The dramatic rise in overweight and obesity in children in the Western world has led to its identification as a major public health concern . Overweight and obesity in children is associated with adverse health outcomes [2, 3], and persists into adulthood . Specific early parental (especially maternal) feeding practices have been proposed as important contributors to the development of obesity and obesogenic eating patterns (e.g., overeating) in children . However, to date little prospective data exist regarding these relationships. Furthermore, the majority of existing research has focused on children over 5 years old and few investigations exist in younger children at the age when eating behaviors develop [6, 7].
Parental feeding practices aim to influence the amount or type of food a child eats and include monitoring intake, restrictive or controlled feeding, pressure to eat, and instrumental and emotional feeding. These practices, although well-intended, can be associated with both child weight and obesogenic eating behaviors  – that is eating behaviors associated with higher calorie intake and overweight, including eating in absence of hunger, lack of caloric compensation, and a fast eating rate . As feeding practices are potentially modifiable risk factors, understanding relationships between parent feeding practices and child weight and eating behaviors is of importance. However, to date, there has been a lack of clearly defined constructs describing maternal feeding practices, as existing measures present considerable overlap. We aim to clarify a set of core independent constructs representing maternal feeding styles and to identify which components most influence child body mass index (BMI) and eating behaviour.
Previous studies focusing on relationships between maternal feeding practices and child outcomes have produced inconsistent findings and focused on disparate concepts. Maternal monitoring, that is keeping track of one’s child’s eating, could be expected to result in healthy weight and eating behaviors, particularly at a very young age when mothers still strongly direct their child’s eating. While self-reported parent monitoring has been shown in one study to prospectively predict lower BMI , more often studies have found no relationship with weight change [8–10].
Restrictive feeding practices involve regulating the type and amounts of foods eaten by children. Although parents may attempt to restrict their children’s eating to reduce weight and maintain health, theorists suggest restriction may lead to the opposite effect by encouraging children to seek out restricted foods and failing to help children regulate eating based on satiety . Longitudinal studies have provided mixed results. Consistent with theorists’ expectations, maternal restrictive practices have predicted increased BMI z-scores from 5 to 7 years , and greater eating when not hungry . However, consistent with likely parental intentions, maternal restriction has also predicted decreased BMI [10, 12]. Finally, other authors have not found maternal restrictive feeding practices predicted changes in child adiposity [8, 9]. One potential explanation for these disparate findings is that different scales have been used to operationalize restrictive feeding. To address this, the present study included five frequently used measures of maternal feeding practices to uncover separate dimensions of maternal feeding before exploring their relationships with child obesogenic eating behaviors and weight change in a sample of younger children.
Parental feeding practices aiming to control when, where and what children eat have also been examined. A distinction has been proposed between overt and covert methods of controlled feeding and their respective influence on child weight and eating behaviors . Overt control refers to explicit control over food consumption, such as being firm about what a child should eat. In one of the few prospective studies of very young children, overt control predicted increased BMI in children aged 12 months . Covert control refers to the extent to which parents manage children’s food environments and restrict access to unhealthy foods, such as avoiding keeping snack foods in the house. Maternal covert control has been cross-sectionally associated with unhealthy snacking in children [13, 15]. However, longitudinal explorations of overt and covert controlled feeding practices in relation to child weight status or obesogenic eating behaviors are lacking.
Another feeding practice, pushing to eat, might be expected to predict weight gain over time, although this has not been observed consistently. Among children from birth to 2 years old, as well as in a sample of 5 year olds, maternal pressure to eat has been found to predict lower BMI one year later [7, 10], increases in child BMI , or to reveal no association with weight change or eating outcomes [8, 9]. Maternal pushing to eat has also been associated with greater fat intake, which in turn was associated with increased BMI . Prompting and encouragement to eat more or try new foods has also been considered in relation to child weight status and obesogenic eating behaviors . However, the little existing cross-sectional research has shown an inconsistent association between encouragement to eat and child BMI [19, 20], and further research is warranted.
Feeding practices which promote the use of food for reward or emotional regulation have also been postulated to influence child weight status and obesogenic eating behaviors. One such practice is instrumental feeding which involves using food as a reward for a desired outcome (e.g., in return for good behaviour). It has been suggested that reinforcing positive associations of palatable foods through instrumental feeding could strengthen the preference for reward (usually high-calorie) foods . In one cross-sectional study, maternal instrumental feeding was associated positively with children’s snacking behaviors . Furthermore, associations between maternal instrumental feeding and food enjoyment and food responsiveness in 3 to 6 year-olds have been found  but no significant association was found between instrumental feeding and child BMI z-scores in the only cross-sectional study that examined this relationship . Although data is lacking, emotional feeding, i.e., using food to help a child regulate emotions, has also been posited to be associated positively with child weight status and obesogenic eating behaviors as a result of encouraging children to eat in the absence of hunger [19, 20, 22].
In summary, the existing literature examining parental feeding practices as predictors of changes in child weight and obesogenic eating behaviors is inconsistent and lacks longitudinal studies in very young children. Measurement issues may partially account for these discrepancies. Furthermore, most studies have focused solely on BMI as an outcome, rather than considering child eating behaviors that potentially contribute to overweight.
The principle aim of the present study was therefore to explore the role of a wide range of maternal feeding practices in relation to weight gain and obesogenic eating behaviors in a sample of 2 year-old children. To date, a variety of potentially overlapping constructs exist in the literature. Consequently, a preliminary aim was to conduct a principal components analysis to clarify a set of core independent constructs representing maternal feeding styles. Next, we aimed to identify which components of mothers’ feeding practices prospectively predicted the development of obesogenic child eating behaviours and BMI. We hypothesized that restrictive and controlling maternal feeding practices would be associated positively with increases in child BMI z-score over time, and we expected these maternal feeding practices to be associated positively, both cross-sectionally and prospectively, with obesogenic child eating behaviors. In order to test the competing hypothesis by which child eating behaviors predict changes in maternal feeding practices, we also examined these relationships prospectively.