Skip to main content
NCBI home page
Maternal & Child Nutrition logoLink to Maternal & Child Nutrition
. 2017 Oct 4;14(2):e12537. doi: 10.1111/mcn.12537

Who knows what: An exploration of the infant feeding message environment and intracultural differences in Port‐au‐Prince, Haiti

Elizabeth L Fox 1,2,✉, Gretel H Pelto 1, Kathleen M Rasmussen 1, Marie Guerda Debrosse 3, Vanessa A Rouzier 3, Jean William Pape 3,4, David L Pelletier 1
PMCID: PMC6866243  PMID: 28976068

Abstract

Worldwide, mothers with young children receive many messages about infant feeding. Some messages are generated by health providers and others by the households, communities, and social contexts in which women live. We aimed to determine the scope of infant feeding messages in urban Haiti and to examine intracultural differences in salience of these messages and their alignment with international guidelines. We applied the method of free listing with 13 health workers and 15 human immunodeficiency virus (HIV)‐infected and 15 HIV‐uninfected mothers with infants 0–6 months old at Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes in Port‐au‐Prince, Haiti. Participants listed all messages women receive about infant feeding and specifically about HIV and infant feeding. Message salience was determined by frequency of mention and recall order; messages were coded for key themes. For all groups, the World Health Organization infant feeding recommendations were salient, especially those related to exclusive breastfeeding. Messages across all groups focused on infant health outcomes, with less emphasis on maternal outcomes. Cultural beliefs were also elicited and showed higher salience for mothers than health workers, particularly for consequences of poor maternal nutrition. Health workers' free lists were poorly correlated to those of mothers, whereas those of mothers were highly correlated, regardless of HIV status. Inasmuch as many salient messages were culturally generated, and differences existed between mothers and health workers, we conclude that it is important for health workers to acknowledge the broader infant feeding message environment, and discrepancies within that environment, to address successes and failures in the messages reaching mothers, given potential consequences for mothers' breastfeeding behaviours.

Keywords: breastfeeding, ethnography, free listing, HIV, infant feeding, intracultural variation

1. INTRODUCTION

The World Health Organization (WHO) currently promotes exclusive breastfeeding for the first 6 months of an infant's life and the introduction of nutritious complementary foods with continued breastfeeding from 6 to 24 months (WHO, 2010). However, only 38% of mothers around the world exclusively breastfeed for this duration (WHO/UNICEF, 2014). Instead, many mothers practice mixed feeding (the provision of foods and liquids together with breast milk) when their infants are less than 6 months of age. There is extensive literature on the factors that influence breastfeeding practices (Allen, Ferris, & Pelto, 1986; Cohen, Brown, Rivera, & Dewey, 1999; Otoo, Lartey, & Perez‐Escamilla, 2009), including studies that describe the role of cultural norms and expectations. In this paper, we use the phrase “infant feeding message environment” to denote the totality of these socially articulated norms, expectations, ideas, and messages mothers are exposed to from all sources, including those from health care providers with which mothers interact.

In the context of human immunodeficiency virus (HIV), biomedical considerations further complicate breastfeeding management (Kuhn & Aldrovandi, 2010), and guidelines promoting breastfeeding in the context of HIV have changed several times over the past two decades (Young et al., 2011). Currently, and since 2009, HIV‐infected pregnant women are counselled to take antiretroviral medications and exclusively breastfeed for 6 months with continued breastfeeding and introduction of complementary foods from 6 to 12 months or until an adequate and sustainable diet is achieved (WHO et al., 2010). The current recommendations for HIV‐infected mothers were preceded by recommendations put forward in 2001 that urged abrupt weaning after 5 months of exclusive breastfeeding and in 1998 that promoted exclusive formula feeding (Moland et al., 2010). These changes in recommendations have resulted in a complex infant feeding message environment for HIV‐infected mothers.

Effective counselling for infant feeding requires knowledge about and attention to the personal and cultural factors informing mothers' infant feeding decisions (Lazarus, Struthers, & Violari, 2013). This is particularly important in situations where mothers receive conflicting advice from family, friends, and health care staff (Larsen, Hall, & Aagaard, 2008). However, strategies to improve infant feeding behaviours often do not adequately account for disparities and contradictions in the messages mothers receive from various sources (Tuthill, Chan, & Butler, 2015). There are multiple reasons for this, including limitations on health worker staffing, space, and time, which often force health workers to rely on prescriptive, rather than maternal‐centred, counselling (de Paoli, Manongi, & Klepp, 2002). Prescriptive counselling is also common in contexts where not following recommendations can lead to serious biomedical consequences, such as vertical transmission of HIV (Desclaux, 2013; Williams, Alderson, & Farsides, 2002). Our research was motivated by our observation that the extent to which women absorb and engage with infant feeding messages, including those received from within and outside of health centres, has not been adequately investigated. To examine this, we must first address knowledge gaps about the messages that compose women infant feeding message environments.

In this study, conducted in a health clinic in Port‐au‐Prince, Haiti, that serves both HIV‐infected and HIV‐uninfected women, we employed systematic, ethnographic methods with the following aims: (a) to determine the broader infant feeding message environment and the salience of different messages in that environment and (b) to identify differences among messages that are salient to HIV‐infected mothers, HIV‐uninfected mothers, and health workers. As defined by Weller and Romney (1988), salient messages are defined as those that are “better known,” “important,” or “familiar” as compared to other messages for a group of individuals, and in this case, each of the three stakeholder groups. For the first aim, we hypothesized that (a) the infant feeding message environment would contain a variety of recommendations, including messages that conflicted with one another. For the second aim, we hypothesized that (b) there would be variation in the messages that were most salient for each stakeholder group and that (c) salient messages for both groups of mothers would be poorly aligned with the WHO's infant feeding recommendations, whereas health workers' salient messages would be well aligned.

Key messages.

  • The infant feeding message environment for mothers contained a large array of messages, which often conflicted with one another.

  • HIV‐infected and HIV‐uninfected mothers shared a similar infant feeding message environment and expressed concerns about maternal nutrition and breast milk quality.

  • Messages about early initiation, on‐demand feeding, and breast milk expression were more salient for health workers than mothers. The economic and developmental benefits of breastfeeding were also more salient for health workers than mothers.

  • Messages related to mothers' health and well‐being were not well reflected in the message environment.

2. PARTICIPANTS AND METHODS

2.1. Study context

The study was conducted from March to April 2014 at the Haitian Group for the Study of Kaposi's Sarcoma and Opportunistic Infections (Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes [GHESKIO]), a Haitian nongovernmental institution whose main mission is to provide free care to individuals affected by HIV/AIDS and tuberculosis in Port‐au‐Prince, Haiti. GHESKIO is the main site of HIV/AIDS care in Port‐au‐Prince, where approximately 2.8% of women ages 15 to 49 are infected with HIV (Cayemittes et al., 2013). GHESKIO also provides services to people in adjacent communities, not only those affected by HIV and tuberculosis. Their programmes include community outreach and health education, screening and counselling for infectious diseases, and reproductive health services.

At the time of the study, GHESKIO had two separate nutrition programmes—one for HIV‐infected mothers and one for HIV‐uninfected mothers. These programmes provided monthly nutrition counselling and monitoring to infants and young children (Ayoya et al., 2013; Heidkamp, Stoltzfus, Fitzgerald, & Pape, 2012). All mothers had access to antenatal and post‐natal care, as well as preventative paediatric services for their children. The nutrition programme for HIV‐infected women was closely integrated with the antenatal and paediatric clinics at GHESKIO that provided care specifically to HIV‐infected mothers and their infants. The nutrition programme for HIV‐uninfected women was integrated with the antenatal and paediatric services delivered by the community clinic that provided care to members of the greater community.

Both nutrition programmes recommended exclusive breastfeeding to participating mothers. HIV‐uninfected mothers were counselled to breastfeed exclusively for their infants' first 6 months of life and to continue breastfeeding through the first 2 years of their infants' life (WHO, 2010). HIV‐infected mothers were counselled to breastfeed exclusively for their infants' first 6 months of life and to continue breastfeeding for a minimum of 12 months, according to the WHO guidelines for prevention of vertical transmission of HIV (WHO et al., 2010). However, the recommendation for exclusive breastfeeding is not aligned with Haitian cultural norms. In Haiti, it is common for infants to receive foods before 6 months of age, and although nearly all (95.7%) infants are breastfed at 4–5 months of age, only 23.9% of infants are exclusively breastfed at that age (Cayemittes et al., 2013).

The changes in WHO infant feeding guidelines in the context of HIV created challenges for health workers (Lazarus et al., 2013; Moland et al., 2010). At GHESKIO, health workers promoted replacement feeding and provided free formula to HIV‐infected women until 2010. In 2010, GHESKIO transitioned to promoting exclusive breastfeeding for HIV‐exposed infants and lifelong antiretroviral therapy for HIV‐infected women in its programmes. GHESKIO health workers who worked at the clinic before 2010 thus changed the set of messages they provided to HIV‐infected mothers, and some of the HIV‐infected mothers, who had children followed at GHESKIO before 2010, were also exposed to varied and conflicting infant feeding messages.

2.2. Study population and recruitment

Using purposive sampling (Pelto, 2013), we recruited 15 HIV‐infected and 15 HIV‐uninfected breastfeeding mothers and 13 frontline health workers. All mothers had infants 0–6 months of age and participated in GHESKIO's nutrition programmes. They were recruited from the GHESKIO clinic during regular health centre or nutrition programme visits. All frontline health workers were GHESKIO employees and provided counselling to HIV‐infected and/or HIV‐uninfected breastfeeding women at the antenatal, paediatric and/or nutrition clinics, or counselling clinic for newly infected patients.

All participants provided oral and written informed consent before their enrolment in the study. The consent process was provided to mothers in Haitian Creole and to health workers in French. Ethical approval was granted by the GHESKIO Human Ethics Committee and by the Cornell University Institutional Review Board for Human Participants.

2.3. Data collection

We used the free‐listing technique to elicit a list of messages from each group of participants. Free listing is an ethnographic cognitive mapping technique used to compile items in a cultural domain. A cultural domain is a set of items that relate to one another in some way (Pelto, 2013), such as the set of messages from all sources that pertain to infant feeding, and that collectively constitute the infant feeding message environment. As a systematic technique, and in contrast to open‐ended questions that elicit unique experiences of individuals, free listing captures items in a domain and allows for comparisons across different groups (Borgatti & Halgin, 2013).

In individual interviews, the first author and the head of the nutrition clinic asked each participant to list all messages they thought women received about infant feeding. We did not ask participants what messages they had heard but rather interviewed them as key informants, asking in Haitian Creole, “What messages do mothers receive about feeding babies less than six months old?” We probed for messages from families and the community, as well as from the health centres. HIV‐infected mothers were specifically encouraged to provide messages related to infant feeding and HIV. We also prompted mothers to describe the meaning of each message they provided. All interviews with mothers were recorded and transcribed. Given scheduling limitations, we provided each health worker with a questionnaire in which they were asked to write down all infant feeding messages received by women and specifically HIV‐infected women. The health workers were also asked for messages they thought women heard from their families and communities, as well as at the health centre.

2.4. Data analysis

To achieve our first aim of characterizing the broader infant feeding message environment, we compiled all messages from the free‐listing exercise into a single document in Haitian Creole, following standard free‐listing analysis procedure (Weller & Romney, 1988). Any duplicates in an individual participant's list were deleted, so that only the first mention of a message was counted. All messages were translated into English by the first author and reviewed for accuracy by collaborators in Haiti. We determined data saturation using FLAME v1.1 (Pennec, Wencelius, Garine, Raimon, & Bohbot, 2012) by evaluating the point at which the consideration of an additional participant generated no new additional messages to the free list. We then determined whether we needed additional informants by assessing the list stability, that is, that the relative frequencies of the top items did not change as more informants were added (Borgatti & Halgin, 2013).

As part of our first aim, we determined the salience of each message in the message environment, considering all participants together and each stakeholder group separately. Salience is a technical construct in cognitive mapping analysis and, for free‐listing data, is measured by both frequency of mention and order of recall of free‐listed items (Borgatti & Halgin, 2013). We used the Smith's S index (S), a measure that accounts for both frequency and recall order, to determine the salience of the listed messages. Smith's S for each message was calculated using the following formula: S = {∑[(Li − Rj + 1)/Li]}/N, where L i represents the length of the list for informant i, R j represents the rank of the message j in that list, and N represents the total number of participants (Smith, 1993; Smith & Borgatti, 1997). Scores range from 0.0 to 1.0; values approaching 0.0 represent low salience (i.e., an item or message is mentioned last and with low frequency); and those approaching 1.0 represent high salience (i.e., an item or message is mentioned first on every participant's list; Sutrop, 2001). We created a scree plot of the messages by plotting the messages from greatest to least salience. A natural break or elbow often occurs in the plot between the few items that are highly salient and mentioned by many respondents and the large number of items mentioned by only one person (Borgatti & Halgin, 2013). We identified the natural break of the plot to determine the appropriate cut‐off for the most salient messages. We considered both the full sample of participants together (i.e., HIV‐infected mothers, HIV‐uninfected mothers, and health workers) and each participant group separately (Borgatti & Halgin, 2013). We used FLAME v1.1 (Pennec et al., 2012) to calculate salience and to produce the plots.

To examine our first hypothesis, we characterized each message in the infant feeding message environment by theme and identified the similarities and differences between the most salient messages. We identified the themes using thematic analysis of the most salient messages (Boyatzis, 1998). We examined the messages in the message environment for conflicts and discrepancies in content, as well as gaps in topics that did not appear in the sets of most salient messages.

To achieve our second aim of determining differences in message salience between stakeholder groups, we examined the specific messages that comprised the most salient messages for each group. Using heat maps, the salience indices of the most salient messages were compared for similarities and differences. Heat maps use colour gradients to graphically represent values in a matrix or table (Perez‐Llamas & Lopez‐Bigas, 2011). In this study, we scaled the heat map from black for the most salient messages (S = 1.0) to white for the least salient messages (S = 0.0), as determined by conditional formatting in Excel v14.5.5 (Microsoft, 2011).

To test our second hypothesis that there would be variation between the groups of participants, we used correlation analysis to compare the most salient messages of each group of participants (Thompson & Juan, 2006). The correlation matrix compared the most salient messages of each participant group to the other participant groups' salience scores for those same messages. For example, the most salient messages for health workers were compared to the salience values for those same messages provided by HIV‐infected and HIV‐uninfected mothers. The analysis for the correlations was conducted using R v3.2.0 (R Core Team, 2015).

Finally, to test our third hypothesis about each group's alignment to the WHO infant feeding recommendations, we compared the degree to which each group's most salient messages corresponded to a list of WHO infant feeding messages. We developed the list of the WHO infant feeding messages by cross‐referencing documents identifying the WHO infant feeding guidelines and Haitian Ministry of Health infant feeding guidelines (Ministère de la Santé Publique et de la Population, 2008, 2011; UNICEF, 2012; WHO, 2015). We assessed alignment by determining whether each participant group's most salient messages mentioned or did not mention messages aligned with the WHO infant feeding guidelines.

3. RESULTS

3.1. Participant characteristics

HIV‐infected and HIV‐uninfected mothers differed in their sociodemographic characteristics (Table 1). Compared to HIV‐uninfected mothers, HIV‐infected mothers tended to be older, have lower levels of education, live in peri‐urban areas of Port‐au‐Prince, and have more experience with the GHESKIO nutrition programmes.

Table 1.

General characteristics of the 30 HIV‐infected and HIV‐uninfected mothers who participated in the free‐listing exercisea

Characteristic Total (n = 30) HIV‐infected (n = 15) HIV‐uninfected (n = 12–15)b
Age, years (mean ± SD) 29.1 ± 7.5 31.1 ± 6.8 26.6 ± 7.8
Education
None 3 (11.1) 2 (13.3) 1 (8.3)
Any pre‐school 1 (3.7) 1 (6.7) 0 (0.0)
Any primary 8 (33.3) 5 (33.3) 3 (33.3)
Any secondary or above 15 (55.6) 7 (46.7) 8 (66.7)
Urban, lives in Port‐au‐Prince 28 (93.3) 13 (86.7) 15 (100.0)
>1 child in nutrition programme 9 (33.3) 8 (53.3) 1 (8.3)
Open in a new tab

Note. HIV = human immunodeficiency virus; SD = standard deviation.

a

Values are n (%), unless otherwise noted.

b

There are missing values for age, education, and prior experience with the nutrition programme for three HIV‐uninfected mothers.

Of the 13 health workers, seven worked in the nutrition programme, three in the paediatric clinic, two in the antenatal clinic, and one in the counselling clinic for newly infected patients. The health worker sample included two nurse practitioners, four nurses, two nurse auxiliaries, two social workers, and three field workers. All health workers had received infant feeding training at GHESKIO and/or from partnering organizations. Health workers from the nutrition programme were primarily responsible for providing infant feeding messages and nutrition education to mothers, whereas health workers from the other clinics were responsible for providing infant feeding messages, in addition to other health messages. Twelve of the 13 health workers were female.

3.2. Data saturation

Participants collectively identified 125 messages about infant feeding, and each participant listed a mean of 17 messages (range: 7–28 messages). List lengths were similar across all participant groups. We achieved saturation of messages in all groups and within each group of participants. Specifically, of the total 43 participants, 26 were sufficient to obtain all the messages (and the remaining 17 participants added no further messages to the free list). Within each participant group, saturation was reached after 13 participants for HIV‐infected mothers (n = 15), after 12 participants for HIV‐uninfected mothers (n = 15), and after 11 participants for health workers (n = 13). The content of the top messages was stable for each group when we reached saturation, and we determined that no additional informants were needed.

3.3. Determining the infant feeding message environment

With regard to our first aim to characterize the infant feeding message environment, we found that all of the participants together identified 125 unique messages. Forty of those messages were mentioned only once. The scree plots of the free‐list messages for all participants and for each participant group showed an elbow at about 15 messages. As such, the top 15 messages were used to represent the most salient messages.

3.4. Content of and conflict among the most salient messages

With respect to the first hypothesis, we found that the infant feeding message environment included a number of messages that represented messages supported by the WHO infant feeding guidelines. However, it also included messages that did not conform to the guidelines. Additionally, messages were often in conflict with one another. The most salient conflict between messages was related to providing the infant with breast milk only versus providing the infant with other foods before 6 months of age. This latter message stemmed from culturally generated norms that were intended to ensure that the infant had enough to eat, given the common perception that breast milk alone was insufficient.

Another aspect of the infant feeding message environment was that messages largely focused on the infant, and there were gaps in content reflecting mothers' experiences. The majority of messages focused on infant outcomes (e.g., satiety and health) and maternal outcomes were often only referenced in relation to the quantity and quality of a mother's breast milk. For instance, “Eat well if you are breastfeeding so that your milk has strength” and “Eat everything you find so that you can continue breastfeeding.”

3.5. Similarities and differences among salient messages in each stakeholder group

Our second aim was to obtain a picture of the extent of similarities and differences among the most salient messages for the different participant groups. The heat maps of the top 15 most salient messages for each participant group (Table 2) showed some variability in the most salient messages of each group. A thematic analysis of similarities and differences between groups showed that (a) neither of the mother groups contained any themes that were unique to their group; (b) there were many aspects of infant feeding that were salient for mothers but not health workers; and (c) health workers attended to more biomedical aspects of infant feeding than either mother group (Figure 1).

Table 2.

Heat map of the 15 most salient free‐listed messages for HIV‐infected mothers (n = 15), HIV‐uninfected mothers (n = 15), and health workers (n = 13)

Salient messages Smith's S index
HIV‐infected mothers HIV‐uninfected mothers Health workers
Give your baby breast milk only, without any other foods or liquids 0.705a 0.447b 0.726c
Breastfeed so the baby does not get any sickness 0.645a 0.723b 0.361c
Eat well if you are breastfeeding so that your milk has strength 0.473a 0.358b 0.145
Give the baby food to eat 0.463a 0.409b 0.068
Breast milk does not do anything for the baby; it needs other food 0.414a 0.339b 0.225c
Breastfeed for 6 months 0.393a 0.280b 0.158
Breastfeed the baby 0.384a 0.240b 0.066
Exclusively breastfeed your baby so that it does not get HIV 0.383a 0.000 0.084
Exclusively breastfeed for 6 months 0.267a 0.205 0.501c
Give the baby other foods when it is 6 months old 0.230a 0.229 0.000
A child gets everything they need from a mother's milk 0.210a 0.256b 0.447c
Exclusively breastfeed so the baby is not sick 0.197a 0.075 0.288c
Do not breastfeed because you are not well‐nourished, and you will become wasted 0.164a 0.148 0.124
Wash your hands each time you are going to breastfeed 0.162a 0.093 0.262c
A mother's milk is very good 0.157a 0.000 0.166
Breastfeed regularly for the baby to gain more weight 0.069 0.371b 0.078
Eat everything you find so that you can continue breastfeeding 0.154 0.338b 0.042
Exclusively breastfeed for the well‐being of the baby 0.017 0.287b 0.132
Breastfeed your baby so that it can develop 0.084 0.254b 0.207c
Giving food early (mixed feeding) is not good for the baby 0.151 0.247b 0.142
Exclusively breastfeed because that is good 0.000 0.244b 0.000
Breastfeeding counts a lot in a mother's life 0.013 0.233b 0.088
Breast milk is the best milk for a baby 0.073 0.094 0.454c
Do not give food with breast milk because you will hurt the baby's stomach 0.106 0.018 0.355c
Exclusively breastfeed because it is economical, and you do not have to buy anything 0.092 0.195 0.301c
Breastfeed the baby as soon as it is born 0.043 0.067 0.294c
Take your medication correctly for the baby not to get sick 0.153 0.000 0.290c
Express your milk into a clean vessel when you leave the house 0.055 0.069 0.197c
Breastfeed when the baby demands/asks for it 0.059 0.041 0.174c
Open in a new tab

Note. HIV = human immundeficiency virus. Messages that were shared across two or more groups are noted once. The following symbols denote the 15 most salient messages for the following groups:

a

HIV‐infected mothers,

b

HIV‐uninfected mothers,

c

Health workers.

The heat map scales from black (S=1.0, a message that is mentioned first on every participant's list) to white (S=0.0, a message that is not mentioned at all).

Figure 1.

Figure 1

Open in a new tab

Venn diagram of themes of the 15 most salient messages for HIV‐infected mothers, HIV‐uninfected mothers, and health workers. HIV = human immunodeficiency virus; WHO IYCF = World Health Organization Infant and Young Child Feeding

Only four messages were shared by all three groups in their most salient messages (Table 2):

Breastfeed so that the baby doesn't get any sickness

Breastmilk doesn't do anything for the baby; it needs other food (<6 months)

Give your baby breastmilk only with no other foods or liquids

A child gets everything they need from a mother's milk.

The shared messages focused on protecting the health of the baby and on exclusive breastfeeding. Only one message shared by all groups was not consistent with the WHO infant feeding guidelines, namely, giving the baby other foods because breast milk was insufficient to satisfy the baby's needs. In addition to the four messages shared across all groups, HIV‐infected and HIV‐uninfected mothers shared four messages, HIV‐infected mothers and health workers shared three messages, and HIV‐uninfected mothers and health workers shared one message (Table 2).

Of the messages and themes that were shared by both groups of mothers, many related to the quality and quantity of breast milk and its ability or inability to be nutritionally adequate for infants less than 6 months of age. Messages such as “Give the baby food to eat” promoted infant feeding practices that were not aligned with the WHO early infant breastfeeding guidelines. Mothers cited more of these messages among their 15 most salient messages than did health workers: HIV‐infected mothers reported five nonaligned early infant feeding messages, and HIV‐uninfected mothers reported six, compared with only one message reported by health workers (Table 2). Mothers more often referred to breastfeeding generally rather than distinguishing between breastfeeding and exclusive breastfeeding. The number of salient messages about exclusive breastfeeding did not differ between the two groups of mothers.

The health workers shared some similarities with the mother groups. Namely, health workers and HIV‐infected mothers both recalled HIV‐specific messages related to taking antiretroviral medication and exclusive breastfeeding. Both groups shared messages related to good hygiene among their most salient messages, specifically “Wash your hands each time you are going to breastfeed.” Health workers and HIV‐uninfected mothers shared only one salient message: “Breastfeed your baby so that it can develop.”

3.6. Statistical variation in salient messages for each stakeholder group

To address the second hypothesis, we examined correlations of the most salient messages to show the degree of variation between the different participant groups (Table 3). The most salient messages for both HIV‐infected and HIV‐uninfected mothers were significantly correlated to one another (p < .01 and p < .001). Health workers' salience indices for mothers' most salient messages were not significantly correlated to either mother group. However, HIV‐infected mothers were correlated to the health workers' most salient messages (p < .05), although the Pearson correlation coefficient was lower than those of the correlations between the mothers' groups.

Table 3.

Correlations of salience for the 15 most salient messages by participant group (Pearson's r coefficient)

Participant group HIV‐infected mothers 15 most salient messages HIV‐uninfected mothers 15 most salient messages Health workers 15 most salient messages
HIV‐infected mothers 1.000 0.681** 0.592*
HIV‐uninfected mothers 0.801*** 1.000 0.383
Health workers 0.344 0.412 1.000
Open in a new tab

Note. Columns represent lists from each group of the most salient messages (ranked by Smith's S index); rows show the correlation of those rankings for the same messages between the participant group originating the list and the results of the other participant groups (Thompson & Juan, 2006). HIV = human immunodeficiency virus.

*

p < .05.

**

p < .01.

***

p < .001 one‐tailed test.

3.7. Alignment with the WHO infant feeding guidelines

In addressing the third hypothesis, we found that the greatest difference between mothers and health workers appeared in messages aligned with WHO recommendations. Health workers listed many messages that were consistent with the WHO infant feeding recommendations (Table 4). Most of these messages were more salient for health workers than for HIV‐infected and HIV‐uninfected mothers, particularly messages on early initiation, on‐demand feeding, and expressing breast milk. However, HIV‐infected mothers elicited more messages that were aligned with the WHO guidelines than HIV‐uninfected mothers: Five of HIV‐infected mothers' most salient messages aligned with WHO recommendations compared to two of the HIV‐uninfected mothers' most salient messages. Messages related to lactation amenorrhea and weaning were not salient for any participant groups.

Table 4.

Alignment of the most salient messages for HIV‐infected mothers, HIV‐uninfected mothers, and health workers with the WHO infant feeding guidelines

WHO infant feeding messages HIV‐infected mothers HIV‐uninfected mothers Health workers
Exclusive breastfeeding: Exclusively breastfeed for first 6 months of life with no additional food or drink, even water X X X
Preventing illness: Exclusively breastfeed for the first 6 months to protect the baby against many different illnesses X X X
Hygiene/sanitation: Practice good hygiene and wash your hands before feeding the baby X X
Early initiation: Initiate breastfeeding within 1 hr of birth X
Responsive feeding: Breastfeed on demand; feed before the baby cries X
HIV and ARVs: If you have HIV, take your medication to reduce the chance that the baby gets the virus X
Expressing milk: Express your milk when you are separated from your baby X
Complementary feeding: At 6 months, you should continue breastfeeding and introduce nutritionally adequate and safe complementary foods X
HIV and exclusive breastfeeding: Mothers who are HIV infected should exclusively breastfeed as long as artificial formula is not AFASS X
Lactation amenorrhea: Breastfeeding contributes to the health and well‐being of mothers by helping to space pregnancies
Weaning: Weaning the baby can take a month; mothers with HIV should wean their babies at 12 months
Open in a new tab

Note. HIV = human immunodeficiency virus; ARV = antiretroviral; WHO = World Health Organization; AFASS = Acceptable, Feasible, Affordable, Sustainable, and Safe.

4. DISCUSSION

In Haiti, mothers interact with an infant feeding message environment that includes many diverse sets of messages. Our findings have revealed the complex message environment of Haitian, breastfeeding mothers. We found that the messages that are salient for mothers are often conflicting and are not always aligned with WHO infant feeding guidelines. Our study has illuminated the importance of investigating message salience as a way to ascertain what information is retained by breastfeeding mothers, as well as the successes and gaps in existing counselling strategies. Understanding the messages that are salient for mothers informs decisions about the tools and counselling practices that could be employed to help mothers succeed with their infant feeding practices in the complex environment in which they reside.

Our free‐listing results also illuminated the heavy emphasis on infant health and developmental outcomes, which overshadowed concerns about mothers' health outcomes. Although mothers mentioned messages related to their own health and nutrition, these primarily related to the quality of their milk and, as such, were still oriented to the baby's nutrition. The fact that messages related to outcomes of mothers' health and well‐being were not highly salient may reflect a feature of the message environment itself. This is not surprising, particularly in clinics and other health care settings, as the majority of tools and recommendations health workers receive for counselling focus on infants (UNICEF, 2012). Other sources have also shown that reproductive health messaging frequently focuses on fetal and infant health rather than on the health of the mother (Parrott & Condit, 1996). The absence of mother‐oriented messages, at least as reflected in the message environment elicited in our study, calls for further investigation. As HIV‐infected mothers face many stressors in addition to preventing mother‐to‐child transmission of HIV (D'auria, Christian, & Miles, 2006), it is important to use these communications as an opportunity to focus on self‐care, as well as on the care of babies.

All participant groups, including both groups of mothers, were familiar with exclusive breastfeeding recommendations. HIV‐infected mothers in our study were also familiar with the current WHO guidelines for HIV and infant feeding and did not reference previous versions of the recommendations among their most salient messages. Findings from other studies in Haiti (Laterra, Ayoya, Beaulière, Bienfait, & Pachón, 2014) and Zambia (Fjeld et al., 2008) indicate that mothers who receive infant feeding counselling are knowledgeable about exclusive breastfeeding. The high salience of these messages among mothers indicates the effectiveness of information transfer from health workers to mothers about this topic. This suggests that gaps in breastfeeding practice among these mothers do not reflect knowledge barriers but rather social and material barriers to implementation. Additionally, HIV‐infected and HIV‐uninfected mothers shared the same number of highly salient messages specific to exclusive breastfeeding. This suggests that health workers do not favour HIV‐infected women compared to HIV‐uninfected women in their transmission of messages about exclusive breastfeeding. This finding is counter to findings in other contexts, where HIV‐infected mothers received more infant feeding counselling and support than HIV‐uninfected mothers (Goga et al., 2012; Orne‐Gliemann et al., 2006).

However, some messages were strikingly absent from mothers message environment. Messages about early initiation, on‐demand feeding, and expressing breast milk were more salient for health workers than for mothers. Other benefits of breastfeeding, including economic benefits and mother–child bonding, were also more salient for health workers than for mothers. Similarly, other research in Haiti describes a greater emphasis on motor skills and development of breastfed infants by mothers compared to benefits of closeness, bonding, and love (Dörnemann & Kelly, 2013). The differences in messages that are more salient for health workers than for mothers may stem from differing perceptions about the relevance of those messages. For instance, some mothers might view exclusive breastfeeding as economically burdensome because they perceive a need to buy food of sufficient quality for themselves if they exclusively breastfeed. On the other hand, health workers may perceive it as economically advantageous because it reduces costs associated with formula and health care. It is important to understand these discrepancies in order to close communication gaps between individuals.

This study also revealed messages in the message environment that originated in the larger cultural environment and reflected long‐standing cultural perceptions and knowledge. For example, breastfeeding is considered a duty by many mothers in Haiti (Dörnemann & Kelly, 2013), and breastfeeding, albeit not exclusive breastfeeding, is seen as a cultural norm. In our study, mothers more often referenced breastfeeding, as a general practice, as opposed to exclusive breastfeeding. Additionally, we found that messages about maternal diet, as it related to the “strength” or quality of breast milk, were highly salient for both mother groups. This aligns with cultural knowledge that emphasizes that what women eat affects the quality and quantity of their milk; as noted in other studies in Haiti, this belief was associated with poor breastfeeding practices and early introduction of solid foods (Balogun, Dagvadorj, Anigo, Ota, & Sasaki, 2015; Dörnemann & Kelly, 2013; Laterra et al., 2014).

Contrary to our second hypothesis, our findings show that HIV‐infected and HIV‐uninfected mothers shared a similar infant feeding message environment, although there were important differences between the mothers groups and health workers. Both groups of mothers identified some messages that were consistent with WHO guidelines and others that were not. In contrast, health workers identified many more messages that reflected the WHO infant feeding guidelines, and many of the non‐WHO‐aligned messages did not fall within the 15 most salient messages for health workers. This finding is consonant with research in other settings that shows strong fidelity in the delivery of infant feeding messages by health workers (Kim et al., 2015). We interpret this as evidence that, though health workers are aware of the wide range of messages women receive about feeding their babies, the community‐generated messages were not as relevant for them.

4.1. Programmatic and policy implications

Given the differences between health workers and mothers and the salience of cultural messages, it is essential that infant feeding recommendations and counselling are provided in a manner that accounts for the local context in which women are functioning—both temporally and culturally (Dörnemann & Kelly, 2013; Zarcadoolas, Pleasant, & Greer, 2006). This requires shifting from framing the question of “Why aren't women doing what we want them to do?” to “How could women be expected to do other than what they are doing?” (Parrott & Condit, 1996). This is particularly relevant in situations in which mothers' experiences do not fit well within the recommendations received from health workers. In such circumstances, mothers may find that the recommendations from the health facility are challenging to implement, in part because they focus more on the product of breastfeeding than the process. This may leave them dissatisfied or uneasy with their counselling interaction and more inclined to embrace other recommendations they receive (Burns, Schmied, Fenwick, & Sheehan, 2012; McInnes & Chambers, 2008).

This study provides a window into the infant feeding message environment of breastfeeding, care‐seeking mothers in Haiti, and the health workers who interact with them. Although our findings are specific to this context and not necessarily generalizable, we believe that the complexity of the infant feeding message environment, as interpreted and recalled by participants of our study, is not unique to Haiti and is relevant to other contexts (Al‐Mujtaba, Sam‐Agudu, & Khatri, 2016; Tuthill et al., 2015). The study also illustrates the utility of free listing to reveal new insights, not only for programmatic formative research but also for ongoing programme assessment. Although we examined issues from the perspective of participants' recall and did not include a behavioural, observational component, we used a mixed‐methods approach to assess variation in a programmatic setting in a fashion that attempts to preserve participants' experiences.

In conclusion, although WHO infant feeding guidelines reach communities all over the globe, they are often not fully aligned with some of the messages mothers receive outside of the health centre. An important consequence of this disconnect is that many infants do not receive the full benefits of breastfeeding, and many mothers do not receive the support they need to achieve optimal breastfeeding practices. Programmes that provide services for women and their infants using multipronged, socioecological approaches have an important role to play in supporting better practices (Aidam, Perez‐Escamilla, & Lartey, 2005; Bhandari et al., 2003). Achieving ambitious public health goals requires further attention to the impediments to effective communication and requires further knowledge about the environments in which caregivers live, as well as about the most effective ways to improve infant feeding practices around the world.

CONFLICTS OF INTEREST

The authors declare that they have no conflicts of interest.

CONTRIBUTIONS

ELF, GHP, KMR, and DLP were involved in the conception and design of the research; ELF, MGD, VAR, and JWP were involved in the acquisition of the data; ELF, GHP, KMR, and DLP were responsible for the analysis and interpretation of the data; ELF wrote the paper and had primary responsibility for the final content; and GHP, KMR, MGD, VAR, JWP, and DLP were involved in providing detailed comments and revising the manuscript for important intellectual content.

ACKNOWLEDGMENTS

The authors thank the mothers who participated in the study, GHESKIO for the opportunity to complete the research, and the GHESKIO team (Edwidge Byron, Ghislaine Saint Louis, Ruthza Mondesir, Evelyne Pierre Solon, Adeline Bernard, Suzette Fleury, Sandra Dorceus, and Vanessa Rivera) for their collaboration in making this study possible. We would also like to thank Barbara Strupp for her feedback on the design of the study and Lynn Johnson and Francoise Vermeylen from the Cornell Statistical Consulting Unit for their guidance on statistical analyses.

Fox EL, Pelto GH, Rasmussen KM, et al. Who knows what: An exploration of the infant feeding message environment and intracultural differences in Port‐au‐Prince, Haiti. Matern Child Nutr. 2018;14:e12537 10.1111/mcn.12537

REFERENCES

  1. Aidam, B. , Perez‐Escamilla, R. , & Lartey, A. (2005). Lactation counseling increases exclusive breast‐feeding rates in Ghana. The Journal of Nutrition, 135(7), 1691–1695. [DOI] [PubMed] [Google Scholar]
  2. Allen, L. H. , Ferris, A. M. , & Pelto, G. H. (1986). Maternal factors affecting lactation In Hamosh M., & Goldman A. J. (Eds.), Human lactation II: Maternal and environmental factors (pp. 51–60). New York, NY: Plenum Press. [Google Scholar]
  3. Al‐Mujtaba, M. , Sam‐Agudu, N. A. , & Khatri, R. (2016). Barriers to the practice of exclusive breastfeeding among HIV‐positive mothers in sub‐Saharan Africa: A scoping review of counselling, socioeconomic and cultural factors. J AIDS HIV Res, 8(6), 70–79. [Google Scholar]
  4. Ayoya, M. A. , Golden, K. , Ngnie‐Teta, I. , Moreaux, M. D. , Mamdoultaibou, A. , Koo, L. , … Marhone, J. P. (2013). Protecting and improving breastfeeding practices during a major emergency: Lessons learnt from the baby tents in Haiti. Bulletin of the World Health Organization, 91, 612–617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Balogun, O. O. , Dagvadorj, A. , Anigo, K. M. , Ota, E. , & Sasaki, S. (2015). Factors influencing breastfeeding exclusivity during the first 6 months of life in developing countries: A quantitative and qualitative systematic review. Maternal & Child Nutrition, 11(4), 433–451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bhandari, N. , Bahl, R. , Mazumdar, S. , Martines, J. , Black, R. , & Bhan, M. (2003). Effect of community‐based promotion of exclusive breastfeeding on diarrhoeal illness and growth: A cluster randomised controlled trial. Lancet, 361(9367), 1418–1423. [DOI] [PubMed] [Google Scholar]
  7. Borgatti, S. P. , & Halgin, D. S. (2013). Elicitation techniques for cultural domain analysis In Schensul J. J., & Lecompte M. D. (Eds.), Specialized ethnographic methods: A mixed methods approach (pp. 80–116). Lanham, MD: AltaMira Press. [Google Scholar]
  8. Boyatzis, R. E. (1998). Transforming qualitative information: Thematic analysis and code development. Thousand Oaks, CA: SAGE Publications. [Google Scholar]
  9. Burns, E. , Schmied, V. , Fenwick, J. , & Sheehan, A. (2012). Liquid gold from the milk bar: Constructions of breastmilk and breastfeeding women in the language and practices of midwives. Social Science & Medicine, 75(10), 1737–1745. [DOI] [PubMed] [Google Scholar]
  10. Cayemittes, M. , Busangu, M. F. , Bizimana, J. D. , Barrère, B. , Sévère, B. , Cayemittes, V. , & Charles, E. (2013). Enquête mortalité, morbidité et utilisation des services, Haiti 2012. Calverton, Maryland, USA: Ministère de la Santé Publique et de la Population, l'Insitut Haïtien de l'Enfance and ICF International. [Google Scholar]
  11. Cohen, R. J. , Brown, K. H. , Rivera, L. L. , & Dewey, K. G. (1999). Promoting exclusive breastfeeding for 4–6 months in Honduras: Attitudes of mothers and barriers to compliance. Journal of Human Lactation, 15(1), 9–18. [DOI] [PubMed] [Google Scholar]
  12. D'auria, J. P. , Christian, B. J. , & Miles, M. S. (2006). Being there for my baby: Early responses of HIV‐infected mothers with an HIV‐exposed infant. Journal of Pediatric Health Care, 20(1), 11–18. [DOI] [PubMed] [Google Scholar]
  13. de Paoli, M. , Manongi, R. , & Klepp, K. (2002). Counsellors' perspectives on antenatal HIV testing and infant feeding dilemmas facing women with HIV in northern Tanzania. Reproductive Health Matters, 10(20), 144–156. [DOI] [PubMed] [Google Scholar]
  14. Desclaux, A. (2013). Senegalese women living with HIV versus the 2009 WHO recommendations for PMTCT: Meanings for resistance regarding infant feeding In Liamputtong P. (Ed.), Women, motherhood and living with HIV/AIDS (pp. 161–176). Netherlands: Springer Science & Business Media. [Google Scholar]
  15. Dörnemann, J. , & Kelly, A. H. (2013). “It is me who eats, to nourish him”: A mixed‐method study of breastfeeding in post‐earthquake Haiti. Maternal & Child Nutrition, 9(1), 74–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fjeld, E. , Siziya, S. , Katepa‐Bwalya, M. , Kankasa, C. , Moland, K. , Tylleskar, T. , & GROUP TP‐ES . (2008). “No sister, the breast alone is not enough for my baby” a qualitative assessment of potentials and barriers in the promotion of exclusive breastfeeding in southern Zambia. International Breastfeeding Journal, 3, 26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Goga, A. E. , Doherty, T. , Jackson, D. J. , Sanders, D. , Colvin, M. , Chopra, M. , & Kuhn, L. (2012). Infant feeding practices at routine PMTCT sites, South Africa: Results of a prospective observational study amongst HIV exposed and unexposed infants—Birth to 9 months. International Breastfeeding Journal, 7, 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Heidkamp, R. A. , Stoltzfus, R. J. , Fitzgerald, D. W. , & Pape, J. W. (2012). Growth in late infancy among HIV‐exposed children in urban Haiti is associated with participation in a clinic‐based infant feeding support intervention. The Journal of Nutrition, 142(4), 774–780. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kim, S. S. , Ali, D. , Kennedy, A. , Tesfaye, R. , Tadesse, A. W. , Abrha, T. H. , … Menon, P. (2015). Assessing implementation fidelity of a community‐based infant and young child feeding intervention in Ethiopia identifies delivery challenges that limit reach to communities: A mixed‐method process evaluation study. BMC Public Health, 15, 316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kuhn, L. , & Aldrovandi, G. (2010). Survival and health benefits of breastfeeding versus artificial feeding in infants of HIV‐infected women: Developing versus developed world. Clinics in Perinatology, 37(4), 843–862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Larsen, J. S. , Hall, E. O. , & Aagaard, H. (2008). Shattered expectations: When mothers' confidence in breastfeeding is undermined—A metasynthesis. Scandinavian Journal of Caring Sciences, 22(4), 653–661. [DOI] [PubMed] [Google Scholar]
  22. Laterra, A. , Ayoya, M. A. , Beaulière, J.‐M. , Bienfait, M. M. , & Pachón, H. (2014). Infant and young child feeding in four departments in Haiti: Mixed‐method study on prevalence of recommended practices and related attitudes, beliefs, and other determinants. Revista Panamericana de Salud Pública, 36(5), 306–313. [PubMed] [Google Scholar]
  23. Lazarus, R. , Struthers, H. , & Violari, A. (2013). Promoting safe infant feeding practices—The importance of structural, social and contextual factors in Southern Africa. Journal of the International AIDS Society, 16, 18037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. McInnes, R. J. , & Chambers, J. A. (2008). Supporting breastfeeding mothers: Qualitative synthesis. Journal of Advanced Nursing, 62(4), 407–427. [DOI] [PubMed] [Google Scholar]
  25. Microsoft . (2011). Microsoft Excel. Redmond, WA: Microsoft. [Google Scholar]
  26. Moland, K. M. I. , De Paoli, M. M. , Sellen, D. W. , Van Esterik, P. , Leshabari, S. C. , & Blystad, A. (2010). Breastfeeding and HIV: Experiences from a decade of prevention of postnatal HIV transmission in sub‐Saharan Africa. International Breastfeeding Journal, 5, 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Ministere de la Sante Publique et de la Population . (2008). Normes nationales relatives à l'alimentation du nourrisson du jeune enfant sains et de ceux nés de mères séropositives. Ministere de la Sante Publique et de la Population Unite de Coordination du Programme National d'Alimentation et de Nutrition. Port‐au‐Prince, Haiti: Servi‐Plus. [Google Scholar]
  28. Ministère de la Santé Publique et de la Population . (2011). Nutrition du nourrisson et du jeune enfant: Bonjan fason pou bay yon tibebe manje depi l fèk fèt jouk li rive 24 mwa. Ministere de la Sante Publique et de la Population Unite de Coordination du Programme National d'Alimentation et de Nutrition. Port‐au‐Prince, Haiti: USAID. [Google Scholar]
  29. Orne‐Gliemann, J. , Mukotekwa, T. , Miller, A. , Perez, F. , Glenshaw, M. , Nesara, P. , & Dabis, F. (2006). Community‐based assessment of infant feeding practices within a programme for prevention of mother‐to‐child HIV transmission in rural Zimbabwe. Public Health Nutrition, 9(5), 563–569. [DOI] [PubMed] [Google Scholar]
  30. Otoo, G. E. , Lartey, A. A. , & Perez‐Escamilla, R. (2009). Perceived incentives and barriers to exclusive breastfeeding among periurban Ghanaian women. Journal of Human Lactation, 25(1), 34–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Parrott, R. , & Condit, C. M. (1996). Evaluating women's health messages a resource book. Thousand Oaks, CA: SAGE Publications, Inc. [Google Scholar]
  32. Pelto, P. J. (2013). Applied ethnography: Guidelines for field research. Walnut Creek, CA: Left Coast Press, Inc. [Google Scholar]
  33. Pennec, F. , Wencelius, J. , Garine, E. , Raimon, C. , & Bohbot, H. (2012). FLAME. Paris: CNRS; Available from: http://www.mae.u-paris10.fr/lesc/spip.php?article63 [Google Scholar]
  34. Perez‐Llamas, C. , & Lopez‐Bigas, N. (2011). Gitools: Analysis and visualisation of genomic data using interactive heat‐maps. Plos One, 6(5), e19541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. R CORE TEAM . (2015). R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; Available at: http://www.R-project.org/ [Google Scholar]
  36. Smith, J. J. (1993). Using ANTHROPAC 3.5 and a spreadsheet to compute a free‐list salience index. Field Methods, 5(3), 1–3. [Google Scholar]
  37. Smith, J. J. , & Borgatti, S. P. (1997). Salience counts—And so does accuracy: Correcting and updating a measure for free‐list‐item salience. J Linguist Anthropol, 7(2), 208–209. [Google Scholar]
  38. Sutrop, U. (2001). List task and a cognitive salience index. Field Methods, 13(3), 263–276. [Google Scholar]
  39. Thompson, E. C. , & Juan, Z. (2006). Comparative cultural salience: Measures using free‐list data. Field Methods, 18(4), 398–412. [Google Scholar]
  40. Tuthill, E. L. , Chan, J. , & Butler, L. M. (2015). Challenges faced by healthcare providers offering infant feeding counseling to HIV‐positive women in sub‐Saharan Africa: A review of current research. AIDS Care, 27(1), 17–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. UNICEF . (2012). Key messages booklet: The community infant and young child feeding counselling package. New York, NY: United Nations Children's Fund. [Google Scholar]
  42. Weller, S. C. , & Romney, A. K. (1988). In Miller M. L., Maanen J. V., & Manning P. K. (Eds.), Systematic data collection (Vol. 10 Qualitative research methods, Series). Newbury Park, CA and London: SAGE Publications, Inc. [Google Scholar]
  43. World Health Organization . (2010) Breastfeeding key to saving children's lives. WHO media centre; Geneva: World Health Organization. [Google Scholar]
  44. World Health Organization . (2015) Infant and young child feeding. Periodical Infant and young child feeding. Available at: http://www.who.int/mediacentre/factsheets/fs342/en. [12 December 2015].
  45. WHO , UNAIDS , UNFPA , & UNICEF . (2010). Guidelines on HIV and infant feeding 2010: Principles and recommendations for infant feeding in the context of HIV and a summary of evidence. Geneva: World Health Organization. [PubMed] [Google Scholar]
  46. WHO/UNICEF . (2014). Global nutrition targets 2025: Breastfeeding policy brief (WHO/NMH/NHD/14.7). Geneva: World Health Organization. [Google Scholar]
  47. Williams, C. , Alderson, P. , & Farsides, B. (2002). Is nondirectiveness possible within the context of antenatal screening and testing? Social Science & Medicine, 54(3), 339–347. [DOI] [PubMed] [Google Scholar]
  48. Young, S. L. , Mbuya, M. N. N. , Chantry, C. J. , Geubbels, E. P. , Israel‐Ballard, K. , Cohan, D. , … Latham, M. C. (2011). Current knowledge and future research on infant feeding in the context of HIV: Basic, clinical, behavioral, and programmatic perspectives. Advances in Nutrition, 2(3), 225–243. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zarcadoolas, C. , Pleasant, A. F. , & Greer, D. S. (2006). Advancing health literacy: A framework for understanding and action. San Francisco, CA: Jossey‐Bass. [Google Scholar]

Articles from Maternal & Child Nutrition are provided here courtesy of Wiley

RESOURCES