Association of low cadmium and mercury exposure with chronic kidney disease among Chinese adults aged ≥80 years: A cross-sectional study : Chinese Medical Journal

Secondary Logo

Journal Logo

Original Article

Association of low cadmium and mercury exposure with chronic kidney disease among Chinese adults aged ≥80 years: A cross-sectional study

Wei, Yuan1,2; Lyu, Yuebin1; Cao, Zhaojin1; Zhao, Feng1; Liu, Yingchun1; Chen, Chen1; Li, Chengcheng1; Gu, Heng1; Lu, Feng3; Zhou, Jinhui1; Wu, Bing1,4; Liu, Yang1,5; Li, Juan2; Shi, Xiaoming1,4

Editor(s): Ji, Yuanyuan

Author Information
Chinese Medical Journal ():10.1097/CM9.0000000000002395, December 30, 2022. | DOI: 10.1097/CM9.0000000000002395

Abstract

Introduction

Chronic kidney disease (CKD) is a leading public health problem characterized by oxidative stress and deteriorated kidney functions. Increased attention has been focused on CKD since its prevalence has notably increased over the past decade, especially in developing countries.[1,2] The overall incidence of CKD in China reached 10.8% in 2009–2010, affecting approximately 119.5 million people.[3] CKD is a multifactorial disease caused by genetic and environmental factors including toxic metals.[4,5] The relationship between toxic metals and kidney damage was historically characterized by either high occupational exposure or association with communities in geographical pockets where high levels of environmental metals naturally occur.[6] However, whether toxic metals are associated with increased risk of CKD at a relatively low, ubiquitous environmental exposure level remains unknown,[7,8] and there have been very few studies focusing on associations between toxic metals and CKD among the oldest old (≥80 years old).

Cadmium (Cd) and mercury (Hg) are potent toxic heavy metals and are widespread environmental contaminants derived from both natural and anthropogenic processes. The kidney is susceptible to the effect of Cd and is one of the main target organs for Hg.[9,10] Substantial evidence suggests that exposure to Cd may cause oxidative stress, inflammation, and lipid peroxidation in kidney.[11,12] Similarly, Hg is known to be toxic even at low concentrations.[13] In addition, previous animal studies have shown that the assimilation efficiency of Hg by green mussels increased significantly following pre-exposure to Cd, presumably as a result of Hg binding with the Cd-induced metallothionein.[14] This suggests that the combined exposure of Cd and Hg may have an interactive effect on CKD.

In this study, we used data from a population-based cross-sectional study conducted in China to explore the associations of low Cd and Hg exposure with CKD among the oldest old. We investigated dose-response relationships between Cd or Hg levels and the risk of CKD respectively by means of restricted cubic spline function. Additionally, we used multiplicative interaction and additive interaction models to determine whether there were interactions between Cd and Hg concentrations in relation to CKD.

Methods

Ethical approval

This study was approved by the Ethics Committee of National Institute of Environmental Health, Chinese Center for Disease Control and Prevention (No. 201922). Informed consent was obtained from all participants or their relatives if they were unable complete the entire survey alone.

Study design and data collection

Participants were recruited for the Healthy Aging and Biomarkers Cohort Study in 2017, an ongoing perspective survey conducted in longevity areas in China involving initially 3016 older adults. We constructed two parallel populations, focused on blood or urine metal levels respectively, with different sample sizes. In Sample 1 (blood group), we excluded 440 participants due to missing data for blood Cd or blood Hg, 83 participants with missing values of serum creatinine, and 958 participants <80 years, leaving 1535 oldest old (640 males and 895 females) eligible for this study [Supplementary Figure 1, https://links.lww.com/CM9/B186]. In Sample 2 (urine group), we excluded 586 participants missing values of urine Cd and urine Hg, 287 participants missing values of serum creatinine, 131 participants missing values of urine creatinine, and 836 participants <80 years, leaving 1176 oldest old (508 males and 668 females) eligible for this study [Supplementary Figure 1, https://links.lww.com/CM9/B186].

Fasting venous blood samples (5 mL without anticoagulant) were collected from all subjects in the morning by healthcare professionals. The blood samples were centrifuged at 3000 × g for 10 min, and part of the serum was stored at −80°C and delivered to the laboratory at Capital Medical University in Beijing for serum creatinine measurement. Urinary albumin was measured using urine samples collected in the morning by the dipstick method. CKD was defined as estimated glomerular filtration rate (eGFR) <60 mL·min−1·1.73 m−2 or the presence of albuminuria in the urine.[3] Participants with the urinary albumin to creatinine ratio (ACR; mg/g cre) >30 mg/g were defined as having albuminuria. eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation.[15] Reduced renal function was defined as an eGFR of <60 mL·min−1·1.73 m−2:

eGFR = 144 × (Scr/0.7)−0.329 × (0.993)Age (if female and Scr ≤62 μmol/L)

eGFR = 144 × (Scr/0.7)−1.209 × (0.993)Age (if female and Scr >62 μmol/L)

eGFR = 141 × (Scr/0.9)−0.411 × (0.993)Age (if male and Scr ≤80 μmol/L)

eGFR = 141 × (Scr/0.9)−1.209 × (0.993)Age (if male and Scr >80 μmol/L)

where Scr is serum creatinine concentration (in mg/dL) and age is in years.

Fasting venous blood samples were collected into 5 mL heparin anticoagulant vacuum tubes and centrifuged at 3000 × g for 10 min. The plasma was isolated and stored at −20°C. Urine samples were collected into 5 mL cryotube and stored at −80°C. Urine creatinine was measured using spectrophotometry after diluted using picric acid. Plasma and urine samples were shipped to the laboratory at Capital Medical University for Cd and Hg analyses. The limits of detection of the plasma Cd, plasma Hg, urine Cd, and urine Hg were 0.07 μg/L, 0.2 μg/L, 0.06 μg/L, and 0.04 μg/L, respectively; and the number of study subjects with the levels below detection limits were 9, 222, 5, and 281, respectively; and the proportions of study subjects with the levels below detection limits were 0.6%, 14.5%, 0.4%, and 23.9%, respectively. If the concentrations were under the limits of detection, we used half of the detection limit for analysis. For our analysis, urine Cd and Hg levels were expressed as urinary Cd-to-creatinine ratio and urinary Hg-to-creatinine ratio (μg/g cre).

Demographics and lifestyle information were acquired through a structured in person questionnaire administered by trained public health practitioners from local Centers for Disease Control and Prevention. One section included sociodemographic characteristics, including age, sex, education level, marital status, residence, health insurance, smoking status, and alcohol intake status. The other section dealt with health-related characteristics; it covered measurements of triglyceride (TG), measured high-density lipoprotein cholesterol (HDLC), body mass index (BMI), and self-reported diseases—for instance, hypertension, diabetes mellitus, cardiovascular diseases, and heart disease.

Residence was categorized as urban or rural. Marital status was categorized as “not in marriage” if a participant had never married or was widowed or divorced, or “in marriage” if a participant was currently married. Current smoking status was assessed by “Do you currently smoke?” Current alcohol intake status was assessed by “Do you currently drink alcohol?” Education level was categorized as “literacy” if a participant had received >1 year of any formal education and “illiteracy” if a participant had received ≤1 year of formal education. BMI was calculated as body weight (kg) divided by squared body height (m2). Health insurance was self-reported based on the question “Do you have any social security or social insurance currently.”

Statistical analysis

The Cd and Hg measurements were divided into quartiles. The characteristics of participants were compared by analysis of variance for continuous variables and chi-squared tests for categorical variables. We used logistic regression models to estimate odds ratios (ORs) with 95% confidence intervals of CKD setting Cd and Hg as categorical variables according to the quartile distribution of Cd or Hg concentrations in blood or urine samples. In Model 1, ORs were adjusted for sex and age. In Model 2, education level, marital status, residence, health insurance, smoking status, alcohol intake status, TG, HDLC, and BMI were adjusted. In Model 3, hypertension, diabetes, cardiovascular disease, and heart disease were further added. Age, TG, HDLC, and BMI were adjusted as continuous variables in the logistic regression models, and other variables were adjusted as categorical variables in the logistic regression models. We used a multiple-imputation procedure to impute values for missing data in variables, including education level, residence, body weight, body height, marital status, TG, and HDLC.[16] Linear trend P-values were estimated by modeling the quartile of the Cd and Hg concentrations in blood or urine as continuous variables in the adjusted models. Then logistic regression with restricted cubic spline was used to characterize a dose-response relationships between Cd or Hg concentrations (continuous variable) and the risk of CKD in our study population. We used the P75 of metals concentration as cutoff value, and divided metals into two groups, respectively: high and low. ORs and P-interaction derived from the models were used to determine whether there was multiplicative interaction. The relative excess risk due to interaction (RERI), the attributable proportion (AP) due to interaction, and the synergy (S) index were used to determine whether there was an additive interaction. We further performed subgroup analyses in terms of sex (male and female); education level (illiterate and literate); and marital status (married and unmarried). We also performed three sets of sensitivity analyses to identify the robustness of our findings. First, we used recalculated eGFR utilizing the modification of diet in renal disease (MDRD) equation as the variable in the models.[17] Second, we excluded participants with hypertension, diabetes, cerebrovascular disease, or heart disease. Finally, we further adjusted vegetable consumption, fruit consumption, cancer, dementia, and respiratory disease based on Model 3.

All P-values were presented as two-sided, and results were deemed statistically significant at P < 0.05 for all analyses. We used SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) to conduct the statistical analyses.

Results

Characteristics of participants

A total of 1535 oldest old [Table 1] were included to analyze the association of blood Cd and blood Hg with CKD (Sample 1), whereas 1176 oldest old were used to analyze the association of urine Cd and urine Hg with CKD (Sample 2). Participants in Sample 1 had an average age of 91.81 ± 7.64 years with 41.69% being males. Of these, 882 (57.46%) had CKD and they were more likely to be females, live in rural areas, be illiterate, be unmarried, never smoked or drank alcohol, and had no health insurance. No significant differences were observed in BMI, HDLC, TG, residence, health insurance, smoking status, alcohol intake status, diabetes, heart disease, and cerebrovascular disease between participants with or without CKD. Participants in Sample 2 had an average age 91.33 ± 7.60 years with 43.20% being males. Of these, 696 (59.18%) had CKD and they were more likely to be females, live in rural areas, be illiterate, be unmarried, never smoked or drank alcohol, and had no health insurance. No significant differences were found in BMI, HDLC, TG, residence, health insurance, smoking status, alcohol intake status, diabetes, and heart disease between participants with or without CKD.

Table 1 - Demographic characteristics of participants for the Healthy Aging and Biomarkers Cohort Study enrolled in the study.
Sample 1 Sample 2


Variables Participant without CKD (N = 653) Participant with CKD (N = 882) Total (N = 1535) P-Value Participant without CKD (N = 480) Participant with CKD (N = 696) Total (N = 1176) P-Value
Age (years) 89.52 ± 7.20 93.50 ± 7.52 91.81 ± 7.64 <0.01 88.86 ± 7.03 93.03 ± 7.51 91.33 ± 7.60 <0.01
Sex, n (%) <0.01 <0.01
 Male 337 (51.61) 303 (34.35) 640 (41.69) 265 (55.21) 243 (34.91) 508 (43.20)
 Female 316 (48.39) 579 (65.65) 895 (58.31) 215 (44.79) 453 (65.09) 668 (56.80)
BMI (kg/m2) 21.58 ± 3.45 21.77 ± 3.92 21.69 ± 3.72 0.32 21.81 ± 3.53 21.93 ± 3.85 21.88 ± 3.72 0.61
Marital status, n (%) <0.01 <0.01
 Married 232 (35.53) 180 (20.41) 412 (26.84) 187 (38.96) 150 (21.55) 337 (28.66)
 Unmarried 421 (64.47) 702 (79.59) 1123 (73.16) 293 (61.04) 546 (78.45) 839 (71.34)
Education level, n (%) <0.01 <0.01
 Illiteracy 424 (64.93) 668 (75.74) 1092 (71.14) 298 (62.08) 517 (74.28) 815 (69.30)
 Literacy 229 (35.07) 214 (24.26) 443 (28.86) 182 (37.92) 179 (25.72) 361 (30.70)
Residence, n (%) 0.46 0.21
 Urban 155 (23.74) 224 (25.40) 379 (24.69) 114 (23.75) 188 (27.01) 302 (25.68)
 Rural 498 (76.26) 658 (74.60) 1156 (75.31) 366 (76.25) 508 (72.99) 874 (74.32)
Have health insurance, n (%) 0.14 0.01
 Yes 77 (11.79) 127 (14.40) 204 (13.29) 42 (8.75) 94 (13.51) 136 (11.56)
 No 576 (88.21) 755 (85.60) 1331 (86.71) 438 (91.25) 602 (86.49) 1040 (88.44)
Smoking status, n (%) 0.66 0.49
 Current smoker 92 (14.09) 129 (14.63) 221 (14.40) 67 (13.96) 112 (16.09) 179 (15.22)
 Former smoker 87 (13.32) 104 (11.79) 191 (12.44) 64 (13.33) 82 (11.78) 146 (12.41)
 Never smoker 474 (72.59) 649 (73.58) 1123 (73.16) 349 (72.71) 502 (72.13) 851 (72.36)
Alcohol intake status, n (%) 0.42 0.35
 Current drinker 103 (15.77) 118 (13.38) 221 (14.40) 83 (17.29) 99 (14.22) 182 (15.48)
 Former drinker 63 (9.65) 88 (9.98) 151 (9.84) 45 (9.38) 70 (10.06) 115 (9.78)
 Never drinker 487 (74.58) 676 (76.64) 1163 (75.77) 352 (73.33) 527 (75.72) 879 (74.74)
TG (mmol/L) 1.35 ± 3.51 1.47 ± 2.60 1.42 ± 3.02 0.47 1.40 ± 4.08 1.52 ± 2.91 1.47 ± 3.43 0.57
HDLC (mmol/L) 1.45 ± 0.38 1.43 ± 0.40 1.44 ± 0.39 0.23 1.46 ± 0.38 1.42 ± 0.41 1.44 ± 0.40 0.17
Hypertension, n (%) 186 (28.48) 332 (37.64) 518 (33.75) <0.01 142 (29.58) 257 (36.93) 399 (33.93) <0.01
Diabetes, n (%) 20 (3.06) 25 (2.83) 45 (2.93) 0.79 18 (3.75) 20 (2.87) 38 (3.23) 0.40
Cerebrovascular disease, n (%) 52 (7.96) 54 (6.12) 106 (6.91) 0.16 46 (9.58) 42 (6.03) 88 (7.48) 0.02
Heart disease, n (%) 57 (8.73) 81 (9.18) 138 (8.99) 0.76 36 (7.50) 69 (9.91) 105 (8.93) 0.15
Continuous variables were presented as mean ± standard deviation. Categorical variables were presented as numbers (%). The missing value for BMI was 108 in Sample 1. The missing value for BMI was 82 in Sample 2.
Used to analyze the association of blood Cd and blood Hg with CKD.
Used to analyze the association of urine Cd and urine Hg with CKD. BMI: Body mass index; CKD: Chronic kidney disease; HDLC: High-density lipoprotein cholesterol; TG: Triglyceride.

Association of cadmium and mercury with CKD

Logistic regression analysis [Table 2] and the logistic regression with restricted cubic spline [Figures 1 and 2] showed that compared with the lowest quartile, higher Cd and Hg levels were associated with increased risk of CKD. Comparing the first quartile of blood Cd, the ORs of CKD for the second, third, and fourth quartiles of blood Cd were 1.28 (0.94–1.74), 1.78 (1.30–2.44), and 1.77 (1.28–2.44). Comparing the first quartile of blood Hg, the ORs of CKD for the second, third, and fourth quartiles of blood Hg were 1.36 (1.00–1.87), 1.38 (1.01–1.88), and 1.57 (1.14–2.14). Comparing the first quartile of urine Cd, the ORs of CKD for the second, third, and fourth quartiles of urine Cd were 1.32 (0.93–1.89), 1.66 (1.15–2.39), and 2.03 (1.38–2.99). Comparing the first quartile of urine Hg, the ORs of CKD for the second, third, and fourth quartiles of urine Hg were 1.41 (0.99–2.02), 1.66 (1.15–2.38), and 1.50 (1.04–2.15) [Table 2]. The P-trends of models for blood Cd, blood Hg, urine Cd, and urine Hg were <0.01, 0.04, <0.01, and <0.01, respectively. In addition, blood Cd and urine Hg were linearly correlated with the risk of CKD in the logistic regression with restricted cubic spline (P for nolinear = 0.12 and 0.10, respectively; Figures 1 and 2). Blood Hg and urine Cd were non-linearly correlated with the risk of CKD (P for non-linear = 0.03 and 0.03, respectively), with a steeper slope at concentrations <2.30 μg/L and 3.30 μg/g cre.

Table 2 - Adjusted ORs (95% CI) for CKD according to quartiles of exposure for metals in the logistic regression models.
Quartiles of metals

Variables Q1 Q2 Q3 Q4 P-Trend
Blood Cd (μg/L) ≤0.72 0.72–1.34 1.34–2.85 >2.85 <0.01
 Model 1 1.00 1.33 (0.99–1.79) 1.85 (1.37–2.51) 1.77 (1.31–2.40)
 Model 2 1.00 1.34 (0.99–1.81) 1.78 (1.30–2.44) 1.73 (1.26–2.38)
 Model 3 1.00 1.28 (0.94–1.74) 1.78 (1.30–2.44) 1.77 (1.28–2.44)
Blood Hg (μg/L) ≤0.57 0.57–1.24 1.24–2.13 >2.13 0.04
 Model 1 1.00 1.53 (1.13–2.07) 1.54 (1.14–2.07) 1.71 (1.27–2.31)
 Model 2 1.00 1.48 (1.09–2.02) 1.50 (1.10–2.04) 1.74 (1.28–2.37)
 Model 3 1.00 1.36 (1.00–1.87) 1.38 (1.01–1.88) 1.57 (1.14–2.14)
Urine Cd (μg/g cre) ≤0.62 0.62–1.13 1.13–2.27 >2.27 <0.01
 Model 1 1.00 1.39 (0.99–1.95) 1.73 (1.22–2.44) 2.17 (1.52–3.08)
 Model 2 1.00 1.34 (0.94–1.91) 1.65 (1.15–2.37) 2.02 (1.38–2.96)
 Model 3 1.00 1.32 (0.93–1.89) 1.66 (1.15–2.39) 2.03 (1.38–2.99)
Urine Hg (μg/g cre) ≤0.08 0.08–0.35 0.35–0.76 >0.76 <0.01
 Model 1 1.00 1.44 (1.02–2.03) 1.82 (1.29–2.58) 1.62 (1.15–2.29)
 Model 2 1.00 1.43 (1.01–2.04) 1.66 (1.16–2.37) 1.52 (1.06–2.17)
 Model 3 1.00 1.41 (0.99–2.02) 1.66 (1.15–2.38) 1.50 (1.04–2.15)
Model 1: Adjusted for age and sex.Model 2: Adjusted for age, sex, education level, marital status, residence, health insurance, smoking status, alcohol intake status, TG, HDLC, and BMI.Model 3: Adjusted for age, sex, education level, marital status, residence, health insurance, smoking status, alcohol intake status, TG, HDLC, BMI, hypertension, diabetes, cardiovascular diseases, and heart disease. BMI: Body mass index; CI: Confidence interval; CKD: Chronic kidney disease; HDLC: High-density lipoprotein cholesterol; ORs: Odds ratios; TG: Triglyceride.

F1
Figure 1:
The restricted cubic spline for the associations of blood metals and CKD. The adjusted ORs were calculated by restricted cubic splines for the log-transformed concentrations of Cd (A) and Hg (B) in the model and the reference value (OR = 1) was set at the 25th percentage of each metal. The variables adjusted were age, sex, education level, marital status, residence, health insurance, smoking status, alcohol intake status TG, HDLC, BMI, hypertension, diabetes, cardiovascular diseases, and heart disease. BMI: Body mass index; CKD: Chronic kidney disease; HDLC: High-density lipoprotein cholesterol; ORs: Odds ratios; TG: Triglyceride.
F2
Figure 2:
The restricted cubic spline for the associations of urine metals and CKD. The adjusted ORs were calculated by restricted cubic splines for the log-transformed concentrations of Cd (A) and Hg (B) in the model and the reference value (OR = 1) was set at the 25th percentage of each metal. The variables adjusted were age, sex, education level, marital status, residence, health insurance, smoking status, alcohol intake status TG, HDLC, BMI, hypertension, diabetes, cardiovascular diseases, and heart disease. BMI: Body mass index; CKD: Chronic kidney disease; HDLC: High-density lipoprotein cholesterol; ORs: Odds ratios; TG: Triglyceride.

Interaction analyses

The interactions between metals and the risk of CKD are shown in Tables 3 and 4 and Supplementary Figures 2 and 3, https://links.lww.com/CM9/B186. Although high Cd levels were associated with increased risk of CKD when the Hg levels were high, there was no significant multiplicative interaction among the two metals. After adjusting for age, sex, education level, marital status, residence, health insurance, smoking status, alcohol intake status TG, HDLC, BMI, hypertension, diabetes, cardiovascular diseases, and heart disease variables, the RERI, AP, and S of blood metals were determined as 0.81 (−0.15 to 1.77), 0.41 (0.07–0.74), and 5.25 (0.38–72.51); and RERI, AP, and S of urine metals were 0.42 (−0.54 to 1.38), 0.24 (−0.23 to 0.70), and 2.18 (0.30–16.05). If there is no additive interaction, RERI and AP are equal to 0 and S is equal to 1.[18] Thus, we did not find that there was additive interaction between the levels of blood or urine Cd and Hg in relation to CKD among our study population. Even though we did not find that there was multiplicative and additive interaction between Cd and Hg in relation to CKD, higher levels of Cd and Hg were still risk factors for CKD in the oldest old. Therefore, the results of this study suggest that in clinical treatment, reducing the level of Cd and Hg in the oldest old may alleviate the occurrence and development of CKD, and provide a scientific basis for preventing the occurrence of CKD in public health.

Table 3 - Adjusted ORs (95% CI) for CKD according to the combined categories of blood metals levels.
Blood metals n (case/control) OR (95% CI) P-multiplicative interaction
Low Cd + Low Hg 498/382 1.00 0.09
High Cd + Low Hg 163/112 1.14 (0.85–1.54)
Low Cd + High Hg 151/122 1.05 (0.78–1.41)
High Cd + High Hg 70/37 2.00 (1.26–3.17)
The combined categories of blood metals levels (Low Cd ≤2.85 μg/L, High Cd >2.85 μg/L; Low Hg ≤2.13 μg/L, High Hg >2.13 μg/L) and adjusted variables (age, sex, education level, marital status, residence, health insurance, smoking status, alcohol intake status TG, HDLC, BMI, hypertension, diabetes, cardiovascular diseases, and heart disease) were included in the multivariable-adjusted model. BMI: Body mass index; CI: Confidence interval; CKD: Chronic kidney disease; HDLC: High-density lipoprotein cholesterol; ORs: Odds ratios; TG: Triglyceride.

Table 4 - Adjusted ORs (95% CI) for CKD according to the combined categories of urine metals levels.
Urine metals n (case/control) OR (95% CI) P-multiplicative interaction
Low Cd + Low Hg 395/310 1.00 0.38
High Cd + Low Hg 115/60 1.35 (0.92–1.99)
Low Cd + High Hg 100/78 0.98 (0.69–1.41)
High Cd + High Hg 86/32 1.79 (1.11–2.87)
The combined categories of urine metals levels (Low Cd ≤2.27 μg/g cre, High Cd >2.27 μg/g cre; Low Hg ≤0.76 μg/g cre, High Hg >0.76 μg/g cre) and adjusted variables (age, sex, education level, marital status, residence, health insurance, smoking status, alcohol intake status TG, HDLC, BMI, hypertension, diabetes, cardiovascular diseases, and heart disease) were included in the multivariable-adjusted model. BMI: Body mass index; CI: Confidence interval; CKD: Chronic kidney disease; HDLC: High-density lipoprotein cholesterol; ORs: Odds ratios; TG: Triglyceride.

Subgroup analysis and sensitivity analysis

Associations between blood Cd, blood Hg, urine Cd and urine Hg and the risk of CKD did not significantly differ in the analyses stratified by education level or marital status [Supplementary Tables 1–4, https://links.lww.com/CM9/B186]. Notably, for the risk of CKD there were significant interactions of blood Cd with sex, urine Cd with sex, urine Hg with sex, with more prominent associations in males than females. The results for the sensitivity analysis were similar after we recalculated eGFR level using the MDRD equation [Supplementary Table 5, https://links.lww.com/CM9/B186]. There were no significant changes observed in the results after excluding the participants with hypertension, diabetes, cerebrovascular disease, or heart disease [Supplementary Table 6, https://links.lww.com/CM9/B186]. Similarly, the associations between blood and urine Cd and Hg and the risk of CKD remained robust after further adjusted for vegetable consumption, fruit consumption, cancer, dementia, and respiratory disease [Supplementary Table 7, https://links.lww.com/CM9/B186].

Discussion

Poisoning due to heavy metals has traditionally been thought to occur only in the setting of industrial or occupational exposure. However, with the advent of more sensitive diagnostic methods and biomarkers of toxicity, it is now possible to identify tissue damage well before the onset of any significant clinical signs or symptoms, and to detect and measure much lower levels of heavy metals in blood and urine.[11] Hg and Cd have been demonstrated to be nephrotoxins at high levels of exposures (e.g., exposure occurring at smelters).[11,19] However, the impact of low level exposure (or non-industrial) on the risk of renal disease is not clear.[20] In this cross-sectional study in Chinese individuals aged >80 years, we found that low Cd and Hg exposure (or non-industrial) were associated with increased risk of CKD after adjusting for confounding factors and this association was further confirmed by a number of subgroup analyses (e.g., with the subgroups of sex, education level, and marital status). Moreover, blood Cd and urine Hg were both significantly linearly correlated with the risk of CKD in the logistic regression with restricted cubic spline model. Blood Hg and urine Cd were non-linearly correlated with the risk of CKD, with a steeper slope at concentrations <2.30 μg/L and 3.30 μg/g cre. Moreover, we did not find any significant multiplicative and additive interaction between levels of Cd and Hg in relation to CKD among our study population.

The possible explanations for our results are that Cd and Hg are prevalent environmental pollutants and nephrotoxic substances. Following chronic exposure, approximately 50% of accumulated body Cd is located in the kidneys.[21] Renal accumulation of Cd leads to reduced GFR, polyuria, and generalized tubular dysfunction (i.e., Fanconi’ s syndrome).[22,23] The earliest sign of Cd-induced kidney injury is microalbuminuria, which is usually characterized by the presence of urinary β2-microglobulin in the urine.[24] In addition, exposure to Cd can cause oxidative stress, inflammation and lipid peroxidation, which are also important causal factors of kidney injury.[11,12] Similarly, the kidney is the primary site for the accumulation of Hb and intoxication by Hg. Exposure to Hg can result in nephrotoxic effects,[25] particularly to the most susceptible proximal tubule.[26] Persistent proteinuria is one of the important indicators of CKD, and proteinuria is a clinical manifestation of Hg poisoning caused by elemental inorganic Hg and ethyl Hg.[27,28]

Notably, we found that even low Cd and Hg exposure were associated with increased risk of CKD. These primary findings were consistent with previous studies.[29-31] For example, a representative cross-sectional study demonstrated positive dose-response relationships between Cd, Hg, and urinary β2-microglobulin (β2M), one of the widely used markers of early kidney disease.[32] In addition, urine and blood Cd were found to be associated with a higher proportion of CKD in the non-institutionalized United States population in a cross-sectional nationally representative survey.[30] However, a number of previous studies showed that evidence of association of Cd and Hg exposure with CKD appeared to be biased.[6,9,20,33] In a cross-sectional representative survey in Korea, environmental low level of Hg and Cd exposure in the general population was not associated with CKD, but Cd exposure was associated with CKD in adults with hypertension or diabetes.[20] The contrasting result may be due to differences in the susceptibility to the effects of metal exposure in participants. Notably, our study explicitly focused on the oldest old, where the average age of participants was 91.81 years, and the prevalence of CKD was 57.46%.

Urinary Cd indicates long-term exposure, while blood Cd reflects more recent exposure.[34] The measurements of blood Hg directly reflect the concentration of Hg in the blood, while the level of urine Hg is affected by reabsorption and metabolism by the kidneys. The main sources of Hg exposure for environmentally exposed individuals are diet, predominately from consumption of fish containing methyl Hg.[35] Methyl Hg exposure in humans has been measured as organic Hg in blood samples, while urinalysis was used to measure excreted, inorganic Hg.[36] This study aimed to reveal associations between Cd and Hg and the risk of CKD under the long-term, low-level exposures. Therefore, in our study, to a certain extent, the analyses using urine Cd and blood Hg were more valuable to reveal the longer-term associations than blood Cd and urine Hg. Urinary concentrations of creatinine above 10 μg/g cre are evident of excessive exposure to Cd, and the cutoff for blood Hg set by the WHO is 5 μg/L.[20,30] Only a few Chinese are exposed to such high levels of Cd and Hg, although our analysis suggested that even low Cd and Hg exposure were associated with increased risk of CKD among the oldest old. Therefore, preventing low exposure of Cd and Hg to the general population has important public health significance.

This study has several strengths. First, the study participants were from the longevity areas of China where the average life expectancy is relatively high. Therefore, our study may uncover patterns of health in the oldest old who are not observed in middle-aged or young elderly, thereby having great significance for promoting the health and longevity of elderly. Second, different statistical models were used to assess the changes in the risk of CKD with increasing metal concentrations; and we used multiplicative interaction and additive interaction models to determine whether there were interactions between Cd and Hg levels in CKD. Third, the method used to estimate the GFR was more suitable for older adults, and we included proteinuria to define CKD to make the outcome definition more accurate. Finally, this study had a relatively large sample size incorporating multiple adjustments for established and potential risk factors, sensitivity analysis, and subgroup analysis.

Nonetheless, there are several limitations in our study. First, as a cross-sectional study, our analysis cannot provide evidence of cause and effect. Therefore, future cohort studies are needed to further study the relationships between Cd and Hg and the risk of CKD among the oldest old in the future. Second, blood and urine Cd and Hg levels were only measured once; and we could not ascertain if these measures were affected by medications or represented a temporary spike in levels after occasional exposure. Third, chronic disease variables adjustments for confounding factors were not medically diagnosed, and self-reported data are well known to be less reliable. Finally, we focused on the Chinese oldest old and therefore it is unknown if our results are applicable to other populations and age groups.

As far as we know, few studies have focused on the association of low Cd and Hg exposure with CKD in a community-dwelling of Chinese oldest old. We explored the linear or non-linear association of low Cd and Hg exposure with CKD using restricted cubic splines instead of using linear regression directly, resulting in a more accurate description of the results. Our research results provide a scientific basis for the formulation of environment-related public health policies and the prevention and control of CKD in Chinese oldest old. Future prospective longitudinal studies with low Cd and Hg exposure with CKD are needed to confirm the association that we found in this study, and we also plan to explore the other underlying factors affecting CKD.

In conclusion, our findings suggest that even low Cd and Hg exposure (or non-industrial) were associated with increased risk of CKD in Chinese oldest old. We did not find that there were significant multiplicative and additive interaction between levels of Cd and Hg in relation to CKD. Considering the ubiquitous nature of Cd and Hg exposure and the increasing prevalence of CKD in China, our study has important public health implications. Future prospective longitudinal studies are needed to monitor the levels of Cd and Hg and their associations with the risk of CKD to develop specific guidelines to help manage this health concern.

Funding

The study was jointly supported by the National Natural Sciences Foundation of China [No. 82025030, 82003550, 81872707, and 81941023] and the Non-profit Central Research Institute Fund of Chinese Academy of Medical Sciences [No. 2021-JKCS-028].

Conflicts of interest

None.

References

1. Couser WG, Remuzzi G, Mendis S, Tonelli M. The contribution of chronic kidney disease to the global burden of major noncommunicable diseases. Kidney Int 2011;80:1258–1270. doi: 10.1038/ki.2011.368.
2. Brück K, Stel VS, Gambaro G, Hallan S, Völzke H, Ärnlöv J, et al. CKD prevalence varies across the European general population. J Am Soc Nephrol 2016;27:2135–2147. doi: 10.1681/ASN.2015050542.
3. Zhang L, Wang F, Wang L, Wang W, Liu B, Liu J, et al. Prevalence of chronic kidney disease in China: a cross-sectional survey. Lancet 2012;379:815–822. doi: 10.1016/S0140-6736(12)60033-6.
4. Ingsathit A, Thakkinstian A, Chaiprasert A, Sangthawan P, Gojaseni P, Kiattisunthorn K, et al. Prevalence and risk factors of chronic kidney disease in the Thai adult population: Thai SEEK study. Nephrol Dial Transplant 2010;25:1567–1575. doi: 10.1093/ndt/gfp669.
5. Erdoğan C, Unlüçerçi Y, Türkmen A, Kuru A, Cetin O, Bekpinar S. The evaluation of oxidative stress in patients with chronic renal failure. Clin Chim Acta 2002;322:157–161. doi: 10.1016/s0009-8981(02)00173-0.
6. Moody EC, Coca SG, Sanders AP. Toxic metals and chronic kidney disease: a systematic review of recent literature. Curr Environ Health Rep 2018;5:453–463. doi: 10.1007/s40572-018-0212-1.
7. Nigra AE, Sanchez TR, Nachman KE, Harvey D, Chillrud SN, Graziano JH, et al. The effect of the Environmental Protection Agency maximum contaminant level on arsenic exposure in the USA from 2003 to 2014: an analysis of the National Health and Nutrition Examination Survey (NHANES). Lancet Public Health 2017;2:e513–e521. doi: 10.1016/S2468-2667(17)30195-0.
8. Tsoi MF, Cheung CL, Cheung TT, Cheung BMY. Continual decrease in blood lead level in Americans: United States National Health Nutrition and Examination Survey 1999-2014. Am J Med 2016;129:1213–1218. doi: 10.1016/j.amjmed.2016.05.042.
9. 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. doi: 10.3109/10408444.2015.1076375.
10. Afrifa J, Essien Baidoo S, Ephraim RKD, Nkrumah D, Dankyira DO. Reduced EGFR, elevated urine protein and low level of personal protective equipment compliance among artisanal small scale gold miners at Bibiani-Ghana: a cross-sectional study. BMC Public Health 2017;17:601. doi: 10.1186/s12889-017-4517-z.
11. Johri N, Jacquillet G, Unwin R. Heavy metal poisoning: the effects of cadmium on the kidney. Biometals 2010;23:783–792. doi: 10.1007/s10534-010-9328-y.
12. Prozialeck WC, Edwards JR, Woods JM. The vascular endothelium as a target of cadmium toxicity. Life Sci 2006;79:1493–1506. doi: 10.1016/j.lfs.2006.05.007.
13. Zahir F, Rizwi SJ, Haq SK, Khan RH. Low dose mercury toxicity and human health. Environ Toxicol Pharmacol 2005;20:351–360. doi: 10.1016/j.etap.2005.03.007.
14. Shi D, Wang WX. Modification of trace metal accumulation in the green mussel Perna viridis by exposure to Ag, Cu, and Zn. Environ Pollut 2004;132:265–277. doi: 10.1016/j.envpol.2004.04.023.
15. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–612. doi: 10.7326/0003-4819-150-9-200905050-00006.
16. Zhou H, Elliott MR, Raghunathan TE. Synthetic multiple-imputation procedure for multistage complex samples. J Off Stat 2016;32:231–256. doi: 10.1515/JOS-2016-0011.
17. Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S, et al. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med 2006;145:247–254. doi: 10.7326/0003-4819-145-4-200608150-00004.
18. Andersson T, Alfredsson L, Källberg H, Zdravkovic S, Ahlbom A. Calculating measures of biological interaction. Eur J Epidemiol 2005;20:575–579. doi: 10.1007/s10654-005-7835-x.
19. Soderland P, Lovekar S, Weiner DE, Brooks DR, Kaufman JS. Chronic kidney disease associated with environmental toxins and exposures. Adv Chronic Kidney Dis 2010;17:254–264. doi: 10.1053/j.ackd.2010.03.011.
20. Kim NH, Hyun YY, Lee KB, Chang Y, Ryu S, Oh KH, et al. Environmental heavy metal exposure and chronic kidney disease in the general population. J Korean Med Sci 2015;30:272–277. doi: 10.3346/jkms.2015.30.3.272.
21. Thévenod F. Nephrotoxicity and the proximal tubule. Insights from cadmium. Nephron Physiol 2003;93:87–93. doi: 10.1159/000070241.
22. Nordberg M, Jin T, Nordberg GF. Cadmium, metallothionein and renal tubular toxicity. IARC Sci Publ 1992;118:293–297.
23. Felley Bosco E, Diezi J. Fate of cadmium in rat renal tubules: a microinjection study. Toxicol Appl Pharmacol 1987;91:204–211. doi: 10.1016/0041-008x(87)90101-3.
24. Tsuchiya K, Iwao S, Sugita M, Sakurai H. Increased urinary beta 2-microglobulin in cadmium exposure: dose-effect relationship and biological significance of beta 2-microglobulin. Environ Health Perspect 1979;28:147–153. doi: 10.1289/ehp.7928147.
25. Orr SE, Bridges CC. Chronic kidney disease and exposure to nephrotoxic metals. Int J Mol Sci 2017;18:1039. doi: 10.3390/ijms18051039.
26. Zalups RK, Lash LH. Advances in understanding the renal transport and toxicity of mercury. J Toxicol Environ Health 1994;42:1–44. doi: 10.1080/15287399409531861.
27. Park JD, Zheng W. Human exposure and health effects of inorganic and elemental mercury. J Prev Med Public Health 2012;45:344–352. doi: 10.3961/jpmph.2012.45.6.344.
28. Devillé WLJM, Yzermans JC, van Duijn NP, Bezemer PD, van der Windt DAWM, Bouter LM. The urine dipstick test useful to rule out infections. A meta-analysis of the accuracy. BMC Urol 2004;4:4. doi: 10.1186/1471-2490-4-4.
29. Edwards JR, Prozialeck WC. Cadmium, diabetes and chronic kidney disease. Toxicol Appl Pharmacol 2009;238:289–293. doi: 10.1016/j.taap.2009.03.007.
30. Ferraro PM, Costanzi S, Naticchia A, Sturniolo A, Gambaro G. Low level exposure to cadmium increases the risk of chronic kidney disease: analysis of the NHANES 1999-2006. BMC Public Health 2010;10:304. doi: 10.1186/1471-2458-10-304.
31. 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;37:276–284. doi: 10.1016/j.clnu.2016.12.025.
32. Kim YD, Yim DH, Eom SY, Moon SI, Park CH, Kim GB, et al. Temporal changes in urinary levels of cadmium, N-acetyl-(-d-glucosaminidase and (2-microglobulin in individuals in a cadmium-contaminated area. Environ Toxicol Pharmacol 2015;39:35–41. doi: 10.1016/j.etap.2014.10.016.
33. Yang F, Yi X, Guo J, Xu S, Xiao Y, Huang X, et al. Association of plasma and urine metals levels with kidney function: a population-based cross-sectional study in China. Chemosphere 2019;226:321–328. doi: 10.1016/j.chemosphere.2019.03.171.
34. Adams SV, Newcomb PA. Cadmium blood and urine concentrations as measures of exposure: NHANES 1999-2010. J Expo Sci Environ Epidemiol 2014;24:163–170. doi: 10.1038/jes.2013.55.
35. Akerstrom M, Barregard L, Lundh T, Sallsten G. Relationship between mercury in kidney, blood, and urine in environmentally exposed individuals, and implications for biomonitoring. Toxicol Appl Pharmacol 2017;320:17–25. doi: 10.1016/j.taap.2017.02.007.
36. Chen G, Chen X, Yan C, Wu X, Zeng G. Surveying mercury levels in hair, blood and urine of under 7-year old children from a coastal city in China. Int J Environ Res Public Health 2014;11:12029–12041. doi: 10.3390/ijerph111112029.
Keywords:

Aged, 80 and over; Cadmium; Chronic kidney disease; Mercury

Supplemental Digital Content

Copyright © 2022 The Chinese Medical Association, produced by Wolters Kluwer, Inc. under the CC-BY-NC-ND license.