Strategies to encourage healthy eating and active living are needed to prevent childhood obesity (1). Although the family is primarily responsible for the health habits of children aged 2 to 5, many preschool-aged children (including those from low-income families) spend a substantial amount of time in child care facilities, making these settings appropriate for initiating obesity prevention practices (2-6). Best practices include creating a supportive mealtime environment that allows children control over their food intake in response to internal cues (1,7,8) and implementing Satter's Division of Feeding Responsibility (9,10). Satter's widely accepted concept postulates a distribution of feeding and eating roles between child and adult. Adult caregivers are responsible for selecting, preparing, and offering healthful, appropriate foods and determining when and where food is served. Children are responsible for how much of these foods they eat and whether they eat at all.

To our knowledge, no published studies define or validate what constitutes a supportive feeding environment. Hence, for the purposes of this study, we define a supportive feeding environment as one that offers children adequate nutritious food in a physical and social setting that enhances their development, establishes routines, and maintains trust (11,12) (Box). In this environment, adults use feeding practices that help children to eat according to their hunger and fullness cues. Physical needs are met with appropriate space, equipment, and utensils. Social components are sharing, togetherness, conversation, and listening. To promote self-help skills, encouraging, friendly staff model healthy eating, serve meals at consistent times and locations, and help children feed and serve themselves and respond to internal cues (13).

Serving appropriate food in amounts consistent with growth and promoting increased physical activity are the cornerstones of preschool childhood obesity prevention at any income level. As a US Department of Agriculture (USDA) supplemental nutrition assistance program, the Child and Adult Care Food Program (CACFP) provides meals and snacks to 3.2 million low-income American children daily (14) through eligible entities (for eligibility requirements see www.fns.usda.gov/cnd/care/CACFP/aboutcacfp.htm). By specifying types of foods and serving sizes (14-16), CACFP has a positive effect on children's nutrient intake (16,17). For more than 10 years, CACFP has suggested feeding strategies that are congruent with messages aimed at reducing childhood obesity (1,7,8,13,18,19). CACFP requires yearly training for participating staff and agencies, although these trainings focus more on program integrity issues than on feeding guidance. Although CACFP does not address physical activity, the program's written feeding suggestions deal with a clean and safe setting, family-style service (defined at the time of this study as allowing children to serve themselves from common bowls and dishes), preparing enough food to meet the needs of all enrolled children, and allowing seconds. However, except for Head Start, CACFP-funded centers are not required to follow these suggestions. Frisvold credits Head Start nutrition services and nutrition education as being responsible for the significantly lower rate of later childhood obesity in black children who participated in Head Start during their preschool years (20). Although not addressed in Frisvold's study, both Satter's Division of Feeding Responsibility and the family-style meal service mandated by some states for Head Start also might have contributed. Because children from low-income households are at higher risk for obesity (1), CACFP-funded centers could play an important role in modulating the prevalence of childhood obesity (4). Given that licensing agencies in most states do not require specific feeding practices, it is unlikely that nonfunded centers serving low-income families would adhere to a formal set of supportive feeding practices. The objective of this study was to compare CACFP-funded to similar nonfunded centers in terms of staff's and directors' reported mealtime feeding practices. In addition, we explored center and staff training in these practices.

Although directors and staff in CACFP-funded centers were significantly more likely than those in nonfunded centers to engage in some supportive feeding practices, not all CACFP-funded centers trained staff in this area. In addition, not all trained staff reported appropriate supportive feeding practices. These findings reveal an important consideration when discussing strategies to prevent obesity in low-income children: lack of training of child care staff about feeding children. If child care centers are to engage in obesity prevention (1,7,8), anyone involved with feeding children must be trained. Feeding practices that support healthy weight should be stressed, whereas practices that could increase the risk of obesity should be discouraged. The latter include restrictive practices, such as "no sweets before you finish your meal," improper use of praise and attention to "good eaters," and strategies to get children to eat what providers believe they should eat. Goodell et al noted that Head Start caregivers' attitudes influenced what they expected a child to consume (23); child size was a factor in how much they served. Overweight children still need access to adequate food and nutrients for proper growth and development (15,24). Without appropriate training to guide caregivers in working with overweight children, unintentional but serious consequences may result, such as inadvertently restricting access to nutritious food while trying to prevent obesity.

It is difficult to prove the effect of allowing children to self-serve on prevention of obesity (25), but evidence suggests that children who serve themselves waste less food and eat as much as 25% less than those for whom food was plated (26,27). Furthermore, when preschool-aged children were served portions double their age-appropriate size, they consumed 25% more of the entrée and increased their energy intake by 15%, compared with when they were served age-appropriate portions (26). Alternative strategies for helping children meet their hunger and satiety needs must be created for centers where preplated food service is used or where only the minimum amount of food is prepared (24).

Whereas overall training about feeding is necessary, our study suggests specific factors should be included when educating child care staff. Frequency of training may not be as important as where the information comes from and who provides it, that is, the orientation and education of the trainer. Information provided by perceived credible sources and people qualified to teach nutrition results in positive practices. Hence, we were surprised that training provided by health departments was negatively associated with practice. We speculate that health department training has more to do with environmental safety than with feeding children or that the training may be delivered by noncredentialed paraprofessionals who do not have expertise in feeding young children. Indeed, some public health department policies prohibit supportive feeding strategies such as allowing children to serve themselves from a common bowl or to participate in food preparation.

This study was limited in geographic scope, preventing generalizability to other states. Despite this limitation, our findings suggest that when trained by nutrition professionals, child care staff can learn, adopt, and operationalize guidelines for a supportive feeding environment in preventing child obesity.

We were disheartened by the number of CACFP-funded centers where staff were not trained in child feeding. The lack of training may reflect the lack of designated CACFP funding set aside for training beyond that addressing compliance and integrity. The required training is often received by the cook or director, rather than teachers. Annual mandatory training for all involved with child feeding could increase knowledge about nutrition and child development, influence caregiver attitudes about feeding, and promote positive practices. CACFP training in child feeding, either in-person, online, or via written materials, should be available to and required for center directors, staff (including cooks), and anyone in the room at mealtimes. Much of the training regarding feeding occurs during site reviews; however, required center audits are conducted a minimum of only once every 3 years and include limited center personnel (oral communication, D. Hogan, MS, RD, CACFP Programs Professional, Nevada Department of Education). CACFP monitors can implement the use of assessment tools (12,28) during site visits and train centers to use these tools to improve feeding practices. External trainers and CACFP monitors need to design creative interventions to increase awareness of the role of self-service (and other feeding skills) in helping children monitor their energy needs and maintain a reasonable, healthy weight.

Findings from this study should be discussed with USDA and others concerned with the influence of CACFP in preventing child obesity. CACFP can serve as a model in developing healthy eaters and preventing childhood obesity. Nutrition regulations that include attention to feeding environments and nutrition standards should be established (4,15). Specific training funds are necessary to increase awareness, demonstrate role modeling, change perceptions, facilitate acceptance of appropriate feeding strategies, and encourage use of self-assessment tools. Additionally, any center eligible for CACFP services should be encouraged to enroll. The advantage to enrolling in CACFP goes beyond reimbursement for food; it can provide exposure to, and support of, education and training to prevent childhood obesity.