A critical analysis of myth-perpetuating research on suicide prevention

In this essay, I will critically review a well-known explanation of variations in rates of death by suicide in First Nations communities in British Columbia and a lesser-known explanation of changes in suicide rates in Inuit communities in Nunavut. In both cases I will comment on the limited ability to make broad generalizations about suicide behaviour among Indigenous populations based on these flawed findings, and the implications of having only incomplete data to draw upon.

The Work of Chandler and Lalonde
Without doubt, the single most frequently cited journal article on suicidal behaviour among Indigenous peoples in Canada is “Cultural continuity as a hedge against suicide in Canada’s First Nations” by psychologists Michael J. Chandler and Christopher E. Lalonde (Chandler and Lalonde, 1998). Despite being widely cited, including in the popular media and by politicians, it has received little critical analysis.

Chandler and Lalonde obtained data on deaths by suicide in British Columbia’s 196 First Nations communities for the period 1987 to 1992, and found that almost 90% of First Nations youth suicides occurred in just 10% of the bands. This was an important finding, and a helpful contribution to public understanding. The authors stated that “Some communities show rates some 800 times the national average, while in others suicide is essentially unknown” (p. 191). Only deaths by suicide were measured, as comparable data on suicide attempts were not available.

Imagine a street in a city in southern Canada with 1,000 residents – roughly the size of some First Nations communities. If it were found that no one living on that street died by suicide during a five-year period, would we therefore describe the street as one where “suicide behaviour is essentially unknown”? Every human society we know of, present and past, experiences some degree of suicide behaviour. Why would Indigenous communities be otherwise? Any very small population group may well go five or more years without a resident dying by suicide, but there will likely have been some amount of suicide ideation – and possibly suicide attempts.

Still, how to explain the variation in rates at the community level that Chandler and Lalonde documented? They analyzed the data on the basis of six community-level “markers of cultural continuity” – “community practices that are interpreted here as markers of a collective effort to rehabilitate and vouchsafe the cultural continuity of these groups” (Chandler and Lalonde, 1998, p. 192). It should be noted that Chandler and Lalonde could only use variables for which they could get binary (yes/no) data for each of the 196 communities in their study.

The binary variables and their indicators are:

  • Land claims: Whether or not each First Nation community has taken “early steps to actively secure title to traditional lands.”1
  • Self-government: Whether or not bands had been “successful in having further established their right in law to a large measure of economic and political independence within their traditional territory.”
  • Education services: Whether or not the majority of children in the community attend a band-administered school.
  • Public and fire services: Whether or not a band has “substantial control over their police and fire services.”
  • Health services: Whether or not a community exercises some direct measure of control over health care services.
  • Cultural facilities: Whether or not a community had “a single facility specifically designated for cultural use.”2

Chandler and Lalonde sorted the communities on the basis of the relative frequency of their “markers of cultural continuity.” Communities with none of the six “cultural factors” had a combined rate of death by suicide of 137.5 per 100,000.3 In contrast, the rate in communities with all six “cultural factors” was “zero (no suicides within the 5-year study window).”

Figure 1. “Youth suicides rates by number of factors present in the community.” Source: Chandler and Lalonde, 1998, p.213.

Figure 1. “Youth suicides rates by number of factors present in the community.” Source: Chandler and Lalonde, 1998, p.213.


The appeal of Chandler and Lalonde’s findings is perhaps self-evident. It makes intuitive sense that Indigenous communities with strong “cultural continuity” would be more cohesive and happy than communities viewed as having weaker “cultural continuity” but it is not clear how varying rates of “cultural continuity” are mediated into decreased or increased levels of suicide. And in addition to the unvalidated way in which “cultural continuity” was measured, there are other critically important variables that Chandler and Lalonde’s approach ignored – including the varying degrees of unresolved historical trauma embedded in different communities, and the accessibility, cultural appropriateness and quality of the mental health care available to the residents of the different communities.

Chandler and Lalonde attempted to understand the factors that shaped suicide rates across a number of communities over a specific time frame using available community-level data. But what of jurisdictions where the variables they employed do not vary between communities? Let’s take Nunavut, for example. None of Chandler and Lalonde’s six variables vary at the community level across the territory. For example, determining “Whether or not the majority of children in the community attend a band-administered school” doesn’t make sense in a context where all children in every community attend schools run by the Government of Nunavut’s Department of Education. There is no community-level difference to measure in how education is delivered. Chandler and Lalonde’s methodology is therefore incapable of explaining the 6.8-fold variation in suicide rates among the Inuit residents of the territory’s 25 communities (Hicks, 2015), especially as ‘cultural continuity’ is visibly high in Inuit communities in comparison to many Indigenous communities in the South. (In fact, many Nunavut communities with higher rates of death by suicide also have high rates of language retention and wildlife harvesting.)

Further, Chandler and Lalonde’s study was not an ‘intervention study’ – they were not attempting to determine whether it was possible to decrease suicide rates by increasing the number of cultural factors present in the communities by means of a targeted intervention (or suite of interventions). They made no claims to that effect.

Others, however, have too often naïvely jumped to conclusions about the role of “cultural continuity” in suicide prevention. In April 2016, for example, Prime Minister Justin Trudeau told the national media that “In indigenous communities where there has been the support and an ability to do language and cultural teaching to an extremely high level, suicide rates have plummeted. So much of it is about a sense of identity, of who they are, how they fit in” (Pedwell, 2016). The assumption is that bolstering one of Chandler and Lalonde’s “markers of cultural continuity” will, independently of other factors, decrease suicide. I am not aware of any evidence to this effect.

The Work of Kral
In a 2009 journal article, psychologist Michael Kral, Health Canada official Patricia Wiebe and other collaborators made two very broad assertions about what does and doesn’t work with regard to suicide prevention in Inuit communities in Nunavut. First, that “Western, professional mental health treatment of Indigenous peoples in North America has, in many cases, met with limited success.” And, second, that when “community-based interventions are in place, suicide, even in communities with very high suicide rates, can stop altogether” (Kral, Wiebe, et al., 2009, p. 299).

Kral, Wiebe, and colleagues failed to explain the basis for their implicit claim that “communities” have to address elevated rates of suicide behaviour on their own, without help from organizations (both Inuit and public) at the territorial level. Why must that be the case? Especially when, in the event of a cluster of suicides, the limited resources in a tiny community are already overtaxed. How could collaborative and coordinated actions at the territorial level, providing resources and support to community groups, possibly be detrimental? Within this worldview, can communities share ideas and knowledge with each other? Can they draw from global knowledge and evidence within this model?

There is also a troubling downside to viewing suicide as a tap which communities can turn on or off depending on the actions they alone are equipped to take (or fail to take). For if it is the case that only community agency can prevent suicide – and long-standing resources gaps and inadequate mental health care have nothing to do with it – it only follows that when a suicide occurs, it is the fault of the community for not having made appropriate interventions. This is classic “blaming the victim” logic.

Such arguments are ideological in nature, justified by anecdotal claims of success and using small and sometimes inaccurate numbers passed off as statistical evidence. In 2011, Kral was quoted as claiming that when a school in Nunavut established a racquetball team: “The students loved it, the team did well against other community teams, and the suicides in their community stopped.” (Tatz, 2011) In an evidence-based discourse on suicide prevention, one might speak of the suicide rate being ‘reduced’ as the result of an intervention. To contend that suicidal behaviour can be “stopped,” especially by an intervention as simple as introducing racquetball to a school, is bewildering.4

In March 2016, Kral told a different tale when interviewed by the Public Broadcasting Service (PBS) in the U.S. He informed readers of the PBS website that “a grant-funded community center [that] gave young people a place to play billiards, watch movies and listen to stories told by village Elders” had resulted in Igloolik being “suicide-free, despite on-going unemployment and poverty” for five years. The lesson that Kral took from this experience was “Don’t give them these Western suicide prevention programs that don’t fit with their culture. Let them do it themselves.” (Santhanam, 2016).

It puzzles and perplexes that someone with such a sweeping grasp of suicide and culture, whose theoretical work implicitly and explicitly acknowledges the huge complexity of the topic, should delude himself and others with the simplistic notion that racquetball, youth groups, etc., are by themselves effective in reducing suicide rates.

Kral again claimed that a miracle had occurred in Igloolik in a 2016 report in the Journal of The American Academy of Child and Adolescent Psychiatry (JAACAP): “the community has not seen a suicide in approximately five to six years” (Kral, 2016). After Kral’s inaccuracy was brought to the attention of the editors of the JAACAP they published a rare Erratum notice – a formal statement that an academic journal had realized that it had published something that was factually inaccurate (Journal of The American Academy of Child and Adolescent Psychiatry, 2017).

Kral’s claims of magical solutions for suicide in Igloolik ignore the evidence. According to the Office of the Chief Coroner of Nunavut, Igloolik has suffered deaths by suicide in 17 of the last 21 years – including seven deaths in the past five years. There has been no “approximately five to six years” time period in which there was no death by suicide in Igloolik. Igloolik has had a somewhat lower suicide rate than its neighbouring communities in the North Baffin area since record-keeping began, and the fluctuations in the number of suicides in Igloolik in recent decades are similar to those seen in almost all other Nunavut communities (Hicks, 2015).

It should be noted that a mayor of Igloolik commented in the media about claims of miracles occurring in his community. “For one individual and-or organization to seek credit and publicity for their success in suicide prevention is rather selfish,” Mayor Aime Panimera wrote, “particularly when they do not provide the facts, but speculations, as a means of media attention. As an Inuk, I would even say that it is unethical, particularly dealing with sensitive and tragic issues, such as suicide” (Panimera, 1999).

Figure 2. Suicide rate for the Inuit populations of the Kitikmeot region. Source: Author’s calculations based on data obtained from the Office of the Chief Coroner of Nunavut and the Nunavut Bureau of Statistics. Five-year rolling averages.

Figure 2. Suicide rate for the Inuit populations of the Kitikmeot region. Source: Author’s calculations based on data obtained from the Office of the Chief Coroner of Nunavut and the Nunavut Bureau of Statistics. Five-year rolling averages.

Figure 3. Suicide rates for the Inuit populations of selected Nunavut communities. Source: Author’s calculations based on data obtained from the Office of the Chief Coroner of Nunavut and the Nunavut Bureau of Statistics. Five-year rolling averages. Data for Canada from Statistics Canada, CANSIM Table 102-0551.

Figure 3. Suicide rates for the Inuit populations of selected Nunavut communities. Source: Author’s calculations based on data obtained from the Office of the Chief Coroner of Nunavut and the Nunavut Bureau of Statistics. Five-year rolling averages. Data for Canada from Statistics Canada, CANSIM Table 102-0551.


The Working Group For A Suicide Prevention Strategy For Nunavut5 considered the work of Chandler and Lalonde, and of Kral and his colleagues, as it developed the framework for the multi-faceted, evidence-informed 2010 Nunavut Suicide Prevention Strategy (which is being implemented by the 2017 Inuusivut Anninaqtuq action plan.)

The Working Group concluded that Chandler and Lalonde’s conclusions made no sense in a Nunavut context. Their approach could explain neither the sharp rise in the rate of death by suicide by Nunavut Inuit over time nor persistent variations in suicide rates across the territory’s 25 communities.

The Working Group rejected Kral’s claims of having participated in suicide prevention miracles, was critical of his failure to acknowledge unresolved historical trauma as a risk factor, and completely disagreed with his dismissal of the need for culturally-appropriate mental health services in the communities.

In both the analyses by Chandler and Lalonde and by Kral, the data used to construct their arguments was a part of the problem. Chandler and Lalonde had a limited number of social variables that they could quantify, and had neither longitudinal data nor data on suicide attempts. The low number of deaths by suicide in Igloolik reported by Kral appear fabricated, differing sharply from the official statistics released by the Office of the Chief Coroner of Nunavut.

It is easy, but misleading, to attribute a short-term variation in the suicide rate of a small community or region to an intervention that a researcher is personally invested in. For example, Figure 2 shows that the Inuit suicide rate in the Kitikmeot region has fluctuated considerably in recent decades, and not because of targeted suicide prevention measures (as there were none that were systematically or longitudinally implemented).

Someone in the Kitikmeot could have claimed, in 1991 or 2005, that a short-term change in the suicide rate was evidence that an intervention they had been involved in had made a significant difference. But just a few years later, the picture would have changed substantially.

One advantage that Nunavut has over many other jurisdictions is a coronorial system that produces data on death by suicide at the community level, disaggregated by age, sex and ethnicity, on an annual basis. This allows us to closely track trends at the community level, as evidenced by Figure 3. There is no excuse for using partial data, or using “alternate facts.”

It would help if we had data of similar quality on suicide attempts, as other jurisdictions do, since attempts are more frequent (and usually higher among women than men) and are also a significant risk factor for later completed suicide. However, the Government of Nunavut’s Department of Health has not yet been able to reliably collect or produce these data.6

Having accurate community-level data on suicide behaviour is critical to understanding variations and changes in suicide rates. But as this review of the research of Chandler and Lalonde, as well as Kral, has demonstrated, questionable methodologies can lead to dubious conclusions.

Taken together, these bodies of writing form an “Indigenous exceptionalist” approach to understanding suicide risk and suicide prevention in Indigenous communities, one which views suicide behaviour by Indigenous peoples as being fundamentally different than suicide behaviour by non-Indigenous peoples. They stand apart from the global evidence base on suicide behaviour and suicide prevention, as summarized in Goldney (2013) and WHO (2014). ◉

Jack Hicks is Adjunct Professor of Community Health and Epidemiology, College of Medicine, University of Saskatchewan. He served as the Government of Nunavut’s Suicide Prevention Advisor during the development of the Nunavut Suicide Prevention Strategy.

Correction: The original print edition of this article mistakenly omitted the author’s endnotes (“notes”) and provided an incorrect reference to Kral (2016).

Canada, House of Commons, Standing Committee on Indigenous and Northern Affairs. 2016 June 14. Evidence. http://www.ourcommons.ca/DocumentViewer/en/42-1/INAN/meeting-20/evidence
Chandler, Michael J., and Christopher E. Lalonde. 1998. “Cultural continuity as a hedge against suicide in Canada’s First Nations.” Transcultural Psychiatry 35:2, 191-219. http://firstnationcitizenship.afn.ca/uploads/A12_Cultural_Continuity_as_a_Hedge_against_Suicide.pdf
—–. 2008. “Cultural continuity as a moderator of suicide risk among Canada’s First Nations.” in: Healing Traditions: The Mental Health of Aboriginal Peoples in Canada. (eds.) Kirmayer, Laurence, and Gail Valaskakis. University of British Columbia Press. pp. 221-48.
Goldney, Robert D. 2013. Suicide Prevention (2nd. Ed.). Oxford University Press. https://global.oup.com/academic/product/suicide-prevention-9780199677580?cc=ca&lang=en&
Hicks, Jack. 2015. “Statistical data on death by suicide by Nunavut Inuit, 1920 to 2014.” Report prepared for Nunavut Tunngavik Inc. http://www.tunngavik.com/blog/2015/09/15/pdf-statistical-data-on-death-by-suicide-by-nunavut-inuit-1920-to-2014/
Journal of the American Academy of Child and Adolescent Psychiatry. 2017. “Erratum.” Journal of the American Academy of Child and Adolescent Psychiatry 56:9, 800. http://www.jaacap.com/article/S0890-8567(17)30329-5/fulltext
Kral, Michael J. 2016. “Indigenized suicide prevention among Inuit in the Canadian Arctic.” Journal of the American Academy of Child and Adolescent Psychiatry 55:10, Supplement, S2. http://www.jaacap.com/article/S0890-8567(16)30786-9/fulltext
—–, Patricia K. Wiebe, et al. 2009. “Canadian Inuit community engagement in suicide prevention.” International Journal of Circumpolar Health 26:3, 292-308. http://www.tandfonline.com/doi/pdf/10.3402/ijch.v68i3.18330
Nunavut, Government of, Nunavut Tunngavik Inc., Royal Canadian Mounted Police, and Embrace Life Council. 2010. Nunavut Suicide Prevention Strategy. http://www.tunngavik.com/files/2010/10/2010-10-26-nunavut-suicide-prevention-strategy-english1.pdf
—–. 2017. Inuusivut Anninaqtuq Action Plan 2017-2022. http://inuusiq.com/wp-content/uploads/2017/06/Inuusivut_Anninaqtuq_English.pdf
Panimera, Aime. 1999 November 5. “Igloolik mayor clarifies suicide situation,” Letter to the Editor, Nunatsiaq News. http://www.nunatsiaqonline.ca/stories/article/igloolik_mayor_clarifies_suicide_situation/
Pedwell, Terry. 2016 April 19. “Governments “only beginning” to realize scope of mental-health struggles, says Justin Trudeau.” Toronto Star. https://www.thestar.com/news/canada/2016/04/19/governments-only-beginning-to-realize-scope-of-mental-health-struggles-says-justin-trudeau.html
Santhanam, Laura. 2016 April 21. “How Canada is working with indigenous peoples to prevent suicide.” PBS Newshour. Accessed April 21, 2016. http://vaunter2.rssing.com/chan-1732028/all_p800.html [Note: The text of this online article was later modified.]
Tatz, Colin. 2011. Sport and Youth Suicide: Submission to the Select Committee on Suicide in the Northern Territory. https://parliament.nt.gov.au/__data/assets/pdf_file/0006/366378/Sub-No.-04,-Prof.-Colin-Tatz-AO,-23-Sept-11.pdf
World Health Organization. 2014. Preventing Suicide: A Global Imperative. http://www.who.int/mental_health/suicide-prevention/world_report_2014/en/

1. In this case, “Land claims” is the binary variable and “Whether or not each First Nation community has taken ‘early steps to actively secure title to traditional lands.’” is the indicator used to create a yes/no response.
2. In a later iteration of their work, they developed additional binary variables: the participation of women in local governance, and the provision of child and family services within the community (Chandler and Lalonde, 2008).
3. A rate about 20% higher than the rate for Inuit in Nunavut in recent years.
4. A similar sports-related anecdote was recently presented as fact to the House of Commons’ Standing Committee on Indigenous and Northern Affairs by psychiatrist Cornelia Wieman, who said that she “recently heard of a small, anecdotal study that came from Nunavut where a community with very high rates of suicide didn’t necessarily do anything too extreme in bringing in crisis teams or getting a psychiatrist to visit that community. The community built a skating rink, and the suicide rates in that community went down” (Canada, 2016, June 14). Questioned about her claims by the author, Wieman was unable to identify the community where this was supposed to have occurred, the time period in question, or any other details about the supposed events.
5. A partnership between the Government of Nunavut, Nunavut Tunngavik Inc. (specifically its Social and Cultural Development division), the Royal Canadian Mounted Police (RCMP)’s ‘V’ Division and the Embrace Life Council.
6. The Nunavut Department of Health has thus far failed to deliver on the commitment it made in the first action plan for the Nunavut Suicide Prevention Strategy. However, one of the Government of Nunavut’s recommendations to the Coroner’s Jury at the 2015 Coroner’s Inquest into suicide in Nunavut was that “The Department of Health shall develop a territory-wide surveillance (information gathering) system to capture suicide attempts data from the health charts by March 2016, with automation of the data to be assessed by April 2017.” The Coroner’s Jury accepted this recommendation.

Download Article