An ecological study of lymphatic cancer mortality in Texas

Abstract

This study investigated environmental factors that are potentially associated with lymphatic cancer mortality in Texas. As a group, lymphatic cancers include lymphosarcoma and reticulosarcoma, Hodgkin's disease, other malignant neoplasms of lymphoid and histiocytic tissue, multiple myeloma and immunoproliferative neoplasms, and leukemia (ICD codes 200-208). An ecological study was conducted of lymphatic cancer deaths from 1986-1994 for the 194 counties in Texas. Death rates of these cancers were correlated with the pounds of Toxic Release Inventory (TRI) carcinogenic chemicals released into the air, water and land in these counties from 1988-1994 and pesticide usage from 1978 to 1992. Possible confounding factors such as manufacturing employment and median family income were analyzed. In the logistic regression model, after controlling for pesticide and income, the odds ratio for QTRI (the top quartile of pesticide releases versus the lower 3 quartiles) if 4.30. The confidence interval ranges from 1.33 to 13.85 and does not include 1.0 (the null value). QPest has an odds ratio of 1.08 with 95% confidence interval of 0.51 to 2.30. These intervals include 1.0, and the odds ratio is very close to 1.0 (no association). Income is a significant confounder, but has an odds ratio of 1.82 with confidence intervals that include 1, indicating that there is not a strong association between Income and Lymcan in this study. The two-variable model with LogTRI and Income is the most parsimonious model. LogTRI is significant at p<0.01, but the adjusted R-square for the model is only 0.04. Pesticide exposure is not significant in any model, which is inconsistent with other literature that used similar measures. A better measure of pesticide exposure is needed. These findings indicate that the top quartile of TRI release counties is associated with counties that have lymphatic cancer mortality rates above the state average rate.