Number 73 September 1995
MAFF UK - SURVEILLANCE FOR OCHRATOXIN A IN RETAIL COFFEE PRODUCTS
Index to MAFF UK Food Surveillance Information
Sheets, 1995
See also:
28: MAFF UK: Surveillance of UK
Cereals for Ochratoxin A (May 1994)
48: MAFF UK: Surveillance of UK Cereals for
Ochratoxin A (January 1995)
80: MAFF UK: Surveillance of
Ochratoxin A in Green (Unroasted) Coffee Beans (March 1996)
95: MAFF UK - Ochratoxin A in
Cereals and Flour, and Carry-Over into Retail Processed Foods (October
1996)
130: MAFF UK - Survey of
Aflatoxins and Ochratoxin A in Cereals and Retail Products (November
1997)
171: MAFF UK - Survey of
Ochratoxin A in Grain Traded by Central Depots 1997-1998 (February 1999)
172: MAFF UK - A Survey of Human
Exposure to Ochratoxin A (April 1999)
185: MAFF UK - 1998 Survey of
Retail Products for Ochratoxin A (August 1999)
192: MAFF UK - Survey for
Aflatoxins, Ochratoxin A, Fumonisins and Zearalenone in Raw Maize
Summary
A survey for ochratoxin A (OA) in retail coffee products has been
completed. OA was not detected in 19 per cent of all samples. OA was
detected in 64 of the 80 soluble decaffeinated and non-decaffeinated
product samples and in 17 of the 20 roast and ground products. Levels
detected ranged from 0.1 to 8 µg/kg. These results are not cause for
concern. They indicate that coffee products are not a major dietary source
of OA. Estimated intakes of OA from coffee are well within safety limits.
Background
OA is a naturally occurring toxicant, produced by Aspergillus
and Penicillium moulds which grow primarily on cereals. OA
contamination may be found in a number of cereals and cereal products,
some pulses and other foodstuffs, including coffee. OA can also be
detected in food products made from non-ruminant animals exposed to OA
from animal feedingstuffs.
OA has been implicated as a cause of kidney damage and is possibly a
carcinogen. The Committee on Toxicity of Chemicals in Food, Consumer
Products and the Environment (COT) has advised that OA should be
considered a genotoxic carcinogen. Following this advice, the Food
Advisory Committee (FAC) recommended that industry should develop
procedures to reduce OA in cereals to the lowest level technologically
achievable. The European Commission's Scientific Committee for Food (SCF)
also considers OA to be a potent nephrotoxic agent, a carcinogen and that
it has genotoxic properties 1.
The World Health Organisation / Food and Agricultural Organisation Joint
Expert Committee on Food Additives (JECFA) has considered a safety limit
for OA and has set a Provisional Tolerable Weekly Intake for OA of 112
ng/kg bw.
The UK is currently involved in discussions with the European Commission
over possible legislation to establish limits for the presence of
mycotoxins in foods, but these are at a very early stage. MAFF is
generating data on the potential range of dietary sources of consumer
exposure to mycotoxins; the surveillance for OA in coffee is one part of
this work.
MAFF has previously carried out surveillance for OA in a range of
foodstuffs. Results of cereals and porcine product surveillance were
published in Food Surveillance Paper No. 36 (HMSO, 1993). A range of
Chinese and Indian foods, including wheat noodles, cornflour, herbs and
spices, have also been surveyed for a range of mycotoxins including OA and
the results published in Food Safety Information Bulletin No. 53,
September 1994. The results of the most recent surveillance exercise on
cereals, carried out in 1993 and 1994, were published in Food Safety
Information Bulletin No. 57, January 1995.
The aim of this current survey was to determine the levels of OA in UK
retail soluble and roast and ground coffee products so that consumer
dietary exposure to the mycotoxin from these sources could be assessed. A
survey for OA in green coffee beans is currently underway and will be
reported later this year.
Current survey
One hundred samples of various coffees were purchased from a range of
different retail outlets in the UK and analysed for OA. Samples were
obtained in duplicate, the second sample being intended for reference or
as a back up. The survey included 80 granulated, powdered and freeze-dried
soluble products (decaffeinated and non-decaffeinated) and 20 roast and
ground coffee samples (non-decaffeinated only).
Method of Analysis
All analyses were carried out at RHM Technology Ltd, High Wycombe, using
an HPLC method with fluorescence detection. A paper fully describing the
method and its validation is currently in preparation for publication in
the scientific literature.
Spiked samples were used to assess recoveries. Recoveries in the range
70-110 per cent were classed as acceptable. All results were corrected for
recovery. The limit of determination (LOD) was 0.1µg/kg for both
soluble and roast and ground coffees. During repeatability studies, a
relative standard deviation of 5.1 per cent and a mean recovery of 91 per
cent were achieved.
Results
OA was not detected in 20 per cent (16/80) of the soluble coffee samples
(i.e. samples contained < 0.1µg/kg) while the remaining 80 per
cent (64/80) of soluble coffee samples contained OA at levels ranging from
0.1 to 8 µg/kg. OA was detected in 17/20 (85 per cent) roast and
ground coffee samples at levels between 0.2 and 2.1 µg/kg. The other
3 samples of roast and ground coffee did not contain detectable levels of
OA. The results of the survey are given in Table 1.
The highest level of OA (8 µg/kg) was detected in a sample of
powdered, non-decaffeinated, soluble coffee. Four of the nine
decaffeinated soluble product samples contained OA at 0.3 to 2.5µg/kg.
Of these positive samples, 1 sample was granulated and 3 were freeze-dried
products.
Interpretation
The results of this survey show that OA contamination of coffee beans is
not destroyed during roasting, decaffeination or the production of soluble
coffee products. OA was detected in the majority of samples analysed but
the levels found during this survey were low and are not a cause for
concern. These results indicate that coffee products are not a major
dietary source of OA in the UK.
Estimates of OA intake from soluble coffee have been made by assuming
the amount of coffee consumed both by average and high level consumers
(97.5 percentile consumer)2
and are based on the upper bound mean level of contamination in soluble
coffee detected during the survey (see Table 2).
JECFA has set a Provisional Tolerable Weekly Intake (PTWI) for OA of 112
ng/kg bw/week. For a high level adult consumer of soluble coffee in the
UK, the weekly intake of OA from this source is estimated to be 1.9 ng/kg
bw/week which is more than 50 times less than (less than 2 per cent of)
the JECFA PTWI.
References
- European Commission Scientific Committee for Food,
Working Group on Contaminants (12 January 1995) Opinion on
Aflatoxins, Ochratoxin A and Patulin, CS/CNTM/MYC/6 Rev.3.
- Gregory, J., Foster, K., Tyler, H. & Wiseman, M.
(1990). The Dietary and Nutritional Survey of British Adults
(HMSO)
Summary of units
kilogramme (kg) : one thousand grammes
microgramme (µg) : one millionth of a gramme
nanogramme (ng) : one thousand millionth of a gramme (or one thousandth
of a microgramme)
Contact Point
Further information can be obtained from:
Dr David Atkins,
MAFF, Joint Food Safety and Standards Group
Food Contaminants Division
Room 208, Ergon House
c/o Nobel House, 17 Smith Square
London SW1P 3JR
Tel: +44 (0)20 7238 6198
Fax: +44 (0)20 7238 5331
|
Table 1: Summary of results of the survey
for OA in retail coffee products |
| Sample |
Frequency of OA occurrence |
Range of OA
(µg/kg) |
Mean of OA
(µg/kg) |
| Lower bound value |
Upper bound value |
| Soluble products (decaffeinated
and non decaffeinated) |
|
Powder |
11/12 |
0.2-8.0 |
2.2 |
2.2 |
|
Granules |
31/36 |
0.1-4.9 |
0.9 |
0.9 |
|
Freeze-dried |
22/32 |
0.2-3.0 |
0.6 |
0.7 |
| Overall
|
64/80 |
0.1-8.0 |
1.0 |
1.0 |
| Roast
and ground (non-decaffeinated only) |
| Overall |
17/20 |
0.2-2.1 |
0.6 |
0.6 |
| Limit of Determination (LOD) = 0.1 µg/kg.
All results are corrected for recovery. The lower bound mean values were
calculated by assuming that samples < LOD contained OA at 0 µg/kg.
The upper bound mean values were calculated by assuming that samples <
LOD contained OA at the LOD level of 0.1µg/kg. |
|
Table 2: Estimates of intake of OA from
soluble coffee based on average and high level daily consumption of
soluble coffee containing the upper bound mean level of contamination
detected during surveillance (1µg/kg). |
|
Consumption of soluble coffee
(g/person/day) |
Estimated Weekly Intake of OA from soluble coffee
(ng/kg bw/week) |
|
Average consumer |
4.5 |
0.4 |
|
High level consumer (97.5 percentile) |
19.3 |
1.9 |
| Assumptions: An adult weighs
70.1kg (the average weight of the adults taking part in the dietary
survey). Daily consumption is 1/7 weekly consumption. Consumption
figures are based on the weight of soluble coffee used to prepare coffee
brew and do not include consumption of soluble coffee which is supplied
made up (e.g. from vending machines). Consumption data for soluble
coffee were recorded as grammes of 'coffee granules' without distinction
between granules, powder and freeze-dried products. It is assumed that
the term 'coffee granules' was applied to all soluble coffee products
during the recording of this data. |
Return to Index to Surveillance Information Sheets, 1995 page
Go to top of page
These pages were last updated on 1 October 1996
