Recruitment
Of the seven nurseries invited, four (57%) gave consent to take part in the nursery food photography feasibility substudy. Reasons for non-participation by nurseries were staff shortages, tight workloads and changes, for example, in management structures. One nursery mentioned not wishing to take part because of a small dining area that would have been a struggle to accommodate any extra adults or equipment. The nurseries were all located in Bristol and differed in their size, IMD and types of meals provided. One nursery permitted packed lunches. One nursery had a chef and three nurseries received their meals from external catering companies. In addition to lunch and snacks, two nurseries provided breakfast and two provided an afternoon meal.
Nursery managers estimated a total of 153 eligible children to be present on the days of data collection. A total of 180 letters were given to parents by nursery staff giving information about the study and the opportunity to opt their child out of the study; seven opt-out consent forms (4%) were received.
On each day of data collection, the two fieldworkers collected data from one group of children throughout the day. In two nurseries, researchers followed children aged 3–4 years, and in two other nurseries researchers followed children aged 2 years. Across the nurseries, data from a total of 51 children were collected on at least one eating occasion. Of these 51 children, 26 (51%) had data for all eating occasions at nursery on the day of data collection.
Staff present at mealtimes and throughout the day of data collection were asked to complete the staff questionnaire (see Appendix 10) at the end of the day. All 19 staff who were approached completed and returned the questionnaire.
A total of 307 photographs were taken across the days of data collection. When staff looked through photographs at the end of the day, one photograph was deleted as the photograph included the participant’s name on the nursery place mat. Figure 5 gives an example of before-and-after photographs.
FIGURE 5 (a) Before and (b) after photographs of a meal in a nursery. (Figure not reproduced here; see the original article.)
Feasibility of data collection
It is feasible to use a tablet in nurseries for photographic data collection. In combination with a written observation sheet to capture anything not photographed, a good estimate of what children aged 2–4 years at nurseries eat and drink throughout the day can be made. The wide screen of the tablet made it easy to capture all food items and the laminated ID card in one photograph. It was often not feasible to take photographs at a 45-degree angle and an arm’s length away, given the small space provided for each plate at the table, but this did not have an impact on the estimation of portion size.
Acceptability to children
The field notes and comments by nursery staff demonstrated the study was acceptable to the children at the nurseries. Children were happy and comfortable with the presence of researchers and taking a photograph before and after of their plate. Some children showed interest in the laminated ID card and played with it. At times, children would look at researchers, speak or wave to researchers and try to look at the photographs that were being taken of their food. However, in general, taking the photographs did not disrupt their eating experience. One child asked us not to take photographs of their plate; this did not discourage or influence the other children at the table and the child was not distressed and was happy for us to continue with the other children.
Acceptability for nursery staff
The questionnaire responses, field notes and comments made by staff throughout the day of data collection demonstrated that nursery staff generally found the study and the process of taking photographs of food acceptable. Staff were happy with the tablet being stored in the room and with researchers taking the tablet from this location prior to each eating occasion. All nursery managers checked the photographs on the tablet at the end of the day and this did not take longer than a couple of minutes. They all stated that they were happy with the process of checking the photographs.
A total of 89% of all staff indicated that they felt well informed about the data collection process. Fourteen out of 18 staff (77.7%) agreed or strongly agreed that the process of taking photographs did not disrupt the staff, and 12 out of 17 staff (70.5%) indicated the process did not disrupt the children. Overall, 17 out of 19 staff (89%) agreed or strongly agreed that they found the process of taking photographs before and after each snack and meal acceptable. Only one member of staff indicated concern about the content of the photographs taken and this member of staff strongly disagreed or disagreed with all other statements that were provided in the nursery staff questionnaire. No other information was provided on the questionnaire to allow these strong views to be interpreted. Thirteen out of 16 staff (81.3%) said that they were aware that a member of staff checked all the photographs at the end of the day. All but one staff member (94.7%) agreed or strongly agreed that they found it acceptable for researchers to be given details of the food eaten. One staff member commented that the use of video could be considered for capturing what children eat.
There were some challenges in working with the staff, such as staff assuming that they were being observed in how they co-operated and dealt with children at mealtimes, and staff being unaware of the study taking place or what was required of them and the children. These challenges were often addressed through speaking with the staff during the break times and developing a rapport with them so that they would ensure that the children waited for us to take the photograph and identified the correct children. All of the staff we worked with were happy to do this and were committed to helping the researchers; however, all staff needed to be briefed on how they should help the researchers.
It transpired that tablets and photographic equipment are commonly used within nurseries to take pictures of the children’s learning and progress. One nursery communicated that they had recently started using tablets to record in written format approximately how much and what children ate at the different mealtimes (not using photographs).
Weighing of sample portions and food ingredients
It was acceptable for nursery staff to weigh a sample portion of each composite meal that was served. When asked, nursery staff were happy to provide a sample portion and answer any questions about the food content. In the staff questionnaire, 14 out of the 16 staff members (85.5%) who answered the question indicated that they found it acceptable for researchers to weigh a sample plate of food served to the children.
Three nurseries used external catering companies to provide the food to children. The information provided to nursery about the exact ingredients at times covered only the allergens present in the meals, not the specific ingredients. Therefore, it is not necessarily possible to obtain this information about every single item of food served to children.
Photographing drinks
It was not feasible to capture what children drank in photographs because drinks were often served following the food being served, or throughout the meal, and children were often able to help themselves. The type of drink and quantity drank can be captured through the observation sheet. In three nurseries, the cups used for children were identical, whereas the cup sizes in one nursery varied. When cup sizes are the same, it is possible to measure the volume of the cup and estimate when a child drinks one-quarter or a half or the volume. At some nurseries, children had access to either individual water bottles or cups of water throughout the day.
Packed lunches
It is uncommon for all children to have packed lunches. Only a small proportion of children registered at the nursery that permitted packed lunches brought packed lunches. The initial estimate by the nursery manager was that seven children (9%) would bring in packed lunches on the day of data collection.
It was feasible to ask nursery staff to collect the packed lunches prior to lunch being served and for fieldworkers to take photographs of the food in a different room. It was acceptable for the fieldworkers to take apart the lunch and for the other fieldworker to take the photograph and to note what the packed lunch contained. In some circumstances, the individual items, particularly homemade unidentified composite foods, could not be identified. It was difficult to take the ‘after’ photograph following lunch as there were usually multiple wrapped items of the packed lunch that took a large space on the table.
Coding by Pennington Biomedical Research Center
It was feasible to upload all photographs from the tablet and to annotate these in accordance with the file requirements of the PBRC. The photographs and the observation sheet, both containing the participant IDs, were submitted for the portion size to be estimated.
Lessons learnt
Taking photographs of meals
The feasibility of taking the before-and-after photographs of the child’s meals is dependent on the way that the meal is served, and a different approach is needed for different meal types. We encountered several types and the methods required for each are outlined in Appendix 10.
Estimates of researchers needed for data collection
Two fieldworkers are needed for data collection. The number of children whom fieldworkers can follow throughout the day is dependent on the style in which the meals are served. As drink consumption needs to be recorded observationally, and some food items (e.g. fruit that children could help themselves to) needs to be recorded on an observation sheet, it is not feasible for one researcher to do both tasks at the same time as this would result in missing data. Thus, we found that two researchers working together, one taking photographs with the tablet and watching for the children to be served and to finish their meal, while the other records all observational data, would optimise the number of children who could be followed.
The serving style and also the number of children per table is still important to estimate how many children could be followed throughout the day by two researchers. Two researchers can usually follow one table of children in order to accurately record the number of additional food portions and drinks of water the children take. We found that we could follow up to 10 children if they were all seated at the same table. If the meal was served in a rolling fashion, as was most common for snacktime, it was possible to follow all the children present at the meal as approximately four children at a time would be given their portion and, only once they had finished, would more children be called up to get theirs.
The co-operation of the staff is essential in identifying which children we are following and asking children to wait for the photograph to be taken. Letting the nursery know which children were going to be followed can help the staff prepare for the arrival of researchers.
Summary
It is feasible to collect diet data in a nursery setting by taking photographs. A combination of photographic and observational data collection using two fieldworkers together means that up to 10 children can be followed in 1 day, providing that the children are seated at the same table for all meals and that children are served sequentially.