Our combined quantitative and qualitative approach enabled us to both quantify breastfeeding practices among young adolescent mothers, and to explore in depth the factors that promote and inhibit breastfeeding. To our knowledge, no previous published reports using population-based data have described breastfeeding practices specifically among adolescents (ages 17 and younger). We found that the proportion of teens in North Carolina that was motivated to try breastfeeding (52%) was quite similar to national data for teens 19 and younger in 2004 to 2008 participating in the National Immunization Survey (53.0%: Confidence Interval (CI) 46.1, 59.9) . However, the duration of breastfeeding was much shorter: Only 28% of teens in North Carolina continued to breastfeed four weeks after delivery, whereas among teens 19 and younger nationally, 19.3% (CI: 13.8, 24.8) were breastfeeding six months after delivery . Furthermore, rates of exclusive breastfeeding were much lower among teen mothers in North Carolina. Less than 17% reported exclusively breastfeeding through four weeks postpartum, whereas nationally the same proportion (17%, CI: 10.3, 23.3) of mothers 19 and younger reported exclusive breastfeeding through three months .
We also observed distinct patterns across racial-ethnic groups. Hispanic teens had the highest rates of breastfeeding initiation and longest duration, followed by White and African American adolescents. Among adult mothers nationally, racial-ethnic patterns are similar, but differences are less pronounced . The exceptionally high breastfeeding rates among Hispanic teens in our study may relate to the fact that nearly 60% of Hispanics in North Carolina are immigrants , and less-acculturated immigrants have higher breastfeeding initiation rates and breastfeed for longer durations than U.S.-born women [15, 16].
The observed differences by race/ethnicity in the quantitative data reflect the social and cultural norms on breastfeeding that emerged in the qualitative study. The impact of teens' social of contexts was most evident in the decision to initiate breastfeeding. In our qualitative sample, more Hispanics than either Whites or Blacks had a family history of breastfeeding and reported the most encouragement from family. African American teens were less able to articulate exact reasons for not breastfeeding, and their discourses reflected family norms in which formula feeding was a more accepted form of infant feeding . Moreover, negative stories of breastfeeding experiences were pervasive among White and Black participants' peer groups. Despite previous studies pointing to the role the baby's father plays in the decision to breastfeed and providing support to breastfeeding teens [5, 8], adolescents in our study did not report receiving information or support from partners.
Given that family support was very influential in teens' decisions to breastfeed in this study and in previous research, measures to increase the involvement of teens' mothers in breastfeeding counseling and education may enhance teens' motivation to breastfeed. Peer counseling by breastfeeding adolescents to share positive experiences and to acknowledge difficulties in advance so that teens are prepared for the commitment that breastfeeding takes--possibly through online groups or social networking sites--may also be a promising strategy warranting more research.
Although the perceived health benefits of breastfeeding and family encouragement motivated many teens to try breastfeeding, they were not necessarily enough to keep them motivated when they encountered difficulty. In both the quantitative and qualitative samples, high rates of discontinuation occurred within the first month postpartum. Reasons for cessation in the first four weeks were largely due to lactation difficulties including sore nipples, perceiving that they were not producing enough milk or that the baby was not satisfied, and trouble getting the baby to latch. These reasons are similar to previously published reports of PRAMS data among women of all ages  and data from the Infant Feeding Practices Study II among adult mothers . This suggests that supplementary professional support in the early postpartum period to address these technical challenges, which has been successful in increasing breastfeeding duration among adult women , may be warranted for teens.
However, few participants in our qualitative study reported receiving professional support for breastfeeding after hospital discharge and none received any hands-on technical support. Strategies that may be helpful include more hands-on tailored support through the early postpartum period, both in the hospital and at home to provide assistance on specific problems that adolescents encounter. These include ways to avoid and treat nipple pain, help the baby to latch properly, stimulate and maintain the milk supply, and use the breast pump. Counseling that emphasizes that breastfeeding is a learning process  and provides adequate information about the frequency and duration of breastfeeding and how to determine whether their infant is getting enough milk may also be beneficial. This concrete help, along with reassurance that the problems are commonplace and that assistance is frequently needed to overcome the challenges among mothers of all ages, may help build teens' confidence and help prevent early cessation.
School was another major barrier to breastfeeding identified by teens in both the quantitative and qualitative studies. The prevailing view was that breastfeeding was not compatible with school. In contrast to our findings, a recent study documented that with adequate support from school and the provision of onsite daycare, teens successfully continued breastfeeding upon return to school . A large portion of teens in our qualitative study received homebound education services from their schools during the first weeks postpartum. This period may provide an opportunity to communicate to teens that breastfeeding for a short period of time is better than not breastfeeding at all, and to help teens prepare for the transition back to school. Once teens return to school, flexibility in their schedule, a private place for expressing milk that is not a restroom, the use of breast pads to cover up leaking, and storage for expressed milk and the breast pump could help teens to succeed in both academics and breastfeeding. As part of the Affordable Care Act enacted in 2010 in the United States, the Fair Labor Standards Act was recently amended to require employers to provide break time and a private place to express milk while at work . Advocacy efforts are needed for teen mothers to receive similar support in schools. Schools could also support breastfeeding among teen mothers by integrating breastfeeding education into school health programs .
Future research should investigate what is the best way and who is best positioned to provide concrete help to breastfeeding teens within the first week postpartum . Since early breastfeeding experiences affect new mothers' confidence in their ability to breastfeed and their motivation to continue , it is important that teens receive support within the first two or three days following hospital discharge. Given previous research, and our findings that teens may not feel as comfortable to ask for help from health care providers [7, 8], further research should also explore whether teens might benefit from being offered extra support regardless of whether or not they are experiencing difficulty. This might consist of a home visit, or an outpatient visit as a second option, that is scheduled prior to hospital discharge . Home visits could be provided by a lactation consultant, certified lactation counselor, or staff from an adolescent parenting program trained in breastfeeding.
Several limitations should be considered when interpreting the findings of this study. The sample size for the quantitative data is relatively small, particularly for sub-groups such as for Hispanic teens, so results should be interpreted with some caution. Although survey data are weighted for nonresponse to reduce bias, maternal age less than 20 years is a predictor of nonresponse in the North Carolina PRAMS, as well as African American or Hispanic background . Given this potential bias, teens participating in PRAMS may be more highly selected and may be more motivated to breastfeed than those who did not respond. Thus our prevalence estimates may be higher than in the general population. In both the quantitative and qualitative study components, respondents provided information about infant feeding experiences retrospectively which may be subject to recall bias. The majority of participants in both the quantitative and qualitative studies either did not breastfeed or stopped within the first month postpartum; information on infant feeding was obtained from PRAMS participants between two and seven months postpartum and from qualitative participants between one and 18 months postpartum.
In our qualitative study, we interviewed a non-representative sample of teen mothers. By recruiting mothers through organizations serving adolescents, we may have missed the most marginalized teens. However, our multifaceted recruitment approach enabled us to interview adolescents who were both in and out of school, from rural and urban areas in North Carolina, with varying levels of programmatic and clinical support, and may be particularly relevant for programs and service providers in understanding the needs of teens they serve.