The Epidemiology of Diabetes among Immigrants to Ontario

by

Maria Isabella Creatore

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Institute of Medical Science University of Toronto

© Copyright by Maria I. Creatore (2013)

The Epidemiology of Diabetes Among Immigrants to Ontario Maria I. Creatore Doctor of Philosophy (PhD) in Epidemiology Institute of Medical Sciences University of Toronto 2013

Abstract

Type 2 Diabetes Mellitus (T2DM) prevalence is increasing globally with roughly 2.4 million people currently living with this condition in Canada. T2DM occurs more commonly in non-European ethnoracial groups, however the distribution of risk by age, sex, ethnicity and immigration status in Canada are not completely understood. The purpose of this thesis is to investigate the epidemiology of diabetes in an immigrant, multi-ethnic population using linked immigration and health data for the province of Ontario. The ultimate goal of this work is to generate information that can be used to design appropriate and effective targeted programs for diabetes prevention, management and control in order to reduce inequities in health. The principal findings of this work indicate that: 1) South Asians had a three-fold higher risk for developing diabetes as compared with people of European ethnicity and this disparity in risk was evident at a very young age; 2) The young age at diabetes onset experienced by many of our high-risk ethnic groups, including South Asians and people of African and Middle Eastern descent, suggest that ii

in order to capture an equivalent risk of disease, screening may be recommended up to 15 years earlier in these groups – which is not reflected in current screening guidelines; 3) Contrary to patterns seen in Western European populations, women belonging to many high–risk ethnicities had equivalent or, in some cases, higher risk than men; 4) Risk varied substantially across country and region of birth making broad definitions of race or ethnicity (eg. ‘Asian’ or ‘Black’) inappropriate. These findings emphasize the heterogeneity of risk experienced by different ethnoracial populations in Canada and suggest that targeted primary prevention programs aimed at young adults and adolescents belonging to high-risk ethnic groups may be warranted. In addition, screening guidelines may need to be updated to reflect the younger age at onset in these populations. Further research is necessary to identify culturally appropriate and effective programs to reduce diabetes risk and associated health problems in these populations.

iii

Acknowledgments There are many individuals whom I would like to acknowledge for their role in supporting me in the creation and completion of this thesis. I would like to thank my supervisor, Dr. Richard Glazier for his constant support and mentorship and his patience and faith in me over these many years. Throughout my PhD journey, juggling work, school, family and children, he was always flexible and ready to suggest creative solutions for how I could manage to continue with my studies while accommodating the rest of life. I would also like to thank my other committee members – Drs. Gillian Booth, Doug Manuel and Rahim Moineddin - for their thoughtful comments, guidance and support throughout this process. In particular I wish to thank Dr. Rahim Moineddin who patiently supervised and advised me on all my analyses, was always available and full of encouragement and from whom I have learned so much. I would also like to thank my friends and colleagues at the Centre for Research on Inner City Health (CRICH) at St.Michael’s Hospital and the Institute for Clinical Evaluative Sciences (ICES) for their encouragement and frequent morale boosts. In particular I would like to thank Flora Matheson, Peter Gozdyra, Jonathan Weyman, Jim Dunn, Chaim Bell, Joel Ray, Mohammad Agha, Anne-Marie Tynan, Donna Hoppenheim, Jane Polsky, Marcelo Urquia and Hadas Fischer. There are many people whose love, support and encouragement made this work possible. First and foremost I owe my gratitude to my wonderful husband whose unconditional love and support makes all things possible. My children, Jacob and Anna Sofia inspired me daily to work hard and do my best to make them proud. Eternal gratitude to my parents, Cheryl and Giuseppe Creatore, who have always had unwavering faith in me and have supported me in everything I have ever attempted. And loving thanks to the rest of my family: Karina, Bryan, Bill, Christine, Myra and Frank.

iv

Contributions

I was responsible for the design, analysis, manuscript writing and preparation however a number of people contributed to the completion of this thesis. Dr. Richard Glazier was my supervisor and provided feedback on study design, analysis, and interpretation of findings as well as feedback on the chapter and manuscript drafts. Committee members Dr. Doug Manuel and Dr. Gillian Booth provided feedback on study design, analysis, and interpretation of findings as well as feedback on the chapter and manuscript drafts. In addition to providing feedback on study design, interpretation and writing, committee member Dr. Rahim Moineddin supervised all statistical analyses. Alexander Kopp and Nadia Gunraj at the Institute for Clinical Evaluative Sciences prepared the datasets. Marie DesMeules and Sarah McDermott contributed to the acquisition of data and creation of the linked datasets. Drs. Jeffrey Johnson, Lorraine Lipscombe and Jan Hux acted as external reviewers of this work, and Dr. Johnson provided valuable written comments and suggestions. I would also like to acknowledge the Public Health Agency of Canada (PHAC) and Citizenship and Immigration Canada (CIC) for their contribution of the data and support of this study.

v

Table of Contents Table of Contents.................................................................................................... v List of Tables ......................................................................................................... ix List of Figures ........................................................................................................ xi List of Appendices .................................................................................................. xii

Chapter 1 Introduction ........................................................................................... 1 1.1 Thesis Overview .............................................................................................. 2 1.2 Background ...................................................................................................... 3 1.3 Theoretical Framework .................................................................................... 19 1.4 Research Questions and Rationale for the Objectives ..................................... 22

Chapter 2 Age and Sex Patterns of Diabetes Among Immigrants to Ontario ........ 27 Abstract .................................................................................................................. 28 2.1 Introduction ...................................................................................................... 30 2.2 Research Design and Methods ........................................................................ 31 2.2.1 Data Sources and Study Population ............................................. 31 2.2.2 Statistical Analysis ........................................................................ 33 2.3 Results ............................................................................................................. 34 2.3.1 Characteristics of the Study Population ........................................ 34 vi

2.3.2 Trends in Diabetes Prevalence ..................................................... 36 2.4 Discussion ........................................................................................................ 44

Chapter 3 Diabetes Screening Among Immigrants: A Population-Based Urban Cohort Study .......................................................................................................... 49 Abstract ................................................................................................................ 50 3.1 Introduction ...................................................................................................... 52 3.2 Research Design and Methods ........................................................................ 54 3.2.1 Study Population ........................................................................... 54 3.2.2 Study Outcomes ........................................................................... 55 3.2.3 Statistical Analyses ....................................................................... 57 3.3 Results ............................................................................................................. 59 3.3.1 Baseline Study Characteristics ..................................................... 59 3.3.2 Diabetes Screening ....................................................................... 61 3.3.3 Screening Efficiency ..................................................................... 61 3.3.4 Predictors of Diabetes Screening .................................................. 64 3.3.5 Undiagnosed Diabetes .................................................................. 67 3.4 Discussion ........................................................................................................ 69

Chapter 4 A Population-Based Study of Diabetes Incidence by Ethnicity and Age: Support for the Development of Ethnic-Specific Age Guidelines for Screening ..... 75 vii

Abstract ................................................................................................................ 76 4.1 Introduction ...................................................................................................... 78 4.2 Research Design and Methods ........................................................................ 80 4.2.1 Study Design ................................................................................ 80 4.2.2 Study Population ........................................................................... 80 4.2.3 Measures ...................................................................................... 82 4.2.4 Study Outcomes ........................................................................... 83 4.2.5 Analysis ........................................................................................ 84 4.3 Results ............................................................................................................. 87 4.4 Discussion ........................................................................................................ 96

Chapter 5 Discussion ............................................................................................ 104 5.1 Main Findings ................................................................................................... 105 5.2 Research Implications for Policy and Practice ................................................. 106 5.3 Interpretation of Findings in the Context of Our Theoretical Framework .......... 112 5.4 Limitations ........................................................................................................ 118 5.5 Unanswered Questions and Future Research ................................................. 123 5.6 Conclusions ..................................................................................................... 127 References ............................................................................................................ 130 Appendices ............................................................................................................ 157

viii

List of Tables Table 2.1 Baseline characteristics of the Ontario long-term resident and recent immigrant study populations, 2005 ........................................................................ 35 Table 3.1 Baseline characteristics of the urban Ontario general population (excluding immigrant cohort) and immigrant study populations†, aged 40 and up and diabetes-free on April 1, 2004 ......................................................................... 60 Table 3.2 Measures of screening uptake and efficiency: The number and percent of the study population (overall, and by sex) with no previous diagnosis of diabetes, having a laboratory test to screen for diabetes in the 3-year period, 2004-2007; the proportion of those screened with newly diagnosed diabetes (screening efficiency); and the number needed to screen to identify one new case ................ 62 Table 3.3 Predictors of receiving a diabetes screen test during the 3-year study period (April 1, 2004 - March 31, 2007): results of regression analyses. Study population limited to immigrants without prior diagnosed diabetes, aged 40 and over

................................................................................................................ 65

Table 3.4 Estimated number and percentage of 'undiagnosed' diabetes cases by world region and immigration status, 2004-2007, among those aged 40 and up with no prior diabetes diagnosis on April 1, 2004 ................................................... 68 Table 4.1 Baseline characteristics of the Ontario long-term residence and recent immigrant diabetes-free study populations, by world region of birth ...................... 88 Table 4.2 Diabetes incidence rates over a 5-year follow-up period by primary covariates, age and sex ......................................................................................... 90 Table 4.3 Age of equivalent diabetes risk by ethnicity. The risk experience by men aged 40 of Western European ethnicity was used as the standard for comparison ............................................................................................................ 95

ix

Tables 5.1 Summary of the elements of the Theoretical Model and how they relate to the current work ................................................................................................. 128

x

List of Figures Figure 1.1 World Health Organization’s Commission on the Social Determinants of Health Conceptual Framework ........................................................................... 21 Figure 2.1 Age-adjusted, sex-specific diabetes prevalence rates (2005) by world region of origin for immigrants (1985-2000) and long-term Ontario residents ........ 37 Figure 2.2 Age-adjusted diabetes prevalence (per 100) with 95% confidence intervals among immigrants to Ontario (1985-2000), comparing the 15 countries with the highest prevalence in 2005 with Ontario’s prevalence in 2005/06 ............ 38 Figure 2.3 Age-specific diabetes prevalence rates for immigrants and long-term residents by sex, 2005 ........................................................................................... 39 Figure 2.4 Age-specific diabetes prevalence rates by WRO (males), 2005 ........... 41 Figure 2.5 Age-specific diabetes prevalence rates by WRO (females), 2005 ........ 42 Figure 2.6 Risk factors for diabetes (2005) among immigrants to Ontario (19852000) by sex .......................................................................................................... 43 Figure 4.1 Adjusted average diabetes incidence rates by ethnicity and age, men (1994-2008) ........................................................................................................... 93 Figure 4.2 Adjusted average diabetes incidence rates by ethnicity and age, women (1994-2008) ............................................................................................... 94 Figure 5.1 The World Health Organization’s Commission on the Social Determinants of Health conceptual framework ...................................................... 113

xi

List of Appendices Appendix A. Data sources ................................................................................... 158 Appendix B. ODD inclusion schema .................................................................... 163 Appendix C. Countries included in the Citizenship and Immigration Canada database and the assigned World Region of Origin ............................................... 164 Appendix D. Incidence cohort creation schema ................................................... 169 Appendix E. Age-Period-Cohort effects ............................................................... 170 Appendix F. Detailed Methodology of the Cox Proportional Hazard model ......... 172 Appendix G. Characteristics of the Ontario long-term resident and recent immigrant study populations, as well as those excluded due to prior diabetes diagnosis, 2010 ...................................................................................................... 183 Appendix H. Diabetes incidence rates before and after restriction to those who have received a diabetes test ................................................................................ 184 Appendix I. Cox Proportional Hazard model sensitivity analyses ........................ 185

xii

1

Chapter 1 Introduction

2

1.1

Thesis Overview

The purpose of this thesis is to investigate the epidemiology of diabetes in an immigrant, multi-ethnic population through an equity lens. The ultimate goal of this work is to generate information that can be used to design appropriate and effective targeted programs for diabetes prevention, management and control in order to reduce inequities in health. To this end, the thesis has been divided into three major sections dealing with distinct yet interconnected objectives. Each objective is dealt with in a separate chapter.

Objective 1: To investigate disparities in diabetes burden among immigrant ethnic groups. (Chapter 2)

Objective 2: To determine rates of screening for diabetes in immigrant ethnic groups and identify whether disparities exist by immigration status, ethnicity and sex. (Chapter 3)

Objective 3: To quantify diabetes risk by ethnicity and sex and explore whether the age of diabetes onset differs between ethnic groups. (Chapter 4)

3

1.2

Background

Type 2 Diabetes Mellitus (T2DM) prevalence is increasing globally, (King & Rewers, 1993) largely in response to rising rates of obesity (Mokdad et al., 2000b; Katzmarzyk, 2002). It was estimated in 2003 that 194 million adults worldwide were living with diabetes and an additional 314 million people had impaired glucose tolerance (putting them at a high risk of subsequent diabetes) (Sicree et al., 2006). One American study recently estimated that the lifetime risk of developing diabetes for Americans born in the year 2000 is one in three (Narayan et al., 2003). In Canada, roughly 2.4 million Canadians are currently living with this condition (Public Health Agency of Canada, 2011). Age and sex-adjusted diabetes prevalence in Ontario increased by 69% in the 10 years between 1995 and 2005 to 8.8% (Lipscombe & Hux, 2007) - an increase that surpassed predictions made by the World Health Organization for prevalence in the year 2030 (Wild et al., 2004). Since diabetes is associated with increased rates of morbidity, mortality and disability (Public Health Agency of Canada, 2009), these global trends have serious implications for population health, the economic sustainability of Canada’s universal health system and the economy in general.

Type 2 diabetes, which accounts for roughly 90-95% of all diabetes (Zimmet et al., 2001) and is the focus of this dissertation, results from a complex interplay between genetic and environmental factors. Throughout this thesis the abbreviated terms T2DM and diabetes will be used interchangeably to refer to Type 2 Diabetes Mellitus. T2DM is a chronic metabolic disorder resulting from increased insulin resistance and defective insulin secretion. This underlying insulin resistance, together with hyperglycemia and

4

other metabolic derangements, impacts the vascular system and other organs. Diabetes is a leading cause of cardiovascular disease (Kapral et al., 2003; Booth et al., 2003), end-stage renal failure leading to dialysis (Oliver et al., 2003), amputations (Hux et al., 2003) and blindness (Klein & Klein, 1995). One-third of all admissions for myocardial infarction and stroke and two-thirds of all non-traumatic amputations in Ontario occur in persons with diabetes (Hux et al., 2003; Kapral et al., 2003; Booth et al., 2003). People with diabetes are at a 50-60% greater risk of depression and experience higher rates of physical disability due to circulatory, nervous and immune system problems (Manuel & Schultz, 2003). Diabetes is associated with a 13 year reduction in life expectancy and mortality rates in adults with the condition are twice as high as among those without diabetes (Manuel & Schultz, 2004; Public Health Agency of Canada, 2009). Diabetes is clearly a clinically important disease, which affects not only general health but also physical functioning, mental health and life expectancy. In 2010 the direct and indirect costs of diabetes in Canada were estimated at $11.7 billion (Canadian Diabetes Association Clinical Practice Guidelines Expert Committee & Diabete Quebec, 2011).

Diabetes Risk Factors and their Distribution Across Populations

Obesity is the most significant risk factor for the development of type 2 diabetes. The likelihood of developing diabetes among individuals who are classified as obese (defined as a body mass index, or BMI, >=30) is more than seven times higher than among those with normal body weight (Abdullah et al., 2010). Increasing age is also associated with an increased likelihood of transitioning from insulin resistance to diabetes, and the highest incidence of diabetes is found in those over age 65 (Public

5

Health Agency of Canada, 2009). In addition to the increased risk associated with obesity and advanced age, T2DM occurs more commonly in non-European ethnoracial groups, largely due to ethnic differences in genetic susceptibility that are not completely understood. This ethnic disparity in risk is the focus of this dissertation.

The epidemiology of diabetes also varies by sex and gender. The prevalence of type 2 diabetes has traditionally been found to be higher in men than in women (Gourdy et al., 2001; Public Health Agency of Canada, 2009; Wild et al., 2004). There is evidence, however, that this may not be the case in some high-risk ethnic groups including Aboriginal, ‘Black’ and Mexican-American populations, in which women have been found to have equally high, or higher risk than men in some studies (Young et al., 2000; Chiu et al., 2010; Cowie et al., 2006; Dyck et al., 2010). High rates of overweight and obesity have also been observed in women in these communities, which may increase their susceptibility to developing disease and be the cause for this departure from traditionally observed sex-specific differences in risk (Flegal et al., 2002; McDermott et al., 2010). Such gender differences in the prevalence of risk factors for the development of diabetes, such as obesity and physical inactivity have been described previously (Matheson et al., 2008; Chiu et al., 2010; Meisinger et al., 2002). Furthermore, women have higher rates of contact with the health care system and are therefore more likely to be screened for diabetes (Wilson et al., 2009). As a result of these gender differences in prevalence, distribution of risk factors and patterns of health services utilization, all analyses in this thesis were stratified by sex.

6

Finally, diabetes risk is unequally distributed across socioeconomic groups, with a larger burden of disease in low income and less educated populations, an association that may be stronger for women (Ross et al., 2010). Due to the higher prevalence of this condition in specific ethno-racial groups and among people of lower socioeconomic status, as well as the possible interaction between socioeconomic status and gender, population health disparities and health equity are underlying themes of this research.

Diabetes and Ethnicity: Genetic Origins and Pathophysiology

A strong genetic component to the risk for type 2 diabetes was established through early epidemiologic studies showing extremely high concordance in monozygotic twins, and a very high prevalence among close family relations (Kumar & Clark, 1999; Pincus & White, 1933). A genetic role in susceptibility to diabetes was further supported by the observation that a high level of variation existed between ethnicities living in the same environment with respect to diabetes prevalence and insulin responses to oral glucose tests in nondiabetic individuals (Rimoin, 1969; Ali et al., 1993). Conversely, other landmark studies have looked at genetically similar populations in different settings, including Japanese migrants living in Hawaii and Los Angeles (Hara et al., 1994), and West African migrants to the Caribbean and Britain (Mbanya et al., 1999), and have found high variability in diabetes prevalence depending on where people live. Therefore taken together, these results support the theory that ethnic differences in diabetes risk have a genetic origin which impacts the predisposition for developing disease; however expression of these differences is dependent on gene-environment interactions. The exact mechanism for the expression of these differences is still not entirely clear.

7

The predominant theories for the higher incidence of diabetes and insulin resistance in certain ethnic groups are the ‘thrifty genotype’ and ‘thrifty phenotype’ hypotheses. The ‘thrifty genotype’ hypothesis states that frequent exposure to periods of starvation or insufficient nutrient intake resulted in a survival advantage for an adaptive genotype that was very efficient at storing nutrients as abdominal fat. Presented as evidence for this is the higher prevalence of abdominal fat in Asians even in persons of normal BMI (McKeigue et al., 1992; Abate et al., 2004; Raji et al., 2001). The ‘thrifty phenotype’ hypothesis postulates a fetal and maternal origin of disease. This theory, supported by recent research, hypothesizes that an adverse intrauterine environment, due to maternal undernutrition, resulting in low birthweight and rapid post-natal growth are associated with later development of obesity and higher diabetes risk (Yajnik & Deshmukh, 2008; Ma & Chan, 2009). This pattern of fetal/post-natal growth is then propagated in subsequent generations. In addition, there is evidence from the Child Heart and Health Study in England (CHASE) that children (= age 65† % Male Income quintile§ of neighbourhood of settlement (%): Q1 (lowest income) Q2 Q3 Q4 Q5 World Region of Birth East Asia & the Pacific South Asia Latin America & the Caribbean Eastern Europe & Central Asia Western Europe & North America North Africa & the Middle East Sub-Saharan Africa Unknown/Stateless None specified Immigration Visa Category Family Economic-Skilled-Independent Refugee Economic-Skilled-Family Economic-Business Other None specified Educational Qualifications at Landing (%) No Education Secondary or Less Non-University Qualifications Some University University Degree or Higher None specified Years Since Arrival (using 2005 as year of reference) (%) 5 - 9 years 10 - 14 years 15 years or more

7,503,085 47 18.2 49.3

1,122,771 43 10.2 49.5

18.2 19.6 19.7 20.4 21.2

26.4 21.7 19.7 18.8 13.0

-

309,043 (27.5%) 217,367 (19.4%) 177,191 (15.8%) 167,456 (14.9%) 98,931 (8.8%) 87,610 (7.8%) 64,367 (5.7%) 795 (0.07%) 11 (0.001%)

-

448,142 (39.9%) 285,322 (25.4%) 184,588 (16.4%) 124,465 (11.1%) 54,056 (4.8%) 26,185 (2.3%) 13 (0.001%)

-

47,604 (4.2%) 608,925 (54.2%) 171,106 (15.2%) 56,021 (5.0%) 239,031 (21.3%) 84 (0.007%)

-

322,047 (28.7%) 384,608 (34.2%) 416,116 (37.1%)

*Population eligible for provincial health care based on administrative databases. Recent immigrant population includes those who obtained legal landed status between 1985 and 2000. **Urban areas identified from first three characters of the postal code of residence (the Forward Sortation Area(FSA)). † Based on age as of March 31st, 2005. § 2001 Census income information was applied based on the individual's postal code of residence in 2005.

36

2.3.2 Trends in Diabetes Prevalence

Immigrants from South Asia, LAC, Sub-Saharan Africa and North Africa and the Middle East all experienced significantly higher diabetes rates than Ontario long-term residents (Figure 2.1). The fifteen countries of birth (out of 239 countries and geopolitical regions) that were associated with the highest rates of diabetes among both sexes were found in South Asia, the Pacific Islands, Latin America, the Caribbean and Africa (Figure 2.2). In the general Ontario population men have higher rates than women (6.5 % versus 6.2 %, respectively) but women who were recent immigrants had rates equal to or higher than immigrant men from the same regions, with the exception of women from Sub-Saharan Africa.

Overall, immigrants of both sexes had statistically higher rates of diabetes than the Ontario long-term population at all ages (except for males aged 75+), with a large disparity between women who were recent immigrants and women who were long-term residents (Figure 2.3).

37

38

39

40

Diabetes prevalence increased sharply with age until age 75 in both sexes and among both recent immigrants and long-term residents (Figures 2.3-2.5). An increase in diabetes prevalence appeared at a young age by immigrants from the highest risk regions and this risk was sustained across all age groups. Men from South Asia had the highest prevalence rates across all age groups (reaching 36.7% in the 65-74 age group) followed by men from Latin America and the Caribbean and Sub-Saharan Africa (35.0% and 33.2%, respectively, in the 65-74 age group) (see Figure 2.4). Women from Latin America and the Caribbean and South Asia had the highest prevalence rates by age (reaching a maximum of 37.0% and 34.8%, respectively, in the 65-74 age group) (Figure 2.5). The lowest rates were found among men and women from Europe, North America and Central Asia.

The multivariate logistic regression analysis results (Figure 2.6) show that after controlling for age, immigrant category, education, income and time since arrival, men and women from South Asia had significantly higher diabetes rates (OR = 4.01, 95% CI 3.82 – 4.21 and OR=3.22, 95% CI 3.07-3.37, respectively) compared to immigrants from Western Europe and North America. The next highest risk was experiences by men and women from Latin America and the Caribbean (OR= 2.18, 95% CI 2.08 – 2.30 and OR=2.40, 95% CI 2.29-2.52, respectively) and Sub-Saharan Africa (OR= 2.31, 95% CI 2.17 – 2.45 and OR=1.83, 95% CI 1.72-1.95, respectively).

41

42

43

Fig. 2.6 Risk factors for diabetes (2005) among immigrants to Ontario (1985-2000) by sex.

44

An income gradient was observed, whereby lower income was associated with higher risk (OR 1.31, 95% CI 1.26-1.36 for men and OR 1.38, 95% CI 1.33-1.44 for women in the lowest, relative to the highest, income quintile neighbourhoods). Women with secondary education or less had the highest risk as compared with those with a university degree or higher (OR = 1.32, 95% CI 1.28 – 1.37). Men with no education had the lowest diabetes risk (OR = 0.56, 95% CI 0.53 – 0.60).

The multiple logistic regression model also displayed a gradient for time since arrival with males and females living in Canada for 15 years or more having the highest diabetes prevalence (OR= 1.52, 95% CI 1.48 – 1.56 and OR=1.40, 95% CI 1.361.44, respectively) compared with those living in Canada 5-9 years.

2.4 Discussion

This study used a unique population-based dataset in a setting with high rates of immigration to describe the epidemiology of diabetes among a heterogeneous immigrant population. The increased relative rates of diabetes among immigrants from South Asia, Latin America and the Caribbean, Africa and the Middle East are particularly striking given that the population of Ontario is itself highly diverse ethnoracially. We also found that the risk for South Asians was at least triple that of immigrants from Western Europe and North America, and the risk for people from Latin America and the Caribbean and Sub-Saharan Africa was roughly double, even after controlling for age, sex, time since arrival, income and immigration-related variables.

45

Our findings support those of three previous Canadian studies that found South Asians had 2-3 times the rate of diabetes as compared with the overall Ontario population, the ”white” Ontario population; or a sample of Canadians of European heritage (Manuel & Schultz, 2003; Shah, 2008; Anand et al., 2000). These studies were all either based on health survey data (Manuel & Schultz, 2003; Shah, 2008), or a relatively small population sample (Anand et al., 2000). Furthermore, all three were focused on ethnic differences in diabetes in the population overall and did not look at immigration status. Research conducted in the UK found people of South Asian descent had diabetes prevalence rates that were 3-6 times than the white British population, which is consistent with our findings (Mather & Keen, 1985). Our prevalence estimates are higher than those generated by the World Health Organization (WHO) in 2004, however those estimates used data derived from a small number of often out-dated studies and were based on extrapolations and assumptions that were likely to lead to underestimations of risk (Wild et al., 2004).

A detailed description of age and sex patterns of diabetes risk among immigrants to Western countries has previously been lacking. Although diabetes prevalence is generally higher in men than women (Wild et al., 2004; Public Health Agency of Canada, 2008), recent immigrant women in our study had prevalence rates roughly equivalent to or higher than men from the same countries. Women from Latin America and the Caribbean experienced particularly high rates relative to men. This pattern of elevated risk in women relative to men has been previously described only in areas with high proportions of Aboriginal populations (Young et al., 2000). We also found that the largest disparity in risk existed between recent immigrant and longterm resident women. These findings suggest that recent immigrant women may be at particularly high risk for diabetes. Combined with the social isolation and barriers

46

to access of services experienced by many recently immigrated women (Bierman et al., 2010), this higher risk may raise important health-related issues for immigrant women and should be of concern to health providers and planners.

We also found an age-related disparity in diabetes prevalence between the highest and lowest risk immigrant groups that became apparent in young adulthood (by age 35 for South Asians) and increased with age. By age 65, more than a third of men and women from these regions had diabetes. Above age 75, a plateau or decrease in risk was seen in people from all regions, which has been described previously in the general Canadian population (Public Health Agency of Canada, 2008).

We found that diabetes risk increased with time since arrival. An increase in prevalence of chronic disease and declining health status over time since immigration has been previously described in the Canadian literature based on health survey data (Perez, 2002; Newbold, 2005; Dunn & Dyck, 2000). Many possible causes of this observed deterioration in health status over time have been suggested including uptake of unhealthy behaviours, acculturation stress, decreased social, economic and political status, barriers to accessing preventative services, and competing priorities resulting in reduced self-care (Newbold, 2009).

Socioeconomic status is known to have an inverse relationship with diabetes risk (Brancati et al., 1996; Robbins et al., 2001), and low education has been previously identified as a correlate of higher diabetes rates in Canada (Manuel & Schultz, 2003). In the current study, we found a gender-education interaction, whereby low educational attainment (less than high school diploma) was a significant independent risk factor for diabetes in immigrant women, but little or no formal education was

47

protective for immigrant men. The latter could reflect more physically demanding employment (e.g. manual labour) among men in this group. We also observed an inverse relationship between income and diabetes risk in women, as has been described before in Canada (Manuel & Schultz, 2003). One explanation for this may be that higher rates of obesity are reported among women with low socioeconomic status than among men (Matheson et al., 2008).

High rates of diabetes in specific ethnic migrant groups are likely to be due to a complex interplay of genetic and environmental factors, including acculturation, stress, social isolation as well as employment and economic challenges (Misra & Ganda, 2007). Despite the limitations of our data in addressing social and experiential exposures, it must be noted that strong ethnic differences continue to be apparent even after controlling for certain pre-migration (country of birth, immigrant category, and education) and post-migration (neighbourhood income, time since arrival) factors.

Additional limitations of this study are related to the use of administrative data. Although we were unable to differentiate type 1 from type 2 diabetes in the administrative data, the former represents a very small proportion of all diabetes (510%) and is therefore unlikely to bias our results (International Diabetes Federation, 2009). Administrative data may also underestimate the prevalence of diabetes. Studies in other jurisdictions suggest that up to 30% of diabetes in the population may be undiagnosed (Harris & Eastman, 2000) and it is possible that the probability of diagnosis may differ by immigration status and country of birth. Finally, a small proportion (< 5%) of health services that are not billed on a fee-for-service basis will not be captured in our data (Williams & Young W, 1996).

48

We found that recent immigrants from South Asia, the Caribbean, South America and Africa, have much higher risk of diabetes than both long-term residents of Ontario and recent immigrants from Europe and Central Asia. This risk begins in young adulthood and continues throughout their life course. The largest disparity in risk between immigrants and the general population was observed among women, the etiology of which should be further explored. Risk increases over time since immigration, but ethnic differences persist even after controlling for this variable, suggesting that acculturation and transition to a ”westernized” diet and lifestyle contributes to, and may exacerbate, but does not explain the differences. Although a few studies have shown promising results, lifestyle interventions aimed at recent immigrants should be explored further (Renzaho et al., 2010). Our findings should also aid policy-makers and planners to develop specific screening guidelines and community-level targeted diabetes educational programs. Finally, this study highlights the critical importance of routine collection of data on immigration status and ethnicity for population health research.

49

Chapter 3 Diabetes Screening Among Immigrants: A Population-Based Urban Cohort Study

Credits This chapter represents a prepublication version of the following article: Creatore MI, Booth G, Manuel D, Moineddin R, Glazier RH. Diabetes screening among immigrants: A population-based urban cohort study. Diabetes Care 2012; 35:754-761.

50

Abstract Background: To examine diabetes screening, predictors of screening, and the burden of undiagnosed diabetes in the immigrant population and whether these estimates differ by ethnicity. Methods: A population-based retrospective cohort linking administrative health data to immigration files was used to follow the entire diabetes-free population aged 40 and up in Ontario, Canada (N = 3,484,222) for three years (2004 to 2007) to determine whether individuals were screened for diabetes. Multivariate regression was used to determine predictors of having a diabetes test. Results: Screening rates were slightly higher in the immigrant versus the general population (76.0% and 74.4%, respectively, p= age 65) were screened less than non-immigrant seniors. Percent yield of new diabetes cases among those screened was high for certain countries of birth (South Asia, 13.0%; Mexico and Latin America, 12.1%; Caribbean, 9.5%) and low among others (Europe, Central Asia, U.S., 5.1-5.2%). The number of physician visits was the single-most important predictor of screening and many high-risk ethnic groups required numerous visits before a test was administered. The proportion of diabetes that remained undiagnosed was estimated to be 9.7% in the general population and 9.0% in immigrants. Conclusions: Overall diabetes screening rates are high in Canada’s universal health care setting, including among high-risk ethnic groups. Despite this finding,

51

disparities in screening rates between immigrant sub-groups persist and multiple physician visits are often required to achieve recommended screening levels.

52

3.1 Introduction

Diabetes is a serious chronic disease that is associated with substantial increases in morbidity and mortality, and imposes a huge economic burden on society. Although screening for diabetes is increasing in Canada (Wilson et al., 2009), up to one-third of all diabetes cases are thought to be undiagnosed in the general population in Canada and the U.S., an estimate that may now be out-ofdate (Cowie et al., 2006; Young, 2001). One significant factor that is likely contributing to increased screening is the rising prevalence of obesity in the population.

Early detection and control of diabetes can potentially reduce the heightened risk of cardiovascular morbidity and mortality associated with this disease. People with screen-detected diabetes have an increased risk of heart disease as compared to the general population, and this risk is modifiable with treatment (Sandbaek et al., 2008; Janssen et al., 2009; Griffin et al., 2011). In addition, timely screening can prevent the onset of common diabetes-related complications that could be avoided through early detection and treatment (eg. retinopathy, peripheral neuropathy, peripheral vascular disease) (Colagiuri & Davies, 2009).

National guidelines in both the U.S. and Canada recommend that diabetes screening should be performed on those age 45 (U.S.) or 40 (Canada) and over every 3 years with more frequent or earlier screening for those with additional

53

risk factors, including belonging to a high-risk ethnic group (Ur et al., 2008; American Diabetes Association, 2010). Ethnic groups that have been shown to display an elevated risk for diabetes include people of South Asian (Dhawan et al., 1994; Creatore et al., 2010; Cruickshank et al., 1991), Aboriginal (Public Health Agency of Canada, 2009) and African-Caribbean descent (Cowie et al., 2006; Creatore et al., 2010). Many of the 250,000 immigrants to Canada every year (Research and Evaluation Branch Citizenship and Immigration Canada, 2009), belong to ethnicities that experience higher rates of diabetes (Creatore et al., 2010), and who therefore should be screened regularly and beginning at a younger age. There is evidence, however, that immigrants may have lower health care utilization (Kliewer & Kazanjian, 2000), which may predispose this group to have lower rates of screening than the Canadian-born population. An important and currently unanswered question therefore, is whether some ethnic or migrant groups are more likely to be ‘under-diagnosed’ than others. In this study, we describe the pattern of diabetes screening among recent immigrants to Ontario by looking at screening rates, screening efficiency/yield, predictors of screening and the burden of undiagnosed diabetes in this population by region of origin.

54

3.2 Research Design and Methods

3.2.1 Study Population

We conducted a retrospective cohort study to examine rates of screening for diabetes among immigrants to Canada compared to those in the general population during the 3-year period from April 1, 2004 (the baseline date for this study) to March 31, 2007. To do so, all adults aged 40 or older (based on Canadian screening recommendations) who were living in Ontario during the 3year period prior to baseline (from April 1, 2001) were identified from the Registered Persons Database (RPDB), an electronic registry of all individuals who are eligible for health coverage in Ontario. In order to identify immigrants to Canada, RPDB records were linked to immigration data from Citizenship and Immigration Canada (CIC), which contains information on all individuals having been granted permanent residency in Canada between 1985 and 2000 (N=1,377,816). This database includes demographic and socioeconomic information collected at the time of application for immigration status. Eighty-four percent of CIC records were linked to the RPDB using probabilistic linkage techniques. Feasibility of linkage between the CIC and health administrative datasets was tested in pilot projects (Kliewer & Kazanjian, 2000), and differences in linkage by immigration variables in these previous studies were found to be small and unlikely to produce significant bias in study results. For the purpose of this study, the general population comprised those who did not have a record of immigration between 1985 and 2000, so individuals having immigrated prior to

55

1985 were included in this group. Furthermore, in order to avoid misclassifying immigrants who were not captured in the CIC data linkage as non-immigrants, individuals in the general population were excluded from the study if they first became eligible for provincial health insurance after 1991. Nineteen-ninety-one is the first date for which administrative data on health insurance eligibility in Ontario is available. The majority of these excluded adults are likely to be external migrants not captured by the CIC data with a small proportion comprised of internal migrants arriving from another province.

Individuals with a diagnosis of diabetes at baseline, which accounted for roughly 11% (422,878 individuals) and 12% (59,766 individuals) of our general population and immigrant cohorts, respectively, were excluded from the study. Those who had no health care contact between April 1, 1999 (5 years before baseline) and March 31, 2007 (end of 3-year observation period) were also excluded. Since 98% of all immigrants in our database settled in urban areas, we excluded rural populations using a Statistics Canada algorithm based on postal codes. This resulted in the further exclusion of 2.0% (12,092) of immigrants and 17.3% (922,028) of long-term residents from the study.

3.2.2 Study Outcomes

Screening rates

Diagnosis of diabetes prior to, or during the study period was established by linking the study population to the Ontario Diabetes Database (ODD), a validated

56

population-based, cumulative, diabetes registry based on physician visits and hospitalizations for diabetes, excluding gestational diabetes (Hux et al., 2002). We determined the percentage of people without prior diabetes diagnosis, who were screened within the 3-year follow-up, along with 95% confidence intervals. Under the universal health insurance program in Ontario, over 95% of health services provided are captured in provincial, administrative data (Williams & Young W, 1996), allowing us to identify what services, including laboratory tests, were billed and when with the exception of a very small proportion of tests conducted in hospitals. Provincial health services data were linked to our study population by encrypted individual health card number. In the 3 year study followup, individuals were considered to be screened for diabetes if they had one or more physician or laboratory billing for a serum blood glucose, hemoglobin A1c or a non-pregnancy related oral glucose tolerance test. Due to our use of administrative data, we could not differentiate whether the test was for screening (in asymptomatic individuals) or diagnosis (in symptomatic individuals).

Screening efficiency

Screening efficiency (defined as the percent positive of the total screened with previously undetected diabetes) was measured. We also calculated the reciprocal of screening efficiency, the number needed to screen (NNS) within each risk group to identify one previously undiagnosed case of diabetes (NNS = Total number screened / Total number of newly diagnosed cases).

57

Burden of undiagnosed diabetes

Finally, based on the yield of new diabetes cases among the screened population, we estimated the number of people with undiagnosed diabetes we would expect to find in the unscreened population on March 31st, 2007 using the formula:.

Undiagnosed cases = Total unscreened population X screening efficiency (Wilson et al., 2010).

The proportion of all diabetes in the population that is undiagnosed was then estimated by dividing the number of undiagnosed cases by the total number of people with diabetes. Total cases of diabetes was calculated as the sum of all diagnosed (both prevalent at baseline as well as newly diagnosed during the study period) and undiagnosed cases.

3.2.3 Statistical analysis

All analyses were performed by world region of origin and were stratified by sex since there is evidence supporting a larger proportion of undiagnosed diabetes in men than in women (Wilson et al., 2010). Comparisons across sub-groups for the descriptive analyses above were conducted using chi-square tests.

Along with the descriptive analyses described above, multivariate log-binomial regression was used to determine the association between receiving a diabetes test within the recommended time frame and the covariates of interest. Three

58

different models were fit: 1) adjusted model to determine characteristics of those having a diabetes test within the recommended period; 2) same as model 1 but including number of primary care physician visits during the study period to adjust for patterns of utilization; 3) adjusted model to determine the predictors of being tested in any one visit (as opposed to being tested at any point in the 3 years of the study observation period, as with models 1 and 2). The latter model was generated using the number of visits up to the first diabetes test as an offset in the model and a Poisson distribution.

Covariates included in the model were age (40-49, 50-59, 60+), world region of birth, immigration visa category, educational qualifications at time of immigration, time in Canada (as of April 1st, 2004), income (based on residential postal code) and number of physician visits (derived from physician billing data and excluding specialist visits), during the study period. Due to the absence of individual-level income information in provincial health administrative databases, residential postal codes were linked to 2006 Canada Census data at the dissemination-area level (an area containing roughly 400-600 people) using a Postal Code Conversion File (PCCF+ v. 5D). Relative income quintiles adjusted for household and community size were then generated and assigned to individuals.

All analyses were performed using SAS (version 9.2). This protocol received ethical approval from the Institutional Review Board at Sunnybrook Health Sciences Centre and the University of Toronto.

59

3.3 Results

3.3.1 Baseline Study Characteristics

A total of 3,927,059 individuals were observed for the 3-year period. Compared with the general Ontario population, immigrants were younger, more likely to be male and more likely to live in low income neighbourhoods (Table 3.1). The largest proportion of immigrants was from Asia and Eastern Europe. The majority of people immigrated under the Economic (including investors, entrepreneurs, skilled workers) and Family (predominantly family reunification and sponsorship) visa categories. Over the 3-year period 212,137 new cases of diabetes were identified.

60

Table 3.1 Baseline characteristics of the urban* Ontario general population (excluding immigrant cohort) and immigrant study populations†, aged 40 and up and diabetes-free on April 1, 2004. Study Population Characteristics Population Median Age‡ % Male Income quintile§ of neighbourhood of settlement: Q1 (lowest income) Q2 Q3 Q4 Q5 World Region of Birth East Asia & the Pacific Eastern Europe & Central Asia South Asia Western Europe & U.S.A. Mexico & Latin America North Africa & the Middle East Caribbean Sub-Saharan Africa Unknown/Stateless Immigration Visa Category Economic Family Refugee Other Educational Qualifications at Landing (%) No Education Secondary or Less Non-University Qualifications Some University University Degree or Higher

General Population

Immigrant Cohort

3,484,222 54 46.5

442,837 48 48.9

17.6 19.1 19.1 20.4 23.6

27.6 23.1 19.7 16.8 12.6

-

133,360 (30.1%) 78,098 (17.6%) 73,212 (16.5%) 37,183 (8.4%) 35,009 (7.9%) 32,596 (7.4%) 29.758 (6.7%) 23,246 (5.2%) 375 (0.1%)

-

194,584 (43.9%) 158,652 (35.8%) 77,680 (17.5%) 11,915 (2.7%)

-

12,469 (2.8%) 204,833 (46.3%) 90,288 (20.4%) 22,277 (5.0%) 112,933 (25.5%)

Years Since Arrival (using 2004 as year of reference) (%) 4 - 9 years 137,339 (31.0%) 10 - 14 years 156,663 (35.4% 15 years or more 148,835 (33.6%) *Urban areas identified from first three characters of the postal code of residence (the Forward Sortation Area (FSA)). †Urban population eligible for provincial health care between April 1, 2004 and March 31, 2007, based on administrative databases. ‡ Based on age as of April 1st, 2004. § 2006 Census income information was applied based on the individual's postal code of residence on April 1, 2004.

61

3.3.2 Diabetes screening

Diabetes testing rates were high. Although statistically significant, the difference in screening rates between immigrants overall and the general population were small (76.0% vs. 74.4%, p=20 years of age) we have further increased the probability that most cases in our study represent type 2 diabetes.

162

References

1. Cernat G, Wall C, Iron K, Manuel D. Initial validation of Landed Immigrant Data System (LIDS) with the Registered Person's Database (RPDB) at ICES. Internal ICES Report to Health Canada. 2002. Toronto, Institute for Clinical Evaluative Sciences (ICES). 2. Hux JE, Ivis F, Flintoft V, Bica A. Diabetes in Ontario: determination of prevalence and incidence using a validated administrative data algorithm. Diabetes Care 2002; 25(3):512-516.

163

Appendix B. ODD algorithm CIHI records with any diagnosis code 250.x (ICD-9)

OHIP physician service claims with diagnosis code 250.x

Candidate cases for DM

Single OHIP claim only

2 OHIP claims or 1 discharge in 2 years

Presumed gestational DM

Previously in ODD?

No

Yes

Incident Cases

Prior prevalent Cases

Total Cases

164

Appendix C. Countries included in the Citizenship and Immigration Canada database and the assigned World Region of Origin. COB code from CIC data 620 621 658 622 610 601 624 650 625 651 626 653 654 602 655 627 652 628 899 656 629 630 631 605 632 633 657 299 399 256 202 840 801 845 832 204 200 222 207 831 257 258 260 261 172 242 262 241 341

Country Name Anguilla Antigua And Barbuda Aruba Bahama Islands, The Barbados Bermuda Cayman Islands Cuba Dominica Dominican Republic Grenada Guadeloupe Haiti Jamaica Martinique Montserrat Netherlands Antilles, The Nevis Ocean Nes Puerto Rico St. Kitts-Nevis St. Lucia St. Vincent and the Grenadines Trinidad & Tobago, Republic of Turks and Caicos Islands Virgin Islands, British Virgin Islands, U.S. Asia Nes Australia Nes Cambodia China, People's Republic of Cook Islands Fiji French Polynesia Guam Hong Kong Hong Kong Sar Indonesia, Republic of Japan Kiribati Korea, People's Democratic Republic of Korea, Republic of Laos Macao Madagascar Malaysia Mongolia, People's Republic of Myanmar (Burma) Nauru

Assigned World Region of Origin Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean Caribbean East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific

165

822 830 342 227 842 903 843 844 246 824 203 267 268 846 823 270 841 081 049 050 051 048 083 043 070 052 025 026 053 054 019 020 055 033 088 056 016 047 057 045 058 059 042 060 044 703 541 751 709 721 722 542 753 543 754 544

New Caledonia Pacific Islands, US Trust Territory of the Papau New Guinea Philippines Pitcairn Island Reunion Samoa, American Samoa, Western Singapore Solomons, The Taiwan Thailand Tibet Tonga Vanuatu Vietnam, Socialist Republic of Wallis and Futuna Albania Armenia Azerbaijan Belarus Bosnia-Hercegovina Bulgaria Croatia Fyr Macedonia Georgia Greece Hungary Kazakhstan Kyrgyzstan Latvia Lithuania Moldova Poland Romania Russia Slovak Republic Slovenia Tadjikistan Turkey Turkmenistan Ukraine Union of Soviet Socialist Republics Uzbekistan Yugoslavia Argentina Belize Bolivia Brazil Chile Colombia Costa Rica Ecuador El Salvador French Guiana Guatemala

East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific East Asia and the Pacific Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Eastern Europe and Central Asia Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico

166

711 545 501 546 548 547 755 723 752 724 725 131 253 255 183 101 223 224 206 225 226 208 171 133 263 213 265 231 210 135 280 274 273 252 212 254 205 901 264 209 201 199 151 160 153 188 154 155 911 157 156 905 159 158 162 161

Guyana Honduras Mexico Nicaragua Panama Canal Zone Panama, Republic of Paraguay Peru Surinam Uruguay Venezuela Algeria Bahrain Brunei Djibouti, Republic of Egypt Iran Iraq Israel Jordan Kuwait Lebanon Libya Morocco Oman Palestinian Authority (Gaza/West Bank) Qatar Saudi Arabia Syria Tunisia United Arab Emirates Yemen, People's Democratic Republic of Yemen, Republic of Afghanistan Bangladesh Bhutan India Maldives, Republic of Nepal Pakistan Sri Lanka Africa Nes Angola Benin, Peoples Republic of Botswana, Republic of Burkino-Faso Burundi Cameroon, Federal Republic of Cape Verde Islands Central Africa Republic Chad, Republic of Comoros Congo, People's Republic of the Congo,Democratic Republic of Eritrea Ethiopia

Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico Latin America and Mexico North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East North Africa and the Middle East South Asia South Asia South Asia South Asia South Asia South Asia South Asia South Asia Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa

167

163 164 165 178 166 167 169 132 152 170 111 173 174 902 175 122 176 177 179 914 180 904 181 182 121 185 186 130 187 136 184 112 113 000 082 305 011 035 012 511 039 009 221 015 014 017 002 018 099 912 021 022 046 024 084 521

Gabon Republic Gambia Ghana Guinea, Equatorial Guinea, Republic of Guinea-Bissau Ivory Coast, Republic Kenya Lesotho Liberia Malawi Mali, Republic of Mauritania Mauritius Mozambique Namibia Niger, Republic of the Nigeria Rwanda Sao Tome e Principe Senegal Seychelles Sierra Leone Somalia, Democratic Republic of South Africa, Republic of Sudan, Democratic Republic of Swaziland Tanzania, United Republic of Togo, Republic of Uganda Western Sahara Zambia Zimbabwe Unknown Andorra Australia Austria Azores Belgium Canada Canary Islands Channel Islands Cyprus Czech Republic Czechoslovakia Denmark England Estonia Europe Nes Falkland Islands Finland France German Democratic Republic Germany, Federal Republic of Gibraltar Greenland

Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Sub-Saharan Africa Unknown origin/Stateless Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S.

168

090 085 027 028 086 013 036 030 087 031 339 006 032 034 089 007 037 915 531 040 041 461 008

Holy See Iceland Ireland, Republic of Italy Liechtenstein Luxembourg Madeira Malta Monaco Netherlands, The New Zealand Northern Ireland Norway Portugal San Marino Scotland Spain St. Helena St. Pierre et Miquelon Sweden Switzerland United States of America Wales

Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S. Western Europe & U.S.

169

Appendix D. Incidence Cohort Creation

Ontario population (RPDB), eligible at any time 1991-2010 N=16,393,637

Immigrant cohort N=1,376,793

Long-term residents N=15,016,843

Exclude rural postal codes

N=1,350,744

N=12,679,645 Exclude those with no record of contact with health system†

N=1,295,841

N=12,253,543 Exclude those with diagnosed diabetes on or prior to March 31st, 1994

N=1,278,140 Exclude those who: arrived prior to 1991, had >2 years between arrival and first health care eligibility, or who lost eligibility prior to baseline.

N= 818,258

N=11,985,081 Exclude those with first ever eligibility after April 1, 1991*

N= 7,312,766

Exclude those less than 20 year of age at baseline

N= 592,376

N= 5,421,654

*In order to remove individuals from Long-term resident cohort who may be recent immigrants (but not in LIDSs). † No date of last contact (DOLC = missing)

170

Appendix E - Age-Period Cohort Effects

Due to the nature of our study cohort where we have multiple waves of immigrants arriving in different years, born in different time periods, and who arrive at different ages, we were faced with the issue of possible age-period-cohort effects. Briefly, there is an intrinsic and mathematical relationship between these three variables whereby the following can be written:

Period = age + cohort

This relationship makes standard regression modeling techniques insufficient for controlling for the impact of all three characteristics on the outcome. In order to investigate the impact of period of arrival and age at arrival, our immigrant study population was divided into multiple groups based on their age at arrival (30-39, 40-49, 50-59, 60-69) as well as their immigration period. Immigration period was categorized into three groups, those arriving between 1991 and 1993, 1994 and 1996, and 1997 and 2000. We then used a simple Poisson regression model with person-year offset to look at whether the probability of being diagnosed with diabetes differed by different immigrant cohorts (was time-dependent) and we saw that, indeed, more recent waves of immigrants were at higher risk of being diagnosed with diabetes than earlier waves of immigrants, both for women (RR with 1991-1993 as baseline, and 95% confidence intervals: 0.93 (0.88-0.98), p= 0.0039 for 1994-96; 0.76 (0.72-0.80), p