Environ. Sci. Technol. 2002, 36, 2684-2692
Nutritional Status Affects the Absorption and Whole-Body and Organ Retention of Cadmium in Rats Fed Rice-Based Diets P H I L I P G . R E E V E S * ,† A N D RUFUS L. CHANEY‡ USDA, ARS, Grand Forks Human Nutrition Research Center, 2420 Second Avenue North, Grand Forks, North Dakota 58203, and USDA, ARS, Animal Manure & Byproducts Laboratory, Room 013, Building 007, BARC-West, 10300 Baltimore Avenue, Beltsville, Maryland 20705-2350
Staple grains such as rice, wheat, and maize consumed by different societal groups differ greatly in their concentrations and bioavailability of the cadmium (Cd) antagonists, zinc (Zn), iron (Fe), and calcium (Ca). We hypothesized that the low nutritional status of rice consumers, which results from an inadequate supply of these minerals from rice, could contribute significantly to a higher apparent susceptibility to soil Cd contamination from rice than the higher nutritional status of those who consume other grains with higher mineral content. To test this hypothesis, a 2 × 2 × 2 factorial study was conducted. Rats were fed diets with adequate or marginal amounts of dietary Zn, Fe, or Ca. The basal diets contained 40% unenriched, milled rice fortified with 0.62 mg of Cd/kg as CdCl2 (0.25 mg of Cd/kg diet). Rat consumed the diets for 5 weeks and then were fed 1 g of a similar diet containing 109Cdlabeled rice. After 2 weeks, whole-body (WB) retention of 109Cd was determine. Rats then were killed, and the organs were removed for total Cd determinations. Rats fed marginal concentrations of dietary Zn had slightly but significantly more WB retention of 109Cd than controls; however, rats fed marginal Fe or Ca had as much as 3-fold higher retention of the label. Rats fed marginal amounts of Zn, Fe, and Ca combined retained as much as 8 times more 109Cd than rats fed adequate minerals. The effects on Cd concentrations in liver and kidney were similar to the effects on 109Cd retention. These results support the hypothesis that populations exposed to dietary sources of Cd and subsisting on marginal mineral intakes could be at greater risk than well-nourished populations exposed to similar amounts of dietary Cd. Thus, different food crops can cause unequal Cd risk at equal Cd concentration if diets containing the food are not balanced to provide adequate interacting mineral concentrations.
Introduction The World Health Organization has established a provisional tolerable weekly intake of cadmium (Cd) at 7 µg/kg body * Corresponding author phone: (701)795-8497; fax: (701)795-8395; e-mail:
[email protected]. † Grand Forks Human Nutrition Research Center. ‡ Animal Manure & Byproducts Laboratory. 2684 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 12, 2002
weight or 490 µg for a 70 kg person (1). It is estimated that continuous Cd intakes at this level or higher for 50 years could increase the body burden of this toxic element and might cause renal tubular dysfunction in sensitive individuals. Food is the major source of Cd exposure with an estimated average daily intake of 70-140 µg of Cd/week for adults in the United States (2-4). Numerous factors affect intestinal absorption and organ retention of Cd, including sex and age; another major factor is the mineral status of the individual. Studies have shown that the nutritional status of animals or humans with regard to zinc (Zn), iron (Fe), and/or calcium (Ca) can have a profound effect on the rate of Cd absorption from the gut. If the long-term intake of one or more of these minerals is low, the nutritional status is reduced and Cd absorption increases; conversely, if long-term intake is high, nutritional status is enhanced and Cd absorption is decreased (5-13). Thus, the mineral nutrient status resulting from the mineral composition of the diet could be an important factor that controls the extent of food Cd absorption and tissue accumulation. However, this factor generally has not been taken into account when interpreting the potential risk of food Cd to humans. Our recent study on the effect of marginal nutritional status of Zn, Fe, and Ca on the bioavailability of Cd in sunflower kernels (SFK) demonstrated a much higher rate of absorption and organ retention of Cd in rats given a marginal supply of these mineral nutrients than in those receiving an adequate supply (5). In addition, it was shown that the intrinsic, natural concentration of Zn, but not Ca and Fe, was enough to reduce the absorption and organ retention of dietary Cd supplied by the SFK. Foods such as rice, on the other hand, contain a very low intrinsic amount of Zn, Fe, or Ca (14). Would this cause an even more pronounced effect of malnutrition on Cd absorption and retention? Previous studies conducted to assess the risk of food chain Cd indicated that rice, because of its poor supply of Zn, Fe, and Ca, could have caused populations of subsisting rice consumers to suffer a high incidence of Cd-induced renal tubular dysfunction. These individuals consumed rice raised in soils that were contaminated by a mixture of ore wastes of Cd and Zn in a ratio of 0.5-1:100 µg. These populations seem to be more susceptible to Cd poisoning than those who consume more nutritious diets but with similar intakes of Cd (15, 16). Chaney et al. (17) noted that the grain of rice grown on flooded contaminated soils differed from other staple food crops by having essentially no increase in grain Zn when grain Cd was increased above safe levels when the rice was grown on soils with geological Cd plus Zn contamination. Vegetables, wheat, soybean, and other crops differ from rice in that they exclude Cd relative to Zn during growth and transport of Cd and Zn to edible plant tissues such that if grain or leaf Cd is increased, Zn is correspondingly increased. With the very strong evidence that mild Zn/Fe/Ca malnutrition increased Cd absorption from SFK (5), we needed to conduct a similar test with polished rice to test the hypothesis that Zn/Fe/Cd malnutrition was causal for Cdrelated diseases in populations that might consume high-Cd rice as a staple food. Several aspects of the current experimental approach should make the results more relevant to human nutrition and risks of dietary Cd than those commonly reported in most toxicology studies: (i) the crops tested have intrinsic or added Cd at concentrations that could occur in human diets; (ii) the ratios between Cd and other nutrients that interact with Cd are at concentrations that usually occur 10.1021/es0158307 Not subject to U.S. copyright. Publ. 2002 Am. Chem.Soc. Published on Web 05/16/2002
TABLE 1. Diet Composition dietary variable
diet 1
Zn Fe Ca
+a + +
+ + -
corn starch cooked riceb egg white, dried hydrolyzed starch sucrose soybean oil R-cellulose mineral mixc vitamin mixd choline bitartrate biotin premixe TBHQ mix f Zn premixg Fe premixh CaCO3
53.25 400 150 132 100 65 42 35 10 2.5 1.0 1.0 1.0 1.0 6.25
59.5 400 150 132 100 65 42 35 10 2.5 1.0 1.0 1.0 1.0 0
total (g)
1000
diet 2
1000
diet 3 + +
diet 4 + -
Ingredients (g) 54.25 60.5 400 400 150 150 132 132 100 100 65 65 42 42 35 35 10 10 2.5 2.5 1.0 1.0 1.0 1.0 1.0 1.0 0 0 6.25 0 1000
1000
diet 5
diet 6
diet 7
diet 8
+ +
+ -
+
-
54.25 400 150 132 100 65 42 35 10 2.5 1.0 1.0 0 1.0 6.25
60.5 400 150 132 100 65 42 35 10 2.5 1.0 1.0 0 1.0 0
55.25 400 150 132 100 65 42 35 10 2.5 1.0 1.0 0 0 6.25
61.5 400 150 132 100 65 42 35 10 2.5 1.0 1.0 0 0 0
1000
1000
1000
1000
Adequate (+) and marginal (-) amounts of test elements as analyzed (+/- Zn; 30/6.3 mg/kg diet; +/- Fe; 32/8.9 mg/kg diet; +/- Ca; 4737/2421 mg/kg diet). The diet contained 0.248 mg of Cd/kg. b Kokuho Rose, U.S. no.1 extra fancy, medium grain rice. The rice was cooked, freeze-dried, and ground to a powder before it was incorporated into the diet. Cadmium as CdCl2 was cooked into the rice to obtain 0.62 mg of Cd/kg dry wt. c The basal mix contained the following minerals (in g/kg mix): CaHPO , 227.72; KH PO , 86.2; K SO , 46.61; MgO, 13.36; Na SiO ‚9H O, 1.45; 4 2 4 2 4 2 3 2 FeSO4‚7H2O, 1.293; ZnCO3, 0.153; MnCO3, 0.318; CuCO3‚Cu(OH)2, 0.257; CrKSO4‚12H2O, 0.275; H3BO3, 0.082; NaF, 0.064; NiCO3, 0.032; NH4VO3, 0.007; (NH4)2MoO4, 0.008; LiCl, 0.018; KIO3, 0.01; Na2SeO4, 0.011; powdered sucrose, 622.132. d AIN-93-VX (26). e Biotin premix contained 1.8 g of D-biotin plus 998.2 g of corn starch. f TBHQ mix contained 50 g of tert-butylhydroquinone (an antioxidant) plus 950 g of soybean oil. g Zn premix contained 44.1 g of ZnCO3 plus 955.9 g of corn starch. h Fe premix contained 124.4 g of FeSO4‚7H2O plus 875.6 g of corn starch. a
in foods; and (iii) the animals are provided with “marginal” concentrations of dietary Zn, Fe, and Ca that do not reduce growth rates or cause untoward illness rather than with severely deficient concentrations that might cause the animals to become moribund. These tests more closely model the marginal nutrition status commonly found in human populations that subsist on rice-based diets (18-21). We believe that this approach models the potential chronic effects of exposure to Cd from human diets and permits accurate measurements of tissue Cd. This will allow us to observe the long-term effects as well as to follow absorption and excretion of 109Cd from a single test meal after the animals have been conditioned to the marginal Zn/Fe/Ca diets. In addition, this approach allows the regulatory responses to marginal Zn, Fe, and/or Ca in the intestine to be expressed in a manner similar to those of subsistence rice consumers.
Experimental Section The experimental design was a 2 × 2 × 2 factorial similar to that used in a previous study by us (5). The three factors were dietary Zn, Fe, and Ca, each at marginal or adequate concentrations. Because female animals tend to absorb Cd more readily than males, the experimental model we chose was the female rat [strain: SAS:VAF (SD), Charles River/Sasco, Willmington, MA] beginning at 3 weeks of age (22). There were 8 groups of 8 rats each. This study was approved by the Animal Use Committee of the USDA-ARS, Grand Forks Human Nutrition Research Center, and was in accordance with the guidelines of the National Institutes of Health on the experimental use of laboratory animals (23). Rice Processing. Milled rice (Kokuho Rose, unfortified, U.S. no. 1 extra fancy, medium grain; Nomura & Co., Inc., Burlingame, CA) was purchased from a local food store. By analysis, the rice contained 7.3 µg of Cd/kg, 10 mg of Zn/kg, 2 mg of Fe/kg, and 50 mg of Ca/kg of edible grain. Before the rice was incorporated into the diet, it was cooked according to package directions, frozen at -80° C, and lyophilized until the moisture content was less than 10%. The dried rice was
then ground to a fine powder before it was incorporated into the diet. To obtain a reasonable but higher than normal amount of Cd in the finished product, we incorporated enough CdCl2 into the cooking water to obtain 0.62 mg of Cd/kg of cooked, dried, ground rice. Diet. The compositions of the diets are shown in Table 1. The formulation of the diets was similar to the AIN-93GEGG diet described by Reeves (24) except that 50% of the carbohydrate source (starch and sucrose) and 5% of the protein source (egg white solids; Harlan Teklad, Madison, WI) were replaced with the cooked-dried rice. The source of fat was soybean oil at 7% of the diet. The mineral supplements to the basal diet were balanced so that none of the minerals in question would be far less than or far in excess of the dietary requirements for the laboratory rat (25). The amounts of Zn, Fe, and Ca in the marginal diets were supplied partially by the rice; however, the concentration of none of the three minerals was high enough to obtain a “marginal” amount. Thus, extra minerals were added as ZnCO3, FeSO4‚7H2O, and CaCO3 to bring them to the marginal amounts. In the adequate diet, the concentrations of Zn, Fe, and Ca were adjusted by the addition of more of the mineral salts to reach the AIN-93G diet concentrations (26). Chemical analysis showed that the marginal diets contained 6.3 ( 1.6 mg of Zn/kg, 8.9 ( 1.5 mg of Fe/kg, and 2421 ( 63 mg of Ca/kg (mean ( SD). These values represent approximately 52%, 25%, and 48% of the requirement of Zn, Fe, and Ca, respectively, for the growing rat according to the U.S. National Research Council (25). The adequate diets contained 30.1 ( 0.8 mg of Zn/kg, 31.6 ( 3.6 mg of Fe/kg, and 4737 ( 62 mg of Ca/kg. Zn concentrations in the adequate diets were higher than the requirement because 30 mg/kg is the value for the standard experimental diet for laboratory rodents, AIN-93G (26). Chemical analysis showed that the diets contained 0.25 ( 0.08 mg of Cd/kg; the unfortified rice supplied only 1.2% of this. Rats were fed their respective diets for 5 weeks, then 5 rats from each group were randomly selected and fasted between 2100 and 0600 h. Each of the fasted rats was given VOL. 36, NO. 12, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2685
FIGURE 1. Plot of the cumulative excretion of 109Cd in the feces of rats and expressed as a percentage of the initial dose. Female rats were fed their respective diets for 35 days and then were given an oral dose of diet containing 109Cd. Feces were collected daily for 15.5 days and assayed for radioactivity. Values represent the mean of 5 rats per group. Downward error bars on the fourth plot to the right in each graph shows the representative error (SEM) for each group.
TABLE 2. ANOVA Table P Values for 109Cd Retention in Whole Body, Whole Intestine, Whole Liver, and Both Kidneys and for Cd Concentration in These Organs and Femur (Figures 2 and 3) Zn
Fe
Zn × Fe
whole body intestine liver kidney
0.037 0.008 -
0.001 0.001 0.001 0.001
0.019 0.014 0.010 0.001
intestine liver kidney femur
0.002 -
0.001 0.001 0.001 -
-
109Cd
a
Ca Retained 0.001 0.001 0.001 -
Cd in Organs 0.001 0.001 0.003 -
Zn × Ca
Fe × Ca
Zn × Fe × Ca
-a -
0.001 0.013 0.001 0.001
-
0.020 -
0.002 0.001
0.034 0.039 -
Dash marks indicate no significant difference (P > 0.05).
a 1-g sample of their respective diets that contained 0.4 g of rice extrinsically labeled with 111 kBq of 109Cd/g as the chloride form. Soon after the rats had completely consumed their labeled meal (about 2 h after feeding), they were placed individually in a small animal whole body counter (WBC) to determine the amount of 109Cd ingested (27). These values were used as the initial doses. Afterward, animals were allowed to resume normal food consumption. After 15.5 days of consuming their respective diets, animals were placed back into the WBC, and the amount of 109Cd remaining was determined. The activity retained in each animal was expressed as a percentage of the initial dose. A phantom test “animal” with geometry similar to a rat, and with known amounts of 109Cd, was prepared and counted to ensure quality control and to determine the efficiency of counting with each runs. To estimate the fractional absorption of Cd, we collected the fecal material daily for 15 days from each animal and measured the amount of 109Cd in each sample. By using the cumulative excretion of daily 109Cd, the percentage of the intake excreted in the feces could be plotted over time (Figure 1). After the collection of 109Cd activity was completed, each rat was anesthetized, and blood was withdrawn from the abdominal aorta until the rat expired. Whole liver, kidneys, one femur, and the entire small intestine were collected for 109Cd activity determination and Cd, Zn, Fe, and Ca concentration analysis. The procedures used to prepare and analyze samples for mineral content were similar to those outlined by Reeves and Chaney (5). To ensure adequate quality control, samples of bovine liver with certified 2686
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 12, 2002
concentrations of minerals were analyzed with each batch of tissues, and all values were within the acceptable range [Cd, 500 ( 30 ng/g; Zn, 127 ( 16 µg/g; Fe, 184 ( 15 µg/g; Ca, 116 ( 4 µg/g (National Institutes of Standards and Technology, Gaithersburg, MD)]. Representative assayed values were Cd, 488 ( 35 ng/g; Zn, 124 ( 3 µg/g; Fe, 205( 16 µg/g; and Ca, 112 ( 6 µg/g (mean ( SD, n ) 6). In our facility, Fe always runs a little higher than the certified standard. The data were statistical analyzed by ANOVA for a full factorial design by using StatView or SAS computer programs (SAS Institute, Inc., Cary, NC). Differences between treatment means were assumed to be significant if P e 0.05.
Results Animal Growth. All rats fed the diets containing 40% cooked, dried rice had similar weight gains over the course of the experiment whether they received adequate or marginal amounts of Zn, Fe, or Ca (data not shown). The average daily weight gain over the 7-week experimental period was 2.4 ( 0.3 g, which is the normal rate of growth for the female rat. There were no effects of dietary treatment on liver and kidney fresh weights (overall mean ( SD; 5.18 ( 0.47 and 1.46 ( 0.13 g, respectively). 109Cd Retention. Whole body retention of 109Cd at day 15.5 after dosing ranged from 0.6 to 4% of the initial dose (Figure 2A). Rats that consumed diets low in either Zn, Fe, or Ca retained significantly more of the label than rats consuming diets with adequate concentrations of these minerals (Table 2). There was a significant (P < 0.02) interaction between Zn and Fe. This showed that although changing dietary Zn had no effect on 109Cd retention overall,
FIGURE 2. Retention of 109Cd in the whole body (A), total intestine (B), total liver (C), and both kidneys (D) of rats. Female rats were fed their respective diets for 35 days and then were given an oral dose of diet containing 109Cd. After 15.5 more days of consuming the diets, the amount of label remaining in the whole body and organs was measured. Bars represent the mean percent retention ( SEM for 5 rats per group. rats fed diets with marginal Zn along with marginal Fe had a greater retention of 109Cd than rats fed adequate Zn and marginal Fe. There also was a significant (P < 0.001) interaction between Fe and Ca, which showed that marginal Ca elevated WB 109Cd retention more when Fe was marginal than when Fe was adequate (Figure 2A). However, in this case, marginal Ca alone also increased WB retention of the label. Many studies use fecal excretion of a radiolabeled dose to estimate apparent absorption or body retention of a dietary ingredient. In this study, we measured fecal 109Cd every day for 15.5 days after dosing (Figure 1). Compared with whole body counting (WBC), this technique tends to overestimate the amount of Cd absorbed and retained. By day 15.5, the animals had lost an average of about 90% of the dose through the feces. However, when WBC was used, they were estimated to have actually lost from 96% to 99.4% of the dose. Part of this difference could have been caused by the loss of 109Cd in the urine. However, urinary Cd was not measured in this experiment. Counting the daily fecal loss of 109Cd shows the different patterns of loss as affected by the status of each mineral and combination of minerals (Figure 1). When animals were fed marginal amounts of Ca and Fe in the diet, the rates of fecal Cd loss between days 2 and 8 were slower than for animals fed adequate amounts of either of these minerals. When they also were fed a marginal Zn diet in addition to marginal Ca and Fe, the rate of loss was even slower (Figure 1). This seems to indicate that Cd from the meal was absorbed into the intestine and then later excreted or absorbed into the blood and then into the organs. Cd seemed to remain in the mucosal cells for a prolonged period when dietary Zn, Fe, or Ca nutrition was marginal (28). Organ Cd. The amount of 109Cd retained in the whole small intestine after 15.5 days ranged from 0.1 to 0.3% of the initial dose (Figure 2B). There was no significant direct effect of Zn on intestinal retention of the label; however, there was a significant (P < 0.015) interaction between Zn and Fe on
the intestinal retention of 109Cd. This was similar to the effect on WB retention of Cd between these two minerals in that rats fed diets with marginal Zn and marginal Fe had a higher intestinal retention of 109Cd than rats fed adequate Zn and marginal Fe. Likewise, the significant (P < 0.015) interaction between Fe and Ca was similar to that for WB retention of the label. There was a somewhat different pattern for nonlabeled Cd concentration in the intestine (Figure 3). In this case, there were only main effects of Fe and Ca where marginal amounts of each in the diet significantly (P < 0.001) elevated Cd concentration. Rats with marginal intakes of each of the three minerals alone had higher amounts of 109Cd in the liver than those given an adequate supply of each of these minerals (Figure 2C). There were significant interactions between Zn and Fe (P < 0.010) and between Fe and Ca (P < 0.001) (Table 2). These interactions were similar to those described above for 109 Cd in the intestine. Nonlabeled Cd in liver showed a different pattern than 109Cd (Figure 3). In this tissue, Cd concentration was significantly elevated by marginal intakes of each of the three minerals. There was no significant interaction between Zn and Fe or between Fe and Ca; however, there was a significant (P < 0.02) interaction between Zn and Ca. When marginal Ca was fed, liver Cd was elevated and similar between rats fed adequate or marginal Zn; however, when adequate Ca was fed, liver Cd was much higher in rats fed marginal Zn than in those fed adequate Zn. A complex interaction among Zn, Fe, and Ca also was revealed (P < 0.034) (Table 2). This three-way interaction is difficult to interpret; however, one salient feature was that rats fed adequate Zn and Ca with marginal Fe had only about 65% as much liver Cd as rats fed adequate Zn with marginal Fe and marginal Ca. In kidney, there was no overall effect of marginal dietary Zn on 109Cd, but there was a strong effect of Fe and Ca (Figure 2D; Table 2). In the latter, marginal intakes of each mineral elevated 109Cd by 2-3-fold. There was an interaction between Fe and Ca (P < 0.002), which showed that when rats were VOL. 36, NO. 12, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2687
FIGURE 3. Concentration (dry weight basis) of Cd in the small intestine, liver, kidneys, and femur of female rats after they had consumed their respective diets for 50 days. Bars represent the mean concentration ( SEM for 5 rats per group. The Cd concentrations usually found in these organs of female rats fed the basal diet without added Cd are 0.023 ( 0.007 mg/kg intestine, 0.034 ( 0.008 mg/kg liver, 0.045 ( 0.003 mg/kg kidney, and 5.2 ( 0.4 µg/kg femur (5). fed adequate Fe, marginal dietary Ca elevated kidney Cd over those fed adequate Ca; however when the rats were fed marginal Fe, there was no difference between dietary Ca concentrations, even though both values were higher than in rats fed adequate Fe (Figure 3). A three-way interaction among Zn, Fe, and Ca was found. As in liver, this complex interaction was difficult to interpret; however, one interesting observation was that rats fed marginal Zn alone had only 50% as much kidney Cd as rats fed marginal Zn, marginal Fe, and adequate Ca. Kidney Cd in the former group also was slightly lower than in the group fed adequate amounts of all minerals. Cadmium-109 in femur was barely above background (data not shown). Cadmium concentration in the femur, however, displayed an unusual pattern among treatment groups (Figure 3). The only significance was an interaction between Fe and Ca (P < 0.001) (Table 2). This showed that when dietary Fe was adequate, femur Cd was higher in rats fed marginal Ca than in those fed adequate Ca, but when dietary Fe was marginal, the effect was reversed. This was the case whether dietary Zn was adequate or marginal. Serum and Organ Zinc. As expected, feeding diets with marginal Zn significantly (P < 0.002) reduced serum Zn concentration (Table 3). On the other hand, rats fed marginal dietary Fe had significantly (P < 0.022) higher serum Zn than those fed adequate Fe. There was a significant (P < 0.013) interaction between Zn and Ca, which showed that when dietary Zn was adequate, marginal Ca elevated serum Zn, but when dietary Zn was marginal, marginal Ca lowered serum Zn. Intestinal Zn was not affected by dietary Zn, but there was a significant (P < 0.006) effect of dietary Fe on intestinal Zn (Table 3), where the concentration was about 10% higher in rats fed marginal Fe than in those fed an adequate amount of Fe. Dietary Ca had no effect on intestinal Zn. The concentration of liver Zn was not affected by dietary Zn or Fe; however, dietary Ca had a significant effect on liver Zn (Table 3). Rats fed diets with marginal amounts of Ca had 2688
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 12, 2002
liver Zn values that were about 5% (P < 0.009) higher than those in rats fed adequate Ca. The concentration of kidney Zn was slightly but significantly (P < 0.037) reduced in rats fed marginal Zn. No other effects were observed. There was a highly significant (P < 0.001) increase in femur Zn in rats fed marginal Ca diets (Table 3). Also, there was a significant (P < 0.001) interaction between dietary Zn and Fe, which showed an increase in femur Zn when dietary Zn was adequate and Fe was marginal, but no effect when dietary Zn was marginal. Marginal dietary Zn alone significantly (P < 0.001) lowered femur Zn by 23%. Marginal dietary Fe alone elevated femur Zn by 8% (P < 0.001), and marginal Ca alone elevated femur Zn by 16% (P < 0.001). Serum and Organ Iron. Rats fed marginal dietary Fe showed signs of marginal deficiency in Fe (Table 3). Serum Fe values were only about 80% of normal, and organ Fe was lower, as discussed below. We attempted to measure serum ferritin, but the values were extremely variable among individual rats. Thus, the data are not presented. Hemoglobin values were not measured in this study; however, in a separate identical study, hemoglobin was not affected by marginal Fe. Rats fed marginal Zn had about 21% lower serum Fe concentrations than controls. Dietary Ca had no effect on serum Fe. The concentration of intestinal Fe was affected by dietary Fe and Ca but not by Zn alone (Table 3). There was a significant (P < 0.050) interaction between Zn and Ca, which showed that when dietary Zn was marginal, marginal Ca elevated intestinal Fe more than when dietary Zn was adequate. Also, there was an interaction (P < 0.001) between Fe and Ca, which showed that rats fed marginal Ca and adequate Fe had much higher intestinal Fe than rats fed marginal Ca and marginal Fe, although marginal Fe itself lowered intestinal Fe by about 50%. Liver Fe was reduced (P < 0.001) to 30% of normal in rats fed marginal Fe in their diets (Table 3) but interacted with dietary Zn so that rats fed adequate Fe and marginal Zn had higher liver Fe than rats fed adequate Fe and adequate Zn.
TABLE 3. Effects of Marginal and Adequate Dietary Zn, Fe, and Ca on Concentration of Zn, Fe, and Ca in Serum and Organs of Female Rats Fed Diets Containing 40% Cooked, Dried, and Ground Ricea ANOVA table P values dietary variable Zn Fe Ca serum (mg/L) intestine (mg/kg) liver (mg/kg) kidney (mg/kg) femur (g/kg)
diet 1
diet 2
diet 3
diet 4
diet 5
diet 6
diet 7
+ + +
+ + -
+ +
+ -
+ +
+ -
+
2.00 ( 0.11 108.6 ( 3.7 103.9 ( 5.0 106.3 ( 3.8 173.6 ( 1.7
2.14 ( 0.08 111.6 ( 4.2 107.1 ( 4.3 110.6 ( 2.2 194.4 ( 6.5
2.12 ( 0.08 122.1 ( 4.3 100.4 ( 1.3 110.8 ( 3.6 199.1 ( 3.7
2.44 ( 0.13 119.2 ( 4.6 107.9 ( 3.5 116.9 ( 3.8 231.4 ( 3.3
1.95 ( 0.11 98.06 ( 3.2 105.1 ( 1.4 106.6 ( 2.6 134.6 ( 2.0
Zinc 1.76 ( 0.11 114.0 ( 4.9 109.3 ( 2.1 105.1 ( 2.7 170.7 ( 3.1
VOL. 36, NO. 12, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
132.3 ( 4.4 273.1 ( 22.2 96.2 ( 4.0 318.6 ( 28.6 246.7 ( 2.9
129.3 ( 5.9 257.9 ( 17.8 87.8 ( 2.7 319.1 ( 18.8 242.2 ( 3.4
136.0 ( 3.8 324.6 ( 31.0 94.6 ( 6.3 283.4 ( 8.9 255.4 ( 3.4
130.6 ( 5.8 244.9 ( 18.8 87.6 ( 4.1 277.2 ( 10.2 246.9 ( 1.6
122.9 ( 5.0 250.8 ( 19.3 91.9 ( 3.8 270.6 ( 5.1 252.1 ( 2.5
Calcium 116.9 ( 5.0 219.1 ( 10.2 141.2 ( 4.1 291.7 ( 13.4 248.8 ( 1.8
124.0 ( 5.2 214.1 ( 8.7 115.8 ( 2.0 278.0 ( 11.0 258.0 ( 1.7
Zn
Fe
Zn × Fe
Ca
Zn × Ca Fe × Ca Zn × Fe × Ca
-
2.08 ( 0.10 1.94 ( 0.13 117.9 ( 8.8 116.5 ( 6.8 103.2 ( 1.9 112.7 ( 3.5 102.3 ( 2.2 110.92 ( 4.3 134.6 ( 2.0 166.1 ( 3.1
Iron serum (mg/L) 5.55 ( 0.53 6.12 ( 0.49 4.29 ( 0.38 4.86 ( 0.45 4.34 ( 0.40 5.06 ( 0.41 3.58 ( 0.42 intestine (mg/kg) 75.3 ( 4.3 102.1 ( 5.3 38.9 ( 2.4 45.8 ( 3.8 70.0 ( 2.6 116.9 ( 6.5 40.8 ( 1.4 liver (g/kg) 1.62 ( 0.11 1.96 ( 0.15 0.49 ( 0.02 0.68 ( 0.04 1.75 ( 0.10 2.33 ( 0.14 0.39 ( 0.03 kidney (mg/kg) 289.2 ( 16.4 362.9 ( 16.6 138.5 ( 9.7 152.4 ( 10.7 268.6 ( 21.1 323.8 ( 8.8 122.6 ( 4.0 femur (mg/kg) 78.9 ( 4.3 101.8 ( 6.4 51.1 ( 3.7 65.9 ( 4.0 83.5 ( 4.8 124.8 ( 5.9 52.2 ( 3.1 serum (mg/L) intestine (mg/kg) liver (mg/kg) kidney (mg/kg) femur (g/kg)
diet 8
0.003 0.022 0.013 0.006 0.003 0.040 0.001 0.001 0.001 0.001 -
3.37 ( 0.25 52.2 ( 5.7 0.64 ( 0.07 160.4 ( 8.4 66.1 ( 5.1
0.001 0.040
0.001 0.001 0.001 0.001 0.001
-
119.0 ( 4.5 229.5 ( 14.9 103.5 ( 2.4 273.5 ( 9.0 250.9 ( 0.9
0.003 0.001 0.001 0.050 0.050 0.010 0.003
-
-
0.001 0.050 0.001 0.001 0.001 -
0.001 0.040 0.020
-
0.050 0.001 0.002 -
0.001 -
0.001 -
a Adequate (+) and marginal (-) amounts of test elements (+/- Zn, 30/6.3 mg/kg diet; +/- Fe, 32/8.9 mg/kg diet; +/- Ca, 4737/2421 mg/kg diet). The diet contained 0.248 mg of Cd/kg. Each value is the mean ( SEM for 8 replicates, expressed on a dry wt basis for the organs and on a liquid volume basis for the serum. Dash marks indicate no significant difference (P > 0.05).
9
2689
Rats fed marginal dietary Ca had 35% more liver Fe (P < 0.001) than rats fed adequate Ca. The effects of dietary manipulations of Zn, Fe, and Ca on kidney Fe concentrations were slightly different than for liver. There was no significant interaction between Fe and Zn, but there was one between Fe and Ca. This showed that feeding rats marginal Ca and adequate Fe elevated kidney Fe, but it was not elevated when fed marginal Ca and marginal Fe. Rats fed marginal dietary Ca alone had significantly (P < 0.001) higher kidney Fe than those fed adequate Ca. The effects of dietary mineral concentrations on femur Fe was almost identical with the effects on kidney except that marginal Zn significantly (P < 0.04) elevated femur Zn. Serum and Organ Calcium. The only effect on serum Ca was caused by dietary Zn (Table 3), where rats fed marginal Zn had serum Ca concentrations about 91% (P < 0.002) of those fed adequate Zn. Marginal dietary Zn also lowered intestinal Ca by about 17% (P < 0.001), and rats fed marginal Ca had slightly lower (P < 0.05) amounts of intestinal Ca than those fed adequate Ca. There was a three-way interaction, one effect of which showed that rats fed adequate Zn, marginal Fe, and adequate Ca had significantly more Ca in the intestine than any other group. The effects of feeding marginal concentrations of Zn, Fe, and Ca on liver Ca was complex. The only significant (P < 0.001) main effect was where marginal dietary Zn caused an elevation in liver Ca by about 20% over the control value. In addition, there were two-way interactions between Zn and Ca and between Fe and Ca, and there was also an interaction among all three dietary nutrients. The most salient feature of the three-way interaction was that when dietary Zn was marginal, adequate Fe and marginal Ca elevated liver Ca, but when dietary Fe and Ca were both marginal, liver Ca was depressed. Rats fed marginal amounts of dietary Zn or Fe had significantly (P < 0.05) lower amounts of Ca in kidney than rats fed adequate amounts of these minerals. Dietary Ca had no effect on kidney Ca. Femur Ca concentrations were elevated in rats fed marginal Zn (P < 0.01) and Fe (P < 0.003) but was lower in rats fed marginal Ca (P < 0.001) (Table 3). However, each change was small, and amounted to no more or less than 2.0%.
Discussion Low dietary intakes of certain minerals, such as Zn, Fe, or Ca, are known to lower the mineral nutrient status of individuals. This in turn leads to an enhancement in the absorption of Cd. In a previous study (5), we showed that female rats fed diets containing highly processed SFK and low concentrations of Cd had 4-fold higher rates of Cd absorption and organ retention when the diets also contained marginal amounts of Ca or Fe than when the diets contained adequate amounts of these essential nutrients. The marginal concentrations of these minerals were lower than that normally required by the rat, but not low enough to cause overt deficiencies. In fact, comparable dietary concentrations could have been encountered very easily in persons consuming subsistence rice diets. These consumers often have strong deficiencies of Zn and Fe, and perhaps Ca, that affect their health (20). The consumption of marginal Zn diets also are known to elevate the absorption of Cd; however, in our earlier experiment, the SFK contained a naturally high amount of Zn, which by itself limited excess absorption and organ accumulation of Cd (5). Would other foods that contain Cd but contain naturally low amounts of Zn, Fe, and/or Cd cause a greater elevation in Cd absorption and organ retention? There are several distinct differences in the manner in which humans can become injured by food Cd. The most obvious ones are the intake of large amounts of Cd over 2690
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 12, 2002
short periods or the intake of smaller amounts for long periods. However, there is a more subtle way the body can gain toxic amounts of Cd from a diet normally considered harmlesssby consuming a diet with low or marginal supplies of certain dietary minerals that are known to act against Cd absorption. Although for decades the Zn, Fe, or Ca nutritional status of an individual has been known to affect the intestinal absorption and organ retention of Cd, this factor seldom has been considered when making assessments about food Cd risk to humans. Because staple grains consumed by different societal groups can differ so greatly in concentrations and bioavailability of Fe, Zn, and/or Ca, we hypothesized that the nutritional status of consumers resulting from an inadequate supply of these minerals from rice, could significantly contribute to the their apparent susceptibility to soil-to-plant Cd transfer (29). This hypothesis was based on several human population studies where it was shown that although the Cd intakes were similar, some populations experienced high incidences of Cd-related disease and others did not. Certain populations of Japan and China who subsisted on rice-based diets experienced high incidences of renal tubular dysfunction when rice soils became contaminated with mine or smelter wastes that contained Cd and Zn (30, 31). These wastes contained Cd and Zn at the normal geological ratio of 5 µg of Cd:1000 µg of Zn, but because of the soil chemistry of the flooded rice paddies and the physiology of the rice plants, the rice grain contained up to 200 times more Cd than the grain from plants raised in uncontaminated paddies. Surprisingly, however, the grain did not contain an increased concentration of Zn. This indicates that rice excluded Zn from its grain (32), thus reducing the potential intake of a mineral that would normally counteract the absorption of Cd in consumers. Rice grain also contains limited amounts of Fe and Ca that would further reduce the intake of these minerals and enhance the absorption of dietary Cd. The current study bears out our conjecture in that marginal dietary intakes of Zn, Fe, and Ca by rats increased the percent of WB retention of dietary Cd from rice by 8-fold as compared to that in rats consuming adequate amounts of these minerals. For a perspective, a typical Cd concentration in the kidney in this study was 1.12 mg/kg dry wt or about 0.34 mg/kg fresh weight. Although the 8-fold increase in percent retention seems high, the amount of Cd in kidney was far less than the 200 mg of Cd/kg fresh weight of kidney cortex at which approximately 10% of a human population is expected to have early adverse effects of Cd. Because of concern about the possible health effects of soil Cd, epidemiological studies were conducted at locations in Western countries where soils had become very highly contaminated because of dispersal of mine wastes and smelter emissions. Three cases are important contrasts to subsistence rice farmers: Shipham, U.K. (33); Stolberg, Germany (34, 35); and Palmerton, PA (36). In these cases, the soils contained approximately 100 mg of Cd and 10 000 mg of Zn/kgsmuch higher than that found in the Japanese and Chinese soils (2-10 mg of Cd and 200-1200 mg of Zn/kg) that caused Cd-related diseases. However, although these populations raised gardens and consumed the foods from them, there was no evidence of proximal tubular dysfunction in older long-term residents, the ones most likely to have developed Cd-related diseases. These unexpected findings of no Cd-related disease in populations who lived for most of their lives in communities highly contaminated with Cd and Zn and who consumed garden foods grown on the contaminated soils stands in remarkable contrast to the findings in Japanese and Chinese farmers who subsisted on diets based on Cd-contaminated rice. Another study with a different source of food Cd was conducted by Sharma et al. (37) and McKenzie-Parnell et al.
(15) on a population in Bluff, New Zealand, who consumed large quantities of Cd from oysters during the harvest season. These oysters contain about 5 mg of Cd/kg fresh weight, higher than that found in most other oysters. Upon examination of blood and urine Cd, they found no large change in Cd concentration as a result of consuming the oyster-Cd in amounts high enough to cause Cd disease in subsistence rice farmers in Japan and China. Further examination of kidney Cd in deceased individuals from this community over time showed that kidney Cd was not unusually high after taking into consideration the effect of smoking on body Cd burden (15). However, smoking cigarettes caused the expected increase in blood Cd in the oyster consumers even though they did not absorb increased Cd from the oysters (15). One explanation for the lack of an effect of consuming high amounts of oyster-Cd is that oysters contain one of the highest concentrations of Zn of any food source. It was hypothesized that the high intake of Zn limited excess Cd absorption and Cd accumulation in the kidneys, a view borne out in the results of the present study. In addition, this population would be unlikely to have low Ca intakes because they consumed milk products and unlikely to have low Fe status because of the high intake of oysters. Although the main thrust of this study was to determine the effects of feeding marginal Zn, Fe, and/or Ca diets on the absorption and organ retention of Cd, there also were interactions among Zn, Fe, and Ca where one mineral affected the organ concentration of the other. These findings again point out the importance of recognizing the fact that individual nutrients, especially the mineral nutrients, are not metabolized in isolation. Absorption and retention of each could depend greatly on the dietary concentrations of other mineral agonists or antagonists. One notable effect in the present study was that of dietary Zn on serum and organ Ca. For the most part, our results were similar to that shown before, that marginal dietary Ca improved Zn status (38, 39). This was especially true for liver and bone and more obvious in those rats receiving a marginal amount of dietary Zn. Generally, because of the interest in Ca supplementation of the diet to help alleviate the incidence of osteoporosis, the effects of dietary Ca on Zn absorption and status have been studied more than the effects of Zn on Ca absorption. However, the present results showed a greater number of significant affects of Zn on Ca parameters than Ca on Zn, and they were variable. For example, marginal dietary Zn reduced the amount of Ca in most organs and serum, except liver and femur where Ca was elevated by marginal Zn. Again, the agonistic effect of certain minerals toward others must be considered along with the antagonistic effects, there are different physiological mechanisms for each, and the interactions cannot be generalized. It is commonly assumed, however, that animals or humans fed diets with very low Zn will have an overall higher Fe status because Zn tends to compete with Fe for absorption (40, 41). Nonetheless, in the present study, rats fed adequate dietary Zn had serum and kidney Fe values significantly higher than those in rats fed marginal Zn. This suggests that in some organs adequate dietary Zn is essential for adequate Fe uptake and metabolism. To summarize, the data presented suggest that a consumer’s mineral nutrient status could play a major role in determining how much dietary Cd is absorbed and retained in the body. Not only is the Cd concentration in food important, the availability of the food Cd to the consumer is important as well. Food Cd availability depends largely upon the type and concentration of dietary minerals that are antagonistic to Cd absorption and utilization. These data support previous findings, which showed that populations exposed to dietary sources of Cd and subsisting on foods that supply low mineral intakes could be at greater risk of developing Cd-related diseases than well-nourished popula-
tions who might be exposed to similar amounts of dietary Cd from other foods, e.g., wheat and sunflower kernels.
Acknowledgments Portions of this paper were presented at the Scientific Committee on Problems of the Environment (SCOPE) Workshop, Brussels, Belgium, September 2000 (42). This research was supported in part by the Northern Plains Area Office, United States Department of Agriculture, Agricultural Research Service, Fort Collins, CO. The study was approved by the Animal Use Committee of the USDA, ARS, Grand Forks Human Nutrition Research Center (GFHNRC) and was in accordance with the guidelines of the National Institutes of Health on the experimental use of laboratory animals. Mention of a trademark or proprietary product does not constitute a guarantee of warranty of the product by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other products that also might be suitable. The authors acknowledge the expert assistance of Jim Lindlauf and Karin Tweton for preparation and analyses of the diets; Denice Schafer and her staff for care of the animals; and the personnel in the Mineral Analysis Laboratory, GFHNRC, for mineral analyses of the diets and animal tissues.
Literature Cited (1) ) Walker, R.; Herrman, J. L. Summary and Conclusions of the Joint FAO/WHO Expert Committee on Food Additives; Report 55; World Health Organization: Geneva, 2000; http:// www.who.int/pcs/jecfa/. (2) Gunderson, E. L. J. AOAC Int. 1995, 78, 1353-1363. (3) Pennington, J. A. T.; Young, B. E.; Wilson, D. B.; Johnson, R. D.; Vanderveen, J. E. J. Am. Diet. Assoc. 1986, 86, 876-891. (4) Gartrell, M. J.; Craun, J. C.; Podrebarac, D. S.; Gunderson, E. L. J. Assoc. Off. Anal. Chem. 1986, 69, 146-161. (5) Reeves, P. G.; Chaney, R. L. Environ. Res. 2001, 85, 215-225. (6) Evans, G. W.; Magors, P. F.; Cornatzer, W. E. Biochem. Biophys. Res. Commun. 1970, 40, 1142-1148. (7) Brzo´ska, M. M.; Moniuszko-Jakoniuk, J. Arch. Toxicol. 1998, 72, 63-73. (8) Fox, M. R. S. J. Environ. Qual. 1988, 17, 175-180. (9) Fox, M. R. S. Fed. Proc. 1983, 42, 1726-1729. (10) Fox, M. R. S.; Jacobs, R. M.; Jones, A. O. L.; Fry, B. E., Jr. Environ. Health Perspect. 1979, 28, 107-114. (11) Flanagan, P. R.; McLellan, J. S.; Haist, J.; Cherian, M. G.; Chamberlain, M. J.; Valberg, L. S. Gastroenterology 1978, 74, 841-846. (12) Kello, D.; Kostial, K. Toxicol. Appl. Pharmacol. 1977, 40, 277282. (13) Koo, S. I.; Fullmer, C. S.; Wasserman, R. H. J. Nutr. 1978, 108, 1812-1822. (14) Pedersen, B.; Eggum, B. O. Qual. Plant.-Plant Foods Hum. Nutr. 1983, 33, 267-278. (15) McKenzie-Parnell, J. M.; Kjellstro¨m, T. E.; Sharma, R. P.; Robinson, M. F. Environ. Res. 1988, 46, 1-14. (16) Tachechi, J. Cadmium Studies in Japan; Elsevier: Tokyo, 1978; pp 283-286. (17) Chaney, R. L.; Ryan, J. A.; Li, Y.-M.; Brown, S. L. Cadmium in Soils and Plants; Kluwer Academic: Dordrecht, The Netherlands, 1999; pp 219-256. (18) United Nations ACC/SCN. Second Report on the World Nutrition Situation. Vol. 1. Global and Regional Results; United Nations: Geneva, Switzerland, 1992. (19) Graham, R. D.; Welch, R. M. Breeding for Staple-Food Crops with High Micronutrient Density: Long-Term Sustainable Agricultrual Solutions to Hidden Hunger in Developing Countries; International Food Policy Research Institute: Washington, DC, 1995. (20) Welch, R. M.; Graham, R. D. Food Nutr. Bull. 2000, 21, 361-366. (21) Underwood, B. Food Nutr. Bull. 2000, 21, 356-360. (22) Reeves, P. G.; Vanderpool, R. A. J. Nutr. Biochem. 1998, 9, 636644. (23) National Research Council. Guide for the Care and Use of Laboratory Animals; National Academy Press: Washington, DC, 1996. (24) Reeves, P. G. Trace Elements in Laboratory Rodents; CRC Press: Boca Raton, FL, 1996; pp 3-37. VOL. 36, NO. 12, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2691
(25) National Research Council. Nutrient Requirements of Laboratory Animals, Vol. 4; National Academy Press: Washington, DC, 1995. (26) Reeves, P. G.; Nielsen, F. H.; Fahey, G. C., Jr. J. Nutr. 1993, 123, 1939-1951. (27) Reeves, P. G.; Johnson, P. E.; Rossow, K. L. J. Agric. Food Chem. 1994, 42, 2836-2843. (28) Shaikh, Z. A.; Smith, C. Mechansims of Toxicity and Hazard Evaluation; Elservier/North-Holland Bomedical Press: Amsterdam, 1980; pp 569-574. (29) Chaney, R. L.; Ryan, J. A.; Li, Y.-M.; Welch, R. M.; Reeves, P. G.; Brown, S. L.; Green, C. E. Sources of Cadmium in the Environment; OECD: Paris, France, 1996; pp 49-78. (30) Nogawa, K. Changing Metal Cycles and Human Health; SpringerVerlag: New York, 1984; pp 275-284. (31) Tsuchiya, K. Cadmium Studies in Japan: A Review; Elsevier/ North-Holland Biomedical Press: New York, 1978. (32) Fukushima, M.; Ishizaki, A.; Sakamoto, M.; Kobayashi, E. Jpn. J. Hyg. 1973, 28, 406-415. (33) Strehlow, C. D.; Barltrop, D. The Shipham Reprot: An Investigation into Cadmium Contamination and Its Implications for Human Health; Elsevier Science Publishers B.V.: Amsterdam, 1988; pp 101-133. (34) Ewers, U.; Brockhous, A.; Dolgner, R.; Freier, I.; Jermann, E.; Bernard, A.; Stiller-Winkler, R.; Hahn, R.; Manojlovic, N. Int. Arch. Occup. Environ. Health 1985, 55, 217-239.
2692
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 12, 2002
(35) Ewers, U.; Freier, I.; Turfeld, M.; Brockhous, A.; Hofstetter, I.; Ko¨nig, W.; Leisner-Saaber, J.; Delschen, T. Gesundheitswesen 1993, 55, 318-325. (36) Sarasua, S. M.; McGeehin, M. A.; Stalling, F. L.; Terracciano, G. J.; Amler, R. W.; Rogue, J. N.; Fox, J. M. Final Report. Biologic Indicators of Exposure to Cadmium and Lead. Palmerton, PA. Part II; Agency for Toxic Substances and Disease Registry, USDHHS: Atlanta, GA, 1995. (37) Sharma, R. P.; Kjellstro¨m, T. E.; McKenzie, J. M. Toxicology 1983, 29, 163-171. (38) Forbes, R. M.; Parker, H. M.; Erdman, J. W., Jr. J. Nutr. 1984, 114, 1421-1425. (39) Wood, R. J.; Zheng, J. J. Am. J Clin. Nutr. 1997, 65, 1803-1809. (40) Hahn, C. J.; Evans, G. W. Am. J. Physiol. 1975, 228, 1020-1023. (41) Crofton, R. W.; Gvozdanovic, E.; Gvozdanovic, S.; Khin, C. C.; Brunt, P. W.; Mowat, N. A. G.; Aggett, P. J. Am. J Clin. Nutr. 1989, 50, 141-144. (42) Reeves, P. G. Proceedings of the SCOPE Workshop on Environmental Cadmium in the Food Chain: Sources, Pathways, and Risks. (Sept. 13-16, 2000; SCOPE: Paris, France, 2001; pp 8286.
Received for review December 6, 2001. Revised manuscript received March 23, 2002. Accepted April 15, 2002. ES0158307
