English * French * German * Spanish
*
*
*
* *
*
YourReminder *
*
*
*
*
EuroFIR Print this page
*
*
Data transfer from industry: why and how?:
 
European consumers today can choose from an enormous variety of manufactured foods and branded products. These may be used as ingredients in home-cooking such as margarines or pasta sauces, or form ready made components of a meal like pizzas, soups or cakes. Their ingredients and formulations often change over time: for example, products are reformulated to remove/reduce trans fatty acids or to reduce salt. In addition, novel ingredients may be developed or isolated and added to foods, such as plant stanols or sterols in margarines and yoghurt drinks.

 

In order to have a clear picture of actual intakes of nutrients and changes in nutrition that may affect our health, it is crucial to know the composition of foods. Public health programmes and enforcement of food legislation within the EU depend on accurate food composition data, but this is also important for other users. The food industry (design of new products, nutrition labelling etc.), health professionals (to assess patients’ diets) and researchers (to determine links between diet and disease) also depend on having accurate compositional information on branded food for their day to day work.

 

EuroFIR, the world-leading European Network of Excellence on Food Composition Databank systems is a partnership between 49 universities, research institutes and small-to-medium sized enterprises (SMEs) from 26 countries. EuroFIR aims to develop and integrate a comprehensive, coherent and valid databank providing a single, authoritative and up-to-date source of food composition data for Europe.

 

One of the Work packages in EuroFIR which has now finished, was entitled ‘Composite, processed and novel foods’ (WP 2.2, until June 2007). It has investigated needs and solutions for more accurate and timely documentation of food composition data. One of its aims was to encourage the flow of information on the composition of branded products, between industry and database compilers. By working closely with the food industry and industrial associations, the project benefits both industrial players and food composition database (FCDB) compilers.

 

A questionnaire has revealed that only half the FCDB compilers analyse foods at their own institutes. In addition, nutrient information on product labels may not be sufficiently detailed for the requirements of FCDB compilers. Methodological information is unlikely to be included on the label and the definition of the nutrient may be unclear. For example, values for vitamin A may or may not include equivalents from b-carotene and the fibre content of a product will differ according to the method of measurement used. There is sometimes difficulty differentiating between a nutrient that occurs naturally in a product and one that has been added to fortify it.

 

Thus, there is a need to improve the flow of information from industry to FCDB compilers. This is important to facilitate an accurate FCDB that can be used effectively for intake evaluations, food regulations and labelling. Therefore more collaboration, including joint data evaluation and data exchange is needed between food industry and database compilers. There are a number of challenges in doing this. As products are constantly changing, it is difficult to keep a FCDB completely up to date. For many large manufacturers the amount of data involved is huge, so that data transfer can be very time consuming both for them and for those inputting the information into the FCDB. Some countries in Europe have developed systems to transfer data from industry to their national FCDBs. These are explored further in the examples below.

 

 

Examples of food composition data transfer in Europe

The Netherlands
The Dutch nutrient database NEderlands VOedingsstoffenbestand or NEVO is owned by the NEVO Foundation, who manage and maintain data in the database and promote its use. The Foundation is supported financially by the Ministry for Public Health, Welfare and Sport.

 

The characteristics of the database are summarized below: 

Characteristics of the NEVO database

2001

2005

Number of food items

1603

2200

Number of nutrients

42

221

Number of food groups

23

25

Number of recipes

 

200

Number of industrial branded foods

 

700

 

The NEVO database is updated every 4-5 years with new data from manufacturers. Data is collected the year before the updated version is published. For the 1996 edition, foods that were most often consumed by the population were selected. Of the 432 items chosen, 25% were manufactured, branded products. 

 

Manufacturers were asked to send nutritional information on these products in their own format. The process of inputting this data from the hard copy forms in which they were received into the database was extremely labour intensive. Thus for the 2001 update, a structured form was developed and sent to manufacturers. However, many of the manufacturers found using this difficult and time consuming, and the response rate decreased. 

 

An overview of the approach to be used in for the 2006 update of the NEVO database is shown in the flowchart below. In view of past experience, it was agreed that manufacturers would be able to use their own format to submit data, but that forms would be available from NEVO on request. 

 

 

As summarised below, 68 manufacturers were selected for the 2006 update of the database. These were chosen according to the foods most commonly consumed by the Dutch population. Also some particular food groups that needed to be updated were selected. Some manufacturers were asked to provide data on all foods produced and others were asked for data on a selection of products.

 

Letters were sent to these companies of which 40 agreed to participate, 10 did not reply and 18 did not wish to be involved. Reasons given for declining involvement included the work being too time consuming and not being a priority. Of the 40 companies who responded positively, 36 (90%) returned their data before the deadline.

 

Industrial data update 2005

 

 

• Manufacturers selected

68

• Positive responses

40

• Negative responses

18

• No response

10

• New / updated data received within deadline:

36

• No data received

4

• Number of food items involved

1260

• Telephone calls / e-mails per manufacturer

2-5

 

Belgium
The Belgian national FCDB (NIMS-database) is updated and managed by Nubel. The database contains both branded and basic foods and is used as the basis for 3 products; a FCDB booklet, including annexes on nutritional recommendations, lists of additives, weights and measures of food products and a food guide pyramid; an online FCDB and a software programme called Nubel voedingsplanner (nutritional planner). This allows users to calculate individuals’ nutrient intake, body mass index and energy requirements. 
 

The food industry and distribution sectors provide nutritional information for the online database in their own format which is then entered manually. Each company signs a contract of participation, taking responsibility for the data they deliver. A small participation fee is charged for each product entered but the companies have free access to the database in return. Other users (dietitians, doctors, schools and consumers) pay a small annual fee for use of the database. Data is currently held on over 3500 products, 2500 of which are branded and this is updated almost every week.

Industry collaborates with Nubel because it receives a number of benefits. Scientific feedback on the nutritional values it provides is given and the Scientific Council validates the data, which is distributed to schools and consumers. In addition, their collaboration with Nubel provides a platform for discussion with governmental and consumer groups.

Italy

Food composition data is produced by the Unit of Food Chemistry, at the Italian National Institute for Research on Food and Nutrition (INRAN). Unlike the previous examples where FCDB operators have requested data from industry, INRAN obtains data through in-house chemical analysis. There are various contexts in which this is done:

 

Agreements with producers or consumer organisations; there have been agreements with a number of organisations in Italy to update compositional information on meat products including Italian salamis, bovine meat and poultry. There have also been projects to look at the composition of milk cheese and eggs.

 

Analytical requests by companies; this may be supported by the food industry or food manufacturers’ organisations and tends to focus on specific foods.

 

Analytical requests by official organisations; usually aimed at evaluating typical/ local primary foods. For example, the Ministry of Agriculture requested data on the composition of cherry tomatoes from Sicily.

 

National, European and International research projects with special interests. These include:

·             ‘Markers for the definition and improvement of the quality of the salami from south Italy’, where data on the composition of typical Italian salami was gathered. This was commissioned by the Operative Multiregional Programme (POM) and supported by the EU (1998-2001).

·             ‘Folate: from food to functionality and optimal health’. The folate content of foods was analysed, supported by the EU (2000-2004).

·             ‘Pulses: characterisation, nutritional quality and utilisation of different varieties’. Data on the composition of fresh and dried pulses and new products were obtained. This was supported by the Ministry of Agriculture and the National Research Council of Italy (1980-1998).

·             ‘TRANSFAIR: trans fatty acids and risk factors for cardiovascular disease, gathered data on the fatty acid composition of foods (1995-1998).

 

Once analysis has been done, the foods in question are coded to allow classification, categorisation and grouping. The data is then entered into the database.

 

In addition to the official food composition tables, INRAN has a food labels database. This includes information about the producer, the importer/distributor, characteristics of packaging, nutrients, ingredients additives and other information such as usage, cooking methods, recipes etc. Some of this information is held as a paper archive but most is available on-line.

 

Finland

 

Fineli is the national FCDB in Finland and is maintained by the National Public Health Institute. Details can be found in Finnish, Swedish and English on-line at www.fineli.fi

 

In 2001 it was agreed to create a database of dietary supplements within Fineli. The National Medicines Agency had an electronic database of all supplements regarded as medicines and agreed to share this data. The data were supplied as Excel files, however, there was no standard format between the two databases, making electronic data transfer impossible. The data had to be inputted manually by the database compilers which was very time consuming.

 

The SINFOS data pool (www.sinfos.fi ) was established in 2004. It is part of an international SINFOS database (www.sinfos.com) used in Austria, Denmark, Finland, Germany, Greece, Ireland and Italy, which aims to be active in 30 European countries by 2008. Manufacturer and trade organisations accepted a common model for the database in 2003. All Finnish retailers supply information to SINFOS on ingredients, nutrients and other dietary and allergenic characteristics.

 

Conclusions

Industrially manufactured foods make up a significant proportion of foods consumed in the EU and it is therefore important to have the latest data on their composition in FCDBs.

 

As outlined above, some countries in Europe are already addressing this need within their national FCDBs. There are however, a number of hurdles that these projects have encountered, and these need to be addressed in order to achieve successful data transfer on a Europe-wide basis.

 

It may be difficult to convince manufacturers that it is worth investing time and effort into data exchange so convenient and fast data exchange methods have to be developed and optimised. Experiences of Nubel in Belgium, who found that industry was willing to provide data, as useful feedback and communication links were received in return, is an important example to consider. In this case, industry was also given free access to the online database.

 

Another major difficulty that often occurs in this area is agreeing a common format with which to transfer data. Without this, data must be inputted manually which is very time consuming and resource intensive, thereby rendering the exercise impossible if it were to be completed on a European scale.

 

Collaboration between industry, researchers, government bodies and end users of FCDBs throughout Europe is needed to facilitate up to date FCDBs, with accurate information on the composition of industrially produced foods.

 

More detailed information about the examples of data transfer discussed in this web feature can be found in the report Pilot Cases on Data Transfer from the Food Industry to European National Food Composition Databases’ below.

 

 

 

 

 

 



More Information:

Pilot cases on data transfer from food industry to European national FCDBs Download
*