Background
Approximately 12% of women of reproductive age have some type of disability. Very little is known about sexual and reproductive health issues among women with disabilities, including what proportion of women with disabilities experience pregnancy. Data on pregnancy are important to inform needs for preconception and pregnancy care for women with disabilities.
Objective
The purpose of this study was to describe the occurrence of pregnancy among women with various types of disability and with differing levels of disability complexity, compared with women without disabilities, in a nationally representative sample.
Study Design
We conducted cross-sectional analyses of 2008–2012 Medical Expenditure Panel Survey annualized data to estimate the proportion of women aged 18–44 years with and without disabilities who reported a pregnancy during 1 year of participation on the survey panel. We used a multivariable logistic regression to test the association of pregnancy with presence, type, and complexity of disability, controlling for other factors associated with pregnancy.
Results
Similar proportions of women with and without disabilities reported a pregnancy (10.8% vs 12.3%, with 95% confidence intervals overlapping). Women with the most complex disabilities (those that impact activities such as self-care and work) were less likely to have been pregnant (adjusted odds ratio, 0.69, 95% confidence interval, 0.52–0.93), but women whose disabilities affected only basic actions (seeing, hearing, movement, cognition) did not differ significantly from women with no disabilities.
Conclusion
Women with a variety of types of disabilities experience pregnancy. Greater attention is needed to the reproductive health care needs of this population to ensure appropriate contraceptive, preconception, and perinatal care.
Nearly 57 million Americans have a disability, constituting approximately 19% of the United States population. Disability is more common among older individuals, but approximately 12% of US women of child-bearing age have some type of disability. Despite the size of the disability population, the sexual and reproductive health needs of this population have largely been ignored. Sexuality in people with disabilities has historically been viewed as either nonexistent or dangerous; thus, there has been little impetus until recently to develop a knowledge base regarding the reproductive health of people with disabilities.
Disability is conceptually defined as the interaction of an individual’s impairments with characteristics of the environment, resulting in restricted ability to carry out social roles or access needed services. In practical measurement terms, however, disability is typically assessed through self-report of difficulty performing basic functions such as movement, vision, hearing, or cognition. Each of these broad categories of disability may also include difficulty with more complex tasks such as activities of daily living (ADLs, eg, bathing or dressing), instrumental activities of daily living (eg, shopping or preparing meals), or participation in social roles such as work or recreation.
Numerous studies have documented disparities between women with and without disabilities in receipt of one aspect of care related to reproductive health: Papanicolaou testing to detect cervical cancer. There is, however, much fewer data available on other components of sexuality and reproductive health of women with disabilities, including sexual behavior, needs for and use of contraception, family planning decision making, pregnancy, and maternal and infant birth outcomes.
Historically, women with disabilities have been discouraged from having children. Although practices such as involuntary sterilization are now illegal, women with various types of disabilities have reported discouraging responses from health care providers and others regarding potential pregnancy and a profound lack of health care provider knowledge about how their disability and pregnancy may interact.
Very little is known about how common pregnancy is among women with disabilities. Recent evidence suggests that women with and without disabilities were equally likely to want children, but women with disabilities were less certain about their intentions to have children. Findings limited to current pregnancy at the time of the interview found a lower prevalence of pregnancy among women with chronic physical disabilities compared with those without disabilities, but pregnancy prevalence was similar after adjusting for other sociodemographic characteristics associated with pregnancy. An analysis of data from women with spinal cord injuries found that 2% reported a pregnancy during the past 12 months, similar to the rate of current pregnancy among all women with physical disabilities.
Important gaps remain in our knowledge of the sexual and reproductive health of women with disabilities. No research has documented the rate of unintended pregnancy among women with disabilities. Nor has research to date examined pregnancy rates among women with cognitive or sensory (vision or hearing) disabilities. Such data are important to understand the needs for pregnancy care, and to ensure adequate access to contraception to prevent unintended pregnancies, among women with disabilities.
Reproductive health care providers need evidence to inform discussions about contraception, child-bearing plans, preconception care, and prenatal care for women with disabilities. The purpose of this study was to describe the occurrence of pregnancy among women with various disability types and differing levels of disability complexity, compared with women without disabilities.
Materials and Methods
Data source
We used existing data from the Medical Expenditure Panel Survey (MEPS) Household Component. MEPS data are collected by the Agency for Healthcare Research and Quality to provide nationally representative data on health and utilization of healthcare among non-institutionalized individuals. The MEPS uses an overlapping panel design with a new panel selected each year from the previous year’s National Health Interview Survey sample. Panel members complete 5 in-person interviews over a 2-year period. Thus, MEPS can capture a relatively broad range of pregnancy occurrences, including those that are short lived.
The Agency for Healthcare Research and Quality creates full-year consolidated files weighted to provide annualized US population estimates. These files can be pooled across multiple years to increase sample size. We conducted cross-sectional analyses of combined full-year files from 2008 through 2012, the available years in which a variable about pregnancy was included in the public use data sets. Because data are deidentified and publicly available, the Institutional Review Board of Oregon Health and Science University classified the study as nonhuman subjects research.
Sample
Our analyses included women aged 18–44 years. The combined 2008–2012 MEPS data set included a total of 31,964 women in this age range, 27,567 of whom had nonmissing data on all variables of interest for our analyses (see Figure ). The proportion of women reporting a pregnancy did not significantly differ between women excluded due to missing data on covariates (described in the following text) and women with complete data.
Measures
Dependent variable
A pregnancy indicator is available for each MEPS round, specifying whether any female household members were pregnant during that interview round (“Since [start date], has anyone in the family been pregnant at any time?”). If yes, a follow-up question was asked to ascertain who in the household was pregnant. We coded as pregnant any females who had a pregnancy indicated in any round.
Disability variables
We created 3 disability variables, each of which served as a primary predictor variable in separate analyses. The first was a dichotomous variable noting presence or absence of any disability as indicated by limitations in physical functions, vision, hearing, or cognition. This operationalization of disability was based on a report of Altman and Bernstein on the concept of basic action difficulties, which are limitations in movement, sensory, cognitive, or emotional functioning. We did not include emotional functioning because the MEPS household interview covers only limitations in movement, sensory, and cognitive functions.
Basic action difficulties were identified by affirmative responses to 1 or more MEPS survey questions about the following: (1) any degree of difficulty with physical functions such as walking, standing, bending, lifting, reaching, or grasping; (2) any difficulty seeing (while wearing glasses, if used); (3) any difficulty hearing (with a hearing aid, if used); and (4) any cognitive limitations such as confusion, memory loss, or difficulty making decisions. Women with no reported limitations were coded as having no disability (reference group).
Second, we created a 6-category variable specifying what type of disability (ie, basic action difficulty) was present, if any. Categories included the following: no disability (reference), physical limitation only, vision limitation only, hearing limitation only, cognitive limitation only, or multiple types of limitations. Disabilities can range considerably in the extent to which they limit activities. Therefore, our third disability variable indicated the extent of impact on daily life and was modeled on the description by Altman and Bernstein of complex activity limitations as restrictions in ability to participate in social roles or self-care activities.
Women were coded as having a complex activity limitation if they had positive responses to the MEPS items about needing assistance with ADLs or instrumental activities of daily living or being limited in work, housework, or social or recreational activities. The variable included the following 3 categories: (1) no disability (reference); (2) basic action difficulties only; and (3) complex activity limitations. This 3-level variable and the 6-level disability type variable are not mutually exclusive; women with any disability type could also have a complex activity limitation.
Covariates
Demographic covariates in our multivariable models included age group (18–24, 25–34, and 35–44 years) and race/ethnicity (non-Hispanic white, all other race and ethnicity groups). Sociodemographic covariates included marital status (currently married or not), family income as percent of the federal poverty level (≥ 400%, 200% to < 400%, 100% to < 200%, < 100%), education (any education beyond high school, high school diploma [HS]/general educational development [GED] or less), and employment status (employed or not employed).
We included 2 health care access variables that could affect the ability to obtain contraception, thereby having an impact on pregnancy. These variables were health insurance (insured all year with any portion of that being private insurance, publicly insured all year, uninsured for part or all of the year) and whether the woman had a usual source of health care. Additional covariates included perceived health status (excellent/very good/good vs fair/poor), region of the United States (Northeast, Midwest, South, West), and survey year.
Statistical analysis
We calculated the proportion (with 95% confidence interval) of women who had been pregnant within each category of our 3 measures of disability. We then developed multivariable logistic regression models with any reported pregnancy as our outcome. Regression analyses examined odds of pregnancy by overall disability (any vs none), disability type, and disability complexity, controlling for covariates described in the previous text. All analyses used Stata version 12.1 (Stata Corp, College Station, TX) with Taylor series linearization and incorporated survey weights to account for the complex survey design of MEPS.
Results
Table 1 shows the proportion of women with reported pregnancy by overall disability and disability type along with other sample characteristics for each group. Table 2 presents similar information by disability complexity. Approximately 12.3% of our sample had some type of disability. Women with disabilities were less likely to be married and more likely to be non-Hispanic white. They also tended to be older, poorer, less educated, less employed, and less healthy than women without disabilities. However, the majority of women with disabilities reported positive perceived health, consistent with prior survey research and with conceptual models distinguishing disability from poor health.
| Variable | None (n = 24,358) a | Any (n = 3209) b | Vision (n = 692) c | Hearing (n = 254) d | Physical (n = 1036) e | Cognitive (n = 417) f | Multiple (n = 810) g | Total (n = 27,567) |
|---|---|---|---|---|---|---|---|---|
| Weighted % (95% confidence interval) | ||||||||
| Pregnant while on panel | 12.3 (11.7–12.9) | 10.8 (9.4–12.3) | 15.3 (11.3–20.2) | 9.8 (6.4–14.6) | 10.6 (8.4–13.2) | 9.2 (6.3–13.1) | 8.4 (6.1–11.3) | 12.1 (11.5–12.7) |
| Age, y | ||||||||
| 18–24 | 26.6 (25.5–27.7) | 18.0 (16.2–19.9) | 26.5 (22.4–31.1) | 14.8 (9.6–22.2) | 14.5 (12–17.3) | 21.8 (16.6–28.2) | 14.4 (10.7–19.2) | 25.5 (24.5–26.6) |
| 25–34 | 37.5 (36.2–38.7) | 33.9 (31.6–36.2) | 38.0 (32.9–43.4) | 33.7 (26.3–41.9) | 33.3 (29.5–37.2) | 34.2 (28.4–40.6) | 30.9 (26.4–35.9) | 37.0 (35.9–38.2) |
| 34–44 | 35.9 (34.8–37) | 48.2 (45.5–50.8) | 35.5 (30.1–41.2) | 51.5 (43.7–59.2) | 52.3 (47.9–56.6) | 43.9 (37.1–51) | 54.6 (48.9–60.3) | 37.4 (36.4–38.5) |
| Married | 48.3 (47.1–49.6) | 40.8 (38.3–43.2) | 43.0 (37.5–48.7) | 49.9 (41–58.8) | 45.1 (41.3–48.9) | 30.3 (23.7–37.9) | 35.0 (29.7–40.7) | 47.4 (46.2–48.6) |
| Family income as percentage of poverty level | ||||||||
| ≥ 400% | 34.7 (33.3–36.2) | 23.8 (21.2–26.6) | 24.0 (19–29.8) | 39.7 (30.9–49.2) | 25.1 (21.1–29.7) | 18.8 (12.2–27.7) | 18.8 (14.1–24.5) | 33.4 (32–34.7) |
| 200% to < 400% | 32.3 (31.3–33.2) | 27.8 (25.6–30.1) | 29.7 (25.3–34.4) | 27.4 (20.7–35.3) | 30.9 (27.4–34.5) | 26.5 (21.1–32.6) | 23.0 (19–3,27) | 31.7 (30.8–32.6) |
| 100% to < 200% | 18.0 (17.2–18.7) | 21.9 (20.1–23.8) | 23.5 (19.6–27.9) | 17.6 (13–23.3) | 19.7 (17–22.8) | 20.1 (15.9–25.2) | 25.6 (21.4–30.2) | 18.4 (17.7–19.2) |
| < 100% | 15.1 (14.1–16.1) | 26.6 (24.4–28.9) | 22.9 (19.4–26.8) | 15.4 (10.7–21.6) | 24.3 (20.4–28.7) | 34.6 (28.5–41.3) | 32.7 (28.1–37.6) | 16.5 (15.5–17.5) |
| Non-Hispanic white | 59.8 (57.6–62) | 66.8 (64.1–69.4) | 63.3 (57.8–68.4) | 75.3 (68.2–81.2) | 68.0 (63.9–71.9) | 62.3 (55.2–68.9) | 67.5 (62.5–72) | 60.7 (58.5–62.8) |
| Greater than high school education | 44.4 (42.8–46.0) | 35.3 (32.4–38.3) | 39.9 (34.3–45.9) | 50.8 (41.7–60.0) | 39.4 (35.5–43.5) | 28.1 (20.9–36.6) | 24.0 (19.7–28.7) | 43.3 (41.7–44.8) |
| Employed | 70.4 (69.1–71.6) | 58.2 (55.4–60.9) | 71.5 (65.9–76.5) | 78.0 (71.4–83.5) | 63.0 (58.6–67.2) | 40.2 (33–47.8) | 42.1 (37.2–47.1) | 68.9 (67.7–70) |
| Health insurance | ||||||||
| Insured all year (any private) | 63.6 (62–65.2) | 49.1 (46.1–52.1) | 56.0 (50.5–61.4) | 63.5 (55.4–71) | 54.9 (50.3–59.4) | 36.8 (29.6–44.6) | 36.5 (31.5–41.7) | 61.9 (60.3–63.4) |
| Publicly insured all year | 9.1 (8.3–10.1) | 24.1 (21.5–26.8) | 9.3 (7.1–12) | 11.7 (7.4–17.8) | 21.3 (17.6–25.4) | 37.7 (30.6–45.5) | 38.4 (33.1–44) | 11.0 (10–12) |
| Uninsured part/all year | 27.2 (25.8–28.7) | 26.8 (24.5–29.3) | 34.7 (30.2–39.6) | 24.8 (19.4–31.2) | 23.8 (20.7–27.3) | 25.5 (20.2–31.5) | 25.1 (20.8–30) | 27.2 (25.8–28.6) |
| Have usual source of care | 71.5 (70.3–72.7) | 78.1 (75.8–80.2) | 67.5 (62.1–72.5) | 84.8 (77.8–89.9) | 83.1 (79.6–86.1) | 74.1 (68.9–79) | 80.4 (75.2–84.7) | 72.3 (71.2–73.5) |
| Fair/poor health | 6.0 (5.6–6.5) | 33.0 (30.7–35.2) | 11.6 (8.8–15.3) | 16.6 (11.6–23.2) | 35.4 (31.5–39.4) | 30.6 (24.7–37.2) | 54.9 (49.7–60) | 9.3 (8.8–9.9) |
| Region | ||||||||
| Northeast | 24.0 (22.3–25.9) | 22.8 (20.5–25.3) | 22.2 (17.7–27.4) | 17.9 (13.3–23.7) | 22.4 (18.4–27) | 27.6 (21.7–34.4) | 23.3 (19.1–28.1) | 23.9 (22.2–25.6) |
| Midwest | 37.4 (35.4–39.4) | 36.9 (33.3–40.7) | 41.2 (35.3–47.2) | 34.6 (26.9–43.1) | 35.4 (30.6–40.5) | 31.4 (24.5–39.3) | 38.6 (32.5–45.2) | 37.3 (35.3–39.3) |
| South | 21.0 (19.6–22.5) | 23.6 (20.1–27.4) | 21.4 (16.7–26.8) | 30.1 (21.8–43.1) | 24.3 (19.6–29.6) | 22.0 (16.2–29.1) | 23.2 (18.3–28.8) | 21.3 (19.8–22.9) |
| West | 17.6 (16.4–18.9) | 16.7 (14.6–19.1) | 15.3 (10.7–21.4) | 17.5 (11.4–25.9) | 17.9 (14.7–21.7) | 19.0 (13.5–26.2) | 14.9 (1.11–19.7) | 17.5 (16.3–18.8) |
| Year | ||||||||
| 2008 | 21.0 (20.2–21.8) | 21.9 (20.2–23.7) | 19.2 (15.9–23) | 24.4 (18.1–31.9) | 22.1 (18.9–25.7) | 21.7 (16.5–27.9) | 23.3 (19.9–27) | 21.1 (20.3–21.8) |
| 2009 | 21.0 (20.5–21.6) | 22.2 (20.7–23.8) | 24.1 (20.3–28.4) | 28.2 (22.5–34.8) | 19.6 (17.1–22.3) | 20.2 (15.9–25.3) | 22.7 (19.3–26.4) | 21.2 (20.7–21.7) |
| 2010 | 21.3 (20.6–22) | 20.1 (18.5–21.9) | 19.8 (16–24.4) | 19.0 (13.9–25.3) | 21.7 (18.4–25.5) | 18.9 (14.9–23.6) | 19.3 (16.6–22.4) | 21.2 (20.6–21.7) |
| 2011 | 18.4 (17.8–19) | 18.3 (16.6–20.1) | 19.4 (15.4–24.1) | 12.5 (9.1–17.1) | 19.0 (16–22.4) | 21.0 (16.1–27) | 17.1 (13.7–21.1) | 18.3 (17.8–18.9) |
| 2012 | 18.3 (17.7–19) | 17.5 (16.1–19.1) | 17.5 (14.4–21) | 15.9 (11.2–22.1) | 17.6 (14.9–20.6) | 18.2 (13.9–23.4) | 17.7 (14.5–21.3) | 18.2 (17.6–18.8) |
| Variable | No disability (n = 24,358) | Basic only (n = 1901) | Complex (n = 1308) | Total (27,567) | |
|---|---|---|---|---|---|
| Weighted % (95% confidence interval) | |||||
| Pregnant while on panel | 12.3 (11.7–12.9) | 12.8 (10.7–15.2) | 7.8 (6.3–9.8) | 12.1 (1.5–12.7) | |
| Age, y | |||||
| 18–24 | 26.6 (25.5–7.7) | 20.2 (17.9–2.26) | 14.7 (11.5–18.6) | 25.5 (24.5–26.6) | |
| 25–34 | 37.5 (36.2–38.7) | 35.7 (32.6–38.9) | 31.1 (27.8–34.6) | 37.0 (35.9–38.2) | |
| 34–44 | 35.9 (34.8–37.0) | 44.2 (40.8–47.6) | 54.2 (49.8–58.5) | 37.4 (36.4–38.5) | |
| Married | 48.3 (47.1–49.6) | 45.4 (42.4–48.4) | 33.8 (29.7–38.2) | 47.4 (46.2–48.6) | |
| Family income as percentage of federal poverty level | |||||
| ≥ 400% | 34.7 (33.3–36.2) | 26.7 (23.5–30.1) | 19.4 (15.7–23.7) | 33.4 (32.0–34.7) | |
| 200% to < 400% | 32.3 (31.3–33.2) | 30.7 (27.9–33.8) | 23.3 (20.3–26.6) | 31.7 (30.8–32.6) | |
| 100% to < 200% | 18.0 (17.2–18.7) | 21.5 (19.2–23.9) | 22.6 (19.9–25.5) | 18.4 (17.7–19.2) | |
| < 100% | 15.1 (14.1–16.1) | 21.1 (18.9–23.5) | 34.7 (30.5–39.2) | 16.5 (15.5–17.5) | |
| Non-Hispanic white | 59.8 (57.6–62.0) | 65.7 (62.6–68.7) | 68.4 (64.6–71.9) | 60.7 (58.5–62.8) | |
| Completed education beyond high school | 44.4 (42.8–46.0) | 41.5 (37.8–45.2) | 26.0 (22.2–30.1) | 43.3 (41.7–44.8) | |
| Employed | 70.4 (69.1–71.6) | 73.4 (70.8–75.9) | 35.2 (31.0–39.6) | 68.9 (67.7–70.0) | |
| Health insurance | |||||
| Insured all year (any private) | 63.6 (62.0–65.2) | 57.7 (54.1–61.3) | 36.1 (31.8–40.6) | 61.9 (60.3–63.4) | |
| Publicly insured all year | 9.1 (8.3–10.1) | 11.9 (10.1–13.9) | 42.5 (37.8–47.3) | 11.0 (10.0–12.0) | |
| Uninsured part or all year | 27.2 (25.8–28.7) | 30.4 (27.4–33.6) | 21.4 (18.1–25.1) | 27.2 (25.8–28.6) | |
| Have usual source of care | 71.5 (70.3–72.7) | 74.8 (72.0–77.5) | 83.0 (79.1–86.3) | 72.3 (71.2–73.5) | |
| Fair/poor perceived health | 6.0 (5.6–6.5) | 20.9 (18.6–23.3) | 51.2 (47.2–55.2) | 9.3 (8.8–9.9) | |
| Region | |||||
| Northeast | 24.0 (22.3–25.9) | 21.2 (18.5–24.2) | 25.1 (21.7–28.9) | 23.9 (22.2–25.6) | |
| Midwest | 37.4 (35.4–39.4) | 37.6 (33.7–41.7) | 35.9 (30.9–41.3) | 37.3 (35.3–39.3) | |
| South | 21.0 (19.6–22.5) | 25.7 (21.2–30.7) | 20.4 (16.7–24.7) | 21.3 (19.8–22.9) | |
| West | 17.6 (16.4–18.9) | 15.5 (12.9–18.4) | 18.5 (14.7–23.1) | 17.5 (16.3–18.8) | |
| Year | |||||
| 2008 | 21.0 (20.2–21.8) | 21.0 (19.0–23.3) | 23.2 (20.3–26.4) | 21.1 (20.3–21.8) | |
| 2009 | 21.0 (20.5–21.6) | 22.2 (20.1–24.5) | 22.1 (19.5–25.1) | 21.2 (20.7–21.7) | |
| 2010 | 21.3 (20.6–22.0) | 20.9 (18.7–23.2) | 19.1 (16.6–21.8) | 21.2 (20.6–21.8) | |
| 2011 | 18.4 (17.8–19.0) | 19.1 (16.7–21.7) | 17.1 (14.9–19.5) | 18.3 (17.8–18.9) | |
| 2012 | 18.3 (17.7–19.0) | 16.8 (14.9–19.0) | 18.5 (16.2–21.1) | 18.2 (17.6–18.8) | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
