More bad epidemiology about BC LNG from the MDs at CAPE

I have written a lot about the BC liquid natural gas (LNG) export industry. I have done so because my examination of the climate math says BC LNG will help reduce global greenhouse gas emissions and will help in the fight against climate change. Many activists disagree with me, with one of the most vocal anti-LNG groups being the good MDs at the Canadian Association of Physicians for the Environment BC (CAPE). CAPE has a history of fighting the natural gas industry. Their most recent effort involves a travelling roadshow called: “Voices from the Sacrifice Zone: Fracking in BC’s North“. Their presentation is summarized in a Narwhal article written by CAPE BC Board Member Melissa Lem. As I have written in the past, the good MDs at CAPE often struggle when trying to translate their professional experience in private practice through the lenses of toxicology and epidemiology. As I will discuss in this blog post, this is again the case with their travelling anti-LNG roadshow. Ultimately, what I intend to show in this blog post is that the anecdotal experiences of the people in the roadshow are trumped by the epidemiological work done to date in BC’s northeast and that any decision on the future of BC LNG should be based on science and not anecdotes.

To explain what I mean by anecdotal experience let me give you an example using one of my favorite hobbies: birding. I love birds and have a backyard feeding station for birds. Before we got our puppy I had black oil sunflower seeds in our station. As a result, I would see dozens of finches daily. Unfortunately, our puppy took to eating the discarded sunflower seed shells so we had to stop including the seeds in our station…and the finches disappeared from my backyard. As a birder I know that all the finches in the neighbourhood didn’t just die off. I am quite sure they simply moved to a feeding station down the street. If we did a regional survey and sent those results to a statistician that statistician would likely find that the number of finches in the region hadn’t changed. In the field of human health risk assessment that statistician is called an epidemiologist.

Going back to the CAPE roadshow. It consists of a small number of practitioners presenting their anecdotal observations to the public. The problem is the anecdotal information they present is not backed up with a look at the bigger picture. This is unfortunate because, as presented in the Narwhal article, a detailed epidemiological study has already been carried out for the BC Northeast and the results were very reassuring. The assessment called: Cancer Incidence in the Peace River South Local Health Area looked at the rate of cancer incidence in the region where LNG is generated and found that cancer incidence was normal or as they put it:

the number of total cancers diagnosed in this region are consistent with Northern Health regional rates; the number expected based on regional rates is almost identical (1193 vs 1201)….overall cancer incidence over the past 10 years in this region is consistent with average cancer rates in Northern BC.

So what are the CAPE doctors using to refute this epidemiology? Why anecdotes of course. In the Narwhal article they note that there were ten incidences of glioblastoma in the area when the average for that population should have been five. The article also notes that this is an “unofficial” count and the numbers presented are not reflected in the cancer incidence study. Not a surprise since the doctor had observed less than 2 incidents a year. Also interesting is the link provided in the piece did not indicate that VOCs were a particular risk factor for this cancer but rather:

The vast majority of glioblastomas occur randomly, without inherited genetic factors. The only confirmed risk factor [my emphasis] is ionizing radiation to the head and neck region. Studies of environmental and genetic factors contributing to glioblastomas have so far been inconclusive or negative. 

Looking at the literature it is understood that radon is a significant risk factor for this cancer while VOCs are not. So when the author points to this cancer she fails to note that the more likely reason for this potentially increased incidence is that Dawson Creek is a hot spot for radon.

They also present the experience of “an internist who diagnosed ten cases of idiopathic pulmonary fibrosis (IPS) in the two short years he’d worked in Dawson Creek“. What they omit is that genetics represents a major risk factor for IPS so in a close-knit community (especially one with First Nations communities that have strong genetic ties) this consideration becomes important and must be considered in the math. This demonstrates why it is so important to use the tools of epidemiology in trying to come to an evidence-based conclusion.

The most telling part of the presentation is CAPE’s go-to study in their fight against LNG; this is the study that CAPE has used in virtually every one of its presentations: Gestational exposure to volatile organic compounds (VOCs) in Northeastern British Columbia, Canada: A pilot study. This study is used because it involves pregnant women and benzene and so pulls at virtually every observer’s heart-strings. The problem is the study is explicitly described as a pilot study meaning it has an exceedingly small sample size and is not able to effectively address critical confounding variables. As such it is absolutely useless for telling us whether LNG is having an effect on the population. Because of its importance to CAPE, I will spend a bit of time explaining where it goes wrong from a decision-making perspective.

This pilot study involved testing for benzene and benzene metabolites in the blood of 29 pregnant women in northeastern BC. Now for the first problem: sample size. Twenty-nine participants is a tiny population to search for trends in any study but this one is particularly notable because of the 29 women, 2 were regular smokers and 4 more were regularly exposed to second-hand smoke. Smoking is a serious confounding variable in a study of this kind because smokers are expected to have higher benzene and benzene metabolite concentrations in their blood as a by-product of exposure to smoking.

In this study, the elevated benzene and benzene metabolite concentrations were only observed in a very small number of participants. Unfortunately, the authors did not confirm whether the elevated participants were among the number that were regularly exposed to cigarette smoke, they only indicate that the highest results were not smokers. Step 1 in any more detailed epidemiological assessment would typically have been to exclude smokers from the study.

The study also does no discuss another major confounding factor: whether the women lived in houses with attached garages. Attached garages you ask? Yes, since the presence of an attached garage is one of the strongest indicators (after smoking) of potential human exposure to volatile organic compounds (VOCs) like benzene. To explain, see this article:Automobile proximity and indoor residential concentrations of BTEX and MTBE where they point out:

Residing in a home with an attached garage could lead to benzene exposures that are an order of magnitude higher than exposures from commuting in a car in heavy traffic, with a risk of 17 excess cancers in a population of a million

Another article (Migration of volatile organic compounds from attached garages to residences: A major exposure source) points out:

A total of 39 VOC species were detected indoors, 36 in the garage, and 20 in ambient air. Garages showed high levels of gasoline-related VOCs, e.g., benzene averaged 37±39 μg m−3. Garage/indoor ratios and multizone IAQ [indoor air quality] models show that nearly all of the benzene and most of the fuel-related aromatics in the houses resulted from garage sources, confirming earlier reports that suggested the importance of attached garages. Moreover, doses of VOCs such as benzene experienced by non-smoking individuals living in houses with attached garages are dominated by emissions in garages, a result of exposures occurring in both garage and house microenvironments.

Put simply, the benzene metabolite study has challenges associated with sample size and confounding variables. These are the sorts of thing a more detailed epidemiological examination would be able to establish. Absent these considerations we have a study that tells us that more study is needed but should not be relied upon for decision-making purposes.

To continue my critique of the Narwhal article; I would also note that in Northeastern BC, where winters are extremely cold, indoor air issues are exacerbated because ventilation is minimized to prevent heat loss. In the Narwhal article the author questions why a study comparing the results to southern BC has not been carried out? The issue is that in southwestern BC we don’t see the same cold as they do in the northeast so people here are more likely to open their windows in winter and thus have fewer indoor air issues. This is why no epidemiologist would blindly compare northern and southern communities in a manner the author of the Narwhal author suggests should be done. It would be bad science.

Ultimately, the “Voices from the Sacrifice Zone” series consists of a collection of anecdotes with the common feature that all the anecdotes involve people who live or work in Northeastern BC. There are no controls, there is no evaluation of the big picture. It is a handful of activists telling their personal stories. As such it makes for a compelling presentation from a human-interest standpoint, and is useless from an environmental decision-making perspective.

The thing we have to keep reminding ourselves is that a presentation that includes a lot of compelling stories is of little use in determining if the LNG industry is safe. For that you need epidemiological research and the epidemiological work to date identifies no cancer hot-sports in need of more detailed assessment. There are no increases in VOC-related cancers in the data. Rather the research shows the exact opposite. The cancer rates are consistent with what is expected in this community. There are simply not enough cancer cases to raise any concerns. The authors of the report note that Acute Myeloid Leukemia (AML, the cancer CAPE associates with benzene) is so rare that the Peace River south Region sees an average of “less than 1 case per year” of AML. In the case of AML, in some years, there is literally no data to crunch.

To end this piece I want to reiterate a simple truth. Decisions about energy policy shouldn’t be made based on anecdotes and first-person narratives, no matter how compelling they may sound. First person narratives can inform further research but decision-makers need to consider real evidence assembled by people experienced in ensuring that the data is not the result of unexamined confounding variables. Epidemiologists have compiled those results and the current output from those experts indicates that the northeast is not a “sacrifice zone” but rather has absolutely typical diseases incidence rates. As such the anti-LNG roadshow really doesn’t inform evidence-based decision-making, rather it muddies the waters by implying a negative trend exists when the data says no such trends exist.

Author’s Note

An earlier version of this blog post had an error with respect to the incidence of glioblastomas. This was the result of my mis-reading the Narwhal article. The error had no significant effect on the piece but has been corrected in the current version.

The text has also been cleaned up with respect to the maternal benzene metabolites study as the original critique could be viewed as too imprecise. I hope the new language addresses those concerns.

This entry was posted in Renewable Energy, Uncategorized. Bookmark the permalink.

1 Response to More bad epidemiology about BC LNG from the MDs at CAPE

  1. Sandra Nishikawa says:

    Hi Blair, I have had similar comments about the incidence of Hepatocellular Carcinoma, lung cancer, in the Fort Chip area. A study done by the Alberta Cancer Board showed similar results that these cancers were no more prevalent than the general population. That the risk factors for these particular cancer were the same a for the general population, namely smoking and alcoholism, poor nutrition and lower socioeconomic level. That the presence of VOC is not the likely cause of the slight increase in three types of cancerHowever, these fallacies continue to make the rounds of the Gang Green talking points. I have included the study for you to read.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.