Study Linking BPA to Heart Disease Has Limitations
Related Materials
A study in the January 2010 edition of Plos One, an online journal, titled, “Association of urinary bisphenol A concentration with heart disease: evidence from NHANES 2003/06” by Melzer et al.reports a statistical association between higher BPA levels and increased incidence of heart disease in the 2005-2006 National Health and Nutrition Examination Survey (NHANES) population.1 The authors conclude that “chance is an implausible explanation for the BPA association with heart disease.” This report echoes findings published in 2008 in the Journal of the American Association (JAMA) titled, “Association of Urinary Bisphenol A Concentration With Medical Disorders and Laboratory Abnormalities in Adults” by Lang et al,2 that reported a relationship between increasing urinary bisphenol A (BPA) concentration and the prevalence of self-reported chronic disease, including diabetes and cardiovascular disease (See IFC Critique titled, “Study that Finds Association with Urinary Bisphenol A Concentration and Chronic Diseases Has Limitations” from September 2008 for more information).
Although these findings received considerable media attention, both of the studies have several limitations that need to be addressed when evaluating their utility. While the authors do attempt some adjustment for body mass index and waist circumference, they make no attempt to adjust for potential dietary confounders. Since BPA is found in liners of canned food, it seems the authors would have examined dietary data for potential confounders, especially since dietary data was readily available in this nutrition survey. It is highly probable that persons who eat more canned food and less fresh fruits and vegetables have a poorer overall diet, which would increase the risk of heart disease. Without controlling for diet in this study population, one cannot rule out that possibility that other dietary factors caused the observed relationship between BPA and heart disease.
In this study, heart disease rates were based on the subjects' self-reporting of whether they had ever been diagnosed with heart disease; there was no verification that the subjects who reported having been diagnosed with heart disease were, in fact, medically diagnosed with heart disease. The term heart disease can represent a wide number of different disorders that have varying etiologies. To truly draw any meaningful conclusions from this study, confirmation is needed to verify that those who self-reported heart disease in fact were medically diagnosed with heart disease; also, for the study to be valuable, standardization is needed regarding a definition of heart disease.
Another concern with this study is that the authors used spot urine collections (time of day for the collection was not controlled) and, by their use of an enzymatic deconjugation procedure, measured only total (conjugated plus “free”) BPA in the urine. In urine samples, the vast majority of BPA is in the conjugated (i.e., bound, or nontoxic) form. Because almost all BPA in the urine is conjugated (e.g, non-toxic), measurements of total BPA in the urine after the deconjugation process are misleading and dramatically overestimate the amount of free BPA that is present. In addition, it is conceivable that, during storage, a portion of the conjugated BPA could break down to “free” BPA, causing a further overestimation of free BPA. To get a true indication of the free BPA in the urine samples, the authors should have done measurements both with and without the enzymatic procedure to deconjugate the BPA. It also would have been of further benefit to measure total and free BPA in a limited number of samples at both the time of urine collection as well as after storage to see what the effects of sample storage were on the breakdown of BPA.
Lastly, it should be noted that the odds ratio of 1.33 for the association between BPA and heart disease was barely statistically significant (P = 0.043 versus the typical P <0.05 cut off point). It is important to consider the biological significance of any odds ratio reported in epidemiological studies, esp. in terms of relative risk. Also, it is important to remember that this study, due to its observational and cross-sectional design, can only show an association; cause and effect between BPA and heart disease cannot be determined from this study. Coronary heart disease is a multifactorial disorder that usually develops over many years. However, the BPA measurements in this paper represent the BPA concentration of the urine not only on just one day, but also on just one particular time of that day. It is to try to draw conclusions on the relationship of urinary BPA concern on heart disease based on this kind of sampling.
References Cited:
1. Melzer D, Rice NE, Lewis C, Henley WE, Galloway TS. Association of urinary bisphenol a concentration with heart disease: evidence from NHANES 2003/06. PLoS One;5(1):e8673.
2. Lang et al. Association of Urinary Bisphenol A Concentration With Medical Disorders and Laboratory Abnormalities in Adults. Journal of the American Medial Association. September 17, 2008.