It has been contended that more healthful food consumption patterns are likely to be established if children are given the responsibility of deciding the types and amounts of foods they wish to eat from among the foods items being served . Theoretically, this approach to meal service allows children to follow their internal cues to match food consumption with hunger level. Also, there is a great deal of scientific evidence suggesting that pressuring children to eat certain foods or eat all the food on their plate may be counterproductive [13–16]. It could be argued, however, that allowing self-selection of foods and food amounts in a preschool program setting will result in poor food choices by children because of their innate preference for sweet and salty foods combined with a home food environment that provides insufficient exposure to food items such as fruits and vegetables. Provider portioned meals, which involves portioning a specific quantity of all menu items on each childs plate rather than allowing the child to self-serve food items, ensures that children consistently receive controlled quantities of fruits and vegetables along with more energy dense food items. Thus, it is possible that this approach may be more conducive to promoting fruit and vegetable intake and concomitantly moderating energy intake.
Results reported in this paper are consistent with the theory that self-selection may help children match food consumption with hunger level, as energy intake was found to be lower when meals were served traditional family style in comparison to being served provider portioned. Inconsistent with the notion that provider portioning may promote consumption of foods with less preferred flavors, consumption of fruits and vegetables were lower when meals were served provider portioned. To our knowledge no previous studies have directly compared traditional family style and provider portioned meal service approaches with regard to influence on food and nutrient intake of children. Thus, findings from this study may not be corroborated.
Traditional family style meal service may have benefits beyond nutrition. Manipulating bowls and serving utensils as is required with traditional family style meal service may help preschool aged children develop fine motor skills and eye-hand perception [17–19]. Also, the act of passing food around the meal table to peers may be helpful in social development. These potential benefits are among the reasons family-style meal service is one of the feeding guidelines recommended by the National Association for the Education of Young Children , the Head Start Program , the American Dietetic Association , and the American Academy of Pediatrics in collaboration with the American Public Health Association .
Study results suggest that a two course traditional family style meal service approach in which fruits and vegetables are served in advance of other menu items may have nutritional benefits. To summarize findings, fruit consumption and intake of some nutrients found in fruits (vitamin A and folate) were found to be higher when fruits and vegetables were served first rather than in tandem with other menu items. Although the magnitude of the effect on fruit intake was modest (25% higher intake of fruit during the fruit and vegetable first in comparison to control condition), the potential cumulative effect of implementing this meal service approach across multiple meals per day and multiple days per week may be considerable. In addition, serving fruits and vegetables in advance of other menu items is a low-cost, easy to implement strategy that may be readily adapted in preschool settings.
Vegetable intake was not found to be significantly higher during the fruit and vegetable first in comparison to control condition. This finding may be due to the fact that vegetables, in minimally processed form, generally do not contain favored flavors (they are neither sweet nor salty flavored) and some contain bitter tastes. As a result, increasing their consumption may require changes beyond manipulating the order in which they are served. Preparation method is an additional change to consider, as the acceptability of vegetables to children may be improved by avenues such as serving fresh vegetables with dip, preparing cooked vegetables with some added fat, and incorporating vegetables covertly in entrees [20–24]. In our study the vegetables served were predominately canned vegetables that may have lacked taste and texture appeal. In addition, the fresh vegetables on the menu were not consistently served with a dip. It could be speculated that serving vegetables in advance of other menu items may be an effective strategy for increasing vegetable consumption if attention is given to maximizing taste appeal.
Strengths of the proposed study are numerous. The study was conducted in a Head Start center so that results are relevant in the context of a preschool program that reaches children who are at high risk for overweight and obesity. The experiment was conducted in a naturalistic setting using the cycle menu in place at the participating center. Thus, external validity of findings is likely high. As a trade-off, there may be concerns with internal validity as potential confounding factors such as pre-meal hunger level were not controlled. The use of randomization in tandem with repeat exposure to each experimental condition (2 one week periods for a total of 10 experimental meals per condition) reduces concern over potential confounders because these factors are apt to be equally distributed across experimental periods. Additional weaknesses include potentially limited generalizability of findings because the study was conducted in one Head Start center located in an urban area. Given the small-scale nature of the study, statistical power to examine possible differences in response by factors such as bodyweight status, race, and sex was limited. A final methodological limitation that should be noted is the reliance on meal observation data as the primary measure of food intake. With the meal observation methodology food amounts may be misestimated by observers, thereby introducing measurement error. Assuming misestimating occurred similarly for each experimental condition, the most likely outcome of this measurement error would be failure to detect effects (type II error).