“A chemical found in food tins and baby’s bottles has been linked to an increased risk of developing heart problems,” The Daily Teleg raph reported. It said that scientists have found that people with high levels of bisphenol A (BPA) in their bodies were a third more likely to develop heart disease than those with low levels.
This study found some associations between BPA levels in the urine and the likelihood of having certain diseases. However, it has several limitations, and cannot prove that BPA caused these diseases.
BPA is commonly found in many household items, and there is likely to be little that individuals can do to reduce their exposure. The US Department of Health and Human Services has information for parents on reducing their child’s exposure.
To date, researchers have found no conclusive evidence that BPA is harmful to humans. Despite this, some countries have taken precautions and Canada has introduced legislation to ban the use of polycarbonate in baby feeding bottles. The European Food Safety Authority (EFSA) stated in 2008 that it considers levels of BPA exposure to be safe, saying "after exposure to BPA the human body rapidly metabolises and eliminates the substance". It continues to monitor the situation and is currently evaluating the study that led to the ban in Canada.
Where did the story come from?
The research was carried out by David Melzer and colleagues from the University of Exeter and University of Plymouth. The study was funded by the Peninsula Medical School and published in the peer-reviewed open access journal PLoSOne .
The Daily Telegraph gave a balanced report of this research, but did not discuss the limitations of the study and the fact that it cannot prove causation. The headline that the chemical is found in babies’ bottles may cause undue concern to parents. BPA is found in many household items, including some baby bottles, and it is probable that there is little that can be done to avoid it.
What kind of research was this?
This cross-sectional study aimed to investigate associations between the concentration of bisphenol A (BPA) in the urine and various health measures. The National Health and Nutrition Examination Survey (NHANES), conducted in 2003/04, reported that there were associations between urinary BPA concentration and heart disease, diabetes and liver enzymes, in adults. It has also been speculated that BPA can affect levels of hormones. The current analysis was a follow-up to this earlier research, and used data from the 2003/04 survey and new data from the 2005/06 survey.
The drawback of cross-sectional studies is that they cannot prove causation. This is because they measure both the exposure and the outcome at the same time. It is therefore not possible to say whether the exposure occurred before the outcome. In this study, BPA levels were only measured on one occasion, and it is not known whether the participants’ levels of BPA remained the same or fluctuated over time, or whether high levels preceded the onset of the disease problems.
What did the research involve?
The study included 1,455 US adults (aged 18-74) assessed in 2003/04 and a further 1,493 assessed in 2005/06. The participants were asked various health questions using computer software called the Computer-Assisted Personal Interviewing system. They were asked whether a doctor had ever diagnosed them with angina, coronary heart disease, heart attack, stroke, asthma, diabetes, emphysema, chronic bronchitis, arthritis, thyroid problems, any kind of liver condition or cancer. They were also given a medical examination and had blood tests for liver enzymes. BPA concentration in their urine was measured in a randomly selected subset of a third of participants.
The researchers assessed the relationships between BPA and diagnoses of heart attack, coronary heart disease, angina, diabetes and liver enzyme levels, adjusting for age, sex, race/ethnicity, education, income, smoking, body mass index (BMI), waist circumference and urinary creatinine concentration (a measure of kidney function).
What were the basic results?
Urinary BPA concentrations were significantly lower in the 2005/06 sample than the 2003/04 sample (a mean (average) of 1.79ng/ml versus 2.49ng/ml). In the 2005/06 sample, higher BPA concentrations had some association with coronary heart disease, with a calculated 33% increase in risk with every incremental rise in BPA concentration, although this increase was only just significant (odds ratio [OR] 1.33, 95% confidence interval [CI] 1.01 to 1.75). There was no significant association with heart attack from the 2005/06 sample alone. The increase in risk for the combined outcome of cardiovascular disease (report of coronary heart disease, heart attack or angina) was significant when data were pooled from both years (OR 1.26, 95% CI 1.10 to 1.44). There was also no significant increase in diabetes risk in the 2005/06 sample, but pooled data from both samples were significant (OR 1.24, 95% CI 1.10 to 1.40).
The researchers also carried out analyses between BPA and all other diseases assessed.
How did the researchers interpret the results?
The researchers concluded that greater BPA exposure, as reflected by urinary levels, is “consistently associated with reported heart disease in the general adult population of the USA”. They say that further studies are now needed to clarify the mechanisms behind these associations.
This study has found some associations between BPA levels in the urine and the likelihood of having certain diseases. This study has several limitations:
The findings do not prove that BPA caused the diseases investigated. The measures of BPA and disease outcomes were only taken at one point in time, and it is not possible to conclude that one thing has caused the other. It is also likely that BPA exposure and levels in the urine vary over time, so the measurements may not reflect the participants’ usual levels.
- The researchers carried out numerous statistical analyses. Not all of the results for the 2005/06 group were significant, and many others for this group were only just significant. Pooling data from both groups showed significant associations between BPA and cardiovascular disease and diabetes, but the high number of statistical tests carried out increases the risk of these findings occurring by chance.
- The disease outcomes were all by self-report only, which increases the likelihood of inaccuracy.
- BPA was measured in only one-third of the participants. Although this subsample was selected at random, assessing the entire sample would be likely to give a more accurate picture of average levels.
As most of the human population is exposed to this chemical, there is little that individuals can do at present to try to reduce their exposure to BPA. It is not known from this study whether particular products give greater exposure. Although The Telegraph specifies babies’ bottles, this study did not look specifically at this. Parents should not be concerned that by feeding their baby from a bottle they are putting them at risk of heart disease or diabetes. The US Department of Health and Human Services has information for parents on reducing their child’s exposure.
BPA is said to be a compound with one of the highest production volumes in the world, and over 90% of humans are exposed. It is commonly found in many household items. If there is a chance that it poses a risk to health, further research is warranted.
There is a growing body of research into the safety of BPA, but none of the research to date has found conclusive evidence that it is harmful to humans. Despite this, some countries have taken precautions and Canada has introduced legislation to ban the use of polycarbonate in baby feeding bottles.
The European Food Safety Authority (EFSA) continues to monitor the situation and is currently evaluating the study that led to the ban in Canada. In July 2008, it stated that, “after exposure to BPA the human body rapidly metabolises and eliminates the substance. Newborns can similarly clear BPA at levels far in excess of the TDI (tolerable daily intake [amount that can be safely consumed in a day]). This means that exposure to BPA is well below the limit that “provides a sufficient margin of safety for the protection of the consumer, including foetuses and newborns”.
The EFSA has also evaluated a 2008 study from the US, which suggested a link between raised levels of urinary BPA and a higher rate of conditions such as heart disease and diabetes. It concluded that the study has no information on long-term exposure to BPA, “which would be important in order to establish a correlation between BPA and the development of the chronic medical conditions in question. The EFSA found that the study did not provide sufficient proof of a causal link between BPA and these health conditions”.
This article has been revised following feedback: March 3, 2010
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Daily Telegraph, 13 January 2010
Links to the science
PLoS One 5(1)