in a 2000 Cochrane review to be associated with a reduction in exclusive breastfeeding but had no significant effect on the earlier termination of nonexclusive breastfeeding.64 Although still relatively common, this practice declined 41% from 2007 to 201365; further contemporary research is needed to evaluate the effects on breastfeeding initiation and continuation. As noted in the Methods section, a comprehensive and relatively recent systematic review of the literature related to maternal obesity and breastfeeding was identified,22 and, thus, maternal obesity was not included in the current metaanalysis. The authors applied rigorous inclusion criteria and provided an assessment of the methodological quality of the studies they included in their review. Their overall conclusion that obese women are less likely to initiate and to continue breastfeeding than women of a healthy weight should be considered alongside the conclusions of the current metaanalysis in developing comprehensive breastfeeding education and support programs. Limitations of our analysis include heterogeneity in measurement and categorization of the factors, particularly maternal education and breastfeeding education. There is additional heterogeneity in the study populations included as well in the timeframes examined for breastfeeding continuation. Not all studies presented RRs or appropriately adjusted THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume 203 194 Cohen et al for confounders nor adjusted results using the same sets of confounders. Another limitation of this study is that these factors were studied in parallel, but in actual clinical practice, they are most certainly related in complex ways that were not considered in this analysis. Furthermore, some of the factors are not dichotomous but may be present in gradations (eg, rooming in, educational interventions). The ability to assess for a doseresponse would have strengthened this analysis; however, most studies do not quantify, for example, the amount of time rooming-in or the intensity of educational interventions. Also, this review did not include community-level factors such as hospital policy initiatives, clinician training, or interventions designed to increase breastfeeding rates. Finally, as noted in the Methods section, we did not include SES as one of our select factors. SES is certainly a potentially important factor in relation to breastfeeding initiation and continuation and is more readily collected in clinical settings (such as by payer identification or WIC status) than some other factors presented in this meta-analysis; however, SES as a construct in the literature published to date is quite heterogeneous, and, thus, SES was not included in this meta-analysis due to these methodologic limitations. In summary, we found scant quality empirical evidence for some factors commonly believed to be important influences of the breastfeeding experience, and there were limitations in methodology for others (such as SES). For the 6 factors included in this review, despite differences in study design, measurement of exposures and outcomes, and underlying demographics of the populations studied, the literature provides relatively consistent indicators as to which factors influence breastfeeding decisions, particularly maternal smoking, maternal educational attainment, and dyad separation. Although some factors are not modifiable, these results may be informative in developing targeted, multifactorial interventions to provide the education and support needed to allow breastfeeding to be successful for more families. ■ This research project about mothers’ early experiences of infant feeding was developed following discussions within the Public Health Children and Young People (PH CYP) team, South Gloucestershire Council about local public health priorities for breastfeeding; current services supporting breastfeeding; and relatively low breastfeeding continuation rates. Consequently, UWE were commissioned to work with the PH CYP team to provide an opportunity for women to voice their experience of infant feeding. This would inform future service planning and commissioning. Locally, quantitative data on breastfeeding initiation and continuation rates is routinely available but there is limited understanding about the reasons why mothers discontinue breastfeeding. We completed this qualitative research with the aim of understanding more about breastfeeding for local mothers, in particular their experience of support, and using the findings to inform and improve local service planning. As a small piece of qualitative research, the findings cannot be generalised within, or beyond, South Gloucestershire, but they are useful in furthering our understanding about how mothers perceive their experience of support for breastfeeding. The research was funded by the Public Health and Wellbeing Division and completed between May-October 2016. The limited budget and tight timescale necessitated the research was planned and completed on a relatively small scale. The findings, nevertheless, provide helpful pointers for future service planning. 2.Background and Context A large body of evidence (outlined in Appendix A) demonstrates that breastfeeding is a major contributor to public health and has an important role to play in reducing health inequalities. It is a matter of concern,