INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY
WORLD HEALTH ORGANIZATION
SAFETY EVALUATION OF CERTAIN FOOD
ADDITIVES AND CONTAMINANTS
WHO FOOD ADDITIVES SERIES: 44
Prepared by the Fifty-third meeting of the Joint FAO/WHO
Expert Committee on Food Additives (JECFA)
World Health Organization, Geneva, 2000
IPCS - International Programme on Chemical Safety
ASSESSMENT OF INTAKE OF SPECIFIC FOOD ADDITIVES
EVALUATION OF NATIONAL ASSESSMENTS OF INTAKE OF ANNATTO EXTRACTS (BIXIN)
First draft prepared by Philippe Verger
French Observatory of Food Consumption, National Agricultural Research
Institute, Paris, France
Introduction
Screening of annatto extracts by the budget method
Assessments of intake of annatto extracts
Assessments based on data on poundage (disappearance)
Assessments based on data from household economic surveys
Assessments based on model diets
Assessments based on individual dietary records
Evaluation of estimates of intake of annatto extracts
Conclusions and recommendations
Bibliography
1. INTRODUCTION
The Committee last examined the toxicity of annatto extracts,
which are food additives used to impart a yellow colour to food, at
its twenty-sixth meeting (Annex 1, reference 59) and allocated an
ADI of 0-0.065 mg/kg bw expressed as bixin. At its present meeting,
the Committee assessed the intake of annatto extracts. National
authorizations are also generally expressed in terms of the primary
chemical colouring agent in the extracts, bixin, which is assumed to
be present at a minimum concentration of 0.2% and a maximum of 3.5% of
the total annatto extracts, with typical concentrations assumed in the
assessments of different countries of 3-3.5%.
Annatto extracts are proposed for use in a wide range of solid
foods in the draft General Standard for Food Additives (GSFA) that is
being developed by the Codex Committee on Food Additives and
Contaminants. Information was received from seven countries:
Australia, Brazil, Canada, France, the Netherlands, the United
Kingdom, and the United States. The assessments were conducted on the
basis of a variety of assumptions about the potential concentrations
of annatto extracts and for various consumption patterns. The
submissions are summarized in Table 1.
2. SCREENING OF ANNATO EXTRACTS BY THE BUDGET METHOD
The Codex GSFA proposes that annatto extracts be permitted for
use in a wide range of solid foods. Table 2 summarizes the data from
the Netherlands, which includes the concentration permitted for use in
Europe, the proportion of the solid food supply likely to contain
annatto, the maximum levels of annatto permitted by GSFA, and a
comparison of these levels to the theoretical maximum level calculated
by the budget method for European countries.
Table 1. Summary of submissions on annatto extracts
Country Budget Poundage FBS/HES/ Model Individual
method data sales data diets dietary
records
Australia ×
Brazil × ×
Canada ×
France × ×
Netherlands × ×
United Kingdom ×
United States × × ×
FBS, food balance sheet; HES, household economic survey; sales, retail stores
The theoretical maximum level of use for annatto in solid foods
was less than the maximum permitted level for the country that
submitted data (25 mg/kg) and was also less than the GSFA level of
1000 mg/kg. A detailed assessment of the intake of annatto from its
use in solid foods is therefore required.
3. ASSESSMENTS OF INTAKE OF ANNATTO EXTRACTS
3.1 Assessments based on data on poundage (disappearance)
Estimates of the amount of annatto extracts available per capita
based on poundage data are given in Table 3 for five countries and
compared with the ADI of 0-0.065 mg/kg bw. The estimated intakes are
less than the ADI, except in Brazil, and vary from 0.005 to 0.14 mg/kg
bw per day (7.6-215% of the ADI). The poundage method shows a
significant increase in the use of this additive in the USA between
1987 and 1996.
3.2 Assessments based on data from household economic surveys
France submitted data from household surveys and data on sales,
and the estimated intake was based on national maximum levels of use
(Table 4). The estimates indicate that the potential intake of annatto
extracts by high consumers (0.034 mg/kg bw per day) represents 52% of
the ADI. Although it is generally not possible to identify high
consumers from household economic surveys, in this submission high
consumption of annatto extracts was estimated by dividing the
household consumption by the number of members of the household in
order to derive the intake of individual consumers.
Table 2. Estimate of a theoretical maximum level for annatto extracts
by the budget method
Country % solid food National GSFA Theoretical
containing maximum maximum maximum
annato extracts level levela level
(mg/kg) (mg/kg) (mg/kg)
Netherlands 25 25 1000 10.4
a The draft General Standard for Food Additives (GSFA) proposes a
maximum use level of 1000 mg/kg for food category 8.1.1.2, mince
of fresh meat, poultry, and game
Table 3. Estimates of intake of annatto extracts based on poundage data
Country Date Consumers Estimated intake % ADIa
(millions) of bixin (mg/kg
bw per day)
Brazil 1997 44.6 (28% of 0.14 215
population)
United Kingdom 1984-86 56 0.0083 12.8
United States 1987 244 Mean, 0.005 7.7
90th percentile, 0.01 15.4
1996 260 Mean, 0.008 12.3
90th percentile, 0.016 24.6
a JECFA ADI, 0-0.065 mg/kg bw
b Assumes that all persons are consumers and that the 90th percentile
intake is twice the mean intake.
3.3 Assessments based on model diets
The USA submitted data based on model diets, details of which are
summarized in Table 5. The model diet was constructed to predict the
intake of annatto by a long-term consumer by using food consumption
data derived from food frequency data for 1982-88 from the Market
Research Corporation of America and average portion sizes from the
1987-88 three-day national food consumption survey of the US
Department of Agriculture. Maximum additive levels were assumed. The
estimates of intake of annatto extracts do not exceed the ADI.
3.4 Assessments based on individual dietary records
Intake estimates for annatto extracts based on individual dietary
records were submitted by six countries. In each case, the mean and
specified percentile estimates were derived for individuals adjusted
for body weight, except in the Canadian assessment. The assumptions
made and the intake estimates derived are summarized in Table 6.
Table 4. Estimates of intake of annatto extracts based on household
economic surveys and sales data
Country Date Survey Assumptions Estimated % ADIa
intake of bixin
(mg/kg bw per day)
France 1995 Sales Maximum European 0.034 52
data Union levels of use,
with adjustment for
catering outside the
home (1994)
a JECFA ADI, 0-0.065 mg/kg bw
The estimates based on individual dietary records and national
use levels indicate intakes of bixin lower than the ADI for mean
consumers in all countries except Australia and Brazil. For Australia,
the estimates were based on the asumption that annatto is used only in
all the products in each food class for which permission has been
granted. For Brazil, the estimate was made for a specific at-risk
population which consumes annatto extracts in large amounts after
direct addition to foods as a condiment; the results do not correspond
to the mean exposure of the total Brazilian population.
For high consumers, the reported intakes were below the ADI in
the USA and about twice the ADI in Brazil, Canada, France, and the
Netherlands. All the countries except Brazil recognized that the
results are overestimates because they assume maximum additive
concentrations in all food categories in which annatto is authorized
for use or might be used. Australia reported intakes that were above
the ADI but recognized that very conservative assumptions were made.
Moreover, the Australian survey is based on a one-day food intake
recall, which would contributes to the overestimate of bixin intake.
Table 5. Estimates of intake of annatto extracts based on model diets
Country Date Survey Assumptions Model Annatto intake % ADIa
(mg/kg bw
per day)
United 1982-88 14-day menu obtained Long-term consumer Long-term consumer
States from MRCA food - maximum additive Mean 0.028 43
frequency data (1982-87) levels (GSFA) 90th percentile 0.73 86
combined with portion - 90th percentile
sizes from USDA/NFCS high consumption
(1987-88); > 2 years level twice mean
consumption
- all respondents
are consumers
- GSFA classification
system
GSFA, General Standard for Food Additives; MRCA, Market Research Corporation of America; USDA/NFCS,
US Department of Agriculture/National Food Consumption Survey
a JECFA ADI, 0-0.065 mg/kg bw
4. EVALUATION OF ESTIMATES OF INTAKE OF ANNATTO EXTRACTS
Screening of additives by the budget method indicated that
annatto extracts require further assessment for use as an additive.
Inclusion of national proportions of the food supply that may contain
annatto extracts did not change this decision.
5. CONCLUSIONS AND RECOMMENDATIONS
The Committee concluded that intake of annatto extracts would
exceed the ADI for bixin if all foods contained annatto extracts at
the levels proposed in the draft General Standard for Food Additives.
Intakes based on national permissions would not exceed the ADI for
most populations. Since estimates of intake that result from use of
the general assumption that all foods in a category are coloured by
the same additive at the maximum level are overestimates, the
Committee recognized that the ADI for bixin is unlikely to be exceeded
as a result of use of annatto extracts. However, information from
Brazil indicated that about 44 million people (28% of the population)
consume annatto seeds directly as a condiment and have done so for
many years, at a level of consumption that is approximately 150% of
the ADI.
Although the results of studies in humans normally take
precedence over those in experimental animals, the submitted reports
were of only limited value. In order to ensure that all of the
relevant data on annatto extrcts have been reviewed, the Committee
recommended their re-evaluation in 2001. The Committee also
recommended that populations with a high intake of annatto extracts
continue to be monitored.
Table 6. Estimates of intake of annatto extracts based on individual dietary records
Country Date Survey Assumptions Model Intake % ADIa
(mg/kg bw
per day)
Australia 1985 National survey; - maximum additive levels Mean intake 0.14 205
24-h recall; (Australia or GSFA) (population = consumers)
adults, > 2 years; - modified GSFA classification Australia
sample, 13 858 system
- maximum additive level Mean intake 0.87 1246
within any one group (population = consumers)
- reports 95th percentile GSFA
consumption
- intakes adjusted for
individual body weight 95th percentile 0.47 672
(consumers) Australia
95th percentile 2.54 3634
(consumers) GSFA
Brazil 1998 Local; food All members of household Mean consumption 0.097- 149-
frequency considered to be consumers 0.103 158
questionnaires;
one-month recall,
202 households,
849 consumers.
High consumers
of 'Colorifico'
Canada 1973 National - maximum additive levels Mean intake of 1.2a 64
- intakes adjusted for all persons 0.042b
individual body weight
- reports 95th, 97.5th
percentile consumption Mean intake in 3.0a 161
- reports by age group worst case 0.105b
Table 6. (continued)
Country Date Survey Assumptions Model Intake % ADIa
(mg/kg bw
per day)
France 1993-94 National; - maximum additive levels (EU) Mean intake (population) EU 0.027 41
7-day record; - adjusted for catering 90th percentile EU 0.052 80
5-75 years; outside the home
sample, 1116 - intakes adjusted for
individual body weight 97.5th percentile EU 0.083 127
- reports 90th, 95th
percentile consumption
- reports by age group Corrected mean intake EU 0.006 9.2
(population)
Netherlands 1992 National; 2-day - maximum additive levels (EU) Mean intake (population) EU 0.028 43
record; 5-65 - intakes adjusted for individual 95th percentile EU 0.104 160
years; sample, body weight
6000 - reports 95th, 97.5th 97.5th percentile EU 0.158 243
percentile consumption
- reports by age group Maximum 0.669
Corrected mean intake < 100
(population) EU
United 1989-92 48 states and Mean intake (total 0.018 28
States Washington DC; population) 29
3-day recall; Mean intake (consumers) 0.019
all age groups;
sample, 11 912 90th percentile (total 0.046 70
population) 0.048 74
90th percentile (consumers)
GSFA, General Standard for Food Additives; EU, European Community
a JECFA ADI, 0-0.07 mg/kg bw; not corrected for bixin
b Assuming 3.5% bixin
6. BIBLIOGRAPHY
Australia-New Zealand Food Authority (1997) Aus-NZ Food Standards
Code, Canberra.
Baines, J. Estimated intake assessment for annatto in Australia.
Personal communication. Australia-New Zealand Food Authority,
Canberra, to FAO, January 1999.
Codex Committee on Food Additives and Contaminants (1997) Report of
the Twenty-ninth Session of the Codex Committee on Food Additives
and Contaminants, Rome, Food and Agriculture Organization of the
United Nations (document ALINORM 97/12A, Appendix V).
Commission of the European Communities (1979) Reports of the
Scientific Committee for Food, Eighth Series, Luxembourg.
European Commission (1995) European Union Directive 95/2/CE,
Brussels
Fisher, C.E. Dietary exposure to the four additives referred to JECFA
by CCFAC. London, Ministry of Agriculture, Fisheries and Food, Joint
Food Safety and Standards Group, submitted to FAO, 22 December 1998.
Rulis, A.M. United State intake data on annatto extract. Food and Drug
Administration, Washington DC, submitted to FAO, 10 December 1998.
Salminen, J.W. Comments about the list of substances scheduled for
evaluation by the JECFA. Health Protection Branch, Ottawa, submitted
to FAO, 16 December 1998.
TNO Nutrition and Food Research Institute (1999) Assessment of the
intake of selected additives with a screening approach, Zeist.
Submitted to FAO by the Ministry of Health, Sport and Welfare, The
Hague, 21 May 1999.
Toledo, C.F. National intake assessment of the intake of bixin in
Brazil. Faculty of Food Enginneering, University of Campinas,
Campinas. Submitted to FAO, 1999.
Verger, P. Estimation of the theoretical maximum intake for certain
food additives in France, Observatory of Food Consumption, Paris.
Personnal communication, submitted to FAO, 1 June 1999.