The prevalence of infertility in American Indian/Alaska Natives and other racial/ethnic groups: National Survey of Family Growth.

We observed increased infertility among black women, and increased infertility and impaired fecundity among AI/AN women, relative to whites. No differences in prevalence of infertility or impaired fecundity were observed for other racial groups compared to whites or by Hispanic ethnicity. Our results demonstrating increased infertility and impaired fecundity among black women compared to white women were consistent with both 2002 (adjusted OR: 1.98, 95% CI: 1.28, 3.06) and 2006–2010 (adjusted OR: 1.84, 95% CI: 1.10, 3.06) NSFG estimates, indicating consistency over time. 8 , 9 Our results were also consistent with a prospective study, which observed an association with infertility among black compared to white women, excluding surgically sterile women (adjusted OR: 1.97, 95% CI: 1.25, 3.12). 10 No other studies have evaluated the prevalence of infertility and impaired fecundity among AI/AN women. Disparities in infertility are likely due to differential distributions of factors such as age, education, socioeconomic status, health behaviors, acute and chronic health conditions, exposure to environmental and infectious agents, access to quality infertility services, and service-seeking behaviors, though research is limited. 1 A strength was the ability to evaluate infertility and impaired fecundity over several NSFG cycles. One limitation is the classification of infertility and impaired fecundity. Measures of infertility were self-reported and do not equate with medical evaluation and treatment. NSFG participants were not queried specifically about infertility. Instead, it was constructed from survey questions addressing marital status, sexual activity, and contraceptive use. In addition, the NSFG definition of infertility was restricted to married or cohabiting women ages 15–44 years who have not used contraception in the previous 12 months, which limits the generalizability of the estimates. Inequalities in the burden of reproductive impairments among blacks and AI/AN women warrant further evaluation of the contribution of underlying causes of infertility/impaired fecundity and opportunities to reduce disparities. Given mounting evidence of maternal/child health disparities among AI/ANs, 13 – 15 it is imperative to understand the contribution of infertility to the reproductive health status of this underserved population.

We used pooled data from the NSFG survey cycles 2002, 2006–2010, and 2011–2013 to examine racial/ethnic variation in infertility prevalence. The NSFG, a national survey using a multistage probability design, is conducted through in-person interviews by the CDC’s National Center for Health Statistics including individual participants in each cycle. 11 , 12 The pooled NSFG cycles gathered information on infertility and impaired fecundity among men and women ages 15–44 years. We analyzed female respondent data from the pooled NSFG data. Racial groups were categorized by NSFG as 1) AI/AN; 2) Asian/PI; 3) black; or 4) white. NSFG respondents self-reported their racial group and were allowed to choose up to four races. Respondents reporting multiple races were asked to choose the racial group that best described them, which we used in our analyses. Ethnicity was categorized as Hispanic or Non-Hispanic (NH). We examined differences in the prevalence of infertility and impaired fecundity by race/ethnic group. The prevalence of infertility, a couple-based measure, was our primary outcome of interest. Consistent with NSFG definitions, 8 infertility was defined as “12 or more months of intercourse without pregnancy and without contraception” among women who were married or cohabiting. Women in this analysis were also classified as surgically sterile or presumed fertile (residual of women who did not meet the definition of infertile or surgically sterile, but are married/cohabiting) and were included in the denominator, which is consistent with previous studies. 8 , 9 For respondents with missing infertility data, the NSFG used regression imputation (n=13) and logical imputation conducted by subject-matter experts at the National Center for Health Statistics (n=49) to impute responses. Impaired fecundity was defined as all women, regardless of marital or cohabiting status, who (i) reported it was not physically possible to conceive (excluding surgical sterilization); (ii) were classified as subfecund (i.e., difficult for a couple to conceive/deliver a baby or received medical advice to not become pregnant); or (iii) had a long interval without conception (i.e., no pregnancy in the 36 months prior to interview when married/cohabiting continuously, without contraception, and no months without intercourse). Women were classified as fecund if the respondent was currently pregnant and her husband/partner was not surgically sterile or she did not meet the criteria for one of the other impaired fecundity categories. Women with impaired fecundity or who were surgically sterile were included in the denominator for consistency with prior reporting. 8 The respondent was classified as surgically sterile if this was reported during the interview. All respondents provided a response to the question on impaired fecundity with no imputation conducted. We obtained demographic characteristics from the NSFG, which included age at interview, marital status (currently married to a person of the opposite sex, not married but living with opposite sex partner, widowed, divorced or annulled, separated for reasons of marital discord, never been married), education (years), poverty level (<150, 150–299, and ≥300 percent of the Federal Poverty Level of the survey year), parity (number of live births), body mass index (BMI), smoking during the last three months (none, <1/2 pack per day, ≥1/2 pack per day), pelvic inflammatory disease treatment (ever/never), use of Depo-Provera (ever/never), age at menarche (years), any medical help to become pregnant (ever/never), total family income of the respondent in the calendar year before the interview (<$14,999, $15,000-$34,999, $35,000-$59,999, ≥$60,000), and gynecologic disorders (ever/never diagnosed with fibroids, endometriosis, or ovulation problems). All data analyses were conducted via remote access to the NCHS Research Data Center (RDC) using SAS v. 9.4. Initial analyses evaluated weighted counts, percentages, and 95% confidence intervals by infertility/impaired fecundity status. We used modified Poisson regression with robust error variance to assess the relationship between race and ethnicity and the prevalence of 1) infertility and 2) impaired fecundity accounting for the complex survey design to estimate prevalence proportion ratios (PPR). Covariates evaluated in the multivariable regression models included age, parity, marital status, education, poverty level, body mass index, smoking, pelvic inflammatory disease treatment, ever use of Depo-Provera, age at menarche, any medical help to become pregnant, income, and gynecologic disorders. For comparison with previous findings, 8 we also evaluated models adjusted for age, parity, marital status, education, and poverty level only. This study was reviewed by the University of Oklahoma Health Sciences Center Institutional Review Board and classified as exempt.

The prevalence of infertility and impaired fecundity was 6.4% (95% CI: 5.7%, 7.0%) and 11.0% (95% CI: 11.0%, 12.2%), respectively. When evaluating descriptive characteristics of women who were infertile or had impaired fecundity, many demographic, behavioral, and reproductive factors differed by infertility and fecundity status ( Table 1 ). Respondents with infertility were more often black (13%) compared to those who were fecund/surgically sterile (10%); approximately 5% of women in each group were AI/AN. Similar proportions (16%) of women with and without impaired fecundity were black; AI/AN women made up 6% and 5% of women with and without impaired fecundity, respectively. Compared to whites, blacks had a 1.45 times greater adjusted prevalence of infertility (95% CI: 1.15, 1.83) ( Table 2 ). AI/ANs had a 1.37 times greater prevalence of infertility compared to whites (95% CI: 0.91, 2.06), though the 95% confidence interval included the reference value of 1.0. The adjusted PPR for infertility among Asian/PI women compared to white women (PPR: 0.89, 95% CI: 0.53–1.49) was not increased. The adjusted PPR for Hispanic women compared to non-Hispanic women was 1.23 (95% CI: 0.95–1.58). In analysis of impaired fecundity, we observed a 1.30 times greater adjusted prevalence among AI/AN women (95% CI: 1.04, 1.62) compared to whites ( Table 3 ). No increased PPRs were observed for black and Asian/PI women compared to whites, or among Hispanic compared to non-Hispanic women.

Infertility is the inability to conceive within one year of unprotected intercourse and has been identified as a public health priority by the Centers for Disease Control and Prevention (CDC). 1 The CDC emphasizes that infertility is more than a quality-of-life issue with considerable public health consequences including psychological distress, social stigmatization, economic strain, and marital discord. Furthermore, infertility is associated with increased risk of subsequent chronic health conditions. 2 – 5 The public health importance of infertility is reinforced by the World Health Organization 6 and American Society for Reproductive Medicine, 7 which define infertility as a reproductive system disease. A report from the 2006–2010 National Survey of Family Growth (NSFG) estimated that 6% of married women aged 15–44 years in the U.S. are infertile and 12% have impaired fecundity, defined as the inability to conceive and carry a baby to term. 8 Racial disparities in infertility have been observed, with the highest prevalence reported for non-Hispanic black women. 8 Comparisons of the racial/ethnic burden of infertility, however, have been mostly limited to assessments of blacks, whites, and Hispanics, excluding subgroups such as American Indian/Alaska Native (AI/AN) and Asian/Pacific Islanders (Asian/PI). 9 , 10 As a result, the prevalence of infertility in AI/AN populations is unknown. Our objective was to evaluate the prevalence of infertility and impaired fecundity in the AI/AN population compared to other racial/ethnic groups to assess potential reproductive health disparities in this underserved population.