Recently Statistics Canada released their Report on Human Biomonitoring of Environmental Chemicals in Canada; in the US a similar report is regularly published by the CDC called the National Report on Human Exposure to Environmental Chemicals. The Canadian report has generated a lot of media attention, specifically regarding bisphenol A (BPA) being measured in 91% of the population. However, the reporting typically doesn’t include any context or information about what this really means – a chemical being detected in a urine sample doesn’t automatically mean it is causing any harm. This is partly because most journalists don’t have the appropriate background to actually understand what they are reporting on when it comes to environmental exposures (or science in general); I suspect sensationalism may also contribute.
The Canadian and US reports actually both do a decent job of providing some background information about the chemicals measured, unlike similar reports published by groups like Environmental Defence that skimp on the background/context (again likely related to a desire for sensationalism). However, even the government reports often don’t have enough information to establish whether what is being measured in the population is really of concern or not.
First, a bit of background on what was done for the Canadian study. Blood and urine samples were collected from approximately 5600 people across Canada, ages 6 to 79 (the plan is for the next survey to go as young as 3). It wasn’t a totally random sample, since their were logistical and cost considerations, but overall the survey is believed to cover approximately 96^ of the Canadian population within the age range sampled. These samples were analyzed for various chemicals that people are exposed to (both natural and man-made chemicals), or metabolites of these chemicals, including several metals and trace elements, organochlorine pesticides (such as DDT), polychlorinated biphenyls (PCBs), polybrominated flame retardants (PBB & PBDEs), perfluorinated compounds (PFCs), BPA, organophosphate insecticies, phenoxy herbicides (2,4-D), chlorophenol, and cotinine (a metabolite of nicotine). The main purpose of the sampling wasn’t really to determine whether these chemicals were causing adverse effects, but rather to establish a baseline so changes over time can be tracked, and for comparison between Canadian subpopulations or with other countries.
The report includes details on how the survey was conducted. It also includes a section on interpreting the data (which I suspect very few journalists read), which says that biomonitoring “cannot ell you what health effects, if any, may result from that exposure. Our ability to measure environmental chemicals at very low levels has advanced. The presence of a chemical in a person’s body does not necessarily mean that it will cause a health effect.” The report also includes some general information about each of the chemicals measured, including what the chemical is, how it is used, how it is released into the environment, what happens to it in the body, and some information on health effects, including (where available) what dose Health Canada considers to be safe. The concentrations measured in blood and/or urine are then provided in tables, including some basic statistics. For any American readers, the US reports are very similar.
Unfortunately, it is very difficult to relate the concentrations measured to actual potential for health effects. For example, consider the case of BPA. Health Canada established a provisional tolerable daily intake (the amount they believe someone can be exposed to over a lifetime without significant adverse effects) of 25 micrograms per kilogram body weight per day. There’s still some controversy about what the actual safe dose is, which I won’t get into today, but for sake of argument let’s use that number for now. The study authors measured the total of BPA and major metabolites of BPA in urine, and found a median concentration of 1.33 micrograms per litre (the median is the value which 50% of the population is above and 50% below, and tends to be a bit more meaningful than a straight average for data like this). Since the amount of urine produced by people is variable, the concentration was also reported as 1.39 micrograms per gram of creatinine (since the amount of creatinine excreted is fairly constant this tends to give a more reliable number). Concentrations tended to be highest in They used the total of BPA and its metabolites because BPA is for the most part rapidly metabolized into a relatively non-toxic form and then excreted, but that metabolite is unstable and can convert back to BPA after sampling. The sampling is further complicated by contamination during sampling and analysis, since BPA is in a lot of plastics; the average BPA concentration in laboratory blanks (basically distilled water run through the same sampling and analytical equipment) averaged 0.41 micrograms per litre, and as high as 1.27 micrograms per litre. The authors attempted to correct for this, but given the variability this means there is a fair amount of uncertainty in the actual concentration in urine.
So we have a concentration in urine, which is subject to a fairly high level of uncertainty due to the high concentrations measured in laboratory blanks, but the value we use to determine whether there might be health effects is an amount ingested per day. How do we compare the two? The short answer is that they can’t be directly compared. However, if we understand the fate of a chemical in the body, we can at least arrive at a rough approximation.
The metabolite of BPA is actually excreted fairly quickly (generally within about 6 hours). This means for it to be measured in 91% of the population, there must be fairly continuous exposure. So, for a quick “back of the envelope” calculation I’m going to assume that the amount of BPA metabolites excreted in a day is equal to the amount ingested (you couldn’t make this assumption for chemicals that stay in the body longer). Normal creatinine excretion is on the order of about 0.020 grams per kilogram body weight per day. So, if BPA is detected at 1.39 micrograms per gram creatinine, that would correspond to excretion of about 0.028 micrograms of BPA per kilogram body weight per day per day. If we assume that the intake rate is the same as the excretion rate, then the average exposure is about 1000 times lower than what Health Canada considers to be a “safe” dose.
The same approach doesn’t really work for some of the other chemicals measured, such as metals or substances like DDT that bioaccumulate. For these chemicals you’d need to use a fairly sophisticated mathematical model of chemical fate in the body, and for many chemicals we don’t know enough about to develop reliable models yet. Also, some of the chemicals included in the study, such as zinc and copper, are essential nutrients which you actually need to have a certain amount of. Others such as arsenic are toxic, but are naturally occurring – background exposures to arsenic from food and water often exceed the amounts that many health agencies would consider “safe”.
So by itself this report doesn’t tell us whether environmental chemicals are causing adverse health effects. It is an important part of the big picture though, and will help us at least understand what chemicals we are being exposed to, whether those exposures are increasing or decreasing, and how we compare to other countries.