Incidence of breast and colorectal cancer among immigrants in Ontario, Canada: a retrospective cohort study from 2004-2014

Background Studies have shown that morbidity and mortality rates due to cancer among recent immigrants are lower than those among the native-born population. The objectives of this study were to describe the incidence of colorectal and breast cancer among immigrants from major regions of the world compared to Canadian-born residents of the province of Ontario and to examine the role of length of stay and neighborhood income. Methods Retrospective cohort study including all individuals 18 years and over residing in Ontario from 2004 to 2014. Age-standardized incidence rates (ASIR) were calculated for immigrants from each world region versus Canadian-born residents and stratified by neighborhood income quintile and length of stay. Binomial regression analysis was used to determine the association of neighbourhood income, length of stay, and location of birth with colorectal and breast cancer incidence. Results Canadian immigrants born in South Asia had the lowest colorectal and breast cancer incidence (colorectal: women: ASIR = 0.14; men: ASIR = 0.18; breast: ASIR = 1.00) compared to long-term residents during the study period (colorectal: women: ASIR = .57; men: ASIR = .72; breast cancer ASIR = 1.61). In multivariate analyses, neighboorhood income did not consistently play a significant role in colorectal cancer incidence; however higher neighbourhood income was a risk factor for breast cancer among immigrant women (adjusted relative risk for highest neighboorhood income quintile versus lowest income quintile =1.21, 95% CI = 1.18–1.24). Increased length of stay was associated with higher risk of cancer. After adjusting for age, neighborhood income, and length of stay, those born in Europe and Central Asia had the highest risk of colorectal cancer compared to those born in East Asia and Pacific and those born in the Middle East had the greatest additional risk of breast cancer. Conclusions After correcting for age, breast and colorectal cancer incidence rates among immigrants differ according to their region of birth and recent immigrants to Ontario have lower colorectal and breast cancer incidence rates than their native-born peers. However, those advantages diminish over time. These findings call for Ontario to generate tools and interventions to maintain the health of the immigrant population, particularly for those groups with a higher incidence of cancer.


Background
Immigrants represent a large, increasing and vital segment of the Canadian population. Most Canadian studies have shown that morbidity and mortality rates of chronic disease among recent immigrants are lower than those among the general Canadian population suggesting that immigrants enjoy the "healthy immigrant effect" whereby they are in better physical condition on arrival than host country inhabitants as a result of selective migration [1][2][3][4].
Ontario, the largest province in Canada, has a large and diverse immigrant population and approximately one-third of the population in Ontario is foreign-born [5]. Ontario also has a provincial cancer registry that includes data on all residents diagnosed with cancer and universal public health care coverage, thus making it an ideal location to explore cancer incidence among immigrants at the population level. However, there has been little recent research examining cancer incidence in the immigrant population [6].
Colorectal cancer is the third most common cancer diagnosed in men and women with 26,800 cases per year in Ontario. Breast cancer is the most common cancer among women with 26,300 cases a year in Ontario [7]. Also, colorectal and breast cancer have clear provincial screening guidelines and are often used as indicators for population health [8]. Therefore, the overall objective of this study was to examine how the incidence of colorectal and breast cancer among immigrants from major regions of the world compare to Canadian-born residents of Ontario. We also explored the roles of age, gender, socioeconomic status and time in Canada.

Methods
The following datasets were linked using unique encoded identifiers and analyzed at the Institute for Clinical Evaluative Sciences (ICES). The Registered Persons Database was used to identify people aged 18 years and over in the province of Ontario eligible for health care. The Registered Persons Database contains basic demographic information for those who have ever received an Ontario health card number for the province's universal health care system (overall linkage rate = 96.5%). All citizens and permanent residents are eligible for health care. The second database was the Immigration Refugee and Citizenship Canada (IRCC) Database [9] which includes individuals who have landed immigrant or permanent resident status at any time from 1985 to 2014. Immigrants were defined as those identified in the IRCC Database, and long-term residents were defined as those not on the IRCC database (Canadian-born and immigrants who arrived before 1985). The IRCC database was also used to identify country of birth, and countries were further collapsed into eight regions, broadly defined according to the World Bank classification (1, Caribbean and Latin America; 2, East Asia and Pacific; 3, Eastern Europe and Central Asia; 4, Middle East and North Africa; 5, South Asia; 6, Sub-Saharan Africa; 7, USA, Australia, and New Zealand; and 8, Western Europe). Third, we identified incident breast and colorectal cancer cases by linking the cohort to the Ontario Cancer Registry from 2004 to 2014. The Ontario Cancer Registry is a passive surveillance patient registry that links data from hospitals, cancer centers and pathology laboratories; incidence data has been previously assessed as having approximately 92% completeness [10]. The Canadian Census was used for calculating the Canadian population standard.

Covariates
Using the postal-code conversion file [11], ecologicallevel measures of income status were estimated using data from the 1996, 2001 and 2006 Canadian census and applied to individual cases according to the dissemination area where the individual resided. Dissemination areas are the smallest geographic census unit for which census data are available, and are uniform in population size, which is targeted from 400 to 700 persons. Individuals were then grouped into income quintiles ranging from 1 (20% lowest income) to 5 (20% highest income). Length of stay was measured by calculating the time since immigration until December 31, 2014 or cancer incidence.

Analysis
The age-standardized annual incidence rates (ASIR) were calculated using the 2010 Canadian population as standard, for long-term residents, for immigrants, and then by world region of origin for immigrants. To assess the effect of neighboorhood income and length of stay in Canada, ASIR were stratified by time since immigration 0-5 years, 6-10 years and 11+ years) and by neighborhood income quintile (1 through 5).
Predictors of breast and colorectal cancer incidence among all residents in Ontario, 2004-2014 were assessed by two binomial regression models, one among the entire cohort and one among only immigrants. Among the entire cohort, predictors entered into the model included age, place of birth and neighborhood income quintile. The second model calculated among only immigrants assessed the effect of age, neighborhood income quintile, length of stay and region of birth. The analyses produced adjusted rate ratios (RR) with corresponding confidence intervals (CI). Statistical significance was determined at the 0. 05 level. All analyses were conducted using SAS statistical software, version 9.4. This study was approved by the institutional review board at Sunnybrook Health Sciences Centre, Toronto, Canada.

Results
Demographic characteristics of the study population are shown in Table 1. Immigrants were younger than long-term residents on average: mean age ranged from 40.2 ± 13.7 for Sub-Saharan Africa to 44.7 ± 15.8 for East Asia and Pacific, whereas longterm residents' mean age was 47.5 ± 2. Those that were born in the Middle East and North Africa had spent the least amount of time in Canada on average (10.7 ± 6.0 years), and those born in Europe and Central Asia had spent the longest amount of time (14.2 ± 6.9). Sub-Saharan Africa, followed by Latin America and the Caribbean, had the greatest percentage of immigrants living in the lowest income quintile ( Table 1).

Place of birth
Age-standardized incidence rates varied by region, with long-term residents consistently having the highest rates and South Asian immigrants consistently having the lowest rates of colorectal and breast cancer (Fig. 1). Among immigrants, incidence of colorectal cancer was highest among Europe and Central Asia for men (ASIR = 0.65) and females (ASIR = 0.51) (Fig. 1). Among women born outside of Canada, the highest ASIR for breast cancer was among those from Middle East and North Africa (ASIR = 1.49, Fig. 1).

Length of stay and income
ASIR were also examined based on length of stay and neighborhood income quintile. We found that the ASIR were not associated with neighborhood income quintile for females and males with colorectal cancer ( Fig. 2a and Fig. 2b). Standardized incidence rates of breast cancer increased for higher neighborhood income quintiles for those born in Europe and Central Asia, South Asia, Sub-Saharan Africa, and New Zealand, Australia and the United States, but did not show trends for the remaining regions (Fig. 2c). There were no clear patterns seen for length of stay for both colorectal and breast cancer in the descriptive analysis and advantages enjoyed by immigrants appeared to disappear after spending over 10 years in Canada for both colorectal and breast cancer incidence (not shown).
In the binomial regression analysis among both long-term residents and immigrants we found that, after controlling for age and neighborhood income, immigrants enjoyed a healthy immigrant effect and were at lower risk of breast and colorectal cancer compared to long-term residents ( Table 2). For colorectal cancer, those in the highest neighborhood  Table 2).
In binomial regression analysis among immigrants only, when controlling for age, neighborhood income, and place of birth, we found that the risk of colorectal and breast cancer increased for each additional five years that immigrants lived in Canada ( Table 3). Length of stay had the greatest effect on risk of breast cancer where risk increased 7% for each additional five years in Canada (p < 0.0001). After adjusting for age, neighborhood income, and length of stay, those born in Europe and Central Asia had the greatest additional risk of colorectal cancer compared to the reference group for this analysis of those born in East Asia and the Pacific. Those born in South Asia had the lowest risk for   those from Europe and Central Asia being at highest risk among immigrants of developing colorectal cancer (incidence among men = .65 and females =0. 51) and those from South Asia having the lowest rates (females = 0.14 and males = 0.18). Those from the Middle East and North Africa were at highest risk among immigrant women to develop breast cancer (incidence = 1.49) and those from South Asia were the lowest (incidence-= 1.00). Third, we saw that neighborhood income did not play a role in colorectal cancer incidence but that higher neighborhood income was a risk factor for breast cancer incidence among immigrant women (RR = 1.21 95% CI = 1.18, 1. 24). Upon examining place of birth, we saw large differences (49-264% difference) in age standardized rates between places of birth. In comparison to long-term residents the largest differences were for those born in South Asia who had the lowest rates for colorectal and breast cancer. These differences are most likely attributed to differences in their home countries where South Asia has among the lowest incidence rates compared to other regions for breast and colorectal cancer [12]. Age-standardized incidence rates for United States, New Zealand and Australia were lower for breast and colorectal cancer compared to long-term immigrants. This was surprising, as Canada is considered to be a historically high-risk area for colorectal and breast cancer, similar New Zealand, United States and Canada, reflecting similar dietary and lifestyle factors [13].
Neighborhood income did not play a significant role in colorectal cancer incidence. Those in the highest neighborhood income quintile had slightly lower rates of colorectal cancer in the regression model including all Ontario residents; however, no effect was seen when examining rates among only immigrants. In contrast, we saw that higher neighborhood income was a risk factor for breast cancer incidence, both unadjusted and in regression models.
Similarly, Canadian and American data have shown that women in neighborhoods with higher neighborhood incomes have a higher risk of developing breast cancer [14,15].
An important dimension of the healthy immigrant effect is that the immigrant advantage we found appeared to disappear after spending over 10 years in Canada for both colorectal and breast cancer incidence. In addition, in the regression model limited to immigrants we saw that risk of cancer increases (5-7%) for each additional five years in Canada for both colorectal and breast cancer. Similarly, researchers have previously found that among Ontarian immigrants, despite the original advantage with immigration, there is a steady decline in survival, and cancer-specific survival, over time [16]. Analysis of Statistics Canada's Longitudinal Survey of Immigrants to Canada showed a decline in selfassessed health, physical health, and mental health among immigrants as little as two years after arrival [17]. Some researchers believe that convergence in health outcomes may stem from the process of acculturation where immigrants begin to take on Canadian habits such as smoking, alcohol consumption and diet [18]. However, Canadian longitudinal data from representative surveys have shown that immigrants did not show higher rates of daily smoking initiation, however, they were much more likely than the Canadian-born population to have had a substantial weight gain since immigrating [19]. Though these habits may be influential, it is unlikely that they are responsible for changes in incidence occurring over the short time observed in this study.
Others maintain that worsening of health status is due to barriers to health services including lack of familiarity with the Canadian health system and language or cultural differences [20] that may lead to the underuse of preventative health screening and treatment of health problems. To that effect,   [21][22][23][24][25][26]. Similarly, research has shown that cancer incidence may increase in the first decade after immigration and it reaches the population level in the host country in 1-2 generations [27,28]. This large cohort study has examined cancer incidence among immigrants in Canada which has not been examined since the 1990s [6]. Due to accurate databases, universal health care and excellent linkage we could consider the effects of region of birth, neighborhood income and length of stay. However, there are several limitations that should be noted. We were not able to determine if this was a cohort effect where those that immigrated more recently were healthier than those who immigrated over ten years ago. Our analyses used the world-region-ofbirth; this method of grouping could be problematic as countries within any region are not homogeneous. Given the absence of individual-level income-related information, we linked residential postal codes to neighborhood -level income which may have misclassification errors in geocoding [11] in rural areas, however, according to our data 99% of all immigrants to Canada settle in urban areas. Additionally, immigrants may have lived in other countries other than their country of birth before they came to Canada, thus possibly reducing the significance of birthplace as a determinant. Finally, our analysis was not able to account for risk factors such as behaviour (i.e. smoking and alcohol) and stress for cancer.

Conclusions
Our analysis showed breast and colorectal cancer incidence rates among immigrants to Ontario, Canada are lower than residents and these rates differ according to region of birth, however, those advantages diminish after arrival. Results from this hypothesisgenerating research initiative hold significant immigration and health policy implications, and add further intricacy to the study of the social determinants of health. The results call for Ontario to generate tools and interventions to maintain the health of immigrant population.
Abbreviations ASIR: age-standardized incidence rates; CI: Confidence Intervals; RR: Rate ratios Funding This study is supported by the Institute for Clinical Evaluative Sciences, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. The opinions, results, and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by the Institute for Clinical Evaluative Sciences or the Ontario Ministry of Health and Long-Term Care is intended or should be inferred. Parts of this material are based on data and information provided by Cancer Care Ontario (CCO). The opinions, results, views, and conclusions reported in this paper are those of the authors and do not necessarily reflect those of CCO. No endorsement by CCO is intended or should be inferred. All inferences, opinions, and conclusions drawn in this paper are those of the authors, and do not reflect the opinions or policies of the data stewards. Immigration data was obtained from the Immigration, Refugees and Citizenship Canada database held at the Institute for Clinical Evaluative Sciences. Aisha Lofters is supported by a New Investigator Award from the Canadian Institutes of Health Research and as a Clinician Scientist by the University of Toronto Department of Family & Community Medicine. This study was funded by a Career Development Award in Cancer Prevention from the Canadian Cancer Society Research Institute (Aisha Lofters).

Availability of data and materials
The data that support the findings of this study are available from the Institute for Clinical Evaluative Sciences of but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of Institute for Clinical Evaluative Sciences.
Authors' contributions J.S: Conceptualization of study, data interpretation and preparation of manuscript. A.L: Conceptualization of study, data interpretation and manuscript review. L.Y.: Statistical analysis and manuscript review. All authors read and approved the final manuscript.