In this secondary dataset analysis using a US representative sample, we found that, after adjusting for covariates, all feeding types that incorporate breastfeeding (exclusive breastfeeding, late mixed feeding, and/or early mixed feeding) were higher in states with paid family leave compared with states without paid family leave policies. These findings offer novel insight into various feeding approaches indicative of real-world infant feeding practices among breastfeeding persons in the labor force. Even after controlling for state-level workplace breastfeeding policies, paid family leave policies are associated with exclusive breastfeeding and with a spectrum of feeding types that incorporate breastfeeding. These results demonstrate that the paid family leave policies are positively associated with breastfeeding, and thus may reflect health benefits for the parent/baby dyad that accompany any degree of breastfeeding [2, 4,5,6].

Paid family leave policies have previously been associated with improvements in breastfeeding duration and with health and economic benefits [31,32,33,34,35,36]. After California was the first US state to provide eight weeks of partial paid family leave in 2004, Huang et al. utilized the CDC Infant Feeding Practices Study and found contemporaneous increases in the rates of breastfeeding in California compared with other US states, with difference-in-differences of 15.8%, 17.4%, and 18.4% in the rates of any breastfeeding at three, six, and nine months, respectively [37, 38]. Globally, parental leave after childbirth has been associated with reduced maternal and infant morbidity and mortality [15, 39,40,41,42]. In high-income countries, paid parental leave has been associated with increases in exclusive breastfeeding, downstream earning potential, workforce retention, and infant vaccination rates and with reductions in maternal medical and mental health morbidity [16, 43]. Data suggest higher degrees of benefit with increasingly generous leave, including longer breastfeeding duration and higher maternal pay [44].

Just over half (56%) of the US workforce qualifies for federal FMLA, which is unpaid [45]. There are also limited opportunities for paid leave in the United States, which results in suboptimal breastfeeding initiation and duration [41]. Cross-sectional US studies have found that 59% of women did not receive paid leave, and, even when it was received, paid leave averaged about three weeks, with reduced salary [46]. Current FMLA policies and qualifications tend to support families who can afford unpaid time off work and have stable employment from large employers [18]. Women facing social and structural barriers to breastfeeding, such as low income, lower educational attainment, and membership in minority racial and ethnic groups, disproportionately do not benefit from federal FMLA policies. In this multivariable analysis, we found disparities in breastfeeding was associated with multiple factors that are also related to disparities in medical leave policies.

When evaluating income, we found that, in the adjusted model, women below the poverty level were less likely to exclusively breastfeed for the first six months before introducing infant formula. Families with lower incomes have previously been shown to have less paid and unpaid leave. One 2014 cross-sectional study found that only 20% of families making under $35,000 per year received paid leave, averaging 1.5 weeks in duration, compared with a respective 55% and 4.5 weeks for families making over $75,000 per year [46]. The 2018 US Department of Labor FMLA surveys demonstrated that low-wage workers making $15 per hour or less were least likely to take needed medical leave, citing the inability to afford unpaid time off from work and fear of job loss [45].

Similar to other studies, we found that marital status was positively associated with breastfeeding [47]. The US Department of Labor reports that approximately 95% of fathers with children under three years of age are working, highlighting the potential importance of parental leave for all caregivers in the workforce [11].

Our findings also reflected known racial and ethnic disparities in breastfeeding stemming from complex policies and histories related to structural and ongoing racism [8, 48, 49]. After adjusting for covariates, we found that, compared with respondents who identified their children as non-Hispanic White, non-Hispanic Black respondents were less likely to exclusively breastfeed. Racial and ethnic disparities extend to FMLA eligibility; studies have found that more Black and Hispanic/Latino workforce members (60.2% and 66.9%, respectively) reported being ineligible for or unable to afford unpaid leave than White workers (55.3%) [39]. Because Black women experience higher rates of pregnancy complications and preterm delivery compared to other races, current federal FMLA policies, which count time from pregnancy complications as part of leave, may further exacerbate racial inequities [50,51,52]. We also found that Hispanic/Latino respondents had higher rates of late mixed feeding but lower rates of exclusive breastfeeding compared with non-Hispanic White women. A 2021 study examining feeding goals found that despite Hispanic/Latina women having higher intentions to breastfeeding compared with non-Hispanic/Latina White women, they had lower odds of meeting their goals [53]. Inequitable access to FMLA policies may be contributing to this gap, as Hispanic/Latina women are less likely to qualify for both paid and unpaid leave, which may be related to part-time work status or working for small employers [39, 54, 55].

Employment and education have also previously been associated with breastfeeding outcomes and may be related to access to FMLA [47]. While NIS-C does not include employment data, in the unadjusted and adjusted analyses, we found that women with lower education levels, which are associated with employment opportunities, were less likely to breastfeed. Return to work is among the top reasons for interrupted breastfeeding [33, 56], and women who take six months or more of leave from work have a 30% higher likelihood of any breastfeeding at six months [57].

We also found that women enrolled in WIC were less likely to report breastfeeding across all adjusted and unadjusted analyses. For breastfeeding individuals, WIC distributes breast pumps, offers nutritional support, and supports breastfeeding peer counselor programs [58]. WIC also subsidizes formula purchases, which can result in sales benefits for formula manufactures [59]. Multi-level strategies have been found to enhance breastfeeding for WIC participants, including supporting early WIC enrollment, assessing breastfeeding intentions, and funding peer counseling [60]. However, formula provision may be an incentive for WIC enrollment for some income-eligible individuals, and prior studies have found that some enrollees perceive WIC as a formula provider and appreciate the financial support for formula supplementation [61]. These findings underscore the need to further enrich the lactation-supporting capacity of WIC while considering financial implications and regulations for formula provision.

Limitations of this cross-sectional, secondary dataset analysis included baseline differences between the states with and without paid family leave policies. Respondents in states with paid family leave policies were more likely to report older age, Hispanic/Latina ethnicity, smaller household size, college degree, married status, higher income levels, and lower WIC enrollment. While we adjusted for these characteristics and for workplace breastfeeding policies, there were potentially other unmeasured confounding factors that may have differed between states with and without paid family leave policies, including maternal employment status. Although more generous leave has been found to be associated with greater benefits for breastfeeding-related outcomes, this study did not account for state-by-state variation in leave policies [44]. Several states had policies that went into effect during the survey lookback period, but sensitivity analyses showed similar outcomes regardless of inclusion or exclusion of these states in analyses. Additionally, there were limited data for families who prefer languages other than English or Spanish, and birthing people with varying gender identities may have been excluded.

Despite the limitations of the cross-sectional analysis, it is important to assess differences in breastfeeding outcomes, including mixed breast and formula feeding, as more states enact paid family leave policies. Future prospective studies can evaluate changes in breastfeeding after policy enactment and can assess breastfeeding prevalence in the workforce.