The COVID-19 pandemic has caused disruption to food security in many countries, including Kenya. However, the impact of this on food provision to children at an individual level is unknown. This small study aimed to provide a qualitative snapshot of the diets of children during the COVID-19 pandemic. During completion of 24-h food recalls, with 15 families with children aged 5–8 years, caregivers were asked about changes they had made to foods given to their children due to the pandemic. Food recalls were analysed to assess nutrient intakes. Qualitative comments were thematically analysed. Most of the families reported making some changes to foods they provided to their children due to COVID-19. Reasons for these changes fell into three themes, inability to access foods (both due to formal restriction of movements and fear of leaving the house), poorer availability of foods, and financial constraints (both decreases in income and increases in food prices). The COVID-19 pandemic has affected some foods parents in rural Kenya can provide to their children.
Participants included 15 families of children aged between 4.8 and 7.6 years. Families were recruited from two rural communities in Laikipia East in central Kenya. Both communities (Chuma and Matanya) are located about 15 km south-west of Nanyuki, the nearest town. Participants were recruited as part of a larger ongoing study funded by the UKRI Global Challenges Research Fund (via ESRC, see funding sources), with 80 families taking part in multiple mealtime observations at home and school in Kenya and Zambia. Researchers used existing contacts to local primary schools, whose preschool teachers verbally invited eligible families to take part. Where families were interested in the study, they were given more details about the procedures by local researchers who shared information sheets and consent forms with participants, reading out and explaining items whenever necessary. The study team have a long-standing relationship with these communities which aided participation. Participating parents gave written informed consent, either by signing or providing a thumb print, depending on literacy levels. For the measures reported in this study, parents provided details in interviews, prior to local COVID-19 lockdown restrictions, or over the telephone after lockdown restrictions. Child anthropometric data were collected in schools by trained teachers and research assistants after the lockdown restrictions were lifted (March 2021). Parents were asked open-ended questions about the impact of the pandemic on the child’s food provision. In particular, as part of each 24-h recall, parents were asked whether the child’s food was typical on that day (and if not, what was different and why), and for each food/drink item whether there was any specific impact of COVID-19 on what the child had consumed (i.e., on the food which the family was able to provide). Data were entered into NViVo version 12 [5] for analysis. The Kenyan Demographics and Health Survey Tool (https://dhsprogram.com/ accessed on 1 June 2020), education and poverty indicators were adapted and used to gain information about maternal and paternal education and family demographics. The dietary recall procedure was adapted to local standards in line with recommendations from the GloboDiet-Africa team [6] and extensively piloted with a team of local preschool teachers and research assistants. Parents completed the 24 h recalls on behalf of their child. Parents completed a 24-h dietary recall over the telephone with a local, trained, researcher. Recalls were completed on three separate days (two weekdays, and one Sunday) in the course of two weeks. Three-repeated 24-h recalls has been shown to yield similar nutrient intakes as three days of prospective weighed food diaries in children in rural Kenya [7], furthermore a review of use of 24 h recalls in low-income countries shows that fewer days of reporting is needed due to the limited variability in individual dietary intakes [8]. Parents provided information about all foods that their child had consumed in the 24-h period preceding the interview. In addition to the information concerning the time and location of snacks and/or meals, the foods consumed (including foods, beverages, condiments, sauces and spreads), brand information, preparation methods, and portion sizes were collected. Grams/day of each food item was calculated for each child and the Food Composition Tables from Kenya [9] were then used to calculate mean daily nutrient intakes. These were compared to nutrient intake recommendations as outlined in the joint FAO/WHO 2002 report [10] and each child was identified as having an adequate or inadequate intake for their age category. The joint FAO/WHO 2004 report on energy requirements [11], was used to determine those with adequate or inadequate energy (kcal) intakes, based on age, sex and weight. Child height and weight were collected in schools by a team of teachers/research assistants trained in using measuring materials. The scales used were Ramtons RM304, which were re-calibrated before each measurement. Height was measured using a yardstick after marking the child’s height on a wall. All measures were taken by one team member and confirmed by a second member of the team. Children were weighed in their school uniforms with shoes removed. Child height and weight were converted to weight for height for age Z scores (WHZ), and height for age (HFA) Z scores, using the WHO AnthroPlus software version 1.0.4 [12]. Children were classed as underweight if their WHZ z scores were ≤−2SD. Having overweight or obesity were defined as >2SD and >3SD, respectively. Stunting was defined as HFA ≤ −2SD. The qualitative data were collected by local, trained fieldworkers in the language requested by the participant (either Kikuyu, Kiswahili, or English). The fieldworkers were fluent in all three languages, and they translated the data from the local language into English. The qualitative data were then analysed by two UK researchers (RC and MJ) through a process of data familiarisation, independent and inductive coding, and grouping of codes into themes. Codes and themes emerging from the data were discussed with researchers in Kenya (PW and HZ) to ensure that results were not biased by cultural assumptions. Consensus on coding and theme generation was reached through discussion. Reports of particular food/drink items that were altered in the diet were extracted from the statements to produce a list of commonly excluded foods. The summary statistics of the quantitative data (n, percentages, mean (sd)) were analysed in Stata version 14.0 [13].
N/A