Note: Dr Kindzierski graciously unpacked his criticism of the report in this article.

Editor’s Note: Scientists bickering between one another is an institution unto itself. I’m sure neither of these gentleman will take anything said personally.

Over the last four days I have driven from Winnipeg to the Toronto region with our kids, which has given me much time to reflect on the amazing media response to our report, ‘Environmental and Human Health Implications of the Athabasca Oil Sands for the Mikisew Cree First Nation and Athabasca Chipewyan First Nation in Northern Alberta’.

The reporting has understandably been concentrated in Alberta, but also across the rest of Canada as well as some additional coverage in the US, Europe, and Asia. This media coverage plays an essential role in getting the word out and – ideally - in affecting change.

Much more disappointing has been the lack of government response, positive or negative. I gather there has been a flurry of internal activity, and scrambling as agency staff and researchers make sense of the report. Ultimately, however, the government tactic seems to hide in the weeds, and to wait out the controversy. Hoping for it all too blow over and for the status quo to resume. This of course only puts the downstream Indigenous communities further at risk. It is a strategy that needs to be rebuked at every turn, by everyone.

Ultimately, however, the government tactic seems to hide in the weeds, and to wait out the controversy.

Generally speaking though, the report and its major, mostly media-mediated outcomes, have been tremendously well received. Even by media in Calgary, which, as one radio jockey reminded me, is still “Oil Town.” There has only been one critical voice, as reported in a generally disappointing CBC article, that of Dr Warren Kindzierski, an Associate Professor at the University of Alberta, in their School of Public Health.

Dr Kindzierski has gone on the record being highly critical of the report and the underlying research design and analysis regarding the health and more specifically the cancer results. Although his criticisms are of course welcome and will only contribute to discussion regarding these important issues, I (perhaps predictably) see them as entirely wrongheaded.

Dr Kindzierski accused the report, and more properly me as the lead researcher, as being biased and subjective. Rather than focusing on any of the major outcomes or conclusions, his criticisms have instead focused on a minor dimension of the cancer outcomes. Indeed, his focus was on the outcomes related to smoking, and then he quite strategically used these criticisms to impugn the credibility of the report, the larger project, and by implication the community concerns as a whole.

Although his criticisms are of course welcome and will only contribute to discussion regarding these important issues, I (perhaps predictably) see them as entirely wrongheaded.

This is of course is his right, and some would say his duty as a university professor. Although there is clearly an element of his own personal bias, which I will elaborate on later.

I have not interacted with Dr Kindzierski directly, and only have the few media tidbits reported on by CBC, in their usual “here is some of the pro, a smidgeon of the con, add water, and mix” kind of reporting. One of the researchers working on this project made a quick visit to Dr Kindzierski, to find out more about his positioning. I’ll use both the CBC report as well as the phone call to inform this blog, although I will only quote the July 7 CBC article.

:

Study suggests link between oilsands and Fort Chip illnesses

“Critics say study’s methodology is not scientifically rigorous enough”

“The conclusions of the study are considered controversial, in part because of the study’s methodology, which some say is not scientifically rigorous enough to support the findings. Warren Kindzierski, an associate professor with the School of Public Health at the University of Alberta, has questions about the model used by researchers, which could create subjective results. For example, people were asked about the “perceived role of smoking in affecting participant health,” which could slant the study’s results, Kindzierski said. “If the subjects do not believe it’s important, it’s given a low weight in the model and that introduces an element of bias.” Kindzierski says he isn’t sure the evidence in the report is strong. “I do not believe that this type of study offers the smoking gun evidence that some people might think,” he said.”

As well as a Radio-Canada piece:

Santé et sables bitumineux : une étude manitobaine critiquée

Warren Kindzierski, un spécialiste en santé et en environnement à l’Université de l’Aberta, n’est pas convaincu de l’intégrité de la méthodologie utilisée. « Je ne crois pas que ce type d’étude offre la preuve criante de ce que peuvent croire certaines personnes », soutient-il. Les auteurs de l’étude promettent que celle-ci sera soumise à une révision scientifique exhaustive par leurs pairs d’ici deux ans.

As indicated in the CBC article, Dr Kindzierski argues that the model used by the researchers could create subjective results.

In particular, “If the subjects do not believe [smoking]’s important, it’s given a low weight in the model and that introduces an element of bias.”

This statement reflects an inadequate understanding of the data and the data analyses reflected in the report, both with respect to the science and Traditional Knowledge, as I describe below.

First of all, had Dr Kindzierski examined the report more thoroughly, particularly Table 9.17, he would have seen that the majority of participants fully acknowledge the importance of smoking in the decline of their own health and that of their community: 71.6 % and 75.6%, respectively. Indeed, with respect to community health, smoking was identified as the third most important factor responsible for this decline, this following “drug abuse” (87.6%), the Oil Sands (82.6%) and then smoking, the latter tied with stress at 75.6%. The role of smoking on the decline of individual health was assessed as of somewhat lower importance, in part likely reflecting the current individual smoking habits of participants.

Most adults in Fort Chipewyan smoke, and almost all have smoked at some point in their lives, and so these insights as grounded in their rich Traditional Knowledge (TK) are well placed. They of course undermine Dr Kindzierski’s contention that communities do not appreciate the adverse health implications of smoking.

It is the ubiquitous community concern regarding smoking and its implications for health that was responsible for the outcomes of the regression analysis, and thus the absence of any significant role in determining cancer occurrence. This again directly contradicts Dr Kindzierski’s claim.

We took an approach in our modeling, which divided participants into two groups: those having contracted cancer and those who have not yet been diagnosed with cancer. We then looked at a wide diversity of (independent) variables that would help explain why those two groups might differ in their cancer occurrence. In so doing, we tested 12 variables that we anticipated would be relevant, including Traditional Knowledge regarding smoking. The outcomes of these data analyses were clear and meaningful; the consumption of traditional foods (the focus on this study) as well as employment in the Oil Sands played a significant role in cancer occurrence. Not so smoking.

That is because the two groups varied in their employment history (cancer survivors were more likely to have worked in the Oil Sands than no-cancer participants) and in traditional food consumption (cancer survivors more frequently ate traditional foods and locally caught fish than no-cancer participants).

But why not smoking? Everyone knows that smoking causes cancer and two participants had contracted lung cancer. Upon further review, those participants recognizing that smoking was important in the decline of their own health were remarkably equally split between the two groups.

Indeed, 29.6% of no-cancer participants vs. 30.0% of cancer survivors somewhat agreed that smoking was important. Moreover, 47.9% of no-cancer participants vs.

45% of cancer survivors strongly agreed that smoking was important in the decline of their own health. Moreover, when I more closely examined the two lung cancer survivors, one somewhat agreed and the other strongly agreed that smoking was important.

Everyone knows that smoking causes cancer and two participants had contracted lung cancer. Upon further review, those participants recognizing that smoking was important in the decline of their own health were remarkably equally split between the two groups.

Again, the majority of all participants had already recognized that smoking was important in health declines. But because the cancer survivors and no-cancer participants were agreed in its importance, there was no ability of this independent variable (smoking) to predict cancer occurrence. This again, is in direct contrast to Dr Kindzierski’s contestation that people did not think smoking was important.

Ultimately, it’s too bad his criticism focused so tightly on this ultimately minor aspect of the report. It would have been better for all if he had focused his criticisms on some of the more important dimensions of the work. That said, it seems like he had very little time to prepare and only read parts of the health chapter, to say nothing of the larger report and the rationale for using both TK and science, which he might have found useful.

He then apparently makes some general and unfounded criticisms of the report.

Kindzierski says he isn’t sure the evidence in the report is strong. “I do not believe that this type of study offers the smoking gun evidence that some people might think,” he said.”

This in part likely reflects the lack of value and credibility that he affords the rich Traditional Knowledge that these communities hold, and which informs their own concerns and many of our outcomes. It also arguably (and ironically) reflects his own bias as a former long-time employee of the Alberta Health and Wellness, the same governmental agency that has been the focus of the tack of this report and the July 7 press release.

Guess what, despite many claims to the contrary, all scientists have values and all are biased to some degree; it’s the human condition. And guess what again, these values and biases shape how we interpret our own data - and those of other scientists. In this case, this is widely reflected in how (quantitative) scientists view Traditional Knowledge (TK). Implied in the quotes above, and as made explicit in the phone interview with Dr Kindzierski, he clearly values data that are quantitative and numerical over TK, which he sees as subjective and as a source (misplaced as I showed above) of bias in the modeling.

The reason we referred to smoking in this holistic sense, rather than as say the number of cigarettes smoked per week, or number of years smoked, or some other indicator is three-fold:

• This approach better resonates with holistic TK, and recognizes that the participants have a much better ability to perceive and assess the role of smoking in their own health and that of their communities than would outside scientists, especially in this initial phase of research.
• This was not a study on the impacts of smoking on cancer, not was it even a cancer study, nor was it technically only a health study. Although people spoke to cancer occurrence throughout the project, it was in the context of larger environmental health declines.

Explicit questions on smoking and other important but background factors will play an important part of the detailed, follow-up medical or patient case histories, which will ideally be conducted with all participants. But these were inappropriate at this first and very sensitive stage of research.

That said, the implications of smoking for human health are made even more complex by the rich and important value that tobacco and smoking often have for the spiritual and other cultural wellbeing of many traditional residents of these communities, which is yet another reason to tread lightly regarding this topic.

• Finally, our regression analysis only allowed for 10-12 independent variables, given the sample size, and we had highlighted the possible role of traditional foods in shaping health throughout the project. Such an integrative TK-based smoking indicator allowed us to accommodate a wide diversity of factors, while providing meaningful insight into the role of smoking in declining individual and community health, and more specifically its role in cancer occurrence. We also took this approach with respect to stress, consumption of traditional foods, the Oil Sands, and even the Bennett Dam in the modelling.

So what about researcher bias, in the report and outside?

Some might accuse me of bias. On one hand our research programme at the Environmental Conservation Lab is quite conventional and is funded by the government (Health Canada, SSHRC, NSERC, CIHR etc.). We focus on the link between environmental and social injustice as it relates to the impacts of industry on the health and wellbeing of Indigenous and rural communities. Our work is conducted in close collaboration with these communities, and is aimed at better understanding any impacts and also at facilitating communication and shaping decision-making regarding these issues. A year ago, you would have seen me at Neil Young’s Honour the Treaties concert in Winnipeg, singing, dancing (badly) and showing my support of the Athabasca Chipewyan First Nation. That is my bias.

Dr Kindzierski is a well-published and established researcher in the field of environmental health, but apparently has his bias as well.

Below, I quote from a 2010 UofA press release entitled, “People living in communities near oilsands can breathe easy: U of A study, regarding one of his recent projects related to the Oil Sands.”

Findings very positive from a public health point of view, says researcher Warren Kindzierski”

Warren Kindzierski, University of Alberta researcher in the School of Public Health, says people living in the communities of Fort McMurray, Fort Mckay and Fort Chipewyan should feel confident that the air they are breathing is safe

That work apparently was prompted by the Wood Buffalo Municipality, reflecting their concerns about declines in the quality of air downwind from the Oil Sands. The findings saw no evidence of any downwind impacts in air quality: indeed, exactly the opposite was shown. The air quality in Fort McMurray was actually better than the comparison communities including Edmonton and Toronto. As Dr Kindzierski indicates in that Press Release, “These findings are very positive from a public health point of view”.

Dr Kindzierski was a long-term employee of the Alberta Health and Wellness, and Head of Chemical Risk Assessment before leaving to work at the University of Alberta. This of course is the same agency that has been issuing these inadequate cancer reports and who both First Nations have been highly critical of.

What is clear from the following Alberta government’s explicitly pro-industry Oil Sands website, is that Dr Kindzierski was a long-term employee of the Alberta Health and Wellness, and Head of Chemical Risk Assessment before leaving to work at the University of Alberta. This of course is the same agency that has been issuing these inadequate cancer reports and who both First Nations have been highly critical of.

That same website described Kindzierski’s work as such:

“Released in April 2010, his 10-year study on the effects of oil sands development on local air quality found that, despite ongoing development, residents in Ft. McMurray and surrounding communities in the oil sands region had better air quality than found in larger cities, including Edmonton, Calgary and Toronto.” (original bold face included).

In conclusion, I have argued that Dr Kindzierski’s critiques are faulty and poorly grounded in the data and reflect his biases with respect to quantitative science. I have also argued that they reflect his own positioning alongside Alberta Health and Wellness, if not the Oil Sands industry.

I welcome the opportunity to debate the outcomes of this report with Dr Kindzierski, and look forward to otherwise discussing these issues with him in the future, in whatever capacity – maybe even at the next Neil Young benefit concert.