Caregivers shape children′s food preferences, consumption and general diet quality through modelling behaviours and the type and quantity of food they make available within the household [1
]. Modelling or observational learning is an important way children learn about food, and this has been demonstrated to shape children′s food preferences and eating practices [2
]. In a recent systematic review, caregiver modelling of food consumption was associated with similar food consumption and food beliefs in children [3
]. Caregivers act as the ‘gatekeepers’ of paediatric nutrition, determining the amount of food to be offered, and developing social norms for the child [4
Children′s meal and snack consumption is influenced by environmental and behavioural factors, such as food availability [5
], the maternal diet [6
], individual differences in eating traits (e.g., satiety responsiveness) [7
] and parental feeding practices (e.g., pressure to eat) [2
]. Caregivers often use specific practices to influence food intake and reinforce the development of eating patterns that they deem appropriate [2
]. Feeding practices fall into two main themes: controlling (e.g., restriction or pressure to eat) or noncontrolling (e.g., provide child autonomy) [9
] and generally are child-centred as an adaptive response to children′s eating traits, food fussiness and specific food problems [10
]. For example, satiety responsiveness, defined as the child′s ability to reduce food intake in response to internal satiety cues [13
], has been inversely related to energy intake [14
]. In contrast, some feeding practices, such as modeling [2
], are successful in promoting healthy consumption [17
] whereas others (e.g., pressure to eat) can reduce desire to eat and actual consumption of a target food [18
Portion size mothers serve their children at an evening meal was found to strongly correlate to the portion size they serve themselves [19
]. However, generally there is a paucity of research exploring the relationship between caregiver portion sizes and child portion sizes in the UK, for both meals and snacks. Snack foods feature significantly in the habitual diet and are reported to contribute around 21% of children′s total daily energy intake in the UK [20
] and USA [21
]. Frequent consumption of large portion sizes of HED foods has been associated with a larger body mass index (BMI) in children aged 1–5 years [22
] and 5–9 years [23
]. Given that larger portion sizes lead to greater immediate energy intakes [24
] and sustained intake over a five days period without compensation in children aged 3–5 years [25
]; a better understanding of the factors that predict snack portion size selection may be useful for developing public health interventions to address child snack intake in the context of healthful eating. Data from the UK suggest that snack foods are offered to children in adult or larger than recommended portion sizes. For example, a UK based survey involving 1000 parents reported that 61% of parents offered their toddlers large portion sizes of jelly sweets (candy), with 24% of parents offering portions of sweets that were the equivalent to three times the recommended weekly amount, in one serving [26
]. Similarly, in Scotland, 29% of parents offered their infants aged 8–12 months high energy dense (HED >2.5 kcal/g as defined by [27
]) snack foods at least once per day, and the frequency of HED snack offerings increased with higher levels of social and economic deprivation [28
Portion size recommendations and daily eating plans have been proposed for preschool children in the UK [29
] and USA [31
], to support caregivers in providing a nutritionally balanced diet to children. In 2015, More and Emmett [32
] proposed evidence-based appropriate portion size ranges for a variety of foods, and a practical, balanced food plan for preschool children by combining published data from two national surveys (National Diet and Nutrition Survey [33
] and Avon Longitudinal Study of Parents and Children [34
]). Foods were allocated into five food groups (1. Bread, rice potatoes; 2. Fruit and vegetables; 3. Milk, yoghurt and cheese; 4. Meat, fish, eggs, nuts and pulses; and 5. Foods high in fat and/or sugar) and two food groups (groups 2 and 5) were split further to provide flexibility in serving frequencies and to reflect snack foods. Despite these guidelines being evidence-based, these recommendations are not easily accessible to the general public, and manufacturers tend not to state portion sizes for children on their products [36
A recent systematic review [37
] revealed that parental portioning practices at meals are influenced by caregiver portion size, perceived child hunger, child body size and caregiver employment status. However, to date, little is known about the associations between children′s snack intake, portion sizes and parental feeding practices. Furthermore, the review by Kairey et al. [37
] comprised only three studies from the UK, of which none focused primarily on snack foods [38
It is not yet clear what determines the snack portion size that caregivers serve to preschool children in the UK. The aims of the current investigation were to compare adult-selected portion sizes for preschool children against the suggested amounts proposed by More and Emmett [32
], and to explore what factors related to child feeding and eating behaviours predict portion size selection of HED and LED snack foods using an online survey.
The aims of the current investigation were to compare adult-selected portion sizes for preschool children against the suggested amounts proposed by More and Emmett [32
] and to explore what factors related to child feeding and eating behaviours predict portion size selection of HED and LED snack foods using an online survey.
Most caregivers identify portion sizes of LED snacks in line with and larger than the suggested amounts for children and themselves. Similarly, the majority of caregivers were likely to serve themselves and their child HED snacks in portion sizes in line with, or smaller than, the suggested amount. However, a proportion of adults were selecting larger than suggested and smaller than suggested portion sizes of HED and LED snacks, respectively. The results of the multinomial logistic regression demonstrate that caregiver portion size, reported child liking, pressure to eat, child satiety responsiveness, caregiver monitoring, child frequency of consumption and child BMI z-score were significant predictors of LED and HED child snack portion size selection.
Overall, caregivers adapted portion sizes for the age and size of their child which were similar to the portion sizes suggested by More and Emmett [32
], thus demonstrating their ability to downsize portion sizes to match their preschool children′s energy requirements. However, 31% and 16% of caregivers selected smaller than recommended portion sizes of LED snacks for themselves and their child, respectively. Almost a third of caregivers (data not shown) selected portion sizes of HED snacks up to four times the recommended amount for adults and children [32
] in one serving, for themselves and their child. These findings are consistent with previous UK survey results [26
] demonstrating that some preschool children are being served large portions sizes of HED snack foods, in many cases the equivalent to three times the weekly recommended amount. Similarly, adults are typically consuming larger amounts of HED snacks than on pack portion size suggestions [67
]. Frequent exposure to large portion sizes is associated with a sustained increase in energy intake over a five day period in children aged 3–5 years [25
] and over 11 days in adults [68
] suggesting that exposure to large portion sizes might contribute towards excess energy intake and, ultimately to a positive energy balance.
Caregiver portion size was positively associated with child snack portion size. This suggests that for all of the included snacks, caregivers tend to judge appropriate portion sizes for their child, in line with their own self-selected portion size. This finding extends previous US-based research that identified a positive association between adult and child portion size at an evening meal [19
]. Positive associations between child and adult portion size may be due to social norms [69
], food availability [5
], parental food liking [19
] or parental hunger [54
]. For example, maternal feelings of hunger influence maternal perceptions of their child′s hunger and thus the amount of food mothers served their children at a buffet style meal, regardless of their child′s actual hunger levels [54
In the present study, increased child food liking was associated with reduced odds of both smaller and larger than recommended HED portion size selections meaning that caregivers are more likely to provide their child with a recommended portion size when the child likes the snack item. Qualitative research suggests that caregivers consider their child′s food preferences and requests when preparing meals by responding to their child′s individual differences [38
]. Some caregivers consider that adjustments to portion size should be made according to nutritional content, such that HED foods should be limited [38
]. Therefore, caregivers may be reluctant to offer large portion sizes of HED snacks, despite them being highly liked, as they choose to prioritize their child′s health and nutritional intake over their child′s food preferences. However, this remains to be further investigated, especially in a more varied sample
Related to child liking is frequency of consumption, since foods that are well-liked by children are generally offered more frequently [71
]. In the present study, frequency of consumption was not a significant predictor of LED snack portion size. However, increased frequency of consumption predicted increased odds of selecting both smaller and larger than recommended portion sizes of HED snack foods. For example, caregivers who report that their child frequently consumes HED snacks might offer this snack in smaller than recommended portions sizes, possibly in an attempt to monitor their child′s snack intake. Conversely, other caregivers are demonstrating more permissive feeding practices and offering frequent and large portion sizes of HED snacks, a strategy which has previously described as a means of controlling preschool children′s behavior [26
Child BMI z-score also predicted larger than recommended portion sizes of HED snacks, so the higher the child BMI z-score the more likely their caregiver was to select a larger than recommended portion size of HED snacks. These findings support previous literature demonstrating a positive association between portion size of HED snacks and BMI [22
]. It is possible that caregivers serve children with a higher BMI z-score larger snack food portion sizes to meet their greater energy needs; however, this warrants further investigation as the direction of causality remains unknown.
Monitoring food intake, a controlling feeding practice, was a significant predictor of smaller than recommended portion sizes of HED snacks. Parental monitoring has been associated with reduced purchases of HED foods [74
] and increased offerings of fruits and vegetables [75
]. Monitoring intake might be a successful strategy to limit overconsumption of HED snacks if not offered frequently. In contrast, pressure to eat was associated with increased odds of selecting large portion sizes of HED and LED snacks, suggesting that caregivers may offer large food portion sizes to promote consumption. Pressure to eat is often demonstrated in circumstances where caregivers want their child to eat a certain type of food (usually fruits and vegetables) or a larger quantity of food [76
]. However, the literature consistently demonstrates counter-productive effects of pressure to eat, whereby children tend to consume less of the target food rather than more [77
Satiety responsiveness, a trait associated with responsiveness to feelings of fullness and good internal self-regulation [79
], was associated with increased odds of selecting smaller than recommended portion sizes of LED snacks, as well as reduced odds of selecting larger than recommended portion sizes of LED snacks. This suggests that parents may be applying a child-centred strategy to portion size for children scoring high on satiety responsiveness. Caregivers learn from past feeding experiences and respond to their child′s appetite to provide portion sizes in line with the quantity they believe their child will accept and consume at meal times [38
]. For example, in a qualitative study exploring the aspirations and challenges of feeding preschool children, mothers stated that they determine mealtime portions by honoring and valuing their child′s food preferences and trusting their child to stop consuming a meal when full [55
]. Similar portioning practices may be apparent when LED snacks are on offer as caregivers may be conscious of food waste from an environmental or financial cost perspective [81
The present study primarily represents maternal portioning practices due to the greater number of female respondents (94%), exemplifying the dominant role female caregivers play in shaping young children′s dietary intake [1
], or alternatively, the increased likelihood of female research participation. Moreover, the sample reflects a white British (87%) and highly educated (72% university degrees) group of mothers. Future work would benefit from recruiting a more heterogeneous sample in order to more fully understand predictors of portion sizes across different groups. The chosen research design allowed for multiple snack foods to be assessed within a single test session, online. The ease of participation increased statistical power meaning the findings could be used to understand the relative importance of variables that influence snack portion size selection.
Despite the advantages, screen-based measures may misinform actual portion size selection and consumption, when a mismatch occurs between expected and actual food properties [82
]. For example, snack food items were removed from their packaging and provided on a plate/bowl, thus observed as 2D objects without exposure to sensory characteristics such as taste, smell and visual cues including packaging which may influence snack food selection and consumption [83
]. Furthermore, an even number of snack food images were presented to reduce a central tendency effect; however, images were displayed in order of size, from smallest to largest, which may have influenced portion size selection towards the middle points of the range. Nevertheless, there was sufficient variability to test differences in portion sizes by a number of child and parent-related predictors. In the current study, manufacturer′s portion sizes were used as the recommended portion sizes for adults. However, Lewis et al. [85
] demonstrated a lack of consistency between portion sizes that are communicated to the public. Future work would benefit from incorporating more than one publicly recommended portion size into the analysis to further validate the current findings.
The present study examined the snack portion size that caregivers select for their young children without addressing possible second servings or snack variety. Research suggests that some mothers choose to offer a small portion size in the first instance, knowing that their child will ask, and thus receive, more [38
]. Therefore, it is possible that the portion sizes selected in this online study may not reflect the entire quantity children receive or consume at one snack occasion. This may be improved in future studies by adapting the scenario description (i.e., telling respondents to consider the full amount of food given). Moreover, whilst BMI was accounted for, the energy needs of participants were not accounted for in the analysis. Similarly, from the current design, we could not establish how the current findings relate to habitual energy intake. These aspects warrant further investigation.
Data on child sex was only collected from 131 participants due to this variable being missing from the early data collection period. Of this smaller sample, 53% of the participant population were female which is a good representation of the UK population, of which 51% are female [86
]. Moreover, a sensitivity analysis was carried out only on those data where sex data was available. Sex was not demonstrated to be a significant predictor in our study population.
Overall, in the current sample of UK-based caregivers, most adapted portion size to the age and size of their child. However, 16% and 31% of the sample selected smaller than recommended portions sizes of LED snacks for their child and themselves, respectively, and 28% selected larger than recommended portion sizes of HED snack foods for themselves and their children. Significant predictors of child portion size selection included: caregiver′s own portion size selection, child characteristics including reported child liking, child satiety responsiveness, child BMI z-score, child frequency of consumption and parental feeding practices such as pressure to eat and caregiver monitoring. These findings suggest that caregivers in this sample used a child-centred approach to portion control, with most offering portions adjusted to child age and size. In the future, interventions could focus on encouraging caregivers to offer more and larger portions of LED nutrient dense foods, such as fruits and vegetables, as snacks, while limiting portion sizes of HED snack foods.