Socioeconomic status and breastfeeding in China: an analysis of data from a longitudinal nationwide household survey

There has been increasing evidence demonstrating the benefits of breastfeeding for both children and mothers. A recent review concluded that breastfeeding was potentially one of the best interventions for reducing mortality in children younger than 5 years of age [1]. Moreover, breastfeeding demonstrated positive long-term effects on childhood obesity, total cholesterol, non-communicable disease occurrences, and intelligence development [2, 3]. For mothers, it can improve birth spacing and reduce the risk of diabetes, ovarian cancer, and breast cancer [4].

Although the health advantages of breastfeeding are well established, the rates of breastfeeding initiation and continued breastfeeding at 2 years, as recommended by the World Health Organization (WHO), are low in most countries [1]. A national survey from the USA reported that 26% of all women, with children aged from 0 to 5 years, did not breastfeed at all [5]. Similarly, in England, 26.1% of mothers did not initiate breastfeeding, and only one third continued breastfeeding at 6 months [6]. Even in Norway, where 98% of mothers initiated breastfeeding, only 35% continued partial breastfeeding for at least a year [7].

China has experienced dramatic economic growth, social polarization, and demographic changes in the past three decades. Its GDP rose from $191,149 billion in 1980 to $11.065 trillion in 2015 [8]. The economic expansion has lifted more than 700 million people out of poverty [9]. Meanwhile, the most salient feature of China’s demographic transformation has been the extensive internal migration from rural to urban areas. Between 2010 and 2015, the number of internal migrants grew from 121 million to 247 million, of which 169 million moved from rural to urban areas [10]. This rapid social and economic transition in China may affect the practice of breastfeeding [11‐13]. A review demonstrated that breastfeeding rates in China fell during the 1970s, especially in big cities, and reached their lowest point in the 1980s. In the 1990s, the breastfeeding rate began to grow, with 80% of mothers breastfeeding at 4 months since 1993 [13]. In 2010, a large study conducted in Central and Western China reported that 98.3% of infants had been breastfed, however, only 28.7% children under 6 months were exclusively breastfed, and 55.5 and 9.4% had continued breastfeeding for 1 and 2 years, respectively [14]. In 2013, a breastfeeding initiation rate of 84.6% was reported in the 5th National Health Survey [15]. Although the increasing initiation rate was encouraging, the continued breastfeeding rate at 1 year was still lower than most Asian countries (e.g., Japan, North Korea, and Mongolia) [4], and very few women continued breastfeeding at 2 years or beyond, as recommended by WHO.

Hence, exploring the factors which affect breastfeeding initiation and cessation are crucial for improving the health of mothers and children in China. Previous studies have indicated that socioeconomic status (SES) was an important factor in breastfeeding initiation and duration in China [12, 16‐20]. However, this association has not been consistently reported and some studies suggest that mothers with higher educational and occupational statuses were less likely to breastfeed [12, 16‐19], while others argued that there was no relationship between the mother’s education or occupation and breastfeeding [14, 20]. Meanwhile, most studies explored the relationship between the mother’s SES and breastfeeding, and seldom considered the father’s SES, which is regarded as a significant factor in determining the initiation and cessation of breastfeeding [21, 22]. Furthermore, there was limited evidence of the association between SES and breastfeeding behavior in China at a national level.

To bridge the gaps in extant literature, this study aims to explore the relationship between SES and the initiation and duration of breastfeeding in China using a nationally representative dataset from a longitudinal household survey. Information from this study will help identify target groups for future breastfeeding promotion projects.

This study included 2938 children, of whom 2658 (90.5%) were breastfed, which was higher than the percentage reported in China’s 5th National Health Survey (84.6%) [15]. As demonstrated in Table 1, the mean duration of breastfeeding was 8.66 months (SD = 6.15). The mean household income per capita was RMB 11,482 (SD = 28,446). Further, the majority of parents had a low educational and occupational status, lived in rural areas, and were married. Most children were ethnically Han, born at a health facility, and had a normal birth weight. The gender and residential region distribution of children were similar in the study sample.

In Fig. 2, the Kaplan-Meier survival curve of breastfeeding indicates that the probability of breastfeeding dropped sharply after 12 months. Rates of breastfeeding at 6, 12, and 24 months were 79.4, 30.1, and 3.2%, respectively. The duration of breastfeeding among participants ranged from 0 to 39 months.

Table 2 reports the marginal effect of SES on the initiation of breastfeeding, obtained using the logistic regression model. It highlights that SES was not associated with the initiation of breastfeeding.

Table 3 presents the results from the Cox regression analysis of the relationship between SES and the risk of breastfeeding cessation. We found that mothers with ISEI scores of 33–43 and 44–58 were more likely to cease breastfeeding when compared to those with an ISEI score of 16–32/unemployment by 1.15 and 1.28, respectively. An inverted U-shaped relationship between the occupational status of mothers and breastfeeding duration was presented. Mothers with a high school or higher education, compared with those with an education level of middle school and below, were more likely to cease breastfeeding by 1.14, while infants whose fathers’ ISEI scores were 59–90 were more likely to experience breastfeeding cessation (versus those whose fathers had ISEI scores of 16–32/unemployment) by 1.29.

Using data from the CFPS, this study demonstrates that SES does not significantly affect women’s choice to initiate breastfeeding. Infants whose mothers have a high school or higher education and a medium occupational status were more likely to experience breastfeeding cessation, as were infants whose fathers had a high occupational status. Our study makes a unique contribution to the existing literature by using longitudinal data from a nationwide survey of China to explore the relationship between SES and breastfeeding initiation and duration, which would greatly improve the understanding of the relationship between SES and breastfeeding behavior.

The following explanations can be applied to the result of SES not being significantly associated with the initiation of breastfeeding for mothers. First, the Chinese government provides a series of maternity protection schemes for women in both urban and rural areas, possibly reducing the gap between women of different SES levels of breastfeeding awareness. For example, women in both urban and rural areas receive antenatal care service at least five times and postnatal care at least two times, which enriches their knowledge regarding breastfeeding [37]. Second, other factors apart from SES, such as preterm birth, insufficient breast milk, and maternal illness can affect the initiation of breastfeeding, hindering the perceived impact of SES [34, 38, 39].

The results also reveal the relationship between the mother’s SES and the duration of breastfeeding. In general, our findings are consistent with previous studies in developing countries, which report that a higher educational and occupational status of the mother could result in a shorter duration of breastfeeding [17, 40]. However, we only found that infants whose mothers had an ISEI score of 33–58 (e.g., shop, stall, and market salespersons and demonstrators, waitresses, and bartenders) were more likely to experience breastfeeding cessation. This suggests that mothers with a medium occupational status were more likely to stop breastfeeding. A partial explanation for this result may lie in maternity employment causing the cessation of breastfeeding.

Owing to economic growth, social polarization, and demographic changes, in 2013, 346.4 million women in China were employed, of which 28.98% had a high school education or above [41, 42]. However, short maternity leaves and lack of accommodation for mothers to express milk in their workplace are an obstacle in continuing breastfeeding. In China, employed women receive a 98-day paid maternity leave, which may be extended by 15 days under special circumstances such as birth complications [43]. Women must return to work after their 3–3.5 months of maternity leave. However, only 2.6% of the workplaces in China have breastfeeding rooms [13], which may impact the mothers’ decision regarding continuing breastfeeding. Another possible reason may be that inappropriate marketing for milk formula influences the mothers’ decision to continue with breastfeeding. While research reports negative health consequences associated with formula use, many women may believe that infant formula is better and more convenient than breastfeeding due to advertising [44]. Therefore, women, especially those with a medium SES [45], are more likely to choose milk formula after returning to work. Mothers with a high occupational status may have better working conditions and better knowledge regarding milk formulas, which may result in continued breastfeeding.

Overall, our findings indicate the importance of the father’s role in breastfeeding duration, which is consistent with previous findings that fathers greatly influence mothers in prolonging breastfeeding [22, 46, 47]. We found that infants of fathers with high occupational status (ISEI score of 59–90, e.g., directors, chief executives, and engineers) experienced shorter breastfeeding duration. Lack of paid paternity leave and emotional support for their partner may be the reasons for this. In China, there was no paid paternity leave for fathers before 2017 [48]. Additionally, fathers with high occupational status may be busier and seldom provide emotional support for their partner. Thus, the father’s role in breastfeeding should be enhanced.

The limitation of this study is that some potential confounding factors could not be controlled for in analyses due to the available data. Future research should consider including more factors, e.g., delivery mode (cesarean section or vaginal delivery), maternal and infant illness, and grandmother’s attitudes towards and prior experience with breastfeeding.

The breastfeeding initiation rate in our study was higher than that reported in the 5th National Health Survey; however it demonstrates a sharp decline in continued breastfeeding at 12 months. Further, the results suggest that SES does not significantly impact women’s choice to initiate breastfeeding. Infants whose mothers have a high school or higher education and a medium occupational status (e.g., shop, stall, and market salespersons and demonstrators, waitresses, and bartenders) were more likely to experience breastfeeding cessation, as were infants whose fathers had a high occupational status (e.g., directors, chief executives, and engineers). Efforts to promote breastfeeding practices should be conducted comprehensively to target mothers with a high school or higher education and a medium occupational status and fathers with high occupational status. Moreover, breastfeeding accommodation at work should be provided, while the milk formula market should be regulated.