Are Causal Associations in Epidemiological Studies of BPA Exposure Plausible?

Saturday, February 16, 2013
Room 302 (Hynes Convention Center)
Richard M. Sharpe , University of Edinburgh, Edinburgh, United Kingdom
Recently, there have been numerous epidemiological studies published which show a significant association between current human exposure to BPA and obesity, type 2 diabetes, liver disease and cardiovascular disease; in other words the major diseases of the modern Western world. Association studies cannot prove ‘cause and effect’, but with so many studies showing similar associations, the natural tendency is to consider it likely that the association is real on the basis of ‘no smoke without fire’. However, this raises issues of plausibility. Most (>95%) human exposure to BPA occurs via our diets, and thus via our gut where most BPA (~70%) is inactivated with further inactivation occurring in the liver (the first port of call after the gut). In effect, this means that the amounts of bioactive BPA that reaches other body tissues via the bloodstream is so low that it cannot be measured. Therefore, if BPA causes the human disorders with which it has been associated, it would mean that BPA was an extremely potent toxic chemical; so potent, that if it was injected into experimental animals (thus bypassing normal breakdown/inactivation) at human exposure levels it would be expected to cause major disorders in the animals, and yet many such studies show no effects or only mild effects. Therefore, there is a plausibility issue that creates uncertainty for health and environmental regulators and for the general public. How can this uncertainty be resolved?
Virtually all of the disorders associated with BPA exposure are known to be causally linked to poor diets – usually too much sugar, fat and processed food. As already mentioned, diet is the main source of BPA. So an obvious possibility is that poorer diets are associated with higher intake of BPA and that, as a result, BPA exposure is associated with the diseases that stem from poor diets. This is plausible, in which case all that the association studies are showing is that poor diet leads to increased risk of ‘Western’ diseases, something that we already accept. Dissecting apart the contribution of diet and BPA exposure in such a situation is, however, extremely difficult, so ultimately the only way to resolve the uncertainty may be on the grounds of plausibility. For precautionary reasons, this decision may have to wait for even more studies, especially longitudinal association studies for BPA, although the confounding role of diet will remain in these studies as is the case in the presently available cross-sectional studies.