Considering the Role of the Persistent Organic Pollutants in Diabetic Nephropathy

Abstract

This journey began with the observation that cross-sectional studies of persistent organic pollutants (POPs) and diabetes have shown significant associations, but several longitudinal studies of POPs and diabetes have not [1]. POPs include dioxin, furans, polychlorinated biphenyls (PCBs), and organochlorine pesticides. An additional factor was a desire to study the full spectrum of diabetes from pre-diabetes to diabetes complications. I noted in Everett [1], that the association of dioxins, furans, and dioxin-like PCBs with diabetic nephropathy (microalbuminuria and macroalbuminuria) was significant and substantial. My initial thinking was that perhaps this relationship could be exploited in longitudinal studies. Investigating dioxin-like chemicals and diabetic nephropathy further, we came to think the relationship between POPs and diabetic nephropathy was both a case of reverse causality and a risk factor for the condition [2]. As an aside, in the revision of Everett and Thompson [2] we hit upon the use of the natural logarithm of toxic equivalency (TEQ) as a continuous measure of dioxin-like chemicals. This gave us a stable result where before [1] we had a very wide confidence interval. The odds ratio for log-transformed TEQ and diabetic nephropathy was 2.35 (95% CI 1.57-3.52), whereas the odds ratio for log-transformed TEQ and diabetes without nephropathy was 1.44 (95% CI 1.11-1.87). Thinking again about longitudinal studies of dioxin-like chemicals and diabetes, I recommended persons with nephropathy be excluded regardless of their diabetes status. Of note, I estimated 12.1% of persons with pre-diabetes had nephropathy [3]. Assuming a patient goes from pre-diabetes to incident diabetes, there would a substantial number of persons with diabetic nephropathy included if there were no exclusion criteria in such studies. Since 2013, we have also studied the relationship of organochlorine pesticides and pesticide metabolites with diabetic nephropathy. We found both dichlorodiphenyltrichloroethane (DDT) and heptachlor epoxide significantly associated with diabetic nephropathy [4]. These relationships do not appear to be cases of reverse causality. Dioxin, furans and dioxin-like PCBs act through the aryl hydrocarbon receptor pathway, and organochlorine pesticides are thought to act through the constitutive androstane receptor/pregnane X receptor pathway. These disease pathways interact, and the relationships observed for organochlorine pesticides may be due to that interaction. The proportion of the US population with elevated DDT and heptachlor epoxide was 7.2%. When both DDT and heptachlor epoxide were elevated the odds ratio for diabetic nephropathy was