efsa, 2005, opinion of the scientific panel on dietetic products, hal 373

8/11/2019 Efsa, 2005, Opinion of the Scientific Panel on Dietetic Products, Hal 373

1/134

Nickel

347 http://www.efsa.eu.int/science/nda/nda_opinions/catindex_en.html

OPINION OF THE SCIENTIFIC PANEL ON DIETETIC PRODUCTS,NUTRITION AND ALLERGIES ON A REQUESTFROM THE COMMISSION RELATED TO THE

TOLERABLE UPPER INTAKE LEVEL OF NICKEL(REQUEST N EFSA-Q-2003-018)

(ADOPTED ON 25 JANUARY 2005 BY WRITTEN PROCEDURE)

SUMMARY

The European Food Safety Authority is asked to derive an upper level for the intake of nickel from food

that is unlikely to pose a risk of adverse health effects.

Nickel has not been shown to be essential for humans. Orally ingested nickel salts can cause adverse

effects on kidneys, spleen, lungs and the myeloid system in experimental animals. Furthermore,

perinatal mortality was reported to be increased in the offspring of female rats ingesting nickel salts,

even at the lowest administered dose (1.3 mg nickel/kg body weight/day). While there is evidence that

inhaled nickel salts are carcinogenic in rodents and humans, orally ingested nickel salts have not been

shown to be carcinogenic; however the data presently available are very limited.

Individuals sensitised to nickel through dermal contact and who have allergic contact dermatitis

(estimated to be up to 15% of women but frequently undiagnosed) develop hand eczema from oral,

as well as dermal, exposure to nickel salts. Oral intakes of nickel as low as about 500 g/day (about 8

g/kg body weight/day) have been reported to aggravate hand eczema in nickel sensitised subjects.

In the absence of adequate dose-response data for these effects, it is not possible to establish a

tolerable upper intake level.

The intake of nickel from the average diet is estimated to be about 150 g/day (about 2.5 g/kg body

weight/day), but may reach 900 g/day (about 15 g/kg body weight/day) or more, when large amounts

of food items with high nickel contents are consumed. In addition, first-run drinking water, which may

contain up to 1000 g/L, and leaching from kitchen utensils into food may also contribute to nickel

intake. Intakes of 150 and 900 g/day are about 500 and 90-fold lower, respectively, than the lowest dose

reported to cause adverse effects in rats. Average intakes from food are about one third of the lowest

intake reported to aggravate hand eczema in nickel sensitised subjects.

KEY WORDS

Nickel, tolerable upper intake level, food safety.

BACKGROUND

In 2002, the European Parliament and the Council adopted Directive 2002/46/EC1 related to food

supplements containing vitamins and minerals.

In addition, and as announced in its White Paper on Food Safety, the Commission aims to put forward

a proposal for harmonising legislation concerning the addition of vitamins and minerals to foods.

With a view to provide scientific support to the European Commissions legislative work in this field,

the Scientific Committee on Food (SCF) issued, from October 2000 to April 2003, a series of opinions

on tolerable upper intake levels of individual vitamins and minerals and safety factors in relation to

1 - Directive 2002/46/EC of the European Parliament and of the Council on the approximation of the laws of the Member States relating to food

supplements. OJ L 183. 12.7.2002, p. 51.


2/134

Nickel

http://www.efsa.eu.int/science/nda/nda_opinions/catindex_en.html 348

their use in fortified foods and food supplements (available on the Internet at: http://europa.eu.int/

comm/food/fs/sc/scf/out80_en.html)

The SCF opinions covered 22 out of the 29 nutrients, which were considered to be within their mandatefor this task. The SCF did not have sufficient time to adopt opinions for the following vitamins and

minerals: vitamin C, chloride, fluoride, iron, phosphorus, potassium and sodium. In addition, during

the decision making process for the adoption of Directive 2000/46/EC on food supplements the

Parliament requested that boron, nickel, silicon, vanadium and tin should be allowed to be used in food

supplements. Therefore, the European Food Safety Authority is asked to provide scientific opinions on

the remaining 12 vitamins and minerals in accordance with the present terms of reference.

TERMS OF REFERENCE

With respect to the outstanding 12 vitamins and minerals, the European Food Safety Authority is asked

1) to review the upper levels of daily intakes that are unlikely to pose a risk of adverse health effects; 2)

to provide the basis for the establishment of safety factors, where necessary, which would ensure thesafety of fortified foods and food supplements containing the aforementioned nutrients.

ASSESSMENT

1. INTRODUCTION

Nickel occurs naturally in soil, water, plants and animals. In its compounds, it has normally the valency

state of +2, but valency states of 0, +1, +3, and +4 also exist.

2. NUTRITIONAL BACKGROUND

2.1. Food levels and dietary intake

The nickel concentration is highest in cocoa (8.2-12 mg/kg), soya beans (4.7-5.9 mg/kg), oatmeal

(0.33-4.8 mg/kg), hazelnuts (0.66-2.3 mg/kg), almonds (1.2-1.3 mg/kg) and legumes. The intake from

a Danish average diet was estimated to be 150 g/person/day, but may reach 900 g/person/day or

more, when large amounts of food items with high nickel contents not included in the Danish average

diet are consumed (Flyvholm et al, 1984). The nickel content of a typical Swedish diet assessed by

analysis of market baskets was on average 82 g/day (Becker and Kumpulainen, 1991), while it was

115 g/day (range 70-170 g/day) in duplicate 7-day diets collected from 15 women in the Stockholm

area (Jorhem et al, 1998). Other estimates of 100-146 g/day (Pennington and Jones, 1987) and 20-406

g/day (Dabeka and McKenzie, 1995) for the average nickel intake by adults are in the same range.

The maximum nickel content in running drinking water from different water works in the greater

Copenhagen area was 35 g/L (Andersen et al, 1983). Similarly, maximum concentrations in drinkingand mineral water found in Germany were 34 g/L and 31 g/L, respectively (Scheller et al, 1988). After

remaining in the tap for 8 hours or overnight, however, levels of 490 g/L (Andersen et al, 1983) and

1000 g/L (WHO, 1996) have been reported, respectively.

Release from kitchen utensils can increase the nickel content of food. The average contribution of this

source to the oral intake of nickel is unknown, but could augment dietary exposure by as much as 1 mg/day

(Grandjean et al, 1989, quoted by IARC, 1990).

2.2. Nutritional requirements and recommendations

Nutritional requirements or recommended dietary allowances for nickel have not been established. The

SCF stated explicitly that the data were not sufficiently conclusive to justify setting any recommended

intakes (SCF, 1993).


3/134

Nickel


2.3. Deficiency

In some species, signs of deficiency have been observed. In rats, nickel deficiency was found to be

associated with growth retardation, impaired reproduction function and lower haemoglobin levels. In

humans, however, nickel deficiency has not been demonstrated.

3. BIOLOGICAL CONSIDERATIONS

3.1. Function

Nickel is essential for the catalytic activity of some plant and bacterial enzymes. It is said to influence

iron absorption and metabolism and the haemopoietic process. However, biochemical functions of

nickel have not been demonstrated in humans and higher animals.

3.2. Absorption, distribution, metabolism and excretion

The rate of absorption of nickel salts can be quite high in the fasting state, but is reduced significantlyin the presence of food, such as milk, coffee, tea and orange juice (Solomons et al, 1982). After an

overnight fast, the absorbed nickel averaged 2717% of a dose of nickel sulphate ingested in water

versus 0.70.4% of the same dose ingested in food (Sunderman et al, 1989). In a study with four

healthy adults who were given after an overnight fast a dose of 10 g/kg body weight of the stable

isotope 62Ni, and using ICP-MS, the absorbed amount was 29 to 40% (mean 33.4) of the dose. Plasma

nickel levels rose rapidly within 1.5 to 2.5 hours to peak levels of 15 to 20 g/L and declined by a factor

of >10 during the next 3-4 days. Between 51 and 82% of the absorbed amount was excreted in urine

over five days, but 34.8 13.4% of the absorbed amount was retained after five days (Patriarca et al,

1997). Nickel binds to albumin, histidine and 2-macroglobulin and is widely distributed in the organism.

Transplacental transfer has been demonstrated in rodents. Absorbed nickel is mainly excreted in the

urine, but to a minor extent also in bile and sweat. It is secreted into human milk (Heseker, 2000).

3.3. Normal levels in human tissues and fluid

The total nickel content of the human body is estimated to be 0.5 mg (Heseker, 2000). Highest levels in human

tissues occur in lungs, thyroid, adrenals, and kidneys with concentrations of 173, 141, 132 and 62 g/kg dry

weight, respectively (Rezuke et al, 1987). In non-occupationally exposed men, the mean concentration of

nickel in whole blood and serum is in the range of 1-5 g/L and in urine less than 10 g/L (ECETOC, 1989).

4. HAZARD IDENTIFICATION

This section focuses on the oral toxicity of nickel compounds and on studies particularly important for

the risk assessment of nickel in food. It does not consider data for other routes with the exception of

carcinogenicity. Data on hazards of inhalation and dermal exposure including aspects of occupational

medicine have been reviewed elsewhere (Coogan et al, 1989; ECETOC, 1989; IARC, 1990; WHO, 1991;

WHO, 1996).

4.1. Acute toxicity

The acute oral toxicity of nickel compounds depends on their solubility. The soluble nickel chloride and

nickel sulphate were found to have LD50

values in rats, equivalent to 42-129 mg nickel/kg body weight

(ECETOC, 1989).

4.2. Subacute/subchronic toxicity

30 male rats were administered by gavage 25 mg nickel sulphate/kg body weight/day equivalent to9.5 mg nickel/kg body weight/day over 120 days (10 males as controls). Testis, livers and kidneys

were examined histologically and histochemically. In treated animals severe lesions in the germ cells


4/134

Nickel


particularly in spermiogenesis were observed. Changes in liver and kidneys were detected only rarely

(Waltschewa et al, 1972).

According to Fairhurst and Illing (1987) quoted and reviewed by ECETOC (1989), a number of oralsubacute and subchronic studies have been performed with nickel carbonate, chloride, and sulphate

mainly in rats. In the most relevant of these studies the following effects were found in rats: (a) decreased

body weight gain and slightly lowered haemoglobin levels (nickel chloride in the diet, equivalent to 0

and 20 mg nickel/kg diet, given for 42 days), (b) reduced body weight gain and elevated serum glucose

at all dose levels (nickel chloride in drinking water, equivalent to 0, 2.5, 5 und 10 mg nickel/L, given for

28 days) and (c) lack of weight gain and extensive proliferation of lymphoid cells and histiocytes as well

as micronecrosis in the intestine at the highest dose level (nickel sulphate given by gavage at doses

equivalent to 0, 0.0005, 0.005, 0.05, 0.5 and 5 mg nickel/kg body weight/day for seven months).

Groups of 7 female B6C3F1 mice were exposed to 0, 1, 5, and 10 g nickel sulphate/L drinking water.

Calculated from the water consumption, these concentrations were equivalent to about 0, 116, 286,

and 396 mg nickel sulphate/kg body weight/day or 0, 44, 109, and 151 mg nickel/kg body weight/day.

Water consumption, blood and tissues nickel concentration, body and organ weights, histopathology,

immune responses, bone marrow cellularity and proliferation as well as cellular enzyme activities were

evaluated. Absolute liver weight was significantly decreased in dosed animals and both the absolute

and relative thymus weight was significantly and dose-related reduced, even at the lowest dose. The

kidney was the major organ of nickel accumulation. The primary toxic effects were expressed in the

myeloid system. There were dose-related decreases in bone marrow cellularity, and in granulocyte

macrophage and pluripotent stem-cells proliferative responses (Dieter et al,1988).

Groups of 30 male and 30 female CD rats were administered nickel chloride by gavage at doses of 0,

5, 35, and 100 mg/kg body weight/day for 90 days. In the mid and high dose group, clinical signs of

toxicity were seen, body weights and weights of kidney, liver and spleen were reduced and mortalityincreased. No adverse effects were seen at the dose of 5 mg/kg body weight/day (ABC, 1988).

Groups of 10 Wistar rats of each sex received nickel sulphate in a concentration of 100 mg nickel/L

drinking water for 3 and 6 months. The average oral intake was calculated on the basis of drinking water

consumption to be 6.9 (males) and 7.6 (females) mg nickel/kg body weight/day. Urinary albumin levels

were significantly increased in females exposed for 6 months. The increase observed in males was not

significant due to outliers in the controls. No effect was seen on urinary levels of 2-microglobulin or total

protein. Kidney weights were significantly increased in the exposed animals (Vyskocil et al, 1994).

Groups of 8 male Sprague Dawley rats were given nickel sulphate hexahydrate in drinking water

in concentrations of 0, 0.02, 0.05, and 0.1%, i.e. 0, 44.7, 111.75, and 223.5 mg nickel/L for 13

weeks. At the highest dose, final body weight, plasma total proteins, albumin, globulins, glutamic

pyruvic transaminase activity and the urine volume were significantly decreased and the lymphocyte

subpopulations (T and B cells) suppressed. At the lower dose levels, T and B cells were induced. No

gross or microscopic changes were seen in any of the various tissues examined. However, the relative

liver weights were significantly decreased in the mid and high dose groups and the relative spleen

weights significantly increased in all treated groups as well as the relative kidney weights at both the

lowest and highest dose and the relative lung weight at the highest dose. Alkaline phosphatase activity

in the bronchoalveolar lavage fluid was significantly decreased at any dose level, indicating a significant

decrease in the activity of type II cells in the alveolar space and some early damage to the rat lung

(Obone et al, 1999).

4.3. Chronic toxicity/Carcinogenicity

The carcinogenicity of nickel and nickel compounds has been assessed by several organisations

including IARC (1990, 1999), WHO (1991) and EC (2004).


5/134

Nickel


4.3.1. Animal studies

The data on the carcinogenicity of nickel compounds in experimental animals following oral exposure

are very limited. The chronic toxicity and carcinogenicity of nickel sulphate by oral route has been

studied in rats and dogs (Ambrose et al, 1976). Wistar rats (25 males and 25 females per group) wereexposed to 0, 100, 1000 or 2500 mg/kg nickel in feed for 2 years. Growth was significantly depressed

at 1000 and 2500 mg/kg; increased heart and decreased liver weights were observed only in females

at 1000 and 2500 mg/kg. No neoplasms or other lesions were observed. Beagle dogs (3 males and 3

females per group) were exposed to 0, 100, 1000 or 2500 mg/kg nickel in feed for two years. Growth

was significantly depressed and lung lesions were observed at 2500 mg/kg. No neoplasms were

observed. Both studies have strong limitations because of the low number of animals (rats and dogs),

the high mortality in rats and the limited reporting of the study design and results. The carcinogenicity

of nickel acetate has been tested in three drinking water studies with rats and mice (Schroeder et al,

1964 and 1974; Schroeder and Mitchener, 1975), receiving 0 or 5 mg/L nickel as nickel acetate from

the time of weaning until death. Histological examinations were limited to the lungs, heart, liver, kidneys

and spleen. No increased incidence of neoplasms was observed in either rats or mice. Also these

studies are strongly limited in the design and in the reported results.

Overall, the available data are too limited for an evaluation of the carcinogenic potential of nickel

compounds in rodents following oral administration. According to information supplied by industry

(NiPERA) to EC (2004), a two-year oral carcinogenicity study with nickel sulphate is planned to be ready

in 2005. No data regarding the carcinogenicity of nickel chloride, nickel nitrate, nickel carbonate and

nickel metal by oral route have been found.

Much attention has been directed towards a series of long-term inhalation NTP studies (1996 a, b, c,).

Rats and mice were exposed to aerosols of nickel subsulfide, nickel oxide or nickel hexahydrate for 2

years. It was concluded that there was clear evidence of carcinogenic activity of nickel subsulfide based

on increased incidences of tumours in the lung and in the adrenal medulla. Another inhalation study (78weeks) of nickel subsulfide in rats, also reported an increase in the incidence of lung tumours (Ottolenghi

et al, 1974, quoted by IARC, 1990). No carcinogenic activity was seen in mice. With regard to nickel oxide

there was some evidence of carcinogenicity in rats based on increased incidences of tumours in the lung

and in the adrenal medulla, with equivocal results in female mice. No carcinogenic activity was seen in

rats or mice exposed to nickel sulphate. No studies regarding carcinogenicity of nickel chloride, nickel

nitrate and nickel carbonate following inhalation exposure or intratracheal instillation in experimental

animals have been located. Injection of various nickel compounds in different ways all have caused

malignant tumours, usually sarcomas but also other types, at the site of application (IARC, 1990).

Injection of nickel produces distant tumours of the liver in some strains of mice (IARC, 1990).

Intraplacental exposure to nickel acetate followed by exposure of the offspring to the promoter barbital

in the drinking water produced renal cortical and pelvic tumours (Diwan et al, 1992). Additionally,

pituitary tumours (combined adenomas and carcinomas) were significantly increased in offspring of

both sexes given prenatal nickel acetate only. The results of this study indicate that nickel acetate

is a transplacental initiator of kidney tumours and a complete transplacental carcinogen of pituitary

tumours. A variety of carcinogenicity studies indicate that metallic nickel can produce tumours when

given by intratracheal instillation, subcutaneous, intramuscular or intraperitoneal injection in rats and

hamsters (IARC, 1990).

According to IARC (1999), there is also sufficient evidence in experimental animals for the carcinogenicity

of implants of metallic nickel and for nickel alloy powder containing approximately 66-67% nickel, 13-

16% chromium and 7% iron.

Three studies in experimental animals indicate a possible promoting effect of nickel sulphate,

when applied locally to the nasopharynx or the oral cavity, or by the feed to pups; however,


6/134

Nickel


the indications are rather weak (IARC, 1990; WHO, 1991). In a two-stage carcinogenesis assay,

orally administered nickel chloride in drinking water (600 mg/L) for 25 weeks enhanced the renal

carcinogenicity of N-ethyl-N-hydroxyethylnitrosamine in rats, but not the hepatocarcinogenicity in

rats after initiation with N-nitrosodiethylamine, the gastric carcinogenicity in rats after initiation withN-methyl-N-nitro-N-nitrosoguanidine, the pancreatic carcinogenicity in Syrian golden hamsters

following initiation with N-nitroso-bis (2-oxopropyl)-amine, or the skin carcinogenesis in mice initiated

with 7, 12-dimethylbenzanthracene (IARC, 1990; WHO, 1991). Nickel metal weakly enhanced the lung

carcinogenicity of 20-methylcholanthrene in rats treated by intratracheal instillation (IARC, 1990). A two-

stage carcinogenesis study was carried out in which nickel acetate tetrahydrate (single i.p. injection of

5.3 mg/kg body weight) was tested as a tumour initiator in male rats using sodium barbital (500 mg/L in

drinking water) as the promoter (Kasprzak et al, 1990). Increased incidences of renal cortical adenomas

and combined adenomas and adenocarcinomas were observed.

In conclusion, the available data indicate that nickel sulphate, nickel chloride and nickel metal may have

a promoting effect in combination with selected initiators. There is also some evidence, again limited,

that soluble nickel compounds may act as promoters also by the oral route.

4.3.2. Human data

Several cohort studies of workers exposed by inhalation to various nickel compounds showed an

increased risk of lung and nasal cancer (IARC, 1990). Although the precise compound responsible of

the carcinogenic effects in humans was not always clear, studies indicated that nickel sulphate and

combinations of nickel sulphides and oxides encountered in the nickel refining industry were responsible

for cancer in humans. An additional study had shown that exposure of nickel refinery workers to soluble

nickel compounds alone or in combination with other forms of nickel caused significant excess risks

for lung and nasal cancer and that smoking and nickel exposure had a multiplicative effect (Andersen

et al, 1996). Nickel exposure in mild-steel welders has been associated with tumours (carcinomas) of

the trachea, bronchus and lung in some cases (Simonato, 1991), although subjects were exposed alsoto chromium, which complicated the results.

4.3.3. Overall conclusion

IARC (1990) made an overall evaluation of nickel compounds as a group (Group 1: Human carcinogens),

based on sufficient evidence of epidemiological information, sufficient evidence in experimental animals

and on indications from mechanistic and animal studies that the event responsible for inducing cancer

is generation of ionic nickel at target sites.

Recently, the European Commission Working Group of Specialized Experts in the fields of

carcinogenicity and mutagenicity, has concluded that nickel sulphate, nickel chloride, nickel nitrate and

nickel carbonate, should be considered as human carcinogens by inhalation (Carc. Cat.1, with the risk

phrase R49 May cause cancer by inhalation) (EC, 2004).

The experimental evidence for carcinogenicity of nickel compounds or metallic nickel following oral

exposure is lacking; however, the data presently available are very limited. A long-term study with

nickel sulphate by oral route, currently being carried out, will provide additional information on which to

evaluate this effect. There is also some evidence, although again limited, that soluble nickel compounds

may act as promoters by oral route.

4.4. Genotoxicity

The genotoxicity of nickel and nickel compounds has been reviewed by several organisations including

ECETOC (1989), IARC (1990), WHO (1991), ATSDR (1997) and EC (2004).


7/134

Nickel


4.4.1. In vitro studies

4.4.1.1. Gene mutation

Most of the presently available data comes from studies in bacteria with nickel chloride. In general, the

nickel compounds tested gave negative results in Salmonella enterica var. Typhimurium (nickel chloride,nickel sulphate and nickel nitrate) and Escherichia coli (nickel chloride and nickel sulphate). Only a

fluctuation test gave a weak positive result with nickel chloride in Salmonella enterica var. Typhimurium.

The overall evidence indicates that nickel compounds are not mutagenic in bacteria. Both nickel

chloride and nickel sulphate have been tested in gene mutation tests with different mammalian cells,

many of which with weakly positive results (e.g. in mouse lymphoma and V79 cells). In at least some

of these assays, the positive results were likely due to genetic events other than gene mutations (e.g.

chromosomal aberrations and DNA methylation). For instance, it has been shown that the increase

in mutation frequency at the gptgene of V79 cells (Christie et al, 1992) were due to changes in DNA

methylation (Klein, 1994). DNA methylation seems to be related to the inhibition of tumour suppressor

genes (Costa and Klein, 1999).

4.4.1.2. Chromosomal effects

Chromosomal aberrations (CA) have been extensively studied with nickel chloride and nickel sulphate

in cultured mammalian cells (IARC, 1990; WHO, 1991). Positive results, although weak, were seen

in almost all studies in the range of 0.59-59 mg nickel/L. A weak (1.5 to 2 fold) increase in sister

chromatide exchange (SCE) was also detected at concentrations of 14 and 19 mg/L. Positive results

were also seen with nickel carbonate (CA and SCE in CHO cells). Induction of chromosomal aberrations

were observed also in cultured human lymphocytes with Ni3S

2 (7.3-73 mg/L), NiCO

3 (0.59-59 mg/L)

and NiSO4 (1.1 mg/L). Most aberrations were gaps rather than breaks or fragments. Disturbance of

spindle function was also seen in rat embryo cells (nickel chloride) and human peripheral lymphocytes

(nickel sulphate), with also weak positive results in the micronucleus test (kinetochore stained) in human

diploid fibroblasts, suggesting that numerical chromosome changes (e.g. aneuploidies) might occur.

4.4.1.3. DNA damage and repair

Most data come from studies with nickel chloride and nickel sulphate (IARC, 1990; WHO, 1991). Both

soluble salts induced mitotic gene conversion in yeast, DNA single-strand breaks (SSB) and DNA-

protein cross-links (DPC) in cultured mammalian cells. The formation of DPC may involve non-covalent

association of DNA and chromatin proteins. SSB are repaired quickly while DPC appear to persist.

Nickel ion binds to chromatin more strongly than to DNA. Nickel sulphate induced inhibition of DNA

synthesis/repair.

4.4.1.4. Cell transformation

Nickel sulphate and nickel chloride have been shown to induce cell transformation in Syrian hamster

cells (SHE), BALB 3T3 and C3H10T1/2 mouse cells (IARC, 1990; WHO, 1991).

Soluble nickel (Ni2+) can act as both an initiator and promoter in SHE cells, although it appears to

be a more potent promoter than initiator. Nickel chloride (1 mg Ni/L) produced a ten-fold increase

in morphological transformation in BALB 3T3 mouse cells. Nickel subsulphide (0.1-7 mg/L), nickel

sulphide (0.3-3 mg/L) and nickel oxide (3-23 mg/L) caused dose-dependent increases in the frequency

of transformation in C3H10T1/2 mouse cells, while nickel sulphate and nickel chloride (0.03-6 mg/L)

were negative. Several studies have measured the transforming potential (as anchorage independance)

in human foreskin cells showing positive effects.

Transformation assays do not directly measure genotoxicity per se, but it is generally believed that cell

transformation involves some form of genotoxicity, including alterations in DNA sequence and expression.


8/134

Nickel


4.4.2. In vivo studies

4.4.2.1. Gene mutations

Weakly positive effects have been seen in one study in Drosophila (Wing spot mutation) with nickel

chloride (Ogawa et al, 1994).

No significant increase in mutation frequency was found in the nasal mucosa or lung cells of transgenic

LacZCD2F1 mice or lacI F344 rats following in vivo inhalation exposure for two hours to nickel

subsulphide (Ni3S

2) at dose levels close to the MTD (Mayer et al, 1998). The results do not support any

conclusion regarding the ability of this compound to induce gene mutations in vivo, due to the short

exposure time, as this model needs 4-6 weeks of exposure for maximal expression of mutations.

4.4.2.2. Chromosomal effects

Most of the studies were carried out with nickel sulphate, chloride and nitrate. Chromosomal

aberrations were induced at high doses in bone marrow cells of mice, rats and hamster via oral,

intratracheal or intraperitoneal administration (IARC, 1990; IPCS 1991). No increases in SCE

were found. The data from micronucleus are conflicting, with negative results (Deknudt and

Leonard, 1982; Covance, 2003) and positive results in Indian studies by i.p. (Dhir et al, 1991)

or by oral gavage (Sharma et al, 1987; Sobti and Gill, 1989). Nickel metal was positive in rat

bone marrow cells by intratracheal administration (Zhong et al, 1990). No significant increase

of dominant lethal mutations were reported (Deknudt and Leonard, 1982; Saichenko, 1985).

A significant increase in sperm abnormalities was shown by Sobti and Gill (1989).

4.4.2.3. DNA damage and repair

There is evidence that both soluble and insoluble nickel compounds can produce DNA single strand breaks

and DNA-protein cross-linksin vivoin rat liver cells (IARC, 1990). An inhalation study in mice and in rats by

Bensonet al(2002) has shown DNA single strand breaks in the comet assay after high doses of nickel sulphate

and nickel subsulfide. There was no indication of oxidative damage, although inflammation was evident.The DNA damage therefore most likely was related to inflammation and/or apoptosis. Nickel subsulfide

but not nickel sulphate was able to induce cell proliferation.

4.4.2.4. Human data

The frequency of chromosomal gaps was significantly increased (3-5-fold) in peripheral T-cells of

workers exposed to nickel (Boysen et al, 1980; Waksvik and Boysen, 1982; Waksvik et al, 1981 a and

b). By contrast, no significant increase in the frequency of chromosomal breaks or SCE was found.

Cytogenetic studies conducted in retired nickel workers 4-15 years after employment revealed increases

both in gaps (1.4-fold) and breaks (8-fold) in peripheral T-cells. No difference in the frequency of SCE

was observed (Waksvik et al, 1984 a and b). In another bio-monitoring study CAs were measured in

peripheral lymphocytes of workers occupationally exposed to oxidic nickel and nickel sulphate in a

Czech Republic chemical plant. A significant although small increase (1.6-fold) in the mean value of

CAs (gaps, chromatid and chromosome breaks) was found in the combined exposed group compared

to the control (Senft et al, 1992). Possible confounding factors were not discussed. A cytogenetic

study in peripheral lymphocytes of workers occupationally exposed to nickel carbonyl did not reveal

significant increases of chromosome breaks or gaps (Decheng et al, 1987).

4.4.3. Mechanisms of genotoxicity

The mechanism of the genetic activity of nickel compounds is not clearly defined. Results ofin vitroand

in vivostudies have shown that nickel compounds produce DNA single-strand breaks either directly or

indirectly. The DNA breaks are the most logic candidates for the initial DNA lesions responsible of the

various effects at chromosome level. However, the mechanisms by which DNA breaks are induced are

not clear. One hypothesis involves the generation of oxygen free radicals in a process analogous to

the Fenton reaction. However, oxidative damage by reactive oxygen species is unlikely to play a major

role, due to the observed genetic profile of nickel compounds, with negative results in tests for hprt


9/134

Nickel


mutations and for oxidative damage of DNA. Inhibition of DNA synthesis seems to be more likely, which

is consistent with the negativity in bacteria. Another possible mechanism may involve inihibition of

DNA repair. Impaired repair function has been seenin vitroat low non-cytotoxic nickel concentrations

(Hartwig and Schwerdtle, 2002). Clearly, more research into the mechanisms of nickel mutagenesis andcarcinogenesis are needed.

4.4.4. Comment

There is considerable evidence for thein vitrogenotoxicity of soluble nickel compounds; the database

for nickel carbonate and metallic nickel is much more limited or inadequate. Positive effects are

generally seen in studies of chromosomal effects (CA and SCE), DNA damage and repair (SSB and

DPC) and cell transformation.

The ability of nickel compounds to induce gene mutations is less clear. The weight of evidence

suggests that nickel compounds are unable to efficiently induce point mutations. The few positive

results reported for gene locus mutations (hprtgene) consist of small effects at high, toxic doses and

where the mutations were not characterized at a molecular level. These results are likely to be due

to other genetic events than points mutations (frame-shift or base substitution type mutations), e.g.

chromosomal aberrations and DNA methylation.

Interpretation of the results ofin vivostudies is more complicated. The in vivoclastogenicity of nickel

chloride is the more convincing; however, when taken together all the data presently available for the

three soluble compounds (nickel chloride, sulphate and nitrate), there is evidence ofin vivogenotoxicity

at chromosome level in somatic cells, although this is manifested at high, toxic doses. There is also

supportive evidence from studies on workers exposed by inhalation to nickel compounds, showing

increased frequencies of chromosomal gaps or aberrations.

Evidence for any possible effect on germ cells is particularly limited. There is evidence that the nickelion reaches the testis in rodents after i.p. administration, but there are few data on possible effects.

Recently (EC, 2004) the European Commission Working Group of Specialized Experts in the fields

of carcinogenicity and mutagenicity has proposed the following classification of nickel sulphate,

nickel chloride and nickel nitrate: Mutag. Cat. 3, with the risk phrase R68 Possible risk of irreversible

effect. This conclusion was based on evidence of in vivogenotoxicity in somatic cells, after systemic

exposure. Hence, the possibility that germ cells are affected could not be excluded. However further

testing of effects on germ cells was not considered practicable. There was insufficient evidence for

classification of the mutagenicity of nickel carbonate.

4.5. Reproductive and developmental toxicity

In a 3 generation study, groups of Wistar rats were fed diets containing 0, 250, 500 and 1000 mg/kg

nickel as nickel sulphate hexahydrate, equivalent to about 0, 12.5, 25 or 50 mg nickel/kg body weight/

day. In all groups, 20 rats of each sex were mated. In the first generation, the number of stillborn rats

was increased at all dietary levels. The number of siblings per litter and siblings weaned decreased with

increasing doses. At the highest dose, body weights of weanlings were markedly reduced. No adverse

effects were noted on fertility, gestation, viability and lactation indices. Teratogenic effects were not

observed (Ambrose et al, 1976).

In a 2 generation study, nickel chloride was administered in drinking water to groups of 30 CD rats of

each sex at dose levels of 0, 50, 250 and 500 mg/L, equivalent to 0, 7.3, 30.8 and 51.6 mg/kg body

weight/day, for 90 days before breeding. At the highest dose, there was a significant decrease in the

maternal body weight, along with absolute and relative liver weights. In the F1a

generation, at this

dose the number of live pups/litter was significantly decreased, pup mortality significantly increased


10/134

Nickel


and average pups body weight significantly decreased. In the F1b

litters, increased pup mortality and

decreased live litter size was also observed in the lower dose groups. These effects are questionable,

however, because the room temperature was higher than normal at certain times along with lower

levels of humidity (RTI, 1987).

Groups of 34 female Long-Evans rats received nickel chloride in drinking water for 11 weeks prior to

breeding and during two successive gestation and lactation periods at doses of 0, 10, 50, and 250

mg nickel/L equivalent to 1.33, 6.8 or 31.8 mg nickel/kg body weight/day. At the highest dose in the

first generation and at all doses in the second generation, a dose-related increase in pups born dead

or dying shortly after birth was observed. Body weight gain was reduced in dams of the mid and top

dose groups (Smith et al, 1993).

4.6. Human data

Twenty of 32 electroplating workers, who inadvertently drank water contaminated with nickel sulphate

and chloride, developed symptoms, such as nausea, vomiting, diarrhoea, giddiness, lassitude,

headache, cough and shortness of breath. Laboratory tests showed elevated levels of blood

reticulocytes, urine albumin, and serum bilirubin. The nickel doses that caused these symptoms were

estimated to be in the range of 7.1-35.7 mg/kg body weight (Sunderman et al, 1988).

Nickel salts are potent skin sensitisers in humans, causing allergic contact dermatitis. Nickel ions

bind to cellular and matrix proteins of the skin and induce a cellular immune response (type IV

hypersensitivity reaction) (Bdinger et al, 2000). The prevalence of nickel sensitivity in the population

is about 8-14.5% for adult women and about 1% for men (WHO, 1996). In sensitised individuals, not

only dermal exposure, but also oral intake of low doses can provoke eczema.

In a number of oral challenge studies, single oral doses of a few mg nickel provoked dermal reactions

in nickel-sensitised subjects (Christensen and Mller, 1975; Kaaber et al, 1978; Gawkrodger et al, 1986;Menne and Maibach, 1991). The lowest oral doses, given to nickel sensitive subjects and reported to

exacerbate hand eczema, were 0.49 mg/day in a high nickel diet (Nielsen et al, 1990), equivalent to

about 8 g nickel/kg body weight/day, and 12 g/kg body weight/day given in drinking water on an

empty stomach (Nielsen et al, 1999).

CONCLUSIONS AND RECOMMENDATIONS

1. DOSE RESPONSE ASSESSMENT AND DERIVATION OF A TOLERABLEUPPER INTAKE LEVEL (UL)

In studies on subchronic toxicity, the main targets for the toxicity of orally ingested nickel salts are

kidneys, spleen, lungs, and the myeloid system. In addition, perinatal mortality has been reportedto increase in rats, even at the lowest administered dose of 1.3 mg nickel/kg body weight/day. The

available studies do not allow the establishment of a NOAEL.

There is evidence that nickel salts are carcinogenic in rodents and humans by inhalation. The evidence

for carcinogenicity following oral exposure is lacking, however the data presently available are very

limited. The Panel notes that a long-term study with nickel sulphate in rats by the oral route, which will

improve the presently limited data-base, is on-going.

The genotoxicity of nickel salts, observed at chromosome level at high, toxic doses is likely due to

indirect mechanisms.

It is not possible to derive a threshold for provoking dermal reactions in nickel-sensitised subjects.

Although only dermal exposure to nickel can lead to sensitisation, oral doses of nickel have been shown


11/134

Nickel


to exacerbate hand eczema in nickel-sensitised individuals. In some studies, as little as 8 and 12 g

nickel/kg body weight provoked such reactions.

In the absence of adequate dose-response data for these effects, it is not possible to establish atolerable upper intake level.

2. RISK CHARACTERIZATION

Nickel has not been demonstrated to be essential for humans.

Estimates of nickel intake from the average diet range from 80 to 150 g/person/day, but may

reach 900 g/person/day or more, when large amounts of food items with high nickel contents are

consumed. An intake of 900 g nickel/person/day, equivalent to 15 g/kg body weight/day, would

be 90-fold lower than the lowest dose of 1.33 mg nickel/kg body weight/day reported to increase the

perinatal mortality of rats. In the worst case, however, the first-run drinking water, which remained in

the tap overnight, would be consumed and the release of nickel from kitchen utensils would cause anadditional alimentary exposure. In that case, the margin between the dietary intake and toxic doses

would be considerably lower.

In the group of nickel-sensitised persons, the margin of exposure is even lower. An intake of 150 g

nickel/person/day from the average diet, equivalent to 2.5 g/kg body weight/day in a 60 kg adult, is

about one third of the lowest reported dose of 8 g nickel/kg body weight (490 g/day) able to cause

flare-ups of hand eczema in sensitised subjects. Consumption of food with high nickel content and

additional exposure from first-run drinking water and kitchen utensils could result in an intake higher

than the critical dose.

Any additional nickel intake from supplements would further increase the risk. In this context, the Panel

draws attention to the high prevalence of nickel sensitisation in the population and to the fact, that

many individuals may not be aware that they are sensitised.

3. REFERENCES

ABC (American Biogenics Corporation) (1988). Ninety day gavage study in albino rats using nickel. Final report submitted to US EPA byResearch Triangle Institute and American Biogenics Corporation (quoted by US EPA Integrated Risk Information System).

Ambrose AM, Larson PS, Borzelleca JF, GR Hennigar (1976). Long term toxicologic assessment of nickel in rats and dogs. Journal ofFood Science and Technology, 13: 181-187.

Andersen KE, Nielsen GD, Flyvholm M-A, Fregert S, Gruvberge B (1983). Nickel in tap water. Contact Dermatitis 9: 140-143.

Andersen A, Berge SR, Engeland A, Norseth T (1996). Exposure to nickel compounds and smoking in relation to incidence of lung andnasal cancer among nickel refinery workers. Occup Environ Med, 53: 708-713.

Arrouijal FZ, Hildebrand HF, Vophi H, D Marzin (1990). Genotoxic activity of nickel subsulfide. Mutagenesis 5: 583-589.

ATSDR (Agency for Toxic Substances and Diseases Registry) (1997). Toxicological profile for nickel. US Department of Health and HumanServices, Public Health Service.

Becker W and Kumpulainen J (1991). Contents of essential and toxic mineral elements in Swedish market-basket diets in 1987 66: 151-60.

Benson JM, March TH, Hahn FF, Seagrove JC, Divine KK, Belinsky SA (2002). Final report for short-term inhalation study with nickelcompounds. Study carried out for NiPERA by Inhalational Toxicology Laboratory, Loveloce Research Institute, Albuquerque, NM, USA.

Biedermann KA and Landolph JR (1987). Induction of anchorage independence in human diploid foreskin fibroblasts by carcinogenicmetal salts. Cancer Res 47: 3815-3823.

Boysen M, Waksvik H, Solberg LA, Reith A, Hgtveit AC (1980). Histopathological follow-up studies and chromosome analysis in nickelworkers. In: Nickel toxicology. Brown SS and Sanders FW Jr (Eds). Academic Press. London, UK, pp. 35-38.

Brooks AL and Benson JM (1988). The induction of chromosome aberrations and cell killing in rat lung epithelial cells by nickel compounds.

Environ Mol Mutagen 11: 17.Bdinger L, Hertl M, Bdinger L (2000). Immunologic mechanisms in hypersensitivity reactions to metal ions: an overview. Allergy 55:108-115.


12/134

Nickel


Chorvatovicov D (1983). The effect of NiCl2on the level of chromosome aberrations in Chinese hamster Cricetulus griseus. Biologia

(Bratislava) 38: 1107-1112 (in Slovak).

Christensen OB and Mller H (1975). External and internal exposure to the ant igen in the hand eczema of nickel allergy. Contact Dermatitis1: 136-141.

Christie NT and Tummolo DM (1987). Synergy between Ni(II) and benzo(a)pyrene in the cell transformation assay using Syrian hamsterembryo cells. Proc Am Assoc Cancer Res 28: 121.

Christie NT, Tummolo DM, Klein CB, Murphy EC Jr (1988). The role of Ni(II) in mutation. In: Nickel and human health: Current Perspectives.Advances in Environmental Science and technology. Nieboer E, Nriagu JO (Eds). John Wiley & Sons, New York, pp 305-307.

Ciccarelli RB and Wetterhahn KE (1982). Nickel distribution and DNA lesions induced in rat tissues by the carcinogen nickel carbonate.Cancer Res 42: 3544-3549.

Conway K, Wang XW, Xu L, Costa M (1987). Effects of magnesium on nickel-induced genotoxicity and cell transformation. Carcinogenesis(London) 8: 115-121.

Coogan TP, Latta DM, Snow ET, Costa M (1989). Toxicity and carcinogenicity of nickel compounds. CRC Critical Reviews in Toxicology19: 341-384.

Costa M and Klein CB (1999). Nickel carcinogenesis, mutation, epigenetics, or selection. Perspectives. Editorial. Environ Health

perspectives 107: 438-439.Costa M, Heck JD, Robison SH (1982). Selective phagocytosis of crystalline metal sulphide particles and DNA strand breaks as amechanism for the induction of cellular transformation. Cancer Res 42: 2757-2763.

Covance (2003). In vivo rat micronucleus assay with nickel sulphate hexahydrate. Study N 7454-100, submitted to NiPERA, 4 August2003. Covance Labs. Incorp. USA.

Dabeka RW and McKenzie AD (1995). Survey of lead, cadmium, fluoride, nickel, and cobalt in food composites and estimation of dietaryintakes of these elements by Canadians in 1986-1988. J AOAC International 78: 897-909.

Decheng C, Jin M, Han L, Wu L, Xie S, Zheng X (1987). Cytogenetic analysis in workers occupationally exposed to nickel carbonyl. MutatRes 188: 149-152.

Deknudt G and Leonard A (1982). Mutagenicity tests with nickel salts in the male mouse. Toxicology 25: 289-292.

Dhir H, Agawal K, Sharma A, Talukder G (1991). Modifying role of Phyllantus embilica and ascorbic acid against nickel clastogenicity inmice. Cancer Letter (Shannon, Ireland) 59: 9-18.

Di Paolo JA and Casto BC (1979). Quantitative studies of in vitro morphological transformation of Syrian hamster cells by inorganic metalsalts. Cancer Res 39: 1008-1013.

Dieter MP, Jameson CW, Tucker AN, Luster MI, French JE, Hong HL, Boorman GA (1988). Evaluation of tissue disposition, myelopoietic, andimmunologic responses in mice after long-term exposure to nickel sulphate in the drinking water. J Toxicol Environm Health 24: 357-372.

Diwan BA, Kasprzak KS, Rice JM (1992). Transplacental carcinogenic effects of nickel (II) acetate in the renal cortex, renal pelvis, andadenohypophysis in F 344/Ncr rats. Carcinogenesis 13: 1351-1357.

ECETOC (European Centre for Ecotoxicology and Toxicology of Chemicals) (1989). Nickel and nickel compounds: review of toxicologyand epidemiology with special reference to carcinogenesis. Technical Report N 33.

EC (European Commission) (2004). Working Group of Specialized Experts in the fields of Carcinogenicity and Mutagenicity. Nickel.Summary Record. European Chemicals Bureau.

Fairhurst S and Illing HPA (1987). The toxicity of nickel and its inorganic compounds. UK Health and Safety Executive, Advisory Committee

on Toxic Substances.Flyvholm MA, Nielsen GD, Andersen A (1984). Nickel content of food and estimation of dietary intake. Z Lebensm Unters Forsch 179:427-431.

Gawkrodger DJ, Cook SW, Fell GS, Hunter JAA (1986). Nickel dermatitis: the reaction to the oral nickel challenge. Br J Dermatol 115: 33-38.

Grandjean P, Nielsen GD, Andersen O (1989). Human nickel exposure and chemobiokinetics. In: Maibach HI, Menne T (Eds.). Nickel andthe skin: Immunology and toxicology. Boca Raton, CRC Press.

Heseker H (2000). Nickel. Funktionen, Physiologie, Stoffwechsel und Versorgung in der Bundesrepublik Deutschland. Ernhrungs-Umschau 47: 483-484.

Howard W, Leonard B, Moody W, Kochhar TS (1991). Induction of chromosome changes by metal compounds in cultured CHO cells.Toxicol Lett 56: 179-186.

IARC (International Agency for Research on Cancer) (1990). Monographs on the evaluation of carcinogenic risks to humans. Volume 49.Chromium, nickel and welding. IARC, Lyon, France.

IARC (International Agency for Research on Cancer) (1999). Monographs on the evaluation of carcinogenic risks to humans. Volume 74.Surgical implants and other foreign bodies. IARC, Lyon, France.


13/134

Nickel


Jorhem L, Becker W, Slorach S (1998). Intake of 17 elements by Swedish women, determined by a 24-hour duplicate portion study.J Food Comp Anal 11: 32-46.

Kaaber K, Veien NK, Tjell JC (1978). Low nickel diet in the treatment of patients with chronic nickel dermatitis. Br J Dermatol 98: 197-201.

Kasprzak KS, Diwan BA, Konishi N, Misra M, Rice J (1990). Initiation by nickel acetate and promotion by sodium barbital of renal corticalepithelial tumours in male F 344 rats. Carcinogenesis 11: 647-652.

Klein CB (1994). Epigenetic control of GPT expression in G12 cells. Environ Mol Mutagen 23: 31.

Larramendy ML, Popescu NC, Di Paolo JA (1981). Induction by inorganic metal salts of sister chromatid exchanges and chromosome aberrationsin human and Syrian hamster cells. Environ Mutagen 3: 597-606.

Lechner JF, Tokiwa T, McClendon IA, Haugen A (1984). Effects of nickel sulphate on growth and differentiation of normal human bronchialepithelial cells. Carcinogenesis 5: 1697- 1703.

Lin X, Sugiyama M, Costa M (1991). Differences in the effects of vitamin E on nickel sulphide or nickel chloride-induced chromosomalaberrations in mammalian cells. Mutat Res 260: 159-164.

Little JB, Frenial JM, Coppey J (1988). Studies of mutagenesis and neoplastic transformation by bivalent metal ions and ionizing radiation.Teratog Carcinog Mutagen 8: 287-292.

Maehle L, Metcalf RA, Ryberg D, Bennett WP, Harris CC, Haugen A (1992). Altered p53 gene structure and expression in human epithelial

cells after exposure to nickel. Cancer Res 52: 218-221.

Mayer C, Klein RG, Wesh H, Schmetzer P (1998). Nickel subsulfide is genotoxic in vitro but shows nu mutagenic potential in respiratorytract tissues of Big BlueTMrats and MutaTMMouse mice in vivo after inhalation. Mutat Res 420: 85-98.

Menne T and Maibach HI (1991). Systemic contact-type dermatitis. In: Dermatotoxicology. Marzulli FN, Maibach HI (Eds.). 4 thEdition.Hemisphere Publishing Corporation, London.

Miura T, Patierno SR, Sakuramoto T, Landolph JR (1989). Morphological and neoplastic transformation of C3H10T Cl8 mouse embryocells by insoluble carcinogenic nickel compounds. Environ Mol Mutagen 14: 65-78.

Mohanty PK (1987). Cytotoxic effect of nickel chloride on the somatic chromosomes of Swiss albino mice. Current Science 56: 1154-1157.

Montaldi A, Zentilin L, Zordan M, Bianchi V, Levis AG, Clonfero E, Paglialunga S (1987). Chromosomal effects of heavy metals (Cd, Cr, Hg,Ni and Pb) on cultured mammalian cells in the presence of nitrilotriacetic acid (NTA). Toxicol Environ Chem 14: 183-200.

Montaldi A, Zentilin L, Venier P, Gola I, Bianchi V, Paglialunga S, Levis AG (1985). Interaction of nitralotriacetic acid with heavy metals in

the induction of sister chromatid exchanges in cultured mammalian cells. Environ Mutagen 7: 381-390.Nielsen GD, Jepsen LV, Jorgensen PJ, Grandjean P, Brandrup F (1990). Nickel-sensitive patients with vesicular hand eczema: oral challengewith a diet naturally high in nickel. Br J Dermatol 122: 299-308.

Nielsen GD, Soderberg U, Jorgensen PJ, Templeton DM, Rasmussen SN, Andersen KE, Grandjean P (1999). Absorption and retention ofnickel from drinking water in relation to food intake and nickel sensitivity. Toxicol Appl Pharmacol 154: 67-75.

NTP (National Toxicology Program) (1996a). Toxicology and carcinogenesis studies of nickel oxide in F 344 rats and B6C3F1 mice(inhalation studies). Technical Report series TR 451.

NTP (National Toxicology Program) (1996b). Toxicology and carcinogenesis studies of nickel subsulfide in F 344 rats and B6C3F1 mice(inhalation studies). Technical Report Series TR 453.

Obone E, Chakrabarti SK, Bai C, Malick MA, Lamontagne L, Subramanian KS (1999). Toxicity and bioaccumulation of nickel sulfate inSprague-Dawley rats following 13 weeks of subchronic exposure. J Toxicol Environm Health A 57: 379-401.

Ogawa HI, Shibahara T, Iwata H, Okada T, Tsuruta S, Kakimoto K, Sakata K, Kato Y, Ryo H, Itoh T, et al(1994). Genotoxic activities in vivoof cobaltous chloride and other metal chlorides as assayed in the Drosophila wing spot test. Mutat Res 320: 133-140.

Ohno H, Hanaoka F, Yamada M (1982). Inducibility of sister chromatid exchanges by heavy metal ions. Mutat Res 104: 141-145.

Ottolenghi AD, Haseman JK, Payne WW, Falk HL, Mac Farland HN (1974). Inhalation studies of nickel sulfide in pulmonary carcinogenesisin rats. J Natl Cancer Inst 54: 1165-1172.

Patriarca M, Lyon TDB, Fell GS (1997). Nickel metabolism in humans investigated with an oral stable isotope. Am J Clin Nutr 66: 616-62.

Pennington JAT and Jones JW (1987). Molybdenum, nickel, cobalt, vanadium, and strontium in total diets. J American Dietetic Assoc 87:1644 - 1650.

Rezuke WN, Knight JA, Sunderman FW (1987). Reference values for nickel concentrations in human tissues and bile. American J IndustrialMed 11: 419 - 426.

Rivedal E and Sanner T (1981). Metal salts as promoters of in vitro morphological transformation of Hamster embryo cells initiated withbenzo(a)pyrene. Cancer Res 41: 2950-2953.

RTI (Research Triangle Institute) (1988). Two generation reproduction and fertility study of nickel chloride administered to CD rats in thedrinking water: Fertility and reproductive performance of the Po generation and F1 generation. Final study report submitted to US EPAby Research Triangle Institute.


14/134

Nickel


Saichenko SP (1985). Experimental evaluation of genetic danger of metals administered with drinking water. In: Domnin S.G. &Shcherbakov S.V. (Eds). Problems of labour hygiene in steel and coloured metals industry. Moscow, Erisman Institute of Hygiene.

Saxholm HJK, Reith A, Brogger A (1981). Oncogenic transformation and cell lysis in C3H10T cells and increased sister chromatidexchanges in human lymphocytes by nickel subsulfide. Cancer Res 41: 4136-4139.

SCF (Scientific Committee on Food) (1993). Nutrient and energy intakes for the European Community. Reports of the Scientific Committeeon Food, Thirty first series. EC, Luxembourg.

Scheller R, Strahlmann B, Schwedt G (1988). Lebensmittelchemische und -technologische Aspekte zur nickelarmen Ernhrung beiendogen bedingten allergischen Kontaktekzemen. Hautarzt 39: 491 - 497.

Schroeder HA, Balassa JJ, Vinton WHJr (1964). Chromium, lead, cadmium, nickel and titanium in mice: effect on mortality, tumours andtissue levels. J Nutr 83: 239-250.

Schroeder HA, Mitchener M, Nason AP (1974). Life-time effects of mercury, methylmercury, and nine other trace metals on mice. J Nutr104: 239-243.

Schroeder HA and Mitchener M (1975). Life-time effects of nickel in rats: survival, tumors, interaction with trace elements and tissue levels.J Nutr 105: 452-458.

Sen P and Costa M (1985). Induction of chromosomal damage in Chinese hamster ovary cells by soluble and particulate nickel

compounds: preferential frequentation of the heterochromatic long arm of the X-chromosome by carcinogenic crystalline NiS particles.Cancer Res 45: 2320-2325.

Sen P, Conway K, Costa M (1987). Comparison of the localization of chromosome damage induced by calcium chromate and nickelcompounds. Cancer Res 47: 2142- 2147.

Senft V, F Loan, Tucek M (1992). Cytogenetic analysis of chromosomal aberrations of peripheral lymphocytes in workers occupationallyexposed to nickel. Mutat Res 279: 171-179.

Sharma GP, Sobti RC, Chandry A, Ahluwalia KK, Gile RK (1987). Effect of some nickel compounds on the chromosomes of mice andmosquitos. La Kromosomo II 45: 1423-1432.

Simonato L, Fletcher AC, Andersen A, Anderson K,Becker N, Changclaude J, Ferro G, Gerin M, Saracci R (1991). A historical prospective-study of European stainless-steel, mild-steel and shipyard welders. Br J Ind Med 48: 145-154.

Singh I (1984). Induction of gene conversion and reverse mutation by manganese sulphate and nickel sulphate in S. cerevisiae. MutatRes 137: 47-49.

Smith MK, George EL, Stober JA, Feng HA, Kimmel GL (1993). Perinatal toxicity associated with nickel chloride exposure. Environm Res61: 200-211.

Sobti RC and Gill RK (1989). Incidence of micronuclei and abnormalities in the head of spermatozoa caused by the salts of a heavy metalnickel. Cytologica 54: 249-254.

Solomons NW, Viteri F, Shuler TR, Nielsen FH (1982). Bioavailability of nickel in man: Effects of foods and chemically-defined dietaryconstituents on the absorption of inorganic nickel. J Nutr 112: 39-50.

Sunderman FW Jr, Dingle B, Hopfer SM, Swift T (1988). Acute nickel toxicity in electroplating workers who accidently ingested a solution of nickelsulfate and nickel chloride. American J Industrial Med 14: 257-266.

Sunderman FW Jr, Hopfer SM, Sweeney KR, Marcus AH, Most BM, Creason J (1989). Nickel absorption and kinetics in human volunteers.Proc Soc Exp Biol Med 191: 5-11.

Vyskocil A, Viau C, Cizkova M (1994). Chronic nephrotoxicity of soluble nickel in rats. Human Exp Toxicol 13: 689-693.

Waksvik H, Boysen M (1982). Cytogenetic analysis of lymphocytes from workers in a nickel refinery. Mutat Res 103: 185-190.Waksvik H, Boysen M, Hgtveit A (1984a). Chromosome aberrations and sister chromatid exchanges in retired nickel refinery workers.Mutat Res 130: 250-251.

Waksvik H, Boysen M, Hgtveit AC (1984b). Increased incidence of chromosomal aberrations in peripheral lymphocytes of retired nickelworkers. Carcinogenesis (London) 5: 1525-1527.

Waksvik H, Boysen M, Broegger A, Klepp O (1981a). Chromosome aberrations and sister chromatid exchanges in persons occupationallyexposed to mutagens/carcinogens. In: Seeberg RE and Kleppe K (Eds). Chromosome damage and repair. Proceedings of a NATOAdvanced Study, Institute and an EMBO lecture course; May-June, 1980, Bergen, Sweden, New York, Plenum Press; pp. 563-566.

Waksvik H, Boysen M, Broegger A, Saxholm H, Reith A (1981b). In vivo and in vitro studies of mutagenicity and carcinogenicity of nickelcompounds. Mutat Res 85: 250-251.

Waltschewa W, Slatewa M, Michailow I (1972). Hodenvernderungen bei weien Ratten durch chronische Verabreichung von Nickelsulfat.Exp Path 6: 116-120.

WHO (World Health Organisation) (1991). Nickel. Environmental Health Criteria 108. Geneva.

WHO (World Health Organisation) (1996). Guidelines for drinking water quality. Volume 2: Health criteria and other supporting information.Geneva.


15/134

Nickel


Zhang A and Barrett JC (1988). Dose-response studies of nickel-induced morphological transformation of Syrian hamster embryofibroblasts. Toxicol In vitro 2: 303-307.

Zhong BZ, Li ZA, Ma GY, Wang BS (1990). Study on mutagenesis and carcinogenesis of productive nickel dust. In: Mendelsohn & Albertini(Eds). Mutation and the environment. Part E. Proceedings of the fifth international conference on environmental mutagens, July 1989.

Wiley Liss Inc., pp. 41-46.

Zhong BZ and Li ZQ (1989). Study of the mutagenicity and carcinogenicity of produced nickel dust. Environ Mol Mutagen 14: 230.

PANEL MEMBERS

Wulf Becker, Francesco Branca, Daniel Brasseur, Jean-Louis Bresson, Albert Flynn, Alan A. Jackson,Pagona Lagiou, Martinus Lvik, Geltrude Mingrone, Bevan Moseley, Andreu Palou, HildegardPrzyrembel, Seppo Salminen, Stephan Strobel, Henk van den Berg, and Hendrik van Loveren.

ACKNOWLEDGEMENT

The Scientific Panel on Dietetic Products, Nutrition and Allergies wishes to thank Jan Alexander, AngeloCarere, Werner Grunow, Andrew Renwick and Gerrit Speijers for their contributions to the draft opinion.


16/134

Nickel



17/134

Fluoride


OPINION OF THE SCIENTIFIC PANEL ON DIETETIC PRODUCTS,NUTRITION AND ALLERGIES ON A REQUESTFROM THE COMMISSION RELATED TO THE

TOLERABLE UPPER INTAKE LEVEL OF FLUORIDE(REQUEST N EFSA-Q-2003-018)

(ADOPTED ON 22 FEBRUARY 2005)

SUMMARY

Fluoride is not essential for human growth and development but is beneficial in the prevention of dentalcaries (tooth decay) when ingested in amounts of about 0.05 mg/kg body weight per day and whenapplied topically with dental products such as toothpaste. Dental enamel which contains fluoride isless likely to develop caries, because of greater resistance to ingested acids or to acids generated fromingested sugars by the oral bacteria. In addition, fluoride inhibits sugar metabolism by oral bacteria.

Fluoride content of the body is not under physiological control. Absorbed fluoride is partly retained in bone

and partly excreted, predominantly via the kidney. In infants retention in bone can be as high as 90% ofthe absorbed amount, whereas in adults retention is 50% or less. Fluoride is also incorporated into dentalenamel during tooth formation.

Excessive intake of fluoride during enamel maturation before tooth eruption from birth to eight years of age,when enamel formation is complete, can lead to reduced mineral content of enamel and to dental fluorosisof deciduous but predominantly of permanent teeth. The incidence and severity of dental fluorosis isdose-dependent. Mild dental fluorosis is not readily apparent and is associated with increased resistanceto caries. The Panel considered moderate dental fluorosis, which is characterised by staining and minutepitting of teeth, to be an adverse effect. On the basis that the prevalence of moderate dental fluorosis ofpermanent teeth is less than 5% in populations ingesting 0.08-0.12 mg fluoride/kg body weight/day, thePanel considered that the upper level (UL) for fluoride is 0.1 mg fluoride/kg/day in children aged 1-8 years.This is equivalent to 1.4 and 2.2 mg fluoride per day in children aged 1-3 years and 4-8 years,

respectively.

Fluoride accretion in bone increases bone density but excessive long term intake reduces bone strengthand increases risk of fracture and skeletal fluorosis (stiffness of joints, skeletal deformities). Studies withtherapeutic oral administration of fluoride in amounts of 0.6 mg/kg body weight/day in postmenopausalwomen over several years increased the risk for non-vertebral bone fractures significantly. The Panelapplied an uncertainty factor of 5 to derive an UL of 0.12 mg/kg body weight/day. This is equivalentto an UL of 5 mg/day in children aged 9-14 years and 7 mg/day for age 15 years and older, includingpregnant and lactating women.

The UL for fluoride applies to intake from water, beverages, foodstuffs, including fluoridated salt, dentalhealth products and fluoride tablets for caries prevention.

Children aged 1-8 years have fluoride intakes from food and water well below the UL provided the

fluoride content of their drinking water is not higher than 1.0 mg/L. An increase in the prevalence ofmild dental fluorosis observed in some countries has been attributed to the inappropriate use of dentalcare products, particularly of fluoridated toothpaste.

The Panel did not establish an UL for infants. Breast-fed infants have very low fluoride intakes fromhuman milk (2-40 g/day) and are not at risk of developing enamel fluorosis even when given fluoridesupplements of 0.25 mg/day. The Panel notes that the Scientific Committee on Food has recommendeda maximum fluoride level of 0.6-0.7 mg/L in infant formula and follow on formula, equivalent to an intakeof about 0.1 mg/kg body weight per day in infants during the first six months of life (body weight 5 kg).For powdered formula, this maximum will be exceeded if water containing more than 0.7 mg/L is usedfor its preparation.

For children older than eight years and adults the probability of exceeding the UL of 5/7 mg fluoride/day

on a normal diet is generally estimated to be low. However, consumption of water with a high fluoridecontent, e.g. more than 2-3 mg/L, predisposes to exceeding the UL.


18/134


19/134

Fluoride


The most important fluorides for human use are sodium and potassium fluoride, which are highlysoluble in water. They are used for addition to foods (e.g. salt), dental products and fluoridation ofwater. They are permitted for use in foods for particular nutritional uses (FPNU) and food supplements(Commission Directive 2001/15/EC; Directive 2002/46/EC).

In Annex III part 1 of the amended Council Directive 76/768/EEC on the approximation of the laws ofthe Member States relating to cosmetic products, 20 fluoride compounds are listed which may be usedin oral hygiene products up to a maximum concentration in the finished products of 0.15% (1500 ppm),calculated as fluorine.

Fluorosilicic acid or hydrofluorosilicic acid (H2SiF

6) or sodiumhexafluorosilicate (Na

2SiF

6) are used for

drinking water fluoridation.

2. NUTRITIONAL BACKGROUND, FUNCTION, METABOLISM AND INTAKE

2.1. Function of fluoride

There is insufficient evidence for the indispensability of fluoride for human health. Because of theubiquity of fluoride it is virtually impossible to create an experimental situation free of fluoride.

Schwarz and Milne (1972) reared several generations of F344 rats in isolators on a fluoride-deficient diet(0.002-0.023 mg/kg/day). Rats on this diet showed decreased gain in weight and bleached incisors.Weight gain was improved by fluoride supplementation of the diet (2.5 mg/kg), tooth pigmentation wasnot. Rats in both the group on the fluoride-deficient and the fluoride-supplemented diet had shaggyfur, loss of hair and seborrhoea, indicative of a probable deficiency of other nutrients in the syntheticdiet as well.

In a cohort study of 109 infants exclusively breast-fed for at least four months (fluoride in breast-milk0.003 mg/L) and living in an area with low fluoride content of the drinking water (0.018-0.166 mg/L),those receiving a fluoride supplement from the 6th day of life onwards in addition to their fluorideintake of less than 0.003 mg/day from human milk, showed a significantly greater increase in length

and weight, especially when the mother had taken fluoride supplements during pregnancy, and asignificantly (by 12 days) earlier eruption of the first tooth in boys, than those who did not receive afluoride supplement during the first six months of life (Bergmann, 1994). Although suggestive, theseresults do not prove an essential role of fluoride in human development and growth.

In vitro, fluoride (0.02-0.1 mg/L) addition to a supersaturated solution of calcium phosphate initiatesthe formation of hydroxylapatite (Ca

3(PO

4)

2Ca(OH)

2) which is the mineral substance of bone and teeth.

With increasing fluoride concentrations fluoroapatite (Ca3(PO

4)

2CaF

2) is formed and results in more

regular and bigger apatite crystals which are less acid soluble (Featherstone et al, 1983; Newesely,1961; Okazaki et al, 1985).

Fluoride in the body is mainly associated with calcified tissue (bone and teeth) due to its high affinity forcalcium. In bone the substitution of fluoride for hydroxyl groups in apatite alters the mineral structureof the bone. This is electrostatically more stable and more compact and results in increased density

and hardness, but not increased mechanical strength in rabbit (Chachra et al, 1999). Both in rats andin humans there is evidence for a biphasic effect of fluoride on bone strength, with increases in bothbone strength and bone fluoride content at moderately high fluoride intake (16 mg/L in drinking waterof rats during 16 weeks) leading to a bone fluoride content of up to 1200 mg/kg and a decrease withhigher fluoride intake (up to 128 mg/L in drinking water) and bone fluoride content up to 10,000 mg/kg(Turner et al, 1992).

Besides the physicochemical effects of fluoride on the bone, fluoride in high doses (0.02-0.2 mg/L) wasfound to be mitogenic in osteoblasts and inhibitory to osteoclasts of chicken embryosin vitro(Farley etal, 1983; 1988; Gruber and Baylink, 1991). The mitogenic effect is restricted to osteoblastic precursors(Bonjour et al, 1993) and the same fluoride dose can be toxic to individual osteoblasts (Chachra et al,1999). Fluoride can activate thyroid adenylate cyclase (ATP pyrophosphate-lyase (cyclizing)) in vitroatvery high concentrations (10 mg/L or 190 mg/L) (Goldhammer and Wolff, 1982).

Fluoride has a cariostatic effect on erupted teeth of both children and adults. A pre-eruptive effect offluoride through increasing fluoridation of the developing enamel is supported by evidence (Groeneveld etal, 1990; Murray, 1993), but difficult to differentiate from the cariostatic effect of fluoride on erupted teeth.


20/134

Fluoride


The prevalence of dental caries in a population was not inversely related to the concentration of fluoride inenamel (Clarkson et al, 1996), which apart from the outmost surface is accumulated through pre-eruptiveenamel development (Richards et al, 1992; Weatherall and Robinson, 1988). Fluoridated enamel is less acidsoluble (Beltran and Burt, 1988). It was also demonstrated that the positive effect on reduction of caries in

both deciduous and permanent teeth was more marked the earlier children were exposed to fluoridatedwater or fluoride supplements (Groeneveld et al, 1990; Stephen et al, 1987). Comparisons of cariesprevalence between two communities in England with water fluoride concentrations of 0.2 and 1.5-2.0 mg/Lshowed that in all age groups (from 15 to >44 years) caries experience of all teeth was significantly lowerin the community with high fluoride water concentration (44% less in persons over 45 years) (Jackson etal, 1973). A similar study in Sweden compared caries prevalence in 30 to 40-years old life-long residentsof Uppsala (n=260; water fluoride concentration 1.0 mg/L) with those of Enkping (n=236; fluoride in water0.3 mg/L) and found 21% units less decayed and filled surfaces in Uppsala. Caries prevalence in that studywas not influenced by other topical fluoride sources (Wiktorsson et al, 1992).

The cariostatic effect of fluoride in saliva or plaque on erupted teeth is due to an inhibition of thedemineralisation of sound enamel by ingested acid foods or acid formed by cariogenic bacteria inthe dental plaque and by enhancing remineralisation of demineralised enamel. Demineralised enameltakes up more fluoride than sound enamel and the resultant structure is more acid resistant andcontains more fluoride (Featherstone, 1999; White and Nancollas, 1990). Moreover, fluoride affects themetabolism of carbohydrates and the production of adhesive polysaccharides by cariogenic bacteria(Hamilton, 1990). However, caries is not a fluoride deficiency disease and no specific fluoride deficiencysyndrome has been found.

2.2. Fluoride homeostasis

Ninety-nine percent of the total fluoride content of the body is concentrated in calcified tissue. Bodyfluid and soft tissue fluoride concentrations are not under homeostatic control and reflect the recentintake (Ekstrand et al, 1977). In blood the fluoride ion concentration in plasma is twice that in bloodcells (Whitford, 1996). Via the plasma fluoride is distributed to all tissues. The ratio fluoride in softtissue to fluoride in plasma is between 0.4 and 0.9. Exceptions are the kidney, pineal gland, brain andadipose tissue. The kidney can accumulate fluoride to higher concentrations than in plasma (Taves et

al, 1983). Experiments with radioactive fluoride have shown that it is not actively transported into thethyroid gland of humans or rats. Nonetheless, after long-term exposure to a high fluoride content infeed or water, the thyroid glands of some animals (cows and rats) have been found to contain increasedfluoride levels compared to their non-exposed controls (Brgi et al, 1984).

2.2.1. Intestinal fluoride absorption

Inhalation of fluoride from the air, as a rule, does not contribute more than 0.01 mg/day to the total intake,except in occupational settings where intake by that route can be several milligrams (Hodge and Smith,1977). For the purpose of setting an UL for oral exposure to fluoride, exposure via inhalation is not relevantand shall not be taken into account.

Readily soluble fluorides (sodium, hydrogen, fluorosilicic, sodium monophosphate) are rapidly almostcompletely absorbed with a plasma peak level occurring after 30 minutes (70, 130, 300, 450 g/L after

single doses of 1.5, 3, 6, 10 mg of fluoride as the sodium salt, respectively), in contrast to the low-soluble fluoride compounds calcium fluoride, magnesium fluoride and aluminium fluoride. Fluoride fromtoothpaste is also absorbed. Sodium monofluorophosphate from toothpaste needs dephosphorylationbefore absorption in the lower intestine. There is variability in the bioavailability of fluoride from differentfoods (Trautner and Siebert, 1983).

Most of fluoride is absorbed as undissociated hydrogen fluoride and absorption occurs by passivediffusion in both the stomach and the small intestine. Higher acidity of the stomach increases absorption.The presence of calcium, magnesium, phosphorus and aluminium decreases the absorption of fluoride(Cerklewski, 1997; Harrison et al, 1984; Kuhr et al, 1987; McClure et al, 1945; Spencer et al, 1981). In thecase of calcium the inhibitory effect depends on the presence of food. Sodium fluoride tablets given inwater on an empty stomach were almost 100% absorbed. The same doses given together with milk were70% absorbed, and were 60% absorbed when given with a meal (Ekstrand and Ehrnebo, 1979; Shulmanand Vallejo, 1990; Trautner and Einwag, 1987). Consecutively faecal fluoride excretion is increased.


21/134

Fluoride


2.2.2. Fluoride distribution and storage in the body

Absorbed fluoride is rapidly distributed by the circulation to the intracellular and extracellular fluid but isretained only in calcified tissues. The fluoride plasma concentration is dependent on the fluoride doseingested, dose frequency and the plasma half-life, which was determined to be 3-9 hours after giving

doses of 3 to 10 mg as tablets orally. The plasma clearance of fluoride ranged between 0.12 and 0.2L/kg/h independent on the dose (Ekstrand et al, 1977). Plasma fluoride occurs in both ionic and non-ionic forms. The non-ionic fluoride in plasma consists mostly of fat-soluble fluorocompounds. Ionicfluoride is not bound to plasma proteins or other compounds. Its level (mol) reflects the recent fluorideintake and the fluoride content of drinking water (in mg/L) when water is the predominant fluoridesource (WHO, 1994). Plasma fluoride levels increase with age and with increasing fluoride content ofbone, and as a consequence of renal insufficiency (Ekstrand and Whitford, 1988; Ekstrand et al, 1978;Singer and Ophaug, 1979).

Fluoride concentrations in plasma ranging from 0.4-2.4 mol/L (7.6-45.6 g/L) have been reported inhealthy adults (IPCS, 2002). Concentrations are lower (


22/134

Fluoride


2200-3200 mg/kg. When water contains 5-7 mg/L of fluoride the concentration in enamel has been 4800mg/kg. Such concentrations usually are accompanied by dental fluorosis (NRC, 1993).

2.2.3. Excretion of fluoride

Absorbed fluoride which is not deposited in calcified tissue is excreted almost exclusively via thekidney. The percentage of absorbed fluoride excreted via the kidney is about 50% in healthy young andmiddle-aged adults, in young infants and children it can be only 10-20%, in elderly persons higher than50%. Fluoride is filtered in the renal glomeruli and reabsorbed in the renal tubuli (10-90%), dependenton the pH of the tubular fluid. The renal clearance of fluoride is 30-50 mL/min in adults (Ekstrand et al,1982; Schiffl and Binswanger, 1982). Fluoride excretion is reduced with impaired renal function (Schiffland Binswanger, 1980; Spak et al, 1985; Torra et al, 1998).

About 10-25% of the daily intake of fluoride is excreted via the faeces (WHO, 1994).

Fluoride concentration in human milk is reported to range between 2 and 95 g/L (IPCS, 2002), whichwide range is probably due to analytical difficulties. Whereas Spak et al (1983) found no correlationbetween the fluoride content of drinking water (0.2 to 1 mg/L) and fluoride content of human milk (7.6

g/L), Dabeka et al(1986) could show a relationship: 32 mothers in an area with fluoride in drinkingwater of


23/134

Fluoride


There is no convincing evidence that health and development of humans depend on the intake offluoride, however, due to the ubiquitous presence of fluoride in the environment a zero exposure is notpossible under normal circumstances.

Based on epidemiological studies of the inverse relationship between dental caries and theconcentration of fluoride in drinking water in the 1940s it was concluded that fluoride has a beneficialeffect in increasing the resistance to dental caries in children (Dean et al, 1942) and at all ages (Russelland Elvove, 1951). In communities with water fluoride concentrations (0.7 to 1.2 mg/L, depending on theaverage regional temperature) the caries prevalence was 40-60% lower than in communities with lowwater fluoride concentrations. The studies of Dean (1942) had also shown that a positive relationshipexisted between water fluoride concentration and the prevalence of dental fluorosis. A concentrationof about 1 mg fluoride/L in drinking water was identified as being optimal both in reducing cariesprevalence and keeping dental fluorosis prevalence below 10% in the population. This fluorosis was ofthe mild to very mild type (see Annex 2) and practically none of the moderate to severe type.

From this optimal water fluoride concentration derives the estimated adequate fluoride intake ofinfants and children above the age of 6 months of 0.05 mg/kg body weight/day (Burt, 1992; Singerand Ophaug, 1979): age 7-12 months 0.5 mg/day; age 1-3 years 0.7 mg/day; age 4-8 years 1 mg/day;age 9-13 years 2 mg/day; age 14-18 years 3 mg/day; for females and males of 19 years and above 3and 4 mg/day, respectively (FNB, 1997). The guidance reference values of the Austrian, German andSwiss Nutritional societies are based on the same calculation (D-A-CH, 2000).There is a difference inthe adequate intake or guidance value for fluoride below the age of six months defined by the FNB andby D-A-CH. The very low fluoride intake of breast-fed infants which is about 0.01 mg/day is defined asthe adequate intake for age 0-6 months by the FNB. Assuming an average body weight of 5 kg for aninfant of that age group and a guidance value of 0.05 mg/kg body weight/day a guidance value of 0.25mg fluoride/day has been calculated (D-A-CH, 2000).

2.3.2. Fluoride intake (exposure)

Fluoride exposure via inhalation and the skin will not be considered, because in normal circumstancesthey contribute little to the total intake. However, the fluoride content of food dried over high-fluoridecoal fires can increase considerably (from 5- to 50-fold) and be a significant source of oral ingestion,

as shown in China (Liang et al, 1997).

Exposure by oral ingestion of fluoride is by water, food (including fluoridated salt available in Austria,Belgium, Czech Republic, France, Germany, Spain and Switzerland), cosmetic dental products andfluoride supplements. Fluoride supplements are considered to be drugs in most countries of theEuropean Community.

2.3.2.1. Water

Among the main sources of total fluoride intake in Europe are drinking and mineral waters with morethan 0.3 mg/L of fluoride. From U.S. and Canadian studies the total fluoride intake of adults in areaswith different fluoride content of drinking water was estimated: 0.3-1 mg/day, 1.4-3.4 mg/day withwater fluoride content


24/134

Fluoride


whereas about half of the cantonal area was served with drinking water containing less than 0.1 mg/L(Rapport Annuel, 1999).

Total tap water intake of adolescents in the UK and in Germany was 676 g/day and 718 g/day, respectively

(Sichert-Hellert et al, 2001; Zohouri et al, 2004). Total fluoride intake from all kind of drinks in Britishadolescents was estimated to be 0.47 mg/day.

Drinking tap water, however, is increasingly replaced by the use of bottled water. Whereas drinkingwater for human consumption according to Council Directive 98/83/EC, following the advice of theScientific Committee on Food (SCF, 1998), may not contain more than 1.5 mg fluoride/L, bottled naturalmineral waters can have higher fluoride levels. Natural mineral waters which contain more than 1 mgfluoride/L can be labelled as contains fluoride. According to Council Directive 88/777/EEC on theapproximation of the laws of the Member States relating to the exploitation and marketing of naturalmineral waters, Member States can make national provisions for labelling a natural mineral water assuitable for the use in infant nutrition. According to Directive 2003/40/EEC the fluoride content ofnatural mineral waters must be not more than 5 mg/L by 1 January 2008. Mineral waters exceeding 1.5mg fluoride/L shall bear on the label the words contains more than 1.5 mg/L of fluoride: not suitablefor regular consumption by infants and children under 7 years of age and shall indicate the actualfluoride content.

A survey of 150 mineral and table waters from the German market measured an average fluorideconcentration of 0.58 0.71 mg/L: 24% had a fluoride concentration below 0.1 mg/L, 43% equal to orbelow 0.3 mg/L, 31% between 0.3 and 0.6 mg/L, and 8 (5%) waters had a fluoride concentration above1.5 mg/L with a maximum value of 4.5 mg/L. The average consumption of bottled water in Germany atthe time of the survey was estimated to be 104 L per year (Schulte et al, 1996). In a similar survey of 33bottled waters from the Swedish market a median fluoride concentration of 0.19 mg/L with a range of0-3.05 mg/L was determined (Rosborg, 2002). The fluoride concentration in 25 commercial brands ofbottled water (spring, mineral or distilled) available in the UK was 0.08 (0.08) mg/L with a range from0.01-0.37 mg/L. The average bottled water intake was estimated to be 108 mL/day in adults (Zohouri etal, 2003) and only 20 mL/day in adolescents (Zohouri et al, 2004). Twenty-four mineral waters availablein Belgium had fluoride concentrations below 1 mg/L in 16 cases, but the highest value found was 5.5

mg/L. A case of dental fluorosis in an eight-year old girl was attributed to the preparation of her infantformula with mineral water containing 1.2 mg fluoride/L. Her fluoride intake from age three months toage 12 months was well above 0.1 mg/kg body weight/day (Bottenberg, 2004)

2.3.2.2 Food

Fluoride intake from food is generally low except when food is prepared with fluoridated water.Exceptions are tea which can contain considerable amounts of fluoride (0.34-5.2 mg/L) (Schmidt andFunke, 1984; Wei et al, 1989; Chan and Koh, 1996), dependent on type, brewing and fluoride contentof water. Some brands of instant teas were reported to be another significant source of fluoride intake(up to 6.5 mg/L when prepared with distilled water) (Whyte et al, 2005).

Vegetables and fruit, except when grown near fluoride emitting industrial plants, contain between 0.02and 0.2 mg/kg fresh weight, milk and milk products 0.05-0.15 mg/kg, bread, cereals and meals 0.1-0.29

mg/kg, meat and meat products 0.15-0.29 mg/kg, eggs 0.18 mg/kg, fish and fish sticks 0.48-1.91 mg/kg(Bergmann, 1994; EGVM, 2001). The fluoride content of both fish and meat depends on the care takenwith deboning, and can be as high as 5 mg/kg. (Bergmann, 1994). Dried herbs contain up to 2.0 mg/kgfluoride. Table 1 summarises the fluoride content in various types of foods from various parts of the worldcompiled by IPCS (2002) as well as Chinese data on corn and vegetables dried naturally or over high-fluoride coal fires (Liang et al, 1997).

efsa, 2005, opinion of the scientific panel on dietetic products, hal 373

Documents