The hormonal effects of long-term DDT exposure on malaria vector-control workers in Limpopo Province, South Africa.

DDT [1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane] compounds, used in many developing countries, including South Africa, for the control of malaria vectors, have been shown to be endocrine disruptors in vitro and in vivo. The study hypothesis was that male malaria vector-control workers highly exposed to DDT in the past should demonstrate clinically significant exposure-related anti-androgenic and/or estrogenic effects that should be reflected in abnormalities in reproductive hormone levels. A cross-sectional study of 50 workers from three camps situated near the Malaria Control Center (MCC) in Tzaneen was performed. Tests included blood sampling before and after a gonadotropin-releasing hormone (GnRH) challenge (100 microg). Serum o'p' and p'p' isomers of DDE, DDT, and DDD and basal and post-GnRH challenge hormone levels, including luteinizing hormone, follicle-stimulating hormone, testosterone, sex hormone-binding globulin, estradiol (E2), and inhibin, were measured. The mean number of years worked at the MCC was 15.8+/-7.8 years and the mean serum DDT was 94.3+/-57.1 microg/g of lipid. Mean baseline E2 levels (62.4+/-29.9 pg/mL) exceeded the laboratory reference range. Associations between DDT exposure measures (years worked at the MCC and DDT compounds) and hormonal outcomes were weak and inconsistent. The most important finding was a positive relationship of baseline E2 and baseline testosterone with DDT compounds, especially with p'p'-DDT and -DDD. The strongest association found, adjusted for age and SHBG, was between baseline estradiol and p'p'-DDT (beta=1.14+/-0.33 pg/mL/microg/ g lipid, P=0.001, R2=0.31, n=46). An overall anti-androgenic mechanism best explains the results, but with a number of inconsistencies. Associations might be due to chance, as multiple comparisons were made. The results therefore do not suggest an overt anti-androgenic or estrogenic effect of long-term DDT exposure on hormone levels, but correlations do exist in a manner that is not understood.