Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Benchmark Dose for Urinary Cadmium based on a Marker of Renal Dysfunction: A Meta-Analysis

  • Hae Dong Woo,

    Affiliation Molecular Epidemiology Branch, National Cancer Center, Goyang-si, Korea

  • Weihsueh A. Chiu,

    Current address: Department of Veterinary Integrative Biosciences, Texas A&M University, College Station, Texas, United States of America

    Affiliation National Center for Environmental Assessment, U.S. Environmental Protection Agency (EPA), Washington, DC, United States of America

  • Seongil Jo,

    Affiliation Molecular Epidemiology Branch, National Cancer Center, Goyang-si, Korea

  • Jeongseon Kim

    jskim@ncc.re.kr

    Affiliation Molecular Epidemiology Branch, National Cancer Center, Goyang-si, Korea

Abstract

Background

Low doses of cadmium can cause adverse health effects. Benchmark dose (BMD) and the one-sided 95% lower confidence limit of BMD (BMDL) to derive points of departure for urinary cadmium exposure have been estimated in several previous studies, but the methods to derive BMD and the estimated BMDs differ.

Objectives

We aimed to find the associated factors that affect BMD calculation in the general population, and to estimate the summary BMD for urinary cadmium using reported BMDs.

Methods

A meta-regression was performed and the pooled BMD/BMDL was estimated using studies reporting a BMD and BMDL, weighted by sample size, that were calculated from individual data based on markers of renal dysfunction.

Results

BMDs were highly heterogeneous across studies. Meta-regression analysis showed that a significant predictor of BMD was the cut-off point which denotes an abnormal level. Using the 95th percentile as a cut off, BMD5/BMDL5 estimates for 5% benchmark responses (BMR) of β2-microglobulinuria (β2-MG) estimated was 6.18/4.88 μg/g creatinine in conventional quantal analysis and 3.56/3.13 μg/g creatinine in the hybrid approach, and BMD5/BMDL5 estimates for 5% BMR of N-acetyl-β-d-glucosaminidase (NAG) was 10.31/7.61 μg/g creatinine in quantal analysis and 3.21/2.24 g/g creatinine in the hybrid approach. However, the meta-regression showed that BMD and BMDL were significantly associated with the cut-off point, but BMD calculation method did not significantly affect the results. The urinary cadmium BMDL5 of β2-MG was 1.9 μg/g creatinine in the lowest cut-off point group.

Conclusion

The BMD was significantly associated with the cut-off point defining the abnormal level of renal dysfunction markers.

Introduction

Cadmium is a naturally occurring element that is widely dispersed in the environment at low concentrations. It is often emitted from industrial activities, such as metal mining and refining, which results in occupational exposure to cadmium [1]. The lungs are the target organs for cadmium in occupationally exposed workers because cadmium enters body through inhalation, which can result in an acute high dose. Cadmium has been classified as a human carcinogen [2]. In the general population, the majority of the cadmium exposure in the human body occurs by the intake of cadmium-containing food such as fish and shellfish, except in the case of smokers [3]. It has been estimated that the amount of cadmium per cigarette is about 1–2 μg and about 10% is inhaled when a cigarette is smoked [4]. However, smoking is often considered as a confounding variable in the assessment of cadmium exposure due to the large number of toxic compounds present [1].

Cadmium is less efficiently absorbed from the gastrointestinal tract than in the lungs. The absorption rate of cadmium in the gastrointestinal tract in humans has been estimated at approximately 3–5% [5] and can increase when the iron stores as assessed by serum ferritin are low [6]. Ingested cadmium accumulates mostly in the liver and kidney; however, the critical target organ for chronic low dose exposure is the kidney [1,7]. Although cadmium exposure in the general population is relatively low, cadmium has a long biological half-life in the body and a slow excretion rate. The estimated half-life in humans is 4–19 years in the liver tissue and 6–38 years in the kidney [1]. Chronic oral intakes of low dose cadmium can lead to cadmium accumulation in the kidney. The metallothionein-cadmium complexes prevent kidney damage, but cadmium become very toxic when the kidney cannot produce enough metallothionein. Chronic exposure to a low level of cadmium is associated with renal tubular dysfunction [8]. Many studies have shown a dose-response relationship between urinary cadmium and renal dysfunction markers, such as low molecular weight proteins [α1-microglobulin (also referred to as protein HC), β2-microglobulin (β2-MG), retinol binding protein (RBP)], intracellular tubular enzyme [N-acetyl-β-d-glucosaminidase (NAG)], and high molecular weight proteins [albumin (ALB)] [9].

To set a reference value for cadmium exposure, the levels of exposure that do not cause appreciable adverse health effects should be estimated. The no-observed-adverse-effect level (NOAEL) approach was traditionally used to set a point of departure for a guideline value of the human exposure limit. Later, the benchmark dose (BMD) was introduced for the risk assessment of contaminants [10,11]. The BMD is the dose level causing a specified change in the response (referred as the benchmark response, BMR) of a health effect, which is derived from the dose-response curve fitted to data. The BMD is generally defined as corresponding to an extra or additional risk of 5 or 10% of the BMR, although the methods to derive BMD and the one-sided 95% lower confidence limit of benchmark dose (BMDL) are varied. The BMD is less dependent on dose spacing and make full use of all dose-response data compared with NOAEL [12,13]. Thus the BMDL has been proposed as an alternative to the NOAEL [14].

Urinary cadmium, which is sensitive biomarker of long-term exposure, has been used as an index of the total body burden, and renal dysfunction has been used as a marker for adverse health effects [6]. Studies have reported the BMD/BMDL derived from urinary cadmium and renal dysfunction markers of a population from cadmium polluted or non-polluted area. Another approach is to conduct meta-analyses to determine the BMDs/BMDLs from a dose-response relationship between urinary cadmium and β2-MG using geometric means and standard deviations observed in each of the reported studies [3,15].

We conducted a meta-analysis using reported BMDs and their corresponding standard errors that were calculated from individual data, and we attempted to compare the results of previous reports as well as to find the factors that affect the BMD in the general population.

Material and Methods

Study selection

A systematic search for relevant studies written in English was conducted with PubMed using the terms (benchmark dose) and (cadmium) up to April 30, 2013. A manual search with a reference list of selected journals was also performed. The inclusion/exclusion criteria were as follows: (1) the article reported the cadmium BMD that was calculated using urinary cadmium concentration and markers of renal damage; (2) the article did not derive BMDs using life-time or rice cadmium intake; and (3) the article reported the cadmium BMD and BMDL.

Data collection

Data on the authors, publication year, country in which the study was performed, environmental cadmium exposure (exposure type), cut-off points for β2-MG (μg/g creatinine) and NAG (U/g creatinine), concentration of β2-MG (geometric mean) and NAG (geometric mean), urinary cadmium concentration (geometric mean, μg/g creatinine), sample size, and BMD and BMDL were collected for the meta-analysis. If only the mean concentrations of each group were reported, the overall mean concentration was calculated based on the number of samples in each group and their mean concentrations for the meta-regression analysis [1619].

Statistical analysis

All statistical analyses were performed using the STATA software package (version 12, Stata Corporation, College Station, TX). Summary estimates of BMDi and BMDLi were calculated by weighting sample size (ni) of the study [20], i.e., summary BMDL = Σ(ni×BMDLi)/Σni. The heterogeneity across studies was tested using the I2 test according to Higgins et al. [21]. A meta-regression was performed to identify both the cause of the heterogeneity and the associated factors. The urinary cadmium level, the cut-off point (as a continuous variable), BMD calculation method (hybrid vs. quantal model), and sex (total and male vs. female subjects) were used for BMDL5 and BMDL10 in the meta-regression.

Results

Characteristics of selected studies

A total of 27 studies were identified and 14 studies were excluded due to the following reasons: 3 studies did not include BMD; 2 studies did not include human population; 2 studies did not include renal damage markers; 3 studies included BMD for lifetime cadmium intake or cadmium concentration in rice; 3 studies did not include cut-off points or calculation method for BMD; and 1 study was a review article. Thirteen studies were finally selected for the review of the BMD derived from cadmium-induced renal damage (Fig 1) [1619,2230]. The most studied renal tubular damage markers were β2-MG and NAG for the urinary cadmium BMD. The BMD calculated from occupationally exposed population were included only in subgroup analysis. Nine studies (21 data points) for β2-MG and 8 studies (16 data points) for NAG were used for the pooled BMD estimates with exception of studies used occupationally exposed population. Selected studies reporting the BMD for the induction of β2-MG and NAG by cadmium were reported (Tables 1 and 2).

thumbnail
Table 1. Studies reporting benchmark dose for cadmium-induced β2-MG.

https://doi.org/10.1371/journal.pone.0126680.t001

thumbnail
Table 2. Studies reporting benchmark dose for cadmium-induced NAG.

https://doi.org/10.1371/journal.pone.0126680.t002

The BMDs were calculated with either 5 or 10% as the BMR level, and the cut-off points for adverse response were defined as corresponding to 84th percentile, 95th percentile, or 97.5th percentile in the control (mainly) or target population. The BMDs were calculated with various models using BMDS software developed by EPA in earlier studies [1619,2225,30], but hybrid approach which allows for calculation of BMD directly from continuous data was applied in more recent studies [2629]. The studies were mostly conducted in Asian populations (China and Japan). The reported or estimated mean urinary cadmium of the studies ranged from 0.8 to 3.49 μg/g creatinine in a non-polluted general population from Japan, from 2.72 to 2.76 μg/g creatinine in occupationally exposed workers from China, and from 1.51 to 8.0 μg/g creatinine in a cadmium polluted/non-polluted mixed area. The BMDs of 5% BMR (BMD5) for urinary cadmium based on β2-MG ranged from 0.5 to 8.7 μg/g creatinine in a non-polluted general population, and from 2.6 to 9.98 μg/g creatinine in a cadmium polluted/non-polluted mixed area. The BMDs of 10% BMR (BMD10s) for urinary cadmium based on β2-MG ranged from 1 to 12 μg/g creatinine in a non-polluted general population, from 4.58 to 5.2 μg/g creatinine for occupationally exposed workers, and from 4.2 to 9.4 μg/g creatinine in a cadmium polluted/non-polluted mixed area. The BMD5s and BMD10s for urinary cadmium based on NAG ranged from 0.3–12 μg/g creatinine and from 0.7 to 16.4 μg/g creatinine, respectively.

Meta-regression analysis for BMD5 and BMD10

Meta-regression analyses were conducted to identify the cause of the heterogeneity across the studies (Table 3). The urinary cadmium level, the cut-off point (as a continuous variable), BMD calculation method (hybrid vs. quantal model), and sex (total and male subjects vs. female subjects) were used in the meta-regression. The meta-regression for BMD of β2-MG showed that the BMD5 and BMD10 were significantly associated with the cut-off point, but BMD calculation method did not significantly affect the results. BMDs for β2-MG were mostly calculated using log-logistic model among conventional quantal analysis. BMD model (probit, quantal linear, logistic vs. log-logistic) was added in the meta-regression, but only cut-off point was significantly associated with BMD for β2-MG (data not shown). The obvious relationships between cut-off point of β2-MG and BMD in both conventional quantal analysis and the hybrid approach were observed (Fig 2).

thumbnail
Fig 2. The relationship between cut-off point for adverse effect and urinary BMD for β2-MG.

BMDS (▲): BMD calculated by EPA BMDS software, Hybrid (○): BMD calculated by hybrid approach. The positive relationships between cut-off point of β2-MG and BMD in both conventional quantal analysis and the hybrid approach were observed.

https://doi.org/10.1371/journal.pone.0126680.g002

thumbnail
Table 3. Meta-regression analysis for BMDL5 and BMDL10 derived from urinary cadmium and β2-MG.

https://doi.org/10.1371/journal.pone.0126680.t003

Pooled estimates of BMD/BMDL

The pooled estimates for the urinary cadmium BMD/BMDL derived from β2-MG and NAG are presented in Table 4. The meta-analyses were conducted separately by BMR % (BMD5 or BMD10) and cut-off (84th or 95th), and further analyses were performed based on BMD calculation method and sex. The BMD/BMDL varied across the studies, resulting in the between-study heterogeneity being significantly high in most of the groups analyzed except for the occupationally exposed group and 95th percentile cut-off group. In conventional quantal analysis with the 95th percentile as a cut off, the pooled BMD5/BMDL5 estimates of β2-MG were 6.18/4.88 μg/g creatinine and 8.30/6.64 μg/g creatinine, and those of NAG were 8.30/6.64 μg/g creatinine and 12.42/8.86 μg/g creatinine. In the hybrid approach with the 95th percentile as a cut off, the pooled BMD5/BMDL5 estimates of β2-MG and NAG were 3.56/3.13 μg/g creatinine and 3.21/2.24 μg/g creatinine. The pooled estimates were calculated according to the strata of the cut-off points of adverse effects. The urinary cadmium BMDL5 of β2-MG was 1.9 μg/g creatinine in the lowest cut-off point group. BMDL5 increased as the cut-off point increased, but it decreased in the highest strata. The pooled estimates were calculated with the studies that reported for both BMDs of β2-MG and NAG in a study, but no significant difference was observed between the two renal markers in both dose groups (Table 5).

thumbnail
Table 4. Pooled estimates of BMD derived from urinary cadmium and β2-MG and NAG.

https://doi.org/10.1371/journal.pone.0126680.t004

thumbnail
Table 5. Pooled estimates of BMD using studies that reported for both BMDs of β2-MG and NAG in a study.

https://doi.org/10.1371/journal.pone.0126680.t005

Discussion

The meta-regression showed that the BMDL was significantly related to the cut-off point for β2-MG, but was not significantly associated with BMD calculation method. The meta-analyses for the BMD/BMDL estimates were conducted using studies that reported BMD and BMDL calculated with urinary cadmium and markers of renal dysfunction. The mean urinary cadmium concentration as well as the BMD of the selected studies varied, and between-study heterogeneity was high in most subgroup meta-analysis.

The studies were mostly conducted in China and Japan. In Japan, a general population aged over 50 years old from either cadmium-polluted or non-cadmium polluted area were included. The study populations were mostly from one of the most cadmium-polluted areas, the Kakehashi river basin. A dose-response relationship between urinary cadmium and renal markers was also shown in the populations from the non-cadmium polluted areas, and the concentration of urinary cadmium varied. In China, the BMDs were calculated using occupationally exposed workers or the general population living in a cadmium polluted area. Studies with occupationally exposed groups showed homogenous results in the summary estimates of the present study. This may be because similar methods for sample collection and bioassay, BMR %, cut-off, and the use of models to estimate the BMD value were used in studies with occupationally exposed workers.

The relationship between cut-off point of β2-MG and BMDL was observed in the meta-regression analysis. However, BMD calculation model, sex, and urinary cadmium concentration were not significantly associated with BMDL. Ethnic differences and age effect on BMDL could not be evaluated in the meta-regression, because the selected populations were mostly Asian populations and only limited number of studies reported the mean age of study participants. The cut-off point in the calculation of the BMD is its critical point. In an analysis by the European Food Safety Authority [9], the cut-off point was determined either statistically or biologically. Statistically, the cut-off point was defined as corresponding to the 95th percentile of β2-MG at background exposure. The biological-based cut-offs for β2-MG were 300 and 1000 μg/g creatinine. A β2-MG level greater than 300 μg/g creatinine is considered adverse, and a level greater than 1000 μg/g creatinine is considered irreversible [31,32]. Nakagawa et al. [33] conducted a prospective cohort study in which the mortality rate was higher in the group with 300 to 1000 μg/g creatinine compared to the group with <300 μg/g creatinine after 9 years of follow up. In the studies selected for the present meta-analysis, a wide range of cut-off points have been used, and all the studies determined the cut-off point statistically. Lower background probability of adverse response yields higher BMD due to the s-shaped dose-response curve [26]. Therefore, higher cut-off point generally yields lower background probability, resulting in a higher BMD. The hybrid approach was applied in several studies for β2-MG [2729]. The hybrid approach allows for calculation of BMDL directly from continuous data, while in the conventional quantal analysis method, continuous data are dichotomized for analysis into quantal data according to cut-off point dividing the population into the affected and unaffected. However, the BMDs from hybrid approach, which fixed a 5% of background probability, were also significantly associated with cut-off point in the present meta-regression analysis. Thus careful consideration should be made to select cut-off point which denotes an abnormal level.

The excretion of urinary β2-MG and NAG, which are often used in cadmium toxicity studies, reflects tubular damage, whereas ALB reflects glomerular damage that is considered to be irreversible [1,34]. Thus, β2-MG and NAG are believed to be early markers of tubular dysfunction, and it has been suggested that NAG is a more sensitive marker for urinary cadmium than β2-MG [3537]. This led us to conduct an analysis to determine whether there were differences between NAG and β2-MG in response to the different levels of urinary cadmium concentration examined in the present study. Meta-analyses were performed by dividing the data into two groups based on urinary concentration. It was hypothesized that NAG showed smaller BMDL values in the lower concentration group (< 2 μg/g creatinine). However, there was not a large difference between β2-MG and NAG for both the low and high cadmium concentration groups, although the cut-off point of 2-MG and UNAG are different in those studies, which might significantly influence the BMDL values.

Conclusion

The BMD for cadmium was not significantly different between hybrid and conventional quantal data approaches, but was significantly associated with the cut-off point defining the abnormal level of renal dysfunction markers.

Supporting Information

Acknowledgments

Disclaimer: The views expressed in this article are those of the authors and do not necessarily represent the views or policies of the U.S. EPA.

Author Contributions

Conceived and designed the experiments: HDW JK. Analyzed the data: HDW SIJ. Wrote the paper: HDW JK. Critically reviewed: WAC.

References

  1. 1. ATSDR. Toxicological Profile for Chromium. Agency for Toxic Substances and Disease Registry, Public Health Service, US Department of Health and Human Services. 2012. http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=62&tid=17 pmid:24049864
  2. 2. IARC. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Beryllium, Cadmium, Mercury and Exposures in the Glass Manufacturing Industry. IARC Scientific Publications No. 58, IARC, Lyon. 1993.
  3. 3. EFSA. Scientific Opinion of the Panel on Contaminants in the Food Chain on a request from the European Commission on cadmium in food. 2009. 1–139.
  4. 4. WHO. World Health Organisation. Environmental Health Criteria 134—Cadmium. International Programme on Chemical Safety (IPCS) Monograph. 1992. http://www.inchem.org/documents/ehc/ehc/ehc134.htm
  5. 5. Morgan H, Sherlock J. Cadmium intake and cadmium in the human kidney. Food Addit Contam. 1984; 1: 45–51. pmid:6537346
  6. 6. Järup L, Åkesson A. Current status of cadmium as an environmental health problem. Toxicol Appl Pharmacol. 2009; 238: 201–208. pmid:19409405
  7. 7. Kotsonis FN, Klaassen CD. The relationship of metallothionein to the toxicity of cadmium after prolonged oral administration to rats. Toxicol Appl Pharmacol. 1978; 46: 39–54. pmid:725949
  8. 8. Satarug S, Moore MR. Adverse health effects of chronic exposure to low-level cadmium in foodstuffs and cigarette smoke. Environ Health Perspect. 2004; 112: 1099. pmid:15238284
  9. 9. EFSA. Technical report of EFSA prepared by Assessment Methodology Unit on Meta-analysis of Dose-Effect Relationship of Cadmium for Benchmark Dose Evaluation. EFSA Scientific Report. 2009; 254: 1–62.
  10. 10. Crump KS. A new method for determining allowable daily intakes. Fundam Appl Toxicol. 1984; 4: 854–871. pmid:6510615
  11. 11. Crump KS. Calculation of benchmark doses from continuous data. Risk Anal. 1995; 15: 79–89.
  12. 12. Davis JA, Gift JS, Zhao QJ. Introduction to benchmark dose methods and US EPA's benchmark dose software (BMDS) version 2.1. 1. Toxicol Appl Pharmacol. 2011; 254: 181–191. pmid:21034758
  13. 13. Barnes DG, Daston GP, Evans JS, Jarabek AM, Kavlock RJ, Kimmel CA, et al. Benchmark dose workshop: criteria for use of a benchmark dose to estimate a reference dose. Regul Toxicol Pharmacol. 1995; 21: 296–306. pmid:7644719
  14. 14. Sand S, Victorin K, Filipsson AF. The current state of knowledge on the use of the benchmark dose concept in risk assessment. J Appl Toxicol. 2008; 28: 405–421. pmid:17879232
  15. 15. JECFA. Evaluation of certain food additives and contaminants (Seventy-third report of the Joint FAO/WHO Expert Committee on Food Additives). WHO Technical Report Series, No. 960. 2011;
  16. 16. Jin T, Wu X, Tang Y, Nordberg M, Bernard A, Ye T, et al. Environmental epidemiological study and estimation of benchmark dose for renal dysfunction in a cadmium-polluted area in China. Biometals. 2004; 17: 525–530. pmid:15688857
  17. 17. Hong F, Jin T, Zhang A. Risk assessment on renal dysfunction caused by co-exposure to arsenic and cadmium using benchmark dose calculation in a Chinese population. Biometals. 2004; 17: 573–580. pmid:15688868
  18. 18. Kobayashi E, Suwazono Y, Uetani M, Inaba T, Oishi M, Kido T, et al. Estimation of benchmark dose as the threshold levels of urinary cadmium, based on excretion of total protein, β2-microglobulin, and N-acetyl-β-d-glucosaminidase in cadmium nonpolluted regions in Japan. Environ Res. 2006; 101: 401–406. pmid:16436274
  19. 19. Shimizu A, Kobayashi E, Suwazono Y, Uetani M, Oishi M, Inaba T, et al. Estimation of benchmark doses for urinary cadmium based on β2-microglobulin excretion in cadmium-polluted regions of the Kakehashi River basin, Japan. Int J Environ Health Res. 2006; 16: 329–337. pmid:16990174
  20. 20. Murata K, Iwata T, Dakeishi M, Karita K. Lead toxicity: does the critical level of lead resulting in adverse effects differ between adults and children? J Occup Health. 2009; 51: 1–12. pmid:18987427
  21. 21. Higgins J, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003; 327: 557. pmid:12958120
  22. 22. Chen L, Jin T, Huang B, Nordberg G, Nordberg M. Critical exposure level of cadmium for elevated urinary metallothionein—an occupational population study in China. Toxicol Appl Pharmacol. 2006; 215: 93–99. pmid:16510161
  23. 23. Kobayashi E, Suwazono Y, Uetani M, Inaba T, Oishi M, Kido T, et al. Estimation of benchmark dose for renal dysfunction in a cadmium non-polluted area in Japan. J Appl Toxicol. 2006; 26: 351–355. pmid:16791912
  24. 24. Lei L-J, Chen L, Jin T-Y, Nordberg M, Chang X-L. Estimation of benchmark dose for pancreatic damage in cadmium-exposed smelters. Toxicol Sci. 2007; 97: 189–195. pmid:17303580
  25. 25. Shao B, Jin T, Wu X, Kong Q, Ye T. Application of benchmark dose (BMD) in estimating biological exposure limit (BEL) to cadmium. Biomed Environ Sci. 2007; 20: 460. pmid:18348403
  26. 26. Suwazono Y, Sand S, Vahter M, Filipsson AF, Skerfving S, Lidfeldt J, et al. Benchmark dose for cadmium-induced renal effects in humans. Environ Health Perspect. 2006; 114: 1072. pmid:16835061
  27. 27. Suwazono Y, Nogawa K, Uetani M, Kido T, Nakagawa H. Reassessment of the threshold of urinary cadmium by using hybrid approach in a cadmium non-polluted area in Japan. Int J Hyg Environ Health. 2011; 214: 175–178. pmid:20889376
  28. 28. Suwazono Y, Nogawa K, Uetani M, Miura K, Sakata K, Okayama A, et al. Application of hybrid approach for estimating the benchmark dose of urinary cadmium for adverse renal effects in the general population of Japan. J Appl Toxicol. 2011; 31: 89–93. pmid:20836141
  29. 29. Suwazono Y, Nogawa K, Uetani M, Nakada S, Kido T, Nakagawa H. Application of the hybrid approach to the benchmark dose of urinary cadmium as the reference level for renal effects in cadmium polluted and non-polluted areas in Japan. Environ Res. 2011; 111: 312–314. pmid:21190684
  30. 30. Uno T, Kobayashi E, Suwazono Y, Okubo Y, Miura K, Sakata K, et al. Health effects of cadmium exposure in the general environment in Japan with special reference to the lower limit of the benchmark dose as the threshold level of urinary cadmium. Scand J Work Environ Health. 2005; 307–315. pmid:16161714
  31. 31. Bernard A. Renal dysfunction induced by cadmium: biomarkers of critical effects. Biometals. 2004; 17: 519–523. pmid:15688856
  32. 32. Kido T, Honda R, Tsuritani I, Yamaya H, Ishizaki M, Yamada Y, et al. Progress of renal dysfunction in inhabitants environmentally exposed to cadmium. Arch Environ Health. 1988; 43: 213–217. pmid:3289506
  33. 33. Nakagawa H, Nishijo M, Morikawa Y, Tabata M, Senma M, Kitagawa Y, et al. Urinary β2-microglobulin concentration and mortality in a cadmium-polluted area. Arch Environ Health. 1993; 48: 428–435. pmid:8250595
  34. 34. Järup L. Cadmium overload and toxicity. Nephrol Dial Transplant. 2002; 17: 35–39. pmid:12386256
  35. 35. Bernard A, Thielemans N, Roels H, Lauwerys R. Association between NAG-B and cadmium in urine with no evidence of a threshold. Occup Environ Med. 1995; 52: 177–180. pmid:7735390
  36. 36. Moriguchi J, Inoue Y, Kamiyama S, Horiguchi M, Murata K, Sakuragi S, et al. N-Acetyl-β-d-glucosaminidase (NAG) as the most sensitive marker of tubular dysfunction for monitoring residents in non-polluted areas. Toxicol Lett. 2009; 190: 1–8. pmid:19467302
  37. 37. Noonan CW, Sarasua SM, Campagna D, Kathman SJ, Lybarger JA, Mueller PW. Effects of exposure to low levels of environmental cadmium on renal biomarkers. Environ Health Perspect. 2002; 110: 151.] pmid:11836143