CHARACTERISTICS OF CHILDREN SELECTING LOW-FAT FOODS IN AN ELEMENTARY-SCHOOL LUNCH PROGRAM

Objective: To describe the demographic characteristics of children tha t were associated with the selection of low-fat entrees available in a school lunch program. Design: For 5 consecutive months, we recorded s tudent entree selections on the 46 days in which one of the two availa ble lunch entrees was low fat (less than or equal to 30% of calories f rom fat). Entree selections were tracked using a computerized meal-car d system. The lunch menus did not indicate that one of the two entrees was low fat. Demographic and family characteristics were obtained fro m the school district's registration database and, in one school, from a household telephone survey. Setting: A school lunch program in two public elementary schools in Bellevue, Wash. Participants: Students wh o regularly eat school lunches (N = 471). Sixty-five percent were from families with incomes less than 185% of the federal poverty level, an d 49% were nonwhite. Intervention: None. Main Outcome Measure: The pro portion of days that each student selected the low-fat entree. Results : Girls selected the low-fat entree more often than boys (33% of days vs 27% of days; P < .001), and the proportion of students selecting th e low-fat entree increased with grade level (P = .003). Children were more likely to select low-fat entrees if a household member was known to have an elevated blood cholesterol level (P = .004). The proportion of students selecting the lowfat entree increased with maternal educa tion level (P = .007), but children receiving free or reduced-price lu nches (<185% of the federal poverty level) chose the low-fat entree as often as those receiving full-price lunches (30% of days vs 29% of da ys; P = .14). There were no significant differences in entree selectio n among races. Conclusions: Given a choice of low-fat school lunch ent rees, girls, older children, and those who had family members with ele vated cholesterol levels were most likely to select these entrees. Whi le parental education level was directly related to the selection of l ow-fat entrees, race and family income had little association with ent ree selection.