The Interaction and Impact of Social Support and Father Absence on Breastfeeding

Abstract

Background:

Behaviors related to early childhood nutrition are influenced by a mother's social environment. In many low- and middle-income countries, breastfeeding rates have steadily declined. At the same time, many communities have a history of domestic or international migration that affects the family support systems for women and children remaining in these communities. While social support has been shown to be important to health behaviors conducive to maternal and child health, scant research examines whether social support moderates the impact of an absent father on breastfeeding.

Objective:

We aim to assess the relationship between father absence and breastfeeding duration and test whether social support moderates the impact of father absence on breastfeeding duration.

Methods:

We use data from the Social Networks and Health Information Survey (n = 292), a random household survey conducted in a municipality in Guanajuato, Mexico, to estimate Poisson regression models of breastfeeding duration.

Results:

In multivariate models, an absent father is negatively associated with breastfeeding, whereas social support is positively associated. A significant and positive interaction between father absence and social support suggests that at high levels of support, breastfeeding duration for women with absent fathers does not appear to be meaningfully different from women with present fathers. This suggests that receiving high levels of social support during pregnancy may mitigate the absence of the child's father.

Conclusions:

Social support interventions for mothers of infants should target mothers and children in households without a father. More research should also be directed at understanding how social support processes during pregnancy can affect breastfeeding in other low- and middle-income countries with high rates of emigration.

Introduction

In setting forth recommendations to exclusively breastfeed an infant for the first 6 months (as the minimum), the World Health Organization (WHO) emphasizes breastfeeding for the health of infants and mothers.¹ Globally, almost two of three infants are not exclusively breastfed.¹ Although high-income countries tend to have a shorter duration of breastfeeding than low- or middle-income countries,² breastfeeding duration represents a significant issue in lower income countries. Only 37% of infants younger than 6 months old are exclusively breastfed in low- and middle-income countries.² This is a serious concern as lower rates of breastfeeding are associated with higher risk of infant and child mortality, especially owing to infectious disease.³

On appearance, it may seem that breastfeeding involves the mother and infant. However, breastfeeding requires education, training, encouragement, and support from family members, health care providers, skilled counselors, employers, and policymakers among others.^4,5 Fathers of the newborn are increasingly viewed as essential to breastfeeding^6,7 and often represent the target population for interventions to improve breastfeeding across global contexts.^8,9 Support from the father is associated with both the decision to breastfeed and breastfeeding duration.^10,11 In general, being married is associated with breastfeeding.^10,12 Having a partner present in the household, particularly a married partner, contributes positively to breastfeeding because this person is an important source of social support conducive to breastfeeding.¹³

Additional family members, especially those residing in a woman's household, represent an important resource because these coresidents are regular sources of social support within a woman's daily life.^14,15 Social support bolsters women's breastfeeding efficacy and helps them overcome barriers to breastfeeding in their immediate environment.¹⁶ In general, higher levels of social support are positively tied to maternal and child health outcomes, including breastfeeding.^16,17 Some research suggests that social support benefits women who are socioeconomically disadvantaged more than other groups.¹⁷

Social support may be vital in contexts in which a large number of households experience the absence of a household member particularly through domestic or international migration. Emigration to high-income countries has become increasingly common for citizens of low- and middle-income countries.¹⁸ The economies of many lower- and middle-income countries depend on remittances (i.e., money) sent home from international migrants with roughly $466 billion received in 2017, an amount that is more than three times the size of official development assistance.¹⁹ Emigration to wealthier countries results in a separation of families across borders, which impacts the social support networks in these communities.²⁰

In Mexico, our specific country of focus, nearly 1 in 6 children live in a household with a parent absent owing to domestic or international migration;²¹ and 1 in 11 children will likely experience father absence because of migration before age 15.²² The presence of fathers in the household is indeed conducive to child health²³ and in particular to breastfeeding where they are seen as a vital source of social support.¹¹ It is possible that the social support from others may mitigate the absence of a father from the household. In other words, the association between the absence of the child's father and breastfeeding duration may vary according to the amount of social support the mother receives.

In this study, we examine the relationship between father absence (i.e., hereafter used to refer to nonresidence with the mother) and breastfeeding in Mexico. Mexico has one of the lowest rates of breastfeeding in Latin America²⁴ and scores poorly in its readiness to promote and support breastfeeding country-wide.²⁵ When looking at breastfeeding trends, Mexico is similar to many other middle-income countries with histories of emigration, such as Nicaragua, Morocco, and the Philippines (∼30% exclusive breastfeeding at 6 months).²⁶ In this study, we assess how the social support women reported receiving during pregnancy benefits women with absent fathers compared with present fathers. We examine this relationship using unique data for women in a medium-sized community in Guanajuato, Mexico. Following calls for social support measures to account for larger sociocultural contexts that shape norms regarding supportive exchanges, social influence, and access to resources,^27–29 we use locally defined measures of social support based on extensive fieldwork in this Mexican community.

Materials and Methods

The data for this study are from the Social Networks and Health Information Survey (SNHIS), a random household survey of mothers conducted in a migrant-sending community in the state of Guanajuato, Mexico. Guanajuato is a traditional sending state for U.S. migrants,³⁰ with the largest share of international migrants of all Mexican states.³¹ In the surveyed municipality of roughly 80,000 residents, >16% of all households reported having an international migrant in 2010.³² The demographics of the selected municipality are comparable with the entire country in terms of age, gender, and civil status based on 2015 Mexican Census data³³ (details available from authors by request).

The SNHIS sample was limited to urban residences in the municipality and proportional to the number of households in a given census block (manzana). Eight households were randomly selected from each block to be a part of the sample. Inclusion was limited to women with at least one child at age 10 years or younger residing in the household. In the case where there were two women with young children in a household, the woman with the most recent birthday was interviewed.

Of the households with eligible respondents, 87% participated in the survey. Respondents received 50 pesos (or roughly $3.50) for participating in the survey. The interviews were conducted in Spanish by female community members and collected information on a woman's social relationships, physical and emotional health, social support, and the health of her children. All questions were translated by a local college-educated educator from the area and validated by other college-educated individuals with Mexican-origin Spanish as their native language. This study was approved by the first author's university institutional review board. All participants gave their consent for their data to be used in the research.

The SNHIS was designed to assess how health information and social support flow within social networks and inform the health of women and their children. As part of this survey, women provided information about their most recent pregnancy resulting in a live birth and early childhood nutrition for this child. The dependent variable, breastfeeding duration, is based on the month a woman reported stopping breastfeeding her youngest child. Women who exclusively formula fed their child (n = 67) and thus did not breastfeed were reported as having 0 months of breastfeeding.

A key focus of this article was on sources of social support in the household during pregnancy. An indicator of father absence was created to capture physical separation from the fathers of the infant (i.e., nonresidence with the mother). Father absence is based on the reported absence of the father of the child at the time of birth. Of this group (n = 54), roughly half were elsewhere in Mexico (n = 30) and half were in the United States (n = 24).

The presence of adult household members has been shown to be important to child health in Mexico,^15,23 whereas studies on breastfeeding have found that support from women may be particularly important.³⁴ For this reason, we have variables for the number of adult females (excluding the mother) and the number of adult males (excluding the spouse) in the household. These variables are based on the reported presence of various types of family members according to gender in the household (mother, mother-in-law, sister, father, brother, etc.). The presence of each relationship type was summed according to gender to generate these two variables.

Women indicated whether they received specific forms of social support during this pregnancy from their family members or friends. These types of support were based on a combination of social support measures adapted from existing research³⁵ and field work in this community that revealed the salience of various forms of support not typically found on social support scales (e.g., advice on what to eat and encouraged me to rest). Nine types of social support in total were listed, and women indicated whether they received each form of support. These included (1) provided emotional support/encouragement, (2) encouraged me to rest, (3) provided advice on how to take care of self/baby, (4) provided advice on what to eat, (5) provided economic assistance, (6) cooked for me or my family, (7) helped with family responsibilities, (8) gave me a ride or accompanied me to doctor's appointments, and (9) helped with things I could not physically do. The nine support items (0 or 1) were summed to create the social support index with a range of 0–9 (alpha = 0.89).

We included demographic controls for the woman's age and education. Age represents the mother's age during the pregnancy of her youngest child. Education is a categorical variable (primary, secondary, preparatory or college education) and the main indicator of socioeconomic status given that household income data were not collected. Pregnancy-related variables in the analysis that have been shown to be related to breastfeeding include male child, cesarean delivery, low birth weight infant (<2,500 g), and parity.

Cases with either missing data (n = 21) or where a mother was still breastfeeding a child (n = 26) were excluded from the analysis through listwise deletion. We used Poisson regression models to analyze the relationship between father absence and breastfeeding duration in months. We interact social support received during pregnancy and father absence to assess how women with absent fathers benefit from social support compared with women with present fathers. All analyses were performed unweighted in Stata 15.

Results

We begin by presenting the descriptive characteristics of the sample and separately by the father absence indicator given in Table 1. Father absence characterizes about one in five of the births in our sample, similar to trends in Mexico and Latin America.²¹ Overall, the differences between the father present and father absent birth samples were modest, and we specify where statistically different. About one-half of the births were delivered through Cesarean section and 10% were low birth weight. The average age of the women in the sample was ∼31 years. Most women had either a primary or secondary level of education (40% in each category); yet, roughly 19% of women with fathers present and only 9% of women with a father absent completed preparatory schooling or more. Households, on average, comprised at least 1 additional female adult and 0.5 adult males. Women reported receiving moderate levels of social support, averaging ∼5.5 (range 0–9). More than three-quarters of women reported breastfeeding after their most recent birth; the average breastfeeding duration was 7 months. This is slightly higher than the median breastfeeding duration of 5.0 for the central region of Mexico³⁶ but similar to other high migration contexts, such as the Phillipines.³⁷

Table 1.

Descriptive Statistics for Most Recent Birth in Social Networks and Health Information Survey (N = 292)

	Entire sample (N = 292)		Father present (n = 238)		Father absent (n = 54)
	Mean/%	SD	Mean/%	SD	Mean/%	SD
Birth characteristics
Cesarean delivery	54.11%		55.46%		48.14%
Low birth weight	10.62%		10.90%		9.26%
Male child	50.34%		48.73%		57.41%
Parity	1.96	0.66	1.62	1.46	1.44	1.46
Mother's sociodemographic characteristics
Age	31.73	7.59	32.05	7.46	30.33	8.06
Education
Primary or less	41.10%		39.92%		46.30%
Secondary	41.78%		41.18%		44.44%
Preparatory or more	17.12%		18.91%		9.26%^a
Household characteristics
No. of adult females (not including self)	1.08	1.44	0.97	1.42	1.53	1.48
No. of adult males (excluding spouse)	0.50	0.68	0.43	0.64	0.79	0.76
Father absence	18.49%
Social support received (range: 0–9)	5.47	3.25	5.36	3.25	5.91	3.27
Decided to breastfeed	78.77%		78.99%		77.78%
Breastfeeding duration	7.30	7.32	7.39	7.51	6.91	6.42

t-tests revealed significant differences according to father presence/absence at p < 0.10 level.

SD, standard deviation.

Next, in Table 2 we present stepwise Poisson models. In the baseline model, Cesarean delivery, maternal age, and having a secondary level of education (relative to primary) are significant and negative, whereas a male birth and parity are significant and positively associated with breastfeeding. The patterns for Cesarean delivery, parity, and secondary education persist after all controls are included, whereas male child and maternal age become insignificant. Father absence (relative to presence), included in the second model, shows a negative association with breastfeeding. This suggests that the difference in the logs of expected count of breastfeeding is ∼0.11 months lower for births with an absent father relative to present father. This pattern holds in the third model with the inclusion of social support. Social support, in contrast, is positively associated with breastfeeding. The receipt of each additional social support item is associated with a 0.03 increase in the log of expected counts for breastfeeding duration.

Table 2.

Poisson Regression Models on Breastfeeding Duration for Social Networks and Health Information Survey Sample Reporting Estimates (SE) (n = 292)

Cesarean delivery	−0.346^*** (0.0450)	−0.351^*** (0.0450)	−0.377^*** (0.0453)	−0.370^*** (0.0455)
Low birth weight	−0.110 (0.0782)	−0.0965 (0.0785)	−0.0635 (0.0785)	−0.0773 (0.0789)
Male child	0.0906^** (0.0441)	0.0905^** (0.0443)	0.0686 (0.0446)	0.0727 (0.0446)
Parity	0.105^*** (0.0182)	0.109^*** (0.0183)	0.121^*** (0.0182)	0.123^*** (0.0183)
Maternal age	−0.00814^** (0.00367)	−0.00710^* (0.00371)	−0.00583 (0.00368)	−0.00599 (0.00367)
Education (primary excluded)
Secondary	−0.105^** (0.0491)	−0.115^** (0.0493)	−0.0934^* (0.0497)	−0.0954^* (0.0498)
Preparatory or higher	0.0602 (0.0673)	0.0394 (0.0681)	0.0321 (0.0683)	0.0461 (0.0689)
Household characteristics
No. of household females		0.0168 (0.0178)	0.0184 (0.0179)	0.0150 (0.0181)
No. of household males		0.0693^* (0.0384)	0.0484 (0.0385)	0.0485 (0.0385)
Father absence		−0.112^* (0.0590)	−0.123^** (0.0592)	−0.330^** (0.134)
Social support received			0.0385^*** (0.00722)	0.0329^*** (0.00788)
Father absence × social support received				0.0340^* (0.0194)
Constant	2.239^*** (0.116)	2.174^*** (0.123)	1.923^*** (0.131)	1.953^*** (0.132)

***

p < 0.01, ^**p < 0.05, ^*p < 0.1.

Model 4 includes the interaction between father absence and social support, revealing a positive and significant coefficient. To aid in the interpretation of the interaction in the multivariate models, we used the margins command in Stata 15 and present a figure representing the predicted duration of breastfeeding by father absence and level of social support. Figure 1 is a plot of social support and breastfeeding by father absence type. There are interesting patterns between and within groups. Generally, births to women in households with a father absent compared with present have lower predicted breastfeeding durations regardless of social support, although at the highest levels of social support, predicted breastfeeding durations are not significantly different for father absent versus present households. In father absent households, those with the lowest levels of support versus highest have significantly shorter breastfeeding durations. Similarly, in father present households, breastfeeding durations are significantly shorter at the lowest levels of social support relative to the highest level. Finally, to reach the goal of breastfeeding for at least 6 months, those with fathers absent require at least a level of social support around four, whereas those with fathers present achieve this without receiving any social support.

FIG. 1.

Marginal effects of father absence on breastfeeding duration by social support received with 95% confidence intervals.

Discussion

Social support is critical to breastfeeding and may be impacted by family separation resulting from increasing international migration from lower- and middle-income countries to high-income countries. We assessed whether social support was associated with the duration of breastfeeding more for women with absent fathers versus those with present fathers in Mexico, a country with an established history of domestic and international migration. We found that father absence was negatively associated with breastfeeding duration, which corroborates other research in Mexico³⁸ and Latin America.³⁹ This indicates that the absence of this person in the household detracts from the time and resources a woman needs to engage in breastfeeding. Social support was positively associated with breastfeeding duration, indicating that expressions of support during pregnancy contribute to the number of months a woman breastfeeds.

The interaction between father absence and social support was positive and significant. At high levels of support, breastfeeding duration for women with absent fathers does not appear to be meaningfully different from those of women with present fathers. This suggests that receiving high levels of social support during pregnancy may mitigate the absence of the child's father when it comes to breastfeeding duration.

While our study may be one of the first to look at father absence and the role of support in breastfeeding, fathers from low- and middle-income countries—where breastfeeding has declined—are migrating all over the world. Research from contexts ranging from Peru to India to Tonga has demonstrated the negative impact of father absence on child health and nutrition.^40,41 In Mexico, the negative impact of parental out-migration on a child's growth was more pronounced when the migration occurred at earlier ages, highlighting how young children are more vulnerable to family disruptions during a time when they have a greater need for intensive feeding.⁴² Our research highlights how high levels of social support can offset the loss of a key social support provider, the father, for mothers during a key developmental period.

Although this study provides insights into social support and breastfeeding, there are limitations. Father absence was a key variable in this study, but other aspects of the relationship, such as relationship quality or communication with the father, are missing from this dataset. Because of power issues associated with the small sample of fathers absent because of migration to the United States and migration elsewhere in Mexico, we grouped all absent fathers. In addition, the social support measure captured social support received during pregnancy, not the postpartum period. However, it is likely that the social environment during pregnancy is consistent with that of the postpartum period and represented consistent relationship dynamics. The father absence measure was constructed at the time of birth. We assumed that he was absent during the pregnancy and postpartum, something that is supported by other research,⁴³ but have no way of assessing potential periods of presence given the structure of the data. The social support indicator measures the receipt of various forms of social support, but not the content of the support or whether this support was desired by the respondent. Given the positive association between social support and breastfeeding duration, we assume that support received contributed to a positive environment surrounding breastfeeding. In addition, small sample size did not allow for us to examine exclusive breastfeeding, but rather any breastfeeding.

This study highlights the salience of social support, particularly for women with absent fathers. Yet, interventions oftentimes focus on father support as a means to promote breastfeeding duration.^8,9 Although these interventions may be effective for women with present fathers, they do not account for women with absent fathers. We argue that the latter group is an important group to focus on breastfeeding interventions as the absence of the father negatively impacts breastfeeding duration. In China, a family-centered intervention that included extended family members was found to be effective at promoting breastfeeding.⁴⁴ Similar such interventions could be developed to include important family members from a woman's social network, especially for women with absent fathers. Finally, interventions utilizing lay community health workers (LCHWs) represent another means to promote breastfeeding given their established effectiveness in low- and middle-income countries.⁴⁵ There is evidence that the social support provided by LCHWs contributes to maternal health as it is an additional source of support outside a woman's network with no expectations of reciprocity.⁴⁶ Interventions that contribute to a woman's social support represent an effective means to promote breastfeeding and should be considered in settings with high rates of international and internal migration.

Conclusions

Breastfeeding is a behavior that is influenced by a woman's social network, both those who are physically present as well as those who are absent. Our research suggests that high levels of support from a woman's network may offset the impact of a key social support provider, the father, for breastfeeding duration. It also highlights that members of women's social networks, aside from the father, should be considered when looking at social support and breastfeeding. Future research should be conducted in other low- and middle-income countries impacted by migration to examine how father absence and social support impact breastfeeding.

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

The authors gratefully acknowledge support provided by the Carolina Population Center as well as the Sociology Department at Rice University.

References

World Health Organization. Infant and young child feeding fact sheet. 2018. Available at https://www.who.int/en/news-room/fact-sheets/detail/infant-and-young-child-feeding (accessed January 20, 2020).

Victora

, Bahl

, Barros

, et al. Breastfeeding in the 21st century: Epidemiology, mechanisms, and lifelong effect. Lancet, 2016; 387:475–490.

Sankar

, Sinha

, Chowdhury

, et al. Optimal breastfeeding practices and infant and child mortality: A systematic review and meta-analysis. Acta Paediatr, 2015; 104:3–13.

Brown

Breastfeeding as a public health responsibility: A review of the evidence. J Hum Nutr Diet, 2017; 30:759–770.

Buccini

, Harding

, Ferré Eguiluz

, et al. An analysis of stakeholder networks to support the breastfeeding scale-up environment in Mexico. J Nutr Sci, 2020; 9:e10.

Carlin

, Mathews

, Oden

, et al. The influence of social networks and norms on breastfeeding in African American and Caucasian mothers: A qualitative study. Breastfeed Med, 2019; 14:640–647.

Wallenborn

, Wheeler

, Lu

, et al. Importance of familial opinions on breastfeeding practices: Differences between father, mother, and mother-in-law. Breastfeed Med, 2019; 14:560–567.

Mahesh

PKB

, Gunathunga

, Arnold

, et al. Effectiveness of targeting fathers for breastfeeding promotion: Systematic review and meta-analysis. BMC Public Health, 2018; 18:1140.

Ajike

, Ogunsanmi

, Chinenye-Julius

, et al. Effect of a breastfeeding educational programme on fathers' intention to support exclusive breastfeeding: A quasi-experimental study. Afr J Reprod Health, 2020; 24:59.

10.

Gibson-Davis

, Brooks-Gunn

. The association of couples' relationship status and quality with breastfeeding initiation. J Marriage Fam, 2007; 69:1107–1117.

11.

Hunter

Breastfeeding initiation and duration in first-time mothers: Exploring the impact of father involvement in the early post-partum period. Health Promot Perspect, 2014; 4:132.

12.

Thulier

, Mercer

. Variables associated with breastfeeding duration. J Obst Gynecol Neonatal Nurs, 2009; 38:259–268.

13.

Maycock

, Binns

, Dhaliwal

, et al. Education and support for fathers improves breastfeeding rates: A randomized controlled trial. J Hum Lact, 2013; 29:484–490.

14.

Clifford

, McIntyre

. Who supports breastfeeding?. Breastfeed Rev, 2008; 16:9.

15.

Donato

, Duncan

. Migration, social networks, and child health in Mexican families. J Marriage Fam, 2011; 73:713–728.

16.

Raj

, Plichta

. The role of social support in breastfeeding promotion: A literature review. J Hum Lact, 1998; 14:41–45.

17.

Laugen

, Islam

, Janssen

. Social support and exclusive breast feeding among Canadian women. Paediatr Perinat Epidemiol, 2016; 30:430–438.

18.

Connor

, Cohn

, Gonzlez-Barrera

Changing Patterns of Global Migration and Remittances. Washington, D.C.: Pew Research Center, 2013.

19.

Vidal

, Tjaden

. Global Migration Indicators. Berlin, Germany: International Organization for Migration, 2018.

20.

Dreby

Divided by Borders. Berkeley: University of California Press, 2010.

21.

DeWaard

, Nobles

, Donato

. Migration and parental absence: A comparative assessment of transnational families in Latin America. Popul Space Place, 2018; 24:e2166.

22.

Nobles

Migration and father absence: Shifting family structure in Mexico. Demography, 2013; 50:1303–1314.

23.

Edelblute

, Altman

. Father absence, social networks, and maternal ratings of child health: Evidence from the 2013 Social Networks and Health Information Survey in Mexico. Matern Child Health J, 2018; 22:626–634.

24.

González de Cosío

, Escobar-Zaragoza

, González-Castell

, et al. Breastfeeding in Mexico was stable, on average, but deteriorated among the poor, whereas complementary feeding improved: Results from the 1999 to 2006 National Health and Nutrition Surveys. J Nutr, 2013; 143:664–671.

25.

González de Cosío

, Ferré

, Mazariegos

, et al. Scaling up breastfeeding programs in Mexico: Lessons learned from the becoming breastfeeding friendly initiative. Curr Dev Nutr, 2018; 2:nzy018.

26.

The World Bank. Exclusive breastfeeding (% of children under 6 months). 20108. Available at https://data.worldbank.org/indicator/SH.STA.BFED.ZS. (accessed January 12, 2020).

27.

Almeida

, Mulready-Ward

, Bettegowda

, et al. Racial/ethnic and nativity differences in birth outcomes among mothers in New York City: The role of social ties and social support. Matern Child Health J, 2014; 18:90–100.

28.

Wakeel

, Witt

, Wisk

, et al. Racial and ethnic disparities in personal capital during pregnancy: Findings from the 2007 Los Angeles Mommy and Baby (LAMB) study. Matern Child Health J, 2014; 18:209–222.

29.

Fleuriet

KJ.

Problems in the Latina paradox: Measuring social support for pregnant immigrant women from Mexico. Anthropol Med, 2009; 16:49–59.

30.

Durand

, Massey

, Zenteno

. Mexican immigration to the United States: Continuities and changes. Latin Am Res Rev, 2001:107–127.

31.

Instituto Nacional de Estadistica y Geografia (INEGI). Censo de Población y Vivienda 2010. Mexico City, Mexico: INEGI; 2011.

32.

Instituto Nacional de Migracion. Estadisticas Migratorias del INM 2009. Mexico City, Mexico: INM-Centro de Estudios Migratorios; 2010.

33.

Minnesota Population Center. Integrated Public Use Microdata Series, International: Version 7.2 [dataset]. Minneapolis, MN: Minnesota Population Center; 2019.

34.

Ekström

, Widström

, Nissen

. Breastfeeding support from partners and grandmothers: Perceptions of Swedish women. Birth, 2003; 30:261–266.

35.

Feldman

, Dunkel-Schetter

, Sandman

, et al. Maternal social support predicts birth weight and fetal growth in human pregnancy. Psychosom Med, 2000; 62:715–725.

36.

González de Cosío

, Escobar-Zaragoza

, González-Castell

, et al. Prácticas de alimentación infantil y deterioro de la lactancia materna en México. Salud Pública Méx, 2013; 55:S170–S179.

37.

Abada

TSJ

, Trovato

, Lalu

. Determinants of breastfeeding in the Philippines: A survival analysis. Soc Sci Med, 2001; 52:71–81.

38.

Afeiche

, Villalpando-Carrión

, Reidy

, et al. Many infants and young children are not compliant with Mexican and international complementary feeding recommendations for milk and other beverages. Nutrients, 2018; 10:466.

39.

Colombara

, Hernández

, Gagnier

, et al. Breastfeeding practices among poor women in Mesoamerica. J Nutr, 2015; 145:1958–1965.

40.

Nguyen

CV.

Does parental migration really benefit left-behind children? Comparative evidence from Ethiopia, India, Peru and Vietnam. Soc Sci Med, 2016; 153:230–239.

41.

Gibson

, McKenzie

, Stillman

. What happens to diet and child health when migration splits households? Evidence from a migration lottery program. Food Policy, 2011; 36:7–15.

42.

Internal migration, international migration, and physical growth of left-behind children: A study of two settings. Health Place, 2015; 36:118–126.

43.

Chávez

, Edelblute

, Korver-Glenn

. Life on the Edge: Balancing Gendered and Occupational Identities among Unauthorized Mexican Migrant Roofers. Qual Sociol, 2016; 39:125–146.

44.

, Ouyang

Y-Q

, Redding

. Family-centered breastfeeding education to promote primiparas' exclusive breastfeeding in China. J Hum Lact, 2018; 34:365–378.

45.

Gilmore

, McAuliffe

. Effectiveness of community health workers delivering preventive interventions for maternal and child health in low- and middle-income countries: A systematic review. BMC Public Health, 2013; 13:847.

46.

Gale

, Kenyon

, MacArthur

, et al. Synthetic social support: Theorizing lay health worker interventions. Soc Sci Med, 2018; 196:96–105.