Serum and Urine Heavy Metal and Fluoride Levels in Patients of Chronic Kidney Disease of Unknown Etiology of Central India: A Case-Control Study

Introduction

The term chronic kidney disease of unknown etiology (CKDu) is used to characterize CKD that cannot be linked to conventional risk factors. Globally, the reports of CKDu are rising, and in several regions of Central America, Eastern Europe, and South Asia, the disease is reaching epidemic proportions.^1-3 Several hypotheses have been proposed, such as prolonged dehydration, high levels of silica and other heavy metals in water, analgesic abuse, folk medicines, heat stress, and high pesticide use.⁴ Despite the logical foundations and some data from other parts of the world supporting many of these notions, epidemiological evidence necessary to establish a strong body of evidence is still lacking. The relationship between certain exposure to elements like lithium, arsenic, chromium, lead, mercury, and fluoride and kidney disease is well established. Whether low to moderate chronic metal and fluoride exposures contribute to CKDu is a critical public health concern.

CKDu at different endemic hotspots could be a heterogeneous disease with varied etiologies. Our center had previously documented a relatively high burden of the CKDu phenotype in certain border districts of Orissa and Chhattisgarh.⁵ A recent case series of 12 patients with CKDu from an endemic village in Chhattisgarh reported high urinary levels of chromium, manganese, and nickel in a few patients, while all patients had elevated urinary fluoride levels.⁶ However, the role of heavy metals and fluoride in the causation of CKDu in central India has not been studied systematically.

The purpose of this study was to assess and compare the levels of select heavy metals and fluoride (F) in the serum and urine of patients with CKDu to those of CKD of known etiology.

Materials and Methods

This comparative study was done at All India Institute of Medical Sciences, Raipur, a tertiary care teaching hospital in Central India, between August 2022 to July 2023. The center has previously reported clustering of cases of CKDu in the Balangir and Nuapada districts of Orissa and Gariaband and Mahasamund districts of Chhattisgarh.⁵ The study was approved by the Institute Ethics Committee (AIIMSRPR/IEC/2022/1181). Informed consent was taken from all the participants.

The study included patients diagnosed with CKD visiting the Nephrology outpatient department, in whom traditional risk factors for CKD were not present (CKDu). The comparative group included age-, sex-, and CKD stage-matched patients with CKD due to a known etiology. The patients already on dialysis were excluded from the study. CKDu was diagnosed as per the case definition published by the Indian Society of Nephrology.⁷ CKDu was defined as renal dysfunction in the absence of known/traditional risk factors and should include all progressive and asymptomatic CKD, sub-nephrotic or absent proteinuria, absence of hematuria, absence of chronic or severe hypertension, snake bite, HIV, glomerulonephritis, or any other urinary tract disease, glycosylated hemoglobin (HbA1c) <6.5%, blood pressure, <160/100 mmHg in untreated patients, 140/90 mm Hg in patients receiving ≤2 anti-hypertensive drugs, and the presence of bilateral contracted kidneys or a biopsy showing features of chronic tubulointerstitial nephritis.

Estimating a moderate effect size for serum fluoride (0.41) from the endemic area in Sri Lanka, as published by Fernando et al.,⁸ between two groups (CKDu and CKD controls) with a confidence level of 95% and power of 80%, the sample size of 99 (50 in each group) was calculated.

The Department of Nephrology runs a special CKDu clinic every Thursday. The first two patients visiting the special clinic who met the diagnostic criteria of CKDu were included in the study; 50 such patients were included in the CKDu group. At the same time, 50 age-, sex-, and CKD stage-matched patients with CKD of known etiology were recruited as controls. Socio-demographic data, medical, personal, and family history were recorded in a structured proforma. Modified Kuppuswamy classification was used for the classification of socioeconomic status.⁹

3 mL of blood was collected in trace element-free K2-EDTA and plain vacutainers; 10 mL of urine was collected in two 10 mL centrifuge tubes. The samples were stored at -80°C until analysis. Serum and urine creatinine were measured using a modified kinetic Jaffe’s method with a Random Access Fully Automated Chemistry Analyzer (Beckman Coulter). Urinary protein was estimated using the pyrogallol method with a Random Access Fully Automated Chemistry Analyzer (Beckman Coulter). HbA1c was analyzed by ion-exchange high-pressure liquid chromatography (HPLC) method using a D10 Hemoglobin testing system (BioRad Laboratories). eGFR calculation was done using the CKD-EPI equation.¹⁰

For heavy metal analysis, samples in the EDTA vacutainer were centrifuged at 3000 rpm for 10 minutes. The supernatant was separated with a micropipette. The samples were deproteinized to remove organic interference, and 5 mL each of concentrated hydrochloric (HCL) and nitric (HNO₃) acid was mixed with 0.5 mL of serum and urine. The mixture was heated to 100°C in a water bath. Following cooling, the liquid was filtered through 2.5 µ filter paper and diluted to 25 mL with deionized water.

Deproteinized samples were analyzed using an Atomic Absorption Spectrophotometer (AAS) after calibration. Concentrations of heavy metals were reported in mg/L. Iron, lead, copper, nickel, manganese, cadmium, and chromium were estimated using the LABINDIA AA8000 Atomic Absorption Spectrophotometer. Serum and urine fluoride were determined by the Ion-selective electrode method using the Orion VersaSTAR. The measurements were reported in mg/dL. The lower limit of detection for all analyzed heavy metals and fluoride was 0.0001 mg/dL. All urine and blood samples were analyzed in triplicate to ensure analytical precision and reliability. Intra-assay variability was assessed by calculating the standard deviation and coefficient of variation (CV%) across triplicate measurements for each analyte.

Results

50 CKDu patients who fulfilled the inclusion criteria were recruited. The control group comprised 50 age-, sex-, and CKD stage-matched participants. The mean age of the CKDu patients was 50.28±7.46 years; 74% (n=37) were males. The median eGFR was 14 mL/min/1.73m² (7,-27.25). Additionally, >50% of CKDu patients had CKD stage V. The control group was well-matched for age, sex, and eGFR [Table 1].

Patients in the CKDu group belonged more commonly to the rural background (70% vs. 38%, p = 0.002), were on less number of antihypertensive medications [1 (1,2) vs. 2 (2,3), p<0.001], had lower median urinary protein creatinine ratio (0.65 vs. 1.82 g/g of creatinine, p<0.001), serum potassium (4.43 ± 1.05 vs. 4.97 ± 0.99 meq/L, p=0.014), and higher serum uric acid (8.67 ± 2.35 vs. 7.139 + 2.06 mg/dL, p=0.008), serum alkaline phosphatase [133 U/L (106,173.5) vs 110 (78.5,154.5), p=0.041], and vitamin D level [26.17 ng/ml (21.23,44.31) vs. 20.22 (16.1,30.61), p=0.034]. The majority of the participants from the CKDu group were from districts Mahasamund (n = 14, 28%) and Balangir (n = 14, 28%). Whereas majority of the participants from the CKD of known etiology were from Raipur (n = 16, 26%) followed by Durg (n=11, 22%).

The most common causes of CKD in the control group were diabetic kidney disease (44%), chronic glomerulonephritis (34%), congenital anomalies of the kidney and urinary tract (CAKUT) (10%), obstructive uropathy (10%), and autosomal dominant polycystic kidney disease (ADPKD) (2%).

Table 2 shows the median serum and urine levels of select heavy metals and fluoride in the study participants. There was no significant difference between the median serum and urine levels of any of the analyzed heavy metals between the CKDu group and the control group.

Serum fluoride was significantly higher in the CKDu group as compared to the control group

[0.022 (0.012, 0.035) vs. 0.009 (0.008, 0.014) mg/dL, p= <0.001]

Discussion

In this study, we used serum and urine levels of select elements known to affect the kidney as a marker of exposure to these elements. The levels of tested heavy metals were similar between the group with CKDu and CKD of known etiology. Our findings are similar to other investigators who did not find evidence to conclude higher levels of heavy metals in patients with CKDu compared to other causes of CKD.¹¹ Atlani et al. reported no significant difference between the blood and urine levels of lead, cadmium, and chromium in CKDu and CKD of known etiology. The study, however, found the levels of heavy metals to be significantly higher in both the CKDu and CKD of known etiologies compared to healthy individuals.¹² They also reported higher levels of Arsenic in patients with CKDu compared to the other causes of CKD. However, the eGFR in the study by Atlani et al. varied significantly between the CKDu, CKD, and healthy groups. Our study has not included healthy controls, and Arsenic levels were not measured. The difference in the level of serum heavy metal between CKD, CKDu, and healthy groups could be due to low GFR.

Our study reported higher serum fluoride levels in patients with CKDu. At the usual fluoride intake of 2 mg/day, the plasma fluoride levels are known to range from 0.001 to 0.002 mg/dL,¹³ which was almost 11 times higher in our patients with CKDu. Fernando et al.¹² have also reported higher levels of serum and urine fluoride in patients with CKDu compared to endemic healthy controls and postulated a nephrotoxic role for fluoride obtained from drinking water in their CKDu cohort from Srilanka.⁸ Further, published recent reviews suggest a possible role of nephrotoxic agents and environmental elements, including fluoride, in the pathogenesis of CKDu in Sri Lanka¹⁴ [Table 3].

The mechanisms of kidney damage due to excess fluoride are mainly due to effects on kidney tubules, changes in urinary ion excretion, inhibiting the tubular reabsorption, disruption of collagen biosynthesis, and enzyme inhibition. This leads to progressive tubulointerstitial fibrosis evolving into CKD.¹⁵

Interestingly, a hygro-geochemistry study by Herojeet et al., in and around Supebeda, a village reporting a large number of CKDu cases to our center, has documented a high level of fluoride in groundwater and expressed its potential role in the development of CKD.¹⁶ While Herojeet et al. have focused on Supebeda, high levels of Fluoride in groundwater have also been reported from other districts reporting patients with CKDu to our center.¹⁷ Further, a case series of CKDu from Supebeda had reported higher urinary fluoride levels in all 12 patients and skeletal fluorosis in at least one patient.⁶ Though urine fluoride is a better indicator of high chronic fluoride intake, the values might have been lower despite having a relatively higher serum fluoride level due to deranged renal functions resulting in decreased fluoride excretion.¹⁸ Consequently, urine fluoride might not be a reliable indicator to identify chronic fluoride exposure in the presence of renal dysfunction.

Fluoride contamination in groundwater is a health problem in central India, with a newspaper report indicating that 17 districts, including the endemic districts of Mahasamund and Gariaband, have high fluoride levels.¹⁷ Even in the present study, the patients with CKDu consumed water more commonly from the non-tap water sources as compared to the participants with CKD with known etiology (72% vs. 52%, p=0.04). Non-tap water sources have been found to have higher fluoride content than tap water.¹⁹ Fluoride can readily find its way into the bodies of humans and animals through diet and water consumption. Groundwater fluoride estimation and fluoride decontamination of drinking water might be an important public health measure in these endemic areas. However, other districts that reported high fluoride content in water have not reported CKDu. This might point to the still undiscovered and complex interplay of other etiological reasons in the causation of CKDu. Further, higher serum fluoride levels may indicate accumulation due to reduced clearance in CKDu patients and making them more vulnerable to skeletal and dental fluorosis.

We also noted lower serum potassium and higher uric acid in patients with CKDu compared to age, sex, and stage of kidney disease matched patients with CKD of known etiology, which might represent predominant tubulointerstitial involvement in CKDu.

The major strength of the study was the comparison of CKDu patients with a well-matched (for age, sex, and CKD stage) control of CKD patients with known etiology, thus removing the confounding effect of renal dysfunction.

A limitation of our study was the analysis of only selected elements; other nephrotoxic elements, such as arsenic, Lithium, and Mercury, were not analyzed. Additionally, heavy metal estimation by more sensitive methods like Inductively Coupled Plasma Mass Spectrometry (ICP-MS) may detect even low levels of potentially toxic elements. Furthermore, exposure to toxic elements may be remote, and serum and urine levels may not accurately predict such exposure. Measurement of tissue levels might provide a better understanding of the exposure to these elements. Furthermore, the sample size calculation was primarily aimed at detecting differences in serum fluoride levels; therefore, they may be underpowered for other heavy metals. Also, the study lacked a healthy control group from the same geographical area that would have helped to address the confounding factors of geography, climate, and water source. As expected, the CKDu group had more participants from the rural area; the confounding effect of area of residence could not be ruled out.

In conclusion, the present study found increased serum levels of fluoride in patients with CKDu as compared to patients with other causes of CKD. Ours is a hypothesis-generating work. Further studies, addressing limitations of our study, are needed to understand the role of nephrotoxic elements in the causation of CKDu in central India.