Authors
M. Lopez-Cervantes, et al. Environmental Health Perspectives: 112:207-214, 2004.

Dichlorodiphenyltrichloroethane Burden and Breast Cancer Risk: A Meta-analysis of the Epidemiologic Evidence.

Dichlorodiphenyltrichloroethane (DDT) is possibly the best known and most useful insecticide. DDT and its main breakdown product - p,p-dichlorodiphenildichloroethylene (DDE) have received much attention because of their persistence in the environment, their ability to concentrate up the food chain, their continued wide spread detection in the food supply and in breast milk, and their ability to be stored in the adipose tissue of animals and humans. DDT use has been prohibited in most developed countries; however, it is still used for disease vector control in tropical areas. The possible contribution of DDT and DDE in increasing risk for breast cancer and its possible role as an endocrine disruptor has been reported but results remain controversial.

M. Lopes-Servantes et al. hypothesized that the previous conflicting results were due to the lack of an adequate DDE exposure gradient among breast cancer cases/controls in the individual studies and different reported measurement units of DDE body burden levels (nanograms per milliliter, nanograms per gram). The objective of the present study was to estimate the strength of the association between DDE and breast cancer on the basis of recently published epidemiologic studies and to identify the gradient of exposure among these studies.

A total of 22 published cohort and case-control studies conducted in 11 countries were accepted for analysis. Nine of these were prospective studies and 13 retrospective. Among the retrospective studies, four were population-based case-control studies and seven were clinic-based case-control studies. The number of cases in each investigation ranged from 58 to 456. Serum or breast/buttock adipose tissue samples were obtained from women from 10-25 years before the diagnosis of breast cancer to the period immediately around the date of diagnosis.

DerSimonian and Laird's method was used to correlate DDE body burden levels to breast cancer. Several sociodemographic, health and reproductive covariates were controlled for in the individual studies (age, body mass index, history of breast feeding, family breast cancer and/or benign breast disease, parity and menopausal status, vital and/or income status, physical activity, tobacco smoking and alcohol consumption, use of hormonal replacement therapy, intake of fruits, vegetables and fat). The Q-statistic was used to identify heterogeneity in the outcome variable across studies including study design, control for breast-feeding and the type of biologic specimen in which the DDT metabolites were measured. The gradient of DDE exposure in epidemiologic studies was homogenized to serum DDE levels in lipid bases (nanograms per gram).

The results of the analysis of the 22 studies showed no evidence for an association between DDE body burden levels and breast cancer risk (summary odds ratio (OR) = 0.97, 95% confidence interval (CI) = 0.87-1.09). The range of the gradient of exposure varied from 84.37 to 12928.08 ng/g. An association between DDE levels and breast cancer was not found for studies with different epidemiologic design or for studies with different biological matrices used to estimate DDE body burden. The summary OR was not different for the studies where breast-feeding was controlled as cofounder (OR=1.01; 95%CI= 0.88-1.16) compared to the studies in which breast-feeding was uncontrolled (OR=0.87; 95%CI= 0.68-1.10). No evidence of publication bias was found (p=0.253).

M. Lopez-Cervantes et al. put forward several issues in the interpretation of their results. They suggested that methodological features among the studies, lack of adjustment by dietary factors and the low estrogenicity of DDE might partially explain the contradictory results.

This study is important because it combines results from 22 large epidemiologic studies that evaluated the association between breast cancer risk and DDE body burden level. This study has some important limitations. There are several known risk factors for breast cancer including occupational, farm and household exposure of pesticides, medical conditions and therapeutic drug use that were not taken into consideration in this analysis. In addition, the percent of recovery of DDE levels was not considered in these studies, however, the differences in analytic methodology, quality control and quality assurance programs may have influenced the observed absolute values of DDE levels reported from different laboratories

Overall, the evidence for an association between the body burden of DDT, DDE level and breast cancer risk from human epidemiologic studies is equivocal. The authors propose that future epidemiological studies should take into account exposure to DDT during critical periods of early human development and possible individual variations in metabolizing enzymes of DDT and its derivatives.