The Impact on School Feeding on Children's Growth and Development

To enable useful comparison, FSNSP will adapt the study design used by World Food Programme to evaluate a similar school feeding intervention undertaken in chronically food insecure areas in rural Bangladesh during the period 2002-04. The baseline study will be conducted using a cluster randomized study design in which schools represent clusters. Seven schools will be randomly selected from each of 7 Upazilas out of 10 that have been selected for intervention. A control group will be included in study design consisting of 7 schools selected randomly from 7 comparable yet adjacent Upazilas not slated for inclusion in the feeding programme. Control groups will share the same socio-economic, geographic background characteristics of adjacent intervention schools, but are located in Upazilas that have not been selected by GoB to be part of the pilot School Feeding Programme. Use of a control group will enable later evaluations to attribute possible improvements in key impact/outcome variables to the intervention itself. A mid-term followup and endline study will be conducted on the same children approximately 6 months and 10-12 months following baseline. The precise timing of follow-up will be a function of when biscuit distribution is initiated, and seasonal period. Care will be taken to capture the same season as the baseline (to control for seasonal availability of fruits and vegetables), and to ensure that follow-up occurs before the Ramadan period when normal food consumption patterns are disrupted.

The baseline study will consist of data collected at individual (child), household (mother) and institutional (school) levels. Study methods will include: 1) a structured survey questionnaire delivered to teachers, school children, and their mothers; 2) In-depth interviews/focus groups with teachers, community members and NGO implementers; 3) data abstraction from school records; 4) anthropometric and cognitive measures of all primary school students, and 5) biological measures on a sub-sample of children. Programme staff will be consulted in the design and piloting of study instruments, and in thinking about how to best accommodate monitoring and evaluation activities into project design.

1. Structured Surveys The student survey questionnaire will comprise of a series of questions that assess what they have eaten recently, as well perceptions of satiety, hunger, energy levels, ability to concentrate, and whether there have been recent school absences.

The mother's questionnaire will cover background information concerning household food security and socio-economic status, as well as information on the child's eating practices, recent morbidity, and personal assessment of child's energy level, concentration and irritability.

A teacher's questionnaire queries their perceptions and experience in the classroom regarding the behavior of children in the classroom.

1. In-depth Interviews In-depth interviews with teachers, school officials and NGO implementer will be undertaken to assess attitudes, expectations and plans regarding the school feeding program (in intervention schools only), and whether any other school-based nutrition or health programmes had ever been implemented in the area.
2. School Records Information on school infrastructure (toilet, water supply), schedules, management structures (school committees) will be collected, and data on school attendance, and enrollment (over the past 3 years), will be extracted from school records.
3. Anthropometry and Cognitive Measures Following the administration of the student survey, measures of height and weight will be taken on each child interviewed, and a short cognitive test "Raven's Matrices" will be administered.

Weight of the children will be measured to the nearest 100 g using a digital scale. Height of the child will be determined with a locally constructed instrument in which a plastic tape measure is extended between a footplate and head bar. Height will be measured within 0.1 cm. The mean of three consecutive measurements (height and weight) will be considered as the observed value. This will be done during data analysis. All measurements will subsequently be compared to the standards according to the WHO 2006 growth standards and the nutritional status will be assessed by z-score. All anthropometric measurements will be performed by trained and experienced field workers following standard procedures.

Raven's Progressive Matrices (often referred to simply as Raven's Matrices) are non-verbal multiple choice measures of the reasoning (or, better, "meaning-making") component of Spearman's g, which is often referred to as general intelligence. The tests were originally developed by John C. Raven in 1936. In each test item, the subject is asked to identify the missing element that completes a pattern. Many patterns are presented in the form of a 4x4, 3x3, or 2x2 matrix, giving the test its name.

It is developed based on simple measures of the two main components of Spearman's g viz. (1) the ability to think clearly and make sense of complexity, which is known as eductive ability (from the Latin root "educere", meaning "to draw out") and (2) the ability to store and reproduce information, known as reproductive ability.

Raven's CPM is designed for young children ages 5:0-11:0 years and older adults. The test consists of 36 items in 3 sets (A, Ab, B), with 12 items per set arranged to assess the chief cognitive processes of which children under 11 years of age are usually capable. The CPM items are arranged to assess cognitive development up to the stage when a person is sufficiently able to reason by analogy and adopt this way of thinking as a consistent method of inference.

The Raven's CPM produces a single raw score that can be converted to a percentile based on normative data collected from various groups.

Although RAVEN was developed for research purposes, because of its independence of language and reading and writing skills, and simplicity of use and interpretation, it has found widespread application internationally (10,11). RAVEN has been deployed successfully among young children in Bangladesh (personal communication with Jena Hamadani, Fahmida Tofail), and Bengali training guides have been developed.

1. Biochemical Measures, Assessments of haemoglobin, Other biochemical tests will be performed on a separate portion of blood (1.5 mL). These include: Assessments of C-reative protein. Assessments of serum ferritin, Assessments of retinol Assessments of folic acid, Assessments of Vitamin B12, We also use pool serum as internal quality control. We participate CDC vital EQA program for serum vitamin B12

Assessments of Zinc, Assessments of 25-Hydroxy Vitamin, Assessments of urinary iodine

Sampling technique To enroll a sufficiently sized random sample of eligible school children, we will obtain lists of children in Classes 1-3 from the school authority. Participating school children will be selected randomly from these lists through computer-generated random numbers. An equal number of boys and girls will be selected from each class to enable gender analysis. Schoolteachers will assist in identifying and contacting all randomly-selected children and their mothers at household level. In the case of a school child who is randomly selected but is not present at the school due to illness, dropout etc., the subsequent child on class list will be selected. Children whose cognitive abilities are are challenged by autism, mental issues or iodine deficiency will be excluded from the study.

Assuming a mean weight of 22.4 kg (SD=3.0 kg); and SFP is expected to increase the mean weight from 22.4 kg to 24.0 kg, thus to detect a difference of 1.6 kg with 90% power and type I error=0.05, design effect=2.0 and 10% drop-outs, the necessary sample-size calculated to be at least 164 school children in each group. This is less than the sample size required for BMI.

For collection of blood and urine sample we will first randomly select a school from the upazila school list. Than again randomization will be performed for selection of children for specimen collection from the study participating children of that school.