ENDOCRINE DISRUPTOR CHEMICALS - Plastics and Environmental Sustainability - page 288

µg/kg body weight/day at 3 months. While the estrogenic effects of BPA are
well known, the full impact of fetal exposure at low concentrations remains
unknown.
7.2.3 Dose-Response Relationships of BPA
Most of the controversy surrounding ED effects of BPA has to do with
the shape of the unusual dose-response curve observed in laboratory and
clinical studies. Conventional toxicological wisdom dictates biological
response (toxicity) to be an increasing linear function of the dose
administered. 4 However, with some EDCs including BPA, the
dose-response function is often nonlinear over the low-dose range of
interest(Lemosetal.,2009;Vandenberg etal.,2009;Wozniak etal.,2005).
Nonlinear dose-response curves (NLDR curves) and life-stage-specific
impacts are quite common with some hormones as well. In fact, with
hormones, it is the low-dose regime that yields the maximum physiological
effects (Welshons et al., 2003); the serum hormone levels (estradiol,
testosterone, cortisol, and thyroid hormones) range only within 0-300 pg/
ml. The low-dose activity of EDCs is therefore anticipated. For instance,
low-dose effects of BPA (such as errors in chromosomal sorting) (Hunt et
al.,2003)arenotnecessarilyobservedatthehigherdosinglevels(Welshons
et al., 2006). It is the complex multimodal action of EDCs (as well as
hormones) that results in the NLDR relationships. These include potential
cytotoxicity, receptor downregulation, receptor competition, negative
feedback loops, and tissue interaction linearity (Myers et al., 2009). The
controversy on EDCs is also due to the lack of good mechanisms to explain
NLDR curves at the present time.
Some NLDR curves can be U-shaped, with high responses at both low and
high doses (Hugo et al., 2008) or inverted U-shaped where the greatest
response is at intermediate concentrations (Jenkins et al., 2011; Newbold
et al., 2007b; Welshons et al., 2006). Figure 7.2 shows two examples of
nonlinear physiological dose-response curves for BPA. Unless the tests are
carried out at the relevant range of concentrations with long enough
durations of observation, conclusions from experimental studies will be
seriously deficient. No extrapolation or interpolation of data can be relied
upon to predict responses for EDCs.