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
SODIUM IRON ETHYLENEDIAMINE TETRAACETIC ACID (EDTA)
First draft prepared by Dr J.B. Greig
Joint Food Safety & Standards Group, Department of Health, London,
United Kingdom
Explanation
Biological data
Toxicological studies
Acute toxicity
Short-term studies of toxicity
Genotoxicity
Developmental toxicity
Special study of nutrition
Observations in humans
Comments
Evaluation
References
1. EXPLANATION
The sodium iron salt of ethylenediamine tetraacetate (EDTA) was
evaluated by the Committee at its forty-first meeting (Annex 1,
reference 107), which provisionally concluded that sodium iron EDTA
that met the tentative specifications prepared at the meeting would
not present a safety problem when used in supervised food
fortification programmes in iron-deficient populations. The Committee
requested the results of additional studies to assess the site of
deposition of iron administered in this form and to assess the
metabolic fate of sodium iron EDTA after long-term administration. The
Committee emphasized that its evaluation applied only to the use of
sodium iron EDTA as a dietary supplement to be used under supervision,
and expressed its concern about the potential for over-fortification
of food because of the enhanced bioavailability of iron in this form.
Several studies were submitted for consideration at the present
meeting in response to the Committee's request.
2. BIOLOGICAL DATA
2.1 Toxicological studies
2.1.1 Acute toxicity
A summary report of the results of determinations of the LD50 of
sodium iron EDTA in rats and mice was submitted (Sichuan Provincial
Sanitary and Anti-epidemic Station, 1998).
2.1.2 Short-term studies of toxicity
Rats
The experiment described below was conducted according to the
OECD principles of GLP. After a pilot experiment, groups of 40 male
Sprague-Dawley Crl:CD(R) BR rats aged five to six weeks were fed
diets containing iron in the form of sodium ferric EDTA or ferrous
sulfate heptahydrate (FeSO4) to provide iron at concentrations of 35,
70, or 140 mg/kg of diet. The mean daily intakes of iron at the low,
intermediate, and high doses were 2.8, 5.7, and 12 mg/kg bw per day
with FeSO4 and 2.8, 5.7, and 11 mg/kg bw per day with sodium iron
EDTA. The animals were fed the diet for 31 days, at which time 20
animals in each group were killed, while the remaining 20 were
continued on the diet for a further 30 days. A group of 10 untreated
male rats were killed at the start of the experiment to establish the
baseline levels of the parameters that were to be measured. The
toxicolo-gical end-points that were studied included clinical signs,
body weights, organ weights, food consumption, food conversion
efficiency, haematological and clinical chemical end-points, and the
pathological appearancce of the liver, spleen, and any gross lesions.
The distribution of iron was evaluated by determining iron and total
iron-binding capacity in plasma and non-haem iron in the liver,
spleen, and kidneys.
No deaths occurred during the experiment, and the clinical signs
observed are commonly seen in laboratory rats. Occasional
statistically significant changes in food consumption were not
obviously related to treatment. Food conversion efficiency was
variable but showed some small, statistically significant changes
during the first four weeks. The iron content of the diets was within
10% of the intended value in 25 of the 26 batches prepared for both
the pilot and main experiments, but the content of the 26th batch was
13% lower than the intended 70 mg/kg of iron.
Administration of either FeSO4 or sodium iron EDTA resulted in
increased non-haem iron concentrations in the liver, spleen, and
kidney when compared with the baseline values after 31 or 61 days of
feeding, and the changes were related to dose except in the spleen at
61 days. The non-haem iron concentra-tion in these organs resulting
from administration of sodium iron EDTA was always either
significantly lower or not significantly different from the
concentration resulting from administration of FeSO4. The plasma iron
concentration was statistically significantly higher after 31 days of
the low dose of sodium iron EDTA than after the intermediate dose, but
no other significant difference was seen at 31 or 61 days. There were
no statistically significant differences in the total iron-binding
capacity of blood plasma after 32 days of feeding, and the
statistically significant differences seen at 62 days were minor and
unrelated to administered dose or the iron compound. After 31 days of
feeding, there were no statistically significant differences in
terminal body weight or absolute or relative organ weights in treated
animals. At 61 days, the absolute and relative thymus weights of rats
given the low dose of sodium iron EDTA were significantly higher than
those at the low dose of FeSO4 or at the other doses of sodium iron
EDTA. No other significant changes in body weight or organ weight were
observed.
The haematological examinations showed a statistically
significant difference in the mean corpuscular haemoglobin
concentration at 32 days only between rats given sodium iron EDTA and
those given the low dose of FeSO4, but the difference was less than
1.5% and no difference was seen between individual groups treated with
sodium iron EDTA or at 62 days. At 32 days, animals given the high
dose of sodium iron EDTA also had a statistically significantly
increased eosinophil count, but it did not appear to be related to
treatment or dose. No significant changes were found in any group at
62 days. Some inconsistent changes were seen in clinical chemical
end-points in blood plasma collected on days 32 and 62. The activity
of alkaline phosphatase was greater in all groups at day 32 than at
day 62 and was greater in the FeSO4-treated groups than in those
given sodium iron EDTA; although statistically significant differences
were seen, there was no evidence of a dose-response relationship and
the effect had disappeared by 62 days. A statistically significant
increase in total bilirubin occurred at 32 days in the group given the
intermediate dose of FeSO4, but no effect was seen at 62 days.
Statistically significant differences in sodium ion concentration at
32 and 62 days and in chloride ion concentration at 32 days did not
appear to be related to biologically significant changes. An
apparently dose-related decrease in calcium ion concentration in both
treated groups at 62 days was considered by the study authors to be
associated with observed changes in albumin and hence total protein
concentrations. Whereas liver damage might have been a cause of the
decreased plasma albumin concentration, the absence of any other
indicators of liver damage (such as liver-derived plasma enzyme
activities and histological changes) led the study authors to conclude
that the changes were of no toxicological significance.
Histopathological examination of sections of the liver and spleen
stained with haematoxylin and eosin did not reveal any dose-or
treatment-related changes. Such observations as were made were
considered to be unremarkable and occurred either sporadically or with
a uniform distribution across all groups. At 32 days, the accumulation
of iron did not result in a positive Prussian blue reaction in stained
sections of the liver. A general increase was seen in the spleen in
all groups, but the reaction tended to be confined to the red pulp and
was classified as very slight to slight, although increased staining
tended to be found in spleens from the groups fed the doses providing
high concentrations of iron. At 62 days, the Prussian blue staining of
the liver sections did not suggest a dose-or treatment-related effect.
The changes seen in the spleens of all animals generally suggested a
dose-and time-related increase in staining, with no evidence of a
difference between the groups fed FeSO4 and sodium iron EDTA (Appel,
1999).
A summary report of a 90-day study in rats given sodium iron EDTA
in the diet was submitted (Su et al., 1999).
2.1.3 Genotoxicity
A summary report of the results of tests for genotoxicity in
vitro and in vivo was submitted (Sichuan Provincial Sanitary and
Anti-epidemic Station, 1998).
Table 1. Results of assays for the genotoxicity of sodium iron ethylenediamine
tetraacetic acid (EDTA)
End-point Test object Concentration Result
Reverse S. typhimurium TA97a, Not reported Negativea
mutation TA98, TA100, TA102,
TA1535, TA1537, TA1538
Gene mutation Tk locus in mouse 0.03-325 mg/ml Positive and
lymphoma L5178Y cells as Fe cytotoxic with S9;
negative without S9
From Dunkel et al. (1999); sodium iron EDTA of 98% purity was used.
S9, 9000 × g supernatant of rat liver
2.1.4 Developmental toxicity
A summary report of the results of a study of teratogenicity in
rats of sodium iron EDTA was submitted (Sichuan Provincial Sanitary
and Anti-epidemic Station, 1998).
2.1.5 Special study of nutrition
Rats
Eight groups of eight weanling male Sprague-Dawley rats weighing
about 45 g were randomized to receive a zinc-sufficient diet
containing zinc at 30 mg/kg of diet or a zinc-deficient diet
containing zinc at 6.1 mg/kg of diet. Three of the groups in each of
these two blocks were fed a diet containing sodium iron EDTA, while
the diet of the other group contained FeSO4. The iron content of all
diets was 50.1 mg/kg, but two groups fed sodium iron EDTA in each
block received a further dose of 300 or 800 mg/kg diet. After
receiving the diets for 18 days, the animals were transferred to
metabolism cages for three days to allow study of the balance of
calcium, copper, and zinc. Animals that did not eat normally during
this period were removed from the analysis, and, on day 21, all
animals were killed and the right femur removed for analysis of
calcium and zinc.
The rats given the zinc-sufficient diets ate significantly more
food, gained more weight, and had a higher femur zinc content than
those given the zinc-deficient diets. Changing from FeSO4 to sodium
iron EDTA had no significant effect on the zinc-sufficient rats but
resulted in a significant increase in food intake and body-weight
gain. Increasing the EDTA content of the diet significantly but
inconsistently reduced the femur zinc content of the zinc-sufficient
rats, while in the zinc-deficient rats increasing the EDTA content
reduced both food intake and body-weight gain with no significant
change in femur calcium or zinc content.
Zinc absorption (intake minus faecal excretion) and retention
(intake minus (faecal excretion plus urinary excretion)) were both
greater in the rats given the zinc-deficient diet. Fortification with
sodium iron EDTA increased both parameters and also urinary zinc
excretion, although only the latter appeared to be related to the dose
of EDTA. The dietary intake of zinc, the nature of the iron compound
used to supplement the diet, and the concentration of EDTA had no
marked effect on the absorption or excretion of either copper or
calcium except for a small but significant dose-related effect of EDTA
concentration on urinary calcium excretion (Hurrell et al., 1994).
2.1.6 Observations in humans
The effect of various ratios of disodium EDTA to FeSO4 on the
absorption of isotopically labelled FeSO4 was studied in 127 women,
some of whom had depleted iron stores and some of whom were also
anaemic. At molar ratios of EDTA:Fe of 1-0.25, the absorption of iron
from a meal with low iron bioavaila-bility was increased (MacPhail et
al., 1994).
In a random, cross-over study, 10 healthy women aged 23-49 and
weighing 40-70 kg received diets fortified with 10 mg Fe from either
FeSO4 or sodium iron EDTA for 14 days, starting after a period of
menstruation and with an interval of about four weeks. Other than the
addition of sugar, butter, or jam to match the caloric requirements of
each subject, each received a basal diet, and all of the water that
was drunk was ultra-pure. The first five days of each arm of the study
were used for dietary adaptation, and then a six-day chemical balance
study was carried out in the course of which stable isotope-labelled
calcium and zinc were administered.
Significant enhancement of apparent zinc absorption was seen with
sodium iron EDTA as compared with FeSO4 (34 ± 17% (SD) and 21 ±
4.4%). Urinary zinc excretion was also significantly increased, but
the difference in zinc retention was not significant. Absorption of
isotopic calcium was not affected by the iron compound used, although
a small but statistically significant increase in urinary excretion of
isotopic calcium was confirmed in the chemical balance study. Overall,
no significant negative effect of sodium iron EDTA on the uptake and
excretion of zinc was observed (Davidsson et al., 1994).
After an overnight fast, test meals containing 5 mg of Fe as
either FeSO4 or sodium iron EDTA and a radioisotopic 59Mn tracer
were administered to one man and nine women, one of whom was
post-menopausal. The activity of 59Mn excreted in urine was measured
in seven consecutive 24-h samples, and whole-body retention of 59Mn
was measured on days 0, 1, 2, 3, 4, 5, 6, 10, 14, 17, 21, 27, and 28.
On day 28, the second arm of the cross-over study was started with
feeding of the other iron compound and a second tracer dose of 59Mn
in a similar protocol. No statistically significant difference in
manganese retention or urinary excretion was seen with the two iron
compounds (Davidsson et al., 1998).
A double-blind intervention study was carried out in Guatemala
over 32 months to test the efficacy in controlling iron deficiency of
the addition of sodium iron EDTA to sugar. The study was carried out
on a community basis and had been planned to involve two highland and
two lowland communities of populations ranging from 1144 to 1726, with
one community in each area as controls. It had been planned to obtain
an age-and sex-stratified sample of 318 persons from each community
for detailed dietary, parasitological, haematological, and biochemical
studies. The detection of a high prevalence of anaemia in one of the
planned control communities, however, resulted in its conversion to a
treatment community. Additionally, although the initial response rates
in the communities were 54-78%, application of exclusion criteria
resulted in failure to achieve the target basal numbers. There were
additional losses of subjects during the study. The sugar supply of
the treated communities, which was already supplemented with vitamin
A, was further fortified with sodium iron EDTA at 1 g/kg sugar,
equivalent to an Fe concentration of 130 mg/kg sugar. Dietary surveys
during the study showed that the mean (± SD) individual sugar intake
ranged from 35 ± 16 to 38 ± 24 g/day.
Biochemical measures of indicators of iron status--haemoglobin
concentration, saturation of serum iron-binding capacity, free
erythrocyte protoporphyin concentration, serum ferritin concentration,
and iron store mass--were reported from the basal and final surveys.
The changes observed indicated an improvement in iron status in the
treated populations, with a reduction in iron-deficiency
erythropoiesis. The authors reported that forti-fication of sugar with
sodium iron EDTA resulting in the absorption of iron at 0.95-3.1
mg/day had no adverse effects on people of each sex of various ages
(Viteri et al., 1995).
3. COMMENTS
A study specifically designed to address the Committee's concerns
involved feeding male rats diets containing iron in two forms, FeSO4
and sodium iron EDTA, each at three doses, for 62 days. The dietary
concentrations provided iron intakes of 2.8, 5.7, and 11 mg/kg bw per
day (sodium iron EDTA) or 12 mg/kg bw per day (FeSO4). Generally,
there was a dose-related increase in the amount of non-haem iron
stored in liver, spleen, and kidney, but smaller or equal amounts of
iron were taken up from the sodium iron EDTA-containing diet in
comparison with that containing FeSO4. There was no evidence that the
total iron-binding capacity of blood plasma was altered by treatment
with sodium iron EDTA. The Committee therefore concluded that there
was no evidence that administra-tion of iron in the form of sodium
iron EDTA would result in greater uptake of iron than from an
equivalent dietary concentration of FeSO4 once the nutritional
requirement for iron is satisfied. There was no evidence of adverse
effects at the highest daily intake of iron, 11 mg/kg bw, which is 55
times greater than the proposed daily human intake of 0.2 mg/kg bw in
food fortification programmes.
Short-term studies in rats and humans have shown no adverse
effects of dietary intake of sodium iron EDTA on the balance of other
minerals such as calcium, copper, manganese, and zinc. The results of
an intervention study in iron-deficient populations in Guatemala
demonstrated the efficacy of a diet supplemented with sodium iron EDTA
in reducing the prevalence of iron deficiency in humans.
4. EVALUATION
The Committee considered that the data submitted satisfied its
concerns about the use of sodium iron EDTA in food fortification
programmes. It was aware of the results of studies of the acute
toxicity, mutagenicity, teratogenicity, and 90-day toxicity in rats of
sodium iron EDTA. Full reports of these studies were not available,
but the information was not considered necessary for evaluating the
safety of this compound. The Committee also received an assessment of
the potential intake of sodium iron EDTA by consumers in the United
States that would result from fortification of foodstuffs. The
Committee was of the view that this assessment was not relevant to any
proposed use of sodium iron EDTA as a food fortifier in areas of iron
deficiency.
The Committee concluded that sodium iron EDTA could be considered
safe for use in supervised food fortification programmes, when public
health officials had determined the need for iron supplementation of
the diet of a population. Such programmes would provide daily iron
intakes of approximately 0.2 mg/kg bw.
5. REFERENCES
Appel, M.J. (1999) A feeding study to examine the disposition and
accumulation of iron fed as sodium iron EDTA in rats. Unpublished
report No. V99.426 from TNO Nutrition and Food Research Institute,
Zeist, The Netherlands. Submitted to WHO by the International Life
Sciences Institute.
Davidsson, L., Kastenmayer, P. & Hurrell, R.F. (1994) Sodium iron EDTA
[NaFe(III)EDTA] as a food fortificant: The effect on the absorption
and retention of zinc and calcium in women. Am. J. Clin. Nutr., 60,
231-337.
Davidsson, L., Almgren, A. & Hurrell, R.F. (1998) Sodium iron EDTA
[NaFe(III)EDTA] as a food fortificant does not influence absorption
and urinary excretion of manganese in healthy adults. J. Nutr., 128,
1139-1143.
Dunkel, V.C., San, R.H.C., Seifried, H.E. & Whittaker, P. (1999)
Genotoxicity of iron compounds in Salmonella typhimurium and L5178Y
mouse lymphoma cells. Environ. Mol. Mutag., 33, 28-41.
Hurrell, R.F., Ribas, S. & Davidsson, L. (1994) NaFe3+EDTA as a food
fortificant: Influence on zinc, calcium and copper metabolism in the
rat. Br. J. Nutr., 71, 85-93.
MacPhail, A.P., Patel, R.C., Bothwell, T.H. & Lamparell, R.D. (1994)
EDTA and the absorption of iron from food. Am. J. Clin. Nutr., 59,
644-648.
Sichuan Provincial Sanitary and Anti-epidemic Station (1998) Acute
toxicity, mutagenicity and teratogenicity of NaFeEDTA in rats and
mice. Translation of unpublished report submitted to WHO by J. Chen.
Su, Y.L., Chen, Z.L., Zhang, J. & Jiang, Z.C. (1999) A 90-day rat
study on subchronic toxicity of NaFeEDTA. Translation of unpublished
report submitted to WHO by J. Chen.
Viteri, F.E., Alvarez, E., Batres, R., Torún, B., Pineda, O., Mejia,
L.A. & Sylvi, J. (1995) Fortification of sugar with iron sodium
ethylenediaminotetracetate (FeNaEDTA) improves iron status in
semirural Guatemalan populations. Am. J. Clin. Nutr., 61, 1153.