Fatal oral doses of stannous chloride or sodium pentafluorostannite initially caused a number of signs of toxicity, including extreme gastrointestinal irritation, anorexia, depression, ataxia, and muscular weakness. Mottling, hyperaemia, and tubular necrosis were seen in the kidneys of rats (Conine et al., 1975). A number of alterations in tissue integrity have also been reported, including necrosis of the liver and spleen in mice (Pelikan et al., 1968).

In a study of the toxicity of tin in various animal species, orange and apple juices and a pear nectar were enriched artificially with tin to concentrations of 500–2000 ppm (orange juice), 600 ppm (apple juice), and 750 ppm (pear nectar). Other foods including evaporated milk, minced mackerel, puréed asparagus, and puréed gherkins containing 400–470 ppm of tin were also tested. No adverse effects were reported in groups of six rats given a single oral dose of tin-containing fruit juices providing 5.4 mg of tin or after 24-h ingestion of a beverage containing tin at 65–190 mg/kg bw or of tin citrate containing tin at 36–300 mg/kg bw (Benoy et al., 1971).

Rabbits receiving a daily oral dose of 1 g of stannous acetate or tartrate for 10 days died within 37–43 days (Cheftel, 1967).

Oral administration of tin at a dose of 45 mg/kg bw induced vomiting and diarrhoea in cats (Omori, 1966). In subsequent studies with adult cats weighing 1.7–2.9 kg, administration of 10 ml/kg bw of orange juice containing 470 ppm of tin (4.7 mg/kg bw) caused vomiting in about 80% of the cats. Oral administration of a tin complex prepared from stannic chloride and sodium citrate induced severe salivation and emesis in all animals given tin at concentrations ³ 9 mg/kg bw (Omori et al., 1973). Cats receiving stannous citrate by gavage at a rate of 9 mg/kg bw showed salivation, vomiting, and diarrhoea but no acute effects at neutral pH (Cheftel, 1967).

Eleven cats of mixed breed and each sex were given juices containing tin at concentrations of 540–2000 ppm by stomach tube. When the cats were given apple juice or pear nectar containing 600 and 750 ppm of tin, respectively, there were no signs of toxicity; the doses corresponded to 3.0 and 3.8 mg/kg bw. However, orange juice containing 540 ppm of tin, providing a dose of 5.4 mg/kg bw, induced vomiting in 1 of 11 cats, and a higher incidence of vomiting was induced by higher concentrations and doses. Thus, 2 of 11 cats were affected at 7.0 mg/kg bw (700 ppm of tin); three of 10 were affected at 14 mg/kg bw (1400 ppm of tin); four of 10 were affected at 10 mg/kg bw (1000 ppm of tin); and two of nine were affected at 20 mg/kg bw (2000 ppm of tin). Administration of tin at 14 mg/kg bw (1400 ppm) 1 or 5 months later did not induce vomiting in 10 cats (Benoy et al., 1971).

No vomiting or signs of toxicity were seen in groups of four dogs that received single doses of beverages containing tin at 2.5–14 mg/kg bw or solid foods containing tin at 4.5–8.0 mg/kg bw (Benoy et al., 1971).

No adverse effects were reported in pigeons given oral doses of fruit juice containing tin at 1.5–3.0 mg/kg bw (Benoy et al., 1971).

2.2.2 Short-term studies of toxicity

Two groups of 10 young male rats were fed diets containing stannous chloride at 0 or 5000 ppm (0 or 0.5%), equivalent to 1500 mg/kg bw, for 1 month. Each animal also received ¹¹³SnCl₂ (2 pCi/day) in an HCl/KCl solution every day. The control animals received the carrier only. Two additional groups of rats were treated as described for 1 month and were used for histological examination of various organs and tissues. Body weight and food consumption were depressed, but food use efficiency, protein use efficiency, and the nitrogen balance were within normal limits. Ninety-nine per cent of the administered radiolabel was excreted in the faeces and less than 1% in the urine. Little was found in the gastrointestinal tract, organs, or carcass. The treated animals developed anaemia characterized by a significant decrease in haemoglobin and haematocrit values. The relative weights of the liver, spleen, and kidney were increased. Histological examination of the treated animals revealed marked congestion of the kidney and adrenal cortex and congestion and desquamation of the mucosa in the upper gastrointestinal tract from the stomach to the ileum (Fritsch et al., 1977).

Groups of 10 male and 10 female weanling Wistar-derived rats were given diets containing stannic oxide, stannous chloride dihydrate, stannous orthophosphate, stannous sulfate, stannous sulfide, stannous oleate, stannous oxalate, or stannous tartrate at a concentration of 0, 0.03, 0.1, 0.3, or 1.0% for 4 weeks. Body weights and food intake were recorded weekly, and blood was sampled on day 27. At autopsy, the liver, kidneys, heart, and spleen were weighed, and samples were processed for histological examination. Stannic oxide, stannous sulfide, and stannous oleate had no deleterious effects, other than a statistically significant increase in haematocrit in male rats fed the highest concentration of stannous sulfide. The rats fed diets containing 0.3 or 1.0% of stannous chloride, orthophosphate, sulfate, oxalate, or tartrate experienced growth retardation with an apparent reduction of feed conversion efficiency. They also had haematological signs of anaemia. Animals at the 1.0% dietary concentration of these compounds had histological changes in the liver, comprising a clearly homogeneous parenchymal cell cytoplasm and slight hyperplasia of the bile ducts. Similar alterations, but of lesser severity and frequency, were found in the rats fed 0.3% of stannous chloride, oxalate, or orthophosphate (De Groot & Feron, 1970; De Groot et al., 1973a).

Stannous oxide and stannous chloride were fed to rats of the same strain, age, and group numbers as in the study summarized above for 13 weeks at a dietary concentration of 0, 0.03, 0.1, 0.3, or 1.0%. Stannous oxide had no toxic effects at any dose. Animals receiving 1% stannous chloride showed effects within the first 7 days, including gross abdominal distension, minimal to no growth, anorexia, and anaemia. By the eighth week, several animals had lost weight and a number had died. This group was terminated in the ninth week. At autopsy, several gross pathological lesions were found, including distension of the intestines, small oedematous pancreases, and greyish-brown livers. Histological evaluation revealed moderate testicular degeneration, severe pancreatic atrophy, spongy white matter in the brain, acute bronchopneumonia, enteritis, and distinct hepatic changes consisting of atypical homogeneous liver-cell cytoplasm and mild proliferation of bile-duct epithelium. In view of the marked reduction in appetite in this group, it is difficult to assess the degree to which tin was responsible for the observed pathological changes. Animals fed the 0.3% diet showed some abdominal distension and loss of appetite during the first 2 weeks, but both appetite and growth returned to normal after the second week. Significantly lower haemoglobin values were determined in animals of each sex between the fourth and ninth weeks. However, by the end of the study, only males at 0.3% stannous chloride had lower (p < 0.05) haemoglobin and haematocrit values. Terminal autopsy and histological evaluation of this group showed only minor treatment-related changes in some of animals of each sex, including atypically homogeneous hepatocyte cytoplasm and bile-duct epithelial proliferation. No treatment-related effects were seen in rats fed 0.03% or 0.1% stannous chloride. The leukocyte, total, and differential blood counts at autopsy were unaffected, as were the serum activities of alanine and aspartate aminotransferases. At termination, both male and female rats showed a dose-related decrease in the activity of serum alkaline phospatase, but this achieved statistical significance only at the 0.3% dietary concentration of stannous chloride (De Groot & Feron, 1970; De Groot et al., 1973a).

The authors indicated that the results of unpublished pair-feeding experiments suggested that the growth depression observed in some of the experiments cannot be explained by lack of palatability of the diet. In addition, on the basis of other unpublished studies, the authors suggested that the toxicity of tin compounds is inversely dependent on the iron content of the diet (De Groot et al., 1973a).

Groups of male and female rats were fed semi-purified diets containing iron at 35 or 250 ppm and tin as stannous chloride at a concentration of 0, 50, 150, 500, or 2000 ppm. Growth depression and reduced appetite and food efficiency were observed in animals at 500 or 2000 ppm of tin. Distinct signs of anaemia occurred in animals at 2000 ppm, but only transitory decreases in haemoglobin were seen in rats receiving 500 ppm. Pancreatic atrophy and histological changes in the liver, kidneys, spleen, testis, and heart were seen in some animals at the highest dose of tin. In all instances in which dietary tin was determined to have effects, the severity was more pronounced in animals receiving the low iron diets. The NOEL for tin was 150 ppm, equivalent to 7.5 mg/kg bw per day (De Groot et al., 1973b)

The effect of stannous chloride on a number of biochemical parameters was investigated in male weanling rats given oral doses of 0 (control), 0.3, 1.0, or 3 mg/kg bw at 12-h intervals for 90 days. A slight, non-significant decrease in the calcium content of the femoral epiphysis was found in rats at 0.3 mg/kg bw. At 1.0 mg/kg bw, significant reductions in succinate dehydrogenase activity in the liver and in the calcium content and alkaline phosphatase activity of the femoral epiphysis were observed. This dose also caused significant decreases in the relative weight and calcium concentration of the femur, in serum lactic dehydrogenase and alkaline phosphatase activities, in hepatic succinate dehydrogenase activity, and in the calcium content and alkaline phosphatase activity of the femoral diaphysis and epiphysis. The decrease in calcium content of the femoral epiphysis was considered to be the most specific biochemical manifestation of the toxic effect of tin. The NOEL was 0.3 mg/kg bw per day (Yamaguchi et al., 1980a,b).

2.2.3 Special studies

2.2.3.1 Interaction of tin with trace elements

Little is known about the biochemistry of the metabolism of inorganic tin in the body or the exact mechanisms by which this element affects physiological processes. It is known, however, that tin interacts with a number of trace elements, many of which have vital functions in the body. For example, during a lifetime study, approximately 500 rats of each sex were given a diet containing tin at a concentration of 0.28 ppm and ‘doubly deionized’ drinking-water supplemented with tin at 5 ppm as stannous chloride. Post-mortem analyses were conducted for a number of elements in the liver, lung, heart, kidneys, and spleen. The regimen significantly increased the concentrations of copper (p < 0.005) and zinc (p < 0.001) in the liver in comparison with controls (Schroeder & Nason, 1976).

At the considerably higher dietary concentration of 210 ppm of tin (as stannous chloride), the concentrations of copper and zinc in rat liver were significantly lower (p < 0.05) than those of control animals receiving tin at a concentration of 1 ppm in the diet at the end of the 21-day feeding period. The animals at 210 ppm also had higher (p < 0.05) concentrations of iron in their livers, while lower (p < 0.05) concentrations of zinc and higher (p < 0.05) concentrations of tin were found in their tibias and kidneys (Greger & Johnson, 1981).

The same authors examined the effects of varying the concentrations of both tin (< 1, 100, 200, 500, and 2000 ppm) and zinc (15, 30, and 52 ppm) on the metabolism of tin, copper, iron, calcium, and zinc in male Sprague-Dawley weanling rats over 3 weeks. The amount of tin that accumulated in the tibia and kidney was proportional to the dietary intake and was not affected by the dietary zinc concentration. The effects of tin on the copper concentration depended on the organ studied. Although there were few changes in iron metabolism, a significant reduction (45%) in the activity of blood delta-aminolaevulinate dehydratase was found in animals at the highest dietary concentration of tin (Johnson & Greger, 1984, 1985).

The effects of dietary tin on copper and zinc metabolism were studied in young growing Long-Evans rats fed diets containing tin at a concentration of 100, 330, or 1100 ppm for 4 weeks. The diet was copper-sufficient. At the lowest concentration of tin, the concentrations of copper in the duodenum, liver, kidney, and femur were significantly reduced, and the zinc concentrations in the kidney and femur were also reduced. Higher concentrations of tin reduced the feed consumption of the rats. In a separate experiment, the effect of tin in weanling rats fed a copper-deficient diet was found to be more pronounced at the lowest dietary concentration of tin (Rader et al., 1990; Rader, 1991)

‘Considerable disturbances’ in zinc and copper metabolism due to tin were reported in rats given seven subcutaneous injections of a stannous chloride solution providing a dose of tin of 2 mg/kg bw, administered every other day. A threefold increase in the concentration of zinc in the liver was observed in treated animals when compared with controls. While they were not significant, there were clear reductions in the zinc concentrations in kidney and heart tissues, and there were significant (p < 0.01) decreases in the zinc concentrations in the lungs, spleen, and brain. The treated animals also showed increased (p < 0.01) concentrations of copper in their livers but a significant (p < 0.01) decrease in those in blood and brain (Chmielnicka et al.,1981).

In addition to its effects on copper, zinc, and iron metabolism, tin has also been shown to interact with calcium. An inhibitory effect of tin on the intestinal absorption of calcium was demonstrated in rats treated orally with stannous chloride at a concentration providing a dose of tin of 30 mg/kg bw, every 12 h for 3 days. The tin content of the duodenal mucosa was a significantly (p < 0.01) increased in treated animals, from 4.7 ± 0.27 to 12 ± 0.80 mg/kg, while the calcium content decreased from 38 ± 9.5 to 14 ± 1.4 mg/kg. The calcium-binding activity and alkaline phosphatase activity of the mucosa were also depressed (p < 0.01) in treated animals. These results were interpreted as evidence of an inhibitory effect of tin on active transport of calcium in the duodenum (Yamaguchi et al., 1979).

In male rats weighing about 120 g that received single intraperitoneal injections of tin as stannous or stannic chloride at doses of 2.5–30 mg/kg bw, the calcium concentration in the kidneys increased in a dose-dependent manner, mainly in the renal cortex. This accumulation of calcium in the kidneys was associated with a significant, dose-related decrease in serum calcium concentrations. The valency form in which the tin was administered did not have a significant effect on the results (Yamamoto et al., 1976). When male rats were given tin at a dose of 30 mg/kg bw by gavage at 12-h intervals for 3 or 10 days, increased renal calcium concentrations were found, with concurrent reductions in serum calcium. The femoral calcium concentration also fell, while the pancreas calcium content was increased (Yamaguchi, 1980). Yamaguchi et al. (1977) proposed that the observed increase in renal calcium concentration in rats dosed with inorganic tin is due to increased synthesis of a calcium-binding protein in the renal cortex, which, according to Piotrowski & Szymanska (1976), is not metallothionein. Tin may also interact with other essential elements and with toxic metals such as lead. The latter effect was postulated by Vander et al. (1979) on the basis of the results of a study on lead transport by renal cortical and medullary tissues in vitro, which indicated that lead and tin have a common renal transport pathway. Confirmation of the competition between these two elements awaits further study.

2.2.3.2 Effect on bone strength

The effects of tin, as stannous chloride, on the mechanical strength of bone was investigated in groups of 22–30 male weanling rats given tin in their drinking-water at a concentration of 0 (controls), 50, 150, 300, or 600 ppm for 4 weeks. The commercial rat chow fed to all animals was found to contain tin at a concentration of 52 ppm. The compressive strength of the distal epiphysis of the femur was significantly decreased in animals at 300 and 600 ppm (Ogoshi et al., 1981).

2.3.1 Episodes of poisoning

Acute toxicity resulting from dissolution of tin from the internal surfaces of tinned cans or copper saucepans was first reported over 110 years ago. Salts of tin were detected in pears stewed in a newly-tinned saucepan, after the nine persons who had eaten similarly prepared pears on two or three occasions had experienced watery diarrhoea, ‘sickness’, and ‘great pain’ in the abdomen. Other instances of painful diarrhoea after consumption of tinned fruits were reported, and the author had detected tin in cans of apricots, pineapples, peaches, and tomatoes (Sedgwick, 1888).

A more detailed report from that period describes the cases of four men aged 18-42 who ate canned cherries and suffered nausea, abdominal pain, vomiting, and diarrhoea about 1.5 h later. Two of the patients became cyanotic, and one was admitted to hospital unconscious. Recovery occurred within 24 h, although at that time two of the patients had traces of albumin in their urine. Analysis of the acidic juice of the remaining cherries indicated the presence of 3400 mg/L of tin (Luff & Metcalfe, 1890).

Ingestion of canned apricots containing 800 ppm of tin by four members of one family resulted in nausea, vomiting and abdominal pain commencing within 30 min to 3 h (Savage, 1939).

An episode was reported in which 31 persons suffered from nausea, abdominal cramps, and/or vomiting within 1–2 h of drinking a fruit punch containing 2000 ppm of tin. The reconstituted pineapple–grapefruit juice used to make the punch had been delivered in a 5-gallon (22.5 L) retinned milk container. No information was available about when the juice was processed or how long and in what kind of container it had been stored before delivery. However, the lining of the container in which the juice was delivered showed obvious signs of corrosion, probably due to the unusually high acidity (pH 3) of the juice. The illness reportedly lasted from 2 to 48 h after the onset of symptoms (Warburton et al., 1962).

Consumption of a canned orange-based drink containing 420 ppm of tin produced a similar outbreak of poisoning, manifested by nausea, vomiting, diarrhoea, fever, and headache, in Japan. Most of the 1838 affected persons recovered within 1–2 days. The inside surface of the cans in which the juice had been stored was covered by a thin, black, rusty layer indicative of corrosion (Omori, 1966; Omori et al., 1973). Similarly, vomiting, diarrhoea, and other signs of distress were observed in 15 students who had consumed a canned orange beverage containing 100–490 ppm of tin and in eight other persons who had consumed tomato juice containing 160–220 ppm of tin (Horio et al., 1967). Eight other cases were reported in 1969 after ingestion of tomato juice containing 250 ppm of tin (Kojima, 1969). The tin content of random samples from the manufacturers of the juices ranged from 75 to about 500 ppm, but epidemiological investigations indicated that only batches with higher concentrations were responsible for the incidents (Ministry of Health and Welfare, Japan, 1969). Fifteen out of 26 persons consuming orange juice containing about 300 ppm tin showed gastric symptoms (Kojima, 1971).

Nausea, vomiting, and diarrhoea were observed in a large, unspecified number of persons in Kuwait who consumed formulated orange juice and apple juice containing 250–390 ppm of tin (Benoy et al., 1971).

In 97 well-documented cases, severe abdominal bloating, vomiting, diarrhoea, and headache were noted after the consumption of canned tomato juice containing tin at concentrations of 141–400 ppm. The mean concentrations in the various lots implicated as the cause of the intoxication were 240–360 ppm. The cans were visibly detinned, an effect that was attributed to the unusually high concentrations of nitrate on the tomatoes used to prepare the juice. Excessive nitrate fertilization of tomato plants was related to complete corrosion of the can lining within 6 months of storage, yielding tin concentrations of 380–480 ppm (mean, 420 ppm), whereas juice from uncorroded cans contained 29–81 ppm (mean, 50 ppm) (Barker & Runte, 1972).

2.3.2 Studies with volunteers

Nine male volunteers weighing 65–83 kg ingested tin at doses of 120–200 mg/day (equivalent to about 1.6–2.9 mg/kg bw per day) for 23 days without adverse effects. Almost all the ingested tin was recovered in the faeces (Calloway & McMullen, 1966). Of eight subjects who ingested a solution of 700 ppm of tin, only two had slight nausea and one had diarrhoea (Cheftel, 1967).

Five volunteers of each sex were given orange juice containing tin at concentrations of 0–1400 ppm on various occasions. The volunteers were unaware of the nature of the test substance. All experienced either nausea (three persons), diarrhoea (one person), or both (one person) when they first drank the juice containing 1400 ppm of tin, which provided a dose of tin of 4.4–6.7 mg/kg bw. Administration 1 month later resulted in only one case of nausea (Benoy et al., 1971).

The effect of tin, with or without iron, on the absorption of radiolabelled zinc was studied in 10 healthy volunteers aged 18–46 years. Consumption of 310 µmol (36 mg) of tin with either a test meal containing zinc chloride or turkey (tin:zinc ratio, 5:1) resulted in a significant decrease (28%) in the absorption of zinc. Partial replacement of the tin with iron (tin:iron:zinc ratio, 2.5:2.5:1) had no significant effect on this reduction (Valberg & Chamberlain, 1984).

A 40-day balance study was carried out in which eight men of a mean age of 24 ± 2 (SD) years and a mean weight of 76 ± 5 kg were given either a control diet containing 0.1 mg of tin daily or a test diet containing tin at a dose of 50 mg daily as stannous chloride, for a mean intake of 0.66 mg/kg bw per day in a cross-over design. The faecal excretion of selenium was statistically significantly increased by 30 % during administration of the test diet, with a corresponding decrease in the apparent retention of selenium during the test period (Greger et al., 1982).

3.1.1 Dietary sources of tin

3.1.1.1 Canned foods

Studies of tin in food and dietary intake of tin were reviewed by Underwood (1977), Fricke et al. (1979), and Greger (1988). Background information was also provided by Rader (1991) and Biego et al. (1999). There is consensus that the natural concentration of tin in plant and animal tissues is low and that the main dietary source of this mineral is foods that have been in contact with metallic tin from the tinplate of cans used to preserve them (Monier-Williams, 1949; Schroeder et al., 1964; Greger & Baier, 1981; Sherlock & Smart, 1984; Rader, 1991; Sumitani et al., 1993; Biego et al., 1996). Tinplate, which is steel coated with a thin layer of metallic tin, has been used to make cans since about 1820. The interior surface of tinplated cans may be coated with a lacquer (also called resin or enamel) to prevent adverse changes in food caused by reaction with the tin; however, use of lacquer coatings on cans is not always practicable or cost effective (Sherlock & Smart, 1984). Particularly high concentrations of tin may be found in canned foods due to the corrosion and leaching of tin from unlacquered or poorly lacquered cans (Monier-Williams, 1949; Schroeder et al., 1964; Fricke et al., 1979; Greger, 1988). More sophisticated canning techniques have decreased the amount of tin contamination over the years, but metal cans are still the main source of tin in the diet (Fricke et al., 1979).

3.1.1.2 Other sources

Little contamination of food comes from tin in air, water, or soil (Hislop et al., 1980). Organic tin-containing food additives, such as plasticizers and fungicides (Sherlock & Smart, 1984; Rader, 1991), and stannous chloride, which is used as a colouring or decolouring agent, preservative, and sequestrant (National Academy of Sciences, 1979; Sistrunk & Gascoigne, 1983), contribute negligible amounts of tin to the diet (Kumpulainen & Koivistoinen, 1977; Greger, 1988). The concentrations of tin residues from an organic tin insecticide used on apples and oranges were < 0.05–0.023 mg/kg of food (Sano et al., 1979).

Tin salts tend to be insoluble in water, and little of the tin that migrates into water supplies from organic tin compounds used in fungicides, insecticides, herbicides, and anthelmintics remains, although tin is present in trace amounts in natural water (National Academy of Sciences, 1977). Hadjimarkos (1967) reported a concentration of 0.006 mg/L in municipal drinking-water in the USA, and Sherlock & Smart (1984) reported that the concentration of tin in drinking-water was generally < 0.010 mg/L. If the daily consumption of water is assumed to be 2 L/day, the intake of tin from water would be 0.012–0.020 mg/day.

3.1.2 Factors affecting the concentration of tin in canned foods

The tin in canned food is likely to be in the form of inorganic tin salts rather than tin in covalently bound organometallic compounds (Sherlock & Smart, 1984; Marro, 1996; Vannoort et al., 2000). The tin content of canned foods varies according to whether the can is lacquered, the pH of the food in the can, the presence of plant pigments, the storage conditions (i.e. time and temperature) of the canned foods, whether the food is stored in opened cans, and the presence of oxygen, reducible organic compounds, and food additives. The reported concentration of tin in foods is also affected by the method of sample preparation and the analytical method used.

3.1.2.1 Lacquer

The tin content of canned foods is lower in cans with an inside lacquer coating (Radar, 1991). Infant foods, soft drinks, and alcoholic beverages are usually contained in fully lacquered cans and generally contain low concentrations of tin. Similarly, meats and meat products are often contained in lacquered cans and are not aggressive to tin (Sherlock & Smart, 1984).

In Finland, it was found that coating tin-plated steel reduced the amount of tin in canned foods by a factor of at least 50, as long as the coating remained intact. They found that the concentrations of tin in foods in cans that were partially or not lacquered were 9–700 mg/kg of food, with median and mean values of 100 and 150 mg/kg of food, respectively, whereas the concentrations in foods in lacquered cans were 0.005–30 mg/kg, with a median value of 0.08 mg/kg (de Goeij & Kroon, 1973).

In the USA, grapefruit juice, orange juice, tomato sauce, and pineapple in unlacquered or partially lacquered cans contained 51–150 mg/kg of food when the cans were first opened. In contrast, foods in lacquered cans contained only 0.1–2.8 mg/kg of food. The mean concentrations of tin were 1.5 mg/kg in foods in lacquered cans, 88 mg/kg in food in partially lacquered cans, and 89 mg/kg in food in unlacquered cans. Storage of the foods in open cans for 7 days increased the tin content of the foods in partially lacquered cans (mean, 360 mg/kg; range, 140–670 mg/kg) and unlacquered cans (540 mg/kg) but not of those in lacquered cans (mean, 1.9 mg/kg; range, 0.1–3.7 mg/kg) (Greger & Baier, 1981). Also in the USA, higher concentrations of tin were reported in canned fruits and vegetables in unlacquered (17 mg/kg; 6.3–32 mg/kg) than in lacquered cans (< 2.0–2.5 mg/kg), the concentration increasing in unlacquered cans with open storage to 22 mg/kg (range, 10–40 mg/kg) on day 2 and 63 mg/kg (19–150 mg/kg) on day 5. No change was seen in the tin content of foods in lacquered cans (Capar & Boyer, 1980).

In a study of the tin content of 17 canned foods in southwest England, the mean concentration and range were 31 (< 10–360) in lacquered, 81 (< 10–300) in partially lacquered, and 48 (< 10–120) in unlacquered canned foods. The higher mean value in partially lacquered than in lacquered cans was explained by the fact that 48% of the foods in partially lacquered cans were tomato-based products and that nitrate in tomatoes may accelerate the corrosion of tin (Meah et al., 1991).

In France, the concentrations of tin were 0.03 ± 0.03 mg/kg in fresh foods, 3.2 ± 2.3 mg/kg in foods stored in lacquered cans, and 77 ± 36 mg/kg in foods preserved in unlacquered cans. The average concentrations in foods in unlacquered tin cans were 24 times higher than those in lacquered cans and 2600 times higher than those in fresh foods. For example, tomatoes in unlacquered tin cans contained 84 ± 61 mg/kg tin, 14 times higher than those in lacquered cans (6.0 ± 3.9 mg/kg) and 1700 times higher than those in fresh tomatoes (0.05 ± 0.01 mg/kg) (Biego et al., 1999).

The tin content of canned US military rations was determined after storage for 20 months at two temperatures. Entrées in lacquered cans contained no detectable tin at 1 °C and only 2.3 mg/kg (range, 0–8 mg/kg) when stored at 37 °C. The entrées in unlacquered cans contained tin at 32 mg/kg (range, 15–48 mg/kg) at 1 °C and 190 mg/kg (range, 43–300 mg/kg) at 37 °C. Fruit in unlacquered cans contained tin at 34 mg/kg at 1 °C and 420 mg/kg at 37 °C (Calloway & McMullen, 1966).

3.1.2.2 pH of food

Foods with a low pH are more aggressive to tin coating than neutral or less acid foods. Acidic foods oxidize the uncoated sides or seals of cans (Fricke et al., 1979), and the dissolution of tin can be rapid in an acid environment (Davis et al., 1979). In an analysis of 15 foods in the New Zealand total diet study, the highest concentrations were found in canned acidic foods such as pineapple, peaches, and tomatoes; the highest value (210 mg/kg) was found in canned pineapple (Vannoort et al., 2000).

3.1.2.3 Plant pigments

Pears, pineapple, grapefruit juice, orange juice, applesauce, and tomato sauce are often contained in unlacquered cans and are weakly to moderately aggressive to tin (Sherlock & Smart, 1984; Greger,1988; Smart & Sherlock, 1989). Highly pigmented red fruits such as blackcurrants, raspberries, and cherries are highly aggressive to tin and are generally contained in fully lacquered cans because the anthocyanin pigments in the fruits accelerate the dissolution of tin and reduce the colour of the fruits (Sherlock & Smart, 1984; Smart & Sherlock, 1989). Nevertheless, even red fruits contained in lacquered cans may have high tin concentrations because they cause rapid tin dissolution at weak points or scratches in the lacquer (Sherlock & Smart, 1984).

3.1.2.4 Storage conditions (time and temperature)

Bigelow (1916) and Adam and Horner (1937) showed that the tin concentration of canned foods increased over periods of a few months to two years. Subsequent investigators also found that canned foods accumulate more tin when stored for several months (Dickinson & Raven, 1962; Calloway & McMullen, 1966; Woolfer & Manu-Tawiak, 1977; Nagy et al., 1980; Greger & Baier, 1981; Sherlock & Smart, 1984). An increasing rate of migration of tin and other metals from packaging materials to foods was found during long storage in warehouses at ambient temperatures (Arvanitoyannis, 1990a,b,c; Cichon, 1995). Biego et al. (1999) also noted that the tin content of canned foods was affected by the length and temperature of storage. The migration of tin was accelerated when the ambient temperature exceeded 40 °C (Calloway & McMullen, 1966; Nagy et al., 1980).

In the study of canned US military rations described above, the tin content of five types of fruit in unlacquered cans was 12-fold higher at 37 °C than at 1 °C (420 and 34 mg/kg, respectively). For seven types of mixed dishes in unlacquered cans, the tin content was sixfold higher at 37 °C than at 1 °C (190 and 32 mg/kg, respectively) (Calloway & McMullen, 1966).

3.1.2.5 Storage of opened cans

Tin can dissolve rapidly in the presence of oxygen (Davis et al., 1979), e.g. when foods are stored in cans after opening, even when refrigerated (Greger & Baier, 1981). With each removal of a portion of food from a can, a new tin–food–air interface is formed, which can result in the dissolution of greater quantities of tin in the food (Capar & Boyer, 1980). The tin content of foods stored in opened cans was reported to increase over periods of up to 7 days (Dickinson & Raven, 1962), and two to five times more tin was found in food in a can that had been open for 48 h (Iwamoto et al., 1970).

The tin content was measured in 11 fruits and vegetables stored in open cans for up to 5 days and six control foods packed in glass. The mean content of eight foods in unlacquered cans increased from 17 mg/kg on day 0 to 22 mg/kg on day 2 and 63 mg/kg on day 5. Thus, the tin content of foods in opened unlacquered cans increased by an average of about fourfold when stored in a refrigerator for 5 days. The tin concentration remained unchanged in three foods in lacquered cans and in the six control foods packed in glass (Capar & Boyer, 1980).

The tin content of five canned fruits and tomato sauce in partially lacquered and unlacquered cans increased by an average of four- and sixfold, respectively, after 7 days of storage in opened cans in a refrigerator. The mean concentrations of tin in four foods in partially lacquered cans increased from 88 mg/kg (range, 51–150 mg//kg) to 360 mg/kg (range, 140–670 mg/kg), and the tin content of one food in an unlacquered can increased from 89 to 540 mg/kg (Greger & Baier, 1981).

3.1.2.6 Presence of oxygen, reducible organic compounds, and food additives

Other factors that may increase the concentration of tin in canned foods are the presence of oxygen in the can headspace and of nitrate, reducible organic compounds, fertilizers, and pesticides in food. The presence of oxygen, nitrate, or reducible organic compounds can increase the rate of dissolution of tin into canned foods to result in unacceptably high tin concentrations within a few months (Sherlock & Smart, 1984), and some pesticides (Boudene, 1979) and nitrates (Davis et al., 1979, 1980) favour corrosion of the internal surface of tin cans. A high nitrate concentration and a low pH in the food and residual oxygen in the headspace of the can can also increase the rate of pitting and increase the migration of tin from unlacquered cans into foods (Davis et al., 1980; Sherlock & Smart, 1984). The ability of acidic foods such as tomatoes, apricots, and peaches to leach tin from cans when they contain high concentrations of nitrate is well recognized (Vanoort et al., 2000). Tomatoes and foods containing them sometimes have high concentrations of nitrate (Sherlock & Smart, 1984; Meah et al., 1991).

3.1.2.7 Sample preparation and analytical method

Schwarz et al. (1970) concluded that many of the reported values for tin in biological materials may be undersestimates because tin is readily lost during drying and ashing because of its volatility, particularly when it is present as organic tin, and because some organic tin compounds are soluble and extractable in lipids. Many of the early estimates of tin in biological materials are unreliable because of the loss of volatile tin compounds or formation of insoluble tin compounds during sample preparation or because of the insensitivity of the analytical methods (Rader, 1991).

3.2.1 Foods not in tin cans

Appendix 1 of this monograph lists the reported tin contents of various foods and food products when not contained in cans. The concentrations in vegetables, fruits and fruit juices, nuts, dairy products, meat, fish, poultry, eggs, beverages, and other foods are generally low, with a mean value < 2 mg/kg of food. Recent studies showed low tin contents in bread and pasta (< 0.003–0.03 mg/kg) (Biego et al., 1999; Ysart et al., 1999), while older studies gave average values for tin in grain products of 3.7–9.5 mg/kg (Zook et al., 1970).

3.2.2 Canned foods

The average tin concentration in foods packed in lacquered cans was 0–6.9 mg/kg of food (Appendix 2 of this monograph), and these foods probably contribute little tin to the diet. Food packed in unlacquered, partially lacquered, or unspecified cans (Appendix 3) have considerably higher concentrations of tin, with mean values ranging from < 1 to 1000 mg/kg. Some of the foods in unspecified cans with lower tin concentrations were probably contained in lacquered cans.

3.2.3 Tin concentrations ³ 250 mg/kg of food

At its thirty-third meeting, the Committee concluded that concentrations of tin as low as 150 mg/kg in canned beverages and 250 mg/kg in other canned foods could produce acute manifestations of gastric irritation in certain individuals (Annex 1, reference 83). None of the foods listed in Appendix 1 (foods not in contact with cans) or Appendix 2 (foods in lacquered cans) contained tin at concentrations > 250 mg/kg, but several foods listed in Appendix 3 of this monograph (foods in unlacquered, partially lacquered, and unspecified cans) had concentrations > 250 mg/kg.

In a study of the tin content of 17 canned foods in southwest England, the concentration in 16 of 856 cans (< 2%) exceeded 200 mg/kg, and that in three of these foods (canned asparagus, 360 mg/kg; tomato soup, 300 mg/kg; and spaghetti, 280 mg/kg) exceeded 250 mg/kg (Meah et al., 1991). After 7 days of open storage in a refrigerator, the tin content of three canned foods in the USA was 670 mg/kg in tomato sauce, 460 mg/kg in grapefruit sections, and 540 mg/kg in crushed pineapple (Greger & Baier, 1981).

The tin content of canned foods available on German markets was analysed between 1993 and 1998 by the laboratories of the enforcement authorities of the Lander. The data, representing over 2000 analyses, are summarized and presented in Appendix 3. The values for 19 of the 85 types of foods analysed exceeded 250 mg/kg (Table 2) and represented about 4% of the values for all foods analysed. The highest values were reported for canned pineapple and samples of champignon mushrooms. More samples of pineapple had values > 250 mg/kg than any of the other foods (Nobel, 2000).

From Nobel (2000)

Table 3 summarizes the estimated daily intakes of tin reported in seven countries (Australia, France, Japan, the Netherlands, New Zealand, the United Kingdom, and the USA). The main variable in these estimates was the amount of canned foods consumed.

3.3.1 Australia

In the Australian total diet study, the tin concentration was examined in only eight canned foods which represented the main source of tin in the diet (Marro, 1996). The estimated intake of tin in six age–sex groups ranged from 0.43 to 5.2 mg/day. The intake of tin at the 95th percentile of consumption was 7.4 mg/day for men and 5.8 mg/day for women (Baines, 2000). Tomatoes contributed 46%, pineapple and juices other than orange juice contributed 46%, and fruit salad contributed 7.8% to the total dietary exposure to tin. In a submission from the Australia–New Zealand Food Authority (Baines, 2000), dietary exposure to tin was calculated by several methods. The DIAMOND (dietary modelling of nutrition data) method provided a mean intake of 1.9–2.4 mg/day for the whole population aged 2–70 years, while that of consumers at the 95th percentile was 8.0–10 mg/day. The main foods that contributed to total dietary exposure in the DIAMOND model were tomato (54%), pineapple (36%), and mushrooms (10%).

The mean intakes of tin in surveys of persons aged 2 years and older were 2.2–2.7 mg/day for all respondents and for consumers only (Baines, 2000). The mean intakes of these two groups were 0.7–1.0 mg/day and 0.7–1.2 mg/day, respectively, while that of consumers at the 95th percentile was 8.0–12 mg/day. The mean maximum intake of Australians aged 2 years and older was 31 mg/day for all respondents and for consumers only, and the median intake for these two groups was 24 mg/day. The 95th percentile maximum intake for consumers only was 83 mg/day.

3.3.2 France

The intake of tin was assessed from the daily average consumption of foods by French adults compiled from several government and industry organizations (Biego et al., 1999). Foods were grouped into nine categories with a separate group for canned foods (Table 4). The daily intake of tin was based on the weight of each food group consumed and the concentration of tin in that food. The estimated intake was 2.7 mg/day or 0.04 mg/kg bw per day. The density of tin in this diet was 1.6 mg/kg of food. The intake of tin depended on the consumption of food stored in tin cans, which represented only 5.6% of the total daily consumption of French adults but 98% of the tin intake. The authors postulated that the dietary intake of tin would be 0.05 mg/day from a diet consisting only of fresh foods, 0.34 mg/day from a diet with usual canned foods in lacquered cans, and 7.4 mg/day from a diet with usual canned foods in unlacquered cans.

From Biego et al. (1999)

3.3.3 Japan

The daily dietary intakes of 28 trace elements by adult Japanese women were estimated from food composition databases. Duplicate samples of all foods were collected from 456 women in 22 cities and villages in 14 prefectures in Japan during the winters of 1990–95. The food items in each duplicate diet were separated, and the weight of each food item was recorded. The average tin intake was 0.64 mg/day with a coefficient of variation of 39% (Shimbo et al., 1996). As the databases did not include all the foods consumed by the women, the intake of trace elements was probably underestimated.

3.3.4 The Netherlands

The Dutch total diet study is based on the consumption by 18-year-old men of 221 food items within 23 food commodity groups which were purchased every 3 months and analysed. The average tin intakes were 1.7 mg/day in 1976–78 and 0.65 mg/day (range, 0–1.8 mg/day) during 1984–86. The food groups that contributed the most tin to the daily diet were canned fruits (82%), milk and dairy products (9%), and vegetables (4%) (van Dokkum et al., 1989).

3.3.5 New Zealand

The intake of tin by young men eating 3.5 kg of food per day (4000 kcal) was estimated in the 1994–95 total diet study to be 15 mg/day. When this figure was adjusted downwards to a food intake of 2.5 kg/day, the intake was 11 mg. The foods were analysed in eight composite groups, one of which consisted solely of canned foods (fruit, fruit juice, jams, baked beans, spaghetti, fish, soft drinks, and beer). The intake of canned foods was 460 g/day, which represented 14% of the weight of the total diet. The investigators stated that most of the tin came from the canned food composite, but the concentrations of tin in the eight composites were not presented (Dick et al., 1978).

In the 1982 total diet study, daily intake of tin was based on three levels of caloric intake and was 6.0 mg for 8.4 MJ (34–50-year-old women), 8.1 mg for 11 MJ (23–50-year-old men), and 12 mg for 17 MJ (extreme intake). The foods were analysed in nine composite groups, and canned foods were included with fruit (10% of the total diet) and vegetables (17% of the total diet). Of the 14 foods in the fruit composite, six (tomatoes, peaches, apricots, pineapple, fruit salad, and apple juice) were canned. Of the 11 foods in the vegetable composite, four (tomato soup, vegetable soup, baked beans, and creamed corn) were canned. The authors stated that the consumption of canned foods had declined from 14% of the diet in 1975 to 8.4% in 1982. The contribution of the nine commodity groups to the total intake of tin at the lower calorie intake (8.4 MJ/day) was 26% from fruit, 20% from vegetables, 19% from beverages, 12% from sweet foods and nuts, 8.0% from cereal-based foods, 7.5% from meat, fish, and eggs, 6.7% from dairy products, < 1% from fats and oils,and < 1% from instant foods (Pickston et al., 1985).

Of the 105 foods analysed for tin in the 1987–88 total diet study, all contained concentrations < 1 mg/kg except for nine canned foods (spaghetti, baked beans, tomato sauce, tomato soup, tomatoes in juice, beets, pineapple, peaches, and fruit salad), which had average concentrations ­ 96 mg/kg. The concentration of tin in most foods in the 1990–91 total diet study was < 1 mg/kg, except for canned foods, which had average concentrations ­ 86 mg/kg.

Only selected foods in the 1997–98 total diet study were analysed for tin, as previous experience had indicated that tin would be undetectable in the other foods (Vannoort et al., 2000). Fifteen of the 114 foods in the Study were analysed, comprising six canned vegetables, three canned fruits, canned spaghetti in sauce, canned salmon, chicken soup (assumed to be canned), honey, marmalade, and tea. A comparison of the tin intake of adult males indicated a decrease from 9.9 mg/day in 1982 to 6.7 mg/day in 1997–98 (Table 3). The estimated dietary intake of tin by six age–sex groups in the 1997–98 total diet study ranged from 2.9 mg/day for 1–3-year-old children to 7.5 mg/day for female vegetarians (Table 3). Canned spaghetti, baked beans, apricots, tomatoes, and peaches contributed 77% of the dietary exposure of men aged 19–24 years and of children aged 1–3 years.

3.3.6 United Kingdom

3.3.6.1 Adults

In 1972, the mean daily intake of tin in the United Kingdom was estimated to be 0.19 mg/day (Hamilton et al., 1972; Hamilton & Minski, 1972–73). Comprehensive data on tin intake is provided by the Total Diet Studies in 1976–94 and 1997, which involved the purchase and preparation of common foods in the national diet. The relative proportion of each food category within a group reflects its importance in the average household diet. Intakes were calculated from the weight of each food group consumed and the concentration of tin in each group. Table 3 shows the mean daily intakes of tin between 1976 (4.4 mg/day) and 1994 (2.4 mg/day) and in 1997 (1.8 mg/day). The upper range of intake in 1994 was estimated to be 7.9 mg/day. The fall in tin intake between 1976 and 1981 may have been due to an increase in the proportion of food cans that were lacquered (Ministry of Agriculture, Fisheries and Food, 1985). The increase in the estimated total dietary exposure observed in the 1991 total diet study was due to an elevated tin concentration in the canned vegetables group (Ysart et al., 1999).

Table 5 shows the concentrations of tin in nine food groups in the 1981 total diet study and the contributions of these food groups to daily tin intake (Ministry of Agriculture, Fisheries and Food, 1985). The tin density of this diet was 1.6 mg/kg of food, and vegetables, fruit, and sugars provided 78% of the intake (2.4 mg/day).

From Ministry of Agriculture, Fisheries and Food (1985)

Table 6 shows that most of the food groups included in the 1982 total diet study contained tin at concentrations < 1 mg/kg and often below the limit of determination of the analytical method. Groups that included some foods in unlacquered cans, especially vegetables and fruit products, contained tin at concentrations > 1 mg/kg. Canned vegetables and fruit contained tin at concentrations of 10–100 mg/kg. Nearly all the tin in the diet in this survey was associated with canned foods, and canned tomatoes, tomato products, pineapple, pears, and similar fruits contained the highest concentrations (Sherlock & Smart, 1984). The tin density of this diet was 2.2 mg/kg food. Canned vegetables (72 g) and fruit products (1400 g) accounted for only 5% (w/w) of the total diet but contributed 82% of the dietary intake of tin.

From Sherlock & Smart (1984)

The 1994 total diet study in the United Kingdom covered 119 categories of foods combined into 20 groups for analysis. Table 7 shows the concentrations of tin in these groups. The highest concentrations were found in canned vegetables (44 mg/kg) and fruit products (17 mg/kg). The concentrations in the other 18 groups were all < 1 mg/kg, and 15 food groups contained ­ 0.03 mg/kg of food. Canned vegetables contributed 66% and fruit products 31% of the estimated total average intake of tin of 2.4 mg/day (Ysart et al., 1999). In the 1997 total diet study, canned vegetables contributed 76% and canned fruits 18% to the average tin intake of 1.8 mg/day (Ministry of Agriculture, Fisheries and Food, 2000).

Table 7. Contributions of various food groups to the intake of tin in the United Kingdom, 1994 Total Diet Study

In a study of 31 women exposed to lead from drinking-water, replicates of whole 1-week diets were collected and the diets were analysed for several metals, including tin. The mean intake was 2.8 mg/day. As canned food was not specifically recorded in the diaries, the source of the tin could not be determined. The distribution of tin intake among the 31 women was 0–1.4 mg/day for 10 women, 1.4–2.8 mg/day for six women, 2.9–5.7 mg/day for 12 women, and 5.7–8.6 mg/day for three women (Sherlock et al., 1983).

Duplicate diets of 35 female vegetarians revealed a mean tin intake of 3.8 mg/day, which was higher than that of the general population reported in the Total Diet Studies. The range of intake by the vegetarians was 0.03–16 mg/day, and concentrations of tin in the duplicate diets were 0.1–46 mg/kg food (Ministry of Agriculture, Fisheries and Food, 2000).

3.3.6.2 Children

A tin intake of 11 mg/day was reported for a 5-week-old infant fed evaporated milk from unlacquered cans (Hamilton et al., 1972).

The dietary intakes of lead and other metals were determined for 2-year-old children in the cities of Birmingham and Harrow and for a control group, on the basis of 7-day duplicate diets. Mothers were asked to collect and weigh duplicate samples of all food and drink, including water, consumed by their children during the week. They were also requested to specify consumption of canned food. The tin density of the duplicate diets of 97 children in Birmingham was 1.8 ±1.8 mg/kg of food, and the average daily intake of tin was 1.7 ± 2.0 mg/kg (range, 0.04–11 mg/day). There was a significant correlation between the amount of canned food consumed and the concentration of tin in the diet (correlation coefficient, 0.40; p < 0.001). Canned foods accounted for 11% of the total weight of food consumed. The PTWI for tin (14 mg/kg bw: Annex 1, reference 83) was exceeded by only one child, who had a tin intake of 75 mg/week (11 mg/day). The tin density of the 64 control duplicate diets was 3.7 mg/kg of food, and that of the 38 duplicate diets in Harrow was 3.9 mg/kg of food, and the daily intake of tin was 2.9 mg/day for the control diets and 1.7 mg/day for those in Harrow (Smart et al., 1988).

3.3.7 USA

Tin was not measured in the total diet study in the USA, and the estimates in Table 3 for the USA are from several small studies. The lowest intakes of 0.11 and 1 mg/day are from diets with no canned foods.

Control and test diets were developed in order to study the effects of dietary tin on the metabolism of tin and calcium in eight men. The control diet, which consisted of typical fresh, frozen, or bottled foods in quantities adequate for adult males (3150–3700 kcal/day), contained tin at a concentration of only 0.11 mg/day, while the test diet contained an average of 50 mg of tin from stannous chloride dissolved in juices served at lunch and dinner (Johnson & Greger, 1982). A diet composed largely of fresh meats, cereals, and vegetables, which usually contain tin at < 1 mg/kg of food, was calculated to supply about 1 mg/day, whereas a diet that included substantial amounts of canned vegetables, fruit juices, and fish could supply as much as 38 mg/day (Schroeder et al., 1964).

The dietary intake and excretion of 22 elements, including tin, was measured for two young men over 50 weeks. One of the men was 23 years old, 187 cm tall, and weighed 100 kg, and the other was 25 years old, 178 cm tall, and weighed 71 kg. Duplicate, self-chosen diets, including water and beverages, were collected and analysed between September 1966 and August 1967. The average daily intake of tin was 5.8 ± 0.7 and 8.8 ± 1.1 mg/day for the two subjects, respectively. Daily varia-tions in tin intake were attributed to tin from cans and utensils (Tipton et al., 1969).

The tin content of three types of military diet were 9.5 mg/day in one composed of fresh foods, 26 mg/day in one of canned rations stored at 1 °C, and 160 mg/day in one of canned rations stored at 37 °C (Calloway & McMullen, 1966).

Individuals who eat four 100-g servings of foods from unlacquered cans could ingest 15–60 mg of tin daily, i.e. each serving could contain 4–15 mg tin (Johnson & Greger, 1982), whereas individuals who eat four 100-g servings of foods from unlacquered cans stored open in the refrigerator could ingest 200 mg of tin daily, i.e. each serving could contain 50 mg of tin (Greger & Baier, 1981).

3.3.8 Comparison with the PTWI

Table 8 provides a summary of the intakes of tin in seven countries, expressed as mg/day, mg/kg bw per day, mg/kg bw per week, and percent PTWI. The estimated intakes ranged from 0 to 12% of the PTWI, except for the maximum Australian intakes, which were 20 and 25% of the PTWI (70% of the PWTI at the 95th percentile of consumption). It should be noted that the PTWI applies to the long-term intake of tin (Annex 1, reference 83).

Estimates of dietary intake of tin were not available from other countries, and it was considered inappropriate to estimate intake on the basis of WHO regional diets because the main variable was the percent of ingested food in canned form and not the type of diet.

3.4.1 Tin density of diets

Table 9 presents the tin density of diets, as calculated from the available information. Tin densities depend on the amount of canned food in the diet and, in particular, the amounts of foods in unlacquered cans. The tin density of the diet is affected by whether the intake of water is included in the weight of the total diet. The densities ranged from 1.2 in the 1994 total diet study in the United Kingdom to 4.4 in the 1974/75 total diet study in New Zealand.

3.4.2 Assumptions about consumption of canned foods

Estimates of the dietary intake of a population group to tin were based on information or assumptions about the intake of canned foods and, in particular, the proportions of the canned foods in lacquered and unlacquered cans. Additional useful information, which was difficult to obtain, is the proportions of acid foods (orange juice, grapefruit, tomatoes) in lacquered and unlacquered cans and storage of canned food in open cans in homes, institutional kitchens, restaurants, and other eating establishments. Unfortunately, the information collected in national food consumption surveys does not contain this level of detail. Survey participants may not know or remember if the food they consumed was canned, fresh, or frozen, and the surveys do not specifically request this information. Furthermore, national food consumption surveys usually do not include individuals living in institutions, who may eat more canned foods.

Another important consideration with regard to the dietary intake of tin is that many foods formerly consumed as "canned" are now also (or primarily) available in bottles. This is true of foods such as tomato sauce, spaghetti sauce, pizza sauce, various soups, many fruit juices and juice drinks, and main dish items such as ravioli, spaghetti, macaroni, and cheese. In addition many foods formerly available in cans are now available in packaging for microwave ovens. These products may be referred to as ‘canned’ because of the processing technique, but they have had little contact with metal and are stored in materials that are not important sources of tin.

Table 10 shows the percentages of various diets estimated to be composed of canned foods. Similar information may be available for other countries but was not identified for this review. Information on the production of canned food may also be available, but a direct link cannot be made between production and human consumption because of the variables of imports/exports, food waste, food fed to pets, and food not intended for human use.

Vegetables and fruits accounted for 78% of the intake of tin in the 1981 total diet study in the United Kingdom (Table 5). As the data for this study were not separated by canned products, the total contribution of canned foods to tin intake could not be determined. Obviously, not all the vegetables and fruits in the diet were canned, but several products in other food groups probably were . Canned vegetables and fruit products accounted for 82% of the tin intake in the 1982 total diet study in the United Kingdom (Table 6), and canned foods accounted for 98% of the tin intake in the estimate for the French diet (Table 4)

In the USA, the National Food Processors’ Association, the Grocers Manufacturing Association, the Canned Manufacturers Institute, the International Life Science Institute, and the Department of Agriculture reported that no specific information on individual intake of canned food was available. The Department of Agriculture collects some information on food packaging in their yearly Continuing Survey of the Food Intakes of Individuals (Perloff, 2000), which determines whether fruits and vegetables consumed by survey participants are fresh, canned, or frozen so that appropriate food codes may be attached. Unfortunately, the term ‘canned’ includes foods in bottles and all types of cans, including those made of aluminium . Some foods are coded as ‘cooked or canned’ or ‘canned or powdered’. In addition, mixed dishes that may be prepared from canned ingredients (e.g. canned tomatoes) are not covered. To date, no one has used these food descriptors to estimate the consumption of ‘canned’ foods in the USA.

Information on the annual per capita consumption (or availability) of various foods in the USA was available from the Economic Research Service (undated) website of the Department of Agriculture. This report included information on the consumption of canned fish, milk, fruits, and vegetables. Although the data were provided in pounds consumed per year, the proportion of canned foods in the total per capita consumption could be estimated. Canned foods accounted for 140 pounds (64 kg) of the total per capita consumption of 2200 pounds (1000 kg) of food, so that average per capita consumption of canned foods may represent about 6.4% of total per capita food consumption. This figure may be an overestimate, because the water weight for coffee and tea and the weight of drinking-water were not included in the total weight of food consumed.

3.4.3 Consumers at risk

3.4.3.1 Consumers of food with a high tin content

The risk of buying and consuming a canned food contaminated with tin is random. The risk would be greater for those who purchase canned acid fruit or juice, tomatoes, or tomato products and who consume canned foods from other sources, such as in other peoples’ homes, in restaurants, and from food vendors.

3.4.3.2 Consumers with a high intake of canned foods

Although most people probably consume a limited amount of canned foods, there may be exceptions. People with low incomes or living in institutions such as nursing homes, boarding schools, or prisons may select or be served canned fruits, vegetables, and juices because of economy and ease of storage. Canned foods are often less expensive than fresh or frozen foods, and they require no refrigerator or freezer storage space. Some elderly individuals are accustomed to, or even prefer, the texture of canned fruits and vegetables (Greger, 1988). As mentioned earlier, national surveys on food consumption do not usually include individuals living in institutions.

Meah et al. (1991) cited an unpublished study from the Ministry of Agriculture, Fisheries and Food on the consumption of canned foods by 5–10-year-old boys in Glasgow, Scotland, in which heavy consumption of canned foods was found, especially by boys from the lowest social class. In this population, the average consumption of canned foods was 4 g/kg bw per day, and 7% of the boys from all social classes ate more than 8 g/kg bw per day of canned foods, which is equal to the PTWI if the concentration of tin is 250 mg/kg of food. Consumers may store canned fruit juice, juice drinks, and other canned foods in opened cans in the refrigerator and thereby increase their intake of tin. Individuals who routinely consume canned fruits, vegetables, and juices from unlacquered cans could ingest 50–60 mg of tin daily, assuming about four servings per day (Johnson & Greger, 1982; Sherlock & Smart, 1984). Those who consume about four servings of foods stored in open cans could consume about 200 mg/day of tin (Greger & Baier, 1981).

The considerations and assumptions used in estimating tin intake include determination of:

• the consumption of ‘canned’ foods by population subgroups, with a focus on high-risk groups such as the poor, institutionalized persons, and the elderly, who may consume more canned foods than the rest of the population;
• a correction for ‘canned’ foods in bottles or jars if these are included in the estimate for ‘canned’ food intake, as bottles or jars do not contribute to tin intake;
• a correction for foods and beverages in aluminium cans (e.g. carbonated soft drinks) if these are included in the estimate for canned food intake, as aluminium cans do not contribute to tin intake;
• the proportion of canned foods in lacquered and in partially lacquered or unlacquered cans, because the tin content of foods in partially lacquered and unlacquered cans is higher;
• the type of food, because acid foods (e.g. fruits, tomatoes, and tomato products) and nitrate in tomatoes tend to increase leaching of tin from a container into the food;
• the practice of open storage of canned foods in a refrigerator, because open storage increases exposure to oxygen and leaching of tin from the can into the food. The practice of open storage could be assumed for large cans of fruit juice or juice drinks.

The estimates of dietary intake provided here do not include detailed information on all the factors listed above. The available information allows estimates of mean intake that are, however, considerably below the PTWI for tin established by the Committee.

Inorganic tin is found in the +2 and +4 oxidation states; it may occur in a cationic form (stannous and stannic compounds) or as inorganic anions (stannites or stannates). Studies in rats provided evidence that the chemical form of inorganic tin is important in determining its toxicity. Inorganic tin compounds generally have little systemic toxicity in animals because of limited absorption from the gastrointestinal tract, low accumulation in tissues, and rapid excretion, primarily in the faeces. Insoluble tin compounds, such as stannous sulfide, had minimal toxic effects in rats when administered for 28 days in the diet at concentrations similar to those at which the soluble tin salts are clearly toxic. In short-term studies in rats, histological changes to the gastrointestinal tract, kidneys, liver, and adrenal cortex were observed. Alterations in haematological parameters indicative of anaemia have also been recorded. The acute toxicity of tin results from irritation of the mucosa of the gastrointestinal tract. Vomiting and diarrhoea were reported in cats given soluble salts of tin, but there was no clear dose–response relationship, and the vehicle in which the tin was administered may have affected its toxicity.

Episodes of human poisoning resulting from consumption of tin-contaminated foods and drinks have resulted in abdominal distension and pain, vomiting, diarrhoea, and headache. These symptoms commonly start within 0.5–3 h, and recovery occurs within 48 h. The doses of tin ingested in such episodes of poisoning were not estimated. In one study, five volunteers experienced symptoms when they ingested juice containing 1400 mg/kg of tin (corresponding to a dose of 4.4–6.7 mg/kg bw). Administration of the same dose to these individuals 1 month later resulted in symptoms in only one person.

The major dietary source of tin is the tinplate of unlacquered or partially lacquered cans used for the preservation of foods. The migration of tin from tinplate into foods is greater (i) into highly acidic foods such as pineapples and tomatoes, (ii) with increased time and temperature of food storage, and (iii) into foods, such as fruit juice, in opened cans. The tin content of canned foods is variable, and some foods may have concentrations high enough to cause an acute toxic reaction. The mean dietary intakes of tin by individuals reported from seven countries ranged from < 1 mg/day to about 14 mg/day and were considerably lower than the PTWI established previously by the Committee. Population groups with higher intakes of canned foods may have higher intakes of tin.

The Committee concluded that insufficient data were available to establish an acute reference dose for tin. It noted that the gastric irritation that may occur after ingestion of a foodstuff containing tin may depend on the concentration and chemical form of the tin. It reiterated its opinion, expressed at its thirty-third meeting, that the limited human data available indicate that concentrations of 150 mg/kg in canned beverages or 250 mg/kg in other canned foods may produce acute manifestations of gastric irritation in certain individuals. In addition, the Committee reiterated its advice, given at its thirty-third meeting, that consumers should not store food in open tin-coated cans. It welcomed the information that estimates of the intake of tin by the populations of several countries do not exceed the PTWI of 14 mg/kg bw.

Adam, W.B. & Horner, G. (1937) The tin content of English canned fruits and vegetables. J. Soc. Chem. Ind., 56, 329t–334t.

Arvanitoyannis, I. (1990a) The effect of storage of canned juices on content of the metals Fe, Cu, Zn, Pb, Sn, Al, Cd, Sb, and Ni. Nahrung, 34, 141–145.

Arvanitoyannis, I. (1990b) The effect of storage of canned meat on concentration of the metals Fe, Cu, Zn, Pb, Sn, Al, Cd, and Ni. Nahrung, 34, 147–151.

Arvanitoyannis, I. (1990c) The effect of storage of canned vegetables on concentration of metals Fe, Cu, Zn, Pb, Sn, Al, Cd, and Ni. Nahrung, 34, 247–253.

Baines, J. (2000) Submission to WHO from the Australia New Zealand Food Authority regarding dietary intake of tin.

Barker, W.H. & Runte, V. (1972) Tomato-juice-associated gastroenteritis, Washington and Oregon, 1969. Am. J. Epidemiol., 96, 219–226.

Barnes, J.M. & Stoner, H.B. (1959) The toxicology of tin compounds. Pharmacol. Rev., 11, 211–231.

Benoy, C.J., Hooper, P.A. & Schneider, R. (1971). The toxicity of tin in canned fruit juices and solid foods. Food Cosmet. Toxicol., 9, 645–656.

Biego, G.H., Biaudet, H., Joyeux, M. & Debry, G. (1996) Determination of tin content in fruits, vegetables and their juices according to packaging. Sci. Aliment., 16, 623–630.

Biego, G.H., Joyeux, M. & Hartemann, P. (1999) Determination of dietary tin intake in an adult French citizen. Arch. Environ. Contam. Toxicol., 36, 227–232.

Bigelow, W.D. (1916) Tin in canned foods. J. Ind. Eng. Chem., 8, 813–815.

Boudene, C.L. (1979) [Recent data on the toxicity of metallic contaminants of preserved foods due to their packaging.] Med. Nutr., 15, 425–430 (in French).

Brown, R.A., Nazario, C.M., de Tirado, R.S., Castrillon, J. & Agard, E.T. (1977) A comparison of the half-life of inorganic and organic tin in the mouse. Environ. Res., 13, 56–61.

Browning, E. (1969) Toxicity of Industrial Metals, 2nd Ed., London: Butterworths.

Calloway, D.H. & McMullen, J.J. (1966) Fecal excretion of iron and tin by men fed stored canned foods, Am. J. Clin. Nutr., 18, 1–6.

Calvery, H.O. (1942) Trace elements in foods. Food Res., 7, 313–331.

Capar, S.G. & Boyer, K.W. (1980) Multielement analysis of foods stored in their opened cans. J. Food Saf., 2, 105–118.

Cheftel, H. (1967) Tin in food. In: Joint FAO/WHO Food Standards Program; Fourth Meeting of the Codex Committee on Food Additives, September 1967 (ALINORM 67/12A). Joint FAO/WHO Food Standards Branch (Codex Alimentarius), Rome: FAO.

Chmielnicka, J., Szymenska, J.A. & Sniec, J. (1981). Distribution of tin in the rat and disturbances in the metabolism of zinc and copper due to repeated exposure to SnCl₂, Arch. Toxicol., 47, 263–268.

Cichon, Z. (1995) Study on the migration phenomena of selected metals in canned fruits during long storage. Nahrung, 39, 219–227.

Codex Alimentarius Commission (1999) Report of the Thirty-first Session of the Codex Committee on Food Additives and Contaminants, The Hague, 22–26 March 1999 (unpublished FAO document ALINORM 99/12A), Rome: Food and Agriculture Organization of the United Nations.

Conine, D.L., Yum, M., Martz, R.C., Stookey, G.K., Muhler, J.C. & Forney, R.B. (1975) Toxicity of sodium pentafluorostannite, a new anticariogenic agent. I. Comparison of the acute toxicity of sodium pentafluorostannite, sodium fluoride and stannous chloride in mice and/or rats, Toxicol. Appl. Pharmacol., 33, 21–26.

Dabeka, R.W. & McKenzie, A.D. (1981) Atomic absorption spectrometric determination of tin in canned foods using nitric acid–hydrochloric acid digestion and nitrous oxide–acetylene flame. J. Assoc. Off. Anal. Chem., 64, 1297–1300.

Davis, D.R., Cockrell, C.W. & Wiese, K.F. (1979) Pitting in canned green beans: Effect of cultural practices, tin coating, vacuum, corrosion accelerators, and storage conditions. J. Food Sci., 44, 241–245.

Davis, D.R., Cockrell, C.W. & Wiese, K.F. (1980) Can pitting in green beans: Relation to vacuum, pH, nitrate, phosphate, copper, and iron content. J. Food Sci., 45, 1411–1415.

Debost, M. & Cheftel, J.C. (1979) Tin binding in canned green beans. J. Agric. Food Chem., 27, 1311–1315.

De Groot, A.P. & Feron, V.J. (1970) Unpublished report No. R 3187. Submitted to WHO by the Centraal Instituut voor Voedingsonderzoek, Zeist, Netherlands.

De Groot, A.P., Feron, V.J. & Til, H.P. (1973a) Short-term toxicity studies on some salts and oxides of tin in rats, Food Cosmet. Toxicol., 11, 19–30.

De Groot, A.P., Til, H.P. & Willems, M.I. (1973b) Short-term (90 day) toxicity of tin in rats on semi-purified diets with excessive or marginal iron content. Unpublished report (No. R 4264) submitted to WHO by the Centraal Instituut voor Voedingsonderzoek, Zeist, Netherlands.

Department of Health & Human Services (1992) Toxicological Profile for Tin, Washington, DC, Department of Commerce.

Dick, G.L., Hughes, J.T., Mitchell, J.W. & Davidson, F. (1978) Survey of trace elements and pesticide residues in the New Zealand diet. I. Trace element content. N. Z. J. Sci., 21, 57–69.

Dickinson, D. & Raven, T.W. (1962) Tin content of canned fruits and vegetables. Food Manuf., 480–484.

van Dokkum, W., de Vos, P.H., Muys, T. & Wesstra, J.A. (1989) Minerals and trace elements in total diets in The Netherlands. Br. J. Nutr., 61, 7–15.

Durrant, S.F. & Ward, N.I. (1989) Multi-elemental analysis of human milk by inductively coupled plasma-source mass spectrometry: Implications for infant health. J. Microsc. Anal., 5, 111–126.

Economic Research Service (undated) Food Consumption, Prices, and Expenditures, 1970–97. Online at www.ers.usda.gov/publications/food/food consumption, prices and expenditures, Washington DC: Department of Agriculture.

Elkins, E.R. & Sulek, A. (1979) Atomic absorption determination of tin in foods: Collaborative study. J. Assoc. Off. Anal. Chem., 62, 1050–1053.

Engberg, A. (1973) A comparison of a spectrophotometric (quercetin) method and an atomic-absorption method for the determination of tin in food. Analyst, 98, 137–145.

Flinn, F.B. & Inouye, J.M. (1928) Metals in our foods, J. Am. Med. Assoc., 90, 1010–1013.

Fricke, F.L., Robbins, W.B. & Caruso, J.A. (1979) Trace element analysis of food and beverages by atomic absorption spectrometry. Prog. Anal. At. Spectrosc., 2, 185–286.

Friel, J.K., Andrews, W.L., Jackson, S.E., Longerich, H.P., Mercer, C., McDonald, A., Dawson, B. & Sutradhar, B. (1999) Elemental composition of human milk from mothers of premature and full-term infants during the first 3 months of lactation. Biol. Trace Elem. Res., 67, 225–247.

Fritsch, P., de Saint Blanquat, G. & Derache, R. (1976) Effect of various dietary components on absorption and tissue distribution of orally administered inorganic tin in rats. Food Cosmet. Toxicol, 15, 147–149.

Fritsch, P., de Saint Blanquat, G. & Derache, R. (1977) [Nutritional and toxicological study in rats of a food contaminant: tin.] Toxicology, 8, 165 (in French).

de Goeij, J.J.M. & Kroon, J.J. (1973) In: IAEA/FAO/WHO Symposium on Nuclear Techniques in Comparative Studies of Food and Environmental Contamination, Vienna: IAEA.

Goss, B.C. (1917a) Absorption of tin by proteins and its relation to the solution of tin by canned foods. J. Ind. Eng. Chem., 9, 144–148.

Goss, B.C. (1917b) Inhibition of digestion of proteins by absorbed tin. J. Biol. Chem., 30, 53–60.

Greger, J.L. (1988) Tin and aluminum. In: Smith, K.T., ed., Trace Minerals in Foods, New York: Marcel Dekker.

Greger, J.L. & Baier, M. (1981) Tin and iron content of canned and bottled food. J. Food Sci., 46, 1751–1754, 1765.

Greger, J.L. & Johnson, M.A. (1981) Effects of dietary tin on zinc, copper and iron utilization by rats. Food Cosmet. Toxicol., 19, 163–166.

Greger, J.L., Smith, S.A., Johnson, M.A. & Baier, M.J. (1982) Effects of dietary tin and aluminium on selenium utilization by adult males. Biol. Trace Elem. Res., 4, 269–278.

Hadjimarkos, D.M. (1967) Effect of trace elements in drinking water on dental caries. J. Pediatr., 70, 967–969.

Hamilton, E.I. & Minski, M.J. (1972–73) Abundance of the chemical elements in man’s diet and possible relations with environmental factors. Sci. Total Environ., 1, 375–394.

Hamilton, E.I., Minski, M.J., Cleary, J.J. & Halsey, V.S. (1972) Comments upon the chemical elements present in evaporated milk for consumption by babies. Sci. Total Environ., 1, 205–210.

Heintze, K. (1959) Ind. Obst. Gemuset., 44, 406.

Heintze, K. (1960) Ueber die Bildung von Metallchelaten bei Abstkonserven in Weissblechdosen. Dtsch. Lebensm. Rundsch., 7, 194.

Hiles, R.A. (1974) Absorption, distribution and excretion of inorganic tin in rats. Toxicol. Appl. Pharmacol., 27, 366–379.

Hislop, J.S., Fisher, E.M.R. & Pickford, C.J. (1980) Multi-element Analysis of Cabbage Samples Grown Near a Heavy Metal Smelter, Report AERE-G 1643, Harwell: Atomic Energy Research Establishment.

Hocquellet, P. & Labeyrie, N. (1977) At. Absorpt. Newsl., 16, 124.

Horio, T., Iwamoto, Y. & Shiga, I. (1967) In: Communications of the 5th International Congress on Canning, Vienna.

Iwamoto, Y., Maeda, Y. & Horio, T. (1970) Internal corrosion of cans. V. Dissolving of tin in cans after opening. Shokuhin Eiseigaku Zasshi, 11, 183–187 (Chem. Abstr., 73, 97519u).

Johnson, M.A. & Greger, J.L. (1982) Effects of dietary tin on tin and calcium metabolism of adult males. Am. J. Clin. Nutr., 35, 655–660.

Johnson, M.A. & Greger, J.L. (1984) Absorption, distribution and endogenous excretion of zinc by rats fed various dietary levels of inorganic tin and zinc. J. Nutr., 114, 1843–1852.

Johnson, M.A. & Greger, J.L. (1985) Tin, copper, iron and calcium metabolism of rats fed various dietary levels of inorganic tin and zinc. J. Nutr., 115, 615–624

Kehoe, R.A., Cholak, J. & Story, R.V. (1940) A spectrochemical study of the normal ranges of concentration of certain trace metals in biological materials. J. Nutr., 19, 579–592.

Kojima, K. (1969) Unpublished report submitted to WHO.

Kojima, K. (1971) Unpublished report submitted to WHO.

Kolmer, J.A., Brown, H. & Harkins, M.J. (1931) A note on tin compounds with chemotherapy of experimental staphylococcus infection. J. Pharmacol., 43, 515–519.

Kumpulainen, J. & Koivistoinen, P. (1977) Advances in tin compound analysis with special reference to organotin pesticide residues. Residue Rev., 66, 1–18.

Le Breton, R. (1962) Tin Toxicology, Thesis, University of Paris.

Luff, A.P. & Metcalfe, G.H. (1890) Four cases of tin poisoning caused by tinned cherries. Br. Med. J., i, 833–834.

Marciniak, M. (1981) Bivalent tin metabolism and toxicity after intravenous injection in rat. Acta Physiol. Pol., 32, 193–204.

Marro, N. (1996) The 1994 Australian Market Basket Survey, Canberra: Australia New Zealand Food Authority.

Martin, A.M., Sanchez, M., Espinosa, P. & Bagur, G. (1994) Determination of tin in canned fruits and vegetables by atomic absorption spectrometry and liquid–liquid extraction. J. AOAC Int., 77, 1627–1630.

Meah, M.N., Smart, G.A., Harrison, A.J. & Sherlock, J.C. (1991) Lead and tin in canned foods: Results of the UK survey 1983–1987. Food Addit. Contam., 8, 485–496.

Meranger, J.C. (1970) Bull. Environ. Contam. Toxicol., 5, 271–?

Meranger, J.C. (1975) J. Assoc. Off. Anal. Chem., 58, 1143–?.

Ministry of Agriculture, Fisheries and Food (1985) Survey of Aluminium, Antimony, Chromium, Cobalt, Indium, Nickel, Thallium and Tin in Food (Food Surveillance Paper No. 15). London: Her Majesty’s Stationery Office, pp . 63–72.

Ministry of Agriculture, Fisheries and Food (1997) Food Surveillance Information Sheet No. 122. http://www.foodstandards.gov.uk/maff/archive/food/infsheet.

Ministry of Agriculture, Fisheries and Food (2000) Food Surveillance Information Sheet No. 193. http://www.foodstandards.gov.uk/maff/archive/food/infsheet/.

Ministry of Health and Welfare, Japan (1969) Unpublished data submitted to WHO.

Monier-Williams, G.W. (1949) Tin. In: Trace Elements in Food, New York: John Wiley & Sons, pp. 138–161.

Nagy, S., Rouseff, R. & Ting, S.V. (1980) Effects of temperature and storage on the iron and tin contents of commercially canned single-strength orange juice. J. Agric. Food Chem., 28, 1166–1169.

National Academy of Sciences (1977) Drinking Water and Health, Washington DC: National Academy of Sciences, Safe Drinking Water Committee, pp. 292–296.

National Academy of Sciences (1979) The 1977 Survey of Industry on the Use of Food Additives, Washington DC: National Academy of Sciences, Committee on the GRAS List Survey–Phase III.

New Zealand (1994) Total Diet Survey Report, 1987/88.

New Zealand (1995) Total Diet Survey Report, 1990/91.

Nobel, P. (2000) Tin content of foods in cans from the German market. Working group of food chemists of the Lander and the Federal Institute for the Health Protection of Consumers and Veterinary Medicine. Submitted to WHO by the Federal Ministry of Health, Bonn.

Ogoshi, K., Kurumatani, N., Aoki, Y., Moriyama, T. & Nanzai, T. (1981) Decrease in compressive strength of the femoral bone in rats administered stannous chloride for a short period. Toxicol. Appl. Pharmacol., 58, 331–332.

Omori, Y. (1966) Tin as a potential cause of intoxication by canned orange juice. Folia Pharmacol. Jpn., 61, 77.

Omori, Y., Takanaka, A., Tanaka, S. & Ikeda, Y. (1973) Experimental studies on toxicity of tin in canned orange juice. J. Food Hyg. Soc., 14, 69–74.

Pelikan, Z., Halacka, K. & Cerny, E. (1968) Acute toxic effects of stannous chloride on white mice. Sci. Med., 41, 351–356.

Perloff, B. (2000) Personal communication.

Pickston, L. & Drysdale, J. (1986) Comparison of lead levels. I. Food stored in welded and soldered cans. Food Technol. N. Z., February, 16–18.

Pickston, L., Brewerton, H.V., Drysdale, J.M., Hughes, J.T., Smith, J.M., Love, J.L., Sutcliffe, E.R. & Davidson, F. (1985) The New Zealand diet: A survey of elements, pesticides, colours, and preservatives. N. Z. J. Technol., 1, 81–89.

Piotrowski, J.K. & Szymanska, J.A. (1976). Influence of certain metals on the level of metallothionein-like proteins in the liver and kidneys of rats, J. Toxicol. Environ. Health, 1, 991–1002.

Price, W.J. & Roos, J.T. (1969) Analysis of fruit juice by atomic absorption spectrophotometry. I. The determination of iron and tin in canned juice. J. Sci. Food Agric., 20, 437–439.

Rader, J.I. (1991) Anti-nutritive effects of tin. In: Friedman, M., ed., Nutritional and Toxicological Consequences of Food Processing, New York: Plenum Press.

Rader, J.I., Hight, S.C. & Capar, S.G. (1990) Copper depletion in Long-Evans rats fed inorganic tin. J. Trace Elem. Exp. Med., 3, 193–202.

Rodushkin, I., Odman, F. & Appelblad, P.K. (1999) Multielement determination and lead isotope ratio measurement in alcoholic beverages by high-resolution inductively coupled plasma mass spectrometry. J. Food Comp. Anal., 12, 243–257.

Sano, M., Furukawa, M., Kourai, M. & Tomita, I. (1979) Flameless atomic absorption spectrophotometric determination of Vendex, an organic tin miticide, in apples, oranges, and tea leaves. J. Assoc. Off. Anal. Chem., 62, 764–768.

Savage, W. (1939) Canned foods in relation to health. Lancet, ii, 991–995.

Schroeder, H.A. & Nason, A.P. (1976) Interactions of trace metals in mouse and rat tissues; zinc, chromium, copper and manganese with 13 other elements. J. Nutr., 106, 198–203.

Schroeder, H.A., Balassa, J.J. & Tipton, I.H. (1964) Abnormal trace metals in man: Tin. J. Chron. Dis., 17, 483–502.

Schryver, S.B. (1909) Some investigations on the toxicology of tin with special reference to the metallic contamination of canned foods. J. Hyg., 9, 253–263.

Schwarz, K., Milne, D.B. & Vinyard, E. (1970) Growth effects of tin compounds in rats maintained in a trace element-controlled environment. Biochem. Biophys. Res. Commun., 40, 22–29.

Sedgwick, L.W. (1888) Noxious salts of tin in fruits prepared in tin vessels. Lancet, i, 1129.

Sherlock, J.C. & Smart, G.A. (1984) Tin in foods and the diet. Food Addit. Contam., 1, 277–282.

Sherlock, J.C., Smart, G.A., Forbes, G.I., Moore, M.R., Patterson, W.J., Richards, W.N. & Wilson, T.S. (1982) Assessment of lead intakes and dose-response for a population in Ayr exposed to a plumbosolvent water supply. Hum. Toxicol., 1, 115–122.

Sherlock, J.C., Smart, G.A., Read, J.I., Brooke, P.J., Forbes, G., Patterson, W., Richards, W., Moore, M. & Wilson, T. (1983) Intakes of copper, zinc, cadmium, tin, iron, nickel and arsenic in a population exposed to lead from water. Sci. Total Environ., 30, 255–260.

Shimbo, S., Hayase, A., Murakami, M., Hatai, I., Higashikawa, K., Moon, C.S., Zhang, Z.W., Watanabe, T., Iguchi, H. & Ikeda, M. (1996) Use of a food composition database to estimate daily dietary intake of nutrient or trace elements in Japan, with reference to its limitation. Food Addit. Contam., 13, 775–786.

Sistrunk, W.A. & Gascoigne, H.L. (1983) Stability of color in ‘Concord’ grape juice and expression of color. J. Food Sci., 48, 430–433, 440.

Smart, G.A. & Sherlock, J.C. (1989) Survey of canned food storage in the home. Food Addit. Contam., 6, 125–132.

Smart, G.A., Sherlock, J.C. & Norman, J.A. (1988) Dietary intakes of lead and other metals: A study of young children from an urban population in the UK. Food Addit. Contam., 5, 85–93.

Sumitani, H., Shekane, S. & Nakatani, A. (1993) Inductively coupled plasma atomic emission spectrometric determination of tin in canned food. J. Assoc. Off. Anal. Chem., 76, 1374–1377.

Theuer, R.C., Mahoney, A.W. & Sarett, H.P. (1971) Placental transfer of fluoride and tin in rats given various fluoride and tin salts. J. Nutr., 101, 525–532.

Tipton, I.H., Stewart, P.L. & Martin, P.G. (1966) Trace elements in diets and excreta. Health Phys., 12, 1683–?.

Tipton, I.H., Stewart, P.L. & Dickson, J. (1969) Patterns of elemental excretion in long term balance studies. Health Phys., 16, 455–462.

Trachman, H.L., Tyberg, A.J. & Branigun, P.D. (1977) Atomic absorption spectrometric determination of sub-part-per-million quantities of tin in extracts and biological materials with a graphite furnace. Anal. Chem., 49, 1090-–1093.

Underwood, E.J. (1977) Trace Elements in Human and Animal Nutrition, 4th Ed., New York: Academic Press, pp. 449–451.

Vaessen, H.A.M.G., van Ooik, A. & van de Kamp, C.G. (1991) Some elements in domestic and imported fresh fruits marketed in The Netherlands. Z. Lebensm. Unters. Forsch., 193, 351–355.

Valberg, L.S. & Chamberlain, M.J. (1984) Effects of iron, tin, copper on zinc absorption in humans. Am. J. Clin. Nutr., 40, 536–541.

Vander, A.J., Mouw, D.R., Cox, J. & Johnson, B. (1979) Lead transport by renal slices and its inhibition by tin. Am. J. Physiol., 236, F373–F378.

Vannoort, R., Cressey, P. & Silvers, K. (2000) 1997/98 New Zealand Total Diet Survey. Part 2: Elements, Wellington: Ministry of Health.

Veerasai, W. (1990) Summary report on speciation of tin(II) and some fruit acid complexes by electrochemical studies. Unpublished report from the Chemistry Department, Mahidol University, Bangkok. Submitted to WHO by the Thai Industrial Standards Institute, Ministry of Industry, Bangkok, Thailand.

Warburton, S., Udler, W., Ewert, R.M. & Haynes, W.S. (1962) Outbreak of foodborne illness attributed to tin. Publ. Health Rep., 77, 798–800.

WHO/IAEA (1989) Minor and Trace Elements in Breast Milk, Geneva, WHO.

Winship, K.A. (1988) Toxicity of tin and its compounds. Adverse Drug React. Acute Poisoning Rev., 7, 19–38.

Woolfer, M.L. & Manu-Tawiak, W. (1977) Tin content of canned evaporated milk manufactured in West Africa. Ecol. Food Nutr., 6, 133–135.

Yamaguchi, M. (1980) Change of calcium content in organs of rats administered stannous chloride orally. Jpn. J. Hyg., 34, 719–722.

Yamaguchi, M., Sato, H. & Yamamoto, T. (1977) Increase in calcium binding activity in renal cortex of rats treated with stannous chloride, J. Toxicol. Environ. Health, 3, 413–420.

Yamaguchi, M., Kubo, Y. & Yamamoto, T. (1979) Inhibitory effect of tin on intestinal calcium absorption in rats. Toxicol. Appl. Pharmacol., 47, 441–444.

Yamaguchi, M., Kitade, M. & Okada, S. (1980a) The oral administration of stannous chloride to rats. Toxicol. Lett., 5, 275–278.

Yamaguchi, M., Saito, R. & Okada, S. (1980b) Dose–effect of inorganic tin on biochemical indices in rats. Toxicology, 16, 267–273.

Yamamoto, T., Yamaguchi, M. & Sato, H. (1976) Accumulation of calcium in kidney and decrease of calcium in serum of rats treated with tin chloride, J. Toxicol. Environ. Health, 1, 749–756.

Ysart, G., Miller, P., Crews, H., Robb, P., Baxter, M., L’Argy, C.D., Lofthouse, S., Sargent, C. & Harrison, N. (1999) Dietary exposure estimates of 30 elements from the UK Total Diet Study. Food Addit. Contam., 16, 391–403.

Zook, E.G., Green, F.E. & Morris, E.R. (1970) Nutrient composition of selected wheats and wheat products. Cereal Chem., 47, 720.