The NOAEL equivalent for the cumulative body burden of cadmium: focus on proteinuria as an endpoint
0 Abstract
The risk of developing chronic kidney disease (CKD), signified by a decrease in the estimated glomerular filtration rate (eGFR), has been linked to long-term exposure to low levels of the metal pollutant cadmium (Cd). Proteinuria is a hallmark of CKD and predicts continued progressive functional decline of the kidney. The aim of this study was to use the extent of proteinuria for Cd health risk assessment. Data were from 405 apparently healthy Thai nationals, of whom 12.6% had an eGFR below 60 mL/min/1.73 m2 (low eGFR), and 16.3% and 13.5% had moderate and severe proteinuria. Urinary excretion of Cd (ECd) and urinary total protein (Epro) were measured and normalized to both creatinine clearance (Ccr) and creatinine excretion (Ecr). We found that the risk of having a low eGFR [prevalence odds ratio (POR) = 12.2, P < 0.001] and severe proteinuria [POR = 10.4, P = 0.001) were increased markedly for every ten-fold increase in ECd/Ccr. However, when ECd was normalized to Ecr, the association between eGFR and ECd was found to be insignificant due to non-differential errors introduced by the Ecr-normalization. Respective benchmark dose limit (BMDL) values of ECd/Ecr that increased protein excretion by 5% and 10% were 0.0536 and 0.1140 µg/g creatinine. The ECd/Ecr at which 5% of the population had Cd-related proteinuria was
Keywords
INTRODUCTION
Dietary exposure to the metal contaminant cadmium (Cd) continues to be one of the most significant public health threats worldwide, given that Cd has no nutritional value or physiological role, and the body burden of Cd increases with age due to a lack of excretory mechanisms[1,2]. Concerningly, the health risk of dietary Cd exposure has been vastly underappreciated because the toxicological risk assessment relied mostly on an increase in urinary excretion of the protein β2-microglobulin (β2M) above 300 µg/g creatinine, termed β2-microglobulinuria[3-6]. Using the β2M as a toxicity endpoint, the Food and Agriculture Organization and World Health Organization (FAO/WHO) Joint Expert Committee on Food Additives and Contaminants (JECFA) found a tolerable intake level of Cd to be 0.83 µg/kg body weight/day (58 µg/day for a 70-kg person), and they identified urinary Cd excretion of 5.24 µg/g creatinine as a toxicity threshold level[4]. The European Food Safety Authority (EFSA) considered the kidneys to be the critical target of Cd toxicity, and employed the β2M as an endpoint. The EFSA designated a Cd excretion rate of
Current evidence, however, has linked kidney malfunction, reflected by a fall of the estimated glomerular filtration rate (eGFR) below 60 mL/min/1.73 m2 (termed a low eGFR) with urinary Cd excretion of 0.27-0.32 µg/g creatinine[7,8], which are much lower than the JECFA’s and the EFSA’s threshold levels mentioned above. Furthermore, in the China Health and Nutrition Survey, a dietary Cd exposure higher than 16.7 µg/day appeared to be sufficient to increase the risk of having a low eGFR[9]. These findings raise a serious concern that the existing Cd exposure guidelines are not protective of population health. Additionally, they question the utility of the β2M endpoint as a basis to define a permissible exposure level for Cd and its toxicity threshold level, given that a low eGFR is a pathological sign for the diagnosis of chronic kidney disease (CKD)[2].
Reductions in eGFR after Cd exposure are irreversible, and further decline is likely if exposure persists[2,8]. A β2M level above 300 µg/g creatinine is indicative of severe kidney pathologies, such as rapid kidney functional deterioration[10,11]. Thus, it is inappropriate to define a toxicity threshold level of Cd based on the β2M excretion exceeding 300 µg/g creatinine. Chronic environmental Cd exposure has been linked to an increased risk of kidney damage[12-14] concurrently with low eGFR and proteinuria[15,16]. The use of proteinuria as an endpoint in the toxicological risk assessment of Cd has never been explored. Indeed, proteinuria is a key biomarker of kidney disease, as it predicts continued progressive functional decline of the kidney or a rapid drop in eGFR to 15 mL/min/1.73 m2, which marks end-stage kidney disease[17-19], a condition that requires dialysis or a kidney transplant for survival. The associated healthcare costs are substantial.
The present study has three major objectives. The first is to explore the dose-response relationship between environmental Cd exposure and urinary excretion of total protein (Epro). The second objective is to define the benchmark dose limit (BMDL) of Cd excretion using Epro as a biomarker of an adverse effect of Cd on kidneys. The BMDL value derived at 5% benchmark response (BMR) has increasingly been used as a replacement for the no-observed-adverse-effect level (NOAEL) which has some shortcomings[20,21]. The NOAEL is defined as the highest experimental dose level at which the response does not significantly differ from the control group[20,21]. For comparison, the dose-response relationship between Cd exposure and eGFR reductions was determined simultaneously. The third objective is to address the “unrecognized” imprecision in the toxicological assessment of Cd exposure and its effects on kidneys. This imprecision arises from the practice of adjusting Cd excretion and urinary biomarkers of adverse effects, such as ECd and Epro, to creatinine excretion (Ecr) as ECd/Ecr and Epro/Ecr. These adjustments introduce non-differential errors, which tend to bias the dose-response relationship toward the null[22]. As a result, current dietary Cd exposure limits, which are computed by adjusting the excretion rates of Cd and biomarkers of kidney effects to Ecr, underestimate the severity of Cd-induced nephrotoxicity. Consequently, these exposure limits are not low enough to afford health protection.
EXPERIMENTAL
Study design
We assembled archived data from large Thai population-based cohorts of residents in a Cd-contaminated area (Mae Sot District, Tak Province) and a low-exposure area (Bangkok)[23-26]. For the Bangkok group, those aged 19 years or older were selected. The health status was ascertained by physician’s examination reports and routine blood and urinary chemistry profiles. For the Mae Sot group, those who had resided at their current addresses for 30 years or longer were selected. Exclusion criteria for both groups were pregnancy, breastfeeding, a history of metal work, and a hospital record or physician’s diagnosis of advanced chronic disease. The sociodemographic data, educational attainment, occupation, health status, family history of diabetes, and smoking status were obtained by structured interview questionnaires. Hypertension was defined as systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg[27], a physician’s diagnosis, or prescription of anti-hypertensive medications.
Measurement of exposure and adverse effects on kidneys
We used urinary excretion of Cd (ECd) as an indicator of cumulative long-term exposure to Cd or body burden[23-26], while urinary Epro and eGFR were employed as kidney function indicators[26,28]. For these measurements, samples of urine, whole blood, and plasma were collected from all participants after an overnight fast, and were stored at -80 ºC for later analysis. Plasma samples were assayed for the concentration of creatinine, while urine samples were assayed for the concentrations of creatinine, Cd, and total protein, as detailed in previous reports[23-26].
The eGFR was computed with equations of the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)[29-31]. CKD stages 1, 2, 3, 4, and 5 corresponded to eGFR of 90-119, 60-89, 30-59, 15-29, and
Normalization of Cd and protein excretion rates
The urinary excretion of x (Ex) was normalized to creatinine clearance (Ccr) as Ex/Ccr = [Cd]u[cr]p/[cr]u, where x = Cd or pro, [x]u = urine concentration of x (mass/volume), [cr]p = plasma creatinine concentration (mg/dL), and [cr]u = urine creatinine concentration (mg/dL). Ex/Ccr was expressed as an amount of x excreted per volume of the glomerular filtrate[32]. This Ccr-normalization corrects for urine dilution and the number of functioning nephrons simultaneously, and it is not influenced by muscle mass.
Ex was normalized to Ecr as [x]u/[cr]u, where x = Cd or pro, [x]u = urine concentration of x (mass/volume), and [cr]u = urine creatinine concentration (mg/dL). Ex/Ecr was expressed as an amount of x excreted per g of creatinine. This Ecr-normalization corrects for urine dilution only, but it introduces non-differential errors due to the variability in muscle mass and Ecr among people. Consequently, a clear dose-response relationship cannot be established[22].
Benchmark dose computation and definitions
We used the web-based PROAST software version 70.1 [https://proastweb.rivm.nl (accessed on 6 May 2023)] to determine the BMD values for Cd exposure as ECd/Ecr and ECd/Ccr that were linked with eGFR, EPro, prevalence of a low eGFR (CKD), and proteinuria[20,21,33]. The BMR was set at 5% and 10%[20,21,33-36].
Datasets were fitted to multiple mathematical dose–response models such as inverse exponential, natural logarithmic, exponential, and Hill models[33-35]. The Akaike information criterion (AIC) was applied to model selection and model comparison of the relative goodness of fit of different models[33]. The lower bound (BMDL) and upper bound (BMDU) of the 95% confidence interval (CI) of BMD were determined from model averaging using bootstrap with 200 repeats[36]. The BMDL could serve as a potential reference point[34,35]. The BMDU was used to calculate the BMDU/BMDL ratio, which reflects the uncertainty in the BMD estimates. The wider the BMDL-BMDU difference, the higher the statistical uncertainty in the dataset[33-36].
The BMDL value of ECd computed at 5% BMR could be considered as the NOAEL equivalent or the level of exposure below which an effect of Cd on eGFR or EPro was discernable[20,21,33]. The BMDL/BMDU values of ECd computed at 5% and 10% prevalence rates of CKD and proteinuria were designated as BMDL5/BMDU5 and BMDL10/BMDU10, respectively. The BMDL5 and BMDL10 values of ECd indicated the exposure levels at which the prevalence of CKD or proteinuria reached 5% and 10%, respectively[35]. BMDL5 could reflect a threshold level, defined as an exposure level at which 5% of the general population showed evidence of Cd-linked CKD or Cd-linked proteinuria.
Statistical analysis
Data were analyzed using IBM SPSS Statistics 21 (IBM Inc., New York, NY, USA). The Mann–Whitney U test. was used to assess the differences between males and females in mean values of continuous variables. Pearson’s chi-squared test was used to determine male-female differences in percentages and prevalences of smoking, hypertension, low eGFR, and proteinuria. The one-sample Kolmogorov–Smirnov test was used to ascertain the conformity to a normal distribution of continuous variables. Logarithmic transformation was applied to ECd and Epro, which showed a right-skewed distribution. For eGFR, no data transformation was required because the distribution of eGFR values was left-skewed. Multiple linear regression was conducted to identify factors affecting eGFR and Epro.
Logistic regression was conducted to evaluate the effects of Cd exposure and other independent variables on the prevalence odds ratio (POR) for two Cd toxicity outcomes, namely CKD and proteinuria. The POR values were adjusted for potential confounders (age, smoking, gender, and hypertension). CKD was defined as eGFR ≤ 60 mL/min/1.73 m2[31]. For Ecr-normalized data, moderate and severe proteinuria were defined as Epro/Ecr ≥ 100 mg/g creatinine and ≥ 150 mg/g creatinine, respectively[19]. For Ccr-normalized data, moderate and severe proteinuria were defined as (Epro/Ccr) × 100 ≥ 100 mg/L and ≥ 150 mg/L filtrate, respectively. A P-value ≤ 0.05 was considered statistically significant for all tests.
RESULTS AND DISCUSSION
Cohort participants
From a total of 2,000 cohort participants, 405 subjects were selected, of whom 190 and 215 were residents of Bangkok and the Mae Sot district of Tak Province, respectively [Table 1]. The overall mean age was 44.6 years, with a range from 19 to 87 years. The respective overall mean values for ECd/Ecr, eGFR, and EPro/Ecr were 5.81 µg/g creatinine, 87 mL/min/1.73 m2, and 43.8 mg/g creatinine. For the Bangkok group, the respective mean values for ECd/Ecr, eGFR, and EPro/Ecr were 0.59 µg/g creatinine, 105 mL/min/1.73 m2, and 4.73 mg/g creatinine. The corresponding mean values for the Mae Sot group were 10.43 µg/g creatinine,
Characteristics of study subjects from Bangkok residential area and Mae Sot District of Thailand
| Parameters | All subjects n = 405 | Bangkok, n = 190 | Mae Sot, n = 215 | ||
| Male n = 97 | Female n = 93 | Male n = 100 | Female n = 115 | ||
| Age, years | 44.6 (16.2) | 29.5 (5.8) | 31.5 (7.2)* | 58.0 (12.4) | 56.2 (9.7) |
| Age range, years | 19-87 | 19-44 | 19-47 | 30-87 | 40-84 |
| Smoking, % | 45.9 | 52.6 | 0.0§ | 86.0 | 42.6§ |
| Hypertension, % | 22.3 | 0.0 | 0.0 | 25.0 | 27.0 |
| [cr]p, mg/dL | 0.98 (0.29) | 0.99 (0.09) | 0.75 (0.08)§ | 1.21 (0.38) | 0.95 (0.26)§ |
| [cr]u, mg/dL | 106 (68) | 117 (79) | 66 (51)§ | 137 (55) | 102 (63)§ |
| [Cd]u, µg/L | 6.54 (10.6) | 0.61 (1.05) | 0.46 (0.47) | 14.0 (15.0) | 10.0 (9.0)* |
| [pro]u | 4.72 (15.1) | 4.64 (4.60) | 3.16 (3.49)# | 112 (236) | 62.7 (158)* |
| eGFR, mL/min/1.73 m2 | 87 (23) | 104 (11) | 107 (13) | 72 (21) | 72 (18) |
| Low eGFRa, % | 12.6 | 0.0 | 0.0 | 26.0 | 21.7 |
| Ex/Ecr normalization | |||||
| ECd/Ecr, µg/g creatinine | 5.81 (7.64) | 0.49 (0.49) | 0.69 (0.49)# | 10.0 (8.3) | 10.8 (7.8) |
| Eprot/Ecr, mg/g creatinine | 43.8 (133) | 4.12 (3.82) | 5.36 (5.94) | 88.4 (189) | 69.4 (162) |
| Eprot/Ecr ≥ 100, % | 16.3 | 0.0 | 0.0 | 18.0 | 14.8 |
| Eprot/Ecr ≥ 150, % | 13.5 | 0.0 | 0.0 | 15.0% | 12.0% |
| Ex/Ccr normalization | |||||
| µg/L filtrate | 6.21 (9.0) | 0.47 (0.42) | 0.51 (0.37) | 12.3 (10.8) | 10.4 (9.0) |
| (Epro/Ccr) × 100, mg/L filtrate | 60.2 (236) | 4.04 (3.67) | 4.11 (4.54) | 156 (428) | 69.5 (160) |
| (ECd/Ccr) × 100 ≥ 100, % | 17.7 | 0.0 | 0.0 | 21.0 | 14.8 |
| (ECd/Ccr) × 100 ≥ 150, % | 13.5 | 0.0 | 0.0 | 16.0 | 11.3 |
Environmental samples such as the paddy soils from the Mae Sot district had Cd concentrations above the standard of 0.15 mg/kg, and samples of household storage rice had Cd concentrations four times above the Thailand permissible level of 0.1 mg/kg[37]. A five-year follow-up study of the Mae Sot residents observed progressive deterioration of kidney function, evident from tubular proteinuria and eGFR endpoints, thereby suggesting these Cd effects on kidneys were irreversible[38].
In another health survey of the Mae Sot residents, a high prevalence of low eGFR of 16.1% was noted along with hypertension, proteinuria (Epro/Ecr ≥ 200 mg/g creatinine), tubular proteinuria (Eβ2M/Ecr ≥ 300 µg/g creatinine), and ECd/Ecr ≥ 2 µg/g creatinine, which were found in 32.5%, 24.1%, 36.1%, and 66.7% of the participants, respectively[39]. In the present study, the cutoff values for moderate and severe proteinuria were 100 and 150 mg/g creatinine and the overall % of low eGFR was 12.6%, while % of moderate and severe proteinuria and those who smoked and had hypertension was 16.3, 13.5, 45.9, and 22.3, respectively.
Source of “unrecognized” error in toxicological risk assessment of Cd
To evaluate the impact of normalization methods applied to Epro and ECd [Table 2], we employed two types of models: ECd was incorporated as log [(ECd/Ecr) × 103] in type A models and log [(ECd/Ccr) × 105] in type B models. All other independent variables in both types of models were identical. The POR values for moderate and severe proteinuria rose by 5% to 6% per every one-year increment of age in both types of models. Gender, smoking, and hypertension did not contribute significantly to the variation in risk of having proteinuria in any model type. In a type A model, per a 10-fold increase in ECd/Ecr, there was a significant increase in the POR for moderate proteinuria only. In comparison, the POR values for moderate and severe proteinuria increased markedly per a 10-fold rise in ECd/Ccr in a type B model. Apparently, Ecr normalization introduced an imprecision to the measurement of both exposure and effects, which predisposed the dose-response relationship between ECd/Ecr and Epro/Ecr to null[22].
Effects of Cd exposure on the POR for proteinuria
| Independent variables/factors | Moderate proteinuria | Severe proteinuria | ||
| POR (95%CI) | P | POR (95%CI) | P | |
| Model Aa | ||||
| Age, years | 1.068 (1.028, 1.110) | 0.001 | 1.065 (1.023, 1.109) | 0.002 |
| Log10[(ECd/Ecr) × 103], µg/g creatinine | 3.685 (1.027, 13.22) | 0.045 | 2.973 (0.766, 11.54) | 0.115 |
| Gender | 1.096 (0.475, 2.528) | 0.829 | 1.137 (0.467, 2.768) | 0.778 |
| Smoking | 1.678 (0.627, 4.486) | 0.303 | 1.942 (0.656, 5.753) | 0.231 |
| Hypertension | 1.113 (0.432, 2.867) | 0.824 | 1.343 (0.473, 3.810) | 0.580 |
| Model Bb | ||||
| Age, years | 1.061 (1.022, 1.102) | 0.002 | 1.051 (1.022, 1.102) | 0.018 |
| Log10[(ECd/Ccr) × 105], mg/ L filtrate | 7.143 (2.133, 23.92) | 0.001 | 10.36 (2.133, 23.92) | 0.001 |
| Gender | 1.117 (0.482, 2.587) | 0.796 | 1.204 (0.482, 2.587) | 0.695 |
| Smoking | 1.947 (0.725, 5.234) | 0.186 | 2.069 (0.725, 5.234) | 0.203 |
| Hypertension | 1.018 (0.410, 2.530) | 0.969 | 0.902 (0.410, 2.530) | 0.839 |
The imprecision of Ecr-normalization also appeared in the logistic regression models of low eGFR [Table 3]. Indeed, the impact of Ecr-normalization on the dose-response relationship between ECd/Ecr and eGFR was even more dramatic than the ECd/Ecrvs. Epro/Ecr: the POR for low eGFR was not statistically associated with ECd/Ecr (POR = 2.638, P = 0.058). In contrast, however, there was a 12.2-fold increase in the POR for low eGFR as ECd/Ccr rose 10-fold. To visualize the source of imprecision, we constructed scatterplots that relate the Cd exposure measure (ECd) to the markers of its adverse effects (Epro and eGFR). As shown in Figure 1, lower coefficients of determination (R2) were evident when ECd and Epro were adjusted to Ecr, compared with the Ccr-normalized datasets.
Figure 1. Dose-response relationships of eGFR and protein excretion rate with Cd excretion rate. Scatterplots relate (A) eGFR reduction and (B) log[(Epro/Ecr) × 103] to log[(ECd/Ecr) × 103] in all subjects. Scatterplots relate (C) eGFR reduction and (D) log[(Epro/Ccr) × 105] to log[(ECd/Ccr) × 105] in all subjects. Coefficients of determination (R2) are provided. In each graph, the line represents mean regression values. eGFR: Estimated glomerular filtration rate; Cd: cadmium.
Effects of Cd exposure on the POR for low eGFR
| Independent variables/factors | Low eGFRa | |||
| POR | 95%CI | P | ||
| Lower | Upper | |||
| Model A | ||||
| Age, years | 1.121 | 1.080 | 1.165 | < 0.001 |
| Log10[(ECd/Ecr) × 103], µg/g creatinine | 2.638 | 0.969 | 7.182 | 0.058 |
| Gender | 1.082 | 0.490 | 2.390 | 0.845 |
| Smoking | 1.425 | 0.596 | 3.406 | 0.426 |
| Hypertension | 2.211 | 1.017 | 4.805 | 0.045 |
| Model B | ||||
| Age, years | 1.118 | 1.073 | 1.165 | < 0.001 |
| Log10[(ECd/Ccr) × 105], µg/ L filtrate | 12.24 | 3.729 | 40.20 | < 0.001 |
| Gender | 0.802 | 0.346 | 1.861 | 0.608 |
| Smoking | 1.335 | 0.546 | 3.262 | 0.527 |
| Hypertension | 2.734 | 1.204 | 6.207 | 0.016 |
Erroneous conclusions on Cd effects resulted from an adjustment of ECd to Ecr, as reported in a 2016 systemic review and meta-analysis, where Cd exposure was not found to be associated with a progressive decline in eGFR[40]. Similarly, another meta-analysis reported that the association between eGFR and urinary Cd was statistically insignificant, while the risk of proteinuria rose by only 35% when comparing the highest category of Cd dose metrics with the lowest Cd exposure category[16]. However, in the latest systemic review and meta-analysis by Doccioli et al. (2024), a significant effect of Cd on eGFR was observed[41]. In the present study, we indicate that the effects of Cd exposure on both eGFR and Epro were demonstratable only when ECd and Epr were normalized to Ccr [Tables 2 and 3].
BMD of Cd exposure derived from Epro and eGFR endpoints
By BMD modeling of Epro/Ecr and ECd/Ecr [Figure 2], an exposure level of Cd at ECd/Ecr 0.0536 µg/g creatinine was the level that produced a negligible effect on protein reabsorption by kidney tubules. By the definition of BMR at 5%[33], this Cd exposure level of 0.0536 µg/g creatinine was the NOAEL for a significant increase in protein excretion. The curves for Epro/Ecr/ECd/Ecr pairs, in the order of highest to lowest model weights, were exponential model (0.6840), Hill model (0.2794), natural logarithmic model (0.0386), and inverse exponential model (0.0017).
Figure 2. BMDL and BMDU of ECd/Ecr producing a 5% increase in protein excretion. BMDL and BMDU of the 95%CI of BMD with a 5% increment of protein excretion were based on (A) exponential, (B) Hill, (C) natural logarithmic and (D) inverse exponential dose-response models. (E-G) BMDL and BMDU values were obtained by bootstrap model weighting and averaging with 200 repeats. BMDL: Benchmark dose lower; BMDU: benchmark dose upper; CI: confidence interval.
BMD modeling of the prevalence of moderate proteinuria [Figure 3] indicates that an exposure level of Cd at ECd/Ecr 1.86 µg/g creatinine was the level at which 5% of the population had moderate proteinuria
Figure 3. BMDL5 values of ECd/Ecr for proteinuria and low eGFR. Bootstrap model weighting and averaging of BMDL and BMDU bounds of the 95%CI of BMD for (A-C) 5% prevalence of proteinuria and (D-F) 5% prevalence of low eGFR. BMDL5 and BMDU5 values of ECd/
In the BMD modeling of the prevalence of CKD, defined as eGFR ≤ 60 mL/min/1.73 m2 [Figure 3D-F], an exposure level of Cd at ECd/Ecr 1.19 µg/g creatinine was the level producing 5% prevalence of CKD. By the BMDL5 definition[33], the threshold of the Cd effect on CKD risk was 1.19 µg/g creatinine. This figure was 36% lower compared to a proteinuria endpoint, thereby suggesting eGFR was more sensitive to Cd than Epro.
Figure 4 provides results of BMD modeling of Ccr-normalized data, where exposure levels of Cd at ECd/Ccr 0.0224 and 0.0152 µg/L filtrate were the levels resulting in 5% prevalences of proteinuria and 5% prevalences of CKD, respectively. Like Ecr normalized data, the eGFR endpoint appeared to be more sensitive than the proteinuria: the BMDL5 value of ECd/Ccr for low eGFR was 32% lower than the BMDL5 value of ECd/Ccr for proteinuria.
Figure 4. BMDL5 values of ECd/Ccr for proteinuria and low eGFR. Bootstrap model weighting and averaging of BMDL and BMDU bounds of the 95%CI of BMD for (A-C) 5% prevalence of protinuria and (D-F) 5% prevalence of low eGFR. BMDL and BMDU values of ECd/Ccr were based on two-stage, logarithmic logistic, Weibull, logarithmic probability, gamma, exponential and Hill dose-response models. BMDL: Benchmark dose lower; eGFR: estimated glomerular filtration rate; BMDU: benchmark dose upper; CI: confidence interval.
Comparing BMD values of ECd/Ecrvs. ECd/Ccr producing a 10% reduction in eGFR
Given that eGFR is a clinically relevant parameter and a diagnostic criterion of CKD, the results discussed above indicate the potential utility of eGFR in defining Cd exposure limits. Thus, additional BMD dose-response models for ECdvs. eGFR were generated and analyzed. As data in Figure 5 indicate, ECd/Ecr of 0.8820 µg/g creatinine was found to be the level at which there was a 10% reduction in eGFR values. The eGFR/ECd/Ecr curves fit mostly the exponential model (0.9535), followed by Hill model (0.0417) and natural logarithmic model (0.0047).
Figure 5. BMDL and BMDU of ECd/Ecr producing a 10% reduction in eGFR. BMDL and BMDU of the 95%CI of BMD with a 10% reduction in eGFR were based on (A) exponential, (B) Hill, (C) natural logarithmic and (D) inverse exponential dose-response models. (E-G) BMDL and BMDU values were obtained by bootstrap model weighting and averaging with 200 repeats. BMDL: Benchmark dose lower; BMDU: benchmark dose upper; eGFR: estimated glomerular filtration rate; CI: confidence interval.
In an equivalent BMD model of eGFR and ECd/Ccr [Figure 6], ECd/Ccr of 0.927 µg/L filtrate was found to induce a 10% reduction in eGFR. The dose-response curve for eGFR vs. ECd/Ccr totally fits the exponential model (1.0). This exponential dose-response curve implies that even a slight increase in ECd/Ccr can result in a significant fall in eGFR, thereby suggesting eGFR to be highly sensitive to Cd.
Figure 6. BMDL and BMDU of ECd/Ccr producing a 10% reduction in eGFR. BMDL and BMDU bounds of the 95%CI of BMD with a 10% reduction in eGFR were based on (A) exponential, (B) Hill, (C) inverse exponential and (D) natural logarithmic dose-response models. (E-G) BMDL and BMDU values were obtained by bootstrap model weighting and averaging with 200 repeats. BMDL: Benchmark dose lower; BMDU: benchmark dose upper; eGFR: estimated glomerular filtration rate; CI: confidence interval.
BMDL10/BMDU10 values for proteinuria and CKD prevalence data
Table 4 provides data on Cd exposure levels, measured as ECd/Ecr and ECd/Ccr, that resulted in a 10% prevalence of moderate proteinuria and a 10% prevalence of CKD. Respective ECd/Ecr values at which 10% of the population had proteinuria and 10% had CKD were 4.7 and 1.35 µg/g creatinine. The corresponding
BMDL and BMDU of Cd exposure producing a 10% prevalence of moderate proteinuria and 10% of CKD
| Parameters | 10% Prevalence of moderate proteinuria | 10% Prevalence of CKD | ||||
| BMDL10 | BMDU10 | BMDU/BMDL ratio | BMDL10 | BMDU10 | BMDU/BMDL ratio | |
| ECd/Ecr, µg/g creatinine | ||||||
| Males | 4.41 | 10.2 | 2.31 | 1.36 | 2.64 | 1.94 |
| Female | 4.05 | 10.3 | 2.54 | 1.36 | 2.34 | 1.72 |
| All subjects | 4.47 | 9.5 | 2.13 | 1.35 | 2.26 | 1.67 |
| ECd/Ccr, µg/L filtrate | ||||||
| Males | 0.0428 | 0.0981 | 2.29 | 0.0330 | 0.0710 | 2.15 |
| Female | 0.0442 | 0.0948 | 2.14 | 0.0330 | 0.0704 | 2.13 |
| All subjects | 0.0486 | 0.0883 | 1.82 | 0.0324 | 0.0614 | 1.90 |
Table 5 provides data on Cd exposure levels, measured as ECd/Ecr and ECd/Ccr, that resulted in 5% and 10% prevalence rates of severe proteinuria, respectively. The % of severe proteinuria rose from 5% to 10% as ECd/Ecr rose by 59% (from 2.30 to 5.67 µg/g creatinine). In comparison, the % of severe proteinuria also increased from 5% to 10% as ECd/Ccr rose by 44% (from 0.0314 to 0.0562 µg/L filtrate). This indicates that a smaller increase in ECd/Ccr than in ECd/Ecr produces the same effect on the prevalence of severe proteinuria. For a more precise risk assessment, Ccr normalization should be adopted.
BMDL and BMDU of Cd exposure producing 5% and 10% prevalence rates of severe proteinuria
| Parameters | 5% Prevalence of severe proteinuria | 10% Prevalence of severe proteinuria | ||||
| BMDL5 | BMDU5 | BMDU/BMDL ratio | BMDL10 | BMDU10 | BMDU/BMDL ratio | |
| ECd/Ecr, µg/g creatinine | ||||||
| Males | 2.25 | 7.64 | 3.40 | 4.77 | 14.1 | 2.96 |
| Females | 2.25 | 6.56 | 2.92 | 5.37 | 13.2 | 2.46 |
| All subjects | 2.30 | 7.13 | 3.10 | 5.67 | 12.2 | 2.15 |
| ECd/Ccr, µg/L filtrate | ||||||
| Males | 0.0236 | 0.0741 | 3.14 | 0.0512 | 0.124 | 2.42 |
| Females | 0.0268 | 0.0718 | 2.68 | 0.0527 | 0.121 | 2.30 |
| All subjects | 0.0314 | 0.0741 | 2.36 | 0.0562 | 0.114 | 2.03 |
Comparing the existing dietary exposure limits for Cd
Using the Eβ2M/Ecr ≥ 300 μg/g creatinine as a toxicity endpoint, JECFA assigned a provisional tolerable weekly intake (PTWI) level for Cd at 7 µg per kg body weight per week[4]. Later, the PTWI was amended to a tolerable monthly intake (TMI) of 25 μg per kg body weight per month (0.83 μg per kg body weight per day), and ECd/Ecr of 5.24 μg/g creatinine was assigned as a threshold level[4]. Notably, however, data in Table 5 indicate that at ECd/Ecr of 5.67 µg/g creatinine, the population prevalence of severe proteinuria due to Cd was as high as 10%. This exceeded the 5% prevalence acceptable for any environment-related disease.
The EFSA’s tolerable exposure level for Cd was 0.36 μg/kg body weight per day, and a threshold level was
Using French population data and bone toxicity as an endpoint, the Cd exposure limit of 0.35 μg Cd per kg body weight per day was derived, and the threshold level of ECd/Ecr was 0.5 μg/g creatinine[44]. In another study using the U.S. population data, Cd exposure limits were found to be between 0.21 and 0.36 μg per kg body weight per day[45]. This study from the U.S. assumed bone and kidney toxicity threshold levels to be similar (0.5 μg/g creatinine)[45]. In summary, existing dietary Cd exposure limits range between 0.21 and
Applying the BMD approach to urinary Cd excretion and diabetes prevalence data in the U.S. population, the BMDL5 and BMDL10 of ECd/Ecr values for diabetes were 0.198 and 0.365 μg/g creatinine, respectively[46]. These BMDL5 and BMDL10 of ECd/Ecr were 3.78% and 6.97% of the JECFA’s threshold level of 5.24 μg/g creatinine, respectively. In the present study, the BMDL value of ECd/Ecr (the NOAEL equivalent of ECd/Ecr) was as low as 0.0536 µg/g creatinine. This figure was obtained, if a 5% increase in Epro was a toxicity endpoint [Figure 2].
Strength, limitation, and recommendation
The strength of our study was that we were able to derive simultaneously the BMDL values of Cd excretion from two clinically relevant endpoints, proteinuria and a low eGFR. The data were obtained from subjects of varying ages and exposure levels, making them representative of the general population. The BMDL values of ECd were determined for all subjects (n = 405), males (n = 197), and females (n = 208). Additionally, we were able to identify the imprecision, which drove the dose-response relationship between eGFR and ECd to null, the phenomenon of non-differential errors, well described by Grandjean and Budtz-Jørgensen (2007)[22]. The limitations were a one-time-only assessment of Cd exposure and its effects, and its inability to determine with certainty the relative contribution of Cd from the diet and smoking to the observed outcomes.
In theory, the basic mechanism of Cd cytotoxicity should be similar across populations, as the amount of Cd causing cellular toxicity is expected to be consistent between men and women. However, when ECd was normalized to Ecr (ECd/Ecr), higher BMDL values for Cd were obtained for females compared to males. In a previous study, ECd/Ecr rates of 6.82 and 2.07 µg/g of creatinine were found to be BMDL values that produced a 5% reduction of eGFR in females and males, respectively[47]. The higher ECd/Ecr values in females are most likely due to their universally lower muscle mass and consequently lower urinary Ecr in women than men. As Ccr-normalization is unaffected by muscle mass and corrects for differences in urine dilution and functioning nephrons[32], the BMDL values for a toxic Cd level in both sexes can be expected to be identical; the BMDL value of ECd/Ccr leading to a 5% reduction in eGFR was 0.0215 µg/L of filtrate for both men and women[47].
As data in Table 5 indicate, the respective BMDL10 values of ECd that produced a 10% prevalence of severe tubular proteinuria in females and males were 5.37 and 4.77 µg/g of creatinine, if ECd and Epro were normalized to Ecr. The corresponding values obtained from the Ccr-normalized data were 0.0527 and
Current environmental exposure data suggest that a significant proportion of the general population is at risk of Cd toxicity. The main source of Cd exposure in non-smoking and non-occupationally exposed people is their diet. However, the current dietary exposure guidelines are not low enough to provide sufficient health protection. New health-protective exposure guidelines should be established, and public health measures should be developed to help minimize Cd contamination of food chains.
CONCLUSIONS
For the first time, a urinary Cd level as low as 0.0536 µg/g creatinine has been found to be the NOAEL equivalent for Cd. This “safe” body burden of Cd is derived from a 5% increase in protein excretion. The small difference between BMDU and BMDL (0.872/0.0536) implies a high degree of statistical certainty in the estimated values. Adjusting urinary Cd and biomarkers of kidney injury and malfunction to Ecr incorporates non-differential errors that bias the dose-response relationship toward the null. Adjusting urinary Cd to Ccr can eliminate such errors and imprecisions. Thus, dietary Cd exposure limits should be derived from the most sensitive endpoint and the BMDL/BMDU values calculated from Ccr-normalized data. An effective chelation therapy to remove Cd from the kidneys does not exist. Avoidance of foods containing high Cd and smoking cessation are commonsense preventive measures.
DECLARATIONS
Authors’ contributions
Conception and design of the study: Satarug S, Vesey DA
BMD modeling and interpretation: Đorđević AB, Satarug S
Data acquisition and analysis: Satarug S
Administrative, technical, and material support: Vesey DA
Availability of data and materials
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Financial support and sponsorship
None.
Conflicts of interest
All authors declared that there are no conflicts of interest.
Ethical approval and consent to participate
The present study involved a retrospective analysis of data from Thai population cohorts that were conducted following the principles outlined in the Declaration of Helsinki. Informed consent to participate in the study was obtained from participants. The Institutional Ethical Committees of Chulalongkorn University, Chiang Mai University and the Mae Sot Hospital approved the study protocol (Approval No. 142/2544, 5 October 2001).
Consent for publication
Not applicable.
Copyright
© The Author(s) 2024.
REFERENCES
2. Satarug S, Vesey DA, Gobe GC, Phelps KR. Estimation of health risks associated with dietary cadmium exposure. Arch Toxicol 2023;97:329-58.
3. Wong C, Roberts SM, Saab IN. Review of regulatory reference values and background levels for heavy metals in the human diet. Regul Toxicol Pharmacol 2022;130:105122.
4. Food and Agriculture Organization of the United Nations/World Health Organization. Summary and Conclusions. In: Proceedings of the Joint FAO/WHO Expert Committee on Food Additives and Contaminants, Seventy-Third Meeting; Geneva, Switzerland; 2011. Available from: https://apps.who.int/iris/handle/10665/44521. [Last accessed on 17 Dec 2024]
5. European Food Safety Authority (EFSA). Scientific opinion: Cadmium in food. 2009. Available from: https://efsa.onlinelibrary.wiley.com/doi/pdf/10.2903/j.efsa.2009.980. [Last accessed on 17 Dec 2024].
6. European Food Safety Authority (EFSA). Scientific opinion: Statement on tolerable weekly intake for cadmium. 2011. Available from: https://efsa.onlinelibrary.wiley.com/doi/pdf/10.2903/j.efsa.2011.1975. [Last accessed on 17 Dec 2024].
7. Satarug S, Đorđević AB, Yimthiang S, Vesey DA, Gobe GC. The NOAEL equivalent of environmental cadmium exposure associated with GFR reduction and chronic kidney disease. Toxics 2022;10:614.
8. Satarug S. Is chronic kidney disease due to cadmium exposure inevitable and can it be reversed? Biomedicines 2024;12:718.
9. Shi Z, Taylor AW, Riley M, Byles J, Liu J, Noakes M. Association between dietary patterns, cadmium intake and chronic kidney disease among adults. Clin Nutr 2018;37:276-84.
10. Kudo K, Konta T, Mashima Y, et al. The association between renal tubular damage and rapid renal deterioration in the Japanese population: the Takahata study. Clin Exp Nephrol 2011;15:235-41.
11. Satarug S, Vesey DA, Nishijo M, Ruangyuttikarn W, Gobe GC. The inverse association of glomerular function and urinary β2-MG excretion and its implications for cadmium health risk assessment. Environ Res 2019;173:40-7.
12. Wei Y, Wang X, Sun Q, et al. Associations of environmental cadmium exposure with kidney damage: exploring mediating DNA methylation sites in Chinese adults. Environ Res 2024;251:118667.
13. Sakuma S, Nogawa K, Watanabe Y, et al. Effect of renal tubular damage on non-cancer mortality in the general Japanese population living in cadmium non-polluted areas. J Appl Toxicol 2023;43:1849-58.
14. Smereczański NM, Brzóska MM. Current levels of environmental exposure to cadmium in industrialized countries as a risk factor for kidney damage in the general population: a comprehensive review of available data. Int J Mol Sci 2023;24:8413.
15. Tsai HJ, Hung CH, Wang CW, et al. Associations among heavy metals and proteinuria and chronic kidney disease. Diagnostics 2021;11:282.
16. Jalili C, Kazemi M, Cheng H, et al. Associations between exposure to heavy metals and the risk of chronic kidney disease: a systematic review and meta-analysis. Crit Rev Toxicol 2021;51:165-82.
17. Makhammajanov Z, Gaipov A, Myngbay A, Bukasov R, Aljofan M, Kanbay M. Tubular toxicity of proteinuria and the progression of chronic kidney disease. Nephrol Dial Transplant 2024;39:589-99.
18. Liu D, Lv L. New understanding on the role of proteinuria in progression of chronic kidney disease. Adv Exp Med Biol ;2019:487-500.
19. Sharma S, Smyth B. From proteinuria to fibrosis: an update on pathophysiology and treatment options. Kidney Blood Press Res 2021;46:411-20.
20. Slob W. A general theory of effect size, and its consequences for defining the benchmark response (BMR) for continuous endpoints. Crit Rev Toxicol 2017;47:342-51.
21. Moffett DB, Mumtaz MM, Sullivan DW, Whittaker MH. Chapter 13 - General considerations of dose-effect and dose-response relationships. In: Handbook on the Toxicology of Metals (Fifth Edition). Elsevier; 2022. pp. 299-317.
22. Grandjean P, Budtz-Jørgensen E. Total imprecision of exposure biomarkers: implications for calculating exposure limits. Am J Ind Med 2007;50:712-9.
23. Satarug S, Swaddiwudhipong W, Ruangyuttikarn W, Nishijo M, Ruiz P. Modeling cadmium exposures in low- and high-exposure areas in Thailand. Environ Health Perspect 2013;121:531-6.
24. Satarug S, Nishijo M, Ujjin P, Vanavanitkun Y, Moore MR. Cadmium-induced nephropathy in the development of high blood pressure. Toxicol Lett 2005;157:57-68.
25. Teeyakasem W, Nishijo M, Honda R, Satarug S, Swaddiwudhipong W, Ruangyuttikarn W. Monitoring of cadmium toxicity in a Thai population with high-level environmental exposure. Toxicol Lett 2007;169:185-95.
26. Honda R, Swaddiwudhipong W, Nishijo M, et al. Cadmium induced renal dysfunction among residents of rice farming area downstream from a zinc-mineralized belt in Thailand. Toxicol Lett 2010;198:26-32.
27. Bloch MJ, Basile JN. Review of recent literature in hypertension: updated clinical practice guidelines for chronic kidney disease now include albuminuria in the classification system. J Clin Hypertens 2013;15:865-7.
28. Satarug S, Boonprasert K, Gobe GC, et al. Chronic exposure to cadmium is associated with a marked reduction in glomerular filtration rate. Clin Kidney J 2019;12:468-75.
29. Levey AS, Stevens LA, Schmid CH, et al; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604-12.
30. White CA, Allen CM, Akbari A, et al. Comparison of the new and traditional CKD-EPI GFR estimation equations with urinary inulin clearance: a study of equation performance. Clin Chim Acta 2019;488:189-95.
31. Levey AS, Becker C, Inker LA. Glomerular filtration rate and albuminuria for detection and staging of acute and chronic kidney disease in adults: a systematic review. JAMA 2015;313:837-46.
32. Phelps KR, Gosmanova EO. A generic method for analysis of plasma concentrations. Clin Nephrol 2020;94:43-9.
33. EFSA Scientific Committee; Hardy A, Benford D, Halldorsson T, et al. Update: use of the benchmark dose approach in risk assessment. EFSA J 2017;15:e04658.
34. Slob W, Setzer RW. Shape and steepness of toxicological dose-response relationships of continuous endpoints. Crit Rev Toxicol 2014;44:270-97.
35. Sand S, Filipsson AF, Victorin K. Evaluation of the benchmark dose method for dichotomous data: model dependence and model selection. Regul Toxicol Pharmacol 2002;36:184-97.
36. Zhu Y, Wang T, Jelsovsky JZ. Bootstrap estimation of benchmark doses and confidence limits with clustered quantal data. Risk Anal 2007;27:447-65.
37. Suwatvitayakorn P, Ko MS, Kim KW, Chanpiwat P. Human health risk assessment of cadmium exposure through rice consumption in cadmium-contaminated areas of the Mae Tao sub-district, Tak, Thailand. Environ Geochem Health 2020;42:2331-44.
38. Swaddiwudhipong W, Limpatanachote P, Mahasakpan P, Krintratun S, Punta B, Funkhiew T. Progress in cadmium-related health effects in persons with high environmental exposure in northwestern Thailand: a five-year follow-up. Environ Res 2012;112:194-8.
39. Swaddiwudhipong W, Nguntra P, Kaewnate Y, et al. Human health effects from cadmium exposure: comparison between persons living in cadmium-contaminated and non-contaminated areas in northwestern Thailand. Southeast Asian J Trop Med Public Health 2015;46:133-42.
40. Byber K, Lison D, Verougstraete V, Dressel H, Hotz P. Cadmium or cadmium compounds and chronic kidney disease in workers and the general population: a systematic review. Crit Rev Toxicol 2016;46:191-240.
41. Doccioli C, Sera F, Francavilla A, Cupisti A, Biggeri A. Association of cadmium environmental exposure with chronic kidney disease: a systematic review and meta-analysis. Sci Total Environ 2024;906:167165.
42. Qing Y, Yang J, Zhu Y, et al. Dose-response evaluation of urinary cadmium and kidney injury biomarkers in Chinese residents and dietary limit standards. Environ Health 2021;20:75.
43. Liu C, Li Y, Zhu C, et al. Benchmark dose for cadmium exposure and elevated N-acetyl-β-D-glucosaminidase: a meta-analysis. Environ Sci Pollut Res Int 2016;23:20528-38.
44. Leconte S, Rousselle C, Bodin L, Clinard F, Carne G. Refinement of health-based guidance values for cadmium in the French population based on modelling. Toxicol Lett 2021;340:43-51.
45. Schaefer HR, Flannery BM, Crosby LM, et al. Reassessment of the cadmium toxicological reference value for use in human health assessments of foods. Regul Toxicol Pharmacol 2023;144:105487.
46. Shi P, Yan H, Fan X, Xi S. A benchmark dose analysis for urinary cadmium and type 2 diabetes mellitus. Environ Pollut 2021;273:116519.
Cite This Article
Open AccessHow to Cite
Satarug, S.; Vesey, D. A.; Đorđević A. B. The NOAEL equivalent for the cumulative body burden of cadmium: focus on proteinuria as an endpoint. J. Environ. Expo. Assess. 2024, 3, 26. http://dx.doi.org/10.20517/jeea.2024.31
Download Citation
Export Citation File:
Type of Import
Tips on Downloading Citation
Citation Manager File Format
Type of Import
Direct Import: When the Direct Import option is selected (the default state), a dialogue box will give you the option to Save or Open the downloaded citation data. Choosing Open will either launch your citation manager or give you a choice of applications with which to use the metadata. The Save option saves the file locally for later use.
Indirect Import: When the Indirect Import option is selected, the metadata is displayed and may be copied and pasted as needed.
About This Article
Copyright
Data & Comments
Data
0
Comments
Comments must be written in English. Spam, offensive content, impersonation, and private information will not be permitted. If any comment is reported and identified as inappropriate content by OAE staff, the comment will be removed without notice. If you have any queries or need any help, please contact us at support@oaepublish.com.