Most Americans have levels of BPA within their serum within the range of concentrations sufficient to suppress adiponectin in these experiments. The BPA effect on adiponectin disappears at higher levels.
What did they do?
Hugo et al. carried out a series of experiments with fat tissue surgically removed from people, exposing the tissue to different amounts of BPA and observing how this affected adiponectin levels. They compared BPA's effect to that of estradiol (E2), a natural form of human estrogen.
The fat tissue was obtained from patients undergoing surgical procedures at The Christ Hospital in Cincinnati, Ohio. The tissue came from three different types of fat deposits: breast, subcutaneous abdominal and fat tissue associated with internal organs. This last tissue type was obtained from morbidly obese patients undergoing bypass surgery.
In one round of experiments they placed small pieces of the tissue in plates and exposed them to different concentrations of BPA and estradiol. After 6 hours they measured the weight of the pieces and collected fluid to measure for adiponectin.
In another round of experiments they collected mature fat cells (adipocytes), placed them in tubes and subjected them to different treatments of bisphenol A and estrogen.
In a third set of experiments they isolated RNA from the different tissues and used real-time PCR to assess changes in gene expression.
What did they find? Background rates of adiponectin release from the different tissues was highly variable among patients. Nevertheless, treatment by BPA and E2 had highly significant effects, suppressing adiponectin release in the tissues of almost all patients even though they may have started from different background rates.
BPA and E2 both suppressed adiponectin release from fat tissue taken from 8 female breasts. This is shown in the graph to the right, plotting average values measured at each concentration. While E2 followed a traditional dose-response curve, BPA showed a non-monotonic dose-response, with the strongest effect, approximately a 50% reduction, occurring at 1 nM (1 part per billion). This is a level commonly seen in human serum (the median is 2 nM)
Adiponectin release was also suppressed in other fat tissue types. In abdominal subcutaneous tissue, BPA's effect was quite similar to the effect in breast tissue. In this tissue, compared to breast fat, E2 had a strong non-monotonic pattern, like BPA.
In experiments with isolated mature adipocytes, Hugo et al. found that exposure to BPA and E2 significantly suppressed adiponectin release at the lowest doses tested (0.1 nM).
In experiments with tissue from a morbidly obese woman, BPA significantly inhibited release of adiponectin. The effect was stronger for visceral fat tissue than for subcutaneous. Both BPA and E2 inhibited release by 65% and 50% after 6 hrs and 24 hrs into the experiment. With tissue from a morbidly obese man, they found that BPA affected relase from subcutaneous tissue but not from visceral tissue.
What does it mean? These results show that BPA at levels well-within the range of common human exposure suppress levels of a hormone that protects people from metabolic syndrome and its consequences: heart disease and Type 2 diabetes. In many of the experiments, BPA's effect was comparable to E2's, and in some even stronger. The effects were seen in all three types of tissues examined.
It is important to note that these results were obtained using human tissues and hence avoid the challenges that often arise when extrapolating results from rodents to people.
Diverse and robust data, both prospective and longitudinal, link low adiponectin levels to increased risk of heart attacks and Type 2 diabetes. This link is thought to be causal: Lower adiponectin levels cause heart attacks and Type 2 diabetes. These results thus predict that BPA, by suppressing adiponectin, is a causal agent for these illnesses. No epidemiological studies published to date have tested whether there is an association between BPA and metabolic disorder, Type 2 diabetes or heart disease. The non-monotonic nature of the dose-response curves reported here will complicate such a test, as epidemiological studies rarely factor in such considerations.
Another notable aspect of these results is that they focus on adult exposures. Most of the scientific literature on BPA showing effects at low levels of exposure has been about developmental exposures, especially fetal and neonatal. Early life stages are usually more vulnerable to endocrine disruption. The sensitivity to BPA observed in these adult tissues emphasizes the potential for exquisite sensitivity in adults. It also suggests that these same physiological processes should be examined in earlier life stages to determine whether fetal and early life effects related to metabolic disorder may be even more sensitive.
These data were published the same week the the US FDA issued a draft conclusion that BPA is safe at current levels of exposure. These data should be integrated into that assessment before it becomes finalized.
Hugo, ER, TD Brandebourg, JG Woo, J Loftus, JW Alexander and N Ben-Jonathan. 2008.
Synopsis by Dr. Pete Myers