harmonised food description – eurofir experitdtiences · pdf fileharmonised food...
TRANSCRIPT
Harmonised food description – EuroFIR i t d texperiences to date
Paul FinglasEuroFIR Coordinator, Institute of Food Research, UK
Jayne Ireland & Anders MøllerDanish Food Information, DK,
EFSA Colloquium on Food Classification, June 2010
Parma ItalyParma, Italy
Food composition data usesMeasurement of dietary exposures in nutritional epidemiology:Measurement of dietary exposures in nutritional epidemiology:
what are food composition databases needed for?
Food composition databases + Food consumption data
Food intakes Nutrient intakes Biomarkers of diet
Plasma serum urine or
Food group intakes
Food Classification
Nutrient intakes as percent of energy
Plasma, serum, urine or tissue biomarkers
Multivariate analyses
p gy(nutrient density)
Absolute, concentration, or percent estimates
Nutrient food sources/ Nutrient patterns
Dietary patterns
28.06.2010
Evidence‐based Policy Decisions
Paul Finglas - IFR, UK
Importance of food nomenclature & descriptionTo ensure data quality & exchange of dataTo ensure data quality & exchange of data
To understand & compare dietary patterns and p y pnutritional status
Food description – standardised thesauri that areFood description standardised thesauri that are maintained & updated
C l ifi ti t ( tibl ithCommon classification system (compatible with COST 99 & EFCOSUM)
EuroFIR chose LanguaL (multilingual & multicultural ) for indexing foods and linking databases
28.06.2010
databases
Paul Finglas - IFR, UK
What has EuroFIR achieved?Standardisation of:Standardisation of:
Food description and classification
Nutrient definitions
Calculation methods for e g energy and otherCalculation methods for e.g. energy and other derived components
Nutrient retention factorsNutrient retention factors
For a variety of uses.....
28.06.2010 Paul Finglas - IFR, UK
CEN/TC387 Project Committee: “Food Data” (SIS/NFA 2008)Food Data (SIS/NFA, 2008)
Requirements on data structure of FCDBs and interchange of food composition data:
Identification, description, classification & labelling of foods
d i di tand ingredientsValues for measurable, estimated or calculated nutrients/componentsnutrients/componentsMethod specificationsReferences to data sources GS1 liaison will enhance coverage & uptake of future standard with industry
28.06.2010 Paul Finglas - EuroFIR 5
EuroFIR Data Quality Evaluation System for
Quality reflects completeness of documentation and its compliance with
Scientific literature & Analytical reports
Quality reflects completeness of documentation and its compliance with standard requirements
Quality ofQuality of…
30000 AT
Foods LanguaL indexed in EuroFIR (2005-10)
25000
BE BGCHCZ DE
20000
s
DK -DFIDK -DTUESFIFR
15000
Inde
xed
food
s
GRIEILISIT -INRAN
10000
IT INRANIT -CSPOLTLV NLNO
5000NOPLPTSESK
DK
FI
FRSE
02005 2006 2007 2008 2009
Years
SRTRUK
28.06.2010Paul Finglas - IFR, UK
Foods
Foods indexed by LanguaL 29 indexed European data sets
Data set Version Foods
indexed AU UVI 2010-03-21 501 CZ UZEI 2009-08-13 77 LT NNC 2009-08-13 133 DK DFI NDS 2007 09 19 339
Over 29,000 indexed foods
Other indexed data sets:
DK DFI_NDS 2007-09-19 339 CH SwissFIR 2010-03-18 912 SE NFA 2010-05-20 1476 LA FVS 2010-03-20 545 ES UGR 2010-03-18 458
EuroFIR eBASIS
French Polyphenols database
BG NCH 2009-08-13 827 GR HHF 2010-03-18 214 IT INRAN 2010-03-21 790 PL NFNI 2010-03-21 932 BE Nubel 2009 08 13 913GEMS/Foods: WHO/FAO Global
Environmental Monitoring System, using Codex CCPR classification
BE Nubel 2009-08-13 913 IS ISGEM 2010-03-18 946 PT INSA 2010-03-20 962 TR TUBITAK 2010-03-21 337 DE BLS 2010-05-18 1034
INFORMALL (food allergens)
USDA data sets
IE UCC 2009-08-13 938 IT ISPO 2010-03-21 935 SR IMR 2010-03-18 1142 NO MVT 2009-08-13 1188 FR AFSSA 2009 08 13 1294FR AFSSA 2009-08-13 1294 SK FRI 2010-06-09 1400 DK DTU_FDB 2010-03-20 1049 UK IFR 2010-03-21 1702 IL BGU 2010-05-18 1925
28.06.2010 Paul Finglas - IFR, UK
HU UB 1995-06-17 2078 FI FINELI 2010-03-20 2095 NL RIVM-NEVO 2009-09-04 1670
Quality assessment of food indexing: a learning process
ReproducibilityReproducibility
Correctness
28.06.2010 Paul Finglas - IFR, UK
Quality assessment of food indexing: a learning process
Improved English translations of food names (when the LanguaL description was found to be more accurate) :
English food name LanguaL Comment Revised English g gdescriptor
gfood name
Salt, mineral REDUCED SODIUM FOOD
How can salt be d d di
Reduced sodium ltSODIUM FOOD
[P0154]reduced sodium food?
salt
Minced fish COW [B1201] Fish? Fish casseroleMinced fish COW [B1201] Fish? Fish casserole (dish)
Bream common CUSK [B2143] BREAM [B1763] Cusk, raw
http://www.eurofir.net/uploads/documents/D1.8.1.pdf
Bream, common, raw
CUSK [B2143] Brosme brosme
BREAM [B1763] Abramis ballerus
http://www.eurofir.net/uploads/documents/D1.8.1.pdfD1.8.1: Report on food record retrieval using proposed description and classificationD1.8.5 update: Results of Food Indexing by EuroFIR Compilers Network
28.06.2010 Paul Finglas - IFR, UK
Quality assessment of food indexing: a learning process
250
Out of 20,000 foods, 1 000 d d t b
200
revi
ewed
~1,000 needed to be reviewed (2007)
100
150
food
s to
be
CorrectnessReproducibility
Physical state at room temperature (ice cream…)
50
100
Num
ber o
f
0A B C E F G H J K M N P R Z
LanguaL Facet
28.06.2010 Paul Finglas - IFR, UK
Compilers feedback: more information in thesaurus
Available food lists
New descriptors
Synonyms, scientific
namesFoodsFoods
selected
Links to authoritativeChosen
d i t authoritative sources
28.06.2010 Paul Finglas - IFR, UK
descrioptors
Translations to national languages
English (US & UK: LanguaL Steering Committee)
Danish (DFVF, updated 2008)
French (AFSSA, updated 2008)
German (Max Ruber Institut & Graz University 2008)German (Max Ruber Institut & Graz University, 2008)
Hungarian (Budapest, 1995)
It li (INRAN & CSPO 2008)Italian (INRAN & CSPO, 2008)
Spanish (Granada, 2008)
Czech (Institute of Agricultural Economics and Information, ongoing)
Portuguese (ongoing: INSA/Latinfoods)
Dutch (proposed by RIVM)
28.06.2010 Paul Finglas - IFR, UK
Where does this lead?
USDA NUTRIENT DATABASE
FOOD INDUSTRY
EUROFIR FOOD INFORMATION PLATFORM
TOTAL DIET STUDIES
CONTAMINANTS
EuroFIR / UNT / LATINFOODS Workshop on Production of Food Composition Data, Tucumán 2009-06-18
European Food Composition InformationEuropean Food Composition on the Internet
Online Databases
S f
Databases underImplementation
Foods – 51000 (total)Generic - 33000Recipes - 14000
GL
FOIS
Static information(data in table format)
Offline Information(publications)
I f ti P di
Recipes - 14000Branded - 4000
DK
SEFI
FO
LVEE
RU
Information PendingIS
Specialized DatasetsSE
SE
DK
NO
LT
LV
DK
DEDE
CZ
BE
NL
USA BE
PLEI
UA
BY
JPCD
USUS
US
NLUK
LT
PL
DKEuroFIR eBASIS
FR RO
AU
BAHR
SI
MD
GE
HU
FR 2008
AUUS
US
CH
ITSR
Oth li
ATSK
PT
SP
ITSP
AL BUMDIT
ES
Other onlinedatabases
IT 2008TR
EuroFIR / UNT / LATINFOODS Workshop on Production of Food Composition Data, Tucumán 2009-06-18GR
GR
CY
MADZ TN
Image © 2005 EarthSat
June 2010
IL
ESGR
eSearch Prototypeyp
Simultaneous online search in national and specialised FCDBs
Conclusions (1)Both EuroFIR & individual compilers benefit fromBoth EuroFIR & individual compilers benefit from standardized food description, allowing foods to be linked and compared across borders and language barriers;
LanguaL thesaurus has been well accepted and appreciated among compilerspp g p
Training sessions for food indexing have been easy to understand for compilers and other expertsunderstand for compilers and other experts
LanguaL thesaurus has benefited from expertise of compilers (new descriptors scope notes widercompilers (new descriptors, scope notes, wider languages)
28.06.2010 Paul Finglas - IFR, UK
Conclusions (2)EuroFIR repository of nutrient retention factors includesEuroFIR repository of nutrient retention factors includes LanguaL facets allowing linking to cooking methods/ parameters & EuroFIR food groups);
LanguaL is integrated in newly developed Food Composition Data Management System (FoodCASE);Composition Data Management System (FoodCASE);
Next challenge will be to index food consumption surveysNext challenge will be to index food consumption surveys, in order to allow automatic matching of foods through food description – more accurate dietary assessment
28.06.2010 Paul Finglas - IFR, UK
Future – EuroFIR Nexus & AISBLEuroFIR Nexus: EuroFIR AISBL:
To further develop, implement and refine European Food Platform to support research
• Thesauri updating/support• Systems development• Repository updating/support
Platform to support research into food, diet and health:
• Training
Providing access to high quality and validated food composition data;Working with the food industry (especially SMEs) in Europe;Increasing understanding of the value of food data -helping consumers make p ghealthier dietary choices.
A /di i ti /A /di i ti / --Effective management,Effective management,Need to implementNeed to implement
Future opportunities: EPIC & Future opportunities: EPIC & EuroFIREuroFIR? ?
Access/dissemination/ Access/dissemination/ maintenance of EPICmaintenance of EPIC--Soft & related facilitiesSoft & related facilities
Effective management, Effective management, updating and extending updating and extending food composition datafood composition data
-- Laboratory and dataLaboratory and data
Need to implement Need to implement the “missing link”: the “missing link”:
--Food Data HandlingFood Data Handling
(search, re(search, re--order, classify, order, classify, export )export )M lti l i t t fM lti l i t t f
Laboratory and data Laboratory and data analysis best practicesanalysis best practices
export…)export…)
--Food Data LinkageFood Data Linkage
(link, algorithm (link, algorithm ))
Multiple interests for Multiple interests for collecting maximum collecting maximum
details on food details on food consumption dataconsumption data
Access >25 national European Access >25 national European FCDBs d documentation FCDBs d documentation
through a virtual platform of through a virtual platform of i ti t t d ti l dt d ti l dcompilation)compilation)
consumption data consumption data (monitoring surveys, (monitoring surveys,
risk assessment risk assessment studies)studies)
interinter--connected national and connected national and specialised databasesspecialised databases
EPIC-Soft Data Collection Common Lib St d di d d d t il d
EuroFIR Food Data Platform( t d di d d d t dLibrary: Standardised and detailed
consumption data(standardised and documented
composition data)
Provide a full integrated dietary methodology in Europe for data collection
(food consumption) and data handling (nutrient/compounds)
Acknowledgements and Further Information
Funding from Food Quality & Safety Priority (FP6-CT-2005-513944)Priority (FP6-CT-2005-513944)
48 EuroFIR partners8 u o pa t e s
Visit www.eurofir.eu
Contact: Paul Finglas ( l fi l @bb k) d Si([email protected]) and Simone Bell ([email protected])