Environmental Risks to Kidney Health

Environmental pollution is well recognized as a threat to public health. The United States Environmental Protection Agency, the federal agency charged with protecting human health and the environment, was established 50 years ago and now has a budget of $8.8 billion and 14,172 employees. Although having a broad focus, particular attention has been paid to the consequences of air and water pollution. The kidney is considered to be a particularly vulnerable target for water pollutants because of its high blood flow and its excretory function, but limited attention has been paid to exposure of environmental toxins as a cause of AKI and CKD (1). Heavy metals, such as lead and cadmium, have long been known to cause occupationally related kidney disease, through both inhalational and dermal exposure (2). There have also been well described environmental catastrophes related to exposure to cadmium and mercury that have resulted in nephrotoxicity. Acute exposure to toxic levels of several chemicals has also resulted in AKI. What remains uncertain is whether chronic exposure to low levels of heavy metals, agrochemicals, or industrial chemicals might lead to CKD.

In this issue of CJASN, Lee et al. (3) examine whether environmental chemicals contribute to CKD. Using an environment-wide association study (EWAS), they identified environmental chemicals associated with CKD. They used data from 46,748 adults who participated in the National Health and Nutrition Examination Survey (NHANES, 1999–2016), and who underwent measurement of 262 chemicals in urine or blood as part of their participation in NHANES. The outcome was CKD, defined as either albuminuria (urine albumin-creatinine ratio ≥30 mg/g) or reduced eGFR <60 ml/min per 1.73 m², or a composite of both. The association of each environmental chemical with CKD was tested using survey-weighted logistic regression adjusted for covariates including age, race, sex, diabetes, hypertension, body mass index, smoking, and socioeconomic status, and using meta-analysis of five discovery data sets. Potential chemical risk factors were confirmed in four validation data sets. In total, 5642 (12%) participants had an increased urine albumin-creatinine ratio and 4315 (9%) had a reduced eGFR. They found an association of albuminuria with increased serum and urine cotinines (markers of exposure to tobacco smoke), blood 2,5-dimethylfuran (a volatile organic compound that may be a marker of exposure to tobacco smoke), and blood cadmium. Increased blood lead and cadmium were associated with decreased eGFR. Increased blood cadmium and lead, and three volatile organic compounds [blood 2,5-dimethylfuran, blood furan (used in the production of other chemicals and occurring naturally in certain foods on heating), and urinary phenylglyoxylic acid (used to synthesize pharmaceutical intermediates and food additives, and as a marker of occupational exposure to toluene or styrene)] were associated with the composite outcome. Notably, increased blood levels of perfluorooctanoic acid and decreased urinary levels of 22 chemicals were associated with lower risks for CKD. As the authors note, several of these chemicals, including lead, cadmium, and furan, have been associated with CKD in other observational studies.

Like genome-wide association studies, an EWAS takes an agnostic approach to the relationship between exposure and outcomes (4). However, whereas genome-wide association study chips may include >1 million single nephron polymorphisms that include most of the genome, an EWAS, because of the requirement for individual chemical determinations, includes only a small proportion of the possible exposures. The study by Lee et al. included 262 chemicals, yet the Chemical Abstracts Service Registry contains >100 million chemicals, and in 2012 the number of industrial chemicals on the global market was estimated at 143,835 (5). In fact, several chemicals that have been associated with CKD in other observational studies, including metolachlor, paraquat, glyphosate, and diazinon, were not included in the study of Lee et al.

The investigators also note several other limitations of their analysis, most prominently the problem of reverse causation. In cross-sectional analyses of chemicals and CKD, those chemicals that are cleared by glomerular filtration will have higher blood levels and lower urinary levels in participants with decreased eGFR. Such confounding may account for the association of higher blood lead and cadmium levels with reduced eGFR, and decreased urinary levels of several heavy metals with decreased eGFR. Their analyses also do not account for the interaction between chemicals. This relationship is particularly complex because some chemicals may have toxic effects in combination, or on the basis of reaction with their solvents or adjuvants. Agrochemicals are only required to include their active ingredients on the label. Some studies have shown greater in vitro toxicity for the commercial product that includes other solvents than for the active ingredient alone (6).

Although an EWAS is a promising option for identifying potential nephrotoxins for further investigation, if a data set of previously measured chemicals such as NHANES exists, it may not be a practical technique if hundreds of new measurements of potentially toxic chemicals are required. So how do we move forward in identifying culprit environmental chemicals? Lebov et al. (7) have used a novel technique to identify potentially toxic agrochemicals. They used data from the Agricultural Health Study, a prospective cohort study of licensed pesticide applicators, linked to outcome data on kidney failure from the U.S. Renal Data Systems. The Agricultural Health Study included data on the use of 39 specific pesticides, with estimates of the frequencies and durations of exposures. Only pendimethalin was associated with kidney failure. Among the wives of pesticide applicators, who may be exposed through contact with contaminated clothing, paraquat and butylate use by their husbands was associated with an increased risk for kidney failure in the wives. Therefore, existing databases of occupational exposures by linkage to the U.S. Renal Data Systems might be useful to identify long-term exposures to chemicals that may result in CKD.

The relationship of agrochemicals with CKD has come to the forefront with identification of the entity of CKD of undetermined origin (CKDu) (8). This entity, first described in sugarcane workers in Central America, but subsequently described in miners, dock workers, and brick makers, as well as in other tropical locations such as Sri Lanka and East India, is characterized as a tubulointerstital nephritis primarily affecting younger men. The cause remains unclear, but hard physical labor in high environmental temperatures is a common thread, and recurrent subclinical heat-related kidney injury has been hypothesized as an inciting pathophysiologic cause. Nevertheless, because it has the highest prevalence in agricultural workers, agrochemicals have been suggested as a possible cause. Some investigators have suggested that, because of increased sweat losses and ingestion of large volumes of fluid, contaminated water supplies may expose the kidney to high concentrations of possible nephrotoxicants. An investigation by Smpokou et al. (9) included urinary measurements of 12 metals and metalloids, and 12 pesticides or their metabolites, but was unable to demonstrate an association with a decline of eGFR over 2 years in 350 young adults from Northwest Nicaragua who were at risk for CKDu. However, this study measured urinary levels of the metals and pesticides, and it is unclear what should be considered toxic levels of these chemicals; whether the values reflect peak exposure or the duration of exposure, and the effects of combinations of chemicals and solvents, might be more relevant. Epidemiologic studies with careful documentation of exposure history with explicit product identification may be more useful than measurement of blood levels of culprit agrochemicals, and provide a causative link between chronic exposures and CKD.

Vervaet et al. (10) have recently published another potential technique for identifying chemical nephrotoxicity. Although several biopsy series in patients with CKDu have been published, these investigators have expanded on these biopsy studies. They examined kidney biopsies in 34 individuals with CKDu and, besides the chronic interstitial nephritis noted by others, they also observed proximal tubular cell dysmorphic lysosomes that contained electron-dense nonmembrane-bound aggregates with remarkable similarities to changes seen in calcineurin nephrotoxicity. The lesion could be reproduced in laboratory rats given the calcineurin inhibitor cyclosporine in their drinking water. The investigators postulated that patients with CKDu have nephrotoxicity from a toxin that affects the same biochemical pathway as do calcineurin inhibitors. This study suggests that kidney biopsy may be useful in other patients with CKD to identify pathologic features of exposures resulting from as yet to be identified chemicals. The Kidney Precision Medicine Project, by encouraging biopsies in patients with CKD, may provide a groundwork for such an investigation.

Investigating possible nephrotoxicity related to environmental chemicals remains a daunting task. An EWAS may be useful for identifying previously unappreciated toxins but other techniques, including careful documentation of occupational exposures with longitudinal follow-up and wider use of kidney biopsies in CKD with uncertain causes, may provide evidence for potential nephrotoxicants. Meanwhile, we need to make sure that workers have appropriate protective equipment when exposed to toxic chemicals and that we all have access to clean, safe water.

Disclosures

Dr. Kaufman has nothing to disclose.

Funding

None.