Can a pragmatic responsive feeding scale be developed and applied globally?

Abstract Responsive feeding (RF) has been recognized as necessary to prevent all forms of malnutrition including stunting and childhood obesity. Specific RF guidelines have been developed, but it is unclear how RF behaviours can be monitored systematically. Therefore, developing valid and reliable abbreviated and pragmatic RF scales is an important global priority. This is challenging, as RF is a construct with multiple dimensions including recognizing and responding to hunger and satiety cues, providing a nurturing environment during feeding episodes, and understanding how feeding needs evolve as a function of the developmental stage of the young child. Further, RF is embedded within the responsive parenting framework that in addition to RF includes sleep, soothing and play routines and the interconnections between them. A recent pioneer study conducted in a rural area of Cambodia validated an 8‐item RF scale through direct feeding observations of 6‐ to 23‐month‐old infants at home, as part of two cross‐sectional surveys conducted before and after a complementary feeding intervention. It is important for similar research to be conducted elsewhere to find out if it is possible or not to develop a core RF scale that is valid and reliable and that has adequate specificity and sensitivity for application in community studies and population surveys globally. As highlighted in this article, different definitions of RF have been used in the field; thus, it is important to reach consensus on a single definition to help move this research area forward.


behaviour, regulation
Responsive feeding (RF) is defined in this commentary as 'feeding practices that encourage the child to eat autonomously and in response to physiological and developmental needs, which may encourage self-regulation in eating and support cognitive, emotional and social development' (Pérez-Escamilla, Segura-Pérez, & Hall Moran, 2019). RF is increasingly being recognized as key to understanding how to help infants and young children develop lifelong healthy dietary habits and to self-regulate food intake (Pérez-Escamilla, Segura-Pérez, & Lott, 2017). In addition, RF has been recognized as necessary to prevent all forms of malnutrition, including stunting, wasting and childhood obesity (Engle & Pelto, 2011;UNICEF, 2020). Therefore, developing valid and reliable RF measures is an important global priority (Bentley, Wasser, & Creed-Kanashiro, 2011;Hurley, Cross, & Hughes, 2011).
The development of RF scales is challenging, as it is a construct with multiple dimensions, including recognizing and responding to hunger and satiety cues, providing a nurturing environment during P E R S P E T I C V E feeding episodes and understanding how feeding needs evolve as a function of the developmental stage of the young child (Pérez-Escamilla et al., 2017). Furthermore, RF falls under the umbrella of the responsive parenting framework that acknowledges that soothing, sleep and play routines are intimately intertwined with feeding routines (Black & Aboud, 2011;Pérez-Escamilla et al., 2017). Each of these dimensions of responsive parenting require valid and reliable scales to measure them (Paul et al., 2014 In their pioneer study, Sall et al. (2020)  how the caregiver spent time while feeding the child, distractions) and three indicators for active feeding (the caregiver talks to the child, verbally encourage him to eat; encourages the child when he is eating well; and motivates the child to eat more using gestures/games or by demonstrating to him how to eat). The fit and reliability of the abbreviated RF scale was confirmed with both baseline and endline data.
The lessons learned from this innovative study include first the importance of using sound RF frameworks and direct observations at different time points to develop RF scales from the ground up and properly test their psychometric validity and reliability. Second, this study highlights the need to recognize that the external validity of this scale cannot be assumed. For example, the items that did not fit well the initial scale in this setting, for example, response to child refusal to eat and child self-feeding, may be important to include in other settings (Jansen, Mallan, Nicholson, & Daniels, 2014 wellbeing of children in the short, medium and long term. One approach could be to establish a multicountry research consortium to assess the fit of the 15 initial items included in the Cambodia study using a standard validation protocol.
Moving forward, and given the complexity of attempting to measure all possible aspects of RF through a pragmatic scale suitable for

Key messages
• A recent study conducted in rural Cambodia validated an 8-item responsive feeding (RF) scale through repeated direct feeding observations of 6 to 23 months old infants.
• Similar research needs to be conducted in other settings to explore developing a valid pragmatic RF scale for use in community studies and population surveys globally.
• It is important to reach consensus on definition of RF to help move the field forward.