WHO FOOD ADDITIVES SERIES: 52

First draft prepared by

Dr D. Bellinger
Harvard Medical School, Boston, MA, USA

Dr M. Bolger, K. Egan
United States Food & Drug Administration, College Park, MD, USA

M. Feeley
Health Canada, Tunney’s Pasture, Ottawa, Ontario, Canada

Dr J. Schlatter
Swiss Federal Office of Public Health, Zürich, Switzerland

and

Dr C. Tohyama
Environmental Health Sciences Division, National Institute for Environmental Studies, Tsukuba, Ibaraki, Japan

Cadmium was evaluated by the Committee at its sixteenth, thirty-third, forty-first and fifty-fifth meetings (Annex 1, references 30, 83, 107, 149). At its sixteenth meeting, the Committee allocated a provisional tolerable weekly intake (PTWI) of 400–500 µg of cadmium per person. At the three subsequent meetings, the Committee retained this PTWI, but expressed it in terms of the intake of cadmium per kg bw (7 µg/kg bw). At its fifty-fifth meeting, the Committee decided that the prevalences of renal tubular dysfunction that correspond to various dietary intakes of cadmium could serve as a reasonable basis for risk assessment, and concluded that the risk of excess renal tubular dysfunction in the population would be negligible below a urinary cadmium excretion of 2.5 µg/g of creatinine. The Committee noted, however, that these estimates are based on a model that is dependent on the values assumed for key parameters (e.g. dietary bioavailability, age dependency of the intake: excretion ratio). Although new information indicated that a proportion of the general population might be at an increased risk of tubular dysfunction at the current PTWI of 7 µg/kg bw, the Committee at the fifty-fifth meeting maintained the PTWI at this value because of lack of precision in the risk estimates. The Committee made several recommendations regarding the data that would be needed in order to reduce the uncertainty in the prevalence estimates. A considerable number of new studies addressed certain aspects of the issues identified in these recommendations and served as the basis for the Committee’s deliberations at the present meeting.

A number of detailed evaluations have reviewed data on the pharmacology of cadmium (WHO, 1992; Agency For Toxic Substances And Disease Registry, 1999; United States Environmental Protection Agency, 1999). Overall, the absorption or bioavailability of cadmium from the gastrointestinal tract is generally considered to be slightly lower in experimental animals than in humans. For the majority of species tested, the absorption of cadmium can range from 0.5% to 3.0% of the dose administered, while in humans a range of 3.0–8.0% can be found (Järup et al., 1998; Agency for Toxic Substances and Disease Registry, 1999). Previous reviews have described how the composition of the diet, including fibre, protein (sunflower seeds) and carbohydrates (rice) can also affect the bioavailability of cadmium (Annex 1, reference 150).

(a) Effect of diet on uptake of cadmium

In groups of five to six female B6C3F₁ mice given drinking-water containing a mixture of metals (arsenic, 18 mg/kg; cadmium, 6 mg/kg; chromium, 150 mg/kg; nickel, 150 mg/kg; and vanadium, 45 mg/kg) for up to 24 weeks, maximum concentrations of cadmium were found in kidney (1466 µg/kg), small intestine (1009 µg/kg), pancreas (<100 µg/kg) and femur (<50 µg/kg). A similar pattern of distribution was observed when mice were fed with a mixture of metals (arsenic, 47 µg/kg; cadmium, 26 µg/kg; chromium, 1105 µg/kg; nickel, 1412 µg/kg; lead, 2376 µg/kg; and vanadium, 1105 µg/kg) in the diet for up to 24 weeks, except that the concentrations of cadmium found in organs were greatly reduced (Radike et al., 2002).

Groups of five female Sprague-Dawley rats received diets containing 40% milled rice supplemented with cadmium at a final concentration of 0.25 mg/kg, equivalent to a daily intake of 25 µg/kg bw, and with zinc, iron and/or calcium at concentrations providing marginal or adequate intake, for 5 weeks. Fractional absorption of cadmium was determined by whole body counts and faecal analysis after a dose of ¹⁰⁹Cd and an additional 15.5 days of test diets. Whole body retention of ¹⁰⁹Cd ranged from 0.6% to 4.0%, with the greatest retention in rats fed the diet containing each of the three additional elements at concentrations providing marginal intake. The least retention was found in rats fed either the diet containing all three elements at concentrations providing an adequate intake, or the diet containing calcium and iron at concentrations providing adequate intake and zinc at concentrations providing marginal intake. Concentrations of cadmium found in the kidney were similar to whole-body retention values (lower for animals receiving diets providing adequate intake of all three elements and highest for animals receiving the diets providing marginal intake of the three elements) (approximately 0.8 mg/kg for animals receiving diet providing adequate intake, and 1.45 mg/kg for animals receiving the diet providing marginal intake) while the diet containing only iron at adequate levels had the largest single effect on intestinal concentration of cadmium (approximately 0.45 mg/kg versus 1.45 mg/kg in diet providing marginal intake). All concentrations are given on a dry-weight basis (Reeves & Chaney, 2002).

In a study with a similar experimental design by the same authors, groups of seven rats fed diets containing 20% sunflower kernels (final concentration of cadmium, 0.18 mg/kg) and adequate or marginal amounts of the same three trace elements retained 0.37–1.46% of a dose of ¹⁰⁹Cd (whole body count only). Lowest retention values were measured in animals fed either adequate amounts of all trace elements or marginal amounts of zinc or calcium in combination with adequate amounts of the other two elements. The authors concluded that the high amounts of zinc found naturally in sunflower seeds minimized the absorption of cadmium (Reeves & Chaney, 2001).

Groups of 10 Wistar rat pups were given cadmium chloride (CdCl₂) orally at a dose of 0.5 mg/kg bw per day for 9 consecutive days prior to sacrifice on day 10 of the experiment. Whole-body analysis of cadmium (organs and carcass) indicated that approximately 2.45% of the total administered dose was retained, while in a group that received the same dose of cadmium chloride in addition to a daily oral supplement of 6% dicalcium phosphate (CaHPO₄) in cows’ milk, only 0.94% of the total dose was retained (Kašuba et al., 2002).

Groups of 10 Wistar rats of both sexes, aged 6 days, were given cows’ milk (approximate calcium content, 0.1%) or milk supplemented with either 1%, 3% or 6% CaHPO₄, in addition to daily oral doses of CdCl₂ (0.5 mg/kg bw) for 9 consecutive days. The estimated total intake of calcium at the end of the dosing period was 108 mg, 176 mg, 311 mg and 515 mg, respectively. Concentrations of cadmium in liver, kidney and brain were significantly reduced in a dose-dependent manner (48–66% reduction in the group receiving supplementation with the highest dose of calcium) (Saric et al., 2002).

Groups of 15 male rats were fed diets that were either supplemented with iron at a concentration of 120 mg/kg or that were deficient in iron (containing iron at a concentration of 2–6 mg/kg), for a period of 4 weeks. The rats were fasted for 18 h and then given a single oral dose of ¹⁰⁹CdCl₂ (0.4 µmol/kg bw or 0.073 mg/kg). Tissue samples were collected 48 h later for analysis of iron, cadmium and divalent metal transporter 1 (DMT1) mRNA. Serum concentrations of iron in rats fed the diet that was deficient in iron were approximately 14-fold lower than those in rats fed the diet supplemented with iron (361 versus 25.4 µg/dl, respectively) while body burdens of cadmium were 10-fold greater (1360 ng/kg bw versus 130 ng/kg bw, respectively, or approximately 1.85% and 0.18% of the administered dose). The level of DMT1 mRNA in duodenum, the organ containing the greatest amount of administered cadmium, was also increased by approximately 15-fold in the rats fed the diet that was deficient in iron (Park et al., 2002).

Groups of six male rats who had been fed diets containing zinc at levels that were either marginal (with or without whole grain wheat, 50 g/kg), adequate (with or without whole grain wheat) or surplus for 7 days were then given cadmium in a test meal (31 kBq ¹⁰⁹Cd plus 562 ng Cd) and assessed for retention and absorption up to 15 days later. There was a reduction in the amount of cadmium absorbed relative to the concentration of zinc in the diet (7.7% for the diet containing a marginal level of zinc versus 2.4% for the diet containing surplus zinc), while the rats fed diets also containing whole wheat absorbed less cadmium (30–40%) than those fed diets containing zinc alone. Less than 2% of the administered dose of cadmium was retained after 15 days, for all the groups. The authors concluded that wheat phytate and/or fibre contributed to the reduced absorption of cadmium (House et al., 2003).

Groups of four female Sprague-Dawley rats were fed diets containing either 72% plain rice supplemented with CdCl₂ at a concentration of 5.1, 20 or 40 mg/kg of diet, or rice contaminated with cadmium at a concentration of 0.02–1.01 mg/kg of diet for up to 8 months, and then assessed for rates of retention of cadmium, and for concentrations of cadmium and metallothionein in liver, kidney and intestine. No toxic effects caused by cadmium were observed during the study period, although there were dose-dependent increases in concentrations of cadmium in the kidney and liver (54 µg/g in kidney and 35 µg/g in liver after 8 months in rats fed the diet containing cadmium at 40 mg/kg diet). Overall rates of cadmium retention ranged from 0.2% to 1.0%, while rates at which cadmium was distributed to the kidney were negatively associated with the concentration of cadmium in the diet (20% at a concentration of cadmium of up to 1.0 mg/kg of diet, versus 10% at 20 and 40 mg/kg of diet) (Hiratsuka et al., 1999).

(b) Distribution

Distribution of cadmium after ingestion has recently been comprehensively reviewed. Briefly, absorption of cadmium from the gastrointestinal tract is characterized by a rapid initial uptake into the intestinal mucosa (mainly duodenum and jejunum in rodents), followed by a slower absorption into the systemic circulation and redistribution to kidney and liver. Transport of cadmium from the small intestine is thought to be facilitated by a number of possible mechanisms, including metal transport proteins such as DMT1, calcium ion channels, amino-acid transporters (as cysteine–cadmium conjugates) and by endocytosis of cadmium–metallothionein complexes. The last mechanism is supported by experimental evidence showing that cadmium induces metallothionein in the intestinal mucosa and that cadmium ingested orally is bound to metallothionein in the small intestine (Zalups & Ahmad, 2003).

Metallothioneins are a family of inducible, cysteine-rich proteins of low molecular mass, ubiquitous to mammals, which are capable of binding with high affinity to a number of divalent and trivalent cations, including zinc, copper, cadmium and other group I/IIb transitional metals. In mammals, metallothioneins are found in the highest concentrations in liver, kidney and intestine; each molecule of metallothionein is able to bind up to seven atoms of cadmium. Besides contributing to possible cellular defence mechanisms by sequestering potentially toxic metals, metallothioneins are also important in the overall homeostasis of zinc (Miles et al., 2000; Coyle et al., 2002).

Once in the systemic circulation, cadmium can be taken up by a number of organs, but the liver generally accounts for the majority of uptake. In the liver, cadmium promotes hepatocellular induction of metallothioneins (MTI and MTII iso-forms). Binding to metallothionein is thought to allow the liver to accumulate potentially cytotoxic concentrations of cadmium without exhibiting adverse effects.

Repeated exposure of both wild-type and MTI/II-null mice to CdCl₂ via daily subcutaneous injections of 0.05–2.4 mg/kg (for wild-type mice) and 0.0125–0.8 mg/kg (for MTI/II-null mice) for up to 10 weeks resulted in dose-dependent increases in the concentration of cadmium in the liver (400 µg/g) and the concentration of metallothioneins in the wild-type mice. Although the concentration of cadmium in the liver in the MT-null mice was only 10 µg/g, approximately, histopathological indices suggested that hepatic damage was more severe than in the wild-type mice (Habeebu et al., 2000).

Groups of eight or nine pregnant MTN (normal MTI and MTII) and MT1,2KO (MTI/II knock-out) mice were given drinking-water containing CdCl₂ at a concentration of 0.15 µg/l for 7, 14 or 17 days during gestation or for 19 days during gestation and 11 days during lactation (a combined period of 30 days). At the end of the period of exposure, dams and/or pups were sacrificed and tissues were analysed for content of cadmium and metallothionein. For dams of both strains, the small intestine retained the highest proportion of the ingested cadmium, 0.94–2.97%, and the liver and kidney retained 0.14–0.49%. The livers of the MTI,IIKO dams contained significantly greater amounts of cadmium than those of the MTN dams (approximately three-fold more in the group exposed for 30 days) while the MTN dams had twice as much cadmium in the kidney. Up to 85–90% of the total amount of cadmium transferred to the pups of both strains was retained by the small intestine, while up to five times more cadmium was found in the liver of the MTI, IIKO pups than in that of the MTN pups. Induction of metallothionein in the placenta of MTN dams appeared to restrict the distribution of cadmium to the fetus; however, approximately equal amounts (0.047–0.062% of the ingested dose) of cadmium were transferred via lactation to pups of both strains (Brako et al., 2003).

(c) Excretion

Cadmium that is released into the systemic circulation, usually in a conjugated form (with albumin, cysteine, glutathione, or metallothionein), can be taken up via endocytosis by proximal tubule epithelial cells in the kidneys. Subsequent lysozymal proteolysis of the complex results in the release of cadmium ions, which can induce the production of metallothionein in situ. When the capacity of metallothionein to bind cadmium is exceeded, cadmium induces cytotoxicity, and eventually causes significant structural and functional damage (Järup et al., 1998).

After oral exposure, most cadmium (up to 90% in experimental animals) passes through the gastrointestinal tract unabsorbed and is excreted in the faeces. A portion of the retained and absorbed cadmium can also be excreted via the faeces after hepatic secretion of cadmium conjugates (e.g. with glutathione) into the bile. When cadmium is administered at a low dose, a minor fraction can also be excreted in the urine (as metallothionein–cadmium). However, as nephrotoxicity progresses, greater amounts of unfiltered cadmium will be eliminated in the urine, owing to declining renal filtration capacity as a consequence of tubule damage (Zalups & Ahmad, 2003).

2.2.1 Acute toxicity

Oral LD₅₀ values for experimental animals (mainly rodents) range from approximately 100 to 300 mg/kg bw and are dependent on the form of cadmium administered (see Table 1).

Table 1. LD₅₀ values for rats and mice treated orally with cadmium

Adapted from WHO (1992) and Agency For Toxic Substances And Disease Registry (1999)

As listed, the LD₅₀ for cadmium sulfide (CdS), a form of cadmium with very low solubility, is approximately an order of magnitude greater that the LD₅₀ for CdCl₂, a very soluble form of cadmium.

At relatively high oral doses of cadmium (slightly below the LD₅₀), histopathological evidence of liver toxicity (fibrosis, necrosis) is a common finding, and testicular atrophy and necrosis, with subsequent decreased fertility, are found in male rats and mice. The latter effect may be related to accumulation of cadmium in the gonads, with subsequent damage to the testicular endothelium (WHO, 1992).

2.2.2 Long-term studies of toxicity and carcinogenicity

Although a variety of toxicological end-points have been observed in experimental animals (reproductive toxicity, neurotoxicity, carcinogenicity), those of relevance to humans are the renal effects which manifest after low-level long-term exposure to cadmium and accumulation of cadmium to a critical concentration in the kidney.

(a) Renal effects

As previously described, long-term exposure to cadmium eventually results in a variety of renal changes involving damage to proximal tubule epithelial cells, degeneration and apoptosis. Morphological changes observed include atrophy, interstitial fibrosis, glomerular sclerosis and focal necrosis. Earlier indications usually associated with renal damage include low-molecular-mass proteinuria, glucosuria and aminoaciduria. The results of relevant experimental studies are summarized in Table 2.

Table 2. NOELs and LOELs for renal effects of cadmium chloride administered in drinking-water

Modified from Annex 1, reference 150

LOEL, lowest-observed-effect level

In general, although some variability exists, renal effects in laboratory animals are associated with kidney concentrations of cadmium of between 200 and 300 µg/g, resulting from long-term exposures of, on average, between 1 and 10 mg/kg bw per day (Agency For Toxic Substances And Disease Registry, 1999).

Rats

Groups of 50 female Sprague-Dawley rats were fed diets containing CdCl₂ at a concentration of 0, 8, 40, 200 or 600 mg/kg for up to 8 months. By 2 months, histopathological examination revealed signs of kidney damage in the form of proximal tubule degeneration and necrosis in animals fed diets containing cadmium at >200 mg/kg. In rats fed diets containing cadmium at 200 mg/kg, the concentration of cadmium in the kidney was, on average, 169 µg/g, reaching an approximate maximum of 250 µg/g by 8 months (Mitsumori et al., 1998). In rats receiving diets containing cadmium at 40 mg/kg for 18 months, no kidney lesions were detected and the maximum concentration of cadmium in the renal cortex was 120 µg/g (Shibutani et al., 2000).

Groups of 36 male Wistar rats given drinking-water containing CdCl₂ at 5 or 50 mg/l for up to 24 weeks exhibited dose-dependent damage to the kidney nephron (primarily the proximal convoluted tubules and glomeruli). The average daily intake of cadmium was estimated to be 320–678 µg/kg bw and 1963–4487 µg/kg bw for the groups receiving 5 and 50 mg/kg, respectively, and the concentration of cadmium in the kidney after 24 weeks of treatment was 10 µg/g and 65 µg/g, respectively, for these two groups. Concentrations of the urinary marker of renal toxicity N-acetyl-beta-D-glucosaminidase (NAG; total and isoenzyme B), were significantly increased after 6 weeks in animals in the group receiving the highest dose (concentration of cadmium in the kidney, 25 µg/g) (Brzóska et al., 2003).

Groups of three to four male Wistar rats, aged 2 months, were treated with a single subcutaneous dose of cadmium–metallothionein (isolated from rabbit liver), equivalent to a dose of cadmium of 0.4 mg/kg bw, and then assessed for cadmium in the kidney and early signs of nephrotoxicity at various times up to 12 h later. By 3 h after dosing, the concentration of cadmium in the renal cortex had increased to 40 µg/g. Various apical-membrane proximal-tubule transporter proteins, in-cluding megalin, Na⁺/H⁺ exchanger fusion protein type 3 (NHE3) and a subunit of vacuolar-type proton-pump ATPase (V-ATPase) underwent a time-dependent redistribution, from the brush-border membrane into the cytoplasm. Additional histopathological signs of tubule necrosis were also observed by 6 h after treatment with cadmium-metallothionein. Nephrotoxicity was induced at lower concentrations of renal cadmium after parenteral exposure to cadmium–metallothionein (Sabolic et al., 2002).

After a single intraperitoneal injection of cadmium (as cadmium–metallothionein) of 0.3 mg/kg bw, indices of renal toxicity (including urinary protein and lactate dehydrogenase activity) significantly increased in female Sprague-Dawley rats after 8–12 h. After 6 h, the kidneys contained 51% of the administered dose, and 8.5% was found in the cortical mitochondria. Mitochondrial function, in terms of oxygen consumption during state 3 respiration (generation of ATP), also decreased in a time-dependent fashion (Tang & Shaikh, 2001).

Non-human primates

In a long-term study, groups of five to eight male rhesus monkeys were fed diets containing CdCl₂ at a concentration of 0, 3, 10, 30 or 100 mg/kg for 9 years (equivalent to 0, 0.12, 0.4, 1.2 and 4.0 mg/kg bw per day, respectively). Decreased growth rate, proteinuria and glucosuria were noted in the monkeys receiving the diet containing the highest dose of cadmium (approximately 4 mg/kg bw per day), while limited or no changes in renal function were seen in monkeys receiving diets containing CdCl₂ at a concentration of <10 mg/kg (<0.4 mg/kg bw per day) (Masaoka et al., 1994).

(b) Effects on calcium metabolism and bone

Long-term oral exposure to cadmium at a dose of 2–8 mg/kg bw per day can produce decreases in the calcium content of bone, and increased calciuria in experimental animals.

Rodents

Bone strength was decreased in weanling rat pups given drinking-water containing CdCl₂ at a concentration of 0, 5 or 10 mg/l for 4 weeks, at both concentrations of CdCl₂ (Agency For Toxic Substances And Disease Registry, 1999).

When CFI female mice were subjected to the combined stresses of age, nutrient-deficiency, multiparity and treatment with cadmium (concentration of CdCl₂ in the diet, 0.25, 5 or 50 mg/kg, equivalent to 0, 0.035, 0.70 or 7.0 mg/kg bw per day), cadmium at a dose of 50 mg/kg was found to reduce the calcium content of the femur by 14–28% (Whelton et al., 1997).

Non-human primates

In groups of 4–10 female rhesus monkeys consuming diets containing CdCl₂ at a concentration of 30 mg/kg for up to 9 years, decreased plasma concentrations of vitamin D₃ and disturbed bone calcification could be induced only when cadmium was administered in combination with a diet containing low levels of vitamin D or low levels of protein, calcium and phosphorus, respectively. When the monkeys were given diets low in protein, calcium, phosphorus and vitamin D, osteomalacia was induced and could be reversed by administration of vitamin D₃ (reviewed in WHO, 1992).

Cadmium is thought to act either directly on bone mineralization, possibly by replacing calcium, or indirectly, through inhibition of the renal conversion of the active form of vitamin D, resulting in decreased absorption of calcium in the intestine.

(c) Additional effects

Mice

In groups of six mice, after a single oral dose of CdCl₂ of 0.26 mg/kg bw (one-tenth of the LD₅₀) and then a subcutaneous inoculum of Semliki forest virus, 100% mortality was observed in the group that received cadmium and virus compared to 20% in the control group that received virus only. In addition, decreased mean survival times were observed and neuronal damage appeared earlier and with greater intensity in groups treated with both cadmium and virus compared with the untreated infected controls (Seth et al., 2003).

Groups of 20 male C3H/HeN, A/J and C57BL/6 mice were fed on diets containing CdCl₂ at a concentration of either 0 or 50 mg/kg (equivalent to 7 µg/kg bw per day) for 54 weeks and then assessed for spontaneous hepatocarcinogenesis (C3H/HeN and C57BL/6 strains) or spontaneous hepatitis (A/J). In C3H/HeN mice, the diet containing cadmium inhibited the high spontaneous rate of hepatocarcinogenesis (all tumours combined) observed in the control animals (incidence of tumours: controls, 79%; cadmium-treated, 10%) while approximately equal numbers of C57BL/6 mice developed hepatic carcinomas. Approximately twice as many A/J mice receiving the control diet developed hepatitis-related lesions when compared with mice receiving the diet containing cadmium (83% versus 47%, respectively). In addition, no hepatitic lesions that could be classified as moderate to severe were observed in mice receiving the diet containing cadmium. In contrast to the C57BL/6 mice, the C3H/HeN and A/J mice receiving the diet containing cadmium had significantly greater concentrations of hepatic cadmium, zinc and metallothionein when compared with their respective controls (Nishiyama et al., 2003).

Non-human primates

One month after undergoing bilateral ovariectomy, groups of two to four female cynomologus monkeys were given CdCl₂ intravenously at a dose of 0, 1.0 or 2.5 mg/kg bw, three times per week for 13–15 months. The frequency of dosing was reduced to twice per week after 9 months, until the experiment was terminated owing to development of severe anaemia. Concentration of cadmium in the pancreas (650 and 900 µg/g at a dose of 1.0 and 2.5 mg/kg bw, respectively) exceeded that in the kidney (500 µg/g), while the highest concentrations were found in the liver (>1000 µg/g). Associated decreases in number of pancreatic islet cells and area of islet tissue relative to pancreatic tissue, and decreases in relative areas of pancreatic insulin-positive tissue, were observed in the animals treated with cadmium (Kurata et al., 2003).

(d) Carcinogenicity

Studies in experimental animals treated by injection or inhalation have provided considerable evidence that cadmium is carcinogenic. In rats, cadmium causes a variety of tumours, including malignant tumours at the site of injection, and in the lungs, after inhalation. Oral intake is associated with proliferative lesions of the ventral lobe of the prostate gland in rats fed diets that are adequate in zinc, whereas deficiency in zinc in the diet appears to inhibit the tumorigenic effect of cadmium. The relevance of these studies to carcinogenesis in the human prostate gland is questionable, because of anatomical differences between the prostate gland in humans and that in rodents. The Committee therefore concluded that cadmium is carcinogenic in experimental animals when given by injection or inhalation, and that workers exposed by inhalation have been shown to develop lung cancer. There was no evidence that cadmium is carcinogenic to humans exposed orally (Annex 1, reference 150).

2.2.3 Developmental toxicity

Mice

Six days after administration of an intravenous dose of ¹⁰⁹Cd, limited amounts of radiolabel could be detected in brain parenchyma of mice. Cadmium could be detected in those areas of the brain that normally lack a blood–brain barrier (the circumventricular organs) (Takeda et al., 1999).

Rats

In a three-generation study in which groups of 10 Wistar rats received CdCl₂ by gavage at a dose of 0, 3.5, 7.0 or 14.0 mg/kg bw (during gestation, lactation and for 8 weeks after giving birth), behavioural (exploration activity) and electrophysiological changes were seen in male offspring at age 12 weeks, mainly in the groups receiving the two higher doses (Nagymajtenyi et al., 1997).

Pregnant Wistar rats (a total of 280 animals) were treated by gavage with CdCl₂ at a dose of 0, 3.5, 7.0 or 14.0 mg/kg bw per day, according to the following protocols: (1) treatment given on days 5–15 of gestation; (2) protocol (1) plus treatment during lacation, on postnatal days 2–28; (3) protocol (2) plus treatment of males after lactation: 5 days per week, for 8 weeks, beginning on postnatal day 29. Male offspring aged 11–12 weeks were tested in a variety of behavioural and neurophysiological experiments. Significant decreases in exploratory activity were observed only in the offspring of the groups treated during both gestation and lactation, protocol (2), at the two highest doses of cadmium. Various electrocortical changes in somatosensory, visual and auditory centres were observed only in rats treated with protocol (3) and receiving the highest dose of cadmium, while decreases in tail-nerve conductance velocity were also only observed in rats treated according to protocol (3) at the two highest doses of cadmium (Desi et al., 1998).

Groups of four pregnant Wistar rats were given drinking-water containing both cadmium acetate (10 mg/kg) and lead acetate (300 mg/kg) throughout gestation and lactation. Average daily intake of cadmium and lead was 1.0–2.1 mg/kg bw and 31.6–64.9 mg/kg bw, respectively. At age 21 or 75 days, pups were sacrificed and the hippocampus and hypothalamus were analysed for various monoamines and metabolites. Male rats aged 74 days were tested on an elevated plus-maze apparatus. Concentrations of dopamine in the hippocampus were decreased by three-fold at 21 days but increased 9.4-fold at age 75 days, and 5-hydroxyindoleacetic acid (5-HIAA) content was significantly increased (approximately two-fold) only at age 21 days. In the hypothalamus, 5-HIAA content was increased by approximately two-fold at age 75 days. The elevated plus-maze results indicated symptoms of increased anxiety (Leret et al., 2003).

2.2.4 Genotoxicity

Equivocal results have been obtained when cadmium is tested for its ability to cause gene mutations in a variety of prokaryotic or mammalian cells. Cadmium induces chromosomal aberrations in both human and rodent cells, but the evidence suggests that the mechanism is not one of direct genotoxicity. DNA damage, in the form of strand breaks and protein cross-links, was induced by cadmium in various rodent cell lines, but only at doses that arrested cell growth (Misra et al., 1998).

In vitro

Cadmium, when given at a dose of 10 µmol/l or 1.83 µg/ml for 2 weeks, can induce malignant transformation of human or rat prostatic epithelial cells in vitro (Nakamura et al., 2002).

In human prostate epithelial cells transformed by continual exposure to CdCl₂ at a concentration of 10 µmol/l, alterations in apoptotic genes and resistance to apoptotic stimuli were observed. The authors concluded that treatment with cadmium increases apoptotic resistance, thereby enhancing tumour initiation and malignant progression (Achanzar et al., 2002).

Long-term exposure (4 days) of yeast strains to CdCl₂ at a concentration of 3 or 5 µmol/l (0.55 or 0.81 µg/ml) caused a decrease in post-replication mismatch repair in homonucleotide runs in yeast genes, resulting in increased mutability. The same concentrations of cadmium also induced mutations in genes involved in mitochondrial function. The effects on mismatch repair and increased rates of mutation were not shown to be related to DNA damage (Jin et al., 2003).

2.3.1 Absorption

The gastrointestinal absorption of cadmium is influenced by diet and nutritional status, with iron status being particularly important. On average, 5% of the total oral intake of cadmium is absorbed, but individual values range from less than 1% to more than 20% (WHO, 1992).

A recent study described the rate of absorption of cadmium from the diet among 38 female farmers who were exposed to cadmium at a level close to the current provisional tolerable weekly intake (PTWI). No statistically significant difference in cadmium absorption was observed between the seven women with diabetes mellitus and their 13 age-matched controls, or between the six anaemic women and their 12 controls. In all women, the rate of absorption of cadmium was significantly correlated with age, serum ferritin, serum iron, blood cadmium and urinary cadmium concentrations. However, multiple regression analysis revealed that only age was a significant predictor of the rate of absorption of cadmium. These results demonstrate that age, rather than iron deficiency, diabetes mellitus or cadmium burden, is the only independent factor to affect the rate of absorption of cadmium, suggesting that there is a large degree of individual variation in the rate at which cadmium is absorbed (Horiguchi et al., 2004b).

An epidemiological study in Japan investigated the effect of deficiency of iron on uptake of cadmium and/or renal tubular dysfunction. A total of 1482 women, aged 20–74 years, living in six prefectures, in areas not known to be excessively polluted with cadmium, participated in the study. Among them, never-smoking, non-pregnant and non-lactating women (1190 subjects) were selected and classified into groups of anaemic women (37), women with iron deficiency (388) and women with normal iron levels (765) on the basis of concentrations of ferritin (<20 ng/ml) and haemoglobin (<10 g/dl). Age and prefecture of residence were used to make matched pairs. In this study, urinary cadmium was considered to be a surrogate for body burden of cadmium, which reflects the overall uptake of cadmium. No significant increases in levels of urinary cadmium or low-molecular-mass protein (alpha₁microglobulin and beta₂-microglobulin) were found in the groups of anaemic and iron-deficient women when compared with the matched control population. The authors concluded that the current level of iron-deficiency among the general female population not exposed to excessive amounts of environmental cadmium may not increase in cadmium body burden or renal tubular dysfunction induced by cadmium (Tsukahara et al., 2003).

The uptake of cadmium in the digestive tract was investigated in a study of 25 nonsmoking women aged 20–23 years. After the women had eaten meals containing low levels of cadmium for 11 days, in order to achieve a steady intake–output balance, rates of absorption of cadmium were calculated by measuring excess excretion during a subsequent period of 1 or 3 days in which the women ate meals containing high levels of cadmium. The overall rate of absorption of cadmium was estimated to be higher than that observed in earlier studies. The mean and its range varied considerably, e.g. 47% (range, -9.4–83.3%) for the group eating meals containing a high level of cadmium for 1 day, and 36.6% (range, -9.2–73.5%) for the group eating meals containing a high level of cadmium for 3 days (Kikuchi et al., 2003).

2.3.2 Renal tubular dysfunction, clinical disease, and mortality

The prevalence of and prognosis associated with renal toxicity caused by cadmium has recently been investigated using human complex-forming glycoprotein (protein HC, also called alpha₁-microglobulin) as the primary effect biomarker. In the assessment of tubular proteinuria, protein HC is regarded as preferable to beta₂-microglobin because it is more stable at the typical pH of urine.

(a) Populations living in the vicinity of industrial sources of cadmium pollution

The OSCAR study involved all people aged 16–80 years who had lived for at least 5 years during the period 1910–1992 in the proximity of a nickel–cadmium battery plant in southern Sweden. An additional group of age- and sex-matched people was randomly selected from a general medical practice register in a nearby area and was included in this "environmentally-exposed" group. A group of workers at the battery plant was also enrolled. The overall participation rate was 60%, resulting in a total sample size of 1021. The 95th percentile of the concentration of protein HC in urine for a Swedish reference population was used to define tubular proteinuria: 0.8 mg/g creatinine in men and 0.6 mg/g creatinine in women. Thus, sex was taken into account by use of values from this reference population, although it was not certain that the age distribution of the reference population resembled that of the OSCAR study population. The mean (10th, 90th percentiles) urinary concentration of cadmium was 0.82 µg/g creatinine (0.18, 1.8) in men and 0.66 µg/g creatinine (0.21, 1.3) in women. A positive, linear relationship was found between urinary concentrations of cadmium and protein HC (p < 0.0001 in men, p = 0.0033 in women). The concentrations of protein HC for a total of 171 people exceeded the cut-offs defining tubular proteinuria, and the prevalence showed a clear dose–response relationship with urinary concentrations of cadmium (trend, p < 0.001 for the entire cohort, p = 0.001 for the environmentally-exposed sub-group only). Table 3 summarizes the prevalence and odds ratios (OR) for tubular proteinuria associated with different ranges of concentration of urinary cadmium, for the cohort as a whole and for the environmentally-exposed subgroup (Järup et al., 2000).

Table 3. Concentration of cadmium in urine and corresponding prevalence and odds ratios of tubular proteinuria

Adapted from Table 3 in Järup et al. (2000)

Additional analyses in which the concentration of cadmium in blood served as the biomarker for exposure were done in response to the criticism that a higher concentration of cadmium in urine might reflect not only higher exposure but might also be a marker for existing renal damage, complicating interpretation. Thus when tubular proteinuria is already present, the concentration of cadmium in blood might be a better estimate of dose. The mean (10th, 90th percentiles) concentration of cadmium in blood was 0.85 µg/l (0.15, 1.91) in men and 0.62 µg/l (0.17, 1.35) in women. Analyses similar to those carried out for urinary cadmium and protein HC and tubular proteinuria were conducted. A significant positive relationship was found between cadmium in blood and protein HC in both men and women, after adjusting for age, smoking, and lead in blood. Stratifying individuals into five groups on the basis of concentration of cadmium in blood revealed a dose–response relationship between cadmium in blood and tubular proteinuria: the concentration of cadmium in blood in the reference group was <0.56 µg/l; the remaining individ-uals were divided among four groups of roughly equal size corresponding to con-centrations of 0.56–0.79, 0.79–1.12, 1.12–1.69, and >1.69 µg/l. Elevated risks for tubular proteinuria, adjusted for age, sex, and smoking, were seen for the three groups with the higher concentrations of cadmium in blood (i.e. >0.79 µg/l) (Alfven et al., 2002).

A complete list was generated of residents (aged 0.5–75 years) of a community contaminated with various heavy metals as a result of past zinc-smelting operations, who had lived in their homes for the previous 0.5 year; residents were randomly selected from this list. Residents of another community 10 km away served as an unexposed comparison group. Untimed ("spot") urine samples were collected from 361 people aged 6–74 years (64% participation rate in the smelter town; 50% in the comparison area). The geometric mean concentration of cadmium in urine was 0.14 µg/g creatinine (95th percentile, 1.01; two individuals had a concentration of >2 µg/g creatinine). Levels of the biomarkers of early kidney damage NAG, alanine aminopeptidase, albumin, and beta₂-microglobulin were measured. Data for children aged 6–17 years and adults were analysed separately because of concerns about the influence of pubertal stage on the biomarkers. In children, concentration of urinary cadmium was not associated with any renal biomarker after adjustment for creatinine, age, and sex. In adults, concentration of cadmium in urine was positively associated with NAG, and alanine aminopeptidase after adjustment for creatinine, age, sex, smoking, and self-reported thyroid disease or diabetes. When adults were classified into five groups on the basis of creatinine-adjusted concentrations of cadmium in urine, NAG and alanine aminopeptidase showed a positive linear relationship, such that the levels of NAG and alanine aminopeptidase in people with the highest concentration of cadmium in urine (>1 µg/g creatinine) were 53% and 43%, respectively, higher than those of people with the lowest concentrations (<0.25 µg/g creatinine). No consistent relationship was found between concentration of cadmium in urine and concentrations of albumin or beta₂-microglobulin (although reduced statistical power might have been responsible for the latter, insofar as almost 50% of urine samples were excluded because of a pH of <6.0). The authors noted that the importance of this study lies in the relatively low exposures to cadmium at which some signs of early kidney damage were evident. The geometric mean concentration of cadmium in urine among the adults in the study sample was 0.23 µg/g creatinine, and all adults except two had a concentration of <2 µg/g, so the dose–effect relationship that was evident for some biomarkers pertained to concentrations of <2 µg/g creatinine. The authors cautioned, however, that it is not clear whether the changes in enzymes noted represent ". . . early, irreversible tubular damage or an overly sensitive indication of subclinical effects that will never progress to actual renal dysfunction . . ." (Noonan et al., 2002).

An evaluation of the association between renal function and relatively low exposures to cadmium was carried out in a population living within 8 km of nonferrous smelters in northern France. Four hundred children (aged 8.5–13 years) and 600 adults (aged 18–54 years) were compared with age- and sex-matched controls drawn randomly from towns of comparable size but without significant sources of exposure to heavy metals. Measurements were made of urinary albumin, transferrin, beta₂-microglobulin, NAG, creatinine, and mercury in urine, while cadmium, lead, and serum creatinine were measured in blood. The mean concentration of blood cadmium was higher in the polluted area (mean, 0.86 µg/l; range 0.04–5.99) than in the control area (mean, 0.64 µg/l; range, 0.04–2.53), but there was no significant difference in the prevalence of elevated concentrations of blood cadmium or indices of renal function between the two areas. Among children, a positive association was found between cadmium in blood and total NAG, although the magnitude of the correlation was modest (r = 0.25). The authors concluded that this finding probably reflects an "early subclinical excretory response of the renal proximal tubule" which, in the absence of increased exposure, is unlikely to progress to overt renal dysfunction. No significant associations between cadmium in blood and indices of renal function were observed among adults. Exposures to heavy metals were lower in this sample than in samples in the Cadmibel or Pheecad studies. The authors concluded that the exposures in this sample were lower than those required to produce a change in renal function (DeBurbure et al., 2003).

In another study of relatively low environmental exposure to cadmium, 800 pig producers in southern Sweden were randomly selected. Of the 48% who volunteered to participate in the study, detailed data from a questionnaire were collected and sampling of blood and urine was performed for 105 persons. Measurements were made of blood cadmium, haemoglobin, serum iron, total iron-binding capacity, serum albumin, serum creatinine, and serum ferritin. Urinary cadmium, beta₂-microglobulin, albumin, creatinine, NAG, and protein HC were also measured. The mean concentration of cadmium in blood was 0.26 µg/l (median, 0.20; range, 0.04–2.02), and the mean concentration of cadmiumin urine was 0.26 µg/g creatinine (median, 0.23; range, 0.07–0.99). beta₂-Microglobulin was significantly positively associated with urinary cadmium, after adjusting for age, whereas protein HC and NAG were not. The authors speculated that ". . . the causal contribution of cadmium to impairments [in HC and NAG] may have been underestimated due to overcontrolling for age, resulting from highly significant correlations between urinary cadmium and age". The most critical aspect of this study is that the concentration of cadmium in urine as <1 µg/g creatinine in all study participants (Olsson et al., 2002).

In a study conducted among 734 individuals aged >36 years in areas contaminated with cadmium in China, cadmium in blood and urine, and urinary retinal-binding protein were measured. Mean concentrations of cadmium among individuals living in a control area were 1.7 µg/l and 2.4 µg/g creatinine in blood and urine, respectively. In the two areas polluted with cadmium, the mean concentrations of cadmium in blood were 4.6 and 12.1 µg/l, and the mean concentrations of cadmium in urine were 4.8 and 14.5 µg/g creatinine. Among both men and women, concentrations of urinary retinal-binding protein were significantly increased at concentrations of urinary cadmium of >10 µg/g creatinine, and at concentrations of blood cadmium of >10 µg/l, using individuals with concentrations of <2 µg/g creatinine as the reference group. In addition, a dose–response analysis was conducted, using urinary retinal-binding protein at a concentration of >3 mg/g creatinine, adjusted for body mass index, to define an elevated level. Among men, again using men with concentrations of cadmium in urine of <2 µg/g creatinine as the reference group, a significantly increased prevalence of elevated concentration of urinary retinal-binding protein was observed among all groups with a concentration of >5 µg/g creatinine. Among women, a significant elevation was observed among all groups with concentrations of urinary cadmium of >10 µg/g creatinine. For both men and women, a significant increase in elevated concentrations of urinary retinal-binding protein was observed in all groups with concentrations of blood cadmium exceeding 10 µg/l (Nordberg et al., 2002).

Additional data are available on the clinical significance of cadmium-associated tubular dysfunction. Previous results from the Cadmibel study indicated that dysfunction was not progressive, i.e. did not lead to glomerular dysfunction, and was perhaps even reversible following reduction of exposure (Holtz et al., 1999). In studies of occupationally-exposed workers, risk was found to depend on the body burden of cadmium and on the severity of tubular proteinuria at the time that exposure was reduced. Tubular dysfunction appeared to be reversible when tubular proteinuria was mild (beta₂-microglobulin >300 but <1500 µg/g creatinine) and concentration of urinary cadmium never exceeded 20 µg/g creatinine. When tubular proteinuria was more severe and concentration of blood cadmium did exceed 20 µg/g creatinine, cadmium-associated tubular dysfunction was progressive, despite reduction in exposure to cadmium (Roels et al., 1997).

An ecological study examined the incidence of end-stage renal disease in the population residing near two nickel–cadmium battery plants in Kalmar county, Sweden. The study population consisted of all people aged 20–79 years who lived in the county between 1978 and 1995, classified into four groups by presumed exposure to cadmium: (1) high exposure: 655 occupationally-exposed workers in the battery plant who were employed for at least 1 year in either of the plants; (2) moderate exposure: 8825 people living within 2 km of either plant; (3) low exposure: 7200 people living 2–10 km from a plant; (4) not exposed: 152 477 people living in other areas of the county. County case registers were used to identify all incident cases of renal replacement therapy (renal dialysis or transplantation) between 1 January 1978 and 31 December 1995. Weighted Mantel-Haenszel rate ratios (MH-RR), stratified by sex and age, and directly age-standardized rate ratios (SRR) were calculated. The weighted MH-RR for renal replacement therapy in the group exposed to cadmium was 1.8 (95% CI, 1.3–2.3) and was greater in women (OR, 2.3; 95% CI, 1.5–3.5) than in men (OR, 1.5; 95% CI, 1.0–2.1). If analyses were restricted to people who were only exposed to cadmium environmentally, the MH-RR was unchanged, at 1.7 (95% CI, 1.3–2.3). When the measure of association was the SRR, a significant trend in the dose–response relationship was found (p <0.001), with an SRR of 1.4 (95% CI, 0.8–2.0) in the group with low exposure, 1.9 (95% CI, 1.3–2.5) in the group with moderate exposure, and 2.3 (95% CI, 0.6–6.0) in the group with high exposure. No measurements were made of bio-markers of exposure to cadmium in any of the participants, making assessment of exposure the weakest aspect of this study. The authors noted that movement into and out of the study area occurs but is limited, noting as well that any misclassification of exposure resulting from such a process would tend to reduce an association between residence and end-stage renal disease, rather than create a spurious association. Cross-sectional surveys conducted among residents in the study area revealed concentrations of urinary cadmium of 1–2 µg/g creatinine. To the extent possible given the limitations of the data, potential confounding of the association by other primary causes of chronic renal failure (e.g, cardiovascular diseases, diabetes, vascular or systemic diseases, use of analgesics) was discounted. As cadmium was not considered to be the primary cause of need for renal replacement therapy in this population, the authors accorded a secondary role to cadmium, suggesting that ". . . accumulated cadmium in the kidneys contributes to the deterioration of renal disease, such as chronic glomerulonephritis or secondary to diabetes" (Hellstrom et al., 2001).

The possible renal effects of combined exposure to cadmium and arsenic were studied using data from three studies: the Cadmibel and Pheecad studies conducted in Belgium, and a study conducted in China. The following biomarkers of exposure were available for all study samples: concentrations of cadmium in blood and urine, and concentrations of inorganic arsenic (including its mono-and dimethylated metabolites) in urine. The following biomarkers of renal function were available: urinary albumin (as an index of glomerular permeability), beta₂-microglobulin and retinal-binding protein (as indices of tubular dysfunction), and NAG. After adjusting for smoking, alcohol consumption, body mass index, diabetes, urinary tract disorders, and use of analgesics, no clear evidence of a significant interaction between cadmium and arsenic was found in either of the studies conducted in Belgium. Among Chinese women who were not exposed occupationally, both biomarkers of tubular dysfunction increased with increasing exposure to cadmium or arsenic alone, but combined exposure to both metals at high levels appeared to act synergistically. The corresponding concentrations of cadmium and arsenic among Chinese women with the highest exposures were: blood cadmium, 10 µg/l; urinary cadmium, 12 µg/g creatinine; urinary arsenic, 35–40 µg/g creatinine. The present reviewer concluded that these findings should be considered preliminary as the authors could not account for the fact that the interaction was significant among Chinese women but not Chinese men, and in Chinese women but not Belgian woman (Buchet et al., 2003).

(b) Studies in Japan

In Japanese research articles, the description "cadmium-nonpolluted area" does not necessarily mean that no cadmium was detected in the soil of the given area. Indeed, the concentration of cadmium in the soil might be higher than those found in the environment in the rest of the world. On the other hand, a "cadmium-polluted area" is defined by administrative authorities, from a risk management point of view, as an area in which the unpolished rice produced in the paddy fields contains a concentration of cadmium of >1 µg/kg. The description of "cadmium-polluted area" in some research articles does not necessarily mean that the given area is so designated by the authorities. The description applied by the authors is adopted throughout this section of the monograph.

General population studies: elevated levels of cadmium in the environment

Since the fifty-fifth meeting of the Committee (Annex 1, references 149, 150), Nogawa and coworkers have published a series of epidemiological studies that re-analyse the data obtained in the health survey performed in 1967 (Osawa et al., 2001; Kobayashi et al., 2002a).

An epidemiological study on the entire population, aged >30 years, of the Jinzu River basin was performed by targeting the subjects participating in the 1967 health survey conducted mainly in the heavily polluted area. From people who had eaten household rice of known concentration of cadmium, 1075 people (634 men and 441 women) who had either resided in their current household since birth or who had moved there from an unpolluted area (group A), and 780 people who had resided in their current household since birth (group B), were selected as the target population. The total intake of cadmium for each person was calculated from the amounts of cadmium ingested from rice and other foods, according to Nogawa’s formula:

(concentration of cadmium in rice × 333.5 g + 34 µg) × 365 days × number of years living in current household + 50 µg × 365 days × non-Jinzu River basin

where the daily mean rice intake, the mean daily intake of cadmium from foods other than rice in the region that was polluted with cadmium and the mean daily intake of cadmium in the region that was not polluted with cadmium were estimated to be 333.5 g, 34 µg and 50 µg, respectively. Logistic regression analysis was performed using the prevalence of abnormal urinary findings (proteinuria, glucosuria or both) as the criterion variable and the total intake of cadmium and age as explanatory variables. In people in groups A and B, the odds ratios became higher as the total intake of cadmium increased. Odds ratios for group A were statistically significant, except for glucosuria in men and proteinuria in women. In group B, the odds ratios were also significant for both proteinuria and glucosuria in men and glucosuria in women. The authors concluded that the greater the increase in total intake of cadmium, the greater the increase in abnormal urinary findings for people having resided in a region contaminated with cadmium in the Jinzu River basin (Kobayashi et al., 2002a).

Another paper on the same epidemiological study performed in 1967 (described above), reported on the relationship between renal dysfunction and concentration of cadmium in rice produced by individual hamlets in the Jinzu River basin, Japan. The authors suggested that people who had lived in the same hamlet since birth, eating rice containing low concentrations of cadmium, did not develop renal dysfunction until the period of residence was prolonged, while people who ate rice containing high concentrations of cadmium developed renal dysfunction even after a short period of residence in the hamlet. For people who had lived in the same hamlet for more than 30 years and were aged >50 years, the prevalence of urinary abnormalities in individual hamlets increased significantly with increases in the mean concentration of cadmium in rice, demonstrating that a dose–response relationship exists. The authors defined the allowable concentration of cadmium in rice as the concentration that does not exceed the reference values for prevalence of proteinuria, glucosuria or both. They calculated that the allowable concentration of cadmium in rice, as derived from the regression line, is estimated to be in the range of 0.05 to 0.20 mg/kg (Osawa et al., 2001). The Committee noted that this study was conducted nearly 30 years ago, that the participants were residing in areas that were polluted with cadmium nearly 20–30 years before the time of the study, and that lifestyles (including intake of rice as well as physical characteristics reflecting nutritional status) have changed. For example, average daily intake of rice in the region polluted with cadmium described above is estimated to have been approximately 340 g per day in the late 1960s, but approximately 160 g per day in 1999, according to the National Nutritional Survey performed by the Ministry of Health, Welfare and Labour (Osawa et al., 2001).

General population studies: normal levels of cadmium in the environment

Several research groups have investigated the renal effects of exposure to cadmium in people living in "cadmium-nonpolluted" areas.

A study of 1501 inhabitants (558 men and 743 women) aged >50 years was conducted in the Boso peninsula in Chiba prefecture. The concentration of urinary cadmium used as a surrogate for the internal dose was found to be 1.0 µg/l for men and 0.9 µg/l for women. The authors adopted total urinary protein, beta₂-microglobulin and NAG as indicators of renal dysfunction. Statistically significant relationships between the amount of cadmium excreted in the urine and these indicators of renal dysfunction were found using multiple regression analysis and a logistic regression analysis. The authors concluded that renal dysfunction was induced by exposure to a normal level of cadmium in the environment (Yamanaka et al., 1998).

Another epidemiological study was conducted by the same research group in two cadmium-nonpolluted areas in the Noto peninsula. Target populations of similar socioeconomic status were selected, comprising 875 people (346 men and 529 women) in area A and 635 people (222 men and 413 women) in area B, all of whom were aged >50 years. The geometric mean concentrations of urinary cadmium were 2.2 µg/l for men and 2.8 µg/l for women in area A, and 3.4 µg/l for men and 3.9 µg/l for women in area B. In the two locations, multiple regression analysis and logistic regression analysis showed that there was a significant association between concentration of urinary cadmium and the indicators of renal dysfunction. The authors concluded that renal dysfunction is significantly related to exposure to cadmium in the environment in cadmium-nonpolluted areas (Oo et al., 2000).

The target populations (1105 men and 1648 women, aged >50 years) in three cadmium-nonpolluted areas were combined for statistical analysis, as these areas are all rural and regarded as similar in terms of socioeconomic characteristics. The urinary cadmium concentrations of people living in the two areas in the Noto peninsula appeared to be higher than those of people living in the Boso penisula. Multiple regression analysis showed a statistically significant relationship between concentrations of cadmium in blood and urine, and indicators of renal dysfunction, while logistic regression analysis indicated that the probability that individual subjects would have abnormal values for the renal parameters was statistically significantly related to concentrations of cadmium in blood and urine (Suwazono et al., 2000).

In contrast to the studies described above, in which positive correlations between normal levels of cadmium in the environment and levels of urinary bio-markers were found, no association between environmental exposure to cadmium and renal dysfunction was been reported. In 2000–2001, 10 753 women (mostly aged 35–60 years), with no history of occupational exposure to cadmium, in 10 prefectures throughout Japan, were requested to provide morning urine specimens. The authors found that the geometric mean creatinine-corrected concentrations of urinary cadmium, alpha₁-microglobulin and beta₂-microglobulin were 1.26 µg/g creatinine (range, below limit of detection–20.9), 2.54 mg/g creatinine (range, below limit of detection–45.0) and 115 µg/g creatinine (range, below limit of detection–3862), respectively. Both multiple regression analysis and logistic regression analysis indicated that age was a confounding factor in evaluating the effect of urinary cadmium on concentrations of alpha₁-microglobulin and beta₂-microglobulin. While the logistic regression analysis including participants of all ages showed a positive influence of log(concentration of urinary cadmium) on log(concentration of alpha₁-microglobulin) and log(concentration of beta₂-microglobulin), this effect disappeared when logistic regression analysis was conducted for each age group separately (ages 41–50 years and 51–60 years (analyses were restricted to these two age groups in order to reduce the effect of age, according to the authors). When each age group was further divided into two subgroups according to concentrations of urinary cadmium, concentrations of alpha₁-microglobulin and beta₂-microglobulin were higher in the groups with a high concentration of urinary cadmium, but the prevalence of alpha₁-microglobulinuria and beta₂-microglobulinuria did not differ between the two groups, depending on the cut-off values used. In the overall evaluation, there was no clear-cut evidence to suggest that environmental exposure to cadmium induced tubular dysfunction among women aged 41–60 years in the general population in Japan (Ezaki et al., 2003).

In a subsequent analysis of the same target population, all women were classified into five groups according to concentration of urinary cadmium, and the geometric mean and its standard deviation for concentration of alpha₁-microglobulin and beta₂-microglobulin in the urine were calculated for each group. The largest geometric mean concentration of urinary cadmium among the 10 prefectures was 3.2 µg/g creatinine, and the maximum concentration of urinary cadmium observed was 20.9 µg/g creatinine. The prevalence of cases for which the cut-off values for each of alpha₁-microglobulin and beta₂-microglobulin were exceeded was found to increase with the amount of cadmium excreted in urine, without an apparent threshold. Since concentrations of alpha₁-microglobulin and beta₂-microglobulin were found to increase not only with increasing concentration of urinary cadmium, but also with increases in the concentration of other elements, such as magnesium and calcium, the authors concluded that no causal relationship exists between urinary cadmium and the low-molecular-mass protein biomarkers. Other contributing factors, such as age and other disease conditions, were not included (Ezaki et al., 2003).

In another epidemiological study, 44 men and 54 women were recruited from a cadmium-polluted area, and 21 men and 49 women from a cadmium-unpolluted area, or reference area. Diagnosis was carried out using biomarkers of exposure to cadmium, such as blood cadmium, as well as biomarkers for renal dysfunction, such as beta₂-microglobulin, alpha₁-microglobulin, NAG, total protein, inorganic phosphorus, lysozyme and creatinine; no cases of severe renal dysfunction caused by cadmium poisoning were observed. The geometric mean concentrations of urinary and blood cadmium in the cadmium-polluted area were 2.69 µg/g creatinine (range, 0.24–11.3) and 0.38 µg/dl (range, 0.2–1.2) for men, and 4.68 µg/g cre-atinine (range, 0.51–22.2) and 0.41 µg/dl (range, 0.1–1.1) for women, respectively. On the other hand, in the reference area the geometric means for these parameters were 1.08 µg/g creatinine (range, 0.20–3.92) and 0.21 µg/dl (range, 0.1–0.6) for men, and 1.69 µg/g creatinine (range, 0.12–5.08) and 0.25 µg/dl (range, 0.1–0.5) for women, respectively. The concentrations of cadmium in both urine and whole blood of the men and women living in the polluted area were significantly higher than those of people living in the reference area, but no difference in concentrations of urinary beta₂-microglobulin was found between the two areas. For women, urinary alpha₁-microglobulin was significantly higher in the polluted area than in the reference area. The correlation analysis, by either Pearson’s or Spearman’s method, depending upon the distribution of data, showed that beta₂-microglobulin, alpha₁-microglobulin and NAG were positively correlated with concentrations of cadmium in both urine and whole blood for men and women in the polluted area, with the exception of urinary beta₂-microglobulin and urinary cadmium in men. The index of estimated intake of cadmium from rice was expressed as the product of concentrations of cadmium in homegrown rice multiplied by daily frequency multiplied by duration (years) of residence in the polluted area. In multiple regression analysis, cadmium in whole blood was independently associated with the index of estimated intake of cadmium from rice, and with age and sex. Variations in whole blood cadmium accounted for a substantial portion of the variance in urinary cadmium; this relationship was weaker in older individuals. The concentration of cadmium in whole blood was the only independent variable that was related to variations in urinary beta₂-microglobulin, suggesting that the consumption of homegrown rice polluted with low concentrations of cadmium resulted in an elevation of concentration of cadmium in whole blood, and a consequent increase in urinary cadmium. On the other hand, no clear-cut relationship between the slight elevation in cadmium body burden and the increased excretion of urinary low-molecular-mass microglobulins was observed (Nakadaira & Nishi, 2003).

The shape of the dose–response relationship between concentration of urinary cadmium as a surrogate for the internal body burden of cadmium and concentration of urinary beta₂-microglobulin was analysed using the results of 12 research articles reporting on epidemiological studies carried out in Japanese populations (32 groups of men and 58 groups of women). The target study populations in these papers comprised four areas that had been officially designated as cadmium-polluted areas (the Jinzu River basin in Toyama prefecture, the Kakahashi River basin in Ishikawa prefecture, Kosaka in Akita prefecture and Tsushima Island in Nagasaki prefecture) and four cadmium-unpolluted areas. The relationship between concentrations of urinary beta₂-microglobulin and urinary cadmium, was found to be not linear, but in the shape of a hockey-stick. When the results for women who did not have elevated concentrations of urinary beta₂-microglobulin, or who had a concentration of either >400 or >1000 µg were examined by regression analysis, a threshold level for concentration of urinary cadmium in relation to an increase in urinary beta₂-microglobulin concentration was apparently present at a urinary cadmium concentration of approximately 11–12 µg/g creatinine. A similar trend and threshold value were found for both men and women (Ikeda et al., 2003). The threshold value estimated by Ikeda et al. (2003) is in agreement with that reported other studies (Van Sittert et al., 1993; Lauwreys et al., 1994; Zhang et al., 2002).

A total of 1482 adult women of various ages in six prefectures in Japan were asked to provide peripheral blood and spot urine samples and to answer questionnaires on their social habits and health conditions. The women were classified into groups of women with anaemia (40), with iron deficiency (526), and normal controls (916), using ferritin (concentration of ferritin, <20 ng/ml) and haemoglobin (concentration of haemoglobin, <10 g/100 ml) as classification indicators. Serum iron was lower and total iron-binding capacity was higher in accordance with levels in ferritin and haemoglobin in the groups of women with anaemia and with iron deficiency, although red blood cells counts stayed either essentially unchanged or only slightly reduced. In contrast, no increases in concentrations of urinary cadmium, alpha₁-microglobulin or beta₂-microglobulin were observed. The comparison of concentrations of urinary cadmium, alpha₁-microglobulin, and beta₂-microglobulin of the cases compared with those of controls matched by age and prefecture of residence showed that there was no substantial difference between women with anaemia and controls (40 pairs) and between women with iron deficiency and controls (391 pairs). For one case of clinical anaemia, however, concentration of urinary cadmium was higher than that among women of the same age range and from the same prefecture, although concentrations of alpha₁-microglobulin and beta₂-microglobulin remained unchanged. The authors concluded that the current level of iron deficiency among women in the general population in Japan might not cause any increase in absorption of cadmium or in tubular dysfunction induced by cadmium (Tsukahara et al., 2002). The re-analysis of the same study population led to the same conclusion (Tsukahara et al., 2003).

The possible confounding effects of age and correction for creatinine on the detection of renal tubular dysfunction were studied using essentially the same data as described above (Tsukahara et al., 2002, 2003), comprising a total of 817 never-smoking women aged 20–74 years. The geometric mean concentration of urinary cadmium was 1.41 µg/l (range, 0.5–7.9) or 1.3 µg/g creatinine. Of the four biomarkers of renal dysfunction used in this study, simple regression analysis showed that NAG had the most significant relationship to concentration of urinary cadmium, followed by alpha₁-microglobulin, beta₂-microglobulin, while urinary retinal-binding protein had the least significant relationship. The authors reported that both concentration and specific gravity (another correction parameter) of urinary creatinine were inversely related to age, and recommended the use of uncorrected observed values for biomarkers for renal dysfunction, rather than the traditionally-used creatinine correction, when a study population consists of a population with a wide range of ages (Moriguchi et al., 2003).

A total of 1381 female farmers living in five different locations around Japan were asked to give samples of their polished rice and bean paste (miso) and were asked what quantities they consumed daily, to determine the level of exposure to cadmium from main food items. One district that had not been known to produce rice that was highly contaminated with cadmium (>0.4 µg/g) was considered to be the reference. The female farmers in the five target populations showed a sequential difference in their levels of exposure to cadmium in the diet, ranging from a level as low as that of the general Japanese population to one close to the current provisional tolerable weekly intake (PTWI) (7 µg/kg bw per day). Urinary excretion of cadmium, an indicator of the accumulation of cadmium in the kidneys, increased along the same sequential pattern as exposure to cadmium in the diet. However, no differences were observed between the populations in the amounts of alpha₁-microglobulin and beta₂-microglobulin excreted in urine. According to multiple regression analysis, age rather than concentrations of cadmium in blood or urine showed larger standard partial regression coefficients. The authors concluded that the current PTWI is adequate to prevent renal dysfunction induced by cadmium in the general population (Horiguchi et al., 2004a).

2.3.3 Calcium metabolism and osteoporosis

It has been established that excessive exposure to cadmium affects the metabolism of calcium, leading to osteomalacia subsequent to proximal tubular dysfunction in the damaged kidneys; in the most severe cases, patients develop Itai-Itai disease, with osteomalacia as well as osteoporosis. Recently, it has been reported that bone mineral loss and disruption of calcium homeostasis has occurred in people in China, and that environmental levels of exposure to cadmium may induce osteoporosis either following or preceding the manifestation of renal dysfunction in China and Japan, as described below.

Urinary excretion of calcium was measured in residents of two villages in China, in an effort to evaluate whether this is a more sensitive and stable biomarker of renal damage associated with exposure to cadmium than beta₂-microglobulin or NAG. One village was located near a smelter and was considered to be polluted with cadmium, while the other village, of similar economic status, was not. Study participants aged >35 years were chosen randomly from village records: 252 residents from the polluted village and 246 residents from the control village. Urinary concentrations of cadmium, calcium, zinc, beta₂-microglobulin, NAG, and creatinine were measured. Most residents of the control village had concentrations of urinary cadmium of <2 mg/g creatinine, while most residents of the polluted village had concentrations of >5 mg/g creatinine. Concentrations of urinary calcium, beta₂-microglobulin, and NAG were significantly higher in the polluted village than in the control village, in both men and women. Using the distributions of values for the control village, cut-off thresholds for identifying residents with abnormal levels of urinary calcium, beta₂-microglobulin, and NAG were determined (calcium, >220 mg/g creatinine; beta₂-microglobulin, >800 and >900 µg/g creatinine for men and women, respectively; NAG, >120 µg/g creatinine). Significant dose–response relationships were found for all three biomarkers. Using individuals with urinary cadmium concentrations of <2 µg/g creatinine as the reference group, the prevalence of hypercalciuria was significantly increased among residents with urinary cadmium concentrations of between 2 and 5 µg/g creatinine. Linear models provided the best fit for all biomarkers of effect. The authors concluded that because concentration of urinary calcium increases in proportion to beta₂-microglobulin and NAG, it is a useful biomarker of early nephrotoxicity at low levels of exposure to cadmium in the environment (Wu et al., 2001).

A sample of 53 women aged 65–76 years (mean, 70.0 ± 3.3 years), with renal tubular dysfunction and living in cadmium-polluted areas in the Jinzu River basin were found to excrete a large quantities of cadmium into urine (mean, 17.2 µg/g creatinine). The women had severe renal dysfunction, characterized by fractional excretion of beta₂-microglobulin (FEbeta₂-microglobulin), with a mean value of 5.1 (range, 0.45–53%) compared to a reference range of 0.01–0.29%. At these high levels of exposure to cadmium, bone metabolism was altered and correlated with markers of renal tubular dysfunction, such as FE beta₂-microglobulin. The authors concluded that postmenopausal women who had been exposed to cadmium in the environment for many years, and who had severe renal dysfunction, also showed altered bone metabolism, potentially resulting in osteoporosis (Aoshima et al., 2003).

Concentrations of cadmium in blood and urine were measured in 734 individuals aged >36 years and living in one of two areas polluted with cadmium or in a control area, in China. Mean concentration of blood cadmium was 1.7 µg/l in controls, and 4.6 and 12.1 µg/l in persons in the two polluted areas. Concentrations of cadmium in urine were 2.4, 4.8, and 14.5 µg/g creatinine in persons in the three areas, respectively. A concentration of urinary cadmium of 20 µg/g creatinine or a blood cadmium concentration of 20 µg/l was considered to be high. Forearm bone mineral density was significantly lower among postmenopausal women with high concentrations of cadmium in blood or urine, and in men with high concentrations of cadmium in blood. A dose–response analysis was also performed, in which a forearm bone density of below the 10th percentile was the adverse outcome. Low bone-mineral density was significantly more frequent among men with either high concentrations of urinary cadmium or high concentrations of blood cadmium, among premenopausal women with high concentrations of blood cadmium, and among postmenopausal women with high concentrations of either blood or urinary cadmium (Nordberg et al., 2002).

Three epidemiological studies have been reported on the possible change in bone metabolism in persons who did not show signs of renal dysfunction. In the OSCAR study, bone mineral density was included among the study protocols (the design of which was described above in the section on renal toxicity). Bone mineral density was measured at a distal site on an individual’s non-dominant forearm while in the supine position. The degree of osteoporosis was expressed as an individual’s z-score, determined by comparing bone mineral density to age and sex norms. The analyses were restricted to study participants aged >60 years, and were adjusted for age, body weight, concentration of blood lead , and smoking. Bone mineral density was negatively correlated with concentration of blood cadmium for both women and men, although the association was statistically significant only for women. Dose–response analyses were conducted, in which the adverse outcome was an age- and sex-adjusted z-score of <-1. The odds ratios, adjusted for age, body weight, and smoking, were 2.0 (95% CI, 1.1–3.9) for individuals with concentrations of blood cadmium of 0.16–1.12 µg/l, and 2.9 (95% CI, 1.4–5.8) for individuals with concentrations of >1.12 µg/l (Alfven et al., 2002).

For a total of 908 women aged 40–88 years living in a urban, cadmium-unpolluted area in Japan, concentration of urinary cadmium, NAG activity, concentration of beta₂-microglobulin, and the stiffness index (STIFF) of calcaneal bone according to an ultrasound method were analysed. The concentration of urinary cadmium in the women (mean, 2.87 µg/g creatinine; range, 0.25–11.4 µg/g creatinine) showed a significant correlation with NAG activity but not with concentration of beta₂-microglobulin. STIFF was significantly inversely correlated with urinary cadmium, and the association remained significant after adjusting for age, body weight, and menstrual status, which suggests that cadmium has a significant effect on bone loss in these women who are without signs of kidney damage induced by cadmium. A two-fold increase in concentration of urinary cadmium was accompanied by a decrease in STIFF corresponding to an increase in age of 1.7 years. The author concluded that these results emphasize the need for reassessment of the significance of exposure to cadmium in the general Japanese population, and the urgent need to study the possible underlying mechanism (Honda et al., 2003).

To examine the hypothesis that low-level environmental exposure to cadmium, independent from renal dysfunction, increases the risk of osteoporosis and bone demineralization, a cross-sectional epidemiological study was conducted in five districts in Japan among a total of 1380 female farmers who had eaten rice contaminated by cadmium. Analysis of the data, grouped by urinary cadmium concentration and age-related menstrual status, suggested that cadmium accelerates the increase in excretion of urinary calcium around the time of menopause and the subsequent decrease in bone density after menopause. However, multivariate analyses showed no significant contribution of cadmium to bone density or urinary excretion of calcium, and the authors reported that the results described above were confounded by other factors, and that exposure to cadmium at a level that is insufficient to induce renal dysfunction does not increase the risk of osteoporosis (Horiguchi et al., 2004c).

2.3.4 Hypertension

Two new cross-sectional studies on exposure to cadmium, and blood pressure were identified, both of which were conducted by a group from Croatia, and included 267 nonsmoking women aged 40–85 years, with a mean blood cadmium concentration of 0.6 µg/l (range, 0.2–4.5). Blood cadmium was a significant predictor of systolic blood pressure, after adjusting for age and body mass index, while the association between blood cadmium and diastolic blood pressure was marginally significant, after adjusting for body mass index, age, area of residence, and alcohol consumption (Pizent et al., 2001).

The associations between blood pressure and several elements (cadmium, lead, copper, zinc and selenium) were studied in 154 men, aged 19–53 years, without known occupational exposures to these elements. The study sample consisted of men randomly selected from those attending a clinic for couples with fertility difficulties or men making voluntary sperm donations. Exclusion criteria included diabetes, coronary heart disease, cardiovascular or peripheral vascular disease, renal disease, or other diseases known to affect either blood pressure or the metabolism of metals. Men taking antihypertension medications were also excluded. The median concentration of blood cadmium was 0.83 µg/l (range, 0.21–11.93) and was highest among smokers. Blood pressure was measured twice in the sitting position. Blood cadmium was not significantly associated with either systolic or diastolic blood pressure in unadjusted or multivariate analyses adjusted for potential confounders. It is possible that when most of the exposure to cadmium within a cohort is the result of smoking, an association between blood cadmium and blood pressure is obscured if adjustment is made for factors, such as smoking, that are risk factors both for higher exposure to cadmium and for the adverse health outcome. If this were the case in this instance, however, it would be expected that the bivariate correlation (i.e. unadjusted) between blood cadmium concentration and blood pressure would have been expected to be significant, and it was not. Therefore, the author concluded that this study provides little evidence that increased blood cadmium concentration is associated with higher blood pressure (Telisman et al., 2001).

2.3.5 Diabetes

Data from the national health and nutrition examination (NHANES) III survey in the United States was used to evaluate the hypothesis, supported by some animal studies, that the risk of type II diabetes is increased by exposure to cadmium. Cadmium was measured in spot urine samples. On the basis of fasting plasma glucose (FPG) (8–24 h), 8722 subjects were classified as normal, having impaired fasting glucose (IFG, or prediabetes, concentration of FPG of 110–126 mg/dl), or diabetes (FPG >125 mg/dl or use of insulin or oral medication for hypoglycaemia). In order to exclude individuals with type I diabetes, analyses were restricted to individuals aged >40 years. Urinary cadmium was modeled both as a continuous variable and as categories (0.0–0.99, 1–1.99, >1.99 µg/g creatinine), and analyses were adjusted for age, race, sex, and body mass index. For both IFG and diabetes, prevalence increased with increasing urinary cadmium in a dose-dependent manner, after adjusting for age, sex, ethnicity, and body mass index. For IFG, the odds ratios were 1.5 (95% CI, 1.2–1.8) and 2.1 (95% CI, 1.4–3.0) for the groups with urinary cadmium concentrations of 1–1.99 and >1.99 µg/g creatinine, respectively (p for trend, <0.0001). For diabetes, the odds ratios were 1.2 (95% CI, 1.1–1.5) and 1.5 (95% CI, 1.1–2.0), respectively (p for trend, <0.0001). Excluding individuals with renal disease did not appreciably affect the results, and similar relationships were seen in smokers and in nonsmokers. The authors concluded that these analyses suggest an increased risk of type II diabetes associated with exposure to cadmium; the association between urinary cadmium and IFG among individuals without renal disease suggests that the increased urinary cadmium was not the result of the development of diabetes but rather preceded it. The present reviewer concluded that the cross-sectional design of this study, and the modest increase in risk noted across urinary cadmium categories, indicates that caution should be used in drawing any causal inferences (Schwartz et al., 2003).

2.3.6 Cancer

The evidence that exposure to cadmium also increases the risk of pancreatic cancer was examined by meta-analysis. Several lines of evidence support the plausibility of this hypothesis: smoking, an important route of exposure to cadmium, is associated with an increased risk of pancreatic cancer; pancreatic function is reduced in patients with Itai-Itai disease; rates of pancreatic cancer are higher in areas with high levels of industrial activity, production of rice or consumption of seafood, and among workers in industries that involve exposure to cadmium (e.g. production of pesticides, manufacture of paint and pigment, metal-working, soldering). The meta-analysis involved the calculation of standardized mortality ratios (SMR) for death from pancreatic cancer in four cohorts of occupationally-exposed workers. Owing to a lack of heterogeneity in SMRs in the separate cohorts, a summary SMR was calculated: 166 (95% CI, 98–280; p = 0.059). In the four cohorts (total n = 1769), a total of 14 deaths from pancreatic cancer occurred, compared to the 8.5 expected (Schwartz & Reis, 2000).

In a large case–control study nested within a large, population-based prospective study conducted in the United States, concentration of cadmium in toenails, determined before the diagnosis of prostate cancer, was not significantly higher among 119 patients with pathologically-confirmed prostate cancer, identified from county and state cancer registries, than in 227 controls matched by age, race, date of blood collection, and size of toenail clipping. Adjustments were made for education, height, current body mass index, body mass index at age 21 years, family history of prostate cancer, smoking, and use of multivitamins. No trend was apparent in the odds ratio of prostate cancer across toenail-cadmium categories. Concentrations of toenail-cadmium were higher, although non-significantly, in patients with extraprostatic prostate cancer or cancer of a high histological grade at diagnosis. Exposures to cadmium were generally low in this sample of the general population, with toenail-cadmium below the limit of detection in 65% of participants. The authors concluded that at exposures to cadmium typically found for the general population living in areas without substantial environmental exposures, the incidence of prostate cancer is unrelated to body burden of cadmium (Platz et al., 2002).

2.3.7 Reproductive toxicity

Maternal blood cadmium concentration is not highly correlated with blood cadmium concentration in the newborn, but it is correlated with cord-blood cadmium concentration, with correlation coefficients of 0.5–0.6 (Galicia-Garcia et al., 1997; Salpietro et al., 2002). Limited evidence suggests that neonatal outcomes are related to indices of prenatal exposure to cadmium (maternal urinary cadmium, maternal blood cadmium), such as reduced birth weight (Galicia-Garcia et al., 1997; Salpietro et al., 2002; Nishijo et al., 2002), and reduced length of gestation (Nishijo et al., 2002). The fact that, in the study by Nishijo et al., urinary cadmium was not significantly associated with infant height and weight after adjustment was made for gestational age suggested that fetuses with higher prenatal exposures to cadmium were of appropriate gestational age for date, but tended to be born earlier. In this study, conducted among women living in the Jinzu River basin in Toyama, the area in which Itai-Itai disease was most common, most adverse fetal outcomes were no longer significantly associated with urinary cadmium after adjustment was made for maternal age. Because the study area was known to have been contaminated with cadmium over a long period, maternal age is likely to have been a marker for duration of exposure and thus for body burden of cadmium. Thus, controlling for maternal age might represent overcontrol, and alternative analytical approaches, such as stratification, might have been preferable.

2.3.8 Neurotoxicity

In a case–control study in 22 patients with sporadic motor neuron disease, and 20 controls (partners of the patients with sporadic motor neuron disease), concentrations of cadmium in whole blood or in erythrocytes did not differ significantly between cases and controls, although plasma concentrations of cadmium were significantly higher in cases than controls. The authors questioned the biological significance of this difference because the distributions of plasma concentrations of cadmium for cases and controls overlapped considerably, and only for two cases of sporadic motor neuron disease were plasma concentrations of cadmium higher than the highest values of the controls. While the authors discounted a simple causal relationship between exposure to cadmium and sporadic motor neuron disease, they cautioned that, because metals enter motor neurons selectively, a defect that interferes with the intraneuronal handling of metals could result in damage to motor neurons, and they identified the need to explore gene–metal interactions that might contribute to interindividual variability in susceptibility (Pamphlett et al., 2001).

It was speculated that cadmium might contribute to the development of multiple sclerosis on the grounds that increased cadmium causes increased production of superoxide, which combines with nitric oxide to form peroxynitrite, a free radical that damages myelin. The present reviewer noted that no empirical evidence to support this hypothesis has been published (Johnson, 2000).

As part of a study of exposure to fluoride and rates of bone fracture among 1016 elderly Chinese people, the association between concentrations of cadmium in water and cognitive function was evaluated, as assessed by the community screening interview for dementia. Concentrations of cadmium in water were generally low (median, 0.25 µg/l; maximum, 0.49 µg/l). No association was observed between concentrations of cadmium in water and cognitive function, either before or after adjustment for age, sex, education, and presence of other metals. A significant interaction with zinc was seen: at low concentrations of cadmium in water, a higher concentration of zinc was associated with better cognition, while at high concentrations of cadmium in water, a higher concentration of zinc was associated with worse cognition (Emsley et al., 2000).

The distribution of cadmium in the brains of 11 patients with Alzheimer disease, 6 patients with senile involutive cortical changes (SICC), and 10 normal controls was compared: no significant differences were found. However, regional variations in the distribution of cadmium in the brains of the normal controls were more pronounced than in the brains of the patients with Alzheimer disease. For example, in the controls, the concentration of cadmium in the thalamus was significantly higher than in the medial temporal gyrus and the hippocampus. In the patients with Alzheimer disease, the concentration of cadmium in the thalamus was significantly greater than in the superior frontal gyrus, the superior parietal gyrus, the medial temporal gyrus, and the hippocampus. In addition, in both normal controls and patients with Alzheimer disease, the concentration of cadmium was significantly positively correlated with the number of neurofibrillary tangles. The Committee concluded that these findings, based on a small number of observations, are difficult to interpret, and do not provide a basis for drawing any inferences about the contribution of exposure to cadmium to development of Alzheimer disease (Panayi et al., 2002).

Previous studies had suggested that concentrations of cadmium were lower in the brains of patients with Alzheimer disease, particularly in the frontal lobes (Spyrou & Stedman, 1996, 1999).

2.3.9 Mortality

Several studies have investigated the association between concentration of cadmium in rice and mortality among the general population living in rural areas of the Jinzu River basin in Japan. A follow-up survey was conducted on 2101 inhabitants (1566 men and 535 women), who participated in a health survey in 1967 and had resided in their present rural community since birth. Communities were divided into two groups: those eating rice contaminated with cadmium at a concentration of <0.30 mg/kg or of >0.30 mg/kg. The association between concentration of cadmium in rice and mortality was expressed as SMRs and hazard ratio on the basis of a Cox proportional hazards model. In both sexes, SMRs tended to be greater in the group eating higher concentrations of cadmium in rice. Cox hazard ratios for men and women eating rice contaminated with cadmium at a concentration of >0.30 mg/kg were 1.42 and 1.10, respectively (significant in males) (Ishihara et al., 2001).

The conclusion that mortality is increased in locations in which the concentrations of cadmium in household rice are higher than normal was affirmed in an additional follow-up study of people living in the Jinzu River basin (Kobayashi et al., 2002b).

In the same target population, mortality (as expressed by SMRs) was analysed with regard to the severity of renal dysfunction characterized as proteinuria, glucosuria or both. The authors concluded that mortality was increased among both men and women living in the cadmium-polluted area in Jinzu River basin (Matsuda et al., 2002).

A possible contribution of employment in agriculture or number of pregnancies/births to mortality in these areas was assessed by dividing the target population by three different river/water basins, i.e. Jinzu River, non-Jinzu River and mixed water system. No association was found between these parameters and mortality (Kobayashi et al., 2002c).

The authors of these studies concluded that the development of renal damage induced by exposure to cadmium is likely to be responsible for the increased mortality, because the mean concentration of cadmium in rice produced in each rural community was closely related to the development of renal injury, in regions with high concentrations of cadmium in rice. The overall conclusion from these studies is that persons having an exceedingly high concentration of cadmium in the diet have an unfavourable life prognosis, although the contributions of socioeconomic factors, lifestyle factors, including smoking and drinking, and the presence of other diseases, must be taken into account.

The critical and most sensitive effect of long-term exposure to cadmium is renal tubular dysfunction, manifested as low-molecular-mass proteinuria. Studies in animals indicate that histological changes in the renal tubules occur at a dose lower than that reported to produce low-molecular-mass proteinuria. An estimate of the range of dietary exposures that result in a renal concentration of cadmium at which a small, but significant fraction of an exposed human population will show low-molecular-mass proteinuria was provided in the report of the fifty-fifth meeting of the Committee (Annex 1, reference 149). This estimate, 260–480 µg/day, was derived on the basis of certain critical assumptions: kidneys account for one-quarter to one-third of the body burden of cadmium; 5% gastrointestinal bioavailability of cadmium; a specified critical concentration of cadmium in the renal cortex; and a single-compartment, metabolic toxicokinetic model.

In the report of the fifty-fifth meeting of the Committee, a synthesis of data from several studies of workers and of general populations exposed environmentally was presented. A model was used to estimate the prevalence of cadmium-induced tubular proteinuria that would be expected to occur in individuals with specific concentrations of urinary cadmium and cadmium intake. These estimates suggested that the risk of tubular dysfunction begins to increase when the urinary excretion of cadmium exceeds 2.5 µg/g of creatinine. It is clear that the concentrations of cadmium in the diet and in urine are empirically related and that age is a significant determinant of the relationship. Estimates of dietary intake were derived from data from Japan, the United States, and Sweden. The mean dietary intake of cadmium by nonsmoking women in many areas of Japan was 25.5 µg/day (range, 19–51 µg/day), and the mean urinary excretion of cadmium was 4.4 µg/g of creatinine (range, 3.6–7.0 µg/g). These results indicate that the ratio of dietary intake of cadmium to urinary excretion of cadmium is 6 (range, 3–15). Using an estimate of the mean dietary intake of cadmium of 5.5 µg/day, from the United States total diet study, and an estimate of the mean urinary excretion of cadmium of 0.5 µg/g of creatinine, obtained (independently) from the United States NHANES survey, results in a cadmium intake:excretion ratio of 11. The data from Sweden indicate that for female nonsmokers, urinary excretion of cadmium is 0.15 µg/g creatinine, and median dietary intake is 10 µg/day (range, 5.7–26 µg/day). The estimated ratio of dietary cadmium intake to urinary excretion of cadmium ranges from 40 to 175.

The predictions used for this calculation were based on several key assumptions regarding values for excretion of creatinine, bioavailability and the amount of absorbed cadmium that was excreted in the urine. Furthermore, the variability in ratios of intake to excretion may partly be caused by the age of each study population. The confidence intervals for the point estimates used are unknown and therefore, as indicated previously (WHO, 2001), these values are associated with considerable uncertainty and might overestimate the risks associated with dietary intakes of cadmium.

In humans, the major route of exposure to cadmium is via food, particularly in the nonsmoking population (Syers & Gochfeld, 2001). The presence of cadmium in food results from contamination of soil and water. Crops differ with respect to absorption of cadmium, and cadmium is known to accumulate in the tissues (particularly in the liver and kidney) of terrestrial animals (particularly in the liver and kidney) and aquatic animals (particularly detritus feeders, such as molluscs). For these reasons, concentrations of cadmium in food vary widely between foods and between geographic regions.

Information on the dietary intake of cadmium was previously reported as part of the safety assessment conducted by the fifty-fifth meeting of the Committee (Annex 1, reference 149). For that assessment, national data on concentration of cadmium in foods and estimates of cadmium intake were summarized. Regional average concentrations were calculated and regional estimates of cadmium intake were calculated using the WHO Global Environment Monitoring System-Food Contamination Monitoring and Assessment Programme (GEMS/Food) regional diets (WHO, 1998). Additional data on concentrations in foods and national estimates of dietary intake of cadmium have become available since the assessment in 2000. These new data are described and summarized below.

3.2.1 National estimates of concentration of cadmium in food

Since the previous evaluation by the Committee, additional data on concentrations of cadmium in foods became available for nine countries and for the European Union. A summary of these data follows.

Australia

For the previous evaluation of cadmium by the Committee, the results of four Australian total diet studies conducted between 1992 and 1998 were submitted by the Food Standards Australia New Zealand (FSANZ). Results for about 120 foods from all four studies were aggregated and the median concentrations for commodity groups were reported. Results of the 20th Australian total diet study conducted in 2000–2001 were recently published (Food Standards Australia New Zealand, 2003). Although it was not possible to combine the most recent data with the aggregated data reported previously, the concentrations found in the 2000–2001 Australian total diet study were comparable to those reported previously to the Committee.

Croatia

In a study of common fresh fish and shellfish from various areas of the Adriatic Sea, a total of 61 samples of fish and shellfish were analysed for mercury, lead, arsenic and cadmium. Concentrations of cadmium were 0.002–0.020 mg/kg for finfish, 0.130 mg/kg for shellfish, and 0.142 mg/kg for mussels (Juresa & Blanusa, 2003).

European Union

In a paper submitted to the Committee at it present meeting (J.C. LeBlanc, submitted to JECFA for evaluation at its current meeting), dietary exposure to cadmium was estimated for the European Union and for France. For estimating intake for the European Union, data on concentrations of cadmium were mainly those included in a report of the Scientific Cooperation (European Commission, 1996a, b). These values represent the best estimates of levels of contamination in 13 Member States of the European Union. Mean levels of contamination were reported for 10 groups of commodities: milk, milk products, fats and oils, cereals, fruit, leafy vegetables, other vegetables, meat/meat products, offal, and fish/fish products. Values ranged from 0.002 mg/kg for milk to 2.8 mg/kg for offal.

France

Mean concentrations of cadmium were reported for 12 food categories (J.C. LeBlanc, submitted to JECFA for evaluation at its current meeting). These values were based on official surveys conducted by the French government and on data from the food industry. Average concentrations ranged from 0.005 mg/kg in milk/milk products to 0.114 mg/kg in offal.

Greece

Samples of 93 different foods and beverages were collected in 2000 and analysed for cadmium content (Karavoltsos et al., 2002). The highest concentrations of cadmium were found in snails (0.642 mg/kg) and mussels (0.130 mg/kg). The results of this study indicate lower concentrations for all food categories than those included in the previous evaluation by the Committee (Tsoumbaris & Tsoukali-Papadopoulou, 1994).

Japan

The government of Japan submitted the results of a number of surveillance studies conducted between 1997 and 2001. Analytical results for 41 903 samples representing 11 major food groups were reported. Of these samples, 37 250 were husked rice, for which the mean concentration of cadmium was 0.061 mg/kg.

Shimbo et al. (2001) reported concentrations of cadmium in foods analysed in 1998–2000. In this study, a total of 4113 samples of rice, bread, noodles and wheat flour were collected in 63 cities throughout Japan. Kikuchi et al. (2002) summarized results for 519 samples of more than 200 different foods and beverages commonly consumed in Japan.

These three sources of data were combined in order to estimate the average concentration of cadmium in Japanese foods. Concentrations submitted by the Japanese government were used where possible because of the large number of samples analysed. For wheat and wheat products, concentrations from the three sources were averaged because sample sizes were comparable. For other foods not reported by the Japanese government, data from Kikuchi et al. were used.

Lithuania

Results of analyses of cadmium in 13 foods, performed in Lithuania in 1999, were reported to the GEMS/Food contaminants database. Highest concentrations were for roots and tubers (0.014 mg/kg).

Nigeria

Samples of beverages (Onianwa et al., 1999) and other types of foods (Onianwa et al., 2000) were analysed for cadmium. In the latter study, single samples of 78 foods were analysed in duplicate. Concentrations for many foods were higher than those reported for most countries. The highest average concentration of cadmium (0.375 mg/kg) was estimated for dairy products.

Slovakia

Data on 147 different foods and beverages were submitted by Slovakia to the GEMS/Food contaminants database. These data summarized the results for more than 8500 samples analysed for cadmium in 2000–2001.

Spain

Samples of rice and wheat products were collected in Madrid in 1997 for analysis of heavy metals (Cuadrado et al., 2000). Average concentrations of cadmium ranged from 0.003 mg/kg in rice to 0.05 mg/kg in pasta. In another study conducted in 2000, 108 samples of foods were collected in seven cities in Catalonia and analysed for arsenic, cadmium, mercury and lead (Llobet et al., 2003). Concentrations of cadmium ranged from 0.002 mg/kg in milk to 0.036 mg/kg in fish and seafood.

3.2.2 Regional average concentrations of cadmium in food

For the previous evaluation by the Committee, aggregated data on typical (mean or median) concentrations in foods reported for each country were grouped by commodity and by geographic region, paralleling the five GEMS/Food regional diets. For the current evaluation, new data were combined with data reported previously, and the average concentrations were recalculated for each of the five GEMS/Food regions. In many cases, the regional average concentrations are derived from data for a limited number of countries and for small sample sizes, thus they may not be representative of the region as a whole. Additional uncertainty in the calculation of regional average concentrations is introduced by combining aggregated data that may have been derived from dissimilar groupings of foods (e.g. all vegetables combined, rather than separating leafy and root vegetables) and food forms (raw and processed). As in the previous assessment, averages were not weighted for sample size, since that information was not always available.

The updated regional average concentrations are presented in Table 4. On the basis of the number of countries and commodities for which data were available and the total number of samples represented, data for the European region were the most robust. Other regions were more limited in terms of the number of countries represented, the number of commodities reported, and the total number of samples analysed.

Table 4. Average concentration of cadmium in foods, by GEMS/Food region

Middle Eastern region

Data on concentrations of cadmium in foods were available only from two studies conducted in Greece, one of which was reported in the previous evaluation. In general, average concentrations reported in the more recent study were lower than those reported previously. The highest concentrations of cadmium were found in molluscs.

Far Eastern region

Average concentrations of cadmium in foods in the Far Eastern region are based on data from Japan and China. Of particular note are those data reported recently by the Japanese government (approximately 42 000 samples in all, of which 37 250 were rice). Average concentrations were highest for molluscs, cattle liver and kidney, and horsemeat.

African region

In the previous evaluation, no data were reported for countries in the African region. For the current assessment, data were available for Nigeria; however, only a single sample of each of 78 different foods was analysed. Owing to the limited amount of data available for this region, average concentrations for the European region were used for estimating dietary intakes.

Latin American region

For both the previous and current assessments, no data were available on concentrations of cadmium in foods for countries in the Latin American region. Regional averages for Europe were used as surrogate data for estimating dietary intakes.

European region

The European region represents European countries as well as other countries with similar dietary patterns (Australia, Canada, New Zealand, and the United States). The highest concentrations of cadmium were seen in molluscs and offal. Particularly high values were reported for horse offal; average concentrations for this commodity were calculated separately from other offal to avoid overestimating the mean concentration in offal as a whole.

3.3.1 Intake estimates based on GEMS/Food regional diets

Intakes of cadmium were estimated for each of the five GEMS/Food regions by multiplying the updated regional average concentrations in a commodity by the amount consumed of the commodity as specified in the GEMS/Food regional diets. As in the previous assessment, if no data on concentrations were available on specific commodities in a region, surrogate data on similar foods within that region were used when estimating dietary intakes. In the absence of data on similar foods, the average concentrations in foods for Europe were used. Intakes from all commodities were summed to estimate the total intake of cadmium, by region. Updated estimates of intake for each of the five regions are summarized in Table 5. Total intakes ranged from 0.378 µg/kg bw per day for the Latin American region to 0.617 µg/kg bw per day for the Far Eastern region. These estimates are similar to those reported by the Committee in its previous evaluation (0.352 to 0.632 µg/kg bw per day).

Table 5. Estimates of dietary intake of cadmium, by GEMS/Food region (µg/kg bw per day)

Estimates of total intake of cadmium were also considered relative to the current PTWI of 7 µg/kg bw per day. Estimated regional intakes were converted from a per capita basis to a body-weight basis by dividing the total intake by standard body weights (55 kg for the Far Eastern region and 60 kg for the other regions). The current PTWI was divided by seven to derive a tolerable daily intake (1 µg/kg bw per day). Total intake of cadmium expressed as a percentage of the daily tolerable intake ranged from approximately 40% for the Middle Eastern region to approximately 60% for the Far Eastern region (Table 6).

Table 6. Intake and approximate contribution of major commodity groups to provisional tolerable daily intake, by GEMS/Food region^a

The Committee also reviewed the contributions of specific commodities to the provisional tolerable intake of cadmium (Table 7); several commodities were noteworthy in this regard. Rice contributed approximately 40% of the tolerable intake of cadmium in the Far Eastern region. Wheat contributed approximately 20% of the tolerable intake in the Middle Eastern region. In the African and European regions, starchy roots and tubers contributed approximately 10% of the tolerable intake. Vegetables (excluding leafy vegetables) contributed more than 5% of the tolerable intake in the Middle Eastern and European regions. Molluscs contributed approximately 10% of the tolerable intake in the European region.

Table 7. Intake and approximate contribution to provisional tolerable daily intake^a for specific commodities, by GEMS/Food region

3.3.2 National estimates of intake

Since the previous assessment by the Committee, estimates of dietary intake of cadmium have been reported for the European Union, France, Japan, and Spain. The more recent estimates of intake are described below. All national intake estimates are summarized in Table 8.

Table 8. National estimates of dietary intake of cadmium

Australia

Intakes of cadmium were calculated from concentrations of contaminants as determined in the 2000–2001 Australian total diet study and estimates of con-sumption from the 1995 national nutrition survey (Food Standards Australia New Zealand, 2003). Dietary intakes of cadmium were reported for six population groups as a range of lower to upper bound estimates, which were based on two average concentrations of cadmium (assuming values of zero and the limit of detection for samples in which cadmium was not detected). At the lower bound, intakes ranged from from 0.07 µg/kg bw per day for adults to 0.18 µg/kg bw per day for children aged 2 years. The estimates are similar to those reported for Australia in the previous evaluation by the Committee.

European Union

Intake estimates for the European Union were derived from data representing its Member States. Food consumption in each country as reported for SCOOP task 4.1 (European Commission, 1996a) was supplemented with information from the GEMS/Food regional diets (J.C. LeBlanc, submitted to JECFA for evaluation at its current meeting). For concentrations of cadmium, those reported by the European Commission (1996b) were considered to be the best estimates in 13 Member States of the European Union. Average intake of cadmium was estimated to be 0.86 µg/kg bw per day for the European Union as a whole.

France

Intakes of cadmium were estimated from the contamination data described above and from data on national food consumption collected in 1993–1994 (Association Sucre/Produits Sucrés, Communication, Consommation (ASPCC); Jean-Luc Volatier, 1994, personal communication). Risk analysis software¹ was used to model the probability distribution of intake (J.C. LeBlanc, submitted to JECFA for evaluation at its current meeting).

Japan

The government of Japan provided estimates of intake of cadmium from the Japanese total diet studies conducted between 1981 and 2001. Average intake over the five-year period of 1997–2001 was 0.52 µg/kg bw per day (assuming a body weight of 60 kg).

Spain

Llobet et al. (2003) reported intakes of cadmium as determined by the total diet study conducted in Catalonia in 2000. Daily intakes were estimated to be 0.020 and 0.026 µg/kg bw per day for adult men and women, respectively.

Overall, national estimates of intake of cadmium ranged from 0.1 to 0.9 µg/kg bw per day (0.7 to 6.3 µg/kg bw per week). These estimates are similar to those calculated from GEMS/Food regional diets and regional average concentrations (0.4–0.6 µg/kg bw per day). Most of the intakes reported in this evaluation are based on average intakes by the population as a whole or by adults. Australia is the only country for which intakes by children were reported. For infants aged 9 months, intakes were estimated to be between 0.13 and 0.68 µg/kg bw per day at the lower and upper bounds, respectively. For children aged 2 years, a similar range of intakes was reported (0.18–0.57 µg/kg bw per day).

3.3.3 High intakes of cadmium

The intake estimates presented above reflect average intakes, using estimates of average food consumption and average concentration of cadmium. Data reported for France also took into account estimates at the upper percentiles of intake for the population as a whole (Table 9).

Table 9. Estimated intake of cadmium by the total population in France

Although the intakes reported by GEMS/Food region (Tables 5 and 6) and by individual countries (Table 8) are estimates of average intake, intakes by high consumers can be approximated from these values. In general, total food consumption at the upper percentiles is approximately twice the mean (WHO, 1985), and high consumption of individual commodities is approximately three times the mean. Based on the range of average intakes reported by GEMS/Food region (0.4–0.6 µg/kg bw per day or 2.8–4.2 µg/kg bw per week) and by individual countries (0.1–0.9 µg/kg bw per day or 0.7–6.3 µg/kg bw per week), it is possible that for some individuals intake of cadmium may exceed the PTWI of 7 µg/kg bw.

Results of studies of processing of rice, wheat and soya beans were submitted by Japan. A study of the effects of cooking on concentrations of cadmium in rice was also carried out. A study of changes in the cadmium content of rice during the polishing process showed that the concentration of cadmium decreased only slightly during milling. Six samples of polished rice were prepared from each of six varieties of brown rice, using a commercial-scale rice mill. The cadmium content of the samples of brown rice ranged from 0.016 to 0.121 mg/kg. The cadmium content of polished rice was reduced by only 3%, regardless of the concentration in the unprocessed samples. Another study measured the effects of cooking (including washing and soaking in water) on the cadmium content of rice. Only slight decreases (average, 5%) in the cadmium content were shown to occur after cooking. In another study on the effect of milling on wheat, the cadmium content of flours was about half that of wheat grain. Concentrations of cadmium in bran were about twice that in wheat grain. In studies of processed soya bean products, concentrations of cadmium in tofu, miso, and soya sauce were about 65%, 80% and 50%, respectively, that of unprocessed soya beans.²

In the animal species tested, the oral bioavailability of cadmium ranged from 0.5 to 3.0%, on average. Experimental studies also identified various factors that can significantly influence the extent of absorption and retention of cadmium from the diet, including sex, developmental stage, and nutritional status. Low dietary concentrations of protein and of essential minerals such as zinc, calcium, copper, and iron have been shown to promote the absorption of cadmium, while, in contrast, high or adequate dietary concentrations reduce absorption and retention. After absorption, cadmium is distributed mainly to the liver, with subsequent redistribution to the kidney in conjugated forms such as cadmium–metallothionein and cadmium–albumin.

Long-term oral exposure to cadmium resulted in a variety of progressive histopathological changes in the kidney, including proximal tubule epithelial cell damage, interstitial fibrosis, and glomerular basal cell damage with limited tubular cell regeneration. Biochemical indications of renal damage were seen in the form of low molecular mass proteinuria, glucosuria and aminoaciduria. Tubular dysfunction also caused the urinary excretion of cadmium to increase. Decreases in bone calcium concentrations and increased urinary excretion of calcium have also been associated with exposure to cadmium. Cadmium induced the malignant transformation of animal and human cells in vitro.

Investigations into the ability of cadmium compounds to induce developmental effects in experimental animals have shown that decreased fetal weight, skeletal malformations and increased fetal mortality are common findings, usually in combination with indices of maternal toxicity. However, developmental neurobehavioural effects, including decreased locomotor and exploratory activity and certain electrophysiological changes, have been seen in the absence of any overt symptoms of maternal toxicity and appear to be a more sensitive indicator of toxicity.

A variety of effects on the immune system have been observed in experimental animals exposed to cadmium, including increased virus-induced mortality in mice co-exposed to non-lethal doses of cadmium and RNA viruses.

A number of new epidemiological studies published since the fifty-fifth meeting have evaluated the relationship between exposure to cadmium and various health effects, particularly renal dysfunction, mortality, and calcium/bone metabolism.

Cadmium accumulates in the kidney and, because of its long half-life in humans, steady-state concentrations in the renal cortex are reached only after about 40 years.

Recent studies conducted in Japan, Europe, China, and the United States have attempted to refine estimates of the dose–effect/dose–response relationship between environmental exposure to cadmium and renal dysfunction. In a Swedish study (the OSCAR study) involving >1000 individuals aged 16–80 years, an increase of nearly three-fold in the prevalence of tubular proteinuria was observed in the group with urinary cadmium concentrations of 0.5–1 µg/g creatinine, compared to the group with a urinary cadmium concentration of <0.3 µg/g creatinine. Above a urinary cadmium concentration of 5 µg/g creatinine, the prevalence of tubular proteinuria was increased five fold. Two studies of populations with low concentrations of urinary cadmium (mean concentrations of 0.23 µg/g creatinine and 0.26 µg/g creatinine) found associations between markers of early kidney damage and urinary cadmium concentration. However, the findings of these two studies were inconsistent; although urinary beta₂-microglobulin and NAG were measured as indices of tubular dysfunction in both studies, in one study, only beta₂-microglobulin was associated with urinary cadmium concentration while in the other study, only NAG was associated with urinary cadmium concentration. In an environmental study, the prevalence of end-stage renal disease was found to be significantly, although modestly, related to the extent of environmental exposure to cadmium, as determined by area of residence. However, individual biomarkers of exposure were not measured in this study. In aggregate, the new data are consistent with the hypothesis that low-level environmental exposure to cadmium is associated with an increased prevalence of renal proximal tubular dysfunction, as assessed by biomarkers.

The epidemiological studies conducted in regions of Japan where levels of environmental cadmium vary identified several issues that complicate the interpretation of studies of renal function and low environmental exposure to cadmium. In some studies, a crude association between urinary cadmium and a biomarker of effect disappeared after adjusting for age. Simple adjustment for creatinine might be misleading if comparisons involve people differing in physique, physical activity, sex, age, and race. The appropriate concentrations of urinary biomarkers to use as cut-off values for identifying tubular proteinuria might also vary depending upon physiological or disease conditions. Finally, the long-term health implications of the changes in renal function observed at low concentrations of urinary cadmium are uncertain.

It is well-established that cadmium-induced low molecular mass proteinuria can progress to an acquired Fanconi syndrome (the continuous loss of calcium and phosphorus into urine) and/or the disturbance of vitamin D metabolism in the damaged kidneys. The latter may eventually progress to Itai-Itai disease, characterized by osteomalacia.

Some recent reports suggest that environmental exposure to cadmium, even at low concentrations, might alter calcium metabolism in bone tissue independently of renal effects, and might increase the risk of osteoporosis and bone demineralization. According to the results of the OSCAR study, the age- and sex-adjusted risk of having a reduced bone mineral density was increased two-fold among individuals with blood-cadmium concentrations of 0.6–1.1 µg/l and three-fold among individuals with blood-cadmium concentrations >1.1 µg/l. This association was corroborated by the results of two earlier studies, one in Belgium and one in Japan, although bone mineral density was correlated with age and body weight, and only weakly with urinary cadmium concentration. Two studies in Japan, one in which environmental exposure to cadmium was moderate and one in which it was high, showed no correlation between exposure to cadmium and bone mineral density or calcium excretion, after adjustment for age, body mass index, and menstrual status. The excretion of calcium was not correlated with exposure to cadmium, but with deterioration of renal tubular function, which was due mainly to ageing.

Bone metabolism is influenced by many factors, including age, estrogen status, physique, physical activity, nutritional status, ethnic group, and environmental factors such as sunlight. None of the studies adjusted for possible confounding by all of these factors. These studies were therefore considered by the Committee to be preliminary.

The Committee reviewed additional studies of the associations between exposure to cadmium and other non-renal health effects, including diabetes, hypertension, carcinogenicity, reproductive outcomes, and neurotoxicity. The Committee found the results of these studies to be too preliminary to serve as the basis for its evaluation. The Committee took note, however, of a study that indicated that the prevalence of type 2 diabetes was significantly increased at urinary cadmium concentrations exceeding 1 µg/g creatinine among individuals without evidence of renal disease. Further work is needed to clarify the contribution of exposure to cadmium to this disease.

At its fifty-fifth meeting, the Committee evaluated the dietary intake of cadmium using data from a number of countries. At its present meeting, the Committee updated its review by adding new information from Australia, Croatia, France, Greece, Japan, Lithuania, Nigeria, Slovakia, Spain, and the European Union. The combined data showed that concentrations of cadmium range from about 0.01–0.05 mg/kg in most foods, although higher concentrations were found in nuts and oil seeds, molluscs, and offal (especially liver and kidney). Estimates of the mean national intake of cadmium ranged from 0.7–6.3 µg/kg bw per week. Mean dietary intakes derived from GEMS/Food regional diets (average per capita food consumption based on food balance sheets) and average concentrations of cadmium in these regions range from 2.8–4.2 µg/kg bw per week. These estimates constitute approximately 40–60% of the current PTWI of 7 µg/kg bw. For some individuals, the total intake of cadmium might exceed the PTWI because total food consumption for high consumers is estimated to be about twice the mean. Regarding the major dietary sources of cadmium, the following foods contributed 10% or more to the PTWI in at least one of the GEMS/Food regions: rice, wheat, starchy roots/tubers, and molluscs. Vegetables (excluding leafy vegetables) contribute >5% to the PTWI in two regions.

The Committee considered an extensive amount of new information, particularly from a series of Japanese environmental epidemiological studies, that addressed issues identified as research needs at its fifty-fifth meeting. The Committee reaffirmed its conclusion that renal tubular dysfunction is the critical health outcome with regard to the toxicity of cadmium. Although the sensitive bio-markers used by some recent studies conducted in Japan, Europe and the USA indicated that changes in renal function and bone/calcium metabolism are observed at urinary cadmium concentrations of <2.5 µg/g creatinine, the Committee noted that appreciable uncertainty remains regarding the long-term health significance of these changes. In addition, the Committee noted inconsistencies between studies regarding the specific biomarkers of renal function that were most commonly associated with urinary cadmium concentrations. Although recent studies suggested that increased concentrations of cadmium biomarkers are associated with health effects such as diabetes, hypertension, pancreatic cancer, fetal growth, and neurotoxicity, the Committee concluded that these data were not, at this time, sufficiently robust to serve as a basis for the evaluation. The Committee reaffirmed its conclusion that an excess prevalence of renal tubular dysfunction would not be expected to occur if urinary cadmium concentration remains <2.5 µg/g creatinine, even under a range of plausible assumptions about the relationship between the amount of bioavailable cadmium in the diet and the urinary excretion of cadmium. Uncertainty remains about how these assumptions affect the predicted excess prevalence of renal tubular dysfunction at concentrations of urinary cadmium of >2.5 µg/g creatinine. The Committee concluded that the new data which became available since its fifty-fifth meeting do not provide a sufficient basis for revising the PTWI, and therefore maintained the current PTWI of 7 µg/kg body weight. No excess prevalence of renal tubular dysfunction would be predicted to occur at the current PTWI under the most appropriate assumptions about the fractional bioavailability of cadmium and the percentage of the absorbed cadmium that is excreted in urine. The Committee noted that two issues being considered by the Joint FAO/WHO Project to Update the Principles and Methods for the Risk Assessment of Chemicals in Food were of particular relevance to the present evaluation: the dose–response assessment of biomarkers of effect and their relationship to disease outcome, and the possible specification of longer tolerable intake periods (e.g. PTMI) for contaminants with longer biological half-lives. The Committee recommended that the evaluation of cadmium should be revisited when this project has been completed.

Achanzar, W.E., Webber, M.M. & Waalkes, M.P. (2002) Altered apoptotic gene expression and acquired apoptotic resistance in cadmium-transformed human prostate epithelial cells. Prostate, 52, 236–244.

Agency for Toxic Substances and Disease Registry (1999) Toxicological profile for cadmium.

Alfven, T., Järup, L. & Elinder, C.-G. (2002) Cadmium and lead in blood in relation to low bone mineral density and tubular proteinuria. Environ. Health Perspect., 110, 699–702.

Aoshima, K., Fan, J., Cai, Y., Katoh, T., Teranishi, H. & Kasuya, M. (2003) Assessment of bone metabolism in cadmium-induced renal tubular dysfunction by measurements of bio-chemical markers. Toxicol. Let., 136, 183–192.

Brako, E.E., Wilson, A.K., Jonah, M.M., Blum, C.A., Cerny, E.A., Williams, K.L. & Bhattacharyya, M.H., (2003) Cadmium pathways during gestation and lactation in control versus metallothoinein 1,2-knockout mice. Toxicol. Sci., 71, 154–163.

Brzóska, M.M., Kamiski, M., Supernak-Bobko, D., Zwierz, K. & Moniuszko-Jakoniuk, J. (2003) Changes in the structure and function of the kidney of rats chronically exposed to cadmium. I. Biochemical and histopathological studies. Arch. Toxicol. Bull., 49, 453–457.

Buchet, J.-P., Heilier, J.-F., Bernard, A., Lison, D., Jin, T., Wu, X., Kong, Q. & Nordberg, G. (2003) Urinary protein excretion in humans exposed to arsenic and cadmium. Int. Arch. Occup. Environ. Health, 76, 111–120.

Chen, J. & Gao J. (1993). The Chinese total diet study in 1990. Part I. Chemical contaminants. JAOAC Int., 76, 1193–1205.

Chen, J. & Gao J. (1997a) [The Chinese total diet study in 1992—Chemical contaminants, a comparison among various regions] J. Hyg. Res., 26, 199–203 (in Chinese).

Chen, J. & Gao J. (1997b) [The Chinese total diet study in 1992—Chemical contaminants, a. comparison among various age groups.] J. Hyg. Res., 26, 255–261 (in Chinese).

Coyle, P., Philcox, J.C., Carey, L.C. & Rofe, A.M. (2002) Metallothionein: the multipurpose protein. Cell. Mol. Life Sci., 59, 627–647.

Cuadrado, C.J., Kumpulainen, A., Carbajal O. & Moreiras (2000). Cereals contribution to the total dietary intake of heavy metals in Madrid, Spain. J. Food Comp. Anal., 13, 495–503.

Dabeka, R.W. & McKenzie, A.D. (1995) Survey of lead, cadmium, fluoride, nickel, and cobalt in food composites and estimation of dietary intakes of these elements by Canadians in 1986–88. JAOAC Int., 78, 897–909.

DeBurbure, C., Buchet, J.-P., Bernard, A., Leroyer, A., Nisse, C., Haguenoer, J.-M., Bergamaschi, E. & Mutti, A. (2003) Biomarkers of renal effects in children and adults with low environmental exposure to heavy metals. J. Toxicol. Environ. Health, 66, 783–798.

Desi, I., Nagymajtenyi, L. & Schulz, H. (1998) Behavioural and neurotoxicological changes caused by cadmium treatment of rats during development. J. Appl. Toxicol., 18, 63–70.

Emsley, C.L., Gao, S., Li, Y., Liang, C., Ji, R., Hall, K.S., Cao, J., Ma, F., Wu, Y., Ying, P., Zhang, Y., Sun, S., Unverzagt, F.W., Slemenda, C.W. & Hendrie, H.C. (2000) Trace element levels in drinking water and cognitive function among elderly Chinese. Am. J. Epidemiol., 151, 913–920.

European Commission (1996a) Reports on tasks for Scientific Cooperation, Task 4.1. Improvement of knowledge of food consumption with a view to protection of public health by means of exchanges and collaboration between database managers.

European Commission (1996b) Reports on tasks for Scientific Cooperation, Task 3.2.4. Dietary exposure to cadmium.

Ezaki, T., Tsukahara, T., Moriguchi, J., Furuki, K., Fukui, Y., Ukai, H., Okamoto, S., Sakurai, H., Honda, S. & Ikeda, M. (2003) No clear-cut evidence for cadmium-induced renal tubular dysfunction among over 10,000 adult women in general Japanese population; a nation-wide large-scale survey. Int. Arch. Occup. Environ. Health, 76, 186–196.

Ezaki, T., Tsukahara, T., Moriguchi, J., Furuki, K., Furkui, Y., Ukai, H., Okamoto, S., Sakurai, H., Honda, S. & Ikeda, M. (2003) Analysis for threshold cadmium levels in urine to induce tubular dysfuncton among women in non-polluted areas in Japan. Int. Arch. Occup. Environ. Health, 76, 197–204.

Food Standards Australia New Zealand (2003) The 20th Australian Total Diet Survey. Canberra: FSANZ.

Galicia-Garcia, V., Rojas-Lopez, M., Rojas, R., Oliaz, G. & Rios, C. (1997) Cadmium levels in maternal, cord and newborn blood in Mexico City. Toxicol. Let., 91, 57–61.

Habeebu, S.S., Liu, J., Liu, Y. & Klaassen, C.D. (2000) Metallothionein-null mice are more sensitive than wild-type mice to liver injury induced by repeated exposure to cadmium. Toxicol. Sci., 55, 223–232.

Hellstrom, L., Elinder, C.-G., Dahlberg, B., Lundberg, M., Järup, L., Persson, B. & Axelson, O. (2001) Cadmium exposure and end-stage renal disease. Am. J. Kidney Dis., 38, 1001–1008.

Holtz, P., Buchet, J.-P., Bernard, A., Lison, A. & Lauwerys, R. (1999) Renal effects of low-level environmental cadmium exposure: 5-year follow-up of a subcohort from the Cadmibel Study. Lancet, 354, 1508–1513.

Hiratsuka, H., Satoh, S., Satoh, M., Nishijima, M., Katsuki, Y., Suzuki, J., Nakagawa, J., Sumiyoshi, M., Shibutani, M., Mitsumori, K., Tanaka-Kagawa, T. & Ando, M. (1999) Tissue distribution of cadmium in rats given minimum amounts of cadmium-polluted rice or cadmium chloride for 8 months. Toxicol. Appl. Pharmacol., 160, 183–191.

Honda, R., Tsuritani, I., Noborisaka, Y., Suzuki, H., Ishizaki, M. & Yamada, Y. (2003) Urinary cadmium excretion is correlated with calcaneal bone mass in Japanese women living in an urban area. Environ. Res., 91, 63–70.

Horiguchi, H., Oguma, E., Sasaki, S., Miyamoto, K., Ikeda, Y., Machida, M. & Kayama, F. (2004a) Environmental exposure to cadmium at a level insufficient to induce renal tubular dysfunction does not affect bone density among female Japanese farmers, Environ. Res. (in press).

Horiguchi, H., Oguma, E., Sasaki, S., Miyamoto, K., Ikeda, Y., Machida, M. & Kayama, F. (2004b) Comprehensive study of the effects of age, iron deficiency, diabetes mellitus and cadmium burden on dietary cadmium absorption in cadmium-exposed female Japanese farmers. Toxicol. Appl. Pharmacol. (in press).

Horiguchi, H., Oguma, E., Sasaki, S., Miyamoto, K., Ikeda, Y., Machida, M. & Kayama, F. (2004c) Dietary exposure to cadmium at close to the current provisional tolerable weekly intake does not affect renal function among female Japanese farmers. Environ. Res. (in press).

House, W.A., Hart, J.J., Norvell, W.A. & Welch, R.M. (2003) Cadmium absorption and retention by rats fed durum wheat (Triticum turgidum L. var. durum) grain. Brit. J. Nutr., 89, 499–509.

Ikeda, M., Ezaki T., Tsukahara, T., Moriguchi, J., Furuki, K., Fukui, Y., Ukai, H., Okamoto, S. & Sakurai, H. (2003) Threshold levels of urinary cadmium in relation to increases in urinary beta2-microglobulin among general Japanese populations. Toxicol. Let., 137, 135–141.

Ishihara, T., Kobayashi, E., Okubo, Y., Suwazono, Y., Kido, T., Nishijyo, M., Nakagawa, H. & Nogawa, K. (2001) Association between cadmium concentration in rice and mortality in the Jinzu River basin, Japan. Toxicol., 163, 23–28.

Järup, L., Berglund, M., Elinder, C.G., Nordberg, G. & Vahter, M. (1998) Health effects of cadmium exposure—a review of the literature and a risk estimate. Scand. J. Work Environ. Health, 24(Suppl 1), 1–51.

Järup, L., Hellstrom, L., Alfven, T., Carlsson, M.D., Grubb, A., Persson, B., Pettersson, C., Spang, G., Schutz, A. & Elinder, C.-G. (2000) Low level exposure to cadmium and early kidney damage: the OSCAR study. Occup. Environ. Med., 57, 668–672.

Jin, Y.H., Clark, A.B., Slebos, R.J.C., Al-Refai, H., Taylor, J.A., Kunkel, T.A., Resnick, M.A. & Gordenin, D.A. (2003) Cadmium is a mutagen that acts by inhibiting mismatch repair. Nat. Genet., 34, 326–329.

Johnson, S. (2000) The possible role of gradual accumulation of copper, cadmium, lead and iron and gradual depletion of zinc, magnesium, selenium, vitamins B2, B6, D, and E and essential fatty acids in multiple sclerosis. Med. Hypoth., 55, 239–241.

Juresa, D. & M. Blanusa (2003). Mercury, arsenic, lead and cadmium in fish and shellfish from the Adriatic Sea. Food Addit. Contam., 20, 241–246.

Karavoltsos, S., Sakellari, A., Dimopoulos, M., Dasenakis, M. & Scoullos, M. (2002) Cadmium content in foodstuffs from the Greek market. Food Addit. Contam., 19, 954– 962.

Kašuba, V., Rozgaj, R., Saric, M.M. & Blanusa, M. (2002) Evaluation of genotoxic damage of cadmium chloride in peripheral blood of suckling Wistar rats. J. Appl. Toxicol., 22, 271–277.

Kikuchi, Y., Nomiyama, T., Kumagai, N., Uemura, T. & Omae, K. (2002) Cadmium concentration in current Japanese foods and beverages. J. Occup. Health, 44, 240– 247.

Kikuchi, Y., Nomiyama, T., Kumagai, N., Dekio, F., Uemura, T., Takebayashi, T., Nishiwaki, Y., Matsumoto, Y., Sano, Y., Hosoda, K., Watanabe, S., Sakurai, H. & Omae, K. (2003) Uptake of cadmium in meals from the digestive tract of young non-smoking Japanese female volunteers. J. Occup. Health, 45, 43–52.

Kobayashi, E., Okubo, Y., Suwazono, Y., Kido, T. & Nogawa, K. (2002a) Dose–response relationship between total cadmium intake calculated from the cadmium concentration in rice collected from each household of farmers and renal dysfunction in inhabitants of the Jinzu River basin, Japan. J. Appl. Toxicol., 22, 431–464.

Kobayashi, E., Okubo, Y., Suwazono, Y., Kido, T., Nishijo, M., Nakagawa, H. & Nogawa, K. (2002b) Association between total cadmium intake calculated from the cadmium concentration in household rice and mortality among inhabitants of the cadmium-polluted Jinzu River basin of Japan. Toxicol. Let., 129, 85–91.

Kobayashi, E., Okubo, Y., Suwazono, Y., Kido, T., Nishijo, M., Nakagawa, H. & Nogawa, K. (2002c) Influence of years engaged in agriculture and number of pregnancies and deliveries on mortality of inhabitants of the Jinzu River basin area, Japan. Occup. Environ. Med., 59, 847–850.

Kreis, I.A., Wijga, A. & van Wijnen, J.H. (1992) Assessment of the lifetime accumulated cadmium intake from food in Kempenland. Sci. Total Environ, 127, 281–292.

Kurata, Y., Katsuta, O., Doi, T., Kawasuso, T., Hiratsuka, H., Tsuchitani, M. & Umemura, T. (2003) Chronic cadmium treatment induces islet B cell injury in ovariectomized cynomolgus monkeys. Japan J. Vet. Res., 50, 175–183.

Leret, M.L., Millan, J.A. & Antonio, M.T. (2003) Perinatal exposure to lead and cadmium affects anxiety-like behaviour. Toxicol., 186, 125–130.

Llobet, J.M., Falco, G., Casas, C., Teixido, A. & Domingo, J.L. (2003) Concentrations of arsenic, cadmium, mercury, and lead in common foods and estimated daily intake by children, adolescents, adults, and seniors of Catalonia, Spain. J. Agric. Food Chem., 51, 838–842.

Masaoka, T., Akahori, F., Arai, S., Nomiyama, K., Nomiyama, H., Kobayashi, K., Nomura, Y. & Suzuki T. (1994) A nine-year chronic toxicity study of cadmium ingestion in monkeys. I. Effects of dietary cadmium on the general health of monkeys. Vet. Hum. Toxicol., 36, 189–194.

Matsuda, T., Kobayashi, E., Okubo, Y., Suwazono, Y., Kido, T., Nishijo, M., Nakagawa, H. & Nogawa K. (2002) Association between renal dysfunction and mortality among inhabitants in the region around the Jinzu River basin polluted by cadmium Environ. Res., 88, 156–163.

Miles, A.T., Hawksworth, G.M., Beattie, J.H. & Rodilla, V. (2000) Induction, regulation, degradation, and biological significance of mammalian metallothioneins. Crit. Rev. Biochem. Mol. Biol., 35, 35–70.

Ministry of Agriculture, Fisheries & Food (MAFF), UK (1999b) 1997 Total diet study—aluminium, arsenic, cadmium, chromium, copper, lead, mercury, nickel, selenium, tin and zinc. Food Surveillance Information Sheet Number 191.

Misra, R.R., Smith, G.T. & Waalkes, M.P. (1998) Evaluation of the direct genotoxic potential of cadmium in four different rodent cell lines. Toxicol., 126, 103–114.

Moriguchi, J., Ezaki, T., Tsukahara, T., Furuki, K., Fukui, Y., Okamoto, S., Ukai, H., Sakurai, H., Shimbo, S. & Ikeda M. (2003) Comparative evaluation of four tubular dysfunction markers, with special references to the effects of aging and correction for creatinine concentration, Toxicol. Let., 143, 279–290.

Nagymajtenyi, L., Schulz, H. & Desi, I. (1997) Behavioural and functional neurotoxicological changes caused by cadmium in a three-generational study in rats. Hum. Exp. Toxicol., 16, 691–699.

Nakadaira, H. & Nishi, S. (2003) Effects of low-dose cadmium exposure on biological examinations. Sci. Total Environ., 308, 49–62.

Nakamura, K., Yasunaga, Y., Ko, D., Xu, L.L., Moul, J.W., Peehl, D.M., Srivastava, S. & Rhim,

J.S. (2002) Cadmium-induced neoplastic transformation of human prostate epithelialcells. Int. J. Oncol., 20, 543–547.

Nishijo, M., Nakagawa, H., Honda, R., Tanebe, K., Saito, S., Teranishi, H. & Tawara, K. (2002) Effects of maternal exposure to cadmium on pregnancy outcome and breast milk. Occup. Environ. Med., 59, 394–397.

Nishiyama, S., Itoh, N., Onosaka, S., Okudaira, M., Yamamoto, H. & Tanaka, K. (2003) Dietary cadmium inhibits spontaneous hepatocarcinogenesis in C3H/HeN mice and hepatitis in A/J mice, but not in C57BL/6 mice. Toxicol. Appl. Pharmacol., 186, 1–6.

Noonan, C.W., Sarasua, S.M., Campagna, D., Kathman, S.J., Lybarger, J.A. & Mueller, P.W. (2002) Effects of exposure to low levels of environmental cadmium on renal biomarkers. Environ. Health Persp., 110, 151–155.

Nordberg, G., Jin, T., Bernard, A., Fierens, S., Buchet, J.-P., Ye, T., Kong, Q. & Wang, H. (2002) Low bone density and renal dysfunction following environmental cadmium exposure in China. Ambio, 31, 478–481.

Olsson, I.-M., Bensryd, I., Lundh, T., Ottosson, H., Skerfving, S. & Oskarsson, A. (2002) Cadmium in blood and urine-Impact of sex, age, dietary intake, iron status, and former smoking-Association of renal effects. Environ. Health Persp., 110, 1185–1190.

Onianwa, P.C., Adetola, I.G., Iwegbue, C.M.A., Ojo, M.G. & Tella, O.O. (1999) Trace heavy metals composition of some Nigerian beverages and food drinks. Food Chem., 66, 275–279.

Onianwa, P.C., Lawal, J.A., Ogunkeye, A.A. & Orejimi B.M. (2000). Cadmium and nickel composition of Nigerian foods. J. Food Com. Anal., 13, 961–969.

Oo, Y.K., Kobayashi, E., Nogawa, K., Okubo, Y., Suwazono, Y., Kido, T. & Nakagawa H. (2000) Renal effects of cadmium intake of a Japanese general population in two areas unpolluted by cadmium. Arch. Envion. Health, 55, 98–103.

Osawa, T., Kobayashi, E., Okubo, Y., Suwazono, Y., Kido, T. & Nogawa K. (2001) A retrospective study on the relation between renal dysfunction and cadmium concentration in rice in individual hamlets in the Jinzu River basin, Toyama Prefecture, Japan. Environ. Res., 86, 51–59.

Park, J.D., Cherrington, N.J. & Klaassen, C.D. (2002) Intestinal absorption of cadmium is associated with divalent metal transporter 1 in rats. Toxicol. Sci., 68, 288–294.

Pamphlett, R., McQuilty, R. & Zarkos, K. (2001) Blood levels of toxic and essential metals in motor neuron disease. Neurotoxicology, 22, 401–410.

Panayi, A.E., Spyrou, N.M., Iversen, B.S., White, M.A. & Part, P. (2002) Determination of cadmium and zinc in Alzheimer’s brain tissue using inductively coupled plasma mass spectrometry. J. Neurol. Sci., 195, 1–10.

Pizent, A., Jurasovie, J. & Telisman, S. (2001) Blood pressure in relation to dietary calcium intake, alcohol consumption, blood lead, and blood cadmium in female nonsmokers. J. Trace Elem. Med. Biol., 15, 123–130.

Platz, E.A., Helzlsouer, K.J., Hoffman, S.C., Morris, J.S., Baskett, C.K. & Comstock, G.W. (2002) Prediagnostic toenail cadmium and zinc and subsequent prostate cancer risk. Prostate, 52, 288–296.

Roels, H.A., Van Assche, F.J., Oversteyns, M., Groof, M., Lauwerys, R.R. & Lison, D. (1997) Reversibility of microproteinuria in cadmium workers with incipient tubular dysfunction after reduction of exposure. Am. J. Ind. Med., 31, 645–652.

Radike, M., Warshawsky, D., Caruso, J., Goth-Goldstein, R., Reilman, R., Collins, T., Yaeger, M., Wang, J., Vela, N., Olsen, L. & Schneider, J. (2002) Distribution and accumulation of a mixture of arsenic, cadmium, chromium, nickel, and vanadium in mouse small intestine, kidneys, pancreas, and femur following oral administration in water or feed. J. Toxicol. Environ. Health A, 65, 2029–2052.

Reeves, P.G. & Chaney, R.L. (2001) Mineral status of female rats affects the absorption and organ distribution of dietary cadmium derived from edible sunflower kernels (Helianthus annuus L.). Environ. Res., 85, 215–225.

Reeves, P.G. & Chaney, R.L. (2002) Nutritional status affects the absorption and whole-body and organ retention of cadmium in rats fed rice-based diets. Environ. Sci. Technol., 36, 2684–2692.

Sabolic, I., Ljubojevic, M., Herak-Kramberger, C.M. & Brown, D. (2002) Cd-mt causes endocytosis of brush-border transporters in rat renal proximal tubules. Am. J. Phys.-Renal Phys., 283, F1389–F1402.

Salpietro, C.D., Gangemi, S., Minciullo, P.L., Briuglia, S., Merlino, M.V., Stelitano, A., Cristani, M., Trombetta, D. & Saija, A. (2002) Cadmium concentration in maternal and cord blood and infant birth weight: a study on healthy non-smoking women. J. Perinat. Med., 30, 395–399.

Sanzo, J.M., Dorronsoro, M., Amiamo, P., Aguinagalde, F.X., Azpiri, M.A. & the EPIC Group of Spain (2001) Estimation and validation of mercury intake associated with fish consumption in and EPIC cohort of Spain. Public Health Nutr., 4, 981–988.

Saric, M.M., Blanusa, M., Piasek, M., Varnai, V.M., Juresa, D. & Kostial, K. (2002) Effect of dietary calcium on cadmium absorption and retention in suckling rats. Biometals, 15, 175–182.

Schwartz, G.G. & Reis, I.M. (2000) Is cadmium a cause of human pancreatic cancer? Cancer Epidemiol., Biomarkers Prev., 9, 139–145.

Schwartz, G.G., Il’Yasova, D. & Ivanova, A. (2003) Urinary cadmium, impaired fasting glucose, and diabetes in the NHANES III. Diabet. Care, 26, 468–470.

Seth, P., Husain, M.M., Gupta, P., Schoneboom, A., Grieder, B.F., Mani, H. & Maheshwari, R.K. (2003) Early onset of virus infection and up-regulation of cytokines in mice treated with cadmium and manganese. Biometals, 16, 359–368.

Shibutani, M., Mitsumori, K., Niho, N., Satoh, S., Hiratsuka, H., Satoh, M., Sumiyoshi, M., Nishijima, M., Katsuki, Y., Suzuki, J., Nakagawa, J. & Ando, M. (2000) Assessment of renal toxicity by analysis of regeneration of tubular epithelium in rats given low-dose cadmium chloride or cadmium-polluted rice for 22 months. Arch. Toxicol., 74, 571– 577.

Shimbo, S., Zhang, Z.-W., Watanabe, T., Nakatsuka, H., Matsuda-Inoguchi, N., Higashikawa, K. & Ikeda, M. (2001) Cadmium and lead contents in rice and other cereal products in Japan in 1998–2000. Sci. Total Environ., 281, 165–175.

Spyrou, N.M. & Stedman, J.D. (1996) Cadmium concentrations in the brains of Alzheimer’s cases. Trans. Am. Nucl. Soc., 74, 1–3.

Spyrou, N.M., Stedman, J.D. & Cutts, D.A. (1999) Trace elements in Alzheimer’s disease. Res. Pract. Alzheimer’s Dis., 2, 125–129.

Suwazono, Y., Kobayashi, E., Okubo, Y., Nogawa, K., Kido, T. & Nakagawa, H. (2000) Renal effects of cadmium exposure in cadmium nonpolluted areas in Japan. Environ. Res., 84, 44–55.

Syers, J.K. & Gochfeld, M. eds. (2001) Environmental cadmium in the food chain: sources, pathways, and risks. Proceedings of the SCOPE Workshop, Brussels, Belgium, 13–16 September 2000. Brussels: Scientific Committee on Problems of the Environment (SCOPE).

Takeda, A., Takefuta, S., Ijiro, H., Okada, S. & Oku N. (1999) ¹⁰⁹Cd transport in rat brain. Brain Res., 49, 453–457.

Takenaka, S., Oldiges, H., Konig, H., Hochrainer, D. & Oberdorster, G. (1983) Carcinogenicity of cadmium chloride aerosols in Wistar rats. J. Nat. Cancer Inst., 70, 367–373.

Tang, W. & Shaikh, Z.A. (2001) Renal cortical mitochondrial dysfunction upon cadmium metallothionein administration to Sprague-Dawley rats. J. Toxicol. Environ. Health, A, 63, 221–235.

Telisman, S., Jurasovic, J., Pizent, A. & Cvitkovic, P. (2001) Blood pressure in relation to biomarkers of lead, cadmium, copper, zinc, and selenium in men without occupational exposure to metals. Environ. Res., 87, 57–68.

Tsoumbaris, P. & Tsoukali-Papadopoulou, H. (1994). heavy metals in common foodstuff: daily intake. Bull. Environ. Contam. Toxicol., 53, 67–70.

Tsukahara, T., Ezaki, T., Moriguchi, J., Furuki, K., Ukai, H., Okamoto, S., Sakurai, H. & Ikeda M. (2002) Effects of iron-deficiency anemia on cadmium uptake or kidney dysfunction are essentially nil among women in general population in Japan. Tohoku J. Exp. Med., 197, 243–247.

Tsukahara, T., Ezaki, T., Moriguchi, J., Furuki, K., Fukui, Y., Ukai, H., Okamoto, S., Sakurai, H. & Ikeda, M. (2003) No significant effect of iron deficiency on cadmium body burden or kidney dysfunction among women in the general population in Japan. Int. Arch. Occup. Environ. Health., 76, 275–281.

United States. Environmental Protection Agency (EPA) (1999) Toxicological review: cadmium and compounds. Integrated risk information system. (http://www.epa.gov/iris/subst/0141.htm).

Van Sittert, N.J., Ribbens, P. H., Huissman, B. & Lugtenburg, D. (1993) A nine year follow-up study of renal effects in workers exposed to cadmium in a zinc ore refinery. Br. J. Ind. Med., 50, 603–612.

Vanoort, R., Cressey, P. & Silvers, K. (2000) 1997/98 New Zealand Total Diet Survey. Part 2: Elements. Report prepared for the Ministry of Health.

Watanabe, T., Nakatsuka, H., Satoh, H., Yamamoto, R.M. & Ikeda, M. (1992) Reduced dietary cadmium intake in past 12 years in a rural area in Japan. Sci. Total Environ., 119, 43–50.

Whelton, B.D., Peterson, D.P., Moretti, E.S., Dare, H. & Bhattacharyya, M.H. (1997) Skeletal changes in multiparous, nulliparous and ovariectomized mice fed either a nutrient-sufficient or -deficient diet containing cadmium. Toxicol., 119, 103–121.

WHO (1985) Guidelines for the study of dietary intakes of chemical contaminants. WHO Offset Publication No. 87, Geneva, p. 54.

WHO (1992) Cadmium (Environmental Health Criteria 134), Geneva.

WHO (1998) GEMS/Food Regional Diets (WHO/FSF/FOS/98.3), GMS/Food Programme, Geneva.

Wu, X., Jin, T., Wang, Z., Ye, T., Kong, Q. & Nordberg, G. (2001) Urinary calcium as a biomarker of renal dysfunction in a general population exposed to cadmium. J. Occup. Environ. Med., 43, 898–904.

Yamanaka, O., Kobayashi, E., Nogawa, K., Suwazono, Y., Sakurada, I. & Kido, T. (1998) Association between renal effects and cadmium exposure in cadmium-nonpolluted area in Japan. Env. Res, Section A, 77, 1–8.

Yang, H.-F., Luo, X.-Y., Shen, W., Zhou, Z.-F., Jin, C.-Y., Yu, F. & Liang, C.-S. (1994). National food contamination monitoring programmes-levels of mercury, lead and cadmium in Chinese foods. Biomed. Environ. Sci., 7, 362–368

Zalups, R.K. & Ahmad, S. (2003) Molecular handling of cadmium in transporting epithelia. Toxicol. Appl. Pharmacol., 186, 163–188.

Zhang, G., Lindars, E., Chao, Z., Bai, Y. & Spickett, J. (2002) Biological monitoring of cadmium exposed workers in a nickel-cadmium battery factory in China. J. Occup. Health, 44, 15–21.

ENDNOTES:

1. @RISK, Version 4.5 (2000), Palisade Corporation, Newfield, NY, USA.
2. Annex 3-1; Moriyama, T., Shindo, K., Taguchi, Y., Watanabe, H., Yasui, A. & Teruo JOH, T. (2003), Changes in the cadmium content of rice during the milling process (in Japanese), The Journal of the Food Hygienic Society of Japan, 41, 145–149.
   Annex 3-3; Moriyama, T., Taguchi, Y., Watanabe, H. & Joh, T. (2002) Changes in the cadmium content of wheat during the milling process (in Japanese). In: Report on Risk Evaluation of Cadmium in Food, Research on Environmental Health, Health Sciences Research Program, Ministry of Health, Labour and Welfare, pp. 153–162.
   Annex 3-2 and Annex 3-4; Shindo, K. & Yasui, A. (2002) The Changes in cadmium content during cooking and processing (in Japanese). In: Report on Risk Evaluation of Cadmium in Food, Research on Environmental Health, Health Sciences Research Program, Ministry of Health, Labour and Welfare, pp. 163–169.

    See Also:
       Toxicological Abbreviations
       Cadmium (EHC 134, 1992)
       Cadmium (ICSC)
       Cadmium (WHO Food Additives Series 4)
       Cadmium (WHO Food Additives Series 24)
       Cadmium (WHO Food Additives Series 55)
       CADMIUM (JECFA Evaluation)
       Cadmium (PIM 089)