A SUBSTANTIAL PROPORTION of women of reproductive age in the United States experience infertility or impaired fecundity. Infertility is generally defined as a lack of pregnancy after 12 months of trying to get pregnant,¹ whereas impaired fecundity is defined as physical difficulty in either getting pregnant or carrying a pregnancy to live birth.² Despite recognition of infertility as a public health concern, considerable gaps and opportunities exist in research, program, and policy development in the prevention, detection, and management of infertility and impaired fecundity.^3,4 A comprehensive public health approach is needed that includes surveillance and monitoring to explore and better understand preventable risk factors and disparities in access to care and treatment use.

Different factors including sociodemographic, behavioral, and physical and mental health conditions have been identified as possible contributors to fertility problems in women. Many of the behavioral factors and health conditions may be preventable, such as smoking, obesity, and diabetes. Disparities by race/ethnicity and socioeconomic status in prevalence, access to care, and use of infertility services/treatments have also been noted in the literature.^5–7 Population level survey data include information that can be used to explore both infertility and impaired fecundity, in addition to sociodemographic factors that may be related to differential access to treatment and potentially unmet treatment needs. One population level survey is the National Survey of Family Growth (NSFG) conducted by the Centers for Disease Control and Prevention’s (CDC) National Center for Health Statistics.⁸ Recent estimates from the 2011 to 2013 NSFG revealed that among currently married, women 15–44 years of age, 6.1% were infertile and 12.3% had impaired fecundity.⁹ However, state-based population level data are lacking. Ultimately, findings from analyses of such data may lead to designing and implementing effective strategies to help close gaps in access to care, prevention, and education.

State-specific estimates may be obtained from other surveillance systems such as the CDC’s Behavioral Risk Factor Surveillance System (BRFSS), which gathers information on the prevalence and distribution of self-reported behaviors, health conditions, and services. The BRFSS was developed in 1984 and is the world’s largest ongoing telephone health survey system. Although the core BRFSS does not regularly collect information on either infertility or impaired fecundity, the survey does allow for the addition of questions related to these issues.

In this article, we use the BRFSS to explore infertility and impaired fecundity. Specific questions were added to the survey with the following objectives: (1) to identify health and behavioral factors and correlates associated with infertility and difficulty staying pregnant and (2) to explore access to care, treatment utilization, and outcomes associated with infertility and difficulty staying pregnant.

The overall percentage of women reporting infertility only was 4.8% (95% CI = 3.6–6.3), with state-specific values ranging from 3.1% (95% CI = 2.0–4.8) in Connecticut to 6.1% (95% CI = 4.7–7.8) in Utah. The overall percentage reporting difficulty staying pregnant only was 4.0% (95% CI = 3.1–5.2), with state-specific estimates ranging from 3.2% (95% CI = 1.8–5.8) in Texas to 5.2% (95% CI = 3.5–7.5) in Ohio.

The distribution of some characteristics of women differed among groups (infertility only, difficulty staying pregnant only, and neither infertility nor difficulty staying pregnant; Table 1). Compared with the percentage of women reporting neither infertility nor difficulty staying pregnant, the percentage of women reporting infertility only was higher for older (67.2% were 35 years of age or older vs. 49.1%), married (74.3% vs. 55.4%), college educated (35.5% vs. 25.3%), covered by healthcare (92.9% vs. 75.5%), and ever told they had a depressive disorder (38.2% vs. 25.0%). The percentage of women ever told they had a depressive disorder also was high among women reporting difficulty staying pregnant only (39.1%). Finally, the percentage ever smoking 100 cigarettes or more among women reporting difficulty staying pregnant only (52.0%) was higher than among women reporting infertility only (35.5%) and higher than women reporting neither (34.1%).

In multivariable analysis, compared with women who reported neither infertility nor difficulty staying pregnant, women ever told they had a depressive disorder were more likely to report either infertility only or difficulty staying pregnant only (adjusted odds ratio [aOR] = 2.02, 95% CI = 1.14–3.59 and aOR = 1.69, 95% CI = 1.07–2.68, respectively), whereas women who had ever smoked at least 100 cigarettes were more likely to report difficulty staying pregnant only (aOR = 1.98, 95% CI = 1.22–3.20) (Table 2). Compared with women with healthcare coverage, women without health-care coverage were less likely to report infertility only compared with those who reported difficulty staying pregnant only (aOR = 0.37, 95% CI = 0.17–0.84) and those who experienced neither (aOR = 0.26, 95% CI = 0.14–0.46).

When the multivariable analysis was restricted to married women 35–50 years of age, the results did not change substantially, but smoking was no longer significantly associated with the infertility group (Table 3). Compared with women who reported neither infertility nor difficulty staying pregnant, women who were ever told they had a depressive disorder were more likely to report either infertility only or difficulty staying pregnant only (aOR = 2.89, 95% CI = 1.22–6.85 and aOR = 2.08, 95% CI = 1.14–3.80, respectively), whereas women without healthcare coverage were less likely to report infertility only or difficulty staying pregnant only (aOR = 0.13, 95% CI = 0.06–0.29 and aOR = 0.24, 95% CI = 0.10–0.55, respectively).

Among all women 18–50 years of age reporting infertility only or difficulty staying pregnant only, the weighted number of women seeking treatment was 334,147 (35.8%), whereas the weighted number of women seeking treatment or consultation was 459,790 (49.3%) (Table 4). Among women who reported difficulty staying pregnant only, the majority did not seek treatment (73.1%); however, 8.0% sought a consultation only, whereas 18.9% sought infertility treatment including infertility drugs, intrauterine insemination, assisted reproductive technology, or surgical intervention. In contrast, almost half (49.6%) of women who reported infertility only sought out infertility treatment, whereas 18.0% sought a consultation only and 32.4% did not seek treatment. Of note, the prevalence of healthcare coverage was lower among those women not seeking treatment (84.5%) compared with those seeking treatment (96.3%) or consultation only (99.0%) (Table 5).

Within our study sample, 4.8% of women reported infertility only and 4.0% reported difficulty staying pregnant only, for a combined 8.8% of women ever experiencing impaired fecundity, defined here as ever having experienced infertility or ever having experienced difficulty staying pregnant (but not both). When including the women experiencing both, the percentages ever experiencing infertility and impaired fecundity are 6.3% and 11.0%, respectively. Population-based estimates of infertility and impaired fecundity are sparse and lack common definitions, limiting the ability to compare estimates across sources. The 2011–2013 NSFG data showed 6.1% of married 15–44-year-old women experiencing infertility and 12.3% of all 15–44-year-old women experiencing impaired fecundity.⁹ However, there are many differences between the NSFG estimates and the BRFSS estimates that limit their comparability, such as different definitions of infertility and impaired fecundity as well as the BRFSS being limited to seven states rather than providing a national estimate.¹³ In addition, the BRFSS analysis explored ever experiencing infertility or impaired fecundity (i.e., lifetime infertility and impaired fecundity), restricted analysis to women having ever tried to get pregnant, and excluded women experiencing both infertility and difficulty staying pregnant. Finally, the age range of included women also differed (18–50 years of age in our study compared with 15–44 years of age in NSFG). The lower age limit in BRFSS is higher; however, as women attempt to conceive at increasing ages, we increased the upper limit for reproductive age to 50 years.

In our study, factors found to be associated with either infertility only or difficulty staying pregnant only included history of depressive disorder, history of smoking, and healthcare coverage. For instance, women with a history of depressive disorder were more likely to report difficulty staying pregnant only and infertility only, women who had ever smoked at least 100 cigarettes were more likely to report difficulty staying pregnant only, and women without healthcare coverage were less likely to report infertility only. Despite controlling for age and marital status in the multivariable model, a secondary analysis of married women 35–50 years of age was conducted to isolate a group more likely to have experienced and identified an infertility problem in their lifetime. The increased risk of impaired fecundity among those with a history of depressive disorder and the decreased risk among those without healthcare coverage persisted; however, the association with smoking did not. Smoking may be related to age and marital status, such that restricting the sample reduced the effects of smoking; however, the reduced statistical power resulting from the smaller sample size may have also played a role.

Because of the cross-sectional nature of this study, we cannot determine the directionality of the observed associations. Although impaired fecundity may lead to depression in women trying to achieve pregnancy, it is also possible that a history of depression may affect fertility. Women without healthcare coverage may be less likely to have discussed issues surrounding impaired fecundity with a healthcare provider, and therefore, be less likely to classify themselves as infertile. Other studies support the findings of an association between infertility and history of depressive disorder as well as associations between smoking and infertility and difficulty staying pregnant.^14–20 Additional female risk factors for infertility and difficulty staying pregnant reported in other studies that were not explored or not found to be associated with infertility with the BRFSS data include abnormal menstrual cycle patterns and menstrual cycle length,^21,22 obesity,^18,23–29 intense physical activity,^28,29 lower folate intake,³⁰ alcohol consumption,^{17–19,28,31} diabetes,²³ cancer,^32–34 karyotypic disorders,¹⁷ immunological rejection,²³ thrombophilias,^17,23 endocrinological issues,^17,35,36 infections,^18,23 environmental pollutants,¹⁸ and delivery history.¹⁹

In our study, the majority of women ever experiencing difficulty staying pregnant only did not seek infertility treatment. In contrast, approximately half of women ever experiencing infertility only did seek infertility treatment, with an additional 18.0% undergoing a treatment consultation only. In addition to the different demographic characteristics of these two groups of women, a reason for this difference also may be that more treatments are designed for infertility and not for difficulty maintaining a pregnancy. The 2011–2013 NSFG showed that 11.3%, or 6.9 million women aged 15–44 years, have ever received any infertility services,⁹ and that the percentage of women aged 25–44 years with current fertility problems seeking treatment to help get pregnant or prevent miscarriage has been slightly >40%.³⁷ Infertility services assessed in the NSFG include advice, infertility testing, ovulation drugs, surgery, artificial insemination, or other services such as assisted reproductive technology. Although the BRFSS statistics are not directly comparable because of differences in the populations and questions, it is important to note that our study yielded similar results with 35.8% of the study sample ever experiencing infertility or difficulty getting pregnant having sought treatment and 49.3% having sought treatment or consultation.

There are several limitations affecting this analysis. First, the BRFSS collects self-reported data from a population 18 years of age and older. The reliability of self-reported data is dependent on the ability of the respondent to accurately recall and relay the questioned information. In addition, the age of the population meant that we were not able to use the classic definition of reproductive age (15–44 years old), which, in combination with the way infertility was defined, made it difficult to compare our results with other studies. Second, the reproductive health questions were included in 2013 BRFSS as state-added questions by seven states that were not randomly selected. It was a convenience sample of states willing to include these questions. As a result, our findings are not generalizable to all states and no inferences related to the U.S. population could be made. Third, although one objective of the study was to explore infertility treatment use among women with infertility or impaired fecundity, the small sample size prevented us from looking in depth at the types of treatment sought and from assessing the relationship between sociodemographic factors and treatment use/types of treatment sought. Fourth, we excluded women reporting both infertility and difficulty staying pregnant. Fifth, the questions only assess infertility as experienced by the female respondent and exclude infertility experienced by the male respondent only. Sixth, the question assessing treatment use was asked of female respondents only, but specified treatment among either the respondent or the partner/spouse. Finally, the response rate for the BRFSS is low.

The study also has several strengths. The BRFSS is a well-known state-based survey that has been widely used by public health professionals and leaders to assess and monitor behaviors and health-related issues at the population level. We were able to use BRFSS data collected in seven states that added reproductive health questions on reproductive history, family planning, and infertility to assess infertility-related issues in this analysis. The findings of this study will be used to review and adjust, as necessary, the reproductive health questions that may be added in the future to BRFSS and by more states, and also to other surveys. The use and expansion of state-based findings can impact state policies, such as insurance coverage and mandates, and can demonstrate variation in infertility itself.

Existing and ongoing surveillance systems such as BRFSS provide an opportunity to obtain population-based measures to estimate the state-specific burden of reproductive health-related issues, including infertility. In addition, they allow us to examine associations between sociodemographic, behavioral and health factors, and infertility and to explore healthcare access and utilization. Gathering state-specific data and expanding the use of such systems enable tracking of trends and allow for comparability of measures across states, and could be adapted for local use. Understanding the burden of infertility and how treatment and services are being accessed could allow states to address unmet need. In addition, understanding how modifiable behaviors or treatable health conditions are related to infertility could help states implement prevention strategies that may reduce the burden of infertility.