Role of Breastfeeding and Complementary Food on Hemoglobin and Ferritin Levels in a Cambodian Cross-Sectional Sample of Children Aged 3 to 24 Months

Background Iron deficiency derives from a low intake of dietary iron, poor absorption of iron, and high requirements due to growth as well as blood loss. An estimated number of about 50% of all anemia may be attributed to iron deficiency among young children in Cambodia. Methods A cross-sectional survey was conducted in rural Cambodia in September 2012. Villages in pre-selected communes were randomly chosen using stunting as a primary indicator of nutritional status. In total, 928 randomly selected households with children aged 3–23 months were included. Hemoglobin, ferritin, soluble transferrin receptor (sTfR), and retinol binding protein (RBP) were assessed from capillary blood samples. In addition, length/height and weight of mothers and children were taken and data on dietary diversity was collected. A child feeding index (CFI) was created. Associations between biomarkers of iron and vitamin A status and nutritional status or food intake were explored. Results Anemia prevalence was highest among 6- to 12-months-olds (71%). Ferritin and sTfR inversely correlated and were significantly associated with hemoglobin concentrations. The consumption of animal source foods (ASF) significantly impacts on the interaction between ferritin, sTfR and hemoglobin. Concentrations of RBP were significantly higher in children who had received a vitamin A supplement. The CFI was associated with sTfR and hemoglobin. Lower length and weight were associated with lower ferritin levels and showed an indirect effect on hemoglobin through ferritin. Conclusion Nutrition programs targeting children under 2 years of age need to focus on the preparation of complementary foods with high nutrient density to sustainably prevent micronutrient deficiency and generally improve nutritional status. Future assessments of the micronutrient status should include identification of hemoglobinopathies and parasitic infections to better understand all causes of anemia in Cambodian infants and young children. Trial Registration German Clinical Trials Register DRKS00004379


Abstract
Undernutrition remains one of the biggest challenges in developing countries.
Children aged 0-23 months are the most vulnerable group with a peak incidence of mortality and morbidity. The Cambodian Demographic and Health Survey 2010 estimated a national stunting prevalence of 40%. Results show that stunting increases after six months of age when most children receive complementary feeding.The promotion of a nutrient-dense diet based on locallyavailable foods is essential to improve the nutritional status of young children. In order to optimize infants and young children's diets, the FAO project "Improving food security and Market Linkages for Smallholders in OtdarMeanchey and PreahVihear Provinces (MALIS)" will implement a food security project with a nutrition education component.
The objective of the research is toevaluate the effectiveness of the behavior change messages which will be delivered through a community based intervention involving community based organizations. The research design encompasses both qualitative and quantitative methods, including two cross-sectional surveys, focus group discussions and key informant interviews. The first cross-sectional survey will take place prior to the intervention to assess the baseline nutrition security situation of families with children below two years in the MALIS project region. The behavior change communication strategies will be implemented in the prior selected intervention areas by the MALIS project. The second cross-sectional surveywill be conducted after at least 18 months of intervention evaluating the impact of the intervention by a research team.
The primary outcome parameter to be measured will be mean Height for Age z-Scores (HAZ). Secondary outcome parameters will be vitamin A deficiency, irondeficiency and behavior change. Methods for data collection will includeanthropometric measurements, questionnaires, collection of blood plasma by finger prick and focus group discussions. Statistical analysis will compare primarily the difference ofthe mean height-for-age z-score of children in intervention and control areas.

Trial of Improved Practices
In order to support countries in their efforts to address problems of food insecurity and malnutritionFAO has been promoting improved complementary feeding in several countries 1 in the past years by teaching families how to enrich young children's diets using locally available nutrient-dense foods. InAfghanistan and Zambia FAO has assisted the Ministries of Agriculture and Health to improve complementary feeding through formative research using Trials of Improved Practices (TIPs -see box below). This approach has been especially set up by FAO to identify improved recipes and ideal messages for programming to improve feeding practices of women or caretakers of infants and young children. While all complementary feeding recipes and recommendations developed during TIPs are targeted to children 6 month of age and higher, it is important to note that exclusive breastfeeding up to 6 month of age is recommended during the TIPs process (1). It is part of FAO's efforts to integrate nutrition into agricultural and rural development activities and to foster linkages between the health and agricultural sector.
Recommended dietary practices (including improved local recipes) have now been published in separate manuals for Afghanistan (2) and Zambia (3). Participatory cooking demonstrations using nutritionally improved recipes in conjunction with dietary counseling at household level to optimize young children's dietary intake have been introduced in various community development projects in these countries. FAO programs using the TIPs approach to improve infant and young child and family nutrition are ongoing in Malawi, Laos and Cambodia.

Box: What are Trials of Improved Practices (TIPs)?
TIPs are a formative research technique used in programs that promote change. The methodology has been well tested and validated, particularly with regard to health and hygiene behavior, and has been used in various countries to develop nutrition behavior change communication strategies, including infant and young child feeding practices(4-6). Using TIPs,program planners gain an in depth understanding of families' preferences and capabilities, as well as the obstacles they face in improving their nutrition and their motivations in trying newbehaviors and practices. TIPs therefore allow program planners to pre-test, adapt and evaluate the actual practices and recommendations in line with local circumstances and needs for eventual dissemination and promotion on a larger scale. The TIPs can be divided into the three steps: 1. Investigation of food security, family feeding and child feeding practices, development of preliminary list of improved feeding recommendations and training of TIPs facilitators, 2. TIPs implementation: participants explore how to improve their child's health and nutrition, 3. TIPs evaluation and development of detailed plan for disseminating acceptable and feasible feeding recommendations and recipes Preliminary evidence from FAO programs that have employed TIPs to improve infant and young child feeding have demonstrated the following: • families' interest in using an increased variety of locally available nutrientdense foods accessible in different seasons to improve the nutritional adequacy of complementary foods, • the relevance of introducing basic nutrition, child feeding and food selection and preparation skills, and • the acceptability and practical feasibility of using improved complementary feeding recipes in the family setting.

Background information Cambodia
According to the Global Hunger Index report 2011 3 (7) Cambodia was able to improve its status from extremely alarming to serious. With 28% of the population living below the international poverty lineof 1.25 US$ per day, Cambodia remains vulnerable to natural disasters, volatile food prices and following this food insecurity (7,8). The coverage with sanitation facilities and the access to improved drinking water sources continues to be low with 18% of the rural population using improved sanitation facilities and 61% using improved drinking water sources. The literacy rate among the 15 to 24 years old in the population is almost 90% (8).
Globally more than one third of child deaths are related to undernutrition (9,76 Focusing on anthropometric data of length/height for age (Height-for-Age Z-Scores), the values decrease dramatically after two months of age (11,12). Around the globe Asian countries show the most intense stunting rates with a mean value between 1.5 to 2SD below the WHO Growth Standard (13).According to the Cambodian Demographic and Health Survey 2010 (10) a total of 40% of children are stunted.
46% are younger than two years and 10% of the stunted children are younger than six months (Table 1). 3 The Global Hunger Index (GHI) 2011 refers to data collected from 2004 to 2009. The GHI categorizes countries by giving scores for the proportion of undernourished in the population, child mortality and child underweight. The scores then allow a distribution of the countries according to their hunger situation into five groups: low, moderate, serious, alarming, and extremely alarming. 4 Low birth weight is defined by the WHO as a weight below 2500g, for a given time period. (74) (14).
Less than one-third of the under twos in Cambodia meet the minimum criteria for dietary diversity and only about 50% received the minimum number of mealsaccording to the WHO guidelines(1,10). The expected outcome of MALIS nutrition component is that the capability of target households is to be strengthened to the extent that the diets of children below two years meet their nutritional needs by the end of the project. The conceptual framework of nutrition (figure 1) shows the different impact factors on nutrition security as well as food and health security, which may be applied on individual, regional and national level. There has been a demand for in depth research in these areas for many years (25) in order to inform policy makers and programrs to ensure funds made available for nutrition interventions are optimally utilized (26).

The importance of complementary feeding
Several reviews of nutrition interventions (11,12,27,28) have shown that increased attention needs to be given to complementary feeding interventions targeted to children aged 6-23 months, which is the period with the peak incidence of growth faltering, micronutrient deficiencies and infectious diseases in developing countries (11). The effects of poor nutrition resulting in stunting may also be associated with delayed motor and mental development (29,30). Therefore, effective interventions that are preventing and reducing stunting during this vulnerable period should be a high priority.
Several interventions, targeting this age group, put the emphasis mainly on micronutrient supplementation and food supplements, as well as therapeutic feeding and care (31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45)(46). A metaanalysis looking at the impact of micronutrient interventions Comprehensive food-based approaches that promote a variety of nutrient-dense local foods to improve children's dietary intake and nutritional status are essential (18,48). Food-based approaches that focus on the use of locally available, 10 affordable, and accepted nutrient-dense foods and recipes are designed to empower local populations to optimally use their resources and limit their dependency on external resources. Therefore, these approaches have a higher potential to improve child nutrition in the longer term compared to programs that mainly rely on donor funding and the distribution of micronutrients or food supplements.

Hypothesis
Assuming that behavior change messages on breastfeeding and complementary feeding practices as well as hygiene aspects as generated by the TIPs are widely accepted and out into practice by the mothers with children ages 0-23 months the data assessment will be based on the following hypotheses: e) TIPs formative research generates nutritionally improved, culturally acceptable and affordable recipes which result in improved feeding practices and food intake which will result in a higher dietary diversity score for the children.

Objectives
The purpose of this study is to observe the impact of promoting improved

Study sites
The intervention and control areas will be selected following the targeted area of the MALISproject in two provinces, OtdarMeanchey and PreahVihear, in Cambodia.
OtdarMeanchey is divided in five districts and PreahVihear is divided in seven districts. The baseline survey will only include the target regions of MALIS.

Participants
Knowing that recommendations on exclusive breastfeeding are included into the behavior change strategies all households with children 0-23 months of age are eligible to participate in the surveys. During the total data assessment period, families with children with a WAZ or WHZ-score <-2SD or sick children will be send for nutrition counseling and/ or treatment according to the countries guidelines.
Inclusion and exclusion criteria for participating in the surveys are: • being resident in the sampled area, • having at least one child 0 -23 month of age, • being randomly selected, • accepting that anthropometric measurements and blood samples will be taken.
Families with a child who does not have a written record of the child's date of birth or the date is not known by anyone in the family or who's age cannot be estimated based on a seasonal calendar of local events around one month will be excluded from the study (see chapter 6.8.3).

TIPs and Intervention
TIPs has already been conducted provinces in Cambodia from 2009 to 2011 be a wider dissemination of the improved recipes and behavior change messages developed through TIPs. During the dissemination phase data collection will include monitoring data on the dissemination and through MALIS.

Sample size for baseline and impact survey
The appropriate sample size for the cross

TIPs and Intervention
TIPs has already been conducted under the FAO EU Food Facilit from 2009 to 2011 (see Chapter 2.2.1). The intervention will be a wider dissemination of the improved recipes and behavior change messages developed through TIPs. During the dissemination phase data collection will include monitoring data on the dissemination and roll out of the behavior chan

Sample size for baseline and impact survey
The appropriate sample size for the cross-sectional baseline survey will be ned largely by three factors (52) Considering that m=19 children per cluster (village) will be measured, and an estimated ICC of 0.03 is applied (see annex 10.3), the DEFF= 1 + 0.03*18=1.54.
Multiplying the DEFF-value with the above calculation this results in 516 children below two for each treatment arm. Adding an extra 10% to account for drop-outs or non-responders results in a sample size of 568 children below two years for each treatment arm.

Randomization procedure
A simple randomization proportional to population per district will be conducted to This implies that randomization will take place after adjusting for certain variables that might interfere with variables of interest to the study. Variables of adjustment are: • district officer • level of stunting • total village population • average land holding size • wealth status • access to health posts, nutrition rehabilitation unit, or nutrition counseling by a

NGO
The selected clusters will be the objective of randomization. Randomization will bedone using the Software package "Experiment" and the operation "randomize".
The "Experiment"-package is a software extension to the statistical software R © . It serves to design and analyze different types of randomized trials, including cluster randomized trials, block, or matched pair designed trials (60).The baseline and the impact survey will randomly select the study participants in the selected clusters.

Statistical methods and analysis
Data assessment and data analysis will follow the adopted UNICEF framework of underlying causes of malnutrition and mortality (see Chapter 3.1). Nutritional status will be interpreted as a result of household food security, i.e. access to food and food availability, adequate social and care environment, i.e. direct caring behaviors, women's role, status and rights, social organizations and networks, and functioning public health system, i.e. health environment and access to health.
The cross sectional surveys will be used to gather information of the above mentioned parameters before and after the intervention. The data will be entered into the SPSS editor (IBM SPSS Statistics version 20) and checked for inconsistencies by two individuals independently. Flagged values will be checked based on the filled questionnaires and if applicable the study participants will be approached to verify the result.
Statistical analysis will be performed using the statistical packages of IBM SPSS statistics (version 20). At first the variables will be tested for normal distribution, followed by a descriptive analysis of prevalence of malnutrition, anemia, vit-A deficiency, respiratory infections and fever (ARI), diarrhea, morbidity rates, dietary diversity, and knowledge of age appropriate feeding. The latter involves a development of a score to describe the knowledge about age appropriate feeding.
Further, regression analysis will be applied looking for causalities of malnutrition in the studied area.
Focus group discussions will be based on a general interview guide approach. The interview guide will include subjects of implementation, acceptance and obstacles of TIPs. They will be analyzed by applying the content analysis according to Mayring (51) and facilitated by the use of MAXQDA©-software (53).

Ethical considerations
Ethical approval will be applied for at the Institutional Review Board (IRB) of the

Faculty of Medicine of the Justus Liebig University Giessen and local National Ethics
Committee for Health Research of the Ministry of Health, Cambodia, prior to commencement of the study.

Recruitment, risks and benefits
List of possible eligible participants will be obtained from official population's list.
There is no risk to the participants involved in the study.
Benefits to Participants: This study has an indirect benefit to the participants. The study does provide anopportunity for the participants to gain information about their and their children's current nutritionaland health status. Participants will receive a "Participants-Card" including their and their children'shealth data available directly in the field (anthropometric data, age and hemoglobin level). The card willprovide information whether the anthropometric measurements or the measured hemoglobin levelindicate poor nutritional status. In case of abnormal results participants will be sent for nutritioncounseling/ or treatment according to the guidelines of Cambodia for treatment of anemia andmalnutrition.

Informed consent
Prior to the surveys, general consent of the EU FAO project communities to participate in this trial will be obtained. Written consent will be obtained by the individual household to be interviewed at the day of data assessment. The participants will be given every opportunity to clarify any points they do not understand and, if necessary, ask for more information. The participant will be given sufficient time to consider the information provided.
The enumerator and the participant have to sign and date the Informed Consent Form to confirm that consent has been obtained. The participant will be provided with a copy of this document.

Data collection
The data collection in the cross sectional surveys will be based on an adaptation of the SMART methodology (49). The SMART methodology includes a questionnaire designed to assess mortality, nutritional status and food security in crisis situations.
The final questionnaire will consist of standardized modules extracted from the questions designed by the SMART initiative, FAO, WHO, GIZ, and the Manoff Group (1,6,49,54,(55)(56)(57)(58)(59). Thus, mothers/caretakers with children below two years will be interviewed about their socio-economic situation, food security, mothers' and children's food intake, care, time availability, access to health, water and sanitation, motor milestones, and access to FAO food security activities by trained enumerators.
Anthropometric measurements will be taken from mothers, their children below two years of age, and, if possible from fathers, by a nutritionist especially trained in anthropometric measurements. Capillary blood samples will be taken from the surveyed mothers (Hb only) and children below two years to assess the micronutrient status (retinol binding protein (RBP), transferring receptor (TfR), hemoglobin as well as the morbidity status (C-reactive protein (CRP), acyl glycoprotein (AGP) by medical trained research staff.

Interview
After written consent (see consent form in annex) the caretaker/mother of the child will be interviewed face-to-face by trained enumerators. Depending on the given infrastructure in the villages, participants are either invited in advance to come to a central meeting point, e.g. community center or primary school, to do the interview, or enumerators are coming to the homestead of the participants. If the interviews are conducted centrally in a public place, privacy will be assured by keeping an adequate distance between the interviewed participants that only the enumerators will hear the answers. In case village lists are not available, and systematic sampling is not possible, the EPI method will be used (49). Participants will then be invited on the day of selection to come to central meeting point at a certain time to avoid any inconvenience by waiting.

Anthropometric measurements
A central weighing and measuring station will be installed in every village. After the interview, mothers/caretakers (if possible the father as well) and their children will be 19 send to the weighing and measuring station. Weight and height of children and adults will be measured according to an anthropometric protocol based on the WHO report "Physical status: the use and interpretation of anthropometry" (1995) (60).
Weight of children will be determined with the child wearing no clothing. Adults' weights will be taken while wearing light indoor clothing and no shoes. Heights and weights will be assessed to the nearest 0.5 cm and 0.1 kg, respectively. To determine the nutritional status of pregnant women, mid-upper arm circumference (MUAC) will be measured. All measures will be taken twice and the mean value is used for analysis (see ISAK manual (61)).
Anthropometric measurements will be taken with standardized equipment from Seca (SecaGmbh& Co KG, Hamburg, Germany). Weight will be measured using standardized digital flatscales (Seca 874, capacity: 200 kg) with mother/child function. Infants and small children are weighed while being held by the mother. The weight of the mother is assessed separately. Then the Mother-Child function ascertains the tare of the weight. Recumbent length will be taken from children with measuring boards (Seca 417, measurement range: 10-100 cm).
The height of adults will be measured with a stadiometer (Seca 213, measuring range: 20 -205 cm). A non-stretchable measuring tape will be used to take MUAC (Unicef). After the weight and length assessment the children will be tested for edema. Edema will be diagnosed by applying moderate finger pressure on the tops of the child's feet. If there is edema, an impression will clearly remain for at least a few seconds (on both feet). Edema will be recorded as absent, mild (both feet/ankles), moderate (both feet/ankles plus lower legs, hands or lower arms) and severe (generalized edema including both feet, legs, hands, arms, and face) (49).
Two pairs of trained research staff, each consisting of a measurer and an assistant, will take all measurements. Privacy of the participants while taking their measurements will be assured.

Date of birth
Children's dates of birth (age) will be recorded either from immunization cards or birth registration cards. If no documents are available and the mother/caretaker doesn't know the child's birth-date, the age willbe estimated based on the FAO Guidelines for Estimating the Month and Year of Birth of Young Children (FAO 2008) (62).

Blood samples
Blood samples will be taken at the central weighing and measuring station. A medical technical assistant from Cambodia will obtain the blood sample for the biochemicalparameters via a finger prick using sterile disposable micro lancets to obtain capillary blood.
The analysis of biochemical measurement will include assessment of retinol binding protein (RBP), hemoglobin, transferrin receptor (TfR), acyl glycoprotein (AGP) and Creactive protein (CRP). RBP, TfR, AGP and CRP will be analyzed using sandwich ELISA technology (63). With this technology it is possible to combine the measurements of the four proteins (TfR, RBP, CRP, AGP). Instead of using four different ELISA methods with different chemicals and procedures only one method is used. This also reduces the necessary amount of blood. A plasma volume of 2 times 7.5 µl is sufficient to do a double measurement of all four proteins. Directly after taking the blood samples they will be centrifuged and serum/plasma is stored in 0.2 mL PCR tubes. These tubes can also be directly used in an automatic pipettor to avoid the tedious and error prone manual pipetting. Samples will be stored on ice in a high efficient styrofoam box with more than 5 cm thick walls and tightly closing lid. All blood samples will be stored on ice until the end of the survey. Analysis of blood samples will be done by Dr. JuergenErhardt, DBS-Tech, Germany.

Motor milestones
Motor milestones will be assessed according to the Motor Development Study component of the WHO Multicenter Growth Reference Study (MGRS) (63). The following six distinct gross motor milestones will be recorded: sitting without support, hands-and knees crawling, standing with assistance, walking with assistance, standing alone, and walking alone. To assess all milestones, the standardized testing procedures from MGRS will be adopted. A milestone will only be considered as achieved, if all given criteria are met.

Data collected Intervention
During the intervention focus group discussions as well as open one-on-one interviews will be held with family members (especially women and elderly), health workers and nutritionists. Focus group discussions and one-on-one interviews are qualitative empirical research methods and imply that questions are posed in a colloquial manner instead of a standardized questionnaire. The focus groupdiscussions will follow a general interview guide approach. During the focus group discussion participants are encouraged to talk freely about the subject. This will enlarge the understanding of how behavior change messages have been integrated in daily life, how behavior has changed and will as well identify obstacles to the implementation. Therefore, focus group discussions will allow gaining deeper understanding of the complex research background. (65-67).

Quality assurance 6.9.1 Statement of Compliance
The study will be conducted in accordance with the design and specific provisions of The principal investigators will assure that no deviation from or changes to the protocol will take place without prior agreement from the sponsor and documented approval from the IRB. The principal investigator will promptly report to the IRB and the sponsor any changes in research activity and all unanticipated problems.

Translation of questionnaire
Questionnaires will be designed in English and translated into Khmer by native speaking nutritionists. The translated questionnaires will be tested on three to five native speakers. To assure analogous translations, independent, native speaking nutritionists will translate both versions of the questionnaire back into English.

Recruitment of field staff
Main requirements for enumerators are: • language skills (Khmer) • minimum of knowledge in nutrition related research and basic interview experiences A total of 12 enumerators will be enrolled in the data collection process. Two medical trained research staff will be recruited from local hospitals to take blood samples and to assist the HemoCue analysis.

Training and monitoring
Enumerators: The enumerators will be trained on the questionnaire. A guideline on how to conduct the interview will be developed and used during the training workshop. Interviews will be conducted pairwise. One enumerator will ask the questions, the other one will record the answers. The training will as well include the correct assessment of motor milestones based on the observation criteria used in the WHO MGRS (64).
During the data collection process enumerator teams will be matched randomly every day. Interviewers will be systematically and frequently monitored and completed questionnaires will be controlled for missing data and consistency on a daily basis.
Anthropometric measurements: In order to assure that all measurements are taken in the same way the training for taking anthropometric measurements will be based on the WHO Child Growth Standards "Training Course on Child Growth Assessment" (68). Staff members of the JLU Giessen and Mahidol University will carry out the training. Furthermore the training curricula will cover issues of sensitivity to local customs, dress, and practices of modesty. Correct handling of equipment will as well be included in the training. To avoid problems in this area, observers of the same sex as the subjects will be employed. Since the research project mainly focuses on young children and their mothers/caretakers, women will be recruited to take the measurements. Survey leaders will frequently visit the weighing and measuring station. These visits will serve to verify that the anthropometry protocol is being implemented properly and consistently. Retraining sessions will be arranged when a lack of standardization is observed among the researchers.

Assurance of communication
Supervisors are responsible for the technical quality of the surveys for which the survey teams under their charge are conducting. Enumerators will be randomly assigned to a supervisor. One supervisor will be in charge of three survey teams.

23
The duties of a supervisor include advice and control of the accuracy of the survey data. The research team will include a translator for the research supervisors from Germany. This will assure the communication between all research team members as well as participants. Survey leaders and survey supervisors will record all important points in a notebook as soon as possible, including observations, ideas, problems, actions taken to address these problems, and the reasoning behind any decisions taken. Survey leaders, supervisors, and surveyors will meet daily after the survey implementation.

Pretest
A pretest of the entire data collection process will be conducted in villages not selected for the baseline survey. The pretest serves to ensure that the questionnaire is fully understood by the enumerators as well as to test the enumerators' behavior in conducting the interviews. The results will be evaluated by the trainers and a final consultation is held with the enumerators. Difficulties encountered by the enumerators with the questionnaire will be discussed and eventually phrasing or translations will be adjusted. The aim is to clarify any remaining uncertain points and to reach an agreement on the final questionnaire to be used.

Registration of the study
The IMCF research study will be registered at the German Clinical Trials Register

Data protection
Data management procedures will protect confidentiality of all data collected on individuals. All Investigators and study site staff involved with this study must comply with the requirements of the respective data protection laws in Cambodia and Germany with regard to the collection, storage, processing and disclosure of personal information. Access to collated participant data will be restricted to the survey management and stored in a locked cupboard.
Each subject will be assigned a unique identification code that will be used for data entry and analysis.To safeguard confidentiality, subject records are accessible only to the team doing the initial data entry, and the individual checking as a part of the dual entry system. Computers used to collate the data will have limited access measures via user names and passwords. Identity information and consent forms are not kept in the main computerized data file, but in a hard copy kept in a locked cabinet available only to the principle investigators. The front page of the individual record with consent and individual identity information is separated from the rest of the record, which contains only the individual code number.
Test tubes and specimens used by laboratory staff will be labeled by individual code numbers only. The collected blood samples will only be used for the specific purpose covered by the informed consent given.
Published results will not contain any personal data that could allow identification of individual participants.

Dissemination of findings
Reports on the project's progress and regular monitoring of the project activities will be provided by the FAO consultant, the Giessen PostDoc and the Principal Investigators. The research project team will be responsible to produce six-monthly progress reports, which will • contain information on main activities and compliance with the work plan; • identify any problems and constrains encountered during the research progress, • provide recommendations for corrective measures; • if necessary, revise the work plan for the following reporting period.
A contact information database will be created and maintained that will be used for group/individual mailings of paper documents and to facilitate telephone and fax communications between the Project Management, Site Management, and TAC. The project visibility will be enhanced by the launching of a FAO project website. The website will be updated at least on a 6 monthly basis.
• Lessons learnt from the project and research results will be shared through:participation and presentations in relevant conferences and technical consultations regarding nutrition and feeding of infants and young children, • preparation of research articles to be submitted to scientific journals, • documentation of case studies, and • preparation of guidelines and technical recommendations on improved complementary feeding using local resources.

Personnel
The research has been developed and will be carried out by the Institute of Nutritional Sciences, Justus Liebig University Giessen, Germany in collaboration with Institute of Nutrition, Mahidol University, Bangkok, Thailand. A PhD student from Mahidol University and a PhD student from JLU-Giessen (AnikaReinbott) will undertake data collection and analysis; they will be supported by MSc students from Germany and Thailand and Cambodia.
Representatives from FAO (Ellen Muehlhoff), Mahidol University, Institute of Nutrition and JLU-Giessen (Dr. Irmgard Jordan) will be involved in project management and technical implementation. They will provide oversight of the project's implementation, approve workplans and associate budgets, and decide upon adjustments to the project implementation strategy as required. Furthermore, they will be responsible for ensuring that information is effectively shared between FAO, Giessen and national research institutes and governments, and between country-level teams and FAO headquarters and JLU Giessen in Germany.