National Food Consumption Survey Lao PDR 2016-2017

The food consumption survey in Lao PDR 2016 was designed as a cross-sectional study to provide separate estimates for three regions of the country.
Based on the wide diversity of the country's regions in terms of climate and agriculture, three zones, representing three regions, were selected: Northern lowland, Central-southern highland and the Mekong corridor. The sampling was performed through a stratified multi-stage sampling technique, following these steps:

1. Selection of two provinces for each region by using simple random sampling
2. Selection of two districts within each province by using simple random sampling (stratified by urban and rural districts)
3. Selection of two villages within each district by simple random sampling
4. Selection of the number of households to be visited (n=42 per village), covering all target groups;
5. Selection of one representative per household; the names of respondents were listed for each one of the 8 target groups, based on age group and sex; the first individual was selected by systematic random sampling method. The selected sample was crosschecked against the household identification in the village registries to ensure that each household member is polled only once with particular attention to the groups for pregnant and lactating women.
   The method for sample selection was supported with information from several sources. In addition to the consultation of the village registries, assistance on the identification of the households and respondents was provided by health workers from the Health center and by the school staff. The districts Oudomxay and Luangprabang in Northern lowland, Vientiane Capital and Bolikhamxay in Central-southern highland, Savannakhet and Champasack in the Mekong corridor were included in the survey.

The target sample size was set at 2016 respondents, and took into consideration the need to capture the variability of diets among the subjects according to the different agro ecological zones, urban and rural areas, and the two seasons of the year. The target sample size was divided between the raining season (n=1008 respondents) and the dry season (n=1008 respondents). The sample for each age group was also stratified by sex to equally represent females and males.

The final total sample size was 2045, with a few additional participants being interviewed in some villages in both seasons.