The Impact of the School Breakfast Program on the National School Lunch Program in an Elementary School

Year: 2008

Research Center: Department of Nutrition at the University of California, Davis

Investigator: Spears, Karen

Institution: University of Nevada, Reno

Project Contact:
Karen Spears
University of Nevada, Reno
1664 North Virginia Street
Reno, NV 89557
Phone: 775-784-1775


An average of about 10 million children participated in the School Breakfast Program (SBP) each school day in 2006. Together with the 30 million children who participated daily in the National School Lunch Program (NSLP), spending for both programs totaled $10.2 billion. The SBP was permanently instituted in 1975 and the combination of the NSLP and SBP was intended to provide “a coordinated and comprehensive child food service in schools.” Yet, the interaction between these two programs has not been examined. The effect of breakfast consumption on subsequent lunch consumption is unknown. Furthermore, little information is known regarding the nutritional impact of the Universal-Free School Breakfast Program or the serving of school breakfast in the classroom as opposed to the cafeteria. One study, the Universal-Free School Breakfast Program Pilot found no significant difference in the average nutrient intakes at breakfast or over 24 hours between SBP participants and nonparticipants. The differences in nutrient intake between SBP participants and nonparticipants generally dissipated over the course of the day.

Studies attempting to evaluate the SBP and the NSLP dietary intakes have been hindered by current dietary assessment methods. Current methods do not address the unique challenges of children who lack the literacy skills, cognitive abilities, and attention span that commonly used dietary intake instruments require. To address these issues, the Spears Point-of-Sale Dietary Assessment Tool (Spears POS-DAT) was developed. The Spears POS-DAT resembles a grocery store checkout arrangement. It uses a bar code scanning system and weighs all food items selected and returned by a student. Actual nutrient intake for a subject is automatically calculated by the difference. The Spears POS-DAT minimizes the burden placed on the students.

A total of 157 children, of which 76 percent were White Hispanic, enrolled in the study. The study was conducted in an elementary school in Reno, NV. One class from each grade kindergarten through fifth was selected. All school meals were provided free of charge to all students in the study. Breakfast and lunch intakes were measured for 9 days at three time periods during school year 2007-08. Breakfast was served in the cafeteria during period 1 and in the classroom during periods 2 and 3. Direct observation assessment method was conducted during period 2 and 3 to determine the validity of the Spears POS-DAT.

The study focused on six issues to assess the impact of SBP on NSLP nutritional intake by students. Results from this study indicate a complex interaction.

The location of breakfast service significantly impacted school lunch intake. Lunch intake was less when breakfast was offered in the classroom compared with breakfast in the cafeteria. This was seen for all nutrient variables, except for calories, after adjusting for confounding factors, such as calories consumed at breakfast and student’s age, gender, body mass index (BMI), appetite rating, and prior consumption outside school.

Although, on average, students’ NSLP calorie intake significantly decreased with SBP participation, higher calorie intake at breakfast showed a tendency to increase the lunch calorie intake. Higher calorie intake at breakfast and at lunch may be an important factor in childhood obesity. For specific nutrients, eating school breakfast significantly increased the total intakes of Vitamin C, carbohydrates, and calcium but not the total intakes of fat, iron, and protein at lunch.

An interesting finding was observed for students who usually skip breakfast. When meal skippers incorporated school breakfast into their diet, their lunch intake of calories and nutrients did not significantly change.

Intuitively, expectations are that students should be hungrier and consume more if they had a longer period between breakfast and lunch. However, the opposite was observed in this study. The longer the time elapsed, a greater reduction in their energy and nutrient intakes at lunch occurred, even after adjusting for potential confounders.

Among the potential confounding variables investigated (calories consumed at breakfast, prior consumption outside school and student’s age, gender, BMI, appetite rating, and time-lapse between breakfast and lunch), the demographic variables of age and sex were the main significant contributors to nutrient intakes in general linear modeling.

Lunch intakes of 700 trays were evaluated using both direct observation and the Spears POS-DAT. Mean nutrient intakes did not significantly differ. The correlation coefficients range was between 0.5-0.8. No systematic or proportional bias was detected by Bland-Altman plots. Study findings indicate that Spears POS-DAT is a valid instrument for measuring dietary intake in elementary children.

Direct inquiries about this study to the Project Contact listed above.