This was a pre–post study conducted during March through June 2010, after approval by the Griffin Hospital Institutional Review Board (IRB) (IRB study no. 2010/03) by the Yale–Griffin Prevention Research Center (PRC) in collaboration with the New Haven Public Schools (NHPS) District Wellness Committee. The PRC provided teacher training and program materials, developed informed consent materials, and collected pre–post data. The NHPS district was responsible for teaching the program, scheduling data collection, distributing informed consent materials, collecting opt-out forms, and informing the PRC about students who had parental permission to participate in the study. To reduce burden on the district in scheduling data collection, no control schools were used for comparison.

The 5 participating schools were a subset of the district’s elementary/middle schools selected by the NHPS District Wellness Committee and whose principals agreed to participate (Table 1). Of the 5 schools, 4 (Schools A, B, C, and D) were magnet schools with a particular instructional focus, and one (School E) was a traditional school. For the overall student population in each school, the percentages of students ranged from 70.6% to 92.2% for eligibility for free or reduced-price meals, 2.2% to 18.5% for students considered gifted or talented, and 74.9% to 90.8% for minority students The percentage of 5th-grade students in participating schools who met state goals for academic achievement in 2009–2010 ranged from 19.4% to 50.0% for reading, 10.0% to 73.6% for writing, 35.1% to 79.6% for mathematics, and 5.0% to 50.9% for science.

Because the program was incorporated into the health curriculum, all 5th-grade students in participating schools who were present when the Nutrition Detectives lesson was taught took part in the lesson. Study inclusion criteria were 1) being a 5th-grade student in a participating school; 2) having parental consent to complete the quizzes; 3) willingness, availability, and ability to complete the quizzes; and 4) presence at school on the day and in the room where the lesson was taught. Exclusion criteria were students who lacked parental consent, were unwilling/unavailable/unable to complete the quizzes, or were not present for the Nutrition Detectives lesson.

Information sheets and consent forms approved by the Griffin Hospital IRB were sent home to parents or legal guardians of 5th-grade students in study schools. The information sheets provided instructions for parents to opt out of having their child participate by a certain deadline date. All documents and participant information were obtained for research purposes only and strictly maintained to ensure confidentiality. All student assessment data were identified by number, not name.

A few months before this study, the PRC had partnered with the NHPS District Wellness Committee to adapt the 90-minute nutrition program to a 45-minute lesson, to serve as the third of 4 nutrition lessons taught by physical education (PE) teachers in all of the district’s elementary schools. The topics of lessons 1 through 4 were 1) how food affects our bodies; 2) the Food Guide Pyramid; 3) making healthful food choices by using nutrition information on food packages and 5 detective “clues”; and 4) balancing caloric intake and expenditure to maintain a healthy body weight. The nutrition lessons were part of a 5th-grade health curriculum developed by the district and community partners that also included lessons on physical activity, reproductive health, and substance abuse. PRC staff had taught the original 90-minute program to 315 5th-grade students in 8 NHPS schools as part of an unpublished feasibility study (D.L.K. and C.S.K., unpublished data) during the 2007–2008 and 2008–2009 school years and used the same tool to evaluate food label literacy.

The original 90-minute nutrition education program, Nutrition Detectives, (http://www.davidkatzmd.com/nutritiondetectives.aspx), was developed in 2002 and has been periodically updated to reflect changes in food products and refined in content and design according to feedback from teachers and students to strengthen the program’s messages and reduce the potential for misconceptions. Nutrition Detectives conveys the idea of a link between food choices and health, in addition to information on how and what nutritious foods to choose. The key concepts are that “we are what we eat”; students can become “detectives” to learn about the foods they eat; and they can apply 5 “clues” to determine whether food products are healthful (“clued-in”) or less healthful (“clue-less”) choices. The objectives are to enable students to 1) explain why making informed food choices can benefit their health; 2) recognize deception on packages of food products; 3) identify 5 clues to make healthful food choices (Figure); 4) demonstrate the location of the Nutrition Facts panel and ingredient list on food packages; and 5) determine whether foods are healthful choices according to the 5 clues, combined with the Nutrition Facts panels and ingredient lists on their packages. The first half of the program is a PowerPoint presentation designed to engage, motivate, and inform the students. The second half is an activity in which teams of students search through a grocery bag containing food products in a given category, such as cereals, crackers, or snack bars, and decide which products are “clued-in” and which are “clue-less.”

The original Nutrition Detectives program included a 96-slide PowerPoint presentation; a 72-page teacher manual with detailed instructions; wall displays; and handouts. The program can be delivered in one 90-minute session, or in 2, 3, or 4 shorter sessions. Nutrition Detectives is freely available to educators; it led to an 18 percentage-point gain in food-label literacy scores (P < .01) in the Independence School District in Missouri (12).

The condensed Nutrition Detectives program used in the NHPS study included a 40-slide PowerPoint presentation and a 20-page lesson plan with program objectives, key concepts, vocabulary words, abbreviated instructions, wall displays, and handouts.

In March 2010, PRC staff provided teacher training on the Nutrition Detectives lesson to PE teachers from all K–8 schools in the district as part of a professional development session, including (but not limited to) the schools that elected to participate in the evaluation of the lesson. In May 2010, NHPS PE teachers taught the lesson in 5th-grade classrooms as part of the curriculum.

The primary outcome measure was food-label literacy and nutrition knowledge on healthful food choices. Children’s ability to choose “better-for-you” foods was assessed by using a previously validated (15) standardized test instrument (Food Label Quiz) that uses Nutrition Facts panels and ingredient lists from 5 categories of food products (breads, crackers, cereals, cereal bars, and cookies) discussed in the Nutrition Detectives program. The quiz consists of 10 questions. Each question asks students to select the more healthful of a given pair of food products on the basis of the nutrition information for each product. The selection of each better-for-you choice is based on one or more of the program’s 5 clues.

PRC research staff administered the Food Label Quiz in participating schools. Pretest and posttest quizzes were scheduled for days and times convenient to school schedules and relative to when the Nutrition Detectives lesson was taught in each school. When the PRC researchers arrived to collect data, the teachers identified the students who would be completing the quizzes. Because of the way informed consent was managed by the schools, the researchers were unable to determine the number of students who lacked parental consent, were unwilling/unavailable/unable to complete the quizzes, or were not present for the Nutrition Detectives lesson (eg, because of absence from school or disciplinary reasons).

Data were entered in an Excel (Microsoft Corporation, Redmond, Washington) spreadsheet and analyzed by using SPSS for Windows version 15.0 (SPSS Inc, Chicago, Illinois). The Food Label Quiz was scored as the percentage of correct responses. Paired t tests were used to evaluate the change in the percentage of correct responses of Food Label Quiz scores from pretest to posttest. Student t tests were used to assess difference in students’ performance between the 90-minute program and the 45-minute lesson. In all analyses, a 2-tailed α of less than .05 was considered statistically significant. This study was powered at 80% to detect a minimal detectable difference of 18 percentage points between pretest and posttest scores.