Livestock

Blood fed An. funestus adult females resting indoor were collected in houses between 06 and 12 AM in Tororo District (0°45′ N, 34°5′E) in Eastern Uganda, an area of high malaria transmission [2] (link). The collection was carried out in April and November 2009. Blood-fed and gravid mosquitoes resting inside houses were collected using aspirators and torches and immediately transported to the laboratory of the National Livestock Resources Research Institute based in Tororo. A new method was used to induce the females to lay eggs. Briefly, the collected blood fed females were stored in net covered paper cups or in cages and provided with cotton wool moistened with sucrose. They were maintained for 4 to 5 days to allow them to fully reach the gravid stage and were checked daily for survival. The gravid mosquitoes were then gently individually introduced into 1.5 ml Eppendorf tubes containing a 1square cm filter paper inserted into the bottom of the tube. The filter paper was moistened and excess water removed. The cap of the Eppendorf tube was pierced with 3 holes to allow air into the tube. The tubes were checked daily for the presence of eggs. Females that laid eggs were carefully removed from the tubes and transferred into Eppendorf tubes with silica gel. Eggs were stored at room temperature or at 4°C for up to 2 days and then sent via courier to the Liverpool School of Tropical Medicine (LSTM) where they were allowed to hatch in small cup and later transferred to larvae bowls for rearing. Apart from 20 families that were reared individually, the egg batches were pooled and reared together. Larvae were comparatively reared in mineral (bottled) and distilled water and fed abundantly with TetraminTM baby fish food every day. Water of each larvae bowl was changed every two days to reduce the mortality. The F₁ adults generated were randomly mixed in cages for subsequent experiments.

We adapted demographic and SES questions from the most recent DHS questionnaires [13 ]. Improved water and sanitation were based on World Health Organization definitions [14 ]. Site investigators reviewed questionnaires and identified items that were problematic in their sites. Each site approved a final list of questions and response categories and the associated data collection procedures. Final demographic questions focused on age and education of the head of household and child’s mother, as well as mother’s fertility history. The SES section assessed household assets, housing materials, water source and sanitation facilities, and ownership of land or livestock. The survey also included a question on monthly household income in local currency. The questionnaire was developed in English and translated into local languages as appropriate and back-translated for quality assurance.

The procedure builds on the method developed by Wint and Robinson [16] for disaggregating livestock statistics, based on environmental and other spatial data. Spatially stratified, statistical regression models are developed using data from a series of sample points within each training data polygon and these models are then applied to the entire one-kilometre resolution set of predictor variables in order to estimate livestock densities, disaggregated over a defined study area. For GLW 2, this basic methodology has been revised and improved in a number of ways and was first assessed by Van Boeckel et al. [19] (link) and Prosser et al. [18] (link) in a detailed analysis of its performances for modelling domestic ducks in Asia and poultry in China respectively. Van Boeckel et al. (2011) looked at how downscaling performance was influenced by the aggregation level of input domestic duck data in Thailand (no data for the country, only one value for the country, administrative level 1 data, and administrative level 2 data) and comparing the predictions to actual admin level 3 data. The result was that downscaling based on the method outlined below was giving relatively good results provided that the training data were available at administrative level 1, and were degraded with coarser (national-level data) input data. In a separate study, Prosser et al. (2011) compared land-use based downscaling with the GLW 2 methodology to predict chicken ducks and geese in China, and found land-use based methods to give lower performances. In another previous study, though based on a non-stratified implementation of regression methods, Newmann et al. (2009) found comparable results between land-use and regression based downscaling [17] . In human population mapping, the AfriPop and AsiaPop databases are still largely based on land-use weighting of human population density per land use class [35] (link)–[36] (link), whereas new developments of the WorldPop consortium (www.worldpop.org.uk) are moving toward the use of machine learning methods such as random forest or boosted regression trees (Andy Tatem, comm. pers).
The stratified regressions are repeated a specified number of times using random selections of pixels from which to extract dependent and independent variables (bootstraps). This produces multiple models from which the variability as well as the mean values of the model predictions can be calculated for each pixel.
Because of the sheer volume of input and covariate data, the modelling must be broken down into continental tiles. Whilst bespoke geographic tiles can be created for specific tasks or projects, there are six continental tiles that are processed independently, and the global dataset is updated every time a new continental tile is processed (see file S1).

At the initiation of the deer feed exposure period, deer were presented exclusively with FDF in an elevated livestock feeder (Additional file 2: Figure S1). A maximum of 1 kg was presented to deer every 24 h, and deer were provided commercial deer diet immediately upon consuming all FDF. Each morning of exposure (08:00 a.m.) the FDF was removed temporarily and weighed to the nearest 0.1 g, after which fresh FDF was immediately presented to the deer. At the conclusion of the exposure period, all FDF was weighed and permanently removed. The above procedures were also followed for the deer in the control group, but they were presented with an untreated placebo (containing all ingredients of FDF minus fipronil) rather than FDF.

Data were double entered using Epi-data 3.1. Data were cleaned, coded, checked for missing and outliers, and analyzed using Stata 14 (College Station, Texas 77845 USA). Frequencies, percentages, summary measures and tables were used to describe and present the descriptive information of respondents. The MUAC is a much simpler anthropometric measure than the BMI, as its use eliminates the need for expensive equipment, such as height charts and scales, and the need for calculations. It is also much easier to perform on a patient who is acutely unwell, bed bound or sedentary. Another important advantage of using MUAC is that there is minimal change in the MUAC during pregnancy, so it may be a better indicator of pre-pregnancy body fat and nutrition than the BMI. The outcome variable (undernutrition) was dichotomized as undernutrition (coded as 1) and normal (coded as 0). Poisson regression analysis models with a robust variance estimate were fitted to identify predictors of undernutrition. Next, the binary analysis variables with a p< 0.25 were entered into the adjusted log-binomial models. Results were presented using the crude prevalence ratio (CPR) and adjusted prevalence ratio (aPR). Akaike’s information criterion (AIC) and Bayesian information criterion (BIC) were used to test for model fitness. The goodness-of-fit was assessed using the Pearson chi-square and deviance tests, with the statistical significance level at alpha = 5%. The explanatory variables were examined for multi-collinearity before taking them into the multivariable model using a correlation matrix for the regression coefficients, the standard errors, and the variance inflation factor value.
The wealth index was employed to estimate the economic level of families. The wealth dispersion was generated by applying the principal component analysis (PCA). The index was calculated based on the ownership of latrines, agricultural land and size, selected household assets, livestock quantities, and source of drinking water, a total of 41 household variables. The previous paper [28 ] described nutritional knowledge and attitudes toward consumption of an iron-rich diet using the Likert scale applying the PCA; the factor scores were totaled and classified into tertiles. Women’s autonomy was evaluated using seven validated questions adopted from the Ethiopian Demographic Health Survey [34 ]. For each question, the response was coded as "one" when the decision was made by the woman alone or jointly with her husband, or "zero" otherwise. The detailed description has been given elsewhere in a previous papers [21 , 23 (link), 24 (link)].

These variables were chosen based on the availability of information in the respective EDHS reports and their possible influence on consumption of ASF from previous studies. Variables such as age [18 (link)], residence [18 (link)], maternal education [30 (link)], household wealth [31 ], household ownership of assets [23 ], and ANC [32 (link)] were reported to influence ASF consumption by children.
Demographic variables included in this analysis were child age, child sex, and maternal age. Child age in months was taken as a continuous variable. Maternal age was categorized into three levels (15–24 years, 25–34 years, and 35–49 years). Besides, family size was grouped into < = 3, 4–6, and > = 7.
Socioeconomic factors included maternal/father’s education, maternal/father’s occupation, household ownership of assets and household wealth. Educational status was dichotomized into no formal education versus formal education. Occupational status was grouped as not working, agricultural works, and non-agricultural works. Household ownership of assets was calculated as an added score from a set of twelve assets including electricity, a watch or clock, a radio, a television, a mobile telephone, a non-mobile telephone, a refrigerator, a table, a chair, a bed with a mattress (cotton/sponge/spring), an electric mitad (a grill or cooktop used for preparing injera or bread), and a kerosene lamp/pressure lamp. Similarly, total number of livestock owned by a household was determined as a summed up variable from a set of livestock including cattle (no cattle data in the 2011 series), cow/bull, sheep, goat, camel, and chicken. Health service factors included frequency of ANC visit (grouped as none, 1–4 visits, or 4 and above visits) and place of delivery (grouped as home or health facility).