Characterization of the uncertainty in exposure assessment of flavourings

1. Characterization of the uncertainty in exposure assessment of flavourings Max FeinbergINRA. Met@riskfeinberg@inapg.inra.fr

2. 1 Exposures = ai  A  1 Foods Contents qj Weights pi Consumptions  Consumers cij I : individuals J : foods General method to calculate exposure

3. Individualconsumptionscij Weights pi Householdconsumptionsmkj Direct methods Purchasing panels Degradation of precision I : individualsK :householdsJ : foods 1 Food production statistics National budget Dimension “reduction”

4. Exposure assessment methods for flavourings

5. Maximized Survey derived Daily Intake or Per-Capita Times Ten (PCTT) • Correction factors • Annual production volume is corrected by 60%. • 10% population are consumers. Flavour production volume v Correctionfactors  = MSDI Consumerspopulation p MSDI

6. Theoretical Added Maximum Daily Intake • “Maximal” Concentration. Upper Use Level (UUL) • Consumer consumes a fixed amount of flavoured food and beverages that contains the flavouring at its specified level. Max. levels (UUL) Foods  = TAMDI Consumptions TAMDI

7. Possible Average Daily Intake • (“Typical” concentration of the flavour agent within specific food categories)  (Daily mean consumption of those food categories). Median values ? • Proposed by Flavour and Extract Manufacturers’ Association (FEMA, USA). “Typical” conc. Foods  = PADI Consumptions PADI

8. Probabilisticevaluation Knowledge ofconsumption Bestestimate Individualrecords Householdsurveys Pointestimates Nationalsurveys TAMDI MSDI PADI First estimate Productiondata Knowledge ofcomposition Regulatoryvalues Imputed contents Analysed contents Ranking exposure estimates

9. Uncertainty and variability A proposal to assess exposure estimate quality

10. Uncertainty (of measurement): parameter, associated with the result of a measurement, that characterises the dispersion of the values that could reasonably be attributed to the measurand. • Measurand: specific quantity subject to measurement. What is uncertainty?

11. Milieu Manpower Method Material Instrument Accommodation and environmental conditions 5.3 Calibration, method and validation 5.4 Personnel5.2 Result +Uncertainty Traceability5.6 Samples andcalibrants handling5.8 Equipment5.5 Sampling5.7 ISO 17025, §5. Technical requirements Uncertainty for analytical measurement

12. Milieu Manpower Instrument Material Method Sampledpopulation Instrument Interviewer Traceability Expertise Consumption +Uncertainty Amount Ability to recall Naming Characteristics Food Consumer Uncertainty sources for dietary recalls

13. Existence of consumption systems inside dietary patterns. Consumptionsystem Dietaryrecord Variability = + Individual + Uncertainty A more complex model

14. Simulated dietary records

15. Estimation of uncertainty Two approaches

16. T Bias B Z L1 Labo1 Repeatability srInter-labs sLReproducibility L2 Labo 2 L3 Labo 3 Type A. Inter-lab analysis

17. Reproducibility U(Z) = 2 sR U(Z) = 2 sr repeatability Example of an inter-laboratory analysis

18. 35 sR 30 25 20 15 10 5 C g/100g 0 0 10 20 30 40 50 60 Reproducibility and concentration (sugars)

19. 100.00 sR sR = 0.005 C 0.649 10.00 1.00 sRLim = 0.04 C 0.849Horwitz model 0.10 0.01 0 1 10 100 C kg/kg Horwitz’s model

20. Z = f (X1, X2, X3,...). • Z : reported result. • Xi : intermediate results, quantities involved in uncertainty, etc... • f : represents the measuring process but does express a physical law. • Z, X1, X2, X3,... random variables. Type B. Variance propagation theorem

21. = + - Z X X X 1 2 3 = + + 2 2 2 u ( Z ) u ( X ) u ( X ) u ( X ) 1 2 3 c × X X = 1 2 Z X 3 2 2 2 æ ö æ ö æ ö u ( Z ) u ( X ) u ( X ) u ( X ) ç ÷ ç ÷ ç ÷ = + + 3 c 1 2 ç ÷ ç ÷ ç ÷ Z X X X è ø è ø è ø 1 2 3 • Linear combination. • Products and/or ratios combination. Simplifications

22. Chemistry of flavouring agents

23. PolyAromatic Hydro. interlaboratory study

24. Flavouring categories (Directive 88/388). • Natural, natural-identical or artificial flavouring substances, flavouring preparations of plant or animal origin, process flavourings, smoke flavourings. • Flavouring substance is obtained : • by appropriate physical … or enzymatic or microbiological processes, from material of vegetable or animal origin … • by chemical synthesis or isolated by chemical processes (chemically identical)… • by chemical synthesis (not chemically identical). Extended chemical definition

25. 2-methyl butyric acid • S-enantiomer pleasant, sweet, and elegant with a fruity note. • R-enantiomer penetrating, cheesy and sweat-like Biological activity and chirality

26. Hyphenated mass spectrometry coupled to chromatography. • GC-MS (quadrupole). • GC-MSn (ion trap). • LC- MSn (ion trap or TOF). • Near infrared spectroscopy. • GC-FTIR. • High resolution NMR spectrometry. • Sensory analysis coupled to GC. Sophisticated analytical techniques

27. Conclusions

28. Flavourings • Urgent needs for “in-food” measured data. • Develop analytical methods (standardization). • Better define some molecules. • Better relate tolerance and le risk. • Food consumption • Needs for harmonisation of dietary records. • Identification of bias and define traceability. • A proposal: uncertainty to organise harmonisation. • Develop methods to assess digestibility. • A new discipline: “Consumetrics” ? Conclusions

30. Trueness Accuracy Error Precision True value + Trueness and precision