This is the first study to describe prevalence of breastfeeding practices among rural-to-urban migrants and examine the association between them. Based on a large, population-based sampling survey, the findings in this article are highly reliable and meaningful. In the present study, the prevalence of breastfeeding practices was not optimal to meet the goal by 2020. The rural-to-urban migration was only inversely related to EBF among those with high education level, living in metropolis and being minority, after adjusting for other covariates.
Prevalence of breastfeeding practices
The overall prevalence of breastfeeding practices found in this study is too low to meet the goal about 50% EBF rate in 2020. Recent study showed that only 37% of children younger than 6 months of age were exclusively breastfed in low-income and middle-income countries [2], higher than what we found in this study. Chinese used to have a good tradition of breastfeeding and more than 80% mothers breastfed their children [19]. But since 1970s, the prevalence of breastfeeding has dropped dramatically [3, 20]. There is urgent need for promotion methods, like personal sessions, family education and social support, to increase breastfeeding rate in China, especially EBF rate [21-23].
Association between breastfeeding practices and rural-to-urban migration
In the present study, we found that prevalence of the four breastfeeding practices (ever breastfeeding, EBF, predominant breastfeeding and age-appropriate breastfeeding) in rural-to-urban migrants was slightly, but not significantly lower than that of local population. Rural-to-urban migrants are more vulnerable to health-related issues because most health services and policies are registration-related in China, making those migrants inconvenient to have access to them. However, breastfeeding is more complicated than those traditional health issues. It also takes consideration of self-efficacy and social environment [23-25]. Some studies concluded that maternal education was a strong indicator for breastfeeding practices [26]. We found no difference in maternal education between the two groups of participants, which could reduce the difference in breastfeeding practices between them. Baby-friendly hospital practices may also help the establishment of breastfeeding [27], particularly giving only breast milk in the hospital [28]. As the new Medicare Reform involved more registration-inconsistent people, rural-to-urban migrants can have similar, even the same access to hospital services like local citizens, resulting in similar prevalence of breastfeeding practices in the two groups.
After adjusting for other confounding variables, rural-to-urban migrant status was solely, significantly and robustly associated with EBF. This indicates the association between rural-to-urban migrant status and EBF may be modified by other variates. In another study conducted in 2013 in China, they examined the association between maternal migrant status and EBF and found it not significant [4]. However, the definition of maternal migrant status was not described in the study and no further multivariate analysis was conducted between them. We found no other study discussing the association between rural-to-urban migrant status and breastfeeding practices in China.
According to previous studies, maternal education level, place of residence and maternal ethnicity were three main factors that associated with breastfeeding prevalence [29]. Stratified by those three factors, we noticed some special, vulnerable groups which should be paid more attention to about EBF. Migrants with high education level gave less exclusive breastfeeding to their children than local mothers with the same diploma. However, such difference was not found within those with low education level. The association between maternal education level and EBF remains inconsistent. According to the systematic review of Boccolini et al., the low education level was associated with the interruption of EBF in Brazil [29]. However, another systematic review of Zhao J et al. found that in the Chinese culture and employment environment, mother with higher education level were less likely to breastfeed their babies compared to those who were less educated [26]. Among higher educated mothers in our study, rural-to-urban migrant status became a risk factor for EBF and we presumed that it was caused by their working status and living places. Migrants are more likely to rent a house, rather than buying one, without enough space for breastfeeding, especially in suburb areas. Higher educated migrant mothers may have better jobs, equivalent to their diploma, in central areas and hire babysitters to looking after their children. As migrants, they may face more pressure than local citizens and have to work hard to avoid being fired and earn enough money for rental and baby caring. Thus, they may have difficulties balancing work and child care and reducing their time with children [30, 31]. The long distance between their rent house and working place also reduced their time for breastfeeding. Migrants with lower education level may just find jobs near their living place for convenience and most of them may be informal employed, which means they have less traffic time and face less pressure than higher educated ones [32]. The low EBF rate of lower educated participants itself may also lighten the influence of rural-to-urban migration on EBF.
Place of residence is another factor that affects the association between rural-to-urban migration and EBF. Overall, living in metropolis rather than medium sized or small cities can promote EBF. Super cities, like capitals, can provide more health-related services, prenatal lessons and peer education to highlight the importance of EBF [33]. In the subgroup of participants living in metropolis, rural-to-urban migrant children are less likely to be exclusively breastfed than local ones. The expense of living in metropolis is much higher for migrants and they are more common to have mental health problems than natives [34]. When local citizens are surrounded by plenty resources of postpartum and neonatal caring, the migrants may be facing heavy working load or traffic jam, resulting in less time for breastfeeding. High prevalence of postpartum mental health problems in migrants can also deter the execution of EBF [35]. Duration of residence can also affect breastfeeding practices. A study conducted in Hong Kong revealed that breastfeeding duration was progressively shorter when the immigrant time increased [14]. However, the difference of EBF rate between rural-to-urban migrants and local people living in medium sized or small cities is slight but not significant. The scale of the cities is not as big as metropolis and the traffic congestion is not so severe, so these migrants waste less time in commuting and have enough time for breastfeeding. Also, migrants living in medium sized or small cities face less stress of living and working than those in big cities and they are less vulnerable to postpartum mental health problems.
The lower prevalence of EBF in rural-to-urban migrants was solely apparent in ethnic minorities (other 55 ethnicities except for Han). This result is consistent with previous findings. Fenglian Xu et al. conducted a survey in Xinjiang, China 2004 and concluded that EBF rate in the Han was significantly lower than “other minority” (excluding Uygur) [36]. There was only 1 Uygur participant in our study, making our “minority” similar to “other minority” in Xu’s article. Another study in China 2014 also found ethnic Han was associated with decreased likelihood of EBF [37]. Acculturation to the culture of not favoring breastfeeding can reduce the prevalence of EBF [38]. We postulated that migrants were likely to abandon their traditional breastfeeding practice and adopt Han’s disfavor of breastfeeding [37, 39]. Local minority citizens are more likely to live in ethnic communities and maintain their traditional breastfeeding habits [40]. However, it might be hard for minority migrants to fit into such ethnic communities, resulting in loss of their traditional breastfeeding practices.
There are still some limitations in our study. First, we didn’t consider the original registration place for migrant population specifically. Where they are from may largely affect their employment status, then make difference to their practice of breastfeeding. Second, causal effect between rural-to-urban migrant status and EBF was difficult to examine because of the nature of cross-sectional study. Further perspective study is needed to clarify the causal relationship. Third, the prevalence of breastfeeding practices was calculated based on a “24-hour recall method”, which could overestimate the prevalence and cause recall bias.