analytic sample

Our sample was restricted to women who (i) were not pregnant at the time of the interview, (ii) gave birth <3 years and/or <1 year after a wave of data collection, and (iii) were at least 18 years old at the time of birth. A total of 3,133 women who were not currently pregnant at the time of the interview responded to a survey 3 years before a live birth, and 862 women responded to a survey 1 year before a live birth. That is, 862 women answered health behavior and health condition questions before they knew they were pregnant and within 1 year of giving birth. Analytic sample sizes vary across outcome variables and are reported in tables, ranging from 2,950 to 3,128 for births in the 3-year window and from 744 to 838 for the 1-year window due to differences in response rates to specific items. Disordered eating and insurance status in the preconception period have an analytic sample size that falls outside this range because questions on disordered eating were only asked at Waves I and III, and insurance status was only assessed at Waves III and IV (once respondents were over the age of 18 years), excluding a larger proportion of women from our analytic time frame.

Measures

All measures were assessed at multiple waves. To reflect correct time ordering, respondents were assigned the value for each variable that most closely preceded the birth of interest. That is, if a woman answered a question about her tobacco use in Wave III at age 24 years and reported a birth at Wave IV that ended when she was 25 years of age, her health behavior measures would come from Wave III of the survey.

Sexual identity was measured using a survey item that asked respondents, “Please choose the description that best fits how you think about yourself: 100% heterosexual (straight); mostly heterosexual (straight), but somewhat attracted to people of your own sex; bisexual, that is, attracted to men and women equally; mostly homosexual (gay), but somewhat attracted to people of the opposite sex; 100% homosexual (gay); or not sexually attracted to either males or females.” We use the sexual identity reported before the birth.

Supplementary analyses were conducted to assess differences in preconception health across sexual identities; however, no differences were detected between the three groups of SMW: mostly heterosexual, bisexual, and gay/lesbian women. Therefore, due to small sample sizes for SMW, we created a dichotomous measure that captures whether, before pregnancy, a woman identified as exclusively heterosexual or as a sexual minority (gay, bisexual, or mostly heterosexual).

Supplementary analyses were conducted to assess differences within the sexual minority population and are discussed in the results section. Moreover, we acknowledge that although sexual identity is not a comprehensive measure of sexual orientation, which also includes dimensions of attraction and sexual/romantic relationships, this measure provides a good starting point for assessing sexual orientation disparities in preconception health.

Outcome variables

Marijuana use was measured as a dichotomous variable that captured whether respondents reported any marijuana use in the past 30 days. Other substance use was measured as a dichotomous variable that captured whether respondents reported using LSD, PCP, ecstasy, mushrooms, speed, ice, heroin, or pills without a doctor's prescription in the past 30 days (yes = 1; no = 0).

Binge drinking was derived from a survey item that asked respondents, “Over the past 12 months, how many days did you drink four (Wave I and III)/five (Wave IV) or more drinks in a row?” A dichotomous measure was constructed that captured whether respondents reported one or more binge drinking events in the prior year (yes = 1) or not (no = 0).

Tobacco use was captured whether respondents were never smokers (referent); former regular smokers; current, but not regular smokers; or current regular smokers (e.g., smoking ≥27 days in a single month).

STI diagnosis was measured using a survey item that asked respondents whether they had received a diagnosis of chlamydia, gonorrhea, or syphilis from a doctor. At Waves I and III, respondents were asked whether this occurred in the past 12 months (yes = 1, no = 0). At Wave IV, 12 months was replaced by “ever.”

BMI was measured using height and weight measurements completed by the survey interviewer. Respondents who had a BMI of ≥25 kg/m² were coded as overweight or obese (yes = 1, no = 0).

Disordered eating was derived from a series of questions that asked respondents whether, in the past 7 days, they had engaged in the following behaviors as a way to lose weight or avoid gaining weight: (i) making oneself vomit, (ii) taking diet pills, and/or (iii) taking laxatives. Respondents who reported yes to any of the behaviors were coded as having disordered eating (yes = 1, no = 0). These questions were only assessed at Waves I and III.

Suicidality was derived from a measure that asked respondents, “During the past 12 months, did you ever seriously think about committing suicide?” A dichotomous measure was created that captured respondents who reported any suicidal ideation in the past 12 months (yes = 1) or not (no = 0).

Self-rated health was derived from a subjective measure of health in which respondents were asked to rate their general health on a 5-point scale that ranged from excellent (1) to poor (5).

Depressive symptoms were measured as a continuous variable using the Center for Epidemiological Studies Depression 5 Item Scale (CES-D), which asks how often a particular feeling was experienced in the past 7 days.⁴⁶ Values were summed across questions and ranged from 0 to 15.

Insurance status was measured with a dichotomous variable that captured whether respondents indicated at the time of interview they had no health insurance (1) or private/publicly funded insurance coverage (0).

Unmet medical need was derived from a variable that asked respondents, “Has there been any time over the past year when you thought you should get medical care, but you did not?” Respondents who answered “yes” were coded as 1 and those who responded “no” were coded as 0.

Control variables

Maternal age was derived from each respondent's date of birth and the month and year that the birth occurred. Maternal age ranged from 18 to 41.

Education level before pregnancy was measured using data from all waves of the Add Health sample. Respondents were asked about the highest degree they had received at all waves, and in Waves IV and V, the year they received that degree. From these data, we were able to calculate the age at which they finished their degree and whether degree completion occurred before each pregnancy or after. Educational achievement was, therefore, measured as a categorical variable that captured whether, before the pregnancy, the respondent had less than a high school degree (referent), a high school degree or equivalent, some college, or a bachelor's degree or more.

Race/ethnicity was measured categorically as non-Hispanic white (referent), non-Hispanic black, Hispanic, or other race/ethnicity.

Adolescent poverty was derived from parents' reported household income and household size at Wave I. A dichotomous measure was constructed in which respondents whose household income was ≤100% of the poverty level for their household size were coded as 1, and all others were coded as 0.

Time to birth was measured as the number of months between the survey date and date of birth.

Analytic strategy

First, we provided descriptive statistics for outcome and control variables for the total population and stratified by sexual identity for births that occurred both 1 and 3 years after a survey interview. Chi-square and t-tests were used to assess the bivariate relationship between respondent sexual identity and our categorical and continuous measures. Second, we conducted appropriate multivariate analyses of preconception health risk factors for heterosexual women versus SMW for births taking place both 1 and 3 years after interview. Logistic regression was utilized for marijuana use, other substance use, binge drinking, STI diagnosis, overweight/obesity, disordered eating, suicidal ideation, insurance status, and unmet medical need. Multinomial logistic regression was used for tobacco use. Ordered logit regression was utilized for self-rated health, and ordinary least squares regression was estimated for depressive symptoms. All models controlled for demographic characteristics including maternal age, race/ethnicity, education, adolescent poverty, and time between survey and birth.

Supplementary models controlled for pregnancy intention. This indicator was considered to be a confounder in the relationship between preconception health and sexual identity for two reasons: (i) poor preconception health is positively associated with unintended pregnancy and (ii) SMW are more likely to report unintended pregnancies than heterosexual women. All results are robust to the inclusion of this additional measure; however, because it is measured retrospectively, and our measures are prospective, it was not included in our final analyses.

Results

Table 1 provides descriptive statistics for the total analytic sample stratified by sexual minority status. Of all the women in the analytic sample, 12.19% identified as a sexual minority (e.g., lesbian, bisexual, or mostly heterosexual) at the survey wave before their births. Among births reported within 1 year of the survey, SMW reported a higher prevalence of marijuana use (21.84% vs. 9.77%, p < 0.01), other drug use (25.98% vs. 8.17%, p < 0.001), and binge drinking (47.15% vs. 29.85%, p < 0.05), and a lower prevalence of reporting never being a smoker (29.21% vs. 46.59%, p < 0.05) relative to heterosexual women. In addition, SMW reported a higher prevalence of STI diagnosis (23.87% vs. 11.48%, p < 0.05) and unmet medical needs (26.5% vs. 15.56%, p < 0.05) relative to heterosexual women, as well as higher mean scores for depressive symptoms (3.97 vs. 2.74, p < 0.01). The differences in preconception health observed 3 years before birth date were similar, although more robust than those observed during the 1-year period. In addition, SMW had higher reported prevalence of disordered eating (8.67% vs. 3.78%, p < 0.01) and suicidal ideation (15.99% vs. 3.25%, p < 0.001) as well poorer self-reported health (2.42 vs. 2.18, p < 0.05) relative to heterosexual women, during the 3-year period.

Table 1.

Descriptive Statistics Stratified by Sexual Identity

	Total (n = 3,133)		Heterosexual (n = 2,751)		Sexual minority (n = 382)
	M (%)	SE	M (%)	SE	M (%)	SE
Age (m)	23.41	0.18	23.38	0.18	23.59	0.35
Race/Ethnicity (%)
White	64.64		63.27		74.75		*
Black	21.4		22.55		13
Latina	11.09		11.27		9.74
Other	2.87		2.92		2.5
Education (%)
<High school	19.6		19.09		23.37
High school graduate	52.18		52.1		52.78
Some college	15.04		15.09		14.65
College graduate	13.18		13.73		9.2
Adolescent poverty (%)	17.13		17.29		15.93
Time to birth in years (m)	1.57	0.02	1.55	0.02	1.64	0.06

	Preconception health, 3 years before birth
	Total (n = 3,133)		Heterosexual (n = 2,751)		Sexual minority (n = 382)
Marijuana use (%)	16.18		14.61		27.65		***
Other drug use (%)	9.53		7.68		23.05		***
Binge drinking (%)	35.49		33.79		47.64		***
Tobacco use (%)
Never	43.45		45.58		27.97		***
Some	11.8		11.87		11.3
Former	19.67		19.34		22.11
Current	25.07		23.21		38.65
STI diagnosis (%)	10.2		9.4		16		*
Overweight or obese (%)	26.05		26.11		25.63
Disordered eating (%)	4.33		3.78		8.67		**
Suicidal ideation (%)	4.75		3.25		15.99		***
Self-rated health (m)	2.2	0.03	2.18	0.03	2.42	0.06	*
CES-D (m)	3.03	0.08	2.88	0.09	4.15	0.23	***
Uninsured (%)	22.08		21.19		28.63		†
Unmet medical need (%)	20.69		19.27		31.1		***

	Preconception health, 1 year before birth
	Total (n = 841)		Heterosexual (n = 748)		Sexual minority (n = 93)
Marijuana use (%)	11.14		9.77		21.84		**
Other drug use (%)	10.19		8.17		25.98		***
Binge drinking (%)	31.73		29.85		47.15		*
Tobacco use (%)
Never	44.62		46.59		29.21		*
Some	8.48		8.9		5.16
Former	23.6		22.92		28.87
Current	23.31		21.59		36.75
STI diagnosis (%)	12.9		11.48		23.87		*
Overweight or obese (%)	36.76		35.94		43.47
Disordered eating (%)	1.9		1.62		4.34
Suicidal ideation (%)	3.13		2.73		6.35
Self-rated health (m)	2.09	0.04	2.06	0.04	2.29	0.12
CES-D (m)	2.9	0.12	2.74	0.13	3.97	0.42	**
Uninsured (%)	15.74		15.02		21.58
Unmet medical need (%)	16.8		15.56		26.5		*

Source: National Longitudinal Study of Adolescent to Adult Health.

^*p < 0.05, ^**p < 0.001, ^***p < 0.0001, ^†p < 0.10.

m, mean; SE, standard error; CES-D, Center for Epidemiological Studies Depression 5 Item Scale; STI, sexually transmitted infection.

Table 2 presents results from our multivariate models detailing sexual identity disparities for the 12 different preconception health outcomes for births that occurred 1 year, then 3 years after survey date. Even after controlling for maternal age, race/ethnicity, adolescent poverty, education, and time between survey and birth, sexual orientation disparities in almost all preconception indicators were observed for births 3 years after survey date. Odds of marijuana use were doubled for SMW relative to heterosexual women for births that occurred 3 years after survey date (odds ratio [OR] = 2.10, 95% confidence interval [CI] 1.42–3.11) and tripled for other drug use (OR = 3.40, 95% CI 2.19–5.26). Increased odds of binge drinking (OR = 1.62, 95% CI 1.16–2.26) and increased risk of being a current smoker (relative risk ratio [RRR] = 2.23, 95% CI 1.45–3.44) or former smoker (RRR = 1.62, 95% CI 1.03–2.54) were also observed among SMW compared with heterosexual women. Odds of reporting an STI diagnosis (OR = 2.14, 95% CI 1.31–3.50) or disordered eating (OR = 2.37, 95% CI 1.17–4.78) were also elevated among SMW. In addition, SMW had higher odds of reporting a poorer self-rated health category than heterosexual women (OR = 1.56, 95% CI 1.21–2.01). Looking at mental health outcomes, SMW reported higher scores on the CES-D scale (B = 1.29, 95% CI 0.80–1.77) and increased odds of suicidal ideation (OR = 5.51, 95% CI 3.14–9.68). SMW were also more likely to report an unmet medical need in the past year (OR = 1.90, 95% CI 1.38–2.63). Only two indicators, being overweight or obese and being uninsured, did not show significant differences between SMW and heterosexual women.

Table 2.

Sexual Orientation Disparities in Preconception Health

	3 Years before birth			1 Year before birth
	OR	95% CI		OR	95% CI
Marijuana (n = 3,109; n = 793)	2.10	(1.42–3.11)	***	2.13	(0.98–4.63)	†
Other drug use (n = 3,128; n = 838)	3.40	(2.19–5.26)	***	3.96	(1.76–8.91)	***
Binge drinking (n = 2,950; n = 789)	1.62	(1.16–2.26)	***	2.04	(1.11–3.74)	*
	RRR	95% CI		RRR	95% CI
Tobacco use (never) (n = 3,115; n = 834)
Some	1.53	(0.91–2.59)		0.68	(0.16–2.80)
Former	1.62	(1.03–2.54)	*	1.99	(0.92–4.28)	†
Current	2.23	(1.45–3.44)	***	2.39	(1.11–5.13)	*
	OR	95% CI		OR	95% CI
STI diagnosis (n = 3,097; n = 827)	2.14	(1.31–3.50)	**	2.70	(1.20–6.10)	*
Overweight or obese (n = 2,895; n = 782)	0.98	(0.67–1.44)		1.32	(0.71–2.47)
Disordered eating (n = 2,606; n = 620)	2.37	(1.17–4.78)	*	2.17	(0.38–12.44)
Suicidal ideation (n = 3,058; n = 744)	5.51	(3.14–9.68)	***	2.31	(0.64–8.40)
No insurance coverage (n = 2,248; n = 696)	1.41	(0.91–2.19)		1.18	(0.48–2.91)
Unmet medical need (n = 3,128; n = 838)	1.90	(1.38–2.63)	***	1.76	(0.89–3.55)	†
Self-rated health (n = 3,127; n = 837)	1.56	(1.21–2.01)	***	1.72	(1.01–2.94)	*
	B	95% CI		B	95% CI
CES-D (n = 3,125; n = 838)	1.29	(0.80–1.77)	***	1.34	(0.49–2.19)	**

Source: National Longitudinal Study of Adolescent to Adult Health.

All models adjust for maternal age, race/ethnicity, adolescent poverty, education, and time between survey and birth

^*p < 0.05, ^**p < 0.001, ^***p < 0.0001, ^†p < 0.10.

OR, odds ratio; RRR, relative risk ratio; B, beta.

For births that occurred 1 year after survey date, SMW's odds of using drugs before pregnancy were nearly four times as high than heterosexual women (OR = 3.96, 95% CI 1.76–8.91). The odds of binge drinking in the past year doubled for SMW relative to heterosexual women (OR = 2.04, 95% CI 1.11–3.74). In terms of tobacco use, SMW had 2.4 times the relative risk of being a current smoker relative to never smoking compared with heterosexual women (RRR = 2.39, 95% CI 1.11–5.13). The odds of reporting an STI diagnosis were nearly three times as high for SMW relative to heterosexual women (OR = 2.70, 95% CI 1.20–6.10). SMW had higher odds of reporting a poorer self-rated health (OR = 1.72, 95% CI 1.01–2.94). Finally, for depression, SMW had, on average, a higher CES-D score relative to heterosexual women (B = 1.34, 95% CI 0.49–2.19).

We conducted additional supplementary analyses that disaggregated sexual identities (mostly heterosexual, bisexual, and gay/lesbian) and found that in only one case did our sexual minority identity groups statistically differ from one another: lesbian women were significantly less likely than bisexual and mostly heterosexual women to report suicidality during the 3-year preconception period.

Discussion and Conclusions

Preconception health is increasingly being recognized as an important issue for maternal, infant, and child health given increases in maternal mortality,⁵ preterm birth,⁴ and low birthweight births.³ In 2006, the select panel on preconception health identified goals for improving maternal and infant health that included increasing preconception health education, improving access to preconception health care, and reducing disparities in adverse birth outcomes.¹⁷ Despite these aims, our findings suggest that SMW are even more vulnerable to poor preconception health than their heterosexual counterparts, which has important implications for maternal and child health. The results presented here indicate that, similar to other studies that have documented sexual orientation disparities across multiple dimensions of health in the general population, these disparities extend to the preconception period. Moreover, these results add to the burgeoning literature on SMW's maternal health that has demonstrated that currently pregnant SMW are more likely to report numerous adverse health behaviors and conditions³⁴ and adverse pregnancy and birth outcomes³¹ relative to heterosexual women. Taken together, these studies provide important evidence for the need to invest in the reproductive health of SMW, particularly in the context of pregnancy and preconceptually.

A number of critiques have been made with regard to the preconception health paradigm,^47–49 and although the CDC now provides preconception health recommendations for both men and women, nonheterosexual identities are not explicitly considered, and therefore differences based on sexual identity have yet to be explicitly addressed. Our findings suggest that future interventions aimed at the preconception period should include SMW. This is especially important given physicians' implicit biases against lesbian and gay patients,⁵⁰ as well as heteronormative standards of pregnancy and birth,^51,52 both of which may negatively impact SMW's preconception health. As previously stated, SMW already face barriers to accessing health care and preventative treatment.^32,35 Interestingly, we did not find significant differences in being uninsured across sexual minority status, but SMW were more likely to report unmet medical needs, suggesting that other barriers to care, such as discrimination, may contribute to disparities in preconception health.

A number of institutions currently provide recommendations for the care of sexual minority populations^53–56; however, none of these recommendations specifically address the preconception period. Providers should discuss pregnancy planning with SMW patients during regular care visits to help SMW who wish to become pregnant in the near future to engage in health-promoting behaviors before conception. Some research has found that culturally tailored health interventions can be beneficial for improving minority women's health.^57–60 Such programs should be developed for lesbian, gay, bisexual, trans, and queer populations to improve maternal and infant health outcomes for this population.

Our findings also suggest that measuring a longer prepregnancy window can illuminate greater preconception health disparities for SMW. Future research should build on this evidence to identify ideal windows of time in which to measure prepregnancy health risks for all women. Although not ideal, we combined lesbian, bisexual, and mostly heterosexual women due to small sample sizes to provide a preliminary exploration of preconception health for this group of women. Moreover, the survey item used to measure sexual identity in the Add Health survey incorporates elements of sexual identity and sexual attraction. It may be possible that some of our respondents do not “identify” with a sexual minority identity while reporting same-sex attraction. Future nationally representative data sets should be designed to measure the fertility and health histories of larger number of SMW, permitting more fine-grained analyses by sexual identity than we could undertake using existing data. We were also unable to analyze several other preconception risk factors, such as vaccinations, dental records, and other important health indicators in our study. Finally, our results are limited by the fact that all behaviors are self-reported. Owing to the fact that our measures are asked before conception, however, we believe that we avoid some recall bias about health status and behaviors before pregnancy that cross-sectional analyses may encounter.

SMW have long been ignored or excluded from discussions of maternal health. The results presented here add to the small body of research documenting that pregnancy across a variety of contexts is common among SMW and that there are stark differences in health conditions and behaviors before and during pregnancy and in birth outcomes. More work is needed to both understand the mechanisms that lead to such disparities and identify ways to promote maternal health equity across all sexual orientations.

Disclaimer

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

Research reported in this publication was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health under Award Number R01HD091405 and by the University of Colorado Population Center (Grant No. R24 HD066613) through administrative and computing support.