Abstract

This study investigates the association between exposure to environmental pollutants鈥攑erchlorate, nitrate, and thiocyanate (PNT)鈥攁nd the prevalence of kidney stone disease (KSD) in US adults, utilizing data from the National Health and Nutrition Examination Survey (NHANES) 2007鈥�2018. With the increasing incidence of KSD, understanding environmental risk factors has become crucial. While PNT compounds are known to be associated with various health issues, their role in KSD remains unclear. A cross-sectional analysis of 11,350 participants measured urinary PNT levels using ion chromatography and electrospray tandem mass spectrometry, adjusting for demographic, lifestyle, and health variables. The analysis found no significant association between urinary perchlorate levels and KSD. However, participants in the highest quartile of urinary nitrate exhibited a significantly higher risk of KSD (OR 1.66, 95% CI [1.27, 2.05]) compared to those in the lowest quartile. Similarly, individuals in the second and third quartiles of urinary thiocyanate also had an elevated risk of KSD (OR 1.39, 95% CI [1.05, 1.85] and OR 1.47, 95% CI [1.08, 2.00], respectively). Restricted cubic spline analysis further demonstrated a nonlinear relationship between both urinary nitrate and thiocyanate levels and KSD risk. While the study benefits from a large, representative sample, its cross-sectional design limits the ability to establish causality, and single urine measurements may not fully capture long-term exposure. Future research should target diverse populations, employ prospective cohort studies, and conduct toxicological experiments to validate these findings and explore the underlying biological mechanisms. The study suggests a complex, non-linear relationship between elevated urinary nitrate and thiocyanate levels and an increased risk of KSD.

Introduction

Kidney stones disease (KSD) ranks among the most prevalent diseases in urology. Worldwide epidemiological studies show a consistent increase in its incidence across different areas in recent decades, now estimated to be around 8.0-12.0%, leading to substantial health and economic impacts [1,2,3]. Despite the rapid development of diagnostic and treatment methods in recent years, the recurrence rate of kidney stones remains high, with up to 50% recurrence within five years after the first surgery [4]. Various risk factors, including lifestyle, associated diseases, nutritional elements, and environmental factors, have been proven to be intimately linked with the onset and progression of kidney stones [5].

Perchlorate, nitrate, and thiocyanate (PNT), as environmental pollutants, have been extensively studied because they can interfere with the sodium-iodide symporter, inhibiting iodine uptake by the thyroid and thus adversely affecting thyroid function [6]. Recent studies have also revealed a significant relationship between high PNT levels and the risk of diseases such as metabolic syndrome, depression, and hypertension [7,8,9]. In addition, Wei Li and colleagues have explored the relationship between PNT and chronic kidney disease, uncovering that PNT is intricately linked to kidney function and might even confer potential benefits to renal health [10]. Despite these findings, the particular function of PNT in the development of kidney stones remains underexplored, with a notable gap in research regarding its correlation with kidney stone prevalence.

The National Health and Nutrition Examination Survey (NHANES) serves as a pivotal cross-sectional survey database, meticulously collecting health and nutrition data from the American household population [11]. Employing a sophisticated stratified multistage sampling method, NHANES ensures the acquisition of representative samples from the US population.

Leveraging this database, the current study provides a new perspective for a deeper understanding of the role of urinary PNT in kidney stones, seeking potential associations between urinary PNT levels and the risk of kidney stones. The insights gained from this study may help guide early prevention and treatment strategies for kidney stones.

Data and methods

Data source and study population

This cross-sectional study utilized data from NHANES, a resource managed by the National Center for Health Statistics at the Centers for Disease Control and Prevention (). Our research focused on the 2007鈥�2018 NHANES dataset, encompassing participants aged 20鈥�80 years and providing comprehensive, reliable information across various domains encompassing population characteristics, dietary and behaviors related to health, body measurements, and information on diseases. The initial participant pool of 59,842 was refined to 10,870 eligible subjects for this study. The refinement process involved screening for completeness in dataset outcomes and exposures (n鈥�=鈥�43,122), relevant covariates (n鈥�=鈥�2385), and eGFR data (n鈥�=鈥�935), as illustrated in Fig.听1.

Assessment of PNT

Urinary PNT was measured using ion chromatography coupled with electrospray tandem mass spectrometry. The chromatographic separation occurred on an Ion Pac AS16 column (Dionex, USA), utilizing sodium hydroxide as the eluent. Post-separation, the eluate underwent ionization through an electrospray interface. We quantified individual analytes by comparing their relative response factors to those of known standard concentrations, specifically the ratio of the natural analyte to its stable isotope-labeled internal standard. For comprehensive details on sample collection and processing, refer to the NHANES website. In instances where values fell below the detection limit (LOD), we applied an imputed fill value, calculated as LOD divided by the square root of 2. Due to skewed distributions in the data, we performed a log2 transformation on the PNT concentrations prior to analysis, enhancing the data鈥檚 interpretability [10].

Assessment of KSD

In the NHANES survey, the primary outcome assessed was the presence of a history of Y disease, dichotomously categorized as 鈥測es鈥� or 鈥渘o.鈥� This information was garnered through a structured questionnaire survey. Participants affirmatively responding to the query 鈥淗ave you ever had KSD?鈥� were categorized as having a history of KSD disease. This method of classification was based on self-reported data [12].

Assessment of covariates

In this study, we comprehensively adjusted for various confounding factors to elucidate the stable relationship between the risk of KSD and PNT. These factors included sociodemographic characteristics, Body Mass Index (BMI, measured in kg/m虏), history of chronic diseases, physical activity levels, cardiovascular diseases, glomerular filtration rate (eGFR), and cycle years. Given the strong association between kidney stones and socioeconomic factors, we included various demographic and sociological covariates in the analysis. These sociodemographic characteristics accounted for age, gender (male or female), race (Caucasian or Non-Caucasian), marital status (unmarried or married), annual household income (less than $20,000 or greater than $20,000), and educational level (below high school, high school graduate, or above high school). The chronic diseases history covered cardiovascular diseases, hypertension, and diabetes, with responses categorized as 鈥渘o鈥� or 鈥測es鈥�. BMI was categorized as less than 25 or greater than 25. Physical activity was assessed through moderate recreational activities and sedentary time, with the former having responses 鈥渘o鈥� or 鈥測es鈥�. Additionally, eGFR was estimated applying the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) expression, as cited in other studies [13]. Sensitivity analyses included the use of eGFR as a covariate. Cycle years were defined as the six biennial cycles from 2007 to 2008 to 2017鈥�2018, providing temporal context to the study [14, 15].

Statistical analysis

This study delineates the presentation and examination of various data types. Uninterrupted variables with normal or near-normal distribution are depicted through average values (standard error, SE) and analyzed using t-tests for two-group comparisons and one-way ANOVA for scenarios involving three or more groups. In contrast, non-normally distributed continuous variables are depicted as median values (interquartile range, IQR) and evaluated using Wilcoxon tests for two groups and Kruskal-Wallis tests for three or more groups. Categorical variables are represented as actual sample numbers (weighted percentages) and analyzed using chi-square tests to ascertain significant distribution differences.

For the analysis of PNT content, we divided these variables into quartiles and utilized a weighted binary logistic regression model. This model assesses associations between the content of these substances and the risk of KSD, presenting the results as odds ratios (OR) with 95% confidence intervals (95% CI). Variance inflation factors were employed to evaluate multicollinearity, with a value above 10 indicating substantial multicollinearity. For handling categorical variables, dummy variables were introduced.

Interaction effects in subgroup analyses were assessed using weighted likelihood ratio tests, which were based on multiplicative interaction terms for each covariate within the fully adjusted model, excluding the stratified covariate. To further ensure the robustness of the PNT model, several sensitivity analyses were performed. These included excluding individuals with eGFR鈥�<鈥�60 mL/min/1.73听m虏, adjusting for cycle years, and incorporating participants with missing or incomplete covariate data. All statistical analyses were conducted using R software (version 4.2.1, R Core Team), and a two-tailed P-value of less than 0.05 was considered statistically significant.

Results

Baseline characteristics of the participants

After excluding participants with incomplete or missing information, data from 11,350 individuals were included (Fig.听1). The relationship between the basic characteristics of the study population and the presence of kidney stones is presented in Table听1. Among the participants, 437 females (42.60%) and 663 males (57.40%) had KSD. Significant differences were observed between the groups with and without kidney stones regarding age, gender, marital status, race, comorbidities (including coronary artery disease, high blood pressure, and diabetes mellitus), and BMI. Compared to those without kidney stones, individuals with kidney stones were generally older, more likely to be of Caucasian race, predominantly married, and had a higher prevalence of hypertension, coronary heart disease and diabetes, higher BMI, and a higher proportion of smokers.

Fig. 1

A flowchart depicting the inclusion and exclusion of participants

Table 1 Characteristics of the study population according to KSD

Table听2 outlines the relevant baseline characteristics of individuals with different concentrations of urinary thiocyanate. Individuals with higher urinary thiocyanate levels, compared to those with lower levels, were typically male, of Caucasian race, no married, with lower household income and education levels. Most of them also reported smoking and alcohol consumption, along with higher BMI. Additionally, higher levels of urinary thiocyanate were observed in individuals with elevated eGFR. Furthermore, individuals with higher urinary thiocyanate levels showed a significant increase in eGFR compared to those with lower levels.

Table 2 Characteristics of the study population according to urinary thiocyanate

We also analyzed the baseline characteristics of individuals with different concentrations of urinary perchlorate (P) and nitrogen (N) (Tables S1 and S2). Individuals with higher urinary nitrogen levels, as opposed to those with lower levels, were predominantly male, older, of Caucasian race, and had a lower ratio of coronary heart disease, hypertension, and diabetes. Additionally, they were more likely to smoke and consume alcohol, with a significantly higher BMI. On the other hand, individuals with higher urinary Perchlorate levels tended to be male, married, having shorter sitting time, and had a higher BMI.

Construction and evaluation of the multifactorial logit model

The association between PNT grouping and the risk of kidney stone disease (KSD) is presented in Table听3. In the original Model 1, the highest quartile (Q4) of urinary thiocyanate (T) showed a significant association with KSD (p-value鈥�<鈥�0.01), while the middle quartiles (Q2 and Q3) of urinary nitrate (N) also exhibited significant associations (p-values鈥�&濒迟;鈥�0.05). In contrast, no significant relationship was observed between KSD and urinary chloride (P) concentrations. Therefore, KSD was significantly associated with urinary nitrate (N) and thiocyanate (T) levels but not with urinary chloride (P).

This significant association between N and T persisted across the other two models, with p-values鈥�<鈥�0.001 or <鈥�0.05 for both N and T groupings. In the multivariate logistic model for P grouping, regardless of the model, the highest quartile (Q4) of T consistently showed the highest risk for KSD when compared to the lower quartiles (Q1, Q2, Q3). In Model 1, the odds ratios (ORs) for Q1, Q2, and Q3 were: OR鈥�=鈥�1.19 (95% CI: 0.92, 1.54), OR鈥�=鈥�1.10 (95% CI: 0.84, 1.44), and OR鈥�=鈥�1.48 (95% CI: 1.16, 1.87) for Q4. Detailed results for Models 2 and 3 can be found in Table听3.

As more factors were adjusted for, the risk associated with the highest quartile of urinary thiocyanate (T) gradually increased. In the final model, the risk for KSD in Q4 was consistent with the original model, with a 6% increase in risk (Model 1: OR鈥�=鈥�1.48, 95% CI: 1.16鈥�1.87; Model 2: OR鈥�=鈥�1.52, 95% CI: 1.20鈥�1.94; Model 3: OR鈥�=鈥�1.55, 95% CI: 1.22鈥�1.97). Additionally, individuals in the highest quartile (Q4) of T had significantly higher effect sizes compared to the lower quartiles, with p-values鈥�<鈥�0.01 or <鈥�0.001. For urinary nitrate (N), the effect sizes were concentrated in the Q2-Q3 range across all three models, with p-values鈥�<鈥�0.05, showing a gradual increase in disease risk with increasing N quartiles (Model 1: P for trend鈥�=鈥�0.01; Model 2: P for trend鈥�=鈥�0.004; Model 3: P for trend鈥�=鈥�0.002).

The reverse relationship between KSD and PNT is also explored in Table听4. Across all models, KSD showed a significant association with increased risk levels of urinary T and N (p鈥�&濒迟;鈥�0.05).

Table 3 Associations between PNT and KSD based on the outcome for KSD status

Table 4 Associations between PNT and KSD based on the outcome for PNT status

Exploring the reciprocal relationship between PNT and kidney stones

We conducted subgroup analyses to explore potential associations between T, N, and the risk of kidney stones in different populations. Interestingly, results from Tables S3 and S4 indicate that there were no statistically significant differences in the risk of kidney stones between individuals with high levels of T and N and those without when stratified by age, race, diabetes, cardiovascular disease, cancer, and smoking history (p-values鈥�>鈥�0.05). No clear mutual effects between T and N were observed, possibly due to a reduction in statistical power.

We confirmed through restricted cubic spline analysis that there were significant nonlinear associations between kidney stone risk and the concentrations of P, T, and N in the NHANES cohort, with all P for nonlinear鈥�<鈥�0.001 (Fig.听2A, B, C). Changes in urinary P concentration had almost no impact on the risk of kidney stones, indicating a weak association between the two. There was a prominent J-shaped nonlinear correlation between urinary T and KSD risk (P鈥�=鈥�0.004, P overall nonlinear鈥�=鈥�0.007). When log2(thiocyanate) was less than 7.0, the correlation between urinary T and KSD risk was relatively low. A U-shaped nonlinear relationship was detected between urinary N and the risk of kidney stones (P鈥�=鈥�0.004, P overall nonlinear鈥�=鈥�0.007).

Predicting the occurrence of KSD by modeling RCS using log2 ion quantities (Fig.听2D, E, F), we observed that the risk of KSD increased with higher P concentrations, exhibiting an inverted U-shaped relationship with urinary T, and a stable and high correlation with urinary N up to a peak of approximately log2(N)鈥�=鈥�16, beyond which the correlation sharply declined.

Fig. 2

The dose-response analysis between KSD and urinary Perchlorate (A), Nitrate (B), and Thiocyanate (C) adjusted for Model 3. P showed the test for nonlinearity

The accuracy of KSD occurrence is predicted by using log2 urinary Perchlorate (D), Nitrate (E), and Thiocyanate (F) ion to model RCS.

Sensitivity analysis

To better understand the association between PNT grouping and KSD, we refined our analysis by excluding participants with an eGFR below 60 and adjusted for cyclic years in repeated modeling efforts (refer to Tables S5 and S6). The results from these models showed consistency; for example, in Main Model 3, the odds ratio (OR) for the highest quartile (Q4) of urinary thiocyanate was 1.55 (p鈥�<鈥�0.001). Sensitivity analysis provided similar findings, with ORs of 1.50 (p鈥�<鈥�0.01) and 1.54 (p鈥�<鈥�0.001). These sensitivity analyses underscore the robustness of our findings, emphasizing a significant association between levels of urinary nitrate (N) and thiocyanate (T) KSD. Notably, no significant correlation was observed with urinary chloride (P).

Discussion

This study reveals that high urinary perchlorate levels do not significantly correlate with an increased risk of KSD. Conversely, exposure to nitrate and thiocyanate appears to elevate KSD risk. Importantly, the relationship between perchlorate, nitrate, and thiocyanate with KSD risk is significantly nonlinear, while the correlation with perchlorate is relatively weaker. This indicates the combined effects of these chemicals on KSD risk are complex and warrant further investigation.

Urolithiasis, often seen as a localized manifestation of systemic disease, is closely linked with various systemic conditions, including obesity, hypertension, diabetes, and metabolic syndrome [16]. Reiner et al.鈥榮 community study further supports this association, suggesting that kidney stones are linked with subclinical atherosclerosis in middle-aged adults [17]. This connection implies shared pathogenic mechanisms between kidney stones and atherosclerosis, such as vascular damage, inflammation, and calcification. Our study鈥檚 findings corroborate these insights, revealing that individuals with a history of kidney stones typically exhibit higher age and BMI, alongside an increased prevalence of coronary heart disease, diabetes, and hypertension.

PNT, widespread environmental pollutants found in water, food, and air, impact health by affecting thyroid function [18]. Their role in regulating thyroid activity has been well-documented, with studies indicating that increased PNT concentrations could lead to reduced parathyroid hormone levels, potentially causing disorders in serum calcium and phosphate ions. Such imbalances are linked to renal dysfunction, nephrocalcinosis, and other complications [19]. King et al.鈥榮 cross-sectional study of 2441 individuals revealed that those with high levels of perchlorate and nitrate tended to be older, predominantly female, and less likely to smoke. In contrast, individuals with elevated urinary thiocyanate levels were significantly more likely to be smokers, aligning with the findings of this study [20]. Additionally, dietary nitrate, particularly from fresh fruits and vegetables, are associated with cardiovascular benefits, including improved endothelial function, reduced ischemia-reperfusion injury, and decreased arterial stiffness [21, 22]. Tang et al. highlighted the health advantages of consuming nitrate-rich vegetables or cured and processed meats [23]. This body of literature, alongside our findings, suggests that high nitrate exposure may offer protective benefits against coronary heart disease, hypertension, and diabetes.

Perchlorate, the anion of perchloric acid, finds widespread use in products like solid rocket fuel, airbag inflation systems, and crop fertilizers, with its potential toxic effects on thyroid function [24,25,26]. Furthermore, Researchers, including Soldin, indicated that there were occasional reports of bone marrow toxicity in the mid-1960s among patients treated with 1000听mg of perchlorate for hyperthyroidism [27].

Similar to perchlorate, the primary sources of nitrate exposure include drinking water and food [28]. Literature highlights that elevated nitrate levels in drinking water are associated with increased risks of methemoglobinemia, adverse pregnancy outcomes, various cancers (including ovarian and bladder cancer), thyroid enlargement, diabetes, and a significant rise in long-term cancer incidence and overall mortality [29, 30]. Research into the correlation between nitrate and cardiovascular diseases has yielded mixed results. Swedish studies have found that coronary heart disease patients treated with nitrate and phosphodiesterase-5 inhibitors face a significantly higher mortality risk [31]. Conversely, other research suggests that dietary nitrate intake might have a neutral or even negative impact on cardiovascular health [32]. Epidemiological data also link high cardiovascular risk with the occurrence of kidney stones, suggesting calcium precipitation in renal tubules and vessels as a potential pathological mechanism [33]. Therefore, high urinary nitrate levels could contribute to the development of cardiovascular diseases, subsequently elevating kidney stone risk, aligning with this study鈥檚 conclusions. Interestingly, specific nitrate like potassium nitrate are recognized for their anti-inflammatory, anti-cancer, and anti-hypercalcemia properties, while gallium nitrate is noted for its antimicrobial activity and may potentially mitigate kidney stone formation by disrupting the pathological calcification process [34, 35]. This indicates that the impact of nitrate on kidney stones may vary depending on the nitrate type.

Thiocyanate, a pervasive environmental pollutant, is found in various foods, raising concerns alongside perchlorate contamination [36]. It is notably one of the primary metabolites of nicotine and serves as a specific biochemical marker for cardiovascular system damage. Studies, including those by Leone et al., have highlighted the irreversible harm thiocyanate, especially from tobacco, can inflict on the vascular endothelium [37, 38]. The link between kidney stones and cardiovascular risk is well-documented, with KSD history also tied to an elevated risk of coronary artery disease [39]. Additionally, thiocyanate鈥檚 role as a systemic uremic toxicant suggests it could be a connecting factor between chronic kidney disease (CKD) and cardiovascular diseases (CVDs) [40]. The relationship between CKD, CVDs, and kidney stones becomes particularly concerning as CKD progresses to end-stage renal disease, exacerbating cardiovascular disease impacts [41, 42]. Thus, thiocyanate鈥檚 contribution to cardiovascular system damage through smoking not only heightens the risk of coronary heart disease and kidney stones but may also, as a toxicant in CKD, harm renal cells, further escalating kidney stone risk.

This study utilized a large, nationally representative sample from the National Health and Nutrition Examination Survey (NHANES), providing a significant advantage in accurately collecting exposure data, covariates, and kidney stone-related parameters from a diverse population. This robust dataset allowed for a detailed analysis of the relationship between PNT exposure and kidney stones. We explored the nonlinear dynamics of these associations and conducted multivariable adjustments and sensitivity analyses to confirm the stability of our findings. The methodological rigor of this study enhances the credibility of its results while offering new insights into the complex relationship between PNT exposure and kidney stone formation.

Regardless of the significant advantages of our research, it is crucial to recognize its limitations. Firstly, while the use of National Health and Nutrition Examination Survey (NHANES) data enhances the representativeness of our sample, there are inherent limitations that may affect the generalizability of our findings. As such, external validation in diverse populations is necessary to confirm our results. Secondly, the cross-sectional nature of our study precludes the determination of causality between PNT exposure and kidney stone occurrence. Additionally, the reliance on single measurements of these compounds in urine may not accurately represent long-term exposure levels of the participants. To address these issues, future epidemiological research in varied conditions or populations, alongside prospective cohort studies and toxicological experiments, is essential. Such studies will help establish causality, explore underlying biological mechanisms, and ultimately offer more targeted recommendations for kidney stone prevention.

Based on our findings, we aim to inspire the development of more comprehensive kidney stone prevention strategies that go beyond individual dietary and lifestyle changes. These strategies should also focus on reducing harmful environmental chemical exposures. For instance, public health policies promoting water purification and encouraging healthier eating habits can significantly lower population-wide exposure to nitrates and thiocyanates, which may, in turn, reduce the incidence of kidney stones. This holistic approach underscores the critical need to address both dietary and environmental factors in the prevention of kidney stone disease (KSD).

Conclusion

Our study demonstrates a significant correlation between nitrate and thiocyanate exposure and kidney stone formation. These insights are pivotal for crafting nuanced kidney stone prevention strategies that incorporate environmental exposure considerations.