Planning for Motherhood: Fertility Attitudes, Desires and Intentions Among Women with Disabilities

Abstract

CONTEXT

An estimated 10% of U.S. women of reproductive age report a current disability; however, the relationship between disability, motherhood attitudes and fertility intentions among these women is largely unknown.

METHODS

Data from the 2006–2010 National Survey of Family Growth were used to examine attitudes toward motherhood and fertility intentions among 10,782 U.S. women aged 15–44. A series of regression models assessed, separately for mothers and childless women, associations between disability status and women's attitudes and intentions.

RESULTS

Women with and without disabilities held similar attitudes toward motherhood. Among women without children, women with and without disabilities were equally likely to want a child and equally likely to intend to have one. However, childless women with disabilities who wanted and intended to have a child were more likely to report uncertainty about those intentions than were childless women without disabilities (odds ratio, 1.7). Mothers with disabilities were more likely to want another child (1.5), but less likely to intend to have a child (0.5), than were mothers without disabilities.

CONCLUSIONS

Deepening understanding of the reproductive health desires, needs and challenges of women with disabilities is essential if the highest quality reproductive health services are to be provided for all.

An estimated 10% of women of reproductive age in the United States have a current disability,¹ the majority of whom have experienced a live birth.^*2,3 Although many types of disabling impairments have become more prevalent in recent decades,^4,5 the relationship between disability, motherhood attitudes and fertility intentions is largely unknown. A better understanding of these attitudes and intentions of women with disabilities is a necessary first step toward adequately addressing the reproductive health needs and challenges related to motherhood for women with disabilities, and toward recognizing and affirming motherhood as a viable option for all women.

Women with disabilities—physical limitations, cognitive limitations, sensory limitations and chronic health conditions—are a heterogeneous group with varying reproductive health needs, abilities in activities of daily living and challenges in health care access. Although underlying impairments differ, these women share many barriers to fertility and motherhood, including social stigma,^6,7 a lack of specialized sex education,^8,9 uninformed health practitioners^10,11 and challenges to legal rights to custody.^12,13 This article uses data from the 2006–2010 National Survey of Family Growth (NSFG) to examine how women with and without disabilities assess the perceived rewards and importance of motherhood in their lives. In addition, we examine women's fertility intentions and their certainty about those intentions. These outcomes are assessed in relationship to variables that may impact attitudes toward parenting and pregnancy—namely, fecundity, parity, partnership status and socioeconomic characteristics.

BACKGROUND

Women with disabilities in the United States have a long history of being denied both the choice to parent and their own reproductive freedom.^14,15 Despite the abolishment of compulsory sterilization programs, U.S. women with disabilities continue to face many barriers to fertility and motherhood. One is stigma. Sexual activity involving people with disabilities is perceived less positively than that involving people without disabilities,^7,16,17 as is parenting among individuals with intellectual, physical and sensory disabilities.^18–20 This lack of acceptance may leave women with disabilities with fewer social resources with which to develop and actualize their intentions.

Other barriers to fertility are likely to emerge in interactions with health care providers. Women with disabilities report being advised by their health care providers to terminate pregnancies with no consideration given to whether the pregnancies are wanted.^21,22 Although women with disabilities have gained greater protection over their right to reproduce since compulsory sterilization for the “feeble-minded” and “socially inadequate” was established by Buck v. Bell in 1927,²³ many caregivers still support it as a contraceptive option.¹⁸ Analysis of the 1992–1996 National Study of Women with Physical Disabilities found that women with physical disabilities had significantly higher rates of hysterectomy —and were more likely to have a hysterectomy for non–medically necessary reasons—than women without disabilities.²⁴ More recently, an analysis of the 2000, 2005 and 2010 National Health Interview Surveys found that women with multiple disabilities experienced higher risk of hysterectomy between ages 25 and 41 than women with one or no disability.²⁵ The focus on controlling women's reproductive capacity instead of supporting their reproductive desires is especially problematic given that health care providers are often unaware of the health care needs of women with disabilities, to the detriment of quality of care.^10,11,26 General medical care providers and obstetrician-gynecologists often lack understanding of the special needs created by women's disabilities, and specialists responsible for the care of disabling conditions are often unaware of women's fertility desires and the reproductive health issues that women with disabilities face during pregnancy.²⁷

Obstacles to receiving appropriate care are particularly encumbering given that many female adolescents with disabilities aspire to motherhood. Analyses of the 1997 National Longitudinal Survey of Youth reveal that female adolescents with either multiple or seriously limiting conditions are more likely than those without disabilities to want to get pregnant at first sexual intercourse²⁸ and more likely to expect a pregnancy by age 20.²⁹ Results from the National Longitudinal Study of Adolescent Health show that female adolescents with severe physical disabilities hold more positive attitudes toward pregnancy than those without physical disabilities.³⁰ Furthermore, Blum and colleagues’ research on male and female adolescents with spina bifida and cerebral palsy indicates that a majority of these youth thought about having children.⁸ Research on adults is limited, but a study from Ireland of pregnant women with physical, sensory or intellectual disabilities suggests that participants “welcomed pregnancy as affirming their identity and worth as women....”^21(p.156) These results indicate that motherhood is an important component of the life course for women with disabilities. It is also biologically attainable for women with many types of disabilities.^27,31,32

Attitudes toward motherhood, fertility desire and birth intentions are distinct outcomes; however, they are related. For example, fertility intentions are predictive of fertility behavior at the individual level in the United States, although this relationship varies by intended parity,³³ certainty about intentions,³⁴ and various family and contextual characteristics.³⁵ Our goal is not to speculate on how having a disability may affect the interrelation between attitudes, desires and intentions, but to better understand how disability might be associated with each of these outcomes in different ways. In sum, our analytic strategy focuses on exploring two questions: Do women's attitudes toward motherhood vary by disability status? Do women's fertility desires and intentions vary by disability status? Because women with disabilities also face barriers to participation in other key markers of the life course— particularly educational attainment, romantic relationships and employment^24,36,37—our analysis also considers these variables.

METHODS

Data and Samples

To answer these questions, it is essential to use data representing the entire population of reproductive-age women in the United States, including those who are and those who are not mothers. The NSFG, which conducts interviews with representative samples of women aged 15–44 living in U.S. households, is the premier survey for producing national estimates of variables associated with reproductive intentions and behaviors.³⁸ We used the most recent NSFG data, which were collected in repeated cross-sectional surveys between June 2006 and June 2010. The annual samples were drawn independently and were pooled and weighted to represent the population of American women of reproductive age from 2006 to 2010.

Prior research suggests that adults with and those without children differ in their evaluations of parenthood.³⁹ Therefore, we stratified participants aged 18 or older into the 6,670 respondents who had children (i.e., reported at least one adoption or live birth) and the 4,119 who did not have children (i.e., indicated no live births, adoptions, current adoption attempts or current pregnancy). Only nine women from the total sample provided invalid data on the disability measure and were excluded. Our final samples included 4,116 childless women and 6,666 mothers with valid information on all covariates. Missing data were negligible for all variables: No more than 3% of the eligible sample for each dependent measure was missing because of respondents’ refusing to reply or replying “don't know.” Not all women were asked all questions used as dependent variables in these analyses; therefore, the number of cases varied across models.

Measures

• Disability

The conceptual model of disability used here is based on the World Health Organization's International Classification of Functioning, Disability and Health model, which has been adopted “as the basis for the scientific standardization of data on health and disability world-wide.”^{40(p. 5)} The model describes disabilities as both impairments in body functions (physiological and psychological) and structure (anatomical), as well as limitations in activities and participation. Our focus is on disabilities that limit activities and participation, which include conditions that affect learning, communicating, walking, carrying, feeding, dressing, toileting, bathing, reading, and involvement in family or community life. The NSFG provides a single item to identify women who have activity limitations: “Are you limited in any way in any activities because of physical, mental, or emotional problems?” The measure refers to the date of interview; the onset of disability is not known. Accordingly, this article focuses on differences in the current attitudes, desires and intentions of women with and without disability at the time of the survey.

• Dependent variables

We examined three questions that measured women's attitudes toward motherhood. The first is asked of all women and assesses level of agreement with the statement “The rewards of being a parent are worth it despite the cost and work it can be.” The original response categories were “strongly agree,” “agree,” “disagree” and “strongly disagree”; the third and fourth categories were combined in the final measure because of small sample size. The second question, asked only in 2007–2010, assesses level of agreement with the statement “People can't be really happy unless they have children.” The original response categories were “strongly agree,” “agree,” “disagree” or “strongly disagree;” the first and second categories were combined because of small sample size. The third question is asked only of childless respondents: “If it turns out that you do not have any children, would that bother you a great deal, some, a little or not at all?” We examined these three indicators separately for two reasons. First, they represent distinct aspects of motherhood: The first two assess global notions of motherhood, while the last assesses personal reactions to childlessness. Second, and likely a reflection of this multidimensionality, Cronbach's alpha for these three items was low (0.31), indicating that combining them into a scale would be inappropriate because of insufficient internal consistency.

We also examined women's fertility desires and intentions. First, a dichotomous measure asks all women if they want a child (or another child). Next, among those who reply that they want a child and are not sterile, single women are asked if they intend to have a child, and married or cohabiting women are asked if they and their husband or partner intend to have a child; response categories are “yes” and “no.” Respondents who report that they intend to have a child are asked two additional questions of interest to our study. First, single women are asked the number of children they intend to have, and married and cohabiting women are asked the number of children they and their husband or partner intend to have; the number of children intended was measured continuously and top-coded in these analyses at six. Finally, these women are asked how sure they are that they will have a child. Response categories for this ordinal indicator are “very sure,” “somewhat sure” and “not at all sure.”

• Independent variables

Our analysis recognizes that women with disabilities and those without disabilities differ on numerous demographic and social indicators. Accordingly, our models controlled for the respondent's race or ethnicity, constructed as Hispanic; non-Hispanic black; or non-Hispanic white (which includes women of other races). Age was included as both linear and squared terms, centered first to reduce multicollinearity. We included this age specification because of evidence that women's fertility intentions change nonlinearly over time and because previous research finds a significant curvilinear relationship between age and fertility intentions.^33,41,42 Results were comparable to those obtained from models using a series of five-year age dummies. Combined family income was measured as percentage of the federal poverty level and was categorized as 0–199%, 200–399%, and 400% or more. Educational attainment measured highest level of schooling: less than high school, high school diploma or GED, or bachelor's degree or higher. Partnership status compared respondents who were currently married or cohabiting with those who were single (widowed, divorced, separated or never-married). Employment status was a dichotomous measure indicating if respondents were employed (including those temporarily on leave from a job at the time of survey) or not employed.

We also included fertility-related indicators that are important to understanding future desires and intentions. Fecundity was a dichotomous indicator; a woman was categorized as not fecund if she had been surgically sterilized, was sterile for other reasons, was subfecund (i.e., she and her partner had had difficulty conceiving, she had had difficulty delivering or a medical doctor had advised her never to become pregnant), or she was infertile (i.e., had not conceived in three years of partnership including monthly unprotected intercourse). Finally, among the sample of mothers, we also included a continuous indicator of number of children, which combined the number of live births to the respondent with the number of adopted children; results using this measure were comparable with those from models using a trichotomous indicator to compare one child with two children or three or more children (not shown).

Analytic Strategy

We examined whether the answers to our research questions varied according to motherhood status by stratifying the sample into childless women and mothers. We began our analysis by examining differences in means and percentage distributions by disability status for all dependent and independent variables. All estimates were weighted to account for survey design, with hypothesis tests reported from adjusted Wald tests for means and Pearson (Rao-Scott correction F statistic) chi-square tests for percentage distributions using Stata version 13.

Ordinal indicators, such as those examined here, are typically modeled using ordinal logistic regression—specifically, the proportional odds model.⁴³ However, the proportional odds assumption of equality between adjacent categories is often violated.⁴⁴ We first tested that we met this assumption in our weighted survey data using Williams’ gologit2 procedure in Stata.⁴⁵ Results indicated that all models met this assumption; therefore, all ordinal outcomes were estimated using ordinal logistic regression.

Binary logistic regression models were fitted for the dichotomous outcomes of wanting a child and intending to have a child. Ordinary least-squares models were fitted for the continuous measure of number of children intended. (Poor fit was achieved when the outcome was modeled as a count measure with a Poisson, negative binomial or gamma distribution; however, results were comparable.) All models were evaluated using robust standard errors and were weighted using the NSFG weight for the full sample of respondents interviewed in 2006–2010. Level of agreement with the statement “people can't be happy unless they have children,” which was asked only in 2007–2010, was alternately weighted using the weight for those years; calculations based on this weighting yielded highly comparable results.

Model diagnostics were assessed for all equations. Variance inflation factors indicated that multicollinearity was not a problem in our final analyses.⁴⁶ They were highest among the education indicators (4.2 and 3.2 for college and high school, respectively) and below 2.5 for all other coefficients. Influential cases in the unweighted proportional odds models were assessed using the cumulative probabilities approach suggested by Hosmer and Lemeshow.⁴⁷ Results for all models in the ordinal and binary logistic models were robust to the exclusion of influential cases, assessed using the cutoff of 1 for Pregibon beta values.⁴⁸ Fox's⁴⁹ cutoff for Cook's D in the linear models flagged most respondents intending to have three or more births; however, given the robustness of results to alternate count, ordinal and nominal model specifications and cutoff points, and previously established modeling strategies using this measure,⁵⁰ we present the linear model for ease of interpretation.

RESULTS

Bivariate Findings

In this sample, 10% of childless respondents and 11% of mothers have a disability. The bivariate analyses provide preliminary evidence that women with and without disabilities are similar in their attitudes about motherhood—but not in their fertility desires and intentions (Table 1). Childless women with disabilities were significantly less likely than childless women without disabilities to want a child (77% vs. 85%) and to be very sure about their intentions if they wanted and intended to have a child (54% vs. 67%). Mothers with disabilities were less likely than mothers without disabilities to intend to have a child if they wanted a child (67% vs. 83%), and less likely to be very sure about their intentions if they wanted and intended to have a child (35% vs. 49%).

TABLE 1.

Selected characteristics of women aged 15-44, by motherhood status and disability status, 2006-2010 National Survey of Family Growth

Characteristic	Childless women		Mothers
	No disability (N=3,690)	Disability (N=426)	No disability (N=5,942)	Disability (N=724)
Dependent variables
Parenting is rewarding
Strongly agree	46.5	39.6†	73.5	73.7
Agree	48.0	52.7	24.4	23.1
Disagree/strongly disagree	5.5	7.7	2.0	3.2
No happiness without children
Strongly agree/agree	3.8	5.4	10.5	7.1
Disagree	49.4	48.2	52.9	51.9
Strongly disagree	46.8	46.4	36.6	41.0
Would be bothered if never had children
A great deal	37.0	35.7	na	na
Some	30.0	29.7	na	na
A little	15.8	13.3	na	na
Not at all	17.2	21.4	na	na
Wants a child	85.3	76.5**	41.3	43.5
Intends to have a child‡	96.5	93.5†	82.6	66.6***
Mean no. of children intended§	1.3	1.4	1.3	1.2
Certainty of intentions§
Very sure	66.8	53.6***	49.4	34.5*
Somewhat sure	28.5	32.8	38.6	44.2
Not at all sure	4.7	13.6	12.0	21.3
Independent variables
Race/ethnicity
Non-Hispanic white††	76.2	81.3†	63.5	75.1***
Non-Hispanic black	11.2	10.7	16.0	14.2
Hispanic	12.6	8.0	20.5	10.7
Mean age	25.8	28.2***	33.8	35.6***
Income as % of poverty level
0-199	32.9	44.7**	48.8	59.2**
200-399	35.6	29.2	38.1	31.2
≥400	31.5	26.1	13.2	9.6
Educational attainment
<high school	9.9	11.5*	19.2	26.1***
High school	56.6	64.5	56.1	60.5
≥bachelor's degree	33.6	24.0	24.6	13.4
Partnership status
Single	68.2	65.7	26.1	35.3**
Married	21.1	22.1	61.2	52.1
Cohabiting	10.7	12.2	12.7	12.6
Employed	78.8	61.5***	69.2	51.1***
Not fecund	14.2	38.6***	47.9	66.4***
Mean no. of children	na	na	2.3	2.3

^*p<.05.

^**p<.01.

^***p<.001.

^†p<.10.

^‡Based on those who want a child and are not sterile.

^§Based on those who intend to have a child.

^††Includes women of other races.

Notes: Unless otherwise noted, data are percentages. Significance levels are results of Pearson chi-square or t tests (two-tailed). All results are weighted. na=not applicable.

Among both childless women and mothers, women with disabilities were significantly older, were more likely to live in or near poverty, had lower levels of educational attainment, were less likely to be employed and were more likely to be not fecund than women without disabilities. Mothers with disabilities were less likely to be married and more likely to be single than mothers without disabilities. Mothers with disabilities were more likely to be Non-Hispanic white or another race and less likely to be Hispanic than mothers without disabilities.

Multivariate Findings

• Motherhood attitudes

Childless women with disabilities do not differ from childless women without disabilities in their evaluation of the rewards of parenting, motherhood as important for happiness or how bothered they would be if they did not have children (Table 2). Similarly, among mothers, disability is not associated with belief in the rewards of parenting or that there is no happiness without children. These results emerge net of race, age, income, educational attainment, partnership status, employment, fecundity and (for mothers) number of children.

TABLE 2.

Odds ratios from ordinal logistic regression models assessing associations between selected characteristics and attitudes toward motherhood, by motherhood status

Characteristic	Childless women			Mothers
	Parenting is rewarding	No happiness without children	Would be bothered if never had children	Parenting is rewarding	No happiness without children
Disability‡	1.28 (0.19)	1.01 (0.19)	0.84 (0.12)	0.93 (0.14)	1.14 (0.17)
Race/ethnicity
Non-Hispanic white§ (ref)	1.00	1.00	1.00	1.00	1.00
Non-Hispanic black	1.27 (0.17)†	1.14 (0.17)	1.04 (0.13)	1.90 (0.20)***	1.25 (0.14)*
Hispanic	0.97 (0.12)	0.43 (0.07)***	1.21 (0.14)†	1.99 (0.21)***	0.35 (0.04)***
Age††	1.03 (0.01)***	1.01 (0.01)	1.10 (0.01)***	1.03 (0.01)***	0.97 (0.01)***
Age-squared††	1.00 (0.01)	1.00 (0.01)*	1.00 (0.01)	1.00 (0.01)	1.00 (0.01)
Income as % of poverty level
0-199 (ref)	1.00	1.00	1.00	1.00	1.00
200-399	0.70 (0.08)**	1.37 (0.18)*	0.92 (0.09)	0.72 (0.08)**	1.45 (0.16)**
≥400	0.85 (0.11)	1.18 (0.17)	0.96 (0.11)	0.89 (0.14)	1.48 (0.24)*
Educational attainment
<high school (ref)	1.00	1.00	1.00	1.00	1.00
High school	0.80 (0.13)	1.44 (0.28)†	0.85 (0.12)	0.58 (0.06)***	1.21 (0.14)†
≥bachelor's degree	0.65 (0.13)*	1.62 (0.36)*	0.54 (0.09)***	0.51 (0.08)***	1.27 (0.20)
Partnership status
Single (ref)	1.00	1.00	1.00	1.00	1.00
Married	0.89 (0.12)	0.68 (0.10)*	0.62 (0.08)***	0.78 (0.80)*	0.68 (0.07)***
Cohabiting	1.04 (0.16)	0.93 (0.16)	0.71 (0.10)*	0.95 (0.13)	0.92 (0.12)
Employed‡	1.05 (0.12)	1.23 (0.16)	0.97 (0.10)	0.88 (0.08)	1.06 (0.10)
Not fecund‡	0.77 (0.11)†	0.95 (0.15)	0.96 (0.13)	0.98 (0.09)	1.58 (0.15)***
No.of children††	na	na	na	1.00 (0.04)	0.95 (0.04)

^*p<.05.

^**p<.01.

^***p<.001.

^†p<.10.

^‡Dichotomous measure.

^§Includes women of other races.

^††Continuous measure.

Notes: Figures in parentheses are robust standard errors. The age measures are centered to reduce multicollinearity. All results are weighted. ref=reference group. na=not applicable.

• Fertility desires and intentions

Childless women with and without disabilities are equally likely to want a child and equally likely to intend to have a child if they want one (Table 3). Among those who want and intend to have a child, the intended number of children is also similar. However, childless women with disabilities are more likely to report uncertainty that they will be able to achieve those intentions: The odds that they are a category indicating being more, rather than less, uncertain about intentions are 1.7 times the odds for childless women without disabilities.

TABLE 3.

Odds ratios and coefficients from regression models assessing associations between selected characteristics and fertility desires and intentions among childless women

Characteristic	Wants a child	Intends to have a child	No. of children intended	Uncertainty about intentions
Disability‡	0.88 (0.21)	0.73 (0.28)	–0.02 (0.05)	1.69 (0.35)*
Race/ethnicity
Non-Hispanic white§ (ref)	1.00	1.00	1.00	1.00
Non-Hispanic black	1.42 (0.30)†	1.83 (0.62)†	–0.01 (0.03)	1.19 (0.19)
Hispanic	2.46 (0.63)***	0.88 (0.31)	0.04 (0.03)	1.02 (0.16)
Age††	0.86 (0.01)***	0.85 (0.02)***	–0.02 (0.01)***	1.12 (0.02)***
Age-squared††	1.00 (0.01)**	1.00 (0.01)	–0.01 (0.01)*	1.00 (0.01)
Income as % of poverty level
0-199 (ref)	1.00	1.00	1.00	1.00
200-399	1.04 (0.19)	1.41 (0.50)	–0.07 (0.03)*	0.89 (0.12)
≥400	1.08 (0.19)	2.31 (0.88)*	–0.10 (0.04)*	0.83 (0.13)
Educational attainment
<high school (ref)	1.00	1.00	1.00	1.00
High school	1.37 (0.35)	1.08 (0.49)	0.11 (0.04)**	0.76 (0.14)
≥bachelor's degree	2.09 (0.62)*	1.53 (0.74)	0.13 (0.06)*	0.49 (0.12)**
Partnership status
Single (ref)	1.00	1.00	1.00	1.00
Married	1.51 (0.32)†	1.05 (0.44)	1.27 (0.07)***	0.60 (0.11)**
Cohabiting	1.16 (0.28)	0.27 (0.09)***	1.13 (0.06)***	0.78 (0.14)
Employed‡	1.03 (0.19)	1.09 (0.34)	0.04 (0.02)	1.09 (0.15)
Not fecund‡	0.88 (0.19)	1.38 (0.52)	–0.06 (0.05)	1.08 (0.20)

^*p<.05.

^**p<.01.

^***p<.001.

^†p<.10.

^‡Dichotomous measure.

^§Includes women of other races.

^††Continuous measure.

Notes: The first two columns show odds ratios from binary logistic regression; the third column shows unstandardized linear regression coefficients; the fourth column shows odds ratios from ordinal logistic regression. Figures in parentheses are robust standard errors. The age measures are centered to reduce multicollinearity. All results are weighted. ref=reference group.

The similarity between childless women with and without disabilities is not observed among mothers (Table 4). Among mothers, those with disabilities were more likely to want another child (odds ratio, 1.5), but less likely to intend to have a child (0.5), than were those without disabilities. Additionally, among mothers who want and intend to have another child, the odds of being in a higher category, rather than a lower category, of uncertainty about intentions are marginally higher for women with disabilities than for others (1.6). These results are particularly of note given that few covariates aside from disability status were significant in these models.

TABLE 4.

Odds ratios and coefficients from regression models assessing associations between selected characteristics and fertility desires and intentions among mothers

Characteristic	Wants a child	Intends to have a child	No. of children intended	Uncertainty about intentions
Disability‡	1.48 (0.22)**	0.46 (0.15)*	0.02 (0.13)	1.62 (0.41)†
Race/ethnicity
Non-Hispanic white§ (ref)	1.00	1.00	1.00	1.00
Non-Hispanic black	1.14 (0.13)	1.85 (0.50)*	–0.06 (0.04)	0.97 (0.17)
Hispanic	1.24 (0.14)†	1.65 (0.44)†	0.01 (0.05)	0.90 (0.16)
Age††	0.91 (0.01)***	0.86 (0.02)***	–0.02 (0.01)**	1.08 (0.02)***
Age-squared††	1.00 (0.01)***	1.00 (0.01)	0.01 (0.01)†	1.00 (0.01)
Income as % of poverty level
0-199 (ref)	1.00	1.00	1.00	1.00
200-399	1.01 (0.11)	1.25 (0.29)	–0.04 (0.06)	1.10 (0.17)
≥400	1.04 (0.15)	1.92 (0.62)*	0.06 (0.07)	0.68 (0.17)
Educational attainment
<high school (ref)	1.00	1.00	1.00	1.00
High school	0.91 (0.11)	1.23 (0.34)	0.08 (0.05)	1.07 (0.19)
≥bachelor's degree	0.95 (0.15)	1.54 (0.53)	0.13 (0.07)*	0.78 (0.20)
Partnership status
Single (ref)	1.00	1.00	1.00	1.00
Married	0.99 (0.10)	1.10 (0.28)	0.39 (0.05)***	0.51 (0.09)***
Cohabiting	1.30 (0.17)*	0.74 (0.22)	0.24 (0.04)***	0.78 (0.15)
Employed‡	0.92 (0.08)	0.64 (0.14)*	–0.03 (0.05)	1.09 (0.16)
Not fecund‡	0.92 (0.08)	1.29 (0.32)	0.12 (0.09)	1.75 (0.35)**
No. of children††	0.71 (0.04)***	0.82 (0.09)†	0.02 (0.04)	1.17 (0.11)

^*p<.05.

^**p<.01.

^***p<.001.

^†p<.10.

^‡Dichotomous measure.

^§Includes women of other races.

^††Continuous measure.

Notes: The first two columns show odds ratios from binary logistic regression; the third column shows unstandardized linear regression coefficients; the fourth column shows odds ratios from ordinal logistic regression. Figures in parentheses are robust standard errors. The age measures are centered to reduce multicollinearity. All results are weighted. ref=reference group.

In supplementary analyses, we assessed whether fecundity mediates the relationship between disability and fertility intentions by repeating the analyses for childless women without including fecundity in our models. If fecundity mediates the relationship for childless women, we would expect disability to be statistically significant in all models. We find no such evidence.^* In sum, the similarity between childless women with and without disabilities in wanting a child, intending to have a child and intended parity cannot be attributed to differences in fecundity.

DISCUSSION

A substantial proportion of women of reproductive age experience disability. Our study about their fertility attitudes reveals two distinct patterns. First, women with disabilities regard the value and importance of motherhood similarly to the way that women without disabilities do. Net of race, age, income, educational attainment, partnership status, employment, fecundity and number of children, motherhood is a salient aspect of the lives of women with disabilities.

Second, mothers’ fertility desires and intentions vary by their disability status. Mothers with disabilities are more likely to want another child than are mothers without disabilities; however, those who want a child are less likely to intend to have a child. Mothers with disabilities experience a mismatch between their fertility desires and intentions that cannot be explained by the social and demographic or fertility characteristics considered in our models. Interestingly, we do not observe this mismatch among childless women with disabilities, who are likely to have lesser knowledge of the challenges of pregnancy and parenting. However, if these women want a child, they are less certain about their intentions than are childless women without disabilities.

Limitations

Although our work takes an important first step toward understanding motherhood attitudes and intentions by disability status from a population perspective, it is not without its limitations. First, the NSFG disability measure is crude; the survey does not include any information that permits distinctions by the type or severity of disability. However, despite their simplicity, even crude disability indicators can capture meaningful distinctions.⁵¹ Next, women with severe cognitive disabilities are not included in the NSFG, because all participants need to be able to participate in the study (proxies are not permitted in the final interview). Also, because the NSFG measure of disability refers to the date of interview and does not indicate disability status at the time mothers in the sample gave birth, it was not appropriate to calculate differences in rates of birth by disability status.

As a general survey of women of reproductive age, the NSFG lacks the data necessary to follow individual women's reproductive histories together with their disability experiences over time. We are unaware of any population-level data source that includes both such detail on reproductive preferences and detail on disability experience among women of reproductive age. It is our hope that the findings reported in this article may lead future studies of persons with disability to include more questions on reproductive health; life-course desires and planning; experiences of stigma and discrimination in regard to childbearing and childrearing; and reproductive health care experiences. Likewise, we hope that future surveys of women's reproductive health (including future cycles of the NSFG) will include a module of questions designed for women with disabilities that would provide detail on disability type and disability trajectories. Because these questions would be asked only of women who indicate a disability in response to the current disability question, adding them would not lengthen the interview for the 90% of women who report no disability. This study also suggests that a careful qualitative research project needs to focus on understanding the tension between fertility desires and intentions among women with disabilities who are already mothers.

Conclusion

The expanded opportunities for persons with disabilities, including legislated supports for enablement, open up social roles that offer those with disabilities a life course more equal to that of those without disabilities. Motherhood is a potential component of this life course. Our analyses indicate that women with disabilities are largely as positive about becoming mothers as are women without disabilities. However, as with many aspects of the life course of persons with disabilities, these desires are not simple to accomplish. Motherhood is a significant, even defining, identity for many women. Societal and personal challenges notwithstanding, women with disabilities want families with children and are actualizing these desires. Although disability need not always do so, it can have implications for pregnancy, birth and parenting. Reproductive health researchers and practitioners are well situated to assure support for all women who desire pregnancy and motherhood. Deepening understanding of the reproductive health needs and challenges of women with disabilities is a facet of such support.