Bisphenol A and Diabetes, Insulin Resistance, Cardiovascular Disease and Obesity: Controversy in a (Plastic) Cup?

Abstract

One of the more notable public health debates in recent times relates to the purported association between the ubiquitous environmental chemical bisphenol-A (BPA) and a host of chronic diseases. The past 5 years have seen a rapid increase in the evidence linking BPA to obesity, diabetes, and cardiovascular disease in human, population-based epidemiological studies. However, 2 recent reports (1, 2) in the JCEM, while both showing modest associations between BPA (measured in urine) and chronic disease, have achieved little in definitively addressing the controversy. Shankar and Teppala (1), using National Health and Nutritional Examination Survey (NHANES) data, showed a positive association between BPA levels and diabetes, and Wang et al (2) report associations between urinary BPA and obesity and insulin resistance in a cohort of adults aged 40 years or older in China. However, the interpretation of these findings is limited by many caveats that relate to BPA metabolism and its analysis, both in the testing methods currently available and with respect to appropriate adjustment for confounders. There is little doubt that most in vitro and animal data support a relationship existing between BPA and adverse physiological effects in humans. In vitro studies by Hugo et al (3) using BPA at environmentally relevant levels have shown that BPA inhibits adiponectin, a key adipokine that increases insulin sensitivity and decreases tissue inflammation. This study provides direct evidence of the health effects of BPA in human tissue. In addition, AlonsoMagdalena et al (4) showed that mice exposed long term to BPA developed hyperinsulinemia, insulin resistance, and glucose intolerance. Skeptics argue that the animal models are too dissimilar to humans to extrapolate the findings with any confidence, not to mention any certainty. But evidence in this regard is becoming more robust: recent reports demonstrate that BPA metabolism toxicokinetics are very similar in humans, monkeys, and mice and suggest also that human exposure is greater than previously estimated (5). These studies achieved plasma levels in monkeys equivalent to those found on many human biomonitoring studies of .3–4.0 ng/mL. The amount of BPA needed to achieve the serum concentrations in monkeys far exceeded the 2007 U.S. Food and Drug Administration human exposure estimate of .16 g/ kg/d as well as the U.S. Environmental Protection Agency’s daily intake dose of 50 g/kg (5), an indication that human exposure is probably underestimated. It is important to note, however, that exposure to BPA is ubiquitous and apart from food and beverage containers, BPA has been detected in dust and air particles, dental sealants, thermal papers, and even water (6). If we look critically at the human data, there are aspects that warrant mention. Much of the data that have shown a positive relationship between BPA and adverse effects in humans have come from one population, the NHANES cohort, across the years 2003–2008. Four of the publications from population-based samples (1, 7–9) have come from this population. Although associations have also been reported in the UK EPIC cohort (10) and two further publications from the same large cross-sectional study in Shanghai, China (2, 11) (with conflicting results), it is important that we gather data from many different countries. This is because median urinary BPA levels and BPA exposure routes may vary widely over these populations and, secondly, using the same set of samples with different outcomes does not necessarily add clarity to literature. We