Acrylamide testing at CFSAN Exploratory survey results Lauren M. Posnick, Sc.D. Public Meeting September 30, 2002.

Acrylamide testing at CFSANExploratory survey results

Lauren M. Posnick, Sc.D.

Public Meeting

September 30, 2002

Outline• Goals of exploratory testing program

• CFSAN method for acrylamide analysis

• Exploratory testing and overall sampling

strategy

• Results of testing to date

• Summary

Goals of exploratory acrylamide testing program

• Initial survey of levels of acrylamide in U.S. foods

• Identify the types of foods that need to be examined in the future

CFSAN acrylamide method

• CFSAN method developed byDr. Steven Musser and colleagues

• Liquid chromatography/tandem mass spectrometry method (LC/MS/MS)

• Method posted on CFSAN website June 20 and updated July 23

CFSAN acrylamide method

• Limit of quantitation (LOQ):– 10 parts per billion (ppb)

• Validation process ongoing– Single-lab validation being completed– Interlaboratory validations being planned– International proficiency testing study

• 33 international laboratories• FDA value within 3.5% of “assigned” value

How did we select samples:Overall Sampling Strategy

1. Starting point

– Foods identified previously

– Determine results for US foods

2. Expand sampling set based on:

– Wide breadth of foods

– Mechanistic insights

– Consumption rates

Overall sampling strategy for U.S. foods

3. Evaluate variability in foods

4. Evaluate exposure in Total Diet Study samples

Where we are now

• FDA is now partway through sampling strategy– Confirmation, breadth, mechanisms

• More than 150 food products

• About 500 analyses

Categories tested• Baby foods• Canned vegetables• Cereals• Chocolate products• Coffee• Cookies• Crackers• French fries

• Gravies and seasonings• Infant formulas• Miscellaneous• Nuts/nut butters• Potato chips• Protein foods

– Meat, Fish, Chicken, Soy• Snack foods (other)

Keep in mind . . .

• These data cover:– A limited number of food categories– A limited number of products in those

categories– A limited number of brands

• These data do not address:– Lot to lot variation– Unit to unit variation

Keep in mind . . .

• We consider the data an initial set of

data.

• We need to test more samples.

What’s in a level?

• Data on acrylamide levels alone do not indicate exposure or risk:

– How many grams of a food are in a serving?

– How much of this food do people eat?

What the graphs will show

• Acrylamide on the vertical axis

• Categories on the horizontal axis

• Each point is an individual result.

Sample graph

010203040506070

Acr

ylam

ide

(pp

b)

Categ

ory A

(4)

Categ

ory B

(3)

What the graphs will show• Some points will

overlap.• The number of

products is given after the category name.

• The categories are grouped on graphs by number of products tested.

Sample graph

010203040506070

Acr

ylam

ide

(pp

b)

Categ

ory A

(4)

Categ

ory B

(3)

What the graphs will show

• Detections below the LOQ of 10 ppb were graphed as 5.

• Non-detects were graphed as 0.

Sample graph

010203040506070

Acr

ylam

ide

(pp

b)

Categ

ory A

(4)

Categ

ory B

(3)

0

200

400

600

800

1000

1200

1400A

cryl

amid

e (p

pb

)

Acrylamide values by food category

Bread

s/ba

kery

(12)Pr

otei

n fo

ods

(12)In

fant

form

ulas

(12)Po

tato

chi

ps (1

6)

Baby

food

s (2

3)

Fren

ch fr

ies (1

6)

0

200

400

600

800

1000

1200A

cryl

amid

e (p

pb

)Acrylamide levels by food category

Cere

als (1

0)

Oth

. sna

cks (1

0)

Gra

v./sea

s. (9

)

Nut

s/bu

tter

s (7

)Cr

acke

rs (7

)

0

200

400

600

800

1000

1200A

cryl

amid

e (p

pb

)

Acrylamide values by food category

Choc

olat

e

(5) Co

okie

s (4

)

Cann

ed V

eg.

(5)

Misce

llane

ous

(5)

Groun

d co

ffee (3

)

Mechanistic research

• FDA is doing mechanistic research

– Time

– Temperature

– Ingredients

0

100

200

300

400

Acr

ylam

ide

(pp

b)

Additional cooking has variable effects on acrylamide levels in purchased bread products

No additionalcooking

Additionalcooking

Summary

• FDA has measured acrylamide levels in a broad range of products.

• Acrylamide was detected in a variety of foods and at a variety of levels. In some foods, no acrylamide was detected.

• We are finding variability between food categories, within food categories, and between different brands.

Conclusions

• We need to test more and different kinds of foods, as specified in the Action Plan.

• Research on mechanisms has begun.

• More mechanistic research is needed, as specified in the Action Plan.