Replication Data for: Urban Lusaka Food Consumption and Nutrition Survey: Role of Fish in Diets of Vulnerable groups

A household survey (cross sectional study) was conducted to establish the consumption of fish, fish products and other food items at household level (N=714). The role of fish and fish products in the diets of urban poor households, and how fish consumption is distributed within the household between women, children and men. Women and children in the first 1,000 days of life were specifically targeted. Children aged 24 – 59 months from participating households were also enrolled in the study. Lusaka district in Lusaka Province was purposively selected as the study area for the following reasons: it is an urban area within Lusaka Province with the highest number of high density settlement townships where the majority of the urban poor live in Zambia. The study targeted low-income settlement localities as the people living in these areas are most vulnerable to food and nutrition insecurity.
To derive the sample size, the formula was applied; n is the minimum required sample size, Z is the Z score for the desired level of confidence (assumed to be 95% or = 0.05), is the population proportion of interest estimated to be 11%, the prevalence of stunted growth among children in Lusaka (27) and d is the margin of error (assumed to be 5%).
The calculated sample size was further adjusted for the design effect and non-response rate (predicted to be 5%), to obtain the optimal sample size of 714 households. A sampling frame was developed from the 2010 Population Census and Housing report, in consultation with the local authorities and the Central Statistics Office (CSO). The sampling process involved, firstly, purposively selecting the three constituencies (Kanyama, Matero and Munali) from Lusaka district. From each constituency, one ward was randomly selected to participate in the study. In each reporting domain, study households were selected using a three-stage randomized cluster approach, with the first two stages using the Ward and Standard Enumeration Area (SEA) sampling frame from the 2010 CSO. A total of 36 SEAs (clusters) were identified and from each, 20 households were selected. Using a determined sampling interval, systematic random sampling was used in the final sampling stage.
Primary data collection was carried out through a tablet-based questionnaire and by the use of the KoBo Toolkit, a platform to customise the survey to collect specific data, in this study:
a) Demographic and socio-economic characteristics, including employment and income generating activities, water and sanitation, and household assets;
b) Dietary diversity questionnaires were developed and used to collect dietary data for children, women and men. Guidelines on food groups to be included in the questionnaire as provided by FAO 2013 were used in developing the questionnaire for women, men and for household level data collection. The WHO 2010 guidelines were used in developing the questionnaire for collecting dietary data for children 6–23 months of age. Dietary diversity is a proxy for adequate micronutrient-density of foods. A 24 hour recall collected data that was used to estimate food intake for two adults within the household (one male and one female), infants aged 6 – 23 months and one child aged 2 – 5 years. Development of the 24 hr recall was based on the methods described by Gibson and Ferguson (2008). In addition, a dietary diversity questionnaire (FFQ) was used collect data on various food groups women, children and men consumed in the last 24 hours prior to the study. With focus on fish in the diet of young children, information was collected on the use of fish in the initiation of complementary feeding, the age at which fish is fed to children, the perceptions of mother and fathers of the importance of fish for growth and development of the young child.
c) Anthropometric measurements such as weight and length/height were taken on the children and mothers/caregivers. This was done to enable determine the nutritional status of children 6 -23 months; 24- 59 months and women aged 19 – 49 years. The weights of children were taken using the SECA electronic scale and for those children, who were unable to stand, the parents/guardians were asked to carry them and their weights were subtracted from the mothers’ weight. The children’s weights were taken to the nearest 0.1 kg with minimal clothes on them. Length/height boards were used to take the length/height to the nearest 0.1 cm. Children’s age was verified using the clinic card. The mothers’ weight and height were also taken using the SECA scales. The measurements were used to determine mothers’ BMI.

本研究开展了一项横断面家庭调查（cross-sectional study），旨在评估家庭层面的鱼类、鱼类制品及其他食品的消费情况，共纳入714户家庭。研究聚焦于鱼类及鱼类制品在城市贫困家庭膳食中的作用，以及家庭内部鱼类消费在女性、儿童与男性间的分配情况，特别关注生命最初1000天内的女性与儿童群体，同时将参与家庭中24~59月龄的儿童纳入研究范围。

本研究选取赞比亚卢萨卡省卢萨卡区作为研究区域，原因如下：该区作为卢萨卡省的城区，拥有赞比亚境内数量最多的高密度聚居城镇，是城市贫困人口的主要聚居地。研究聚焦低收入聚居社区，因为该区域居民最易遭受粮食与营养不安全问题的影响。

样本量计算采用如下公式：n为所需最小样本量，Z为对应目标置信水平的Z统计量（本研究假设置信度为95%，即α=0.05），p为目标总体比例，预估为11%（即卢萨卡地区儿童生长迟缓患病率，文献27），d为边际误差（假设为5%）。计算得到的样本量进一步通过设计效应与无应答率（预估为5%）进行校正，最终得到最优样本量714户家庭。

研究结合当地政府与中央统计署（Central Statistics Office, CSO）的协作意见，基于2010年人口普查与住房报告构建抽样框。抽样流程如下：首先，从卢萨卡区中有意选取3个选区（卡尼亚马、马特罗与穆纳利）；随后从每个选区中随机抽取1个社区参与研究。各报告域内的研究家庭采用三阶段整群随机抽样法选取：前两阶段基于2010年CSO的社区与标准枚举区（Standard Enumeration Area, SEA）抽样框。最终共确定36个标准枚举区（群），每个群中抽取20户家庭；最后一阶段采用系统随机抽样，通过预设抽样间隔完成选取。

本研究采用基于平板电脑的问卷结合KoBo Toolkit开展原始数据收集：KoBo Toolkit是一款可自定义调研方案以收集特定数据的平台，具体收集内容如下：
a) 人口与社会经济特征：涵盖就业与创收活动、饮水与卫生设施、家庭资产等内容；
b) 膳食多样性问卷：针对儿童、女性与男性分别设计膳食多样性问卷以收集膳食数据。针对女性、男性及家庭层面的数据收集，问卷编制采用联合国粮食及农业组织（Food and Agriculture Organization, FAO）2013年发布的食物分组指南；针对6~23月龄婴幼儿的膳食数据收集，则采用世界卫生组织（World Health Organization, WHO）2010年发布的指南。膳食多样性可作为食品微量营养素密度充足性的替代指标。本研究采用24小时膳食回顾法收集数据，用于估算家庭内2名成人（1男1女）、6~23月龄婴幼儿及1名2~5岁儿童的食物摄入量；24小时回顾法的设计参考了Gibson与Ferguson（2008）的研究方法。此外，本研究还采用膳食频率问卷（Food Frequency Questionnaire, FFQ）收集研究前24小时内女性、儿童与男性各类食品的消费数据。鉴于本研究聚焦幼儿膳食中的鱼类摄入，问卷还收集了如下信息：鱼类在辅食添加初期的使用情况、儿童首次食用鱼类的年龄，以及父母对鱼类促进幼儿生长发育重要性的认知。
c) 人体测量学指标：对儿童及其母亲/照料者开展体重、身长/身高的测量，以评估6~23月龄、24~59月龄儿童及19~49岁女性的营养状况。儿童体重采用SECA电子秤进行测量：对于无法站立的儿童，由其父母/监护人抱着称重，再减去母亲的体重得到儿童实际体重；测量时儿童仅着少量衣物，体重精度精确至0.1kg。身长/身高采用身长/身高测量板进行测量，精度精确至0.1cm；儿童年龄通过诊疗卡进行核实。母亲的体重与身高同样采用SECA秤进行测量，测量结果用于计算母亲的身体质量指数（Body Mass Index, BMI）。