The investigators hypothesize that school-based nutrition education will increase dietary diversity among the adolescent girls. The investigators propose a matched, pair-cluster randomized controlled trial to measure the efficacy of school-based nutrition education on dietary diversity of the adolescent girls in Bangladesh. This study will have two arms (one intervention and one control arm). After screening, based on exclusion criteria, the investigators will prepare two separate lists (one for urban and one for rural) of schools in Rangpur district. From each list, clusters (schools) will be paired based on monthly tuition fees provided by the students (as a proxy indicator of socio-economic status of the students) and infrastructure of the schools. The investigators will randomly select one pair from each list and within each pair one school will be assigned to intervention arm and another one will be assigned to control arm through randomization. Targeting an effect size of 20 percentage point reduction of inadequate dietary diversity, a minimum of 148 adolescent girls will be required for each arm. Eleven to fifteen years old adolescent girls studying in grade six, seven and eight will be recruited from each school. To ensure household level participation and support, caregivers will be invited to the school for a discussion at the beginning of the intervention. Nutritional education will be delivered using audio-visual techniques (audio-visual presentation) once in a week for each class, for 3 months. Individual (IDDS) and household dietary diversity scores (HDDS) will be used for measuring dietary diversity at individual and household level, respectively. IDDS and HDDS data will be collected at recruitment, at the end of education intervention and again after 3 months of the completion of intervention. Following national guidelines, weekly iron-folic acid (IFA) supplementation will be provided to both intervention and control arm for 3 months. For identifying the barriers to and facilitators of intake of diversified food, a qualitative research will be conducted after the intervention. Adolescent girls having improved and girls showing no improvement in individual dietary diversity score will be recruited purposively for the qualitative assessment.

The investigators propose a matched, pair-cluster randomized controlled trial to measure the efficacy of school-based nutrition education in improving dietary diversity among Bangladeshi adolescent girls. After screening, based on exclusion criteria the investigators will prepare two separate lists (one for urban and one for rural) of schools in Rangpur district. From each list, clusters (schools) will be paired based on monthly tuition fees provided by the students (as a proxy indicator of socio-economic status of the students) and infrastructure of the schools. The investigators will randomly select one pair from each list and within each pair one school will be assigned to intervention arm and another one will be assigned to control arm through randomization. Randomization will be done by computer-generated random numbers using STATA. Eleven to fifteen years old adolescent girls studying in grade six, seven and eight will be recruited from four selected schools (2 urban and 2 rural schools). Nutrition education will be given in intervention arm and information on water, sanitation and hygiene (WASH) and iron-folic acid will be provided in both the arms. Although the investigators will provide intervention to all the students in each grade, the investigators will select an equal number of girls from each grade to assess the outcomes of intervention. The investigators will also conduct surveys at the households of those selected girls to capture data on household level dietary diversity, water, sanitation and hygiene, and socio-economic status. Three surveys will be conducted among the same adolescent girls and at their household level at three different time points-before rolling out of the intervention, after accomplishment of the intervention and after three months of completion of intervention. At the end of intervention, in-depth interviews with some purposively selected adolescent girls will be conducted to explore the barriers and facilitators which may influence the intake of diversified food among the adolescent girls.

At the beginning of the intervention, the investigators will arrange a parents' meeting at the schools. In parents' meeting, a facilitator (research team member) will discuss about the objective of the study, overview of dietary diversity, health benefits of dietary diversity (DD) on adolescent health and parent's role in ensuring intake of diversified food of adolescent girls. An hour-long nutrition education session will be conducted to each class/grade of the schools under intervention arm. Eight such educational discussion sessions will be held in first 2 months and during month three the previous eight education sessions will be repeated. Nutrition education session will be delivered using audio-visual techniques (audio-visual presentation). In addition, the investigators will arrange quizzes to inspire them and to reduce monotony of the sessions. The investigators will provide pamphlets for their remembrance. Components of eight educational sessions will include overview of 16 food groups of individual dietary diversity score (IDDS) chart, detail description of 16 food groups presenting dietary diversity, overview of basic food components, importance of consuming diversified foods, strategies to select a diversified meal and remove monotony in food, proper timing and frequency of taking meals during 24 hours, selecting diversified meal in low expense, consequences of absence or scarcity of dietary diversity in meal.

Following national guideline, weekly iron-folic acid (IFA) supplementation will be provided to both intervention and control arm for 3 months. The investigators will provide capsule Ferocit Z (provided by government of Bangladesh in different government medical college and healthcare facilities) as Iron-folic acid (IFA) supplementation. It contains ferrous sulphate - 150Mg, folic acid - 0.5Mg and zinc sulphate monohydrate - 61.8Mg. Iron-folic acid (IFA) supplementation will be provided among the girls of each class/grade (six, seven and eight) of all selected schools regardless of intervention and control arm of the study. In the intervention arm, apart from nutrition education the adolescent girls will also receive messages of water, sanitation and hygiene (WASH). But in control arm, the girls will receive messages on WASH only (along with IFA). The message on WASH will include safe source of drinking water, improved toilet facility and how to maintain good sanitation and five critical times of handwashing.

The "State of Food Security and Nutrition in Bangladesh, 2015" reports the nationwide prevalence of inadequate dietary diversity among 10-49 years old women to be 66%. Considering this prevalence, the investigators assume that this intervention would reduce 20 percentage points. However, considering this study as a superiority trial and an effect size of 10 percentage point the investigators also estimated sample size using the formula for sample size estimation in superiority trial. It yields the equal sample size, 93 in each arm. It will also represent the sample size calculated based on the effect size of 20 percentage point using the formula of sample size estimation for conventional cluster randomized control trial. After considering design effect 1.5 and 5% attrition rate, desired sample size will be 148 in each arm. Since the investigators will have two schools in each arm, the investigators will collect data of 74 girls from each school. Thus, to reach the desired sample size the investigators will collect data of 25 girls from each grade in a school.

Dietary diversity scores will be calculated by summing up the number of food groups consumed by the individual participant over the last 24 hours. There is no established cut off point in terms of food groups to indicate adequate or inadequate dietary diversity for the IDDS and HDDS. So, it is recommended to use the mean score for analytical purpose. Dietary diversity scores have been validated for several age/sex groups including non-breast-fed children, adolescents and adults. Categorical variables will be presented as frequency and percentage. Continuous variables will be presented as mean with standard deviation. To see the relationship with study group, t-test and Mann-Whitney test (skewed data) will be used. Linear mixed effect model- a longitudinal data analysis technique will be used to measure the role of intervention on individual dietary diversity and household level dietary diversity. Intention-to-treat and per protocol analysis will be done to assess the effect of nutritional education on changes in the mean dietary diversity scores among the participants. The investigators will perform pairwise comparison tests to see whether the baseline survey data are comparable between intervention and control groups. If the baseline survey reports are found to be not comparable, the investigators will adjust the variables using difference-in-differences analysis.

The transcripts of the interviews will be read thoroughly for familiarity with the data. Furthermore, the field impression notes or memos will be sent to the investigator by the interviewers in order to get feedback on what issues they need to investigate more in-depth. The investigator will provide feedback on those immediately so that the interviewers can investigate more in-depth on those issues from the next interviews. In this iterative way, initial analysis will begin during data collection. After data collection, recorded in-depth interviews will be transcribed and read thoroughly for familiarity with the data. Team members will code these data from transcripts based on a priori and inductive code. A priori code will be based on topic guide considering existing literature. Coded transcripts will be exchanged to check consistency. Common patterns will be identified from coding and data will be analyzed thematically.

Study data collection and administrative forms will be identified by coded number to maintain participants' confidentiality and to enable tracking throughout the study. All information regarding study subjects will be kept in password-protected computer files or in locked file cabinets that can be accessed only by authorized study personnel. Chart information and information from study records will not be released without written permission from the participant's legal guardian (school headmaster). These records will be kept in locked file cabinets. However, records may be reviewed by representatives from the Research Review Committee and Ethical Review Committee of icddr,b. All study related documents will be kept in locked cabinets in locked rooms with limited access. Information in the electronic database established at icddr,b will be password protected and access will be available only to authorized research team members, any information printed from the database will be stored in locked files until its use is complete and then shredded. The study investigators will be responsible for ensuring complete and accurate documentation for the study and for each subject including records detailing each participant's progress through the study, signed informed consent forms, correspondence with IRB (s), adverse event reports and information regarding participant's discontinuation and completion of the study.

Inclusion Criteria:

• Never married adolescent school girls studying in grade six, seven and eight (age range 11-15) at the selected schools
• Household(s) belong to selected study participants
• Girl(s) willing to give assent

Exclusion Criteria:

• Adolescent girls who are < 11 years and > 15 years
• Presence of any kind of chronic disease among study participants
• Presence of any major psychiatric illness

Girls are the future mothers and adolescent malnutrition is directly related to poor reproductive health outcome in future. Moreover, from previous studies it is seen that maternal height is directly related to offspring's height but paternal height is not related to offspring's growth. We cannot increase height of a mother because Linear growth spurt lasts 24-26 months during early adolescence and final adult weight & height gain will cease after early adolescence. So, adolescence is the last opportunity to intervene and break the vicious cycle of inter-generational malnutrition and as we cannot increase a mother's height, we have to go down to the life cycle and put intervention at early adolescence among adolescent girls.