Caregiver–infant's feeding behaviours are associated with energy intake of 9‐11 month‐old infants in rural Ethiopia

Abstract

Inappropriate complementary feeding, both in quantity and quality, is a major determinant of undernutrition. However, little is known about how infant–caregiver's feeding behaviours affect infants' energy intake. Therefore, the objective of this study was to characterize infant–caregiver feeding behaviours and investigate their association with infants' energy intake. The study involved 106 mother–child pairs recruited from seven randomly selected kebeles of Mecha district, West Gojam, Ethiopia. The feeding styles were assessed through observations of 1‐day, in‐home, feeding episodes that were videotaped and coded into self‐feeding, responsive, active, distracting, and social feeding behaviours. Infants' haemoglobin and anthropometric measurements were taken. The association between feeding behaviour scores and energy intake per meal was investigated. The mean food intake of the infants was very low (11.4 ± 7.0 g/kg body weight per meal) compared to the minimum theoretical gastric capacity (30 g/kg body weight per meal). Infants' haemoglobin concentration was negatively associated with energy intake (ρ = 0.178, p = .03). Infants' responsive and active positive feeding styles were positively associated with energy intakes (ρ = 0.258 and 0.432, p = .004 and p < .001, respectively) as well as caregivers' responsive positive feeding styles (ρ = 0.237, p = .007). Both haemoglobin concentrations and feeding styles were associated with infant's energy intake. Anaemia prevention and control measures should be reinforced. Current nutrition education programmes should give emphasis on ways to effectively incorporate culturally adapted responsive feeding messages in this and similar settings.

1. INTRODUCTION

Undernutrition and micronutrient deficiencies are still highly prevalent in developing countries. In children, these deficiencies are associated with poor growth, impaired cognitive development, and poor health status and thus can impede the achievement of the Sustainable Development Goals (Baye, 2017; Black et al., 2008). The overwhelming impact of growth faltering is usually irreversible after the age of 2, thereby leaving a small window of opportunity for intervention (Baye & Faber, 2015; Martorell, Khan, & Schroeder, 1994). In this regard, the role of adequate complementary feeding, both in quantity and quality, is of great importance. Much of the discourse in the area of complementary feeding remained on “what” and “when” a child should be fed and has given little attention to the notion of “how” the child is fed (Stewart, Iannotti, Dewey, Michaelsen, & Onyango, 2013). This is unfortunate because in many low‐income countries, the amount of food consumed by infants and young children is lower than their theoretical gastric capacity estimated at 30 g/kg of body weight (BW) per meal (WHO, 1998), making energy and nutrient requirements even more difficult to meet (Baye, Guyot, Icard‐Verniere, & Mouquet‐Rivier, 2013; Gibson et al., 2009; Mouquet‐Rivier, Traoré, Soma, Kaboré, & Trèche, 2016).

Difficulties in early feeding evoke strong emotions in parents and can undermine parenting confidence and parents' sense of competency (Black & Aboud, 2011). Thus, feeding difficulties must be addressed in a timely manner. By 9 months, most infants start to sit without support, put objects in their mouth, open their mouth when food is provided, and can lean forward when they want to be fed and lean backward to refuse. Besides, infants at this age can control head and neck muscles, turn their head to refuse mouthfuls, or can stop the protrusion reflex when food is put in their mouth (Birch & Doub, 2014). Despite these advantages, the transition to complementary foods being new to mothers can be overwhelming, which can be exacerbated by the infants' inability to talk and the mothers' failure to identify and respond to infants' hunger and satiety cues (signals). Very few studies have investigated the prevailing child feeding style in low‐ and middle‐income countries (Bentley, Wasser, & Creed‐Kanashiro, 2011; Moore, Akhter, & Aboud, 2006).

In Ethiopia, child malnutrition is of public health concern. With 38% of children under 5 years being stunted, 10% wasted, and 29% underweight, the country has one of the highest malnutrition rates in sub‐Saharan Africa (CSA, 2016). The few existing quantitative dietary intake surveys on children have indicated that energy and nutrient intakes were suboptimal (Abebe, Haki, & Baye, 2017b; Baye et al., 2013; Gibson et al., 2009; Mengistu, Moges, Samuel, & Baye, 2017). Although inadequate food intake may partly be related to food insecurity, the high stunting rates in food surplus regions of Ethiopia (i.e., West Gojam) suggest that inadequate feeding styles may also significantly contribute to the high rates of malnutrition (Abebe, Haki, & Baye, 2017a). The few studies that tried to characterize the feeding style behaviours of Ethiopian mothers had a wide age range (Aboud & Alemu, 1995; Wondafrash, Amsalu, & Woldie, 2012). This is unfortunate as feeding styles are dynamic and evolve quickly with the age of the child. Besides, the study by Wondafrash et al. (2012) was based on caregivers' report, which may not be the most suitable design to minimize bias. More importantly, studies that relate feeding behaviours in infancy to food and energy intakes in Ethiopia do not exist yet. Therefore, the present study aimed to characterize the feeding style of infants aged 9–11 months and its association with energy intake in West Gojam, Ethiopia.

Key messages.

• Infants' food intakes were far lower than their theoretical gastric capacity.

• Energy intake was associated with infant's haemoglobin concentration.

• Caregivers' responsive and active feeding behaviours were associated with increased energy intake.

• Feeding behaviours, independent from haemoglobin concentrations, are associated with energy intake.

2. METHODS

2.1. Location of the study

The study was conducted in Mecha district, West Gojam zone, Amhara region. This district is located at 524 km north‐west of Addis Ababa and 34 km south‐east of Bahir Dar, the capital city of Amhara region.

2.2. Study participants

From seven kebeles (smallest administrative unit in Ethiopia) randomly selected from the district, 9‐ to 11‐month‐old infants (n = 150) were identified on the basis of the health centre's record that were completed through a census. The inclusion criteria were the following: the children had to be already introduced to complementary feeding and their families to be permanent residents in the area of the study. Infants affected by severe malnutrition with length for age or weight for age or weight for length z‐score less than −3 based on the WHO (2006) growth standard were excluded. These infants were referred to the closest health centre for follow‐up. Finally, 106 singleton infants and their caregivers were eligible and agreed to participate in the study.

2.3. Ethical aspects

Ethical approval was obtained from the Research Ethics Review Committee of the College of Natural Sciences of Addis Ababa University. Informed consent was obtained from legal guardians in the presence of health extension workers.

Consent forms and questionnaires were translated in the local language (Amharic) prior to the survey.

2.4. Sociodemographic and anthropometric characteristics

The sociodemographic characteristics of the subjects were assessed using a pretested questionnaire. Data on livelihood activities, level of education of caregivers, household assets, and size of land owned were collected.

Length and weight were measured in duplicate using standardized techniques with the children wearing light clothing and no shoes. Z‐scores for length for age, weight for age, and weight for height were calculated using ENA software, on the basis of the WHO multicentre growth reference data (WHO, 2006). All anthropometric measurements were made by the same person to avoid inter‐examiner errors.

2.5. Feeding practices

Information related to feeding practices such as breastfeeding, introduction to complementary foods, and who often feeds the child were collected through a questionnaire. Hunger and satiety signals (cues) and caregivers' reaction to food refusal were recorded.

2.6. Standard complementary food provision and preparation

Differences in food characteristics may significantly affect the energy intake and hence the overall appetite of infants. To better isolate the effect of feeding style and control for confounders such as food characteristics (Rolls, 2012), a standard commercially available complementary food Cerifam™ was used. This commercial complementary food was not accessible in this rural setting. We thus have provided to households the Cerifam™ 1 week prior to the actual survey to allow them to be accustomed to the taste. At the time of the complementary food distribution, the preparation of standardized porridges with appropriate consistency (i.e., thick enough to stay on a spoon and not drip off) was demonstrated to the caregivers. They were also recommended to feed the complementary food to their 9‐ to 11‐month‐old child at least once a day until the next visit. The provision of the Cerifam™ served the key purpose of standardizing the food eaten, which allowed to analyse the effect of feeding style independent of that of the food matrix.

2.7. Feeding event observation

Feeding event was observed early in the morning. The data collectors sat in a nonintrusive position and videotaped the feeding episode (first meal of the day).

The complementary food was prepared by the caregiver following the previously demonstrated standardized preparation procedure. The amounts of flour and water used and the final porridge were weighed to estimate the dry matter (DM) content. The caregivers were asked to feed their child as they usually do. A sufficient amount of porridge was served to allow the caregiver to feed the child ad libitum. No other food was served. The amount of food consumed was measured by weighing the plate before and after the feeding episode. The duration of the feeding episode was recorded.

The videos were analysed by two different assessors, and feeding behaviours of both the caregiver and the child were coded according to the coding system developed by Moore et al. (2006). This includes five behavioural categories: (a) self‐feeding, (b) responsive, (c) active feeding, (d) social behaviour, and (e) distraction, which were further characterized as positive or negative behaviours. A positive behaviour was one that promoted consumption, whereas a negative behaviour interrupted the feeding. A preliminary analysis of 10 videos selected at random was performed jointly by the assessors, and consensus was reached on the definition and coding of every behaviour (Tables  S1 and S2).

The videos were coded at two levels of analysis. The first level of analysis was each smallest meaningful action or utterance that led to the acceptance or refusal of a mouthful. A feeding behaviour was considered to be present if it was observed at least twice during the feeding episode. The number of times each feeding behaviour was observed was then determined. The second level of analysis was the overall feeding episode that was categorized into one feeding behaviour category to reflect the most dominant feeding behaviour based on the empirical judgement of the coder/observer.

2.8. Haemoglobin screening

Haemoglobin was assessed using a portable photometer (Hemocue HB 301). After wearing gloves for protection, the infants' middle or ring finger was cleansed with a disinfectant wipe, and the side of it was pricked by using a lancet. After wiping away the first two to three drops of blood, light pressure was reapplied towards the fingertip, and a drop of blood was collected directly into the testing cuvette, and haemoglobin concentrations were recorded. Measurements were then adjusted for altitude (1,500 m above sea level) by subtracting 0.5 (WHO, 2011).

2.9. Estimation of energy intakes

The energy value of the complementary flour was calculated on the basis of its proximate composition on DM basis using Atwater's conversion factors and taking into account 2 kcal/g for dietary fibre (Charrondiere, Chevassus‐Agnes, Marroni, & Burlingame, 2004), as follows:

Energy (kcal/100 g DM) = (% available carbohydrate × 4) + (% protein × 4) + (% fat × 9) + (% fibre × 2) Energy intake (kcal per meal) was then calculated by multiplying the energy value of the flour by the food intake and the DM content.

2.10. Data analysis

Data were analysed using SPSS version 20. The normality of the data distribution was checked using Shapiro–Wilk test. Spearman correlations were used to investigate associations of feeding behaviour scores, haemoglobin, and infant's anthropometry with energy intake. Partial correlations with and without adjustment for haemoglobin were also performed. Differences were considered statistically significant for p‐values < 0.05.

3. RESULTS

3.1. Sociodemographic and anthropometric characteristics

Almost all mother–child pairs included in this study were from farming households (Table 1). The average land size was less than 1 ha, and this was used to feed an average of six family members. Most mothers were in their twenties, had two to three children, were mainly housewives, but also involved in farming activities. Radio (32%) and TV (9%) ownership was low. A significant proportion of the children suffered from underweight, stunting, and anaemia.

Table 1.

Sociodemographic and anthropometric characteristics of mother–child (9–11 months) pairs ( n = 106) in Mecha district, West Gojam, Ethiopia

Variables	Mean ± SD	Frequency (%)
Sociodemographic characteristics
Farming households		95 (89.0)
Family size	5.7 ± 2.1
Boys : Girls	1.03
Infant's age (months)	10.3 ± 0.8
Mother's age (years)	28.2 ± 6.3
Mother's education (illiterate)		76 (71.7)
Farming land size (ha)	0.81 ± 0.7
Caregivers' occupation (housewife)		83 (78.3)
Owns a radio		34 (32.1)
Owns a TV set		9 (8.5)
Nutritional characteristics
Weight (kg)	7.9 ± 0.9
Length (cm)	70.1 ± 3.4
LAZ	−1.0 ± 1.2
WAZ	−1.1 ± 0.9
WLZ	−0.7 ± 1.0
Stunted		27 (25.4)
Underweight		16 (15.1)
Wasted		6 (5.7)
Haemoglobin (g/dl)a	11.7 ± 1.3
Anaemia		28 (26.4)

Note. TV = television; LAZ = length for age; WAZ = weight for age; WLZ = weight for length.

^aThreshold used for moderate anaemia <11 g/dl anaemic, corrected for altitude (1,500 m) based on (WHO, 2011).

3.2. Self‐reported and observed infant feeding practices

According to the caregivers ‘declarations, all infants but one were still breastfed at the time of the survey (Table 2). but complementary foods were introduced fairly late (mean age > 8 months). Mothers, grandmothers, and sisters were most often responsible for feeding the child; more occasionally fathers and other siblings were also involved. The caregivers identified making noise (45%), crying (43%), and restlessness (15%) as infants' behaviours expressed in response to hunger. In contrast, spitting out food (38.7%), playing (7.0%), clenching teeth (14.5%), and pushing food away (40.6%) were identified as behaviours related to fullness, and mothers' response to child food refusal was to take food away.

Table 2.

Infant feeding practices reported by caregivers in Mecha district, West Gojam, Ethiopia

Variables	Mean ± SD	Frequency (%)
Infant currently breastfed		105 (99.1)
Age of introduction to CF (months)	8.3 ± 0.7
Who feeds the child
Mother		27 (25.5)
Mother and grandmother		38 (35.8)
Mother and sister		35 (33.0)
Others		6 (5.7)
Infant's reaction when hungry
Cries		43 (40.6)
Disturbs		15 (14.2)
Makes noise of anger		48 (45.3)
Infant's reaction when full
Refuses		24 (22.6)
Spits food		17 (16)
Plays		7 (6.6)
Clenches teeth		15 (14.5)
Pushes food away		43 (40.6)
Caregiver's reaction to food refusal
Takes away food		106 (100)

During the feeding observation, most infants were fed by their mothers (Table 3). An average 90 g of complementary food was consumed per meal, which represented an average of 11 g/kg BW per meal, 51 kcal, and 17% of the estimated needs from complementary foods. More than 85% had a food intake less than 10 g/kg BW per meal. Food intakes were highly variable due to interindividual variation as frequently observed in similar studies. Besides, a significant proportion of caregivers provided breastmilk during complementary feeding.

Table 3.

Infant's feeding characteristics during meal observation ( n = 106)

	Mean ± SD	Frequency (%)
Who fed the child
Mother		83 (78.3)
Sister		9 (8.5)
Others		14 (12.5)
DM content of porridge (g DM/100 g)	15.0 ± 2.5
Duration of the feeding episode (min)	12.5 ± 4.5
Food intake (g/meal)	89.9 ± 55.1
Food intake (g/kg BW/meal)	11.4 ± 7.0
[0–10[		93 (87.7)
[10–20[		10 (9.4)
[20–30[		2 (1.9)
≥30		1 (0.9)
Energy intakea (kcal/meal)	51 ± 35
Mean contribution to estimated energy needsb (%)	17.0
Breastfeeding during complementary feeding		24 (22.6)

Note. DM = dry matter; BW = body weight; CF = complementary foods.

^aCalculated as DM content multiplied by the energy value of the complementary flour (370 kcal/100 g DM);

^bEstimated energy needs from CFs for 9‐11 month‐old children = 300 kcal/day (Dewey & Brown, 2003); for any given [x, y[,the inversed bracket refers that it is not inclusive of y.

3.3. Caregiver–infant feeding behaviours

Both active positive and negative feeding behaviours were dominant among caregivers (Table 4). Caregivers showed more negative self‐feeding (50%) and responsive feeding (59%) behaviours than positive forms of these behaviours. Overall, the general appreciation of the observation suggests that controlling feeding style, followed by laissez‐faire, and active negative were dominant (Figure 1). Feeding behaviours were not associated with family size and maternal age (P > .05). Infants showed more positive self‐feeding, responsive, and active feeding than negative feeding behaviours.

Table 4.

Feeding behaviours of caregivers and infants ( n = 106) during a test meal feeding episode scored when observed at least two times

Feeding behaviour	Median scoresa (% of caregivers/infants' exhibiting the feeding behaviour)
	Caregiver		Infant
	Positive	Negative	Positive	Negative
Self‐feeding	0.0 (14)	0.0 (50)	4.0 (54)	0.0 (14)
Responsive	3.0 (35)	2.0 (59)	4.0 (48)	0.0 (32)
Active	3.0 (67)	1.0 (75)	3.0 (39)	1.0 (24)
Social	2.0 (66)	0.0 (20)	2.0 (40)	0.0 (20)
Distractive	1.0 (5)	0.0 (37)	0.0 (11)	0.0 (19)

^aRefers to the median of the number of times (frequency) the feeding behaviour was observed during the feeding episode. The unit of analysis was each smallest meaningful action or utterance that led to the acceptance or refusal of a mouthful.

Figure 1.

General appreciation of caregivers' feeding style based on observation of a feeding episode

3.4. Factors associated with energy intake from test meal

Anthropometric indices were not associated with energy intakes, but haemoglobin concentration was (ρ = 0.178, P < .03; Table 5). Among caregivers' feeding behaviours, negative responsive feeding and distraction positive were inversely associated with energy intake. In contrast, positive responsive feeding behaviours were associated with increased energy intake from the test meal. Results for distraction either positive or negative were inconsistent, maybe due to the difficulty in classifying them into factors that promoted or prevented food intake. Sociodemographic variables were not associated with energy intake from test meal.

Table 5.

Factors associated with energy intakes from the test meal

Factors	Energy intake (kcal/test meal)
	Correlation coefficienta	p‐value
Socio‐demographic characteristics
Caregivers' age (years)	.025	.40
Land size (ha)	.133	.09
Family size	.022	.41
Child's characteristics
Age (months)	.132	.09
LAZ	.005	.60
WAZ	−.008	.94
WLZ	−.003	.94
Haemoglobin (g/dl)	.178	.03*
Mother's feeding behaviour score
Self‐feeding positive	.052	.30
Self‐feeding negative	−.199	.11
Responsive positive	.237	.007**
Responsive negative	−.179	.03*
Active positive	−.003	.49
Active negative	−.088	.19
Social behaviour positive	.016	.44
Social behaviour negative	.055	.29
Distraction positive	−.199	.02*
Distraction negative	.021	.42
Overall positive scores	.056	.28
Overall negative scores	−.131	.09
Infant's feeding behaviour score
Self‐feeding positive	.019	.43
Self‐feeding negative	−.271	.002**
Responsive positive	.258	.004**
Responsive negative	.100	.16
Active positive	.432	<.001**
Active negative	−.542	<.001**
Social behaviour positive	.291	.001**
Social behaviour negative	−.155	.06
Distraction positive	.097	.16
Distraction negative	.095	.17
Overall positive scores	.137	.08
Overall negative scores	−.220	.01*

Note. LAZ = length for age; WAZ = weight for age; WLZ = weight for length.

^aBivariate correlation: Spearman correlation coefficient (one‐tailed) with energy intakes (kcal) from test meal as a response variable.

Correlation is statistically significant:

^* p < .05 level;

^** p < .01 level.

Among infant feeding behaviours, overall negative scores and particularly negative behaviours of self‐feeding and active feeding were significantly associated with lower energy intakes. Conversely, positive behaviours of responsive, active, and social feeding behaviours were associated with increased energy intake from test meal.

4. DISCUSSION

The present study characterized the caregiver–infant feeding behaviours and investigated its association with infants' energy intake. Mothers were the most responsible for feeding the infants. Continued breastfeeding was the norm, but complementary feeding started late. A quarter of the infants were anaemic and stunted. Caregiver–infant's feeding behaviours and infants' haemoglobin concentration were associated with the energy intake. Maternal responsive positive feeding behaviour score was positively associated with the energy intake even after controlling for haemoglobin concentrations.

The meal observations conducted in the present study revealed that infants' food intake in the study area was very low. Although based on few earlier studies, the minimal gastric capacity of infants is estimated to be 30 g/kg BW per meal (PAHO/WHO, 2003; WHO, 1998). Relative to this figure, most if not all, had food intakes that were inadequate. This is in line with previous studies that also reported low food intakes among infants and young children in Ethiopia and other developing countries (Baye et al., 2013; Mouquet‐Rivier et al., 2016). A national food consumption survey also reported the low energy intake of Ethiopian infants and young children (EPHI, 2013). Such low food intake from complementary foods resulted in low energy and micronutrient intakes, which, along with the frequent infections, may explain the high prevalence of stunting and anaemia among these infants (Danaei et al., 2016). No association between feeding styles and anthropometric indices was found in this study, but a recent study in the same district revealed associations between stunting and feeding behaviours among older children (12–23 months; Abebe et al., 2017a). This suggests that the consequences (e.g., stunting) of inappropriate feeding behaviours may take time to display, but longitudinal studies are needed to establish causality.

In addition, anaemia can also contribute to poor food intake. Indeed, earlier studies have shown that treatment of anaemia or iron deficiency is associated with improved appetite and food intake (Stoltzfus et al., 2004). The infants did not have access to iron supplements or iron‐fortified foods. Unfortunately, we did not collect information about the iron status of the mothers nor about factors such as iron supplementation during pregnancy or lactation that may influence the iron status of the mothers/infants and their feeding behaviours. Given the limited access to iron supplements in this rural setting, we expect that the proportion of mothers taking supplements, if any, will be very low and limited to pregnancy. Iron is a type II nutrient, and thus, maternal iron status will have little effect on breastmilk iron composition (WHO, 1998). Besides, past the age of 6 months, only 2% of the infants' requirement is expected to be covered from breastmilk (FAO/WHO, 2004). From the present study, it is impossible to exactly know if the low energy intake is due to iron deficiency/anaemia or vice versa. To what extent feeding behaviours are also mediated by poor maternal nutritional status remains unknown; but, like in previous studies (Aboud & Akhter, 2011; Bentley et al., 2011; Vazir et al., 2013), caregivers' negative feeding practices such as active and responsive negative behaviours were common and were associated with lower energy intakes. Such nonresponsive feeding practices, including negative self‐feeding behaviours of caregivers, may have long‐term consequences on appetite regulation by impairing recognition of hunger and satiety cues (Brown & Lee, 2015).

In contrast, caregivers' positive responsive feeding scores were associated with increased energy intake even after controlling for haemoglobin concentrations. This is in line with the report of a cluster randomized controlled trial from India that showed higher energy and nutrient intakes among toddlers that received training on responsive feeding and optimal breast and complementary feeding compared to those that received training on optimal breast and complementary feeding alone (Vazir et al., 2013). Similarly, Mouquet‐Rivier et al. (2016) showed that encouraging feeding increased energy intakes among infants and young children in Burkina Faso. Despite the positive effects that responsive feeding may have, a recent study from Mecha district (present study site) indicated that health extension workers knew little or gave little attention to responsive feeding (Abebe, Haki, & Baye, 2016).

Most infants were fed while seating on the lap of the mother, a position that hindered interactions such as eye‐to‐eye contact that favour responsiveness. Another practice that may have negative effect on food intake is the intermittent offering of breastmilk during complementary feeding. This habit can divert the infant from the offered food, as the taste of breastmilk is often preferred. Thus, it could be beneficial to recommend mothers to separate the timing of breastfeeding from that of complementary feeding. On the other hand, the observation of responsive feeding practices among some caregivers suggests the possibility of designing interventions that promote them.

The present study had several limitations that need to be considered when interpreting the findings. First, the study relied on a narrow age range, and thus, the findings should not be extrapolated to children outside 9–11 months of age, particularly because feeding behaviours in early childhood is dynamic. Nevertheless, focusing on an age range in which the child is unable to talk thus requiring more engagement and responsiveness from the mother increases the programmatic implication of the present study. The observation of only one meal per infant may not be adequate to classify the feeding style adopted at the individual level (mother/child) but is enough to identify the dominant feeding styles of the study population and investigate its association with infants' energy intake. Several factors could have been associated with poor feeding style (e.g., maternal depression or nutrition), but because of the cross‐sectional design of the study, it is difficult to relate the observed feeding behaviours to the various possible causes. Although who feeds the child may have an effect on feeding behaviour and energy intake, our study was not adequately powered to provide statistically valid conclusions. The videotaping of the meal observations is a major strength of this study, as this allowed more settled observations of each mouthful and thus minimized missing feeding behaviours from being coded. However, this was also a source of limitation because it was impossible to avoid self‐consciousness among some caregivers, which might have affected the feeding observation.

Notwithstanding the above limitations, the findings of the present study identified that inadequate complementary feeding style including controlling and negative feeding behaviours are prevalent in this setting. Infants' haemoglobin concentrations are negatively associated with energy intake, whereas caregivers' responsive feeding behaviour scores were positively associated. These findings indicate that along with anaemia prevention and control measures, nutrition programmes that aim to prevent undernutrition should integrate the promotion of culturally adapted responsive feeding messages. Such responsive feeding messages could include the promotion of eye‐to‐eye contact, allowing the child to self‐feed and avoiding distraction during feeding. Studies that evaluate the cultural acceptability of such messages and test their effectiveness in improving the child's appetite are needed.

CONFLICTS OF INTEREST

The authors declare that they have no conflicts of interest.

CONTRIBUTIONS

KB and CMR were involved in developing the study design. AT coordinated and supervised the fieldwork. AT, CMR, and KB analysed and interpreted the data. KB wrote the first draft of the manuscript. All the authors contributed to the manuscript preparation.

Supporting information

Table S1. Definition and classification of the main caregivers' feeding behaviors encountered in the study^a

Table S2. Definition and classification of the main child's feeding behaviors encountered in the study^a

Baye K, Tariku A, Mouquet‐Rivier C. Caregiver–infant's feeding behaviours are associated with energy intake of 9‐11 month‐old infants in rural Ethiopia. Matern Child Nutr. 2018;14:e12487 10.1111/mcn.12487