Detailed guidance on food lists in the Intake24 system
9 min read
Essential information Aim to produce a comprehensive list of the foods consumed by the population. Each food in the list should be unique and easily identifiable to the general public. Local food composition tables, if available, should be used as the starting point for the list. This should then be added to, using robust data on the food intakes of the population, food purchasing data and expert local knowledge. Each food then needs to be linked to a nutrient composition code, at least one method of portion size estimation and categorised into a food group/s. Note - The creation of the food list should be conducted by local nutrition professionals with good knowledge of the local food environment and food composition.
After completing the quick list in Intake24, the user matches each food or drink reported to one of the items returned in the search list. The food list is the list of foods contained within the system against which the users’ search term is matched. It is important that the foods in the list are easily identifiable and unique.
The aim is for the list to be as comprehensive as possible so that the user is able to find all of the foods and drinks, which they want to report, within the system, without making the list too long and therefore difficult to search.
The approach to creating the food list for the initial development of the UK version of Intake24 is detailed below in order to give some insight into the process required (click the arrow to expand).
Background - Approach to developing the UK food list
Assessing foods already in the system
The Intake24 system contains a comprehensive list of foods for adaptation of the system for another locale the food list is provided as an excel spreadsheet.
In adapting the system for use in other locations or populations first go through the current list of foods to identify whether the existing foods should be:
deleted as the food is not consumed by the target population
kept maintaining the existing links to food composition and food portion size data
kept, but linked with local food composition data and/or food photographs.
The food list for adaptation is available here. They include the following fields:
Action should be completed in order to indicate whether a food should be deleted (1), kept in the database with the existing links (2), kept in the database but linked to local food composition data (3) or where a new food is to be added (4) - see instructions below on adding new foods.
Quality of match is to be completed only once the food list has been linked to food composition data. It is used to indicate the confidence in the food composition data match (1 for an exact match, 2 for a good match and 3 for an approximate match).
Frequency of consumption is used to indicate how important a food is in the local diet (1 high, 2 medium, 3 low and 4 rare). This can be used to identify where to invest time e.g. in searching multiple food composition databases for a close match or in conducting recipe calculations.
Frequency of consumption and quality of match can also be used at a later date to identify where to invest time on refining the data.
Intake24 code should be a unique codeof up to 10 characters and should start with the 2 digit country code (See https://www.iso.org/obp/ui).
English description should remain as is and should be created where possible for new foods added to the system.
Local description should be the name of the food in the local language, it should be specific and in layman’s terms so that it is easily recognisable to the end users.
Food composition table and food composition record ID These relate to the food composition database (food tables) used and the unique ID for the food item within that database. For food to be kept in the database with the existing links this should not be changed. For new foods and foods kept in the database but linked to local food composition data this should detail the local composition table to be used. Where recipe calculations are to be conducted “Recipe Calculation” can be inserted in the food composition table column.
Where local composition tables are to be used please add in a local food code in column H (format does not matter i.e. can be numbers letters or combination). If food composition data are still to be calculated/gathered simply add a code to signify the country e.g. MY for Malaysia or ID for Indonesia.
Surveys can be set to ask additional questions about specific foods. The ready meal option is used to identify foods for which the question “was this part of a ready meal?” should be asked if this question is enabled. This should be set to TRUE if it is something which could be part of a ready meal. The same as before option should be TRUE if you would like users to be able to say “It was the same as before”. This is useful if they report an item multiple times during the recall e.g. cups of tea.
The associated food/ food category column is used to generate the prompts within the system for foods commonly consumed together. For example, in the UK version associated prompts ask users whether they added milk or sugar to their tea and coffee, butter or jam to their toast and sauce or cheese to their pasta. Look through the associated prompts within the system and decide whether they are relevant to the local population and add any other common food combinations. Associated foods can be added at the category level e.g. by adding milk at the category level the food list returned will include both cow’s milk and plant based milks. To add a new associated food add the category code and then the question to be asked e.g. BTTR({en: Did you have any butter or margarine on your bread?}). The Associated food name column is a vlookup which brings in the name of the category based on the four letter code.
The brand names column is used where for certain foods or food groups additional questions about the brand might be useful, for example in a study looking at sugars intake we asked about the brand of fruit juices, cereals and biscuits. The Synonyms column should be used to enter alternative names for the same food item but where everyone would recognise the name of the food in the food list. For example bread roll in the food list would have the synonyms buns, baps, cobs, stottie, batch but people from across the UK would all recognise bread roll as the correct match for their food. The brand names as a search term column is used for brand names which may be entered by the user in the quick list but that you do not want to be displayed in the food name. For example, for cereals such as coco pops you could include the supermarket’s own versions e.g. coco krispies, choco snaps, choco rice pops etc.
*Note – The Portion size estimation method column will be populated from the info entered using the template on the portion size data sheet.
The Categories columns are used to assign foods to a category e.g Meat or Fruit. There is a drop down with the four letter category codes and category name columns with vlookup. The Locales column lists all of the locales which that food entry is used in.
The Green section at the bottom of the spreadsheet is for adding new foods.
Identifying regional foods to add
Details for the new foods should be added to the bottom of the base locale food list spreadsheet. Some of the cells are drop downs or have validation to limit errors in data entry.
If the country has its own comprehensive food composition tables these can be used as the starting point for the local food list as this negates the step of matching the food item to the composition code. The next stage in identifying foods to add should be food consumption data for the country or population.
Countries vary in terms of the extent to which data on food intakes of the population is available. The following is a hierarchy of sources for intake data starting with the most robust source first. In reality though a combination of these data sources should be used in order to make the food list as comprehensive as possible.
National dietary data e.g. Government run national surveys. Ideally recent data collected using dietary recalls or food diaries. If possible data on the foods consumed should be used rather than the food code assigned to the item as it is acknowledged that food composition tables are not comprehensive enough to include all items consumed and that nutritionists therefore will assign the closest match.
Large research datasets.
Food purchase data (although this will not allow determination of how foods were prepared or cooked).
Local knowledge/ Dietician / Nutritionist.
Supermarket websites and restaurant menus.
Local knowledge on home grown and foraged items (this information may be gathered through focus groups with the local population)
Bear in mind that there may be cultural or regional variations in the way food is prepared which need to be captured by the system. In this case it may be necessary to include several versions of the food or dish including the details of the region.
Similar foods If there is a similar food you can add this by duplicating the row. For example, if you have spaghetti bolognaise and spaghetti with mince in the local cuisine you can copy the bolognaise row and insert as new row below. Change the English and local names for the food, put the local FCT and food code and assign a new INTAKE24 code. This will mean the links to the portion size images, estimation methods and associated foods from the original food item are maintained.
Food description and associated data
The description of the food should be as clear and as detailed as possible in order that the item can be easily identified by the user. For meat, for example, the cut of the meat and the cooking method should be given, for fruit the variety and colour should be given. There needs to be a balance between the user being able to select the specific food which they consumed and the length of the search list returned. For foods which are commonly consumed and where different varieties/brands/cuts of meat/cooking methods make a significant difference to the nutritional content it is worth including a number of options. For foods less commonly consumed the most frequently consumed item could be included e.g. Gala apple, or a generic term such as Red apple average could be used. Synonyms or regional variations in the naming of foods and brands can be included in the food name or could be added as metadata, which do not display in the food name returned in the food list, but do result in that food being returned. For example, if Braeburn was included in the metadata for Red apple then Red apple would be returned in the food list when Braeburn was entered in the free text search. If there are regional differences in the name given to a food and people wouldn't recognise the other region's name as the same food then duplicate the entry in the food list so you have an entry for each name (linked with the same food composition data and the same portion size data). Synonyms could also be used for foods which are very similar and are eaten too infrequently to justify having separate entries. e.g. Fried rice with prawns and fried rice with crab against a food list entry of fried rice with seafood.
Ontology/ categorisation (Food category)
The food category assigned within Intake24 is used in the categorisation of the food lists. In addition to using the free text search, users can search by food category. The original structure of this was based on a survey of supermarket websites and the intention of the food category is to help the user to find the correct food, drink or meal rather than for use in data analysis.
These are different from the food groups (which are in the food composition tables) that are for use in data analysis, for example, to look at intakes of high sugar cereals, fruit, vegetables, red meat and oily fish. Mixed dishes are broken down into their percentage contribution to different food groups. Further food groups can be set up in the composition tables and added to the data extract if needed.
Foods should be added to an existing food category within the database or, if necessary, a new food category can be created.
Once the adaptation of Intake24 is complete and the system is in use there will be occasions where foods need to be added to the database. This can be done using the Intake24 Editor Tool (see instructions in the section on using the Editor tool). Each new food item added to the system will need to be linked to a food group, food composition data, food category, portion size estimation method and associated prompts.
