The Importance of Gene X Environment Interactions in Diabetic Nephropathy

A year ago I was working on a manuscript revision for the journal Environmental Pollution. A reviewer had objected to my characterization of a finding as a case of reverse causality. I was looking at diabetic nephropathy in Mexican Americans using the 1999-2004 National Health and Nutrition Examination Survey [1]. In our study nephropathy was defined as microalbuminuria or macroalbuminuria. The data showed an odds ratio of 14.95 (95% CI 2.96-75.48) for the fourth quartile of p,p’-DDE (dichlorodiphenyldichloroethylene) compared to less than the median. DDE is a metabolite of the pesticide DDT (dichlorodiphenyltrichloroethane), and p,p’-DDE is a common isomer of DDE. I had assumed the result had to be due to reverse causality because the odds ratio was so large. For this to be true diabetic nephropathy would have to precede a rise in p,p’-DDE concentrations. As part of the reviewer’s argument he suggested a gene X environment interaction hypothesis would be more plausible and cited Siddarth et al. [2]. Siddarth et al. studied polymorphism of the xenobiotic metabolizing enzyme glutathione S-transferase (GST) genotypes and compared p,p’-DDE in chronic kidney disease patients, that did not have diabetes, to age and sex matched healthy controls in India (N=540). The GSTM1(-)/GSTT1(-) genotype (absence of both) was associated with chronic kidney disease having an odds ratio of 1.81 (95% CI 1.08-3.03), and the third tertile of p,p’-DDE had an odds ratio of 2.70 (95% CI 1.04-7.02) for chronic kidney disease compared to the first tertile.