Full Text
Among children under the age of 5 years globally, rates of stunting (22.0%) and wasting (6.7%) remain high with a concurrent steady rise in the rate of obesity (5.7%) (Global Nutrition Report, 2021). This double burden of malnutrition is being observed in many low- and middle-income countries (LMIC) that are experiencing a global shift in the way that individuals meet dietary requirements and experience illness (Popkin et al., 2020). In South Asia, a region home to eight countries, overall rates of economic development are increasing while food environments are changing to accommodate more processed and ultraprocessed food options that consumers increasingly demand (Popkin et al., 2020).
Sri Lanka is no exception: despite its rapid economic growth and improvements in health care delivery in the previous two decades, child malnutrition remains a national public health challenge. The 2016 Demographic Health Survey indicates that 15% of Sri Lankan children under the age of 5 years are wasted while 17% have stunted growth, despite a 66% decrease in stunting prevalence between 1978 and 2016. (Department of Census and Statistics, 2017). While substantial progress has been made to improve the national nutrition situation for vulnerable communities, Sri Lanka is currently not on track to meet the World Health Organization's (WHO) Global Nutrition Targets by 2025 (Global Nutrition Report, 2021).
Infant and young child nutritional status are determined most proximally by infection and dietary intake, the latter of which relies on several underlying factors such as maternal feeding practices. The WHO recommends that infants and young children be fed in a timely, adequate, and safe manner according to their hunger and satiety cues. Doing so is considered to be a responsive feeding style, which has been positively associated with a normal weight status of children under the age of 2 years within high-income contexts (Spill et al., 2019). Conversely, nonresponsive feeding practices, defined as a lack of reciprocity (e.g., food restrictions, pressuring while feeding) between an infant and young child and their caregiver, has been associated with suboptimal weight status among young children of the same age (Spill et al., 2019). Studies conducted specifically in LMIC settings have also provided evidence that responsive feeding is one of several feeding practices important for improving the likelihood of optimal infant and young child nutrition (Bentley et al., 2011; Boucheron et al., 2020; Chowdhury, 2016; Robert et al., 2020).
To that effect, the Sri Lanka national responsive feeding recommendations indicate that caregivers should feed their children responsively, with specific feeding guidance promoted at national and subnational levels of the health system (Ministry of Healthcare & Nutrition, 2007). To date, the research to understand infant and young child feeding in Sri Lanka has primarily relied on caregiver self-reporting (Dharmasoma et al., 2020; Pallewaththa et al., 2019, 2021). While these studies revealed forceful and indulgent feeding styles among caregivers in Sri Lanka, there has been limited data derived from direct meal observations focused on child feeding to inform related policies and guidelines.
Therefore, we conducted a mixed methods study including both caregiver reports and direct observations to understand how and why caregivers feed children aged 6–23 months in urban, rural and estate sector households of Sri Lanka. This manuscript firstly aims to describe the extent to which caregivers followed current national responsive feeding recommendations and secondly to explain factors limiting and enabling those behaviours.
METHODS Study designThis ethnographic substudy was conducted using a four-phase, mixed methods formative research design across six sites located in the rural, urban and estate (i.e., areas primarily growing marketable crops such as tea for export), sectors of Sri Lanka (Bernard, 2011; Creswell, 2008; Dharmadasa & Polkotuwa, 2016). Data were collected between September 2020 and September 2021 by a team of locally-hired data collectors who had postgraduate educations and language proficiencies needed for fieldwork among both Tamil and Sinhalese-speaking community members. The data collection team was trained for 2 weeks before fieldwork and refresher training were conducted before the commencement of each phase.
Data collection methods and sampling procedures Direct observationsLocal health workers helped to purposively recruit households with infants and young children aged 6–23 months (n = 72) from low-income households by age range (e.g., 6–11, 12–23 months) and by sector (urban, rural, estate) (Patton, 2014). Where possible, the same households were observed at two different time points to decrease the reactivity of participants in the presence of an observer (Gittelsohn et al., 1997). Each observation lasted between 1 and 6 h during morning, midday, and afternoon/evening mealtimes. Observers used a semistructured form to continuously record target behaviours using both events (i.e., when behaviours of interest were observed) and time-based (i.e., at minimum, every 10 min during continuous behaviours such as boiling water) approaches (S. Kodish, Aburto, Dibari, et al., 2015; S. Kodish, Aburto, Hambayi, et al., 2015; S. R. Kodish et al., 2019). Each observation captured the timing of key behaviours, the persons responsible for the behaviours (i.e., caregiver, infant/child, or other household members), the type of food or drink consumed, and other behaviours relevant to understanding caregiver feeding styles (e.g., methods of encouragement, responses to food refusals). While recording behaviours, observers sat in nonintrusive areas of each household and limited their interactions with study participants. After each observation, data collectors wrote detailed field notes in response to a series of semistructured questions pertinent to the guiding study aims (Supporting Information: Appendix 1).
InterviewsCommunity leaders (n = 10), mothers of children aged 6–23 months (n = 34) and community members (n = 37) who are known influencers of child feeding were purposively sampled with the help of health workers familiar to those communities (Table 1). Interview participants were purposively sampled based on their role (leader, caregiver, other community members), by sector (urban, rural, estate) and by spoken language (Tamil, Sinhalese). Sample sizes were determined a priori based on the number of interviews expected for reaching saturation of key themes by research question (Guest et al., 2006; Patton, 2014). Interviews were conducted in Tamil or Sinhalese languages using semistructured guides and recorded using digital recorders (Supporting Information: Appendices 2–4)
Table 1 Participant characteristics.
| Method | Participant type | Participant description | n |
| Interviews | Community leaders | Samurdhi officers | 2 |
| Resident officer | 1 | ||
| Grama Niladhari | 3 | ||
| Leader of a community-based organization | 1 | ||
| Religious leaders | 2 | ||
| Government officer | 1 | ||
| Mothers | Nonworking/housewives | 30 | |
| Formally employed | 4 | ||
| Community members known to influence child feeding | Grandmothers | 8 | |
| Mothers support group leaders | 7 | ||
| Early childhood development officer | 2 | ||
| Social worker | 2 | ||
| Estate welfare officer | 1 | ||
| Preschool teacher | 5 | ||
| Ayurvedic Doctor | 1 | ||
| School principal | 1 | ||
| Public Health Midwives | 8 | ||
| Community health volunteer | 1 | ||
| Leader of a community-based organization | 1 | ||
| Direct child observations | Infant and young childrena | 6–11 months | 20 |
| 12–23 months | 52 | ||
Infants and young children who participated in this study were an average of 15.30 (SD 5.70) months of age, children in the 6–11-month age range were an average age of 8.80 (SD 1.96) and children in the 12–23 month age group were an average age of 17.79 (SD 4.57).
Data analysis Direct observationsA simple coding scheme was developed to identify whether and how specific target behaviours described in observational field notes occurred. Guidance by Sall and Kodish was used to develop behavioural codes for capturing both the feeding practices observed (responsive, active, self-feeding, feeding situation) and characteristics of the feeding environment (e.g., kitchen cleanliness) (Sall et al., 2020; S. R. Kodish et al., 2019) (Supporting Information: Appendix 5).
Numerical data were first analysed to determine the proportion of participants carrying out targeted feeding behaviours based on Sri Lanka's national responsive feeding recommendations, specifically: (1) keep regular mealtimes and learn to recognize child's hunger and satiety cues (2) avoid force-feeding (3) feed infants directly and assist older children when they feed themselves (4) talk lovingly to your child during feeding and try different methods of encouragement and praise (5) minimize distractions during meals, and (6) keep a fixed place for feeding (Ministry of Healthcare & Nutrition, 2007).
InterviewsInterview data were translated verbatim from Sinhalese and Tamil into English transcripts during the data collection period. English transcripts were uploaded to Dedoose analytic software for data management and analysis (SocioCultural Research Consultants LLC, 2018). Textual analysis was a stepwise process: Step 1: An initial codebook was developed in Dedoose and included a minimal number of code categories that were chosen based on the guiding research questions. Step 2: Transcripts were read line-by-line and chunks of text were coded. The codes are labels applied to text that help for assigning meaning to the information collected (Miles, 2014). Step 3: Themes and subthemes were identified based on those codes that repeated between and among participant transcripts. Step 4: The codebook was further developed with additional codes and subcodes based on emergent themes from interviews and observations, as well as discussions among the analytic team members who were reviewing the data set (Supporting Information: Appendix 6). Step 5: Salient themes and subthemes were then extracted in line with national responsive feeding recommendations and triangulated among participant types and in consideration of the observational findings. This approach followed a combination of well-established qualitative data analysis procedures (Bernard, 2011; Charmaz, 2014; Miles, 2014). Interview findings presented throughout this manuscript reflect those stemming from shared themes across participant types.
RESULTSThe results from this study are presented in line with six national responsive feeding recommendations in Sri Lanka (Ministry of Healthcare & Nutrition, 2007).
National responsive feeding recommendation 1. Aim to maintain a regular mealtime and recognize your infant and young child's hunger and satiety cues
Regular mealtimesWe observed three regular mealtimes for infants and young children across sectors. Morning meals were served between 6 and 10:30 am; midday meals were served from 12 to 2 pm; and evening feeding occurred between 4:30 and 6:30 pm. During interviews, caregivers reported following regular, yet flexible feeding schedules overall, explaining that the educational materials provided by the Family Health Bureau, often reinforced by Public Health Midwives, helped them ensure mealtime consistency.
Recognizing hunger and satiety cuesCaregivers that reported not following strict meal schedules indicated that their infants or young children did not always want to eat according to their ‘timetable’ and that trying to do so led to increased fussiness. Those caregivers reported relying more heavily on hunger and satiety cues (e.g., child crying) to inform feeding times throughout the day. Twelve feeding cues were identified through interview data (Figure 1).
Enlarge this image.Figure 1. Reported hunger and satiety cues used by caregivers to inform feeding time by frequency of mention during interviews. †Factors were mentioned an equal number of times across interviews.
During observed feeding episodes, nearly all food requests (87.2% [34/39]) made by infants and young children across sectors were met by caregivers, especially among children from 12–23 months of age (87.9% [29/33]). A variety of different child food requests, mostly sugary snacks (e.g., biscuits, cake) and Thriposha (i.e., a locally produced nutritional supplement for children and pregnant & lactating women in Sri Lanka), were observed in estate households. Regardless of sector, children were frequently observed snacking throughout the day, most often requesting store-bought snacks. Snack requests were commonly made in the presence of family members who were consuming those items at that time, as well as when a door-to-door snack cart (e.g., ‘bread vans’ which are typical to Sri Lanka) passed nearby (Table 2).
Table 2 Observed food refusals and forceful feeding behaviours during meals by sector and child age in Sri Lanka.
| National responsive feeding recommendation 1. Aim to maintain a regular mealtime and recognize your infant and young child's hunger and satiety cues | |||
| Sector | Child age (months) | Infant/child requested specific type of fooda | Caregiver provided the requested fooda |
| % | % | ||
| Urban | 6–11 | 100.0% (1/1) | 100.0% (1/1) |
| 12–23 | 88.9% (8/9) | 87.5% (7/8) | |
| Overall | 90.0% (9/10) | 88.9% (8/9) | |
| Rural | 6–11 | 100.0% (4/4) | 75.0% (3/4) |
| 12–23 | 88.0% (22/25) | 90.9% (20/22) | |
| Overall | 89.7% (26/29) | 88.5% (23/26) | |
| Estate | 6–11 | 100.0% (1/1) | 100.0% (1/1) |
| 12–23 | 100.0% (3/3) | 66.7% (2/3) | |
| Overall | 100.0% (4/4) | 75.0% (3/4) | |
| Overall | 6–11 | 100.0% (6/6) | 83.3% (5/6) |
| 12–23 | 89.2% (33/37) | 87.9% (29/33) | |
| Overall | 90.7% (39/43) | 87.2% (34/39) | |
For indicator, ‘Infant/child requested specific type of food’ only n = 43 children were recorded as requesting or not requesting a specific food during the observation. For indicator, ‘caregiver provided the requested food’, the total number of observations was n = 39 as data collectors only recorded when the caregiver did or did not respond to the child's request (i.e., the request did not occur in all observations).
National responsive feeding recommendation 2. Aim to avoid forceful feeding practices
During meal observations, most caregivers (63.9% [39/61]) avoided using forceful feeding practices. Among the one-third (36.1% [22/61]) of caregivers employed forceful feeding practices. We observed a higher proportion in rural households (45.5% [15/33]) than in urban (30.0% [6/20]) and estate (12.5% [1/8]) households. No substantial difference was found comparing infants 6–11 months (33.3% [6/18]) to young children 12–23 months (37.2% [16/43]) (Table 3).
Table 3 Observed feeding behaviours and national responsive feeding recommendations by sector and child age in Sri Lanka.
| Sector | Child Age (months) | National responsive feeding recommendation 2: Aim to avoid forceful feeding practices | National responsive feeding recommendation 3: Aim to feed infants directly and assist older children when they feed themselves | ||||
| Observed food refusals | Subsequently observed forceful feedinga | Infant/child is fed only by caregiver (child does not touch utensil) | Infant/child is mostly fed by caregiver, but sometimes self feeds | Infant/child mostly self feeds, but receives some help | Infant/child fully self-feeds without any help | ||
| % | % | % | % | % | % | ||
| Urban | 6–11 | 88.9% (8/9) | 25.0% (2/8) | 88.8% (8/9) | 0.0% (0/9) | 11.1% (1/9) | 0.0% (0/9) |
| 12–23 | 75.0% (12/16) | 33.3% (4/12) | 31.3% (5/16) | 37.5% (6/16) | 31.3% (5/16) | 0.0% (0/16) | |
| Overall | 80.0% (20/25) | 30.0% (6/20) | 52.0% (13/25) | 24.0% (6/25) | 24.0% (6/25) | 0.0% (0/25) | |
| Rural | 6–11 | 100.0% (8/8) | 50.0% (4/8) | 62.5% (5/8) | 37.5% (3/8) | 0.0% (0/8) | 0.0% (0/8) |
| 12–23 | 86.2% (25/29) | 44.0% (11/25) | 27.6% (8/29) | 44.8% (13/29) | 24.1% (7/29) | 3.4% (1/29) | |
| Overall | 89.2% (33/37) | 45.5% (15/33) | 35.1% (13/37) | 43.2% (16/37) | 18.9% (7/37) | 2.7% (1/37) | |
| Estate | 6–11 | 66.7% (2/3) | 0.0% (0/2) | 66.7% (2/3) | 0.0% (0/3) | 33.3% (1/3) | 0.0% (0/3) |
| 12–23 | 85.7% (6/7) | 16.7% (1/6) | 28.6% (2/7) | 28.6% (2/7) | 28.6% (2/7) | 14.3% (1/7) | |
| Overall | 80.0% (8/10) | 12.5% (1/8) | 40.0% (4/10) | 20.0% (2/10) | 30.0% (3/10) | 10.0% (1/10) | |
| Overall | 6–11 | 90.0% (18/20) | 33.3% (6/18) | 75.0% (15/20) | 15.0% (3/20) | 10.0% (2/20) | 0.0% (0/20) |
| 12–23 | 82.7% (43/52) | 37.2% (16/43) | 28.8% (15/52) | 40.4% (21/52) | 26.9% (14/52) | 3.8% (2/52) | |
| Overall | 84.7% (61/72) | 36.1% (22/61) | 41.6% (30/72) | 33.3% (24/72) | 22.2% (16/72) | 2.8% (2/72) | |
For indicator, ‘Subsequently observed forceful feeding’, the total number of observations was n = 61 as data collectors only recorded when the caregiver did or did not respond to the food refusal with forceful feeding behaviours.
Forceful feeding was most observed during the latter half of meals when food refusals were more frequent and when caregiver encouragement (positive or negative) was no longer working. Themes from interview data help explain forceful feeding practices. First, caregivers expressed good intentions behind forceful feeding, most of whom explained that doing so would help their children ‘maintain adequate weight’. Second, caregivers explained using forceful feeding approaches when they were certain their child was hungry with an empty stomach (i.e., in the child's best interest to eat). If we know that his belly has been almost empty for some time, then we try to force feed him. For example, if he didn't eat his dinner sufficiently, we try to feed him more of his breakfast meal the next morning.
Third, caregivers explained that their infants were typically ‘never hungry’ and thus forceful feeding is the only option to ensure adherence to meal schedules. But nowadays, our kids never feel hungry. We try to feed them on time thinking they are hungry…. Sometimes a child may not be hungry, but since we feel, or think, they are hungry then we try to feed them. It is like mothers are following a timetable for their children nowadays, but their babies don't like it.
National responsive feeding recommendation 3. Aim to feed infants directly and assist older children when they feed themselves
Self-feeding was observed during just 2.8% (2/72) of meals across sectors (Table 3). Among the 52 observed children older than 12 months, only two (3.8%) self-fed, while the majority (96.2% [50/52]) received some level of caregiver help. No infants aged 6–11 months self-fed. Interview participants explained that a focus on optimal feeding style, including the appropriate level of infant/child involvement during feeding episodes, is relatively new for caregivers and health promoters in Sri Lanka. So, things like that [responsive feeding] fails…then again, responsive feeding is a new concept. Because I think, in our culture, not only the mother, but the mother and extended family help. You know, they [all] believe they must feed the child. Like…the adult is [responsible for] feeding the child…is the concept that used to be. But the new concept is the child is an active participant…in our Asian culture, that is a little harder to accept for the older generation. When the child is an active participant, we have to respect the child…the food [that we feed] …if they say no, we have to stop…But I do feel we are working our way through it.
National responsive feeding recommendation 4. Aim to talk lovingly to your child and encourage your child to eat using different forms of encouragement and praise
In line with national responsive feeding recommendations, most caregivers (61.1%, 44/72) used positive verbal communication during meal observations to encourage their children to eat (i.e., ‘good boy’). Overall, positive verbal encouragement was observed more frequently among children within the 6–11-month age range than the 12–23-month age range (Table 4).
Table 4 Positive and negative verbal communication observed during mealtimes with infants and young children by sector.
| National responsive feeding recommendation 4. Aim to talk lovingly to your child and encourage your child to eat using different forms of encouragement and praise | |||
| Child Age (months) | Caregiver talks positively to the infant/child to encourage eatinga | Caregiver talks negatively to the infant/child to encourage eatinga | |
| Sector | % | % | |
| Urban | 6–11 | 88.9% (8/9) | 22.2% (2/9) |
| 12–23 | 62.5% (10/16) | 18.8% (3/16) | |
| Overall | 72.0% (18/25) | 20.0% (5/25) | |
| Rural | 6–11 | 75.0% (6/8) | 25.0% (2/8) |
| 12–23 | 44.8% (13/29) | 37.9% (11/29) | |
| Overall | 51.4% (19/37) | 35.1% (13/37) | |
| Estate | 6–11 | 66.7% (2/3) | 0.0% (0/3) |
| 12–23 | 71.4% (5/7) | 14.3% (1/7) | |
| Overall | 70.0% (7/10) | 10.0% (1/10) | |
| Overall | 6–11 | 80.0% (16/20) | 20.0% (4/20) |
| 12–23 | 53.8% (28/52) | 28.8% (15/52) | |
| Overall | 61.1% (44/72) | 26.4% (19/72) | |
During any single meal episode, both positive and negative communication could have occurred, thus categories are not mutually exclusive.
Approximately one quarter (26.4%, 19/72) of caregivers observed spoke negatively (i.e., ‘we will give your food away if you do not eat’). Negative verbal communication during feeding was more frequently observed in rural (35.1%, 13/37) than urban or estate households. Observations revealed that most caregiver communication with infants and children started positively but regressed when feeding became more difficult. At the beginning of meals, most caregivers used positive communication approaches, such as offering praise. The mother tried to feed [the child] by telling a story and saying [that the] child is a good child.
After initial food refusals, most caregivers utilized innocent distractions (e.g., showing toys) or promised treats (e.g., sugary snacks) to encourage eating. In cases of repeated food refusals, we observed changes in caregiver communication style, with a transition to more negative language that included threats of not finishing food. Then the mother reacted by telling the child to eat properly otherwise he will have to drink some medication, showing a bottle of cough syrup [to the child].
Findings were similar regardless of child age in months (6–11: 20.0% [4/20]; 12–23: (28.8% [15/52]).
National responsive feeding recommendation 5. Aim to minimize distractions during meal
Most caregivers (81.9% [59/72]) used distractions during mealtimes to encourage their children to eat, more frequently observed in urban (88.0% [22/25]) than in rural (81.1% [30/37]) or estate (70.0%, [7/10]) households. Typical distractions used by caregivers while feeding included the provision of toys, turning on the TV, pointing to nearby animals, showing off a mobile phone and wandering around the house together. Community health workers who often conduct home visits supporting maternal and child health corroborated observational findings during interviews. …but practically, mothers feed their children by showing mobile phones or the television. Actually it [responsive feeding] is practiced very rarely…
In interview data, six primary reasons help explain why caregivers use distractions during feeding episodes: (1) fear that the child is not consuming enough food, (2) to avoid being scolded by the Public Health Midwife, (3) distractions are considered a better alternative than forceful feeding, (4) children are just generally difficult to feed, especially once they are able to walk, (5) distractions reduce time spent encouraging feeding during meals, and (6) children get bored of monotonous diets. We have to go behind the children and feed them. I mean we cannot do anything else, right? It's really a nightmare. We need to divert their attention. We have to spend at least forty-five minutes feeding them. Sometimes, we even need an hour.
Adhering to infant and young child feeding guidelines was more difficult for employed mothers in the workforce who reported very limited time availability for optimal feeding and care at home.
National responsive feeding recommendation 6. Aim to keep a fixed place for eating
Most children who were observed in this study did not have a fixed place for eating. Caregivers were observed following children throughout the house during most feeding episodes when they were old enough to walk. We have to run behind them and feed them. For my son, we bought a toy car and a bicycle which he can ride…so each day we put him in the toy vehicle and try to feed him there.
Even when infants and younger children were only able to crawl, caregivers followed them around the house during feeding episodes. In most cases, caregivers did so in an effort to distract their attention from eating by pointing to nearby animals, for instance.
DISCUSSIONThis manuscript aimed to describe the extent to which caregivers followed the national responsive feeding recommendations and the factors limiting and enabling those behaviours. We found that caregivers across Sri Lanka will nearly always respond to their children's food requests, overly control meal episodes by limiting child self-feeding, positively encourage their children to eat, especially in the beginning, and use distraction as a strategy to encourage intake. Doing so was in most cases not a result of limited caregiver feeding knowledge, but instead to help ensure children maintained adequate weight.
For instance, we observed most caregivers fully meeting infant and young child food requests. However, in most cases, those requests were met using unhealthy, store-bought snacks. In 2016, 33.6% of children aged 6–23 months consumed a sugary food in the past 24 h in Sri Lanka, with a similar proportion also found in Nepal and Cambodia (Department of Census and Statistics, 2017; Huffman et al., 2014; Pries et al., 2017). Although explicit child demands for such snack foods were not made in Nepal and Cambodia studies, similar trends may exist as these South Asian countries experience a nutrition transition and changing food environments (Pries et al., 2017). Responding to children's unhealthy snack demands may result in displacement of nutrient-rich food intake, as well as excess intake, but more research is needed in LMIC settings, specifically (Anzman-Frasca et al., 2012; Pries et al., 2019).
Our findings of forceful feeding practices align with those of similar research in Sri Lanka (Agampodi, 2014; Dharmasoma et al., 2020; Pallewaththa et al., 2019). Caregivers in similar South Asian contexts such as Bangladesh have also used force-feeding to encourage eating, as food refusals are often perceived by caregivers to be associated with undernutrition (Affleck & Pelto, 2012; Naila et al., 2018). Children who are undernourished, across contexts including in Sri Lanka, may be more likely to have poor appetites and refuse foods, lending credibility to caregiver concerns during feeding (Abebe et al., 2017; Jayatissa et al., 2012; Moore et al., 2006; Mutoro et al., 2020). Thus, caregivers in our study may have adopted forceful feeding as a protective behaviour. However, forceful feeding is a type of controlling feeding practice, which is positively associated with heightened childhood obesity risk (Cristina Lindsay et al., 2017). Controlling feeding practices may limit an infant and young child's innate ability to self-regulate, thus contributing to obesity risk into childhood and adolescence (Cristina Lindsay et al., 2017; Fries et al., 2017). Maternal control during feeding may also result in heightened picky eating behaviours in combination if present in combination with other environmental and biological factors (Cole et al., 2017).
At 12 months of age, infants and young children are developmentally ready to self-feed with a spoon, their hands, or a cup using both hands (Pérez-Escamilla et al., 2017). We found that more than one-third of children in this age range did not self-feed at all. Our findings align with those in Indonesia, Peru and Nicaragua where self-feeding among children aged 12–23 months was not commonly practiced (Robert et al., 2020). A study from Bangladesh showed that the barriers to child self-feeding included a lack of time to supervise children who are self-feeding, a higher risk of wasting food, a lack of confidence in a child's ability to self-feed, and an increased risk of messiness during mealtime (Affleck & Pelto, 2012). Caregivers from multiple contexts who do allow self-feeding reported that doing so saved their own time needed for other household chores (Affleck & Pelto, 2012; Mwase et al., 2016). Self-feeding, when developmentally appropriate, may help improve food acceptance and increase overall energy intake, a message that may resonate with mothers concerned with undernutrition in this setting (Flesher et al., 2020; Ha et al., 2002; Moding et al., 2020; Pallewaththa et al., 2021).
During observations, we found that most (61.1% [44/72]) caregivers used positive communication to encourage eating throughout entire meal episodes, while a smaller proportion turned to negative communication after repeated child food refusals. A similar feeding study in Kenya also found changes in caregiver communication style from the beginning to the end of meals. (Mwase et al., 2016). Both styles may have differential effects on child diets. Positive verbalization during feeding episodes may more than double the likelihood that children accept bites, while negative verbalization may decrease the likelihood (Dearden et al., 2009). Not only does positive verbalization encourage food intake, but it is recognized as a form of positive stimulation by the Family Health Bureau, Sri Lanka to promote infant and young child development (Family Health Bureau & Ministry of Health, 2014).
Most caregivers (82%) we observed used distractions (e.g., giving toys, showing mobile phones, television, wandering around the house, singing or telling stories) to encourage eating, a finding aligned with that of previous research in Sri Lanka (Dharmasoma et al., 2020; Lucas et al., 2021). Not all distractions may be considered to be negative, however: short, but intentional distractions may be positive if effectively used to refocus a child during feeding (Moore et al., 2006). In our study and others in Sri Lanka, however, distractions were not brief, but employed throughout entire feeding episodes and often resulted in children moving around the house (Lucas et al., 2021; Senevirathne et al., 2015). Ensuring consistent mealtime structures, such as having a fixed location for eating and using minimal distractions, may be associated with less food fussiness during feeding (Finnane et al., 2017).
This study had both strengths and limitations. First, the study was designed to ensure both methodological and participant triangulation to corroborate data sources and perspectives (Patton, 1999). What people say and what people do can often be at odds with one another, especially in health behaviour research (Johnson & Van de Vijver, 2003). Second, the study was designed with multiple iterative phases allowing findings from each phase to be incorporated into subsequent phases of data collection. Such an iterative and emergent design is a primary strength of qualitative research. Despite these strengths, our study was not without limitations. Direct observations were used as the primary method to assess feeding, but when people are observed they may change their behaviours (Gittelsohn et al., 1997). To limit such reactivity, we conducted two rounds of observations among the same households (Gittelsohn et al., 1997). We also did not assess all possible factors that could help explain feeding behaviours reported in this study, for instance, child appetite. Finally, fieldwork was conducted during the COVID-19 pandemic when caregivers may have been facing additional hardships that more negatively affected feeding and care practices.
CONCLUSIONWe found that strict adherence to Sri Lanka's national responsive feeding recommendations was challenging for most caregivers due to real-life factors such as recurring child food refusals during mealtime. Using context-specific findings from this study to tailor social and behaviour change communication materials may aid Public Health Midwives whose job is to facilitate optimal health and nutrition practices during community outreach throughout Sri Lanka.
AUTHOR CONTRIBUTIONSStephen R. Kodish was responsible for study design conceptualization and presentation of findings. Upul Senarath and Dhammica Rowel were responsible for management of fieldwork. Hiranya Jayawickrama, Chithramalee de Silva and Safina Abdulloeva provided technical feedback on data collection forms and aided in interpretation of findings. Olivia Romano was responsible for data analysis and interpretation. Teresa R. Schwendler was responsible for data analysis and writing of this manuscript.
ACKNOWLEDGEMENTSThis research was produced under the leadership of the Ministry of Health, Sri Lanka and in collaboration with the UNICEF Sri Lanka Country Office, South Asia Infant Feeding Research Network (SAIFRN), and The Pennsylvania State University. Without this team, this research would not have been possible. This work was supported by UNICEF's regular resources and global thematic funding for nutrition.
CONFLICT OF INTEREST STATEMENTThe authors declare no conflict of interest.
DATA AVAILABILITY STATEMENTThe data that support the findings of this study are available from the corresponding author upon reasonable request.
ETHICS STATEMENTThis research was approved by both the Pennsylvania State University Institutional Review Board and the Ethics Review Committee at the University of Colombo in Sri Lanka
© 2023. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Rowel, Dhammica 2 ; Safina Abdulloeva 2 ; Jayawickrama, Hiranya 3 ; Chithramalee de Silva 3 ; Romano, Olivia 1 ; Senarath, Upul 4 ; Kodish, Stephen R 1 1 Department of Nutritional Sciences, Pennsylvania State University, State College, Pennsylvania, USA
2 United Nations Children's Fund, Sri Lanka, Colombo, Sri Lanka
3 Family Health Bureau, Ministry of Health, Colombo, Sri Lanka
4 Department of Community Medicine, University of Colombo, Colombo, Sri Lanka