Health Reports
Cancer in First Nations people in Ontario, Canada: Incidence and mortality, 1991 to 2010
by Sehar Jamal, Carmen Jones, Jennifer Walker, Maegan Mazereeuw, Amanda J. Sheppard, David Henry, and Loraine D. Marrett
DOI: https://www.doi.org/10.25318/82-003-x202100600002-eng
There are three distinct Indigenous groups in Canada: First Nations, Inuit and Métis (FNIM).Note 1 They have unique histories, languages, cultural practices and beliefs. First Nations comprise the largest of these groups, with over 230,000 people residing in Ontario (about 2% of the provincial population and about 24% of the national First Nations population).Note 2 Many First Nations people face significant barriers to achieving good health, stemming in part from colonialism, racism and social exclusion.Note 3 As a result, many experience poorer health compared with the general population (e.g., lower life expectancy, higher avoidable mortality).Note 4 Other factors may also contribute to poorer health outcomes, including geographic challenges that limit accessibility to health services, affordable food and education.Note 3
Cancer is among the leading causes of death in First Nations people.Note 5 However, a lack of Indigenous identifiers in health administrative databases, including cancer registries, has limited the measurement of health outcomes of First Nations people and the required conducting of special studies. Through record linkage, recently accumulated evidence has demonstrated differing patterns of cancer incidence in Indigenous populations compared with non-Indigenous populations. In Canada, it has been shown that First Nations people have a higher incidence of colon and rectum, kidney, cervix, and liver cancers, and a lower incidence of prostate, breast, bladder, uterus and brain cancers, as well as non-Hodgkin lymphoma, leukemia and melanoma.Note 6Note 7Note 8
In Ontario, Canada’s most populous province, available data were very out of date; coverage was only as recent as 1991.Note 7 These data showed rising cancer incidence rates in First Nations people in Ontario, especially for lung and colorectal cancers.Note 7 The purpose of this study is to present updated cancer burden information for First Nations people living in Ontario and to look at trends over time, particularly in cancer rates that are targeted by screening programs (cervical, breast, colorectal) or public health interventions (smoking).
Data and methods
As mandated by the Ontario First Nations Chiefs in Assembly in 2009, three organizations (Chiefs of Ontario, ICES and Ontario Health (Cancer Care Ontario)) collaborated on the development of a strategy to estimate cancer burden in First Nations people in Ontario. Chiefs of Ontario is the secretariat for the 133 First Nations communities in Ontario and led the effort in gaining access to the federal Indian Register (IR). ICES is an independent data research organization in Ontario that holds, maintains and uses the IR under authority from Chiefs of Ontario. Ontario Health (Cancer Care Ontario) is the government’s cancer advisor and was responsible for analyzing and interpreting the cancer data.
The project strategy involved linking multiple databases to create a First Nations cohort, along with developing data governance and sharing agreements that detailed how the organizations would collaborate to manage data under the First Nations data-sovereignty principles of ownership, control, access and possession (OCAP®).Note 9Note 10 The Cancer Surveillance Working Group was formed to guide project development and implementation. Results were initially delivered to First Nations communities in Ontario using a knowledge translation and exchange framework with support from a First Nations Knowledge Translation Specialist.
Cohort creation
The Cancer Surveillance Working Group submitted an application to Indigenous Services Canada requesting access to the IR for the years 1991 to 2010 (“study period”). The IR is the official record of all “status Indians” (“Indians” are now referred to as First Nations people) who meet defined eligibility criteria under the federal Indian Act.Note 11 A copy of the file with selected IR variables was sent to ICES, a prescribed entity under the Personal Health Information Protection Act of Ontario that is legally able to receive and link files with individual identifying information. ICES acted as data custodian on behalf of the First Nations in Ontario for the purpose of this project.Note 12 ICES created a cohort of Registered First Nations people (aged 0 to 99) living in Ontario (“First Nations cohort”) by linking the IR file with the Ontario Registered Persons Database (RPDB); the latter includes all Ontario residents entitled to publicly funded health care since 1991.Note 12 Ontario residents must have their health cards updated every five years; the RPDB includes specific years of health coverage and, therefore, assumed residence in the province. The linkage was performed using a combination of deterministic and probabilistic methods.Note 12 The First Nations cohort thus included all First Nations people in the IR (irrespective of the location of their band) who were also in Ontario’s RPDB (and therefore assumed to be Ontario residents) for at least one year between 1991 and 2010. These datasets were linked using unique encoded identifiers and analyzed at ICES. Analysis was also done at Ontario Health (Cancer Care Ontario).
Cancer cases and deaths
The Ontario Cancer Registry (OCR) includes information on all diagnoses of invasive cancer in Ontario residents since 1964, excluding non-melanoma skin cancers.Note 13 Cancers were classified according to the International Classification of Diseases for Oncology, Third Edition.Note 14 Vital status and cause of death were available in the OCR. Deaths were classified according to the International Classification of Diseases, 10th Revision.Note 15 The First Nations cohort was linked with the OCR using a combination of deterministic and probabilistic methods. Cohort members who matched with the OCR were First Nations people with cancer or dying from cancer; all other people diagnosed with or dying from cancer (i.e., non-Registered First Nations, other Indigenous and non-Indigenous people) over the same time period were classified as “other people in Ontario”.
Statistical analysis
Person-years at risk for First Nations people were calculated from the first year that an individual appeared in the First Nations cohort file to the last year they were determined to be an Ontario resident, the year of cancer diagnosis, the year of death, or the end of the follow-up period (December 31, 2010), whichever was earliest. Annual population estimates for First Nations people in Ontario were created from the First Nations cohort, based on specific years of provincial residence imputed from the RPDB. Person-years for other people in Ontario were calculated using Statistics Canada censuses from 2001, 2006 and 2011; intercensal estimates; and Ontario Ministry of Finance projections (Fall 2014). Annual population estimates for “other people in Ontario” were formed by subtracting the number of people in the First Nations cohort from annual population estimates for all of Ontario. Population sizes were estimated by age and sex, as well as calendar year.
Sex- and site-specific cancer incidence and mortality rates were calculated as the respective number of cases and deaths per 100,000 person-years, age-standardized to the World Standard Population.Note 16 Cancer types were grouped according to the Surveillance, Epidemiology, and End Results Program site recode.Note 17 Rates are presented for age, sex and cancer combinations with 30 or more cases or deaths. The three most common cancer types for each sex were further stratified by age. Estimates for First Nations people and other people in Ontario were compared using the ratio of age-standardized rates, for which 95% confidence intervals (CIs) were calculated.Note 18 Rates between the two populations were considered to differ significantly if rate ratio (RR) CIs excluded 1. These analyses were performed using SEER*Stat software (version 8.3.2).
Incidence rates for the six most common cancer types in each sex were also estimated by year of diagnosis. For cervical and kidney cancers, incidence rates are presented in groups of two and three years, respectively, because of small numbers. Trend lines were produced, and annual percent change (APC) estimates and 95% CIs were calculated using Joinpoint software (version 4.2). The model recommended by Joinpoint was used, with a maximum of three joinpoints.
Results
The First Nations cohort comprised 194,392 individuals, among whom 6,859 cancers were diagnosed during 3.1 million person-years of follow-up between 1991 and 2010. Of other people in Ontario, 1,069,446 cases of cancer were diagnosed during more than 246 million person-years at risk. The First Nations cohort was younger than other people in Ontario. Only 15% of First Nations people were older than 50, compared with over 25% of other people in Ontario. Table 1 shows the number of new cancer cases, age-standardized incidence rate (ASIR) and RR by cancer type and sex.
Cancer incidence for all sites combined was significantly higher for First Nations females compared with other females in Ontario (RR=1.05, 95% CI: 1.02 to 1.09), although incidence was significantly lower for females younger than 30; it was significantly higher for those aged 50 to 64 and 65 to 74 (Table 1). The all-sites ASIR for First Nations males was similar to that of other males in Ontario, although it was significantly lower for age groups younger than 50.
The most commonly occurring cancers in First Nations males were prostate, colorectal, lung and kidney cancers. In First Nations females, the most common were breast, lung, colorectal, cervix and kidney cancers. First Nations males and females had significantly higher incidence of colorectal, lung, kidney and liver cancers compared with other people in Ontario. First Nations females also had significantly higher incidence of cervical cancer, myeloma, stomach, gallbladder and vulvar cancers. First Nations males had significantly lower incidence of prostate, bladder and brain cancers, as well as non-Hodgkin lymphoma, leukemia and melanoma of the skin. First Nations females had significantly lower incidence of breast, uterus, thyroid and brain cancers, as well as melanoma of the skin.
Colorectal and lung cancer rates were significantly higher in First Nations males 50 or older compared with other males of the same age group in Ontario. Among First Nations females, colorectal cancer was significantly higher in all age groups younger than 75, and lung cancer was significantly higher in all age groups, compared with other Ontario females in those respective age groups. Kidney cancer RRs were significantly higher in First Nations males and females in all age groups younger than 75 compared with other males and females in Ontario younger than 75. Cervical cancer rates were significantly higher for all age groups in First Nations females, excluding those aged 50 to 64, compared with other females in Ontario.
Figures 1 to 5 illustrate fitted time trends for incidence rates of the most common cancers from 1991 to 2010. Rates increased significantly for lung cancer in First Nations females (p < 0.05), compared with other females in Ontario, whose rates remained relatively stable, thereby diverging over time. Lung cancer rates decreased in both First Nations and other males in Ontario. Colorectal cancer rates in First Nations females and males remained higher than for other people in Ontario, with no significant upward or downward trends. Although breast cancer incidence was lower in First Nations females, rates increased steadily and approached those of other females in Ontario (APC 1.19; APC -0.09). Prostate cancer rates were stable in First Nations males; other males in Ontario experienced increasing rates until 2006.
Kidney cancer rates in First Nations females notably increased over time compared with other females in Ontario (APC 17.43; APC 4.13). First Nations males had higher rates than other males in Ontario but exhibited a declining rate until 2003 (APC -9.64), after which rates rose rapidly compared with males in Ontario, potentially because of small numbers (APC 38.39; APC 3.26). Cervical cancer incidence in First Nations females decreased significantly (APC -9.53), converging with the rate for other females in Ontario.
Mortality rates for all sites combined were significantly higher among First Nations people (especially in females and older age groups) compared with other people in Ontario (males: RR=1.08, CI: 1.03 to 1.14; and females RR=1.23, CI=1.16 to 1.29) (Table 2). Mortality rates were significantly elevated for lung and colorectal cancer in First Nations males and females aged 30 or older, and in all ages for kidney cancers, compared with other people in Ontario. Deaths from cervical cancer (30 years or older) were almost three times higher in First Nations females cwompared with other females in Ontario.
Discussion
For several cancer types, incidence and mortality rates were higher in First Nations people compared with other people in Ontario. Similar findings have been reported for Indigenous populations in other parts of Canada (British Columbia, Alberta, Quebec) and the United States, Australia and New Zealand.Note 6Note 8Note 19Note 20 Rising incidence rates of kidney cancers in First Nations males and females, and lung cancer rates in First Nations females, were observed. Rates of lung cancer incidence among First Nations males declined over time. Encouragingly, a fall in cervical cancer incidence was also documented, resulting in a rate close to that of other Ontario females by the end of the study period. These positive trends may be a result of measures taken to reduce smoking and encourage cervical screening in First Nations communities.Note 21
The incidence for all cancers combined was similar in First Nations males compared with other males in Ontario. However, rates of all cancers were higher in First Nations females compared with other females in Ontario. This finding, as well as the observation of higher cancer mortality rates, highlights the particular vulnerabilities of First Nations females.
Lung cancer was the most common cancer in First Nations people in Ontario, affecting 1,059 First Nations people between 1991 and 2010; incidence was significantly elevated in both sexes and across most age groups. Higher incidence of lung cancer was also found in the Indigenous populations in New Zealand, the United States (Alaska) and Queensland compared with the respective general populations.Note 19 However, lung cancer incidence was lower among First Nations people in British Columbia and Alberta compared with non-First Nations people.Note 6Note 19 In this study, lung cancer incidence declined among First Nations males (paralleling the trend of other Ontario males) but increased in First Nations females. Higher lung cancer rates in First Nations people may be expected in the future, given the current high prevalence of smoking in First Nations youth compared with non-Aboriginal youth.Note 21 Smoking cessation and prevention programs remain crucial to prevent or reduce commercial tobacco consumption among Indigenous youth in Ontario. The Indigenous Cancer Care Unit (ICCU) within Ontario Health (Cancer Care Ontario) has developed an Indigenous Tobacco Program to collaborate with communities for the purpose of enhancing knowledge and building capacity through culturally sensitive approaches to address commercial tobacco cessation and prevention.
Colorectal cancer incidence was substantially higher among First Nations people than other people in Ontario. Higher colorectal cancer incidence has also been observed in First Nations people in other regions of Canada.Note 6Note 19 By contrast, colorectal cancer incidence was lower in Indigenous populations in New Zealand and Australia, although mortality was significantly higher and survival was poorer.Note 22Note 23 Specific to the findings of this study was that the risk of colorectal cancer in younger age groups (younger than 50 years) was higher in First Nations people, but only statistically significant in females. In Ontario, colorectal cancer screening guidelines suggest screening every two years for adults aged 50 to 74 who are at average risk.Note 24 However, the higher level of risk and earlier age of onset in First Nations people suggest that screening guidelines may need to be reconsidered for this population. This has been done for African Americans and Alaska Natives through organizations in the United States.Note 25Note 26 More research on the apparently younger age of diagnosis for colorectal cancer is needed in Indigenous populations.Note 27 First Nations have a higher prevalence of obesity and a lower consumption of fruits and vegetables compared with non-Aboriginal Indigenous people—strong risk factors for colorectal cancer.Note 21 In self-reported surveys of age-eligible First Nations people, 39% of females and 53% of males were under-screened for colorectal cancer.Note 21
Historically, and as observed in this study, cervical cancer incidence has been significantly higher in First Nations females than in the non-Indigenous population.Note 7 However, from 1991 to 2010, incidence has decreased and converged with that of the Ontario population. Higher cervical cancer incidence has also been observed in Indigenous populations in other Canadian jurisdictions, as well as in New Zealand, Australia and parts of the United States (Alaska).Note 6Note 19 Based on self-reported data, First Nations women and non-Aboriginal women had similar Pap test uptake in more recent years,Note 21 which may have contributed to the strong downward trend of cervical cancer incidence in First Nations females and its convergence with that of other females in Ontario.
Breast cancer incidence has historically been lower among First Nations women in Canada, although it has increased since the 1980s.Note 7Note 8 Lower breast cancer incidence has also been observed in most Indigenous populations outside Canada, except in Alaska Natives in the United States and Maori females in New Zealand.Note 19 While breast cancer incidence was lower in First Nations people in Ontario from 1991 to 2010, it has increased over time and approaches that of other females in Ontario. This increase may be related to changes in risk factors such as diet and physical activity levels, or initiatives to improve screening (e.g., mobile screening bus or coach that made mammography more accessible). More research on understanding the increase in breast cancer among First Nations females is important to reverse its rise.
Kidney cancer was the fifth most common cancer affecting First Nations people. By contrast, it was the 12th most common cancer in other people in Ontario; however, it increased over time in both populations. High rates of kidney cancer have also been reported in Alaska Natives, who are more likely to be diagnosed with late-stage disease compared with White people.Note 28 Higher kidney cancer incidence was also observed in First Nations females in British Columbia.Note 6 Some partial explanations for increased kidney cancer incidence may include improved diagnostic imaging and higher prevalence of risk factors, including cigarette smoking, obesity, diabetes and hypertension. In the United States, over 40% of renal cell carcinomas are attributed to overweight and obesity.Note 28 Follow-up research is urgently needed, especially because of the steep increasing trend among First Nations people relative to other people in Ontario.
Overall cancer mortality was significantly higher in First Nations people compared with other people in Ontario; higher mortality from lung, colorectal, cervical and kidney cancers contributed to this. Previous studies of a similar yet older cohort from 1968 to 1991, as well as national cohorts in Canada, indicated survival was worse for First Nations people for many cancer types.Note 29Note 30 This was possibly explained by a higher prevalence of comorbidity and later-stage diagnosis.Note 31 Poorer access to medical care may lead to later presentation and impaired response to treatment.
The higher incidence and mortality of certain cancers may be due to the higher prevalence of some behavioural risk factors (e.g., cigarette smoking, obesity, lower fruit and vegetable consumption) in First Nations people in Ontario.Note 21Note 32Note 33Note 34 These differences in risk factors are rooted in the determinants of Indigenous peoples’ health, which also include historical trauma, marginalization and poor access to the healthcare system.Note 3 Equitable access to a conducively built environment, affordable healthy foods, and a culturally safe and respectful health system are important in addressing the differences in the cancer experience and risk of First Nations people compared with others in Ontario. Indigenous Navigators support First Nations people and other Indigenous groups in Ontario make their way through the cancer system, providing access to Indigenous healers and well-being resources, as well as participating in medical appointments.Note 35
Policies and programs must take a holistic approach to modifying the economic, physical and social environments that influence health-related behaviours.Note 4 Recommendations for such policies in Ontario have been made through extensive outreach and engagement with First Nations communities.Note 36
Strengths and limitations
The work involved close collaboration between the Chiefs of Ontario, Ontario Health (Cancer Care Ontario) and ICES. The work was led by First Nations and conducted according to First Nations OCAP® principles.Note 10 The First Nations Cohort included information about all Registered First Nations people in Ontario who were linked to the RPDB (a database that identifies Ontarians who have accessed health care). Subsequent linkage to a population-based cancer registry (OCR) provided details regarding cancer diagnoses between 1991 and 2010. This enabled estimation of incidence for over 15 cancer types, with relatively stable estimates of incidence over a 20-year time period, including trends for more common cancers. Data linkage was completed by experienced staff using both deterministic and probabilistic methods.Note 12 This study has demonstrated that through partnership there is an opportunity to continue First Nations cancer surveillance using high-quality data to help inform cancer programs and funding for First Nations in Ontario.
Generalizability is limited to First Nations people who are registered under the Indian Act; according to the 2016 Census, at least 36% of self-identified First Nations people in Ontario are not registered.Note 37 Diversity among First Nations people across different communities in Ontario is not reflected in the results of this study.
Conclusions and impact
Overall, colorectal and kidney cancer incidence and mortality in First Nations people were notably higher than those of other people in Ontario. Conversely, lung cancer incidence in First Nations males and cervical cancer incidence in First Nations females have been declining. The declines in incidence demonstrate the success of efforts to enhance prevention policies (e.g., smoking cessation) and programs (e.g., cancer screening) in First Nations communities. Developing culturally relevant programs in collaboration with First Nations communities may be a successful way to reduce the prevalence of risk factors and improve screening. The ICCU at Ontario Health has developed policy recommendations to reduce population-level exposure to four key risk factors (tobacco, alcohol, diet and physical activity), to build capacity for chronic disease prevention and to work toward health equity.Note 36 The education and prevention programs, research, and policy changes supported by the ICCU have been developed through strong partnerships among First Nations communities and organizations. Continued partnerships are needed to leverage First Nations health data to measure the entire cancer journey, including cancer screening and follow-up, to further understand and inform decisions.
- Date modified: