- Email: [email protected]
Significance of cadmium concentrations in blood and hair as an indicator of dose 15 years after the reduction of environmental exposure to cadmium
Toxicology Letters 123 (2001) 135– 141 www.elsevier.com/locate/toxlet
Significance of cadmium concentrations in blood and hair as an indicator of dose 15 years after the reduction of environmental exposure to cadmium Xiao Jie Liu a, Kokichi Arisawa a,*, Atsuhiro Nakano b, Hiroshi Saito a, Tatsuya Takahashi a, Akiko Kosaka a a
Department of Pre6enti6e Medicine and Health Promotion, Nagasaki Uni6ersity School of Medicine, 1 -12 -4 Sakamoto, Nagasaki 852 -8523, Japan b Department of Basic Medical Sciences, National Institute for Minamata Disease, Minamata, Japan Received 30 March 2001; received in revised form 24 May 2001; accepted 25 May 2001
Abstract To evaluate the significance of cadmium (Cd) concentrations in blood (B-Cd) and hair (H-Cd) as an indicator of dose, a cross-sectional study was performed on 40 residents in a Cd-polluted area, Nagasaki Prefecture, Japan, in 1996. In the study area, soil replacement of Cd-polluted rice fields ended in 1981. B-Cd and H-Cd were significantly higher in the study population than in the control subjects. B-Cd was positively correlated with urinary Cd (U-Cd) (Spearman r=0.50, P=0.06 for males and r= 0.72, P= 0.0001 for females), while H-Cd was weakly or moderately correlated with U-Cd. After adjustment for gender using logistic regression analysis, log(B-Cd) and log(U-Cd), but not log(H-Cd), were significantly associated with the prevalence of increased urinary b2-microglobulin (P for trend B 0.05). These findings suggest that B-Cd is a good indicator of cumulative dose many years after the reduction of environmental exposure to Cd. H-Cd may be weakly or moderately correlated with body burden. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: b2-microglobulin; Blood; Cadmium; Hair; Kidney function
1. Introduction Long-term exposure to high levels of cadmium (Cd) causes kidney damage. The renal dysfunction induced by Cd is characterized by an increased * Corresponding author. Tel.: + 81-95-849-7062; fax: + 8195-849-7064. E-mail address: [email protected] (K. Arisawa).
urinary excretion of low-molecular weight serum proteins, such as b2-microglobulin (b2M) (World Health Organization, 1992). In addition, a decrease in the glomerular filtration rate (GFR) has been reported among Cd-exposed workers and residents in Cd-polluted areas. Impaired renal functions caused by Cd are known to persist long after the cessation of exposure (Roels et al., 1989; Iwata et al., 1993).
0378-4274/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 8 - 4 2 7 4 ( 0 1 ) 0 0 3 8 1 - 2
136
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
It has been traditionally considered that urinary Cd (U-Cd) reflects total body burden of Cd and blood Cd (B-Cd) reflects recent exposure to this metal (Elinder et al., 1988). However, a recent follow-up study on Cd-exposed workers showed that B-Cd was a good indicator of cumulative dose many years after the cessation of exposure (Ja¨ rup et al., 1997). No definite conclusion has been reached regarding the significance of Cd in hair (H-Cd) as an index of exposure. An experimental study reported a strong positive correlation between H-Cd and total body burden in mice (Nordberg and Nishiyama, 1972). On the other hand, Ellis et al. (1981) reported that H-Cd was not a good indicator of body burden in occupationally exposed workers. In the present study, we evaluated the significance of B-Cd and H-Cd, as indicators of dose 15 years after the reduction of environmental exposure to Cd. The relationships of B-Cd and H-Cd to U-Cd and renal function were evaluated.
2. Materials and methods
2.1. Study area and population (Cd-polluted area) Kashine is a small village located on Tsushima Island, Nagasaki Prefecture, in southwestern Japan. Cd discharged from a zinc mine (Taishu Mine) contaminated this district, and the average dietary Cd intake among the residents exceeded 200 mg/day in the 1960s. However, it decreased to approximately half that amount in 1983, because the soil of the polluted paddy fields was replaced with nonpolluted soil from another island in 1981 (Iwata et al., 1993). In 1997, the average Cd intake from foods further decreased to roughly 30 mg/day (Harada, 1999). In July 1979, a survey of the Cd concentration in scalp hair was performed in the Kashine district. A total of 94 residents (37 men and 57 women) took part in this survey. In December 1996, a follow-up study was conducted on these 94 subjects. Nineteen subjects had already died, 13 had moved out of the study area, and the remaining 62 subjects served as the population for
the present study. Urine and hair samples were collected from 46 subjects, and venous blood was obtained from 41 subjects. The analysis was restricted to 40 subjects (15 men and 25 women) for whom urine, hair and blood samples were all available. In this paper, we used the data collected in 1996 only.
2.2. Collection and analysis of urine, hair and blood (Cd-polluted area) The purpose of the study was explained verbally, and informed consent was obtained from each subject. First-voided morning urine specimens were collected in wide-neck polyethylene bottles previously washed with nitric acid (3N). The pH of samples for determination of b2M was adjusted to approximately 7.0 by adding NaOH– NaH2PO4 buffer, and the samples for analysis of U-Cd were acidified by adding nitric acid. All samples were stored at −40 °C until analysis. Scalp hair within 3 cm from the skin was taken from behind the ear. Approximately 0.2 g of hair sample was washed with 0.4% polyoxyethylene lauryl ether solution in an ultrasound washer, rinsed with distilled water and dried at 100 °C for 4 h. Fasting blood was taken by venipuncture. Approximately 1 ml was directly put into graduated glass tubes that were previously washed with concentrated nitric acid at 110 °C for 4 h. b2M in plasma (P-b2M) and urine (U-b2M) was measured by radioimmunoassay (Phadebas kit, Pharmacia/Shionogi, Osaka, Japan). Plasma levels of ferritin and iron were analyzed by latex agglutination assay (LX test Eiken FER, Tokyo, Japan) and a colorimetric method (L-type Wako Fe, Osaka, Japan), respectively. Creatinine concentrations in plasma (P-Cr) and urine were measured by an enzyme method (L type Wako Creatinine F, Osaka, Japan) and Jaffe’s method, respectively. U-Cd was determined by flameless atomic absorption spectorophotometry (FAAS, Perkin Elmer 4110ZL model), after wet ashing with nitric acid and sulfuric acid, followed by extraction with ammonium pyrrolidine dithiocarbamate-metyl-isobutyl ketone. H-Cd and B-Cd were measured by FAAS after wet ashing with nitric acid and hydrogen peroxide, and dilution
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
with 0.1 N nitric acid. For quality assurance of the determination of b2M, U-Cd, H-Cd and B-Cd, control serum (Bio-Rad Laboratories, Lyphocheck Tumor Marker Control, Levels 1 and 2), urine (National Institute of Standards and Technology SRM 2670, USA), hair (National Institute for Environmental Studies No.13, Tsukuba, Japan), and blood (Biological Control Materials in the Danish National Institute of Occupational Health, AMI B 1001, B 1005 and B 1702, Denmark) were simultaneously analyzed. The certified and the estimated values agreed well with each other, and the coefficients of variation were within 6% (Table 1).
2.3. Collection of hair and blood, and determination of H-Cd, P-Cr and B-Cd in control areas During December 1997– March 1998, scalp hair samples were collected from 73 women (aged 20– 78) living in a non-Cd-polluted area of Nagasaki Prefecture. In addition, in September and October 2000, fasting blood samples were collected from 26 males (aged 28–77) and 23 females (aged 24– 83) living in another non-Cd-polluted area. H-Cd, P-Cr and B-Cd were determined by the same methods as used in the Cd-polluted area. Because of financial constraints, it was not possible to collect and analyze urine samples for control subjects.
2.4. Statistical analysis U-b2M and U-Cd were adjusted to urinary creatinine concentrations. B-Cd, H-Cd and U-Cd
137
were skewed to the right. Therefore, they were log-transformed when considered necessary. The Spearman rank correlation and logistic regression analysis were used to evaluate the relationships among B-Cd, H-Cd, U-Cd and renal function. All statistical analyses were performed using the SAS software package (version 6.12, SAS Institute, Inc. 1997), and all P-values reported are two-sided. 3. Results Table 2 shows the characteristics of the study population in the Cd-polluted area, according to gender. Because of the limited volume of samples, data on plasma iron and ferritin were not available for every individual. The median age of the study subjects was 70.0 for males and 67.8 for females. The median duration of residence in the Cd-polluted area (until 1981) was approximately 20 years shorter among females than males. The Spearman rank correlation coefficient between age and residential period for males and females was 0.99 and 0.74, respectively. The median B-Cd was 4.7 mg/l (range, 1.0–7.2) for males and 6.5 mg/l (2.1–16.2) for females, and the median H-Cd was 34.5 mg/kg (6.6–149.2) for males and 81.3 mg/kg (8.8–364.2) for females. B-Cd, H-Cd, U-Cd and U-b2M were significantly higher, while P-Cr was significantly lower, among females than males (PB 0.05 by Wilcoxon rank sum test). Plasma iron and ferritin levels were lower among females than males, although the difference was not significant. The prevalence of low-molecular weight proteinuria, as defined by U-b2M\ = 1,000 mg/g cr., in males and females was 4/15 (26.7%) and
Table 1 Comparison of certified and estimated values of b2-microglobulin and cadmium concentrations in reference materials
Plasma b2-microglobulin (mg/l) Plasma b2-microglobulin (mg/l) Cd in blood (mg/l) Cd in blood (mg/l) Cd in blood (mg/l) Cd in hair (mg/kg) Cd in urine (mg/l) a
CV, Coefficient of variation.
Certified value
Estimated value
CVa(%)
1,400 4,500 m0 9 0.0 m0+19.1990.01 6.3 90.6 230930 8.8090.3
1,405 965 5,300 9 130 8.86 9 0.30 29.37 90.82 6.68 9 0.36 237 9 9 8.78 9 0.40
4.6 2.5 3.3 2.8 5.3 3.7 4.7
138
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
Table 2 Characteristics of the study population in the Cd-polluted area, according to gender Variablesa
Age (years) Years of residencec B-Cd (mg/l) H-Cd (mg/kg) U-Cd (mg/g cr.) U-b2M (mg/g cr.) C-b2M (% CCr) P-b2M (mg/l) P-Cr (mg/100 ml) Plasma iron (mg/l) Plasma ferritin (mg/l)
Males
Females
No.
Median (range)
No.
Median (range)
P-valueb
15 15 15 15 15 15 15 15 15 11 10
70.0 (25.5–81.0) 54.2 (10.5–66) 4.7 (1.0–7.2) 34.5 (6.6–149.2) 4.6 (0.8–10.4) 80 (11–21,818) 0.05 (0.01–12.95) 1.3 (0.6–2.3) 0.77 (0.57–1.00) 920 (300–1,900) 60 (3–142)
25 25 25 25 25 25 25 25 25 21 18
67.8 (24.6–76.9) 33 (6–62) 6.5 (2.1–16.2) 81.3 (8.8–364.2) 6.9 (1.4–23.8) 266 (32–66,061) 0.10 (0.02–34.22) 1.3 (0.7–3.9) 0.62 (0.42–2.02) 825 (150–2,090) 37 (5–167)
0.40 0.11 0.01 0.03 0.009 0.02 0.07 0.51 0.0002 0.70 0.20
a B-Cd, blood cadmium; H-Cd, hair cadmium; U-Cd, urinary cadmium; U-b2M, urinary b2-microglobulin; cr., creatinine; C-b2M, b2-microglobulin clearance; CCr, creatinine clearance; P-b2M, plasma b2-microglobulin; P-Cr, plasma creatinine. b By the Wilcoxon rank sum test. c Years of residence in the cadmium-polluted area until 1981.
7/25 (28.0%), respectively. The prevalence of reduced GFR, as defined by increased P-b2M ( \ = 2.3 mg/l), in males and females was 1/15 (6.7%) and 2/25 (8.0%), respectively. The proportion of those with increased P-Cr (male, \ = 1.1 mg/dl; female, \ =0.8 mg/dl) was 0/15 (0%) and 2/25 (8.0%), respectively. Among the control subjects, the median B-Cd was 1.25 mg/l (range, 0.6– 5.2) for males and 1.4 mg/l (0.7–4.4) for females, which was significantly lower than that in the Cd-polluted area (P B0.05 by Wilcoxon rank sum test). The prevalence of increased P-Cr was 1/26 (3.8%) for male and 1/23 (4.3%) for female subjects, which was not significantly different from that of the Cd-polluted area in either gender (by Fisher’s exact test). The median H-Cd among 73 women in the control area was 22.7 mg/kg (range, 4.4– 204.5), which was significantly lower than that in the Cd-polluted area (PB 0.05 by Wilcoxon rank sum test). In the Cd-polluted area, there was a moderate to strong positive correlation between B-Cd and U-Cd (Fig. 1A, Spearman r = 0.50, P =0.06 for males and r=0.72, P = 0.0001 for females). On the other hand, H-Cd was weakly or moderately correlated with U-Cd (Fig. 1B, Spearman r =0.46 for males and 0.33 for females) or B-Cd (Fig. 1C,
Spearman r=0.07 for males and 0.29 for females). Table 3 shows the prevalence of increased Ub2M (\ = 1,000 mg/g cr.) in relation to B-Cd, H-Cd and U-Cd. After adjustment for gender using logistic regression analysis, there were significant dose–response relationships between log(B-Cd) and log(U-Cd) and the prevalence of low-molecular weight proteinuria. However, the trend for the association between log(H-Cd) and the prevalence of increased U-b2M was not significant. When 300 mg/g cr. was used as the cutoff point of U-b2M, the P for trend was 0.01 for log(B-Cd), 0.07 for log(H-Cd) and 0.002 for log(U-Cd). 4. Discussion The medians of B-Cd and H-Cd among the Kashine district residents were more than threefold higher than those of the control subjects, even long after the environmental Cd exposure had decreased. These results suggest that B-Cd and H-Cd are influenced by the body burden of Cd on a group basis. There was a significant dose–response relationship between B-Cd and the prevalence of in
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
creased U-b2M. In addition, B-Cd was positively correlated with U-Cd, an indicator of total body or kidney burden. These results suggest that, on an individual basis, B-Cd reflects cumulative dose long after the reduction of environmental exposure to Cd. A study on the biological half-life of B-Cd identified two components, i.e., one with a half-life of several months, and the other with a half-life of 7.4–16 years (Ja¨ rup et al., 1983). In
Fig. 1. (A) Correlation between cadmium concentrations in blood and urine. ( , males, r =0.50, P= 0.06; , females, r= 0.72, P = 0.0001). (B) Correlation between cadmium concentrations in hair and urine. ( , males, r = 0.46, P= 0.08; , females, r= 0.33, P =0.11). (C) Correlation between cadmium concentrations in hair and blood. ( , males, r= 0.07, P= 0.81; , females, r = 0.29, P= 0.16).
139
the present study, B-Cd was measured approximately 15 years after the reduction of exposure. Thus, the influence on B-Cd of the second component, which is related to Cd content in internal organs, may have been greater. This may be the reason why B-Cd was closely associated with the prevalence of low-molecular weight proteinuria. The present results were consistent with those of Ja¨ rup et al. (1997). They reported a correlation coefficient of 0.79 between B-Cd and U-Cd, and significant dose–response relationships between B-Cd and prevalences of increased urinary excretion of b2M and a1-microglobulin among Cd-exposed workers, who were followed for 15 years after the cessation of exposure. In the present study population, age was strongly correlated with duration of residence in the Cd-polluted area. In addition, all subjects resided in a narrow district with a small variation in the pollution level. These facts indicated that age was essentially an indicator of external dose. For these reasons, we did not adjust for the effect of age in dose–response analyses. An age-dependent rise in the prevalence of increased U-b2M (\ = 1,000 mg/g cr.) has been reported among residents in non-Cd-polluted areas. However, the prevalence is generally lower than 20% even among those aged 80 years or older (Hayano et al., 1996). Thus, the dose–response relationship between B-Cd and the prevalence of low-molecular weight proteinuria observed in this study cannot be explained by the confounding effect of aging alone. The trend for the association with the prevalence of low-molecular weight proteinuria was not as clear for H-Cd as for U-Cd or B-Cd. In addition, the correlation between H-Cd and U-Cd was weak to moderate. From these results, it may be concluded that H-Cd does not serve as a very good indicator of total body burden among persons previously exposed to environmental Cd. The Cd in hair may originate from Cd in circulating blood. Therefore, it is expected that H-Cd is closely correlated with B-Cd. However, the relationship between H-Cd and B-Cd was weaker than that between U-Cd and B-Cd. Some other factors, such as variation in the uptake of Cd
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
140
Table 3 Prevalence of increased urinary excretion of b2-microglobulin (\=1,000 mg/g creatinine) in relation to cadmium concentrations in blood, hair and urine Variablesa
P for trendb
Males
Females
Total
Prevalence (%)
Prevalence (%)
Prevalence (%)
B-Cd (vg/l) B5 5–9.9 10B=
2/10 (20.0) 2/5 (40.0) 0/0 –
0/5 (0.0) 4/15 (26.7) 3/5 (60.0)
2/15 (13.3) 6/20 (30.0) 3/5 (60.0)
0.009
H-Cd (vg/kg) B50 50–99 100B=
3/12 (25.0) 1/2 (50.0) 0/1 (0.0)
1/9 (11.1) 2/6 (33.3) 4/10 (40.0)
4/21 (19.0) 3/8 (37.5) 4/11 (36.4)
0.43
U-Cd (vg/g creatinine) B5 1/9 (11.1) 5–9.9 2/5 (40.0) 10B= 1/1 (100.0)
0/7 (0.0) 3/11 (27.3) 4/7 (57.1)
1/16 (6.3) 5/16 (31.3) 5/8 (62.5)
0.01
a b
Abbreviations are as shown in Table 2. By logistic regression analysis adjusted for gender.
from blood to hair root and external contamination, may have contributed to the variations in H-Cd. The prevalence of increased P-b2M or P-Cr was not high in the Kashine district residents. This was because subjects with a marked decrease in GFR had already died before the survey; increased P-b2M and P-Cr were strong predictors of total mortality in this population (Arisawa et al., 2001). Therefore, it was difficult to evaluate the associations of B-Cd, H-Cd and U-Cd with a reduction in GFR. The median duration of residence in the Cdpolluted area was approximately 20 years shorter among females than males. Nevertheless, B-Cd and H-Cd were significantly higher among females than males. It has been reported that gastrointestinal absorption of Cd is higher in females than in males, especially when the body iron stores are low (Buchet et al., 1990; Berglund et al., 1994). A somewhat higher B-Cd among female than male nonsmokers has also been reported in rural areas of Japan (Watanabe et al., 1983). In the present study population, plasma iron and ferritin levels were lower in females than males, as in the general population. However, plasma ferritin levels were not negatively correlated with
B-Cd, H-Cd, or U-Cd (data not shown). One reason might be that a single measurement of plasma ferritin does not necessarily represent the body iron store status from birth to the time of the health survey. In summary, the present results suggest that B-Cd is a good indicator of cumulative dose many years after the reduction of environmental exposure to Cd. H-Cd may be weakly or moderately correlated with total body burden. Acknowledgements We are grateful to Drs R. Oshibuchi and H. Oshibuchi for their co-operation in this study. This work was supported in part by a Grant-inAid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan (No. 11670345). References Arisawa, K., Nakano, A., Saito, H., Liu, X.-J., Yokoo, M., Soda, M., Koba, T., Takahashi, T., Kinoshita, K., 2001. Mortality and cancer incidence among a population previously exposed to environmental cadmium. Int. Arch. Occup. Environ. Health 74, 255 – 262.
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141 Berglund, M., Akesson, A., Nermell, B., Vahter, M., 1994. Intestinal absorption of dietary cadmium in women depends on body iron stores and fiber intake. Environ. Health Perspect. 102, 1058 – 1066. Buchet, J.P., Lauwerys, R., Roels, H., Bernard, A., Bruaux, P., Claeys, F., Ducoffre, G., de Plaen, P., Staessen, J., Amery, A., Lijnen, P., Thijs, L., Rondia, D., Sartor, F., Remy, A.S., Nick, L., 1990. Renal effects of cadmium body burden of the general population. Lancet 336, 699 – 702. Elinder, C.G., Gerhardsson, L., Oberdoerster, G., 1988. Biological monitoring of toxic metals-Overview. In: Clarkson, T.W., Friberg, L., Nordberg, G.F., Sager, P.R. (Eds.), Biological Monitoring of Toxic Metals. Plenum Press, New York and London, pp. 1– 71. Ellis, K.J., Yasumura, S., Cohn, S.H., 1981. Hair cadmium content: Is it a biological indicator of the body burden of cadmium for the occupationally exposed worker? Am. J. Ind. Med. 2, 323 – 330. Harada, T., 1999. Changes in the dietary intake of heavy metals and follow-up of persons with renal tubular dysfunction in a cadmium-polluted area. Kankyo Hoken Report (in Japanese) 65, 41 –44. Hayano, M., Nogawa, K., Kido, T., Kobayashi, E., Honda, R., Turitani, I., 1996. Dose –response relationship between urinary cadmium concentration and (b2-microglobulinuria using logistic regression analysis. Arch. Environ. Health 51, 162 – 167. Iwata, K., Saito, H., Moriyama, M., Nakano, A., 1993. Renal
141
tubular function after reduction of environmental cadmium exposure: a ten-year follow-up. Arch. Environ. Health 48, 157 – 163. Ja¨ rup, L., Rogenfelt, A., Elinder, C.G., Nogawa, K., Kjellstro¨ m, T., 1983. Biological half-time of cadmium in the blood of workers after cessation of exposure. Scand. J. Work Environ. Health 9, 327 – 331. Ja¨ rup, L., Persson, B., Elinder, C.G., 1997. Blood cadmium as an indicator of dose in a long-term follow-up of workers previously exposed to cadmium. Scand. J. Work Environ. Health 23, 31 – 36. Nordberg, G.F., Nishiyama, K., 1972. Whole-body and hair retention of cadmium in mice. Arch. Environ. Health 24, 209 – 214. Roels, H.A., Lauwerys, R.R., Buchet, J.P., Bernard, A.M., Vos, A., Oversteyns, M., 1989. Health significance of cadmium induced renal dysfunction: a five year follow up. Br. J. Ind. Med. 46, 755 – 764. SAS Institute Inc. 1997, SAS/STAT software: Changes and Enhancements through Release 6.12. SAS Institute, Inc, Cary, NC. Watanabe, T., Koizumi, A., Fujita, H., Kumai, M., Ikeda, M., 1983. Cadmium levels in the blood of inhabitants in nonpolluted areas in Japan with special references to aging and smoking. Environ. Res. 31, 472 – 483. World Health Organization. 1992, International Programme on Chemical Safety. Environmental Health Criteria 134 Cadmium, Geneva, pp. 131 – 146.
Significance of cadmium concentrations in blood and hair as an indicator of dose 15 years after the reduction of environmental exposure to cadmium Xiao Jie Liu a, Kokichi Arisawa a,*, Atsuhiro Nakano b, Hiroshi Saito a, Tatsuya Takahashi a, Akiko Kosaka a a
Department of Pre6enti6e Medicine and Health Promotion, Nagasaki Uni6ersity School of Medicine, 1 -12 -4 Sakamoto, Nagasaki 852 -8523, Japan b Department of Basic Medical Sciences, National Institute for Minamata Disease, Minamata, Japan Received 30 March 2001; received in revised form 24 May 2001; accepted 25 May 2001
Abstract To evaluate the significance of cadmium (Cd) concentrations in blood (B-Cd) and hair (H-Cd) as an indicator of dose, a cross-sectional study was performed on 40 residents in a Cd-polluted area, Nagasaki Prefecture, Japan, in 1996. In the study area, soil replacement of Cd-polluted rice fields ended in 1981. B-Cd and H-Cd were significantly higher in the study population than in the control subjects. B-Cd was positively correlated with urinary Cd (U-Cd) (Spearman r=0.50, P=0.06 for males and r= 0.72, P= 0.0001 for females), while H-Cd was weakly or moderately correlated with U-Cd. After adjustment for gender using logistic regression analysis, log(B-Cd) and log(U-Cd), but not log(H-Cd), were significantly associated with the prevalence of increased urinary b2-microglobulin (P for trend B 0.05). These findings suggest that B-Cd is a good indicator of cumulative dose many years after the reduction of environmental exposure to Cd. H-Cd may be weakly or moderately correlated with body burden. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: b2-microglobulin; Blood; Cadmium; Hair; Kidney function
1. Introduction Long-term exposure to high levels of cadmium (Cd) causes kidney damage. The renal dysfunction induced by Cd is characterized by an increased * Corresponding author. Tel.: + 81-95-849-7062; fax: + 8195-849-7064. E-mail address: [email protected] (K. Arisawa).
urinary excretion of low-molecular weight serum proteins, such as b2-microglobulin (b2M) (World Health Organization, 1992). In addition, a decrease in the glomerular filtration rate (GFR) has been reported among Cd-exposed workers and residents in Cd-polluted areas. Impaired renal functions caused by Cd are known to persist long after the cessation of exposure (Roels et al., 1989; Iwata et al., 1993).
0378-4274/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 7 8 - 4 2 7 4 ( 0 1 ) 0 0 3 8 1 - 2
136
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
It has been traditionally considered that urinary Cd (U-Cd) reflects total body burden of Cd and blood Cd (B-Cd) reflects recent exposure to this metal (Elinder et al., 1988). However, a recent follow-up study on Cd-exposed workers showed that B-Cd was a good indicator of cumulative dose many years after the cessation of exposure (Ja¨ rup et al., 1997). No definite conclusion has been reached regarding the significance of Cd in hair (H-Cd) as an index of exposure. An experimental study reported a strong positive correlation between H-Cd and total body burden in mice (Nordberg and Nishiyama, 1972). On the other hand, Ellis et al. (1981) reported that H-Cd was not a good indicator of body burden in occupationally exposed workers. In the present study, we evaluated the significance of B-Cd and H-Cd, as indicators of dose 15 years after the reduction of environmental exposure to Cd. The relationships of B-Cd and H-Cd to U-Cd and renal function were evaluated.
2. Materials and methods
2.1. Study area and population (Cd-polluted area) Kashine is a small village located on Tsushima Island, Nagasaki Prefecture, in southwestern Japan. Cd discharged from a zinc mine (Taishu Mine) contaminated this district, and the average dietary Cd intake among the residents exceeded 200 mg/day in the 1960s. However, it decreased to approximately half that amount in 1983, because the soil of the polluted paddy fields was replaced with nonpolluted soil from another island in 1981 (Iwata et al., 1993). In 1997, the average Cd intake from foods further decreased to roughly 30 mg/day (Harada, 1999). In July 1979, a survey of the Cd concentration in scalp hair was performed in the Kashine district. A total of 94 residents (37 men and 57 women) took part in this survey. In December 1996, a follow-up study was conducted on these 94 subjects. Nineteen subjects had already died, 13 had moved out of the study area, and the remaining 62 subjects served as the population for
the present study. Urine and hair samples were collected from 46 subjects, and venous blood was obtained from 41 subjects. The analysis was restricted to 40 subjects (15 men and 25 women) for whom urine, hair and blood samples were all available. In this paper, we used the data collected in 1996 only.
2.2. Collection and analysis of urine, hair and blood (Cd-polluted area) The purpose of the study was explained verbally, and informed consent was obtained from each subject. First-voided morning urine specimens were collected in wide-neck polyethylene bottles previously washed with nitric acid (3N). The pH of samples for determination of b2M was adjusted to approximately 7.0 by adding NaOH– NaH2PO4 buffer, and the samples for analysis of U-Cd were acidified by adding nitric acid. All samples were stored at −40 °C until analysis. Scalp hair within 3 cm from the skin was taken from behind the ear. Approximately 0.2 g of hair sample was washed with 0.4% polyoxyethylene lauryl ether solution in an ultrasound washer, rinsed with distilled water and dried at 100 °C for 4 h. Fasting blood was taken by venipuncture. Approximately 1 ml was directly put into graduated glass tubes that were previously washed with concentrated nitric acid at 110 °C for 4 h. b2M in plasma (P-b2M) and urine (U-b2M) was measured by radioimmunoassay (Phadebas kit, Pharmacia/Shionogi, Osaka, Japan). Plasma levels of ferritin and iron were analyzed by latex agglutination assay (LX test Eiken FER, Tokyo, Japan) and a colorimetric method (L-type Wako Fe, Osaka, Japan), respectively. Creatinine concentrations in plasma (P-Cr) and urine were measured by an enzyme method (L type Wako Creatinine F, Osaka, Japan) and Jaffe’s method, respectively. U-Cd was determined by flameless atomic absorption spectorophotometry (FAAS, Perkin Elmer 4110ZL model), after wet ashing with nitric acid and sulfuric acid, followed by extraction with ammonium pyrrolidine dithiocarbamate-metyl-isobutyl ketone. H-Cd and B-Cd were measured by FAAS after wet ashing with nitric acid and hydrogen peroxide, and dilution
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
with 0.1 N nitric acid. For quality assurance of the determination of b2M, U-Cd, H-Cd and B-Cd, control serum (Bio-Rad Laboratories, Lyphocheck Tumor Marker Control, Levels 1 and 2), urine (National Institute of Standards and Technology SRM 2670, USA), hair (National Institute for Environmental Studies No.13, Tsukuba, Japan), and blood (Biological Control Materials in the Danish National Institute of Occupational Health, AMI B 1001, B 1005 and B 1702, Denmark) were simultaneously analyzed. The certified and the estimated values agreed well with each other, and the coefficients of variation were within 6% (Table 1).
2.3. Collection of hair and blood, and determination of H-Cd, P-Cr and B-Cd in control areas During December 1997– March 1998, scalp hair samples were collected from 73 women (aged 20– 78) living in a non-Cd-polluted area of Nagasaki Prefecture. In addition, in September and October 2000, fasting blood samples were collected from 26 males (aged 28–77) and 23 females (aged 24– 83) living in another non-Cd-polluted area. H-Cd, P-Cr and B-Cd were determined by the same methods as used in the Cd-polluted area. Because of financial constraints, it was not possible to collect and analyze urine samples for control subjects.
2.4. Statistical analysis U-b2M and U-Cd were adjusted to urinary creatinine concentrations. B-Cd, H-Cd and U-Cd
137
were skewed to the right. Therefore, they were log-transformed when considered necessary. The Spearman rank correlation and logistic regression analysis were used to evaluate the relationships among B-Cd, H-Cd, U-Cd and renal function. All statistical analyses were performed using the SAS software package (version 6.12, SAS Institute, Inc. 1997), and all P-values reported are two-sided. 3. Results Table 2 shows the characteristics of the study population in the Cd-polluted area, according to gender. Because of the limited volume of samples, data on plasma iron and ferritin were not available for every individual. The median age of the study subjects was 70.0 for males and 67.8 for females. The median duration of residence in the Cd-polluted area (until 1981) was approximately 20 years shorter among females than males. The Spearman rank correlation coefficient between age and residential period for males and females was 0.99 and 0.74, respectively. The median B-Cd was 4.7 mg/l (range, 1.0–7.2) for males and 6.5 mg/l (2.1–16.2) for females, and the median H-Cd was 34.5 mg/kg (6.6–149.2) for males and 81.3 mg/kg (8.8–364.2) for females. B-Cd, H-Cd, U-Cd and U-b2M were significantly higher, while P-Cr was significantly lower, among females than males (PB 0.05 by Wilcoxon rank sum test). Plasma iron and ferritin levels were lower among females than males, although the difference was not significant. The prevalence of low-molecular weight proteinuria, as defined by U-b2M\ = 1,000 mg/g cr., in males and females was 4/15 (26.7%) and
Table 1 Comparison of certified and estimated values of b2-microglobulin and cadmium concentrations in reference materials
Plasma b2-microglobulin (mg/l) Plasma b2-microglobulin (mg/l) Cd in blood (mg/l) Cd in blood (mg/l) Cd in blood (mg/l) Cd in hair (mg/kg) Cd in urine (mg/l) a
CV, Coefficient of variation.
Certified value
Estimated value
CVa(%)
1,400 4,500 m0 9 0.0 m0+19.1990.01 6.3 90.6 230930 8.8090.3
1,405 965 5,300 9 130 8.86 9 0.30 29.37 90.82 6.68 9 0.36 237 9 9 8.78 9 0.40
4.6 2.5 3.3 2.8 5.3 3.7 4.7
138
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
Table 2 Characteristics of the study population in the Cd-polluted area, according to gender Variablesa
Age (years) Years of residencec B-Cd (mg/l) H-Cd (mg/kg) U-Cd (mg/g cr.) U-b2M (mg/g cr.) C-b2M (% CCr) P-b2M (mg/l) P-Cr (mg/100 ml) Plasma iron (mg/l) Plasma ferritin (mg/l)
Males
Females
No.
Median (range)
No.
Median (range)
P-valueb
15 15 15 15 15 15 15 15 15 11 10
70.0 (25.5–81.0) 54.2 (10.5–66) 4.7 (1.0–7.2) 34.5 (6.6–149.2) 4.6 (0.8–10.4) 80 (11–21,818) 0.05 (0.01–12.95) 1.3 (0.6–2.3) 0.77 (0.57–1.00) 920 (300–1,900) 60 (3–142)
25 25 25 25 25 25 25 25 25 21 18
67.8 (24.6–76.9) 33 (6–62) 6.5 (2.1–16.2) 81.3 (8.8–364.2) 6.9 (1.4–23.8) 266 (32–66,061) 0.10 (0.02–34.22) 1.3 (0.7–3.9) 0.62 (0.42–2.02) 825 (150–2,090) 37 (5–167)
0.40 0.11 0.01 0.03 0.009 0.02 0.07 0.51 0.0002 0.70 0.20
a B-Cd, blood cadmium; H-Cd, hair cadmium; U-Cd, urinary cadmium; U-b2M, urinary b2-microglobulin; cr., creatinine; C-b2M, b2-microglobulin clearance; CCr, creatinine clearance; P-b2M, plasma b2-microglobulin; P-Cr, plasma creatinine. b By the Wilcoxon rank sum test. c Years of residence in the cadmium-polluted area until 1981.
7/25 (28.0%), respectively. The prevalence of reduced GFR, as defined by increased P-b2M ( \ = 2.3 mg/l), in males and females was 1/15 (6.7%) and 2/25 (8.0%), respectively. The proportion of those with increased P-Cr (male, \ = 1.1 mg/dl; female, \ =0.8 mg/dl) was 0/15 (0%) and 2/25 (8.0%), respectively. Among the control subjects, the median B-Cd was 1.25 mg/l (range, 0.6– 5.2) for males and 1.4 mg/l (0.7–4.4) for females, which was significantly lower than that in the Cd-polluted area (P B0.05 by Wilcoxon rank sum test). The prevalence of increased P-Cr was 1/26 (3.8%) for male and 1/23 (4.3%) for female subjects, which was not significantly different from that of the Cd-polluted area in either gender (by Fisher’s exact test). The median H-Cd among 73 women in the control area was 22.7 mg/kg (range, 4.4– 204.5), which was significantly lower than that in the Cd-polluted area (PB 0.05 by Wilcoxon rank sum test). In the Cd-polluted area, there was a moderate to strong positive correlation between B-Cd and U-Cd (Fig. 1A, Spearman r = 0.50, P =0.06 for males and r=0.72, P = 0.0001 for females). On the other hand, H-Cd was weakly or moderately correlated with U-Cd (Fig. 1B, Spearman r =0.46 for males and 0.33 for females) or B-Cd (Fig. 1C,
Spearman r=0.07 for males and 0.29 for females). Table 3 shows the prevalence of increased Ub2M (\ = 1,000 mg/g cr.) in relation to B-Cd, H-Cd and U-Cd. After adjustment for gender using logistic regression analysis, there were significant dose–response relationships between log(B-Cd) and log(U-Cd) and the prevalence of low-molecular weight proteinuria. However, the trend for the association between log(H-Cd) and the prevalence of increased U-b2M was not significant. When 300 mg/g cr. was used as the cutoff point of U-b2M, the P for trend was 0.01 for log(B-Cd), 0.07 for log(H-Cd) and 0.002 for log(U-Cd). 4. Discussion The medians of B-Cd and H-Cd among the Kashine district residents were more than threefold higher than those of the control subjects, even long after the environmental Cd exposure had decreased. These results suggest that B-Cd and H-Cd are influenced by the body burden of Cd on a group basis. There was a significant dose–response relationship between B-Cd and the prevalence of in
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
creased U-b2M. In addition, B-Cd was positively correlated with U-Cd, an indicator of total body or kidney burden. These results suggest that, on an individual basis, B-Cd reflects cumulative dose long after the reduction of environmental exposure to Cd. A study on the biological half-life of B-Cd identified two components, i.e., one with a half-life of several months, and the other with a half-life of 7.4–16 years (Ja¨ rup et al., 1983). In
Fig. 1. (A) Correlation between cadmium concentrations in blood and urine. ( , males, r =0.50, P= 0.06; , females, r= 0.72, P = 0.0001). (B) Correlation between cadmium concentrations in hair and urine. ( , males, r = 0.46, P= 0.08; , females, r= 0.33, P =0.11). (C) Correlation between cadmium concentrations in hair and blood. ( , males, r= 0.07, P= 0.81; , females, r = 0.29, P= 0.16).
139
the present study, B-Cd was measured approximately 15 years after the reduction of exposure. Thus, the influence on B-Cd of the second component, which is related to Cd content in internal organs, may have been greater. This may be the reason why B-Cd was closely associated with the prevalence of low-molecular weight proteinuria. The present results were consistent with those of Ja¨ rup et al. (1997). They reported a correlation coefficient of 0.79 between B-Cd and U-Cd, and significant dose–response relationships between B-Cd and prevalences of increased urinary excretion of b2M and a1-microglobulin among Cd-exposed workers, who were followed for 15 years after the cessation of exposure. In the present study population, age was strongly correlated with duration of residence in the Cd-polluted area. In addition, all subjects resided in a narrow district with a small variation in the pollution level. These facts indicated that age was essentially an indicator of external dose. For these reasons, we did not adjust for the effect of age in dose–response analyses. An age-dependent rise in the prevalence of increased U-b2M (\ = 1,000 mg/g cr.) has been reported among residents in non-Cd-polluted areas. However, the prevalence is generally lower than 20% even among those aged 80 years or older (Hayano et al., 1996). Thus, the dose–response relationship between B-Cd and the prevalence of low-molecular weight proteinuria observed in this study cannot be explained by the confounding effect of aging alone. The trend for the association with the prevalence of low-molecular weight proteinuria was not as clear for H-Cd as for U-Cd or B-Cd. In addition, the correlation between H-Cd and U-Cd was weak to moderate. From these results, it may be concluded that H-Cd does not serve as a very good indicator of total body burden among persons previously exposed to environmental Cd. The Cd in hair may originate from Cd in circulating blood. Therefore, it is expected that H-Cd is closely correlated with B-Cd. However, the relationship between H-Cd and B-Cd was weaker than that between U-Cd and B-Cd. Some other factors, such as variation in the uptake of Cd
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141
140
Table 3 Prevalence of increased urinary excretion of b2-microglobulin (\=1,000 mg/g creatinine) in relation to cadmium concentrations in blood, hair and urine Variablesa
P for trendb
Males
Females
Total
Prevalence (%)
Prevalence (%)
Prevalence (%)
B-Cd (vg/l) B5 5–9.9 10B=
2/10 (20.0) 2/5 (40.0) 0/0 –
0/5 (0.0) 4/15 (26.7) 3/5 (60.0)
2/15 (13.3) 6/20 (30.0) 3/5 (60.0)
0.009
H-Cd (vg/kg) B50 50–99 100B=
3/12 (25.0) 1/2 (50.0) 0/1 (0.0)
1/9 (11.1) 2/6 (33.3) 4/10 (40.0)
4/21 (19.0) 3/8 (37.5) 4/11 (36.4)
0.43
U-Cd (vg/g creatinine) B5 1/9 (11.1) 5–9.9 2/5 (40.0) 10B= 1/1 (100.0)
0/7 (0.0) 3/11 (27.3) 4/7 (57.1)
1/16 (6.3) 5/16 (31.3) 5/8 (62.5)
0.01
a b
Abbreviations are as shown in Table 2. By logistic regression analysis adjusted for gender.
from blood to hair root and external contamination, may have contributed to the variations in H-Cd. The prevalence of increased P-b2M or P-Cr was not high in the Kashine district residents. This was because subjects with a marked decrease in GFR had already died before the survey; increased P-b2M and P-Cr were strong predictors of total mortality in this population (Arisawa et al., 2001). Therefore, it was difficult to evaluate the associations of B-Cd, H-Cd and U-Cd with a reduction in GFR. The median duration of residence in the Cdpolluted area was approximately 20 years shorter among females than males. Nevertheless, B-Cd and H-Cd were significantly higher among females than males. It has been reported that gastrointestinal absorption of Cd is higher in females than in males, especially when the body iron stores are low (Buchet et al., 1990; Berglund et al., 1994). A somewhat higher B-Cd among female than male nonsmokers has also been reported in rural areas of Japan (Watanabe et al., 1983). In the present study population, plasma iron and ferritin levels were lower in females than males, as in the general population. However, plasma ferritin levels were not negatively correlated with
B-Cd, H-Cd, or U-Cd (data not shown). One reason might be that a single measurement of plasma ferritin does not necessarily represent the body iron store status from birth to the time of the health survey. In summary, the present results suggest that B-Cd is a good indicator of cumulative dose many years after the reduction of environmental exposure to Cd. H-Cd may be weakly or moderately correlated with total body burden. Acknowledgements We are grateful to Drs R. Oshibuchi and H. Oshibuchi for their co-operation in this study. This work was supported in part by a Grant-inAid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan (No. 11670345). References Arisawa, K., Nakano, A., Saito, H., Liu, X.-J., Yokoo, M., Soda, M., Koba, T., Takahashi, T., Kinoshita, K., 2001. Mortality and cancer incidence among a population previously exposed to environmental cadmium. Int. Arch. Occup. Environ. Health 74, 255 – 262.
X.J. Liu et al. / Toxicology Letters 123 (2001) 135–141 Berglund, M., Akesson, A., Nermell, B., Vahter, M., 1994. Intestinal absorption of dietary cadmium in women depends on body iron stores and fiber intake. Environ. Health Perspect. 102, 1058 – 1066. Buchet, J.P., Lauwerys, R., Roels, H., Bernard, A., Bruaux, P., Claeys, F., Ducoffre, G., de Plaen, P., Staessen, J., Amery, A., Lijnen, P., Thijs, L., Rondia, D., Sartor, F., Remy, A.S., Nick, L., 1990. Renal effects of cadmium body burden of the general population. Lancet 336, 699 – 702. Elinder, C.G., Gerhardsson, L., Oberdoerster, G., 1988. Biological monitoring of toxic metals-Overview. In: Clarkson, T.W., Friberg, L., Nordberg, G.F., Sager, P.R. (Eds.), Biological Monitoring of Toxic Metals. Plenum Press, New York and London, pp. 1– 71. Ellis, K.J., Yasumura, S., Cohn, S.H., 1981. Hair cadmium content: Is it a biological indicator of the body burden of cadmium for the occupationally exposed worker? Am. J. Ind. Med. 2, 323 – 330. Harada, T., 1999. Changes in the dietary intake of heavy metals and follow-up of persons with renal tubular dysfunction in a cadmium-polluted area. Kankyo Hoken Report (in Japanese) 65, 41 –44. Hayano, M., Nogawa, K., Kido, T., Kobayashi, E., Honda, R., Turitani, I., 1996. Dose –response relationship between urinary cadmium concentration and (b2-microglobulinuria using logistic regression analysis. Arch. Environ. Health 51, 162 – 167. Iwata, K., Saito, H., Moriyama, M., Nakano, A., 1993. Renal
141
tubular function after reduction of environmental cadmium exposure: a ten-year follow-up. Arch. Environ. Health 48, 157 – 163. Ja¨ rup, L., Rogenfelt, A., Elinder, C.G., Nogawa, K., Kjellstro¨ m, T., 1983. Biological half-time of cadmium in the blood of workers after cessation of exposure. Scand. J. Work Environ. Health 9, 327 – 331. Ja¨ rup, L., Persson, B., Elinder, C.G., 1997. Blood cadmium as an indicator of dose in a long-term follow-up of workers previously exposed to cadmium. Scand. J. Work Environ. Health 23, 31 – 36. Nordberg, G.F., Nishiyama, K., 1972. Whole-body and hair retention of cadmium in mice. Arch. Environ. Health 24, 209 – 214. Roels, H.A., Lauwerys, R.R., Buchet, J.P., Bernard, A.M., Vos, A., Oversteyns, M., 1989. Health significance of cadmium induced renal dysfunction: a five year follow up. Br. J. Ind. Med. 46, 755 – 764. SAS Institute Inc. 1997, SAS/STAT software: Changes and Enhancements through Release 6.12. SAS Institute, Inc, Cary, NC. Watanabe, T., Koizumi, A., Fujita, H., Kumai, M., Ikeda, M., 1983. Cadmium levels in the blood of inhabitants in nonpolluted areas in Japan with special references to aging and smoking. Environ. Res. 31, 472 – 483. World Health Organization. 1992, International Programme on Chemical Safety. Environmental Health Criteria 134 Cadmium, Geneva, pp. 131 – 146.