It has been shown that exposure to mycotoxins through complementary foods and breast milk had profound effects on growth of children from 6 months to 1 year of age. In Tanzania maize is the main ingredient for complementary food, however, the crop is vulnerable to mycotoxins contamination. it has been reported that the most effective way to reduce mycotoxin exposure is to lower the mycotoxins contamination of maize. This study has developed a post-harvest intervention package for use in rural Tanzania to reduce contamination of maize and breast milk with two forms of mycotoxins; aflatoxins and fumonisins. It is composed of 5 components; 1) hand sorting, 2) drying surface, 3) proper drying 4) application of insecticide, 5) dehulling and was introduced in three main maize producing agro-ecological zones. The intervention used randomised controlled trial design where infants less than 8 six months of age were recruited and followed up 6 months after recruitment. The intervention expects to demonstrate the effectiveness of the post harvest intervention package in reducing aflatoxins and fumonisins contamination of maize and subsequent exposure of these toxins to infants and young children through maize based complementary food and breast milk. The findings will be used in developing guidelines for farmers and extension officers to reduce aflatoxins and fumonisins contamination of maize and improve the health of the infants and the whole population.

The study hypothiseses that introducing post-harvest intervention package can reduce aflatoxins and fumonisins contamination of maize and subsquent exposure of these toxins to infants and young children through maize based complementary foods and breast milk than routine agriculture extension services offered to the farmers in rural Tanzania. The specific objectives include;

• To evaluate the effectiveness of post-harvest intervention package on reduction of aflatoxins and fumonisins contamination of maize and in breast milk
• To perform exposure assessment of aflatoxins and fumonisins to infants and young children through maize based complementary food and breast milk
• To perform risk characterization of aflatoxins and fumonisins exposure to infants and young children through maize based complementary foods and breast milk
• To assess the association between exposure to aflatoxins and fumonisins and child growth

Study sites , selection criteria and settings:

This study was conducted in three different agro-ecological zones (AEZ) of Tanzania; these are Northern Highlands, South-western Highlands and Eastern Lowland zones. For each zone, one maize-producing district, primarily for home consumption, was purposively selected, Hanang' district representing the Northern Highlands, Kilosa district represent Eastern Lowland zone and Rungwe represents the South Western Highlands. For each district, ten villages; five interventions and five control was randomly selected. From each village, 10 households were chosen. The selection criterion of the households were to be a local maize cultivars /grower, capacity of storing maize for the period not less than five to six months after harvest and having lactating mother of a child less than 8 months of age. Therefore, a total of 300 households; 100 selected from each of districts were involved in the study. The three districts each are divided into a divisions, wards and villages. A village is the lowest administrative unit and is considered as a geopolitical unit and intact social group. It has a local government authority and committees. Each village has one agriculture extension officer who is a Government employee under the Local Government Authority. The extension services provided includes the provision of farmers with knowledge, information, experiences and technologies needed to increase and sustain productivity and for improved wellbeing and livelihoods.

Design:

Study used a Cluster Randomized Controlled Trial. Villages were randomly assigned to either the intervention or control study groups. The intervention duration was 6 months. The control group continued to receive routine agriculture extension services on good practices for handling crops , an agricultural extension service for farmers offered regularly by village agriculture extension officer at vilage level. The extension services provided includes the provision of farmers with knowledge, information, experiences and technologies needed to increase and sustain productivity, avoid crop spoilage during storage for improved wellbeing and livelihoods. The intervention group received post-harvset intervention package in addition to the routine agriculture education.

Recruitment:

One month before harvest , infants aged less than 8 months were recruited from the register of births in child and reproductive health clinics in each village. Infants of less than 8 months of age were identified using their registration number and date of birth. In Tanzania, all infants born in clinics are registered soon after birth. In case of home deliveries, registration is done on the day the child is taken to the clinic for immunization. On registration, each child is allocated a registration number and the child's particulars including date and place of birth recorded. Inclusion criteria were to be a breastfeeding infant aged 0-8 months at the time of recruitment, parents to be local residence, maize grower and capacity of storing maize for the period not less than six months after harvest. An exclusion criterion was abnormalities impairing feeding or physical growth measurements of an infant.

Ethical considerations:

Application requesting ethical approval was made to the National Institute of Medical Research in Tanzania. Local government nutritionists, resident nurses and sociologists were involved in this study. With the help of village executive officers, mothers of the eligible infants were requested to gather at a nearby health facility or school to have informal meetings. During the meetings, the objective of the research, use of the results, benefits of the research to them and the procedure used to select them were explained and their formal consent sought. Mothers were also provided with written forms containing the same information. For the mothers who were not able to read and write, the resident nurse read the consent form and asked them for their verbal consent. For those mothers who gave their verbal consent, the local nurse wrote her/his name on the consent form and signed on it as a witness.

Intervention:

The developed post-harvest intervention package composed of 5 components, namely 1) hand sorting, 2) drying surface, 3) proper drying 4) application of insecticide, 5) dehulling was elaborated and discussed among Local Government agriculture workers. They later provided guidance to farmers explained the intervention strategy and demonstrated different techniques included in the package. Farmers were provided with guidelines on how to apply the developed package. The language of communication during the intervention delivery was Kiswahili, a written and spoken language in the country.

Component 1: Hand sorting

Separating damaged/infected maize kernel and cobs from health normal one helps to reduce spread of mould during storage. Farmers were shown how to identify damaged maize kernel and that are visibly mouldy. They were encouraged to remove and discard mouldy and damaged maize kernels before storage.

Component 2: Drying practice

Maize are commonly spread on the bare ground for sun drying, making them susceptible to humidity, fungal contamination from the soil, and difficult to gather in the event of unexpected rain. Farmers were trained on proper drying practices that include drying maize on mats. We provided intervention group with the drying mats.

Component 3: Proper sun drying

Incomplete sun drying leaves residual humidity in the maize during storage. Maize should be dried in such a manner that moisture levels are lower than those required to support mould growth during storage (generally less than 13%). This is necessary to prevent further growth of a number of fungal species that may be present on fresh grains. Farmers were shown how to judge the completeness of sun drying by applying the "salt and bottle technique"; with this technique, salt and maize grain are put together in the transparent glass bottle, then shaken for about 5 minutes, if salt sticks on the walls of the bottle and on maize grain surface, it indicates that the right moisture content has not been attained, when no stickiness is noted, then it indicates the completeness of drying. The technique was validated in the laboratory.

Component 4: Application of insecticides

One of the main factors affecting mycotoxins formation is the presence of insects in storage facilities, which produce humidity via metabolic activity and spread fungal spores. This may be minimised by use of suitable, registered insecticides and fungicides. Farmers were provided with packs of insecticides. Explanations on the application of insecticides were given to farmers according to the instruction of use for a particular insecticide.

Component 5: Dehulling

Mycotoxins are likely to be more concentrated in the outer parts (pericarp and embryo) of the maize grain so that removal of these parts would result in a reduction of the toxin level in maize. Importance of de-hulling was explained to farmers and were encouraged to de-hull maize before milling.

Data collection and measuring impact of the intervention package:

Sample collection:

Maize grain samples intended for human consumption were collected at harvest and six months after harvest, while ready to cook maize flour and breast milk samples were taken six months after harvest. All samples were transported to Tanzania Food and Drugs Authority in Dar es Salaam for laboratory analysis. Maize grains were finely grounded before the analysis. An information sheet were prepared and completed for samples from each household to collect all the information on post-harvest practices.

Dietary assessment:

A food intake survey was conducted between five to six months after harvest. Replicas, 24h dietary recall technique was used to estimate amount of maize based food consumed by the infants and mothers. Two visits at an interval of 1 to 2 weeks were made to the home of each of the infants. In addition to 24 dietary recalls, food frequency questionnaire was administered to get information of the frequency of consumption of maize per week and other foods that are commonly consumed by the mother and the child. A questionnaire that was previously used by other researchers in communities in Tanzania which have food consumption habits similar to the communities in the area of our study was used. The amount of maize flour consumed by an infant from a thin (Uji) or stiff (ugali) porridge will be assessed using the Lucille food intake software of Ghent University. This will also be carried out in accordance to the quantity of food consumed and the food preparation techniques described in the Tanzania food composition tables. Breast milk intake will be estimated as documented by United State Environmental Protection Agency (US-EPA, 2011).

Anthropometric measurement:

Each child's weight and date of birth was obtained from his or her clinic card or aregister of birth in local health clinics.

Mycotoxin analysis:

Chemical and reagent: Aflatoxins total, aflatoxin M1 and Fumonisins standards. Acetonitrile, Acetic acid and Methanol of HPLC grade, Methanol, HCL, KOH, BORAX, OPA.

Aflatoxin B1, B2, G1 and G2 were determined in the maize in accordance with the method described by Stroka et al (2000). It was extracted and cleaned using AflaStar immunoaffinity column (AIC), AFM1 was analysed in breast milk as described by (Magoha et al. 2014) extraceextracted and cleaned using immunoaffinity column for AFM1 (AFLAPREP M) and fumonisin analysis in breast milk and maize flour was analysed as described by (Magoha et al. 201a) and (Sydenham et al. 1992) respectively, in both cases samples were extracted and cleaned using SAX column, the extract were dried under nitrogen and reconsituted by ACN:H20 (1:1), and derivatization was done by OPA. Quantification of all toxins was done using HPLC with RF-10AXL detector.

Exposure assessment and risk characterisation:

The exposure of the children to mycotoxins will be assessed by combining the maize consumption data with the contamination data from the analyses. The exposure will be assessed individually, using individual consumption data and body weight (bw) of each infant according to the following formula:

Ei,j = Ci x Lj/Bwi

Where Ei,j is the exposure to the contaminant j (μg/kg bw/day) for the subject i, Ci, is the consumption level of the maize by the subject i (kg/ day , Lj is the level of contaminant j in the maize (μg/kg) , BWi is the body weight of the subject i (kg). The individual exposure to each mycotoxin will be compared with the health based guidance value (TDI, PMTDI, or PTWI), and the children rate exceeding the health-based guidance value will be calculated with the 95% confidence interval.

Data management and statistical analysis:

Data will be entered in double in Epipdata 3.1 (Odense Denmark). Data analysis will be carried out in SPSS version 16.0 for Windows . Statistical significance will be set at P<0.05 for all tests. Chi square test will be used to determine association between post-harvest handling practices of maize and fumonisins and aflatoxins content in maize and breast milk and exposure to infants.

Inclusion Criteria:

• Infants aged 0-8 months at the time of recruitment into the study
• Infants currently being breastfed
• Parents anticipated local residence, maize grower and capacity of storing maize for the period not less than six months after harvest

Exclusion Criteria:

• Abnormalities impairing feeding or physical growth measurements