Abstract
Objectives. To describe trends in sexual and reproductive health behaviors and service utilization among young women in the United States.
Methods. We analyzed data from 8835 female respondents aged 18 to 25 years from 4 cycles of the National Survey of Family Growth, a nationally representative cross-sectional survey, from 2002 to 2015. We used bivariate and multivariable logistic regression to compare rates of self-reported sexual activity, sexually transmitted infection–related care, and contraception use over time and by race/ethnicity.
Results. Sexually transmitted infection–related care and human papilloma virus vaccination increased from 2002 to 2013–2015, whereas sexual activity and contraception use remained stable. Compared with White women, racial/ethnic minority women were less likely to report effective contraception use, and Black women were less likely to report human papilloma virus vaccination; these differences did not change over time.
Conclusions. Sexual and reproductive health service utilization increased from 2002 to 2015 among young women, whereas sexual activity remained stable. Overall, rates of recommended care were low, and racial and ethnic disparities persisted.
Public Health Implications. Young women could benefit from clinical interventions and health policies to increase recommended care and reduce disparities.
Young adult women (typically defined as those aged 18–25 years) in the United States face a unique set of health and health care issues, including higher sexual and reproductive health (SRH) risks and lower levels of health care access compared with other age groups.1 Young women aged 18 to 24 years had the highest rate of unintended pregnancy of all age groups in 2008 and 2011,2 and those aged 15 to 24 years accounted for nearly half of all incident sexually transmitted infections (STIs) in 2014.3 Black and Hispanic young adult women have generally fared even worse on major SRH indicators; they have been shown to experience lower rates of contraception use,4 higher rates of unintended pregnancy,2 higher rates of STI diagnoses,3 and lower rates of human papilloma virus (HPV) vaccination5 compared with their White peers.
Lower health care access and utilization among young adult women likely compound their SRH risks. Data from 2005 to 2012 showed that young adults in the United States had lower rates of overall and preventive health care utilization than did persons in other age groups6 and lower rates of health insurance coverage than did adolescents.1 Health care utilization and spending were lower among Black and Hispanic young adults than among White young adults, even after adjusting for insurance, usual source of care, and income.6
Several events since the early 2000s have had the potential to improve the health of young adult women in the United States, especially in the domain of SRH (Appendix A, Figure A [available as a supplement to the online version of this article at http://www.ajph.org]). Relevant advances in clinical guidelines included recommendations for highly effective long-acting reversible contraceptives (LARCs),7 HPV vaccination,8,9 and routine HIV testing.10 Recent health policy changes through the Patient Protection and Affordable Care Act (ACA) also aimed to improve young adults’ access to SRH services. In 2010 the ACA expanded public insurance, mandated private insurance plans to extend coverage to dependents aged 26 years and younger, and made preventive health services free for anyone with insurance.11 Furthermore, a contraception coverage mandate, added in 2012 and implemented by most plans in 2013, required insurance companies to cover all Food and Drug Administration–approved contraceptive methods at no cost to women.
Since 2011, studies have shown expanded insurance coverage among young adults,12 increased use of prescription contraception among reproductive-aged women,13 and narrowed gaps in access to care between Black and Hispanic and White populations.14 However, studies have yielded mixed results on recent trends in SRH behaviors and care8,12,15 and few conclusions about whether improved access to care has reduced racial and ethnic disparities,16 particularly among young adults.17 There also remains political controversy about whether increased access to SRH services affects sexual behavior and risk.18 Understanding trends in SRH services and sexual behaviors over a time of increasing access to care could help clarify this issue.
We aimed to define trends in sexual activity, STI-related health care, and contraception use among young adult women in the United States from the early 2000s through 2015, a period of multiple significant changes in women’s health policy and practice. We also aimed to understand changes in the relationship of race and ethnicity with SRH services over time. The causes of racial and ethnic health disparities are complex and include differential access to health care and unequal content and quality of care received.19,20 We hypothesized that updated SRH care guidelines, increased contraceptive choice, and policies to improve access to health care would increase rates of SRH service utilization among young adult women and would reduce racial and ethnic disparities.
We analyzed data from the National Survey of Family Growth (NSFG), a nationally representative multistage probability-based survey of the US household population of reproductive age (15–44 years).21 The NSFG collects data primarily through face-to-face interviews and additionally via audio computer-assisted self-interview software for sensitive information. The NSFG has been administered in cycles since 1973, with oversampling of Black, Hispanic, and adolescent participants since 2002. We used NSFG data from surveys conducted in 2002, 2006 to 2010, 2011 to September 2013, and September 2013 to 2015. Response rates were 80%, 78%, 73%, and 71% for all women in the 4 cycles, respectively.
To provide more data points for the trend lines, we divided responses from 2006 to 2010 into 2 periods, separated at June 30, 2008, using procedures recommended by the NSFG.22 This resulted in 5 nationally representative survey populations: 2002, 2006 to 2008, 2008 to 2010, 2011 to 2013, and 2013 to 2015 (Figure 1). We applied weights provided by the NSFG to produce nationally representative estimates.21,23
FIGURE 1— Age-Adjusted Rates (99.9% CI) of Selected Sexual Activity Measures Among Female National Survey of Family Growth Respondents Aged 18–25 Years: United States, 2002–2015
Note. CI = confidence interval.
We used data from respondents who identified as female and were aged 18 to 25 years at the time of interview. Of 31 222 female respondents from 2002 to 2015, we analyzed sexual activity among those aged 18 to 25 years (n = 8835). From this age-restricted sample, we analyzed other outcomes as follows: STI-related care among those who reported sexual activity in the past 12 months (78%; n = 6853); contraception use among those at risk for unwanted pregnancy, as defined in Methods (77%; n = 6764); and HPV vaccination among those interviewed on the topic (aged 18–24 years starting in 2007) regardless of sexual activity (57%; n = 5047).
All outcomes were self-reported. Our key outcomes were sexual activity, STI-related service utilization, and contraception use, which we defined as follows:
We defined sexual activity by heterosexual sexual activity (“Have you ever had sexual intercourse with a man?”). We assessed rates of sexual initiation before the age of 15 years, sexual activity in the past 12 months, and having more than 1 sexual partner in the past 12 months. We were unable to analyze nonheterosexual sex because this was not assessed at the same level of detail in the survey.
We defined STI-related service utilization with a focus on guideline-recommended interventions24: annual STI screening and counseling and HIV testing for sexually active young adults, in concordance with recommended care, and HPV vaccination for all. Among respondents who reported sexual activity in the past 12 months, we assessed rates of receiving any STI-related care (i.e., any counseling, examination, testing, or treatment) and of HIV testing (excluding testing for blood donation purposes only) in the past 12 months. The HIV survey questions changed over time: in 2002, respondents were asked the month and year of their last HIV test; in subsequent years, ambiguous responses were followed up with a direct question, “Were you tested for HIV in the 12 months?”
We assessed rates of any HPV vaccination at any time in the past (“Have you ever received the cervical cancer vaccine, also known as the HPV shot or Gardasil?”), regardless of sexual history. The NSFG asked about HPV vaccination only among respondents younger than 25 years, in 2007 and later, and did not ask the number of HPV vaccinations received. From 2007 to 2010 the NSFG assessed knowledge of the HPV vaccine, and respondents who had never heard of it were not assessed for vaccine receipt. We assumed that those reporting no knowledge of the HPV vaccine (15%; n = 366 from 2007 to 2010) had not received it. Questions specifically about STI diagnosis and treatment appeared only in the last 2 survey cycles, so we were unable to analyze trends in STI incidence.
We defined contraception use by use at last incidence of sexual intercourse, and we analyzed it for respondents at risk for unwanted pregnancy to assess contraception use for contraceptive purposes. Respondents at risk for unwanted pregnancy included those who reported sexual activity in the past 12 months, excluding those who reported not using contraception because they were “pregnant,” they were “seeking pregnancy,” or they or their partners were “sterile [for noncontraceptive reasons].” Respondents could report up to 4 different methods, all of which we included in our analysis. We considered any method associated with fewer than 10 unintended pregnancies per 100 women per year of typical use “effective.”25 By this definition, effective contraception included short-acting hormonal methods (i.e., pills, patches, rings, injections), long-acting reversible contraceptive methods (i.e., LARC methods, which are intrauterine devices and subdermal contraceptive implants), and sterilization for contraceptive purposes. A respondent used effective contraception if she reported using any of these methods at last instance of sexual intercourse. All so-called effective methods required a prescription in most states.
Key covariates included race/ethnicity and insurance status, which were self-reported and defined using methods from previous studies,12,26 as follows:
We categorized race/ethnicity according to the 1997 Office of Management and Budget standards27: we analyzed any respondent reporting Hispanic ethnicity as “Hispanic”; the remainder were “White,” “Black,” or “other” race. We categorized those reporting multiple races as other. The 2002 survey did not allow multiple race reporting, thus we categorized respondents in 2002 according to the 1977 standards.
We summarized insurance status by type (i.e., private, public, or none) and duration of coverage (i.e., full year, partial year, or none). Respondents were able to name up to 4 types of health insurance coverage over the past 12 months.27 The NSFG defined private insurance as “Private health insurance plan (from employer or workplace; purchased directly; through a state or local government program or community program).” We considered all other insurance types public. We grouped respondents with full-year mixed insurance with those with full-year public insurance because they were demographically similar.
Additional covariates included language spoken at home, household income, education level, marital status, number of previous births, urban versus rural place of residence, and having a usual source of care. Definitions of these covariates can be found in Appendix A.
We compared population-weighted demographic and socioeconomic characteristics of respondents in the 5 study samples. We tested for changes in means over time using linear regression with survey cycle indicators as predictors (2002 was the reference year) and proportions over time using the Pearson χ2 test. We then calculated population rates of sexual activity, STI-related service utilization, and contraception use in each survey cycle. To assess the effects of survey cycle and race/ethnicity, we performed individual-level bivariate and multivariable logistic regression on all outcomes. We tested for interactions between survey cycle and 2 predictors of interest (i.e., race/ethnicity and insurance status), using a threshold of 10% change in the main effect or a Wald test P < .001 to be included in the model. When a sampling stratum consisted of only 1 observation, we held the singleton observation at the overall sample mean.28 Sensitivity analyses revealed no significant differences when we scaled the singleton units to strata with multiple observations or treated them as certainty units.
Unadjusted and adjusted models revealed similar trends. We present only the age-adjusted outcomes by survey cycle and the fully adjusted effects of race/ethnicity. Interaction terms for survey cycle with insurance status and race/ethnicity were not significant (Appendix A, Tables A and B [available as supplements to the online version of this article at http://www.ajph.org] provide unadjusted rates and complete model results).
We conducted all analyses between June 2017 and March 2018 using Stata version 13.0 (Stata Corp, College Station, TX) with survey commands that accounted for the complex survey design. We determined statistical significance by a 2-sided P < .001, on the basis of P < .05 for a single comparison and a Bonferroni correction for a total of 70 comparisons (10 outcomes over 4 survey cycle comparisons and 3 race/ethnicity group comparisons). We present corresponding 99.9% CIs.
As shown in Table 1, sample characteristics such as age, race/ethnicity, and having a usual source of care were stable from 2002 to 2013–2015. Over the same period, the lowest-income group grew from 26% to 33% of the population, whereas the higher-income groups shrank. In later years, a higher proportion reported year-round insurance coverage (66% and 68% in 2011–2013 and 2013–2015, respectively, compared with 62% in 2008–2010).
TABLE 1— Demographic Characteristics of Female National Survey of Family Growth Respondents Aged 18–25 Years: United States, 2002–2015
| Characteristic (n = 8835) | 2002 (n = 2106) | 2006–2008 (n = 1681) | 2008–2010 (n = 1853) | 2011–Sep 2013 (n = 1610) | Sep 2013–2015 (n = 1585) | P |
| Mean age, y | 21.5 | 21.5 | 21.4 | 21.5 | 21.6 | .18 |
| Race/ethnicity,a weighted % | .80 | |||||
| Non-Hispanic White (single race) | 62.7 | 62.8 | 61.0 | 56.8 | 55.3 | |
| Hispanic | 16.4 | 16.1 | 17.0 | 20.3 | 21.8 | |
| Non-Hispanic Black (single race) | 14.6 | 14.6 | 15.2 | 16.2 | 16.3 | |
| Non-Hispanic other or mixed | 6.4 | 6.5 | 6.9 | 6.7 | 6.6 | |
| Home language, weighted % | < .001 | |||||
| English | 94.8 | 95.0 | 97.0 | 92.2 | 92.0 | |
| Non-English | 5.2 | 5.0 | 3.0 | 7.8 | 8.0 | |
| Household income, % of FPL, weighted % | .002 | |||||
| ≥ 400 | 21.2 | 17.4 | 11.9 | 14.4 | 18.1 | |
| 200–399 | 29.6 | 29.4 | 35.3 | 27.8 | 26.6 | |
| 100–199 | 23.3 | 24.1 | 25.0 | 23.6 | 22.4 | |
| < 100 | 26.0 | 29.0 | 27.7 | 34.2 | 32.8 | |
| Education level, weighted % | .004 | |||||
| Less than high school | 17.9 | 19.8 | 18.9 | 12.3 | 10.6 | |
| High school graduate or GED | 31.1 | 29.8 | 28.6 | 33.3 | 32.2 | |
| Some college | 37.1 | 35.5 | 38.6 | 42.0 | 42.2 | |
| College graduate | 13.9 | 14.9 | 13.9 | 12.3 | 15.0 | |
| Marital status, weighted % | < .001 | |||||
| Never married | 76.0 | 78.1 | 86.0 | 85.3 | 85.8 | |
| Currently married | 21.1 | 19.3 | 11.6 | 13.2 | 12.0 | |
| Divorced/separated/widowed | 2.9 | 2.6 | 2.4 | 1.4 | 2.2 | |
| Prior births, mean | 0.48 | 0.44 | 0.41 | 0.39 | 0.37 | .002 |
| Place of residence, weighted % | .005 | |||||
| Urban | 56.8 | 55.9 | 65.2 | 62.3 | 59.3 | |
| Rural | 43.2 | 44.1 | 34.8 | 37.7 | 40.7 | |
| Usual source of care, weighted % | .69 | |||||
| Yes | . . . | . . . | 79.8 | 79.4 | 81.0 | |
| No | . . . | . . . | 20.2 | 20.6 | 19.0 | |
| Insurance typeb, weighted % | .001 | |||||
| Full-year private | 49.5 | 45.3 | 44.6 | 42.7 | 45.7 | |
| Full-year public | 14.6 | 16.2 | 17.7 | 23.2 | 22.4 | |
| Full-year mixed private/public | 5.6 | 5.3 | 4.3 | 5.4 | 8.2 | |
| Partial-year uninsured | 18.4 | 19.4 | 19.4 | 14.4 | 12.0 | |
| Full-year uninsured | 11.9 | 13.9 | 14 | 14.3 | 11.7 |
Note. FPL = federal poverty level; GED = general equivalency diploma. FPL determined by US Census Bureau in the year before the interview.
a 2002 survey on the basis of 1977 Office of Management and Budget standards, 2006 onward on the basis of 1997 Office of Management and Budget standards.
b 30 respondents were missing insurance status (n = 8805).
In all survey cycles, about 13% of respondents reported sexual initiation before the age of 15 years, 75% reported sexual activity in the past 12 months, and 18% reported more than 1 sexual partner in the past 12 months (Figure 1). There were no significant changes in sexual activity over time.
Self-reported STI-related care increased over time but remained well below recommended levels (Figure 2). Among respondents reporting sexual activity in the past 12 months, rates of STI counseling, testing, or treatment rose from 21% in 2002 to 45% in 2013 to 2015 (age-adjusted odds ratio [AOR] = 3.5; 99.9% CI = 2.4, 5.1), and HIV testing rates were stable around 25%. HPV vaccination rates rose from 13% in 2007 to 2008 to 50% in 2013 to 2015 (AOR = 6.6; 99.9% CI = 3.1, 14.3) among all respondents aged 18 to 24 years.
FIGURE 2— Age-Adjusted Rates (99.9% CI) Among Female National Survey of Family Growth Respondents Aged 18–25 Years of (a) Sexually Transmitted Infection (STI) Service Use and (b) Contraception Use: United States, 2002–2015
Note. CI = confidence interval; LARC = long-acting reversible contraceptives. Contraception use defined as use at last instance of sexual intercourse in past 12 months.
Among respondents at risk for unwanted pregnancy, an increasing percentage reported LARC use at last instance of sexual intercourse (1% in 2002 to 13% in 2013–2015; AOR = 9.9; 99.9% CI = 4.1, 23.9). Rates of any contraception use, effective contraception use, and condom use were stable around 80%, 50%, and 35%, respectively.
Race and ethnicity were significant predictors of STI-related service utilization and contraception use (Table 2), with effects that did not vary over time (Appendix A, Figure B [available as a supplement to the online version of this article at http://www.ajph.org]). Even after adjusting for age, insurance status, survey cycle, socioeconomic status, and sexual activity, Black respondents had nearly 3 times the odds of reporting HIV testing than did their White counterparts, roughly half the odds of HPV vaccination, and half the odds of effective contraception use. Hispanic and other or mixed race respondents had similar adjusted odds of STI-related care, HIV testing, and HPV vaccination as did White respondents but about half the odds of effective contraception use. Race and ethnicity had no relationship to LARC or condom use in the adjusted model.
TABLE 2— Adjusted Effects (99.9% CI) of Race and Ethnicity on Selected Sexual and Reproductive Health Outcomes Among Female National Survey of Family Growth Respondents Aged 18–25 Years: United States, 2002–2015
| Outcome | Non-Hispanic White (Ref), AOR (99.9% CI) | Hispanic, Any Race, AOR (99.9% CI) | Non-Hispanic Black, AOR (99.9% CI) | Non-Hispanic Other, AOR (99.9% CI) |
| Sexual activity (n = 8805) | ||||
| Sexual initiation before aged 15 y | 1.0 | 0.7 (0.5, 1.1) | 1.2 (0.8, 1.7) | 0.8 (0.4, 1.5) |
| Sexual intercourse within the past 12 mo | 1.0 | 1.0 (0.6, 1.4) | 1.1 (0.7, 1.6) | 0.6 (0.4, 1.1) |
| > 1 sexual partner within the past 12 mo | 1.0 | 0.7 (0.5, 1.1) | 0.9 (0.7, 1.3) | 0.7 (0.4, 1.2) |
| STI-related servicesa (n = 6841) | ||||
| Any STI-related care in past 12 mob (n = 6832) | 1.0 | 1.0 (0.6, 1.4) | 1.3 (1.0, 1.8) | 0.7 (0.4, 1.3) |
| HIV testing in past 12 mo | 1.0 | 1.4 (0.9, 2.1) | 2.9 (2.1, 3.9) | 1.2 (0.7, 2.2) |
| HPV vaccination everc (n = 4960) | 1.0 | 0.9 (0.5, 1.3) | 0.6 (0.4, 0.9) | 0.8 (0.4, 1.8) |
| Contraception use at last sexual intercoursed (n = 6753) | ||||
| Any contraception | 1.0 | 0.8 (0.5, 1.3) | 0.6 (0.4, 0.9) | 0.7 (0.4, 1.3) |
| Effective contraception | 1.0 | 0.6 (0.5, 0.9) | 0.5 (0.3, 0.7) | 0.4 (0.2, 0.7) |
| LARC | 1.0 | 1.2 (0.6, 2.2) | 0.6 (0.3, 1.2) | 1.0 (0.4, 2.5) |
| Condom | 1.0 | 1.1 (0.7, 1.5) | 1.3 (1.0, 1.9) | 1.3 (0.7, 2.4) |
Note. AOR = adjusted odds ratio; CI = confidence ratio; HPV = human papilloma virus; LARC = long-acting reversible contraceptives (implants, intrauterine devices); STI = sexually transmitted infection. Model adjusted for survey cycle (year), age, insurance category, usual source of care, home language, household income category, education level, marital status, parity, and rural residence. For human papilloma virus vaccination only, adjusted for sexual activity ever. We did not include interaction terms for survey cycle and insurance category or survey cycle and race/ethnicity, as they were tested and not found to be significant. Thirty respondents were missing insurance information (n = 8805).
a 21 respondents were missing insurance status (n = 6841).
b 9 respondents were missing STI-related care response (n = 6832).
c 66 respondents were missing the human papilloma virus vaccination response (n = 4981), and an additional 21 were missing insurance status (n = 4960).
d 11 respondents were missing insurance status (n = 6753).
In repeated cross-sectional nationally representative samples of US women aged 18 to 25 years, we found increasing rates of SRH service utilization from 2002 to 2013–2015, as we hypothesized. Specifically, we observed increases in self-reported HPV vaccination, STI service utilization, and LARC use. At the same time, we found no increases in self-reported sexual activity. In addition, contrary to our hypothesis that disparities would improve over time, we identified persistent disparities in HPV vaccination for Black respondents and in effective contraception for respondents in all racial and ethnic minority groups. Our results are consistent with others’ findings: young adults used more preventive health care over the period when the ACA came into effect8,12,13; sexual activity among US adults has been stable for decades,29 including since the passage of the ACA30; and racial and ethnic disparities in SRH service utilization have persisted despite population-wide increases in SRH utilization.9,31
Our study adds to and updates previous work in several ways. First, it provides a high-level picture of SRH trends in the United States from 2002 to 2015, a time of significant changes in medical practice and health policy relevant to young adult women’s SRH. Second, by contrasting trends in sexual activity and SRH service utilization in nationally representative samples of young adult women, we have shown that increases in SRH service utilization and insurance expansions under the ACA have not been associated with changes in sexual activity at the population level. Finally, we analyzed the effects of race and ethnicity on receipt of several SRH services over time, and we showed that racial and ethnic disparities in HPV vaccination and effective contraception use persisted despite increasing contraceptive choice and expanded access to care after health insurance reforms.
In STI-related care, we identified a steady rise in HPV vaccination since 2007 to 2008 and an apparent increase in STI counseling, testing, and treatment between 2011 to 2013 and 2013 to 2015. As others have found,30 the increase in STI-related care did not appear to be related to increased risk, as measures of sexual activity and rates of STI diagnoses (data not shown) were stable in the study sample over the same period. Rather, the increase in STI-related care might be attributable to recent health insurance expansions that enabled more preventive care utilization, such as counseling and testing. The increasing proportion of young women reporting full-year insurance coverage between 2008 to 2010 and 2011 to 2013 supports this hypothesis. A previous study found that post-ACA increases in young adult preventive care utilization were largely explained by gains in insurance coverage.12 Future research will need to determine if these increases in STI care persist over time or lead to reductions in STI incidence and severity.
Contraception use, conversely, was generally stable from 2002 to 2013–2015; the only significant change was a steady rise in LARC use since 2002, consistent with observed national trends. The rise in LARC use may be explained by the increasing number and strength of recommendations for LARCs since 2000. If and how the ACA affected contraceptive use was beyond the scope of our analysis. Nonetheless, we did not observe any significant changes in overall rates of contraception use or method choice following the ACA’s insurance expansions in 2010 or contraception coverage mandate in 2012 to 2013. Longer follow-up and more detailed analyses may be needed to understand the effects of the ACA on contraceptive choice and use.
We hypothesized that SRH disparities would be reduced by updated clinical guidelines, increased contraceptive choice, and policies to improve access to care. Although we observed favorable overall trends in HPV vaccination, STI care, LARC use, and insurance coverage, there were persistent disparities by race and ethnicity. Even after multivariable adjustment, Black respondents were less likely than were White respondents to report guideline-recommended24 HPV vaccination. In addition, being in any racial or ethnic minority group was associated with lower rates of effective contraception use. Our findings were similar to those of previous studies,31,32 in which racial disparities in SRH care were not fully explained by differences in health care access. The fact that Black respondents reported higher rates of HIV testing than did White respondents suggests that they had access to some SRH care and could have accessed a broader range of SRH services. It is concerning that Black women received higher rates of clinical care in some domains of STI care (e.g., HIV testing) but lower rates in others (e.g., HPV vaccination). Further research is needed to explore the etiology of racial disparities in SRH care processes, including the possible roles of health system resource distribution and provider bias.19
Meanwhile, across all years and racial/ethnic subgroups in our study, sexual activity was stable. Understanding trends in sexual behavior among young adult women is particularly important in the current health policy context. In October 2017, the US Department of Health and Human Services passed a bill to allow more employers and insurers to claim a religious or moral objection to the ACA’s contraceptive coverage mandate,33 citing research to suggest that access to contraception may lead to more adolescent premarital sex and pregnancy.18 Although several studies have found no link between access to SRH services and sexual risk taking,5,30,34 none to our knowledge has examined this question for young adult women in the United States over the past decade or more. By contrasting trends in sexual activity and SRH service utilization in nationally representative samples of young adult women, we were able to show no association between increasing STI care and contraception use and sexual activity at the population level.
Our study had several important limitations. First, we derived the data from independent cross-sectional surveys that used somewhat different instruments and sample designs in different years and that were not equally spaced over time. However, NSFG provides guidance that enabled us to derive consistent national estimates with known sampling error. Second, all data were self-reported and therefore subject to inaccurate reporting, particularly for the sensitive subject matter included in our analysis. We would not expect rates of inaccurate reporting to change over time, although they might vary by subgroup.
Third, HIV status and HIV test results were not asked about in the survey, and STI diagnoses were only asked about in the last 2 rounds, limiting assessment of the burden of STIs over time. In addition, we would have liked to assess the impact of the dependent care mandate on insurance coverage and health care utilization in this population, but parental insurance was only asked about in the final survey round. Finally, we did not focus on persons of other or mixed race in subgroup analyses because of within-group heterogeneity that reduced our ability to understand patterns in this subgroup.
We found that young adult women in the United States utilized more SRH preventive services and highly effective contraception from 2002 to 2015. However, overall rates of recommended STI care and contraception use remained low, and racial disparities persisted. Continued efforts to address disparities and increase uptake of SRH services, in particular HPV vaccination and effective contraception use, are needed. Offering HPV vaccination and comprehensive contraceptive counseling when and where young adult women go for other types of STI care (e.g., HIV testing) is a possible strategy to improve the equity and quality of SRH care among young adult women in the United States.
In addition, the notion that contraception leads to greater sexual risk taking is not supported by our or others’ research.5,30,34 Rather, restricting access to contraception may lead to lower rates of effective contraception use, increased rates of unintended pregnancy and abortion, higher medical costs, and worse socioeconomic disparities.35 Policymakers should consider these risks when crafting policy that affects access to SRH care.
Young adult women have been shown to have lower levels of health care utilization and worse SRH outcomes than have other age groups. We found that young adult women in the United States utilized more SRH services in 2013–2015 than in 2002 (specifically LARCs, HPV vaccination, and STI-related care) without any associated changes in sexual activity. Yet rates of recommended care remained low, and racial and ethnic disparities persisted. Future research, targeted interventions, and health policies should explicitly address these gaps and disparities in care.
ACKNOWLEDGMENTS
This research was supported by funding to M. E. M. H. through an Institutional National Research Service Award from the National Institutes of Health (award 2-T32-HP10251) and the Harvard Pilgrim Health Care Institute.
We would like to thank Isaedmarie Febo-Vázquez at the Centers for Disease Control and Prevention, National Center for Health Statistics for her invaluable technical support with the National Survey of Family Growth (NSFG) data.
HUMAN PARTICIPANT PROTECTION
The NSFG obtained written informed consent from all respondents. The use of publicly available de-identified data in this study was considered exempt from approval by the Harvard Pilgrim Health Care Institute institutional review board.
