Cumulative dietary exposure assessment

Introduction

The goal of this exercise is to perform a probabilistic cumulative dietary exposure assessment, illustrating all data needed. In Example 1 we will upload and use nine different files containing the data. In Example 2 we will upload and use a single data file for the same purpose. In the example the exposure will be characterised by upper tail percentiles, and the risk driving substances and foods can be examined. In Example 3 an uncertainty analysis is added.

Preparation

In the workspace browser ( icon), create a new workspace Examples, using the  button in the bottom right corner.

Example 1

Calculate a cumulative chronic dietary exposure for Dutch young adults in 2003 regarding a group of eight triazole substances according to the basic optimistic model of the EFSA 2012 guidance document. Use liver steatosis as a focal effect and Cyproconazole as an index substance. The data files are already available in the data folder Documentation-Examples / Exercise Dietary Exposure Assessment.

Detailed steps are as follows.

• In the Examples workspace, create a new action using the  button in the bottom right corner.

  • Select action type Dietary exposures

    • Name it, e.g. Triazoles exposures

    • (Optional) You can also add tags (e.g. triazoles, NL, steatosis) as labels that can be used later to find similar actions

    • (Optional) You can add a description for further information

    • Click Next

  • Specify Dietary exposures settings

    • Tier: EFSA 2012 Optimistic

    • Risk type Chronic

    • Click Create

You are now directed to the main page of the new action. You can always return to this main page by clicking Action settings  or the action type name (Dietary exposures) in the green bar.

The main page contains at least three blocks of information: Scope, Inputs and Settings. We will now first link all nine data files needed for this cumulative assessment. For most settings we will use default values in accordance with the chosen tier (EFSA 2012 Optimistic).

Scope of the assessment:

• Click Effects (path in the green bar changes Total Dietary exposures / Effects)

  • At Effects data source, click  and browse to the file Effect - Steatosis.xlsx, then click Select

  • At Effect Settings for focal effect select Steatosis-liver and click  Save Changes

  • In the green navigation bar, click Dietary exposures to go up one level.

• Click Foods (path: Dietary exposures / Foods)

  • At Foods data source, click  and browse to the file Foods.xlsx, then click Select

  • In the green navigation bar, click Dietary exposures to go up one level

• Click Populations (optional) (path: Dietary exposures / Populations)

  • At Populations data source, click  and browse to the file Populations.xlsx, then click Select

  • This file contains two populations, only one is allowed. Click  under Populations selection, this opens a pop-up window. Deselect NL_2006, then click Save. The red warning signs  should now be gone. (Note: green warning signs  point at details and can usually be ignored)

  • In the green navigation bar, click Dietary exposures to go up one level.

• Click Substances (path: Dietary exposures / Substances)

  • At Substances data source, click  and browse to the file Substances - Triazoles.xlsx, then click Select

  • At Substance settings for Index substance select Cyproconazole and click  Save Changes

  • In the green navigation bar, click Dietary exposures to go up one level

Next we choose the other input data:

• Click Consumptions by modelled foodd (path: Dietary exposures / Consumptions by modelled food)

  • Click Consumptions (path: Dietary exposures / Consumptions by modelled food / Consumptions)

    • At Consumptions data source, click  and browse to the file FoodConsumptions.xlsx and Select

    • At Consumptions data selection, with  open the food consumption surveys selection.

      • The file contains two surveys, but only one is allowed. Click  under Consumptions data selection, this opens a pop-up window. Deselect VCP-kids, then click Save (the red warning  should now be gone)

    • In the green navigation bar, click Consumptions by modelled food to go up one level

  • Click Food conversions (path: Dietary exposures / Consumptions by modelled food / Food conversions)

    • Click Modelled foods (path: Dietary exposures / Consumptions by modelled food / Food conversions / Modelled foods)

      • Click Concentrations (path: Dietary exposures / Consumptions by modelled food / Food conversions / Modelled foods / Concentrations)

        • At Concentrations data source, click  and browse to the file ConcentrationData.xlsx, then click Select

        • In the green navigation bar, click Food conversions to go up two levels

    • Click Food recipes (path: Dietary exposures / Consumptions by modelled food / Food conversions / Food recipes)

      • At Food recipes data source, click  and browse to the file FoodTranslations.xlsx. then click Select

      • In the green navigation bar, click Dietary exposures to go up three levels

• Click Concentration models (path: Dietary exposures / Concentration models)

  • Click Relative potency factors (path: Dietary exposures / Concentration models / Relative potency factors)

    • At Relative potency data source, click  and browse to the file RPFs.xlsx, then click Select

    • In the green navigation bar, click Dietary exposures to go up two levels

• Click Processing factors (path: Dietary exposures / Processing factors)

  • At Processing factors data source, click  and browse to the file ProcessingFactors.xlsx, then click Select

  • In the green navigation bar, click Dietary exposures to go up one level

• Click Active substances (optional) (path: Dietary exposures / Active substances)

  • In this example we have a fixed list of relative potency factors for the eight substances, and don’t need point of departure (POD) data to decide which substances are active with respect to the health effect and therefore belong to the cumulative assessment group. Deselect the setting “Derive memberships from POD presence”, then click  Save Changes

  • In the green navigation bar, click Dietary exposures to go up one level

Now run the model, either by clicking the  run icon in the grey bar, or by clicking the  run icon in the green bar (Note:  in the green bar can also be used to run subactions on their own).

The  icon is replaced by the text “Running”. When the run has finished, the interface automatically changes to the Results screen. You can also click the Results icon  to go there.

As an exercise, try find the following results:

1. The 99th percentile of cumulative exposure

2. The substance(s) with highest contribution to the total exposure

3. The food(s)-as-measured with the highest contribution to the upper tail of the exposure distribution

Answers:

• In the grey bar, browse to the results panel by clicking the  icon and click on the latest output (path: Results / Dietary exposures)

  • In the Dietary exposures tab, browse in the tree (unfold by clicking  where necessary) to  Dietary exposures  Distribution (OIM)  Percentiles

    • In the table it states that the 99% exposure percentile is at an exposure of 0.02127 µg/kg bw/day.

  • In the Dietary exposures tab, browse in the tree (unfold by clicking  where necessary) to  Dietary exposures  Details  Exposures by substance  Total distribution

    • From the pie chart it is clear that Tebuconazole contributes the most to the total exposure distribution with 32.7%. In the table below the graph more details can be found.

  • In the Dietary exposures tab, browse in the tree (unfold by clicking  where necessary) to  Dietary exposures  Details  Exposures by food and substance  Risk drivers upper tail

    • From the pie chart it is clear that Flusilazole in grapefruit contributes the most (16.7%) to the upper tail exposure distribution

Example 2

We will create a new action to demonstrate uploading all the data at once. All data is now contained within one file, MCRA-Documentation Example Dietary exposures.xlsx.

Detailed steps are as follows.

• In the Examples workspace, create a new action (using )

  • Select action type Dietary exposures

  • Name it, e.g. Triazoles exposures from one data file

  • Click Next

• Specify Dietary exposures settings

  • Tier: EFSA 2012 Optimistic

  • Risk type Chronic

  • Click Create

• Then go to the actions settings  of this action (path: Dietary exposures)

  • Click Effects (path: Dietary exposures / Effects)

    • At Effects data source, click  and browse to the file MCRA-Documentation Example Dietary exposures.xlsx. Click Toggle all, then Select. This will load all available data tables for all subactions of Dietary exposures.

You still need to specify the focal effect (under Effects), index substance (under Substances), and food surveys (under Consumptions by modelled food / Consumptions). You also need to deselect the “Derive memberships from POD presence” setting under Active substances. Navigate to the subaction where these changes have to be made using the green bar.

You now have achieved the same as in Example 1, only with the upload of one single file. You can now run the model, and inspect the results, which should be the same as for Example 1.

Example 3

Repeat the run of the previous task, but in addition to the nominal run, perform an uncertainty analysis as well.

• Click on the  icon (in the grey bar) to open the uncertainty settings panel

  • At Uncertainty settings, check  Perform uncertainty analysis

    • For Monte Carlo iterations per uncertainty run choose 100, and press  Save Changes

• Now run the model, by pressing the  run icon in the grey bar. Note that the run will take much more time.

Compare with the previous results, to find:

1. 95% uncertainty bounds for the 99% exposure percentile

2. 95% uncertainty bounds for the highest contribution from a substance to the total exposure distribution

3. 95% uncertainty bounds for the highest contribution from a food to the total exposure distribution