In this prospective cohort study, higher serum TCDD levels in the comparison group are associated with decreased risk of being diagnosed with BPH. Serum TCDD is also associated with lower testosterone levels in both Ranch Hand and comparison veterans.
The TCDD exposure levels in the comparison group are similar to the ‘background’ exposure levels in the general population (4.22 ppt according to the 1987 NHATS) (Orban et al. 1994
). These results are consistent with the results of our previous cross-sectional study in which we found a decrease in the odds of having BPH with increasing TCDD body burden at general population exposure levels (Gupta et al. 2006
). To our knowledge, the present study is the only prospective study that has examined the association between serum TCDD and BPH.
In the present study we showed an inverse association between serum testosterone and TCDD levels. Other investigators have also reported similar results (Egeland et al. 1994
; Johnson et al. 2001
). Egeland et al. (1994)
studied 231 controls and 248 chemical production workers who were occupationally exposed to TCDD and found an inverse association between TCDD and serum testosterone. Johnson et al. (2001)
, in their study of 37 workers exposed to TCDD through spraying of herbicides, found a statistically significant inverse relationship between TCDD and testosterone in some of their analyses. The present study has the largest sample size compared to prior studies that have investigated the same hypothesis.
The strengths of our study are that it was prospective in nature and the loss to follow-up was minimized. We included two groups: the comparison veterans and the Ranch Hand veterans. The comparison veterans were exposed to the background exposure levels in the general population, whereas the Ranch Hand group was exposed to the background level plus a varying amount of TCDD through exposure to Agent Orange. This enables us to study the effects of dioxin exposure in two comparable populations with two different mechanisms of exposure. The prospective nature of this study resolves the temporal ambiguity inherent in cross-sectional studies because the serum dioxin levels were measured before the veterans were diagnosed with BPH. More than half of the participants—56% of the comparison veterans and 57% of the Ranch Hand veterans—had experienced the outcome of interest (i.e., had been diagnosed with BPH over the follow-up period). This provides our study with adequate statistical power to evaluate the relationship between serum TCDD levels and BPH. The results of the study were consistent when different exclusion criteria were applied ( and ).
Our study is limited by the fact that serum levels were measured only for TCDD and we did not have data on the levels of other dioxin and dioxin-like congeners. TCDD was the major dioxin in Agent Orange, and other dioxin-like compounds were not considered in the initial study design. In the general population, TCDD accounts for < 5% of the total dioxin toxic equivalents in the body (Schecter and Gasiewicz 2003
). The study population was predominantly composed of whites; thus, the results may not be generalized to the entire population. BPH was determined by use of medical records, which may result in some misclassification. However, any misclassification is unlikely to be differential with respect to serum TCDD levels and thus is expected to bias the study results toward the null.
BPH was assessed as a dichotomous outcome in this study. Correlating a continuous outcome measure such as prostate volume with TCDD levels is expected to be a more sensitive measure of the effect of dioxins on prostate growth. Although prostate volume measurement is invasive, it merits consideration for further studies. The data we used for assessing the relationship between testosterone and TCDD is cross-sectional in nature because both TCDD and testosterone were measured in 1987. Thus, the results represent associations and do not prove causation.
Prostatic growth in rats is sensitive to TCDD exposure (Appendix
). The mechanisms of the effect of TCDD exposure on rat prostate might help in explaining the observed association between TCDD exposure and BPH (Appendix
TCDD is also known to decrease testosterone levels in adult male rats. In adult rats exposed to TCDD, testosterone decreased in a dose-dependent fashion, and there was a dose-dependent decrease in volume per testis of Leydig cell cytoplasm, nuclei, or total Leydig cell volume (Johnson et al. 1992
). TCDD exposure also decreased the number of Leydig cells, size of individual Leydig cells and the volume per testis of smooth endoplasmic reticulum and mitochondria (Johnson et al. 1994
). Moreover, TCDD also inhibits the compensatory rise in the concentration of luteinizing hormone in plasma in response to low testosterone levels in rats (Bookstaff et al. 1990a
The difference in the results for the comparison and the Ranch Hand groups with respect to the association between serum TCDD levels and the risk of being diagnosed with BPH is surprising and not readily explainable. The results for the quartiles I–III of the Ranch Hand veterans were consistent with the results of the comparison group (). However, the quartile IV showed an increased risk that was not statistically significant when compared with the referent category, but it was statistically significant if the comparison veterans were not used as the referent group and TCDD was treated as a continuous variable (). Also, there was a statistically significant trend toward higher risk of BPH with increasing TCDD levels when certain exclusion criteria were applied (). This increased risk was confined exclusively to the TCDD quartile IV. The reason for this finding is not known. The finding may have occurred due to chance, but a few alternate explanations are also plausible.
First, the results are almost U-shaped, with a decrease in risk followed by an increasing risk on BPH. Other investigators studying endocrine-active chemicals have also noticed such results whereby the initial increase or decrease was followed by a subsequent reversal. For example, Eskenazi et al. (2005)
studied the risk of early menopause with exposure to TCDD and found a nonmonotonic dose-related association. They divided the data into quintiles based on serum TCDD levels. The risk ratio for the second, third, and fourth quintiles compared with the first quintile was 1.1, 1.4, and 1.6, respectively (test for trend, p
= 0.04); but for the fifth quintile the risk ratio was 1.1. Similarly, in another study (Markowski et al. 2001
), a curvilinear association between body weight and TCDD dose was seen in both male and female Holtzman rats; the body weight of rats exposed to lower dioxin doses (20 ng/kg and 60 ng/kg) was higher than in controls and rats exposed to a higher TCDD dose (180 ng/kg). Hormones and endocrine-disrupting chemicals are thought to have a U- or inverted U-shaped response because lower concentrations of a hormone can stimulate a tissue, whereas higher concentrations can have the opposite effect (vom Saal et al. 1997
). Mice exposed to lower concentrations of estradiol or diethylstilbestrol had higher prostate weights compared with controls and mice exposed to higher concentrations of estradiol and diethylstilbestrol (vom Saal et al. 1997
). Similarly, lower concentrations of bisphenol A (an estrogenic compound) produced greater increases in body weight and uterine weight than higher doses (Rubin et al. 2001
). Other studies have also shown similar trends (vom Saal et al. 1995
). Thus, it is possible that the effect of TCDD exposure on the human prostate follows a U-shape, whereby the initial decrease in BPH with lower doses is followed by increased occurrence of BPH at higher doses.
Second, the mechanism of exposure to TCDD differs between the comparison and Ranch Hand veterans. The comparison group was exposed to continuous background levels of dioxins, whereas the Ranch Hand group was exposed to a “bolus” of dioxins (specifically TCDD) while involved in the spraying of Agent Orange, in addition to exposure to background levels of dioxins. A possible explanation of the observed difference is that the reproductive effects of dioxins may be most pronounced when exposure occurs earlier in life. Thus, the background exposure levels at an early age may have a greater influence than a bolus TCDD exposure later. We consider the serum TCDD levels in the comparison group representative of the exposure levels experienced at a much younger age; however, the TCDD levels in the Ranch Hand group are sums of background exposure and bolus exposure from TCDD-contaminated Agent Orange. This bolus exposure may have masked the effects of the earlier background exposure and would make assessing the effects of TCDD exposure difficult. This difference in the mechanism of TCDD exposure may explain why the steady decrease in risk of BPH observed in the comparison group is not seen in the Ranch Hand veterans. Evidence from prior studies shows that age at TCDD exposure is an important determinant of the effects. The median effective dose (ED50
) of TCDD that produces decreases in testosterone and dihydrotestosterone levels in adult rats is 15 μg/kg TCDD (Moore et al. 1985
), whereas the ED50
for in utero
and lactational TCDD exposure of 0.16 μg/kg TCDD can produce a spectrum of adverse effects such as decreased weight of ventral prostate and seminal vesicles and decreased epididymal sperm numbers (Mably et al. 1992
). Hardell et al. (2003)
reported that mothers of men with testicular cancer had higher PCB levels than controls. The men themselves did not have high PCB levels. This suggests that TCDD exposure during development is more predictive of future outcomes. Further studies examining age in relation to TCDD exposure and future outcomes are needed.