Malnutrition is the largest risk factor in the world for disability and premature mortality among young children, especially in developing countries [1, 2]. Although the condition is entirely preventable, malnutrition is a significant underlying factor in more than half of the deaths of young children in these countries [1, 3]. Malnutrition is a background factor for deaths from diarrhea, measles, acute respiratory infection, meningitis and malaria [4]. In Ethiopia, 47%, 11% and 38% of children under five years of age are stunted, wasted and underweight, respectively [5]. An Ethiopian child is 30 times more likely to die by his or her fifth birthday than a child in Western Europe and the most common cause of child death is the interacting combination of malnutrition and infection [6].

Malnutrition is a complex phenomenon that stems from various underlying determinants, including a lack of optimal feeding practices for infants and young children. In UNICEF's conceptual framework for determinants of nutritional status, maternal and child care practices have been given due attention in addition to sufficient food supply at the household level, access to health services and a clean environment [7]. Some children in low-income countries with high rates of malnutrition grow normally due to better education and household management, or coping skills of their mothers [8]. Moreover, proper feeding practices, which ensure intake, are as important as the provision of complementary foods that meet nutritional requirements.

The caregiver-child interaction (feeding style) critically influences dietary intake on top of the dietary aspect of child feeding [9]. However, comparing feeding styles between studies prior to 1995 was difficult due to a lack of consistency in terminology and unclear definitions of feeding styles [10]. Later on, feeding styles were conceptualized and defined by Birch and Fisher as controlling, laissez-faire, and responsive feeding styles [11], and have been used in different researches [9, 12, 13].

The "controlling" feeding style occurs when a caregiver has complete control of when, what and how much the child eats, and includes restriction and force-feeding the child. In the "laissez-faire" feeding style the caregiver makes little effort to encourage eating; feeding is not encouraged even when the child may be marginally nourished [14]. Birch and Fisher defined the term "responsive" feeding as a condition in which "the caregiver is in close proximity to the young child during the meal and responds to the child's hunger cues in a reasonable time" [11]. Responsiveness is now viewed by different authors as a three-step process, namely observation of child cues by the caregiver, interpretation of the signs and taking action as quickly as possible in response to those cues [8, 13, 15].

Appropriate child feeding practices and behaviors of parents have a positive effect on growth of infants and young children [16]. For instance, an analysis of data sets from several Latin American countries demonstrated that appropriate breastfeeding and complementary feeding practices were positively associated with child height-for-age in most of the countries studied [17]. However, the majority of the literature on child feeding and parental practices are based on women in affluent societies and their emphases have been on child overweight or obesity [18–20]. Control over feeding appears to vary among cultures, socioeconomic status and child's gender [21]. Hence, results of the effect of different parental or caregiver feeding behaviors among various socio-cultural settings should be used cautiously.

Most studies carried out in Ethiopia have focused mainly on identifying factors associated with undernutrition instead of overweight or obesity [5, 22]. To the best of our knowledge, no studies have been conducted to assess child feeding behavior of caregivers and the association between these behaviors and child growth and development, dietary intake, eating behavior or morbidity in Ethiopia. Current interventions mainly address the issue of what to feed young children instead of how to feed them [8]. The aim of this study was to identify the different types of child feeding behaviors (styles) of caregivers and the factors that influence these behaviors.

This study examined three types of caregivers' feeding styles (laissez-fair, controlling, responsive) from a socio-demographic, child feeding and caregivers' behavior perspective. The majority of the studies investigating these three child feeding styles were carried out in developed countries and used a variety of techniques and instruments. This made direct comparisons of the findings from this study with other findings difficult.

In this study, 51% of the children were exclusively breastfed at 6 months of age. This is much higher than the findings reported in the Ethiopian Demographic and Health Survey 2005 (EDHS 2005). The EDHS 2005 report showed that only one in three children of 4-5 months of age were exclusively breastfed and that the rate of exclusive breastfeeding for those less than 6 months was 49% [5]. This might be because EDHS 2005 was a nationally representative survey with a wide range of exclusive breastfeeding proportions in different regions of Ethiopia.

The WHO recommends children of 6-8, 9-11 and 12-23 months of age to be fed 2-3, 3-4 and 3-4 times per day, respectively, with the addition of 1-2 snacks [9]. In this study, 95% of the children within each of these respective age groups received complementary foods as frequent or more than the WHO recommendation. The mean (± SD) complementary feeding frequency per day was 4.6 ± 2.0, 5.2 ± 1.8 and 5.2 ± 2.0 times for children in the respective age groups. This result was comparable to the findings from rural Malawi where the mean (± SD) complementary feeding frequency was 3.1 ± 1.2, 4.1 ± 1.1 and 4.4 ± 1.4 times, respectively [26]. The majority of caregivers were classified as practicing a responsive child feeding style (75.7%) which is consistent with the findings from a study in Vietnam [12].

In this study, 80.6% of caregivers were biological mothers. A study conducted in Nicaragua showed that mothers, as caregivers, were more likely than other caregivers to offer encouragement for eating [27]. Most of the caregivers in this study were rural dwellers but practiced a responsive feeding style most of the time. Two thirds of rural mothers in Bangladesh practiced active feeding ("behavior that encourages the child to eat or the mother to feed, either indirectly through words or directly through forcing food into the child's mouth") but only a third of them were responsive [28].

Caregivers' behavior in child feeding are part of the broader socio-cultural context of societies in different geographical locations [29]. Several studies suggested that appropriate child feeding practices have an impact on child nutritional status [10, 15–17, 20, 30]. Responsive or active feeding was found to be more beneficial to the child in some low- or middle-income countries [15, 27, 31, 32] and non-responsive feeding was associated with child BMI and overweight or obesity in high-income countries [20] though, cross-study definitions of responsive feeding were inconsistent. Responsiveness was also observed in reaction to a child's decreased appetite during diarrheal episodes in Guatemala [33] and Peru [34]; and low interest in food in Nicaragua [27]. However, recent reviews concluded that the effect of responsive feeding on child weight or growth was mixed and failed to isolate the effect of responsive feeding, as most of these studies were done together with broader health or nutrition education and food supplementation programs [15] and used cross-sectional designs [20]. Moreover, the mechanism of linking feeding behavior and child weight or growth differs between low- and middle-income and high-income countries [15, 20].

In developing countries, there is very little information on the practice of controlling or laissez-faire feeding and implications for child nutrition. In Ghana, it was found that using force during feeding was common among caregivers with negative deviance in child feeding and had a direct effect on child's nutrient intake [32]. In Nigeria, force-feeding and pressurizing was followed by food refusal by the child [35]. In Vietnam, non-responsive feeding, characterized by pressure or force-feeding, was associated with the child's rejection of the food [36]. A recent study in Malawi revealed that controlling feeding behavior was negatively associated with the acceptance of food by the child [37].

Unlike the present study, a laissez-faire feeding style was the most frequently observed behavior among caregivers in Peru and Guatemala [33, 34]. Encouraging feeding in these circumstances was a response only to a decrease in child food intake because of illness and hence was not pro-active.

This study did not assess the outcome of a particular feeding style practiced by caregivers, though the available literature on feeding styles underlined the association of feeding style with child growth and development, acceptance of food or child eating behavior and overweight or obesity [14, 30, 38–41]. As a limitation of the cross-sectional study design, temporality could not be established. Hence, it is difficult to determine whether a positive child response to the provided food leads to responsive feeding by caregivers or vice versa [15]. Nevertheless, behavior at any point in time could reflect prior caregiver-child interactions. The results of this study are also dependent on the reporting of the caregiver. Most studies used additional observations through videotaping of caregivers feeding behaviors which can increase the validity of the finding from interview-based studies [38].

Responsive feeding was the commonest style practiced by the studied caregivers. In this study, rural residence and being a biological mother were found to be independent predictors of responsive feeding after controlling for potential confounders. As this study was the first of its kind in the country, more research is needed to check the validity and reliability of the Caregiver's Feeding Styles Questionnaire in Ethiopia and to evaluate the effect of caregivers' feeding style on child dietary intake and nutritional status in different socio-cultural settings within the country. Moreover, longitudinal research starting from early infancy is recommended to see the factors affecting caregiver's child feeding behavior and other care practices.