We analyzed data from the Pregnancy Risk Assessment Monitoring System (PRAMS), a population-based, site-specific surveillance system, which collects self-reported data on respondent’s behaviors and experiences before, during, and shortly after pregnancy. Women with a recent live birth were randomly sampled from site birth certificate files and contacted 2 to 6 months after delivery to participate. PRAMS is conducted by the Centers for Disease Control and Prevention (CDC) in collaboration with site health departments. Data are collected using both mail and telephone survey methods. PRAMS does not collect data on gender identity; therefore, the term “women” was used to align with the birth certificate.

The PRAMS survey includes a core set of questions on all site surveys and, to address emerging topics, sites periodically add supplements to their survey.²³ The PRAMS disability supplement²⁴ was made available to sites in 2018 to address the gap in availability of population-based data on disability among women. The number of sites that implemented the disability supplement varied during 2018 to 2020. PRAMS methodology²⁵ and the disability supplement have previously been described.²³

Among 39 673 respondents, 6.0% reported disability (Table A, available as a supplement to the online version of this article at https://www.ajph.org). Difficulty with remembering (3.7%) was most common, followed by difficulty with seeing (1.6%), walking or climbing stairs (0.6%), hearing (0.6%), communicating (0.5%), and self-care (0.3%). A higher prevalence of respondents with disability than without were non-Hispanic American Indian/Alaska Native (1.3% vs 0.8%), aged 20–24 years (26.8% vs 17.8%), had completed less than high school (15.6% vs 10.1%) or had a high school diploma or GED (36.5% vs 24.7%), were multiparous (65.7% vs 59.6%), participated in WIC prenatally (47.8% vs 31.0%), and had infants who were preterm (11.1 % vs 9.0%; Table 1). Prevalence of disability was higher among respondents who were non-Hispanic American Indian/Alaska Native (9.8%) versus those who were non-Hispanic White (5.8%) and non-Hispanic Asian/Pacific Islander (4.0%); were younger than 20 years (8.1%) or aged 20 to 24 years (8.8%) versus aged 25 to 34 years (5.3%) or 35 years or older (5.0%); had completed less than high school (9.0%) or who had a high school diploma or GED (8.7%) versus some college or higher (4.5%); were multiparous (6.6%) versus primiparous (5.2%); had participated in WIC prenatally (9.0%) versus those who had not (4.6%); and whose infants were preterm (7.3%) versus those whose infants were term (5.9%). The prevalence of disability type by select characteristics generally followed a similar pattern as disability overall (Table A).

In unadjusted analyses, respondents with disability had a lower prevalence of breastfeeding initiation (83.1% vs 87.0%) and breastfeeding at 1 (68.7% vs 76.7%), 2 (56.1% vs 67.0%), and 3 (48.3% vs 60.0%) months than respondents without disability (Table 2). By type of disability: breastfeeding initiation ranged from 72.4% to 84.1% among respondents reporting difficulty communicating and remembering or concentrating, respectively; breastfeeding at 1 month ranged from 57.6% to 71.4% among respondents reporting difficulty communicating and hearing, respectively; breastfeeding at 2 months ranged from 51.5% to 61.6% among respondents reporting difficulty communicating and hearing, respectively; and breastfeeding at 3 months ranged from 45.7% to 54.1% among respondents reporting difficulty remembering or concentrating and hearing, respectively (Table B, available as a supplement to the online version of this article at https://www.ajph.org).

In adjusted analyses, breastfeeding initiation (86.7% vs 86.6%; adjusted PR [APR] = 1.00; 95% CI = 0.97, 1.02) and at 1 month (74.2% vs 76.4%; APR = 0.97; 95% CI = 0.94,1.01) did not differ by disability status. However, breastfeeding remained lower among respondents with disability at 2 (62.6% vs 66.6%; APR = 0.94; 95% CI = 0.89, 0.99) and 3 months (54.7% vs 59.6%; APR = 0.92; 95% CI = 0.86, 0.98) than among respondents without disability.

Overall, 92.2% of respondents reported receiving breastfeeding information from health care providers, with respondent’s doctor being most common (78.3%), followed by nurse, midwife, or doula (75.4%), and baby’s doctor (71.0%; data not shown). Precisely 75.5% of respondents received information from breastfeeding or lactation specialists. Overall, 72.3% of respondents received information from both health care providers and breastfeeding or lactation specialists, 19.8% received information from only health care providers, 4.7% did not receive information from either source, and 3.1% received information from only breastfeeding or lactation specialists (Table 3). In adjusted analyses, regardless of disability, respondents who received information from both health care providers and breastfeeding or lactation specialists or received information from only breastfeeding or lactation specialists were more likely to breastfeed across all outcomes than respondents who had not received information from either source.

Respondents with disability were less likely to receive breastfeeding information from health care providers (89.3% vs 92.3%; APR = 0.97; 95% CI = 0.94, 0.99) and as likely to receive information from breastfeeding or lactation specialists (78.1% vs 75.3%; APR = 1.04, 95% CI = 1.00, 1.08) compared with respondents without disability (Table 4). Among respondents without disability only, those who received information from health care providers were more likely to be breastfeeding at 1 and 2 months than respondents who did not receive information from health care providers (Table C, available as a supplement to the online version of this article at https://www.ajph.org). Respondents who received information from breastfeeding or lactation specialists were more likely to be breastfeeding across all outcomes than respondents who did not receive information from breastfeeding or lactation specialists, regardless of disability status.

Breastfeeding initiation and breastfeeding at 1, 2, and 3 months were lower among women with disability than without disability in unadjusted analyses by 4 to 12 percentage points. In adjusted analyses, breastfeeding at 2 and 3 months remained lower among women with disability by 4 to 5 percentage points.

Although the absolute differences in prevalence of breastfeeding outcomes were modest, our findings suggest that women with disability might face additional challenges to breastfeeding than those without disability. Regardless of disability, many barriers to breastfeeding exist, including lack of knowledge, poor family and social support, lack of parental leave and lactation support at work, lactation problems, social norms, embarrassment, and health services.²⁸ In addition, additional challenges accessing breastfeeding support might exist, such as the availability of community support and distance to lactation counselors.²⁹ Limited evidence exists on breastfeeding experiences of women with disability in the United States. Data on breastfeeding experiences are primarily from qualitative studies with small sample sizes (n ≤ 25) among women who are predominantly White, in their thirties, married, with at least some college education. In these studies, women with a range of disabilities reported challenges, including lack of support, limited information on disability-related health considerations, milk supply and latch problems, and challenges communicating lactation difficulties with lactation consultants.^20,21 Furthermore, women with disability might have difficulty accessing health care,¹ which could impact their ability to seek guidance when they encounter breastfeeding difficulties. Women with disability have also reported breastfeeding facilitators, such as peer support, positive support from health care providers, use of a breast pump, and physical support to breastfeed.²⁰ In a recent cohort study, conducted in Canada, women with an intellectual or developmental disability or multiple disabilities were less likely to receive in-hospital breastfeeding assistance than women without disability.³⁰

Overall, we found respondents who received information from health care providers were more likely to initiate breastfeeding and be breastfeeding at 1 and 2 months than those who did not, suggesting that receiving information from providers may facilitate breastfeeding. However, this association was not found among women with disability at 1 and 2 months, suggesting women with disability may not receive the same benefit from information from providers alone. Regardless of disability status, we found respondents who received information from both health care providers and breastfeeding or lactation specialists and those who received information from only breastfeeding or lactation specialists were more likely to breastfeed than those who received information from only health care providers or neither source. However, respondents with disability were less likely to receive breastfeeding information from health care providers compared with those without disability.

Health care providers might need additional resources to support women who choose to breastfeed, regardless of disability.^11,31 Providing referrals to lactation or breastfeeding specialists for patients, when appropriate, might be another opportunity for health care providers to offer breastfeeding support when they are unable to provide the care directly.^14,16 In a survey of obstetrician-gynecologists, pediatricians, and family physicians, medical education related to breastfeeding was found to be inadequate, and several gaps in training were identified, including the need for breastfeeding education to be prioritized throughout training.³¹ Providers might need additional support to follow guidance from professional organizations to support patients in overcoming breastfeeding challenges.^11,14,15 Efforts to improve breastfeeding education among physicians, such as AAPs’ Physician Engagement and Training Focused on Breastfeeding Project, a collaborative effort between multiple medical and health organizations, aim to improve physician training to support patients in reaching their breastfeeding goals.^13,32 Provider resources specific to breastfeeding support for women with disability may be needed. Beyond breastfeeding, qualitative studies have demonstrated women with disability face challenges in receiving reproductive health care that is inclusive of their disability.^20,21 Challenges in seeking and receiving health care for women with disability include inaccessible equipment, lack of transportation and lack of provider experience.^7,8,33 In a survey of obstetrician-gynecologists, providers reported that they are not trained to properly care for the needs of women with disability.³⁴

Receipt of information on breastfeeding from health care providers that is tailored to the needs of women with disability and referrals to breastfeeding or lactation specialists when warranted might be important strategies to improve breastfeeding outcomes.

This analysis is subject to several limitations. Our findings were limited to PRAMS sites that included the disability supplement on their site-specific survey and met CDC’s 50% response rate threshold criteria, limiting the generalizability of our findings. Findings were also subject to recall and social desirability biases. Because we did not disaggregate Asian and Pacific Islander, we may have masked important differences among this population. Due to the small sample size of women with disability, we report the prevalence of each breastfeeding outcome among respondents reporting each type of disability; however, we did not conduct further analyses. Additionally, our definition of disability status is heterogenous and may have masked variability in the associations. While the WG-Short Set is intended to identify persons who report difficulty across 6 functional domains, it might not capture all persons with disability.³⁵ Our estimates of disability might be an underestimate as the prevalence of disability was lower than what other sources have reported among reproductive-aged women (6% vs 12 – 18%).^1–3 This could be due to differences in how disability was measured or defined, surveys were administered, or differences in disability prevalence in the general versus postpartum population. While PRAMS offers multiple modes to complete the survey,²⁵ persons with disability might need additional accommodations that are unavailable. We were unable to assess whether there was a difference in survey responses or reasons for nonparticipation by disability status.

The PRAMS core questionnaire does not measure breastfeeding intentions, reasons women breastfed or did not breastfeed (any or exclusivity) and we did not analyze breastfeeding outcomes beyond 3 months. We were therefore unable to assess whether respondents’ decision to breastfeed or not was influenced by their disability or other reasons impacted their decision. We were also unable to assess the mode of delivery (in-person, phone, or virtual), content, timing, or frequency of the information received. The category of health care providers included nurse, midwife, and doula; however, the support offered by doulas might differ from other providers. There could be variation in the quality of the information received from breastfeeding or lactation specialists as the training required varies widely.³⁶ We were also unable to assess respondents’ access to breastfeeding information. It is possible that respondents who reported they did not receive information did have access to information, but declined or did not seek or utilize the information for various reasons (e.g., past breastfeeding experiences). There are individual reasons (e.g., stronger desire to breastfeed) and structural drivers (e.g., socioeconomic inequities) that could affect a person’s ability to seek out and access additional support if they were experiencing problems. It’s also possible that women with disability might be more likely to be taking medications than women without disability and were advised to pause or stop breastfeeding as a result. However, PRAMS does not have detailed information on medical conditions, including medication use. Few medications are contraindicated while breastfeeding, and in most cases if contraindicated, a safe substitute can be found.¹¹ However, some evidence suggests women are often advised unnecessarily to stop breastfeeding.³⁷

Another limitation of these data is that the latter portion of data were collected duringthe early part of the COVID-19 pandemic when access to health care was interrupted.³⁸ While telehealth consultations were more common during the COVID-19 pandemic,³⁹ we were unable to assess the mode through which women received information on breastfeeding. Lastly, while we were able to adjust for many covariates reported to be associated with breastfeeding^1,2,22 (e.g., race or ethnicity, age), we could not adjust for all factors that could have contributed to differences in breastfeeding outcomes (e.g., breastfeeding intentions).⁴⁰ Despite these limitations, this study provides valuable information on breastfeeding outcomes and sources of breastfeeding information by disability status using population-based data.