DIETARY PATTERNS OF ETHNIC NON-INDIGENOUS POPULATIONS IN NORTHERN BRITISH COLUMBIA: A CROSS-SECTIONAL SURVEY STUDY by Yaser Ahmed M.B., B.Ch. (M.D.), Cairo University, 2007 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN COMMUNITY HEALTH SCIENCES UNIVERSITY OF NORTHERN BRITISH COLUMBIA April 2018 © Yaser Ahmed, 2018 DIETARY PATTERNS OF ETHNIC POPULATIONS Abstract Dietary transitions and acculturation of different ethnic and immigrant subpopulations in western countries are associated with change in diet patterns and increased obesity risk. Loss of the “healthy immigrant effect” have been observed in such populations, so further research is critical. The study objectives were to a) explore dietary patterns for a sample from various ethnic backgrounds in northern British Columbia, b) examine the association between dietary patterns and healthy body weight status. A cross-sectional survey instrument including a food frequency questionnaire was developed to collect data from 444 participants. Two predominant patterns were identified and termed “western” and “prudent” dietary patterns. There was a positive association between a “western” diet pattern and increased BMI (p < .05). Men had higher adherence to “western” diet patterns, and higher risk of overweight and obesity (p < .05). Interventions promoting adherence to prudent diet and retention of ethnic food habits are warranted. ii DIETARY PATTERNS OF ETHNIC POPULATIONS TABLE OF CONTENTS Abstract..................................................................................................................... ....................ii Tables of Contents........................................................................................................... ............iii List of Tables...............................................................................................................................vi List of Figures.............................................................................................................. ...............vii Glossary.....................................................................................................................................viii Acnowledgments.........................................................................................................................ix Chapter 1: Introduction.................................................................................................................1 1.1 Rationale of the Study........................................................................................................5 1.2 The Purpose of the Study...................................................................................................5 Chapter 2: Literature Review........................................................................................................7 2.1 Dietary Acculturation.......................................................................................................10 2.2 Review of Major Dietary Patterns in Literature...............................................................11 2.2.1 Mediterranean Diet……………………………………………………………….…...11 2.2.1.1 Review of Studies: Mediterranean Diet and Cardiovascular Disease....................14 2.2.1.2 Review of Studies: Mediterranean diet and Type 2 Diabetes................................16 2.2.1.3 Review of Studies: Mediterranean diet and Metabolic Syndrome.........................17 2.2.2 South Asian Diet...........................................................................................................18 2.2.3 Chinese Diet..................................................................................................................19 Chapter 3: Design and Methods..................................................................................................21 3.1 Recruitment Strategy.......................................................................................................21 3.1.1 Sample Size and Sampling Method..........................................................................23 3.1.2 Response Rate...........................................................................................................25 iii DIETARY PATTERNS OF ETHNIC POPULATIONS 3.2. Data Collection...............................................................................................................26 3.2.1. Survey Questionnaire Development........................................................................27 3.2.2 Modifications to the Cross-Sectional Survey Instrument.........................................32 3.2.3. Pilot Testing of the Survey Instrument....................................................................35 3.3 Data Entry........................................................................................................................36 3.4. Data Recoding and Analysis...........................................................................................37 3.4.1 Recoding of Dietary Variables.................................................................................37 3.4.2 Deriving Body Mass Index.......................................................................................38 3.4.3 Deriving Age in Years..............................................................................................38 3.4.4 Deriving Length of Stay/Residence in Years...........................................................39 3.4.5 Data Analysis Techniques........................................................................................39 Chapter 4: Results......................................................................................................................41 4.1 Descriptive Statistical Analysis.......................................................................................41 4.1.1 Socio-demographic characteristics...........................................................................41 4.2 Body Mass Index.............................................................................................................44 4.3 Multivariate Analyses Assumptions................................................................................46 4.4 Principal Components Analysis.......................................................................................47 4.5 Multiple Regression.........................................................................................................61 Chapter 5: Discussion & Conclusion..........................................................................................66 5.1 Dietary Patterns for Study Sample...................................................................................67 5.2 Healthy Body Weight Status and Association with Dietary Patterns..............................72 5.3 Study Limitations.............................................................................................................73 5.4 Future Research................................................................................................................74 iv DIETARY PATTERNS OF ETHNIC POPULATIONS 5.5 Conclusion.........................................................................................................................75 References................................................................................................................... .................78 Appendix A: Dietary Pattern Survey Questionnaire....................................................................96 Appendix B: Number of Missing Responses by Dietary Item...................................................127 Appendix C: Communality for 63 Dietary Items following PCA with Varimax Orthogonal Rotation......................................................................................................................................129 Appendix D: UNBC Research Ethics Board Approval Letter...................................................131 v DIETARY PATTERNS OF ETHNIC POPULATIONS List of Tables Table 1. Recruitment Event Date and Number of Participants.................................................26 Table 2. Food Groupings Adapted from Hu et al. (2000).........................................................30 Table 3. Sociodemographic Characteristics of Study Sample...................................................43 Table 4. Mean, Standard Deviation & Categories of BMI and BMI by Gender.......................45 Table 5. Correlation among Six Dietary Components...............................................................50 Table 6. Six Dietary Components Extracted by PCA with Beta Component Loading and Percent of Variance and Covariance Using Varimax Orthogonal Rotation..............................52 Table 7. Five Dietary Components Extracted by PCA with Their Beta Component Loading and Percent of Variance Using Varimax Orthogonal Rotation..................................54 Table 8. Four Dietary Components Extracted by PCA with Their Beta Component Loading and Percent of Variance Using Varimax Orthogonal Rotation...................................56 Table 9. Zero Order Correlations among Outcome and Covariate Variables...........................62 Table 10. Zero Order Correlations among Outcome and Predictor Variables..........................63 Table 10. Zero Order Correlations among Covariate and Predictor Variables.........................64 Table 12. Coefficients, p Value, Semipartial Correlations, and Unique Variance Contribution for Each Covariate and Predictor Variable in the Regression Analysis…..........65 vi DIETARY PATTERNS OF ETHNIC POPULATIONS List of Figures Figure1. Number of non-official mother tongue population in British Columbia in 2011……..2 Figure 2. Mediterranean diet pyramid…………………………………………………………15 Figure 3. Scree plot produced by orthogonal Principal Components extraction method….. …48 Figure 4. Western diet patterns identified from the food frequency questionnaire (n = 410)…59 Figure 5. Prudent diet patterns identified from the food frequency questionnaire (n = 410)….60 vii DIETARY PATTERNS OF ETHNIC POPULATIONS Glossary Caucasian: It refers to people whose ancestors originate from Europe or Western countries as distinguished from the Indigenous, Eastern and Oriental people Chinese: It refers mainly to greater china including mainland Hon Kong, China, Taiwan, and Macau. First Generation Canadian Immigrant: An individual born outside Canada who is a landed immigrant or permanent resident, a non-permanent resident who has a work/study permit, or a refugee claimant Northern British Columbia (BC): It refers to the northern part of BC province which receives healthcare services through Northern Health Authority. Geographically, this region includes the communities of Altin, Burns Lake, Chetwynd, Dawson Creek, Fort Nelson, Fort St. James, Fort St. John, Haida Gwaii, Hazelton, Houston, Hudson’s Hope, Kitimat, Mackenzie, Masset, McBride, Pouce Coupe, Prince Rupert, Prince George, Quesnel, Sandpit, Smithers, Terrace Tumbler Ridge, Valemount, and Vanderhoof Second Generation Canadian Immigrant: An individual born in Canada and has at least one parent born outside Canada South Asian: It refers mainly to the Indian subcontinent including Bangladesh, Nepal, Pakistan, India, Sri Lanka, Bhutan, and Maldives Third Generation Canadian Immigrants: An individual born in Canada, and has both parents born in Canada. One or more grandparents were born outside Canada viii DIETARY PATTERNS OF ETHNIC POPULATIONS ACKNOWLEDGEMENTS The work presented in this thesis would not have been possible without my close association with many people. I take this opportunity to extend my sincere gratitude and appreciation to all those who made this master’s thesis possible. I would like to express my sincere gratitude to my advisor Dr. Mamdouuh Shubair for the continous support of my master’s study and related research. His mentorship, guidance and encouragement have been remarkable. I would like to deeply thank Dr. Henry Harder and Dr. Peter MacMillan for their invaluable insight and expertise that greatly assisted my research. I must thank the many volunteers including Ahmed, Farzana, Khaled, Quinn, Belal, Ali, Dhruv, Florence, Vishal, Priya, Penny, Karlene, Sunita, and Rosemary who gave their time and graceful volunteer efforts and assistance whenever required. In particular, these colleagues have done great job during planning and conducting the various data collection events for this study, without whom this study would not have been possible. I owe my deepest gratitude towards my better half for her eternal support, patience, and encouragement. I am also thankful for my two sons, Zain and Mazen, for making me a happy proud father during all this time. I feel a deep sense of gratitude to my mother, father, and my teachers throughout my life who made it possible for me to see this day. Special thanks to Immigrant Multicultural Service Society (IMSS), UNBC graduate and undergraduate societies, Vaisakhi celebration organizers, Holi celebration organizers, Prince George Islamic Centre, and the local Sikh temples for their collaboration that made this research possible. ix DIETARY PATTERNS OF ETHNIC POPULATIONS Dietary Patterns of Ethnic Non-Indigenous Populations in Northern British Columbia Chapter 1: Introduction Ethnic groups in Canada comprise a noteworthy segment of the population. In 2006, the foreign-born population represented 19.1% of the general Canadian population, and 27.5% of the total British Columbia (BC) population (Statistics Canada, 2008). Immigration to Canada is an important component of the country’s economic development policy. It is estimated that Canada receives over 200,000 immigrants annually representing 60% of the annual population growth (Chui, Tran, & Maheux, 2006). According to Statistics Canada (2012), the three most common non-European ethnicities in BC are Indian, Chinese, and Filipino (Figure 1). Ethnic populations in northern BC are further marginalized and more studies are needed to explore their health status and needs (Sarkar, Chau, Lix, Bruce, & Young, 2010). Health disparities associated with ethnic minorities are attributed to several factors such as cultural & linguistic barriers and lower socio-economic status, which may contribute to unhealthy eating habits (Lai & Chau, 2007). This research study explored the current dietary patterns of ethnic populations in northern BC. Diet plays a significant role in the development of obesity and obesity-related chronic diseases such as type 2 diabetes (T2D), metabolic syndrome (MetS), atherosclerosis, hypertension, and cardio-vascular disease (CVD) (Remington & Brownson, 2011). High energy diet and physical inactivity are the main determinants of the global obesity epidemic phenomenon, most notably in western industrialized countries (Kimokoti & Millen, 2011; World Health Organization, 2003). In Canada, obesity rates have tripled between 1985 and 2011 (Twells, Gregory, Reddigan, & Midodzi, 2014). Currently, it is estimated that approximately 27% of the adult Canadian population suffer from obesity (Public Health 1 DIETARY PATTERNS OF ETHNIC POPULATIONS 400,000 365,870 350,000 300,000 250,000 200,000 150,000 193,985 138,845 124,580 100,000 96,420 77,745 77,100 50,000 50,600 43,965 37,470 28,135 0 Figure1. Number of non-official mother tongue population in British Columbia in 2011 (Statistics Canada, 2012). Agency of Canada, 2015). Furthermore, it is predicted that, by 2019, half of the Canadian provinces will have more overweight or obese adults than normal-weight adults with an alarming disproportionate increase in excessive obesity rates (Twells, et al., 2014). In North America and Europe, immigrants have higher morbidity and mortality rates due to nutrition related diseases compared to the mainstream population, as well as the population of their native country despite the “healthy immigrant effect” that places recent immigrants at a better health status than that of average population in the host country (Gilbert & Khokhar, 2008; Satia et al., 2001). In Canada, candidates for immigration have to pass medical tests to screen out those who have significant health problems. However, there is abundant evidence in the literature that attribute both the loss of the “healthy immigrant effect” and the paradoxical adverse health outcomes of Canadian immigrants to transitions in 2 DIETARY PATTERNS OF ETHNIC POPULATIONS dietary habits (Sanou et al., 2014). More details about the dietary acculturation associated with immigration to a western country and more specifically to Canada are discussed in chapter two. There is also an important evidence that area of origin plays a significant role in the healthfulness of dietary intake for ethnic groups. For example, some traditional diet patterns are considered to be heart healthy, such as the Mediterranean diet (Trichopoulou & Lagiou, 1997), whereas other traditional diet patterns are associated with an increased risk of MetS, such as the Mexican diet (Denova-Gutierrez et al., 2010). Traditional Indian diets are high in grains, fruits and vegetables, and low in meat and meat products (Varghese & Moore-Orr, 2002). Furthermore, traditional Chinese diets are high in fiber and carbohydrates (e.g., rice and noodles), and low in fat (Gilbert & Khokhar, 2008). Both traditional Indian and Chinese diets are considered healthier than western diet (Gilbert & Khokhar, 2008; Varghese & Moore-Orr, 2002). Further details on ethnic diet characteristics, associated dietary behaviors, and food beliefs that are relevant to this study are further discussed in chapter two. The Western diet pattern (WDP) is considered as one of the less healthy diets, which includes whole fat dairy products, processed meat, processed food, high intake of energy rich foods and junk food (i.e., food containing high levels of calories or fat with little nutritional value such as candy and carbonated beverage), red meat, and refined sugars (Cordain et al., 2005). Some think further epidemiological research on the increasing incidence of coronary heart disease and certain types of cancer amongst populations adopting WDP is needed (Jemal, Center, DeSantis, & Ward, 2010; Yusuf, Redyy, Ounpuu, & Anand, 2001). For example, incidence and mortality rates of breast cancer in the U.S. and many other European countries are on average five times the rates in Africa, Asia, and South America (Jemal et al., 3 DIETARY PATTERNS OF ETHNIC POPULATIONS 2010). More interestingly, the researchers found the rates amongst U.S. Hispanics and Asians to be significantly higher compared to the rates in Latin America and Asia, respectively. Another study found out that the combined obesity and overweight rates amongst Filipino women living in the US have almost doubled compared to women living in the Philippines (Vargas & Jurado, 2016). Healthy diet is an important pillar for wellbeing. It supports normal growth and development throughout different stages of life, promotes healthy body weight, and reduces chronic disease risks (Rogol, Clark, & Roemmich, 2000). Analysis of a single or few nutrients or foods in relation to a disease has been traditionally used in nutritional epidemiological studies (Tucker, 2010). However, such a method of analysis has several limitations including the inability to account for in-vivo chemical interaction between different nutrients and the failure to eliminate confounders associated with dietary patterns when performing “single” nutrient analysis (Hu, 2002). Dietary patterns are defined as “measures of the total usual intake of food combinations in individuals and groups” (Tucker, 2010, p. 212). The “dietary pattern” analysis is a holistic approach to determine the dietary habits and characteristics of food and nutrient intake. Recently, many authors have suggested to study the overall dietary patterns instead of a “single” nutrient analysis (JonesMcLean, Shatenstein, & Whiting, 2010; Quatromoni et al., 2002). Furthermore, some researchers ascertain the feasibility of introducing healthy “dietary patterns” as a successful new approach of nutrition intervention to chronic disease prevention and control, such as the DASH (Dietary Approach to Stop Hypertension) study and the Lyon Diet Heart Study (Hu, 2002). Dietary patterns are mainly identified through multivariate analysis techniques such as principal component analysis (PCA) and cluster analysis (Hu, 2002). In this study, PCA was 4 DIETARY PATTERNS OF ETHNIC POPULATIONS utilized to analyze average monthly frequency of food consumption. Information on food and nutrient intake, dietary acculturation, and general health and lifestyle of ethnic subpopulations was collected. Body weight status and risk factors for chronic disease were the main outcome measures. Further details about methods of data collection and analysis are discussed in chapter three. 1.1 Rationale of the Study The increasing epidemic prevalence of “diseases of modernization” such as type 2 diabetes, high blood pressure, and cardiovascular disease is already being observed amongst northern rural and remote populations, particularly amongst low-income ethnic minority populations when compared to Caucasian, urban and higher socio-economic status population counterparts in northern BC (Hill, Chau, Luebbering, Koivras, & Zoellner, 2012; Sarkar et al., 2010). Although the literature shows some evidence that the pattern of dietary changes for ethnic groups differs from one country to another, and even from one region to another, only few studies were conducted in Canada to assess the dietary patterns of ethnic groups; and none of these studies were conducted in northern communities where there are knowledge gaps as to food patterns and lifestyle of ethnic populations. 1.2 The Purpose of the Study The primary objective of this quantitative epidemiological study is to explore the dietary patterns of ethnic non-Indigenous populations in northern BC. This quantitative study included administering surveys to a convenience and snowballing sample to identify the dietary intake characteristics for such populations , and the healthy versus unhealthy patterns of their diet. In this community-based study, development of survey instrument and data collection was supported by key individuals and community leaders who played 5 DIETARY PATTERNS OF ETHNIC POPULATIONS significant role in raising the sense of ownership to the research project within the community. Key individua ls in each ethnic community were key to assist in the recruitment of other members of each ethnic group, and also for gaining the trust and support of other possible participants in their community at large. In this study, the following were the main objectives of the study:  To identify the characteristics of food & nutrient intake, and dietary pattern structure for the overall study sample based on responses to the semi-quantitative food frequency questionnaire. This objective is explored using principal components analysis (PCA).  To examine the influence of sociodemographic variables on the dietary behaviour of the study sample, and to test whether there are any gender differences regarding body weight status.  To examine the relationship between dietary patterns of the study sample and body mass index (BMI) used as a measure of healthy body weight status after controlling for sociodemographic confounders which include age, gender, education, income, and marital status. The purpose of this objective is test dietary patterns as predictors for BMI using multiple regression model.  To provide recommendations for ethnically appropriate public health promotion strategies through tailoring dietary interventions and nutritional education programs to promote healthy body weight and reduce the risk of obesity related chronic diseases. 6 DIETARY PATTERNS OF ETHNIC POPULATIONS Chapter 2: Literature Review Recently, there has been great emphasis on studying the overall dietary patterns rather than individual nutrients when examining chronic disease epidemiology in relation to diet through (Tucker, 2010). Conceptually, there are two main approaches to derive dietary patterns in nutritional epidemiology depending on whether the study is being driven by a specific dietary guideline (i.e., hypothesis oriented) or exploratory in nature (Schulze & Hoffmann, 2006). In hypothesis-oriented approach, dietary indices or scores following specific pre-determined dietary guidelines are used to assess the degree to which an individual or a group of individuals adhere to the recommended dietary pattern (Désilets, Rivard, Shatenstein, & Delisle, 2007). For example, the Healthy Eating Index (HEI) is used as a measure to assess the degree to which the food consumed by an individual adheres to the recommendations of the US Department of Agriculture Food Guide Pyramid and the recommendations in the US Dietary Guidelines for Americans (Hu 2002; Kennedy, Ohls, Carlson, & Fleming, 1995; U.S. Department of Agriculture, 1992). There have been previous modifications to the 1995, 2005, and 2010 versions of the HEI to develop Canadian adaptations of the measure following Canada’s Food Guide recommendations, known as Health Eating Index-Canada (HEI-C) (Jessri, Ng, & L’Abbe, 2017). In exploratory approach, existing dietary patterns are derived empirically from the available data without any guiding hypothesis (Schulze & Hoffmann, 2006). This is applied using exploratory analysis methods such as PCA and factor analysis, or cluster analysis. Cluster analysis is only preferred in case of planning nutritional intervention targeted to specific subgroups of the population (Tucker, 2010). On the other side, PCA has higher power when comparing linear variables with health outcomes, relative to multiple subgroups generated by cluster analysis (Tucker, 2010). 7 DIETARY PATTERNS OF ETHNIC POPULATIONS Exploratory PCA and factor analysis are considered as multivariate data reduction methods that produce components or factors, which account for the largest amount of variation in consumed food amongst individuals (Schulze & Hoffmann, 2006). The significant principal components or factors can produce component scores or factor scores, which could be used in multiple regression analysis to investigate the relationship between dietary patterns and health outcomes (Tabachnick & Fidell, 2013). PCA is more commonly used to identify dietary patterns because principal components scores represent mathematical transformation of the observed variables, whereas factor scores are only estimates of where each individual stand on the unobservable factor (Hu, 2002). To identify dietary patterns, Information on food consumption could be obtained through various data collection tools such as diet records, multiple 24-h recalls, and food frequency questionnaires (FFQ) (Jones-McLean et al., 2010). Although all of these data collection methods have been previously used in different dietary analysis studies, FFQ has advantage over other methods because it takes into account that consumed food changes from one day to another. Thus, it produces a better representation of the long term dietary pattern rather than pattern of a single day diet (Tucker, 2010). Furthermore, FFQ is considered as a cost-effective method for assessing dietary intake over a period of time in any given population (Quatromoni et al., 2002). Further details about methods used for data collection and analysis are discussed in chapter three. In a review of the literature, few dietary patterns were repeatedly identified during dietary pattern analysis. Most of the studies originating from North America and Europe that focused on the Caucasian populations has identified two main dietary patterns, namely “prudent” healthy and “western” unhealthy pattern (Kant, 2004). The prudent diet pattern is 8 DIETARY PATTERNS OF ETHNIC POPULATIONS rich in vegetables, fruits, legumes, whole grains, and fish, while the western diet pattern is rich in red meat, processed meat, refined grains, potatoes, and high-fat dairy products (Hu et al, 2000). Another important heart healthy diet pattern identified in the literature is the Mediterranean diet pattern, which is rich in vegetables, fruits, olive oil, nuts, and fish (Whayne, 2014). The Mediterranean diet pattern proved to be superior to other diet patterns regarding its health benefits as there is epidemiological evidence on the cardiovascular protective effect of the Mediterranean diet pattern (Enas, Senthilkumar, Chennikkara, & Bjurlin, 2003), which has been previously concluded through multiple studies including the randomized trials performed in the Lyon Diet Heart Study (de Lorgeril et al., 1999) and the PREDIMED study (Sala-Villa et al., 2014 & Toledo et al., 2013). Despite the fact that Mediterranean diet have deep roots in the south European region, further focus on the Mediterranean diet is warranted due to the study inclusion of significant number of ethnic participants from different regions that adopt Mediterranean or Mediterranean like diets including North Africa, Middle east, and western Asia (Altomare et al., 2013). Add to that, the promising prospect of using this healthy dietary pattern in nutritional interventions to high risk groups in North America. Beyond North America and Europe, several traditional ethnic diet patterns have been identified in the literature. For example, the tradition Korean pattern, which is rich in anchovies, Soybean, and Kimchi sea (Yang et al., 2005), the traditional Latin American pattern, which is rich in bean and rice (Marchioni et al., 2005), and the Chinese diet pattern which is rich in vegetables, soy, and fish (Nettleton et al., 2006). In my review of the literature, there will be further focus on the characteristics of the traditional diets for the two 9 DIETARY PATTERNS OF ETHNIC POPULATIONS largest ethnic subpopulations in northern BC, namely the traditional Chinese and south Asian diets. 2.1 Dietary Acculturation Acculturation is referred to as a mechanism by which various groups and individuals adapt to the norms, values and beliefs of a host culture while having the freedom to adhere to the original ethnic cultural norms and values (Broesch & Hadley, 2012). More specifically, dietary acculturation is determined by “the relationship between the uptake of ‘new’ nontraditional foods common in host culture, and the use of traditional food items” (Khamis, 1996, p. 46). Dietary acculturation and adoption of nutritional practices and diet similar to the host is closely related to the length of residence in the host country, which is considered as a marker of dietary acculturation. Sanou and colleagues found the duration of stay in Canada to be the primary measure of acculturation in most of the studies that investigated dietary habits of immigrants. Other significant measures of acculturation included country of origin, age at arrival to Canada, and official language proficiency Sanou et al., (2014). Recently, there has been an increased attention to the effect of acculturation on diet and health in minority groups (Ngo, Gurinovic, Anderson, & Majem, 2009). According to a recent Scopus review by Gilbert and Khokhar (2008), dietary habits of ethnic communities in Europe shift towards unhealthy patterns due to increased consumption of processed foods (e.g., processed meat, sweetened breakfast cereal), which are high in sugar and fat, rather than healthy diet (e.g., fruits, vegetables). This shift results in predisposition to obesityrelated morbidities due to increased fat, sugar, and salt intake (Ngo et al., 2009). These findings are consistent with the association between the increasing prevalence of obesity among Canadian immigrants and the longer duration of stay in Canada (Wahi, Boyle, 10 DIETARY PATTERNS OF ETHNIC POPULATIONS Morrison, & Georgiades, 2014). However, there seems to be lack of consensus in the literature regarding the risk of obesity amongst South Asian children in the UK relative to Caucasians (El-Sayed, Scarborough, & Galea, 2011). 2.2. Review of Major Dietary Patterns in Literature 2.2.1 Mediterranean Diet Furthermore, the literature shows some evidence that more acculturated immigrants (e.g., those who have longer durations of stay in a western host country) have mixed food habits where some healthy dietary habits such as vegetables and fruit consumption are replaced by unhealthy dietary habits such as fast foods consumption (Vargas & Jurado, 2016); This is explained by the adoption of more westernized lifestyle, including western dietary pattern, associated with acculturation of immigrants (Davis et al., 2013; Ghaddar, Brown, Pagan, & Diaz, 2010; Landman & Cruickshank, 2001). Dietary acculturation amongst ethnic groups is mainly influenced by food availability, socio-economic status, and cultural factors (Ottesen & Wandel, 2012; Satia, Patterson, Neuhouser, & Elder, 2002; Shatenstein, & Ghadirian, 1998). Some researchers attribute dietary changes to the unavailability of traditional cooking equipment or ingredients, and the availability of fast food outlets in the host countries, especially in rural environments (Dunn, Sharkey, & Horel, 2012; Varghese & Moore-Orr, 2002). Other researchers indicate that altered meal pattern is a key factor in changing ethnic dietary pattern (Garduño-Diaz & Khokhar, 2013). A recent study about the dietary patterns of Indian immigrants in Newfoundland revealed that some of the traditional dietary habits were abandoned (Varghese & Moore-Orr, 2002). There was also a trend of increased fat and energy intake, reduction of carbohydrates, 11 DIETARY PATTERNS OF ETHNIC POPULATIONS switch from whole grains into more refined sources of carbohydrates, and increased intake of meat and dairy products. These dietary changes were consistent with results from other studies of south Asian immigrants in Europe (Gilbert & Khokhar, 2008; Ottesen & Wandel, 2012). A study conducted in United Kingdom (UK) to identify dietary patterns of South Asian ethnic population revealed that dietary acculturation was associated with significant changes in the ethnic food habits of immigrants who resided in the host country for longer periods of time (Garduño-Diaz & Khokhar, 2013). These changes included lower proportion of immigrants abstaining from meat consumption (i.e., less number of vegetarians), change in eating habits by having two meals per day instead of the traditional three meals per day pattern, and increased intake of caffeinated drinks. Such changes were associated with higher predisposition of type 2 diabetes (T2D) and Metabolic Syndrome (MetS) among South Asian individuals in the UK who shifted to a WDP, when compared to recent South Asian immigrants or the populations in their respective native countries. A study by Lesser, Gasevic & Lear (2014) revealed that the dietary behaviour of South Asian immigrants in Canada is largely affected by dietary acculturation. In this study, there has been an indication regarding the shift into healthier methods of food preparation (e.g., stir-frying and grilling instead of deep frying), and higher intake of vegetables and fruits. However, there was also an indication of unhealthy food habits including increased consumption of convenience foods, sugar-sweetened beverages, red meat, and increased dinning out instead of home meals; which increased the risk of T2D and cardiovascular disease (CVD) in South Asian immigrants. Satia et al. (2001) found that western dietary acculturation of Chinese women who resided in Vancouver BC, and Seattle WA, was associated with higher intake of fatty and junk foods. The study also revealed an important 12 DIETARY PATTERNS OF ETHNIC POPULATIONS association between dietary acculturation and certain socio-demographic characteristics such as young age, high education, employment, and longer residence in North America. Another study revealed that longer length of residence (i.e., number of years lived in the host country) among Chinese immigrants is associated with an increased intake of juice, soft drinks, and chips (Rosenmoller, Gasevic, Sidell, & Lear, 2011). The Mediterranean diet is considered as a heart-healthy dietary pattern when compared to other traditional diets (Panagiotakos, Polystipioti, Papairakleous, & Polychronopoulos, 2007). This Traditional Mediterranean diet is rich in vegetables, fruits, nuts, healthy plant fats (i.e., extra virgin olive oil), cereals and grains, with high legume consumption. It is also characterized by moderate intake of fish and dairy products, and low intake of sugary food products, red and processed meats, and dairy products as shown (Whayne, 2014). Figure 2 shows the Mediterranean diet pyramid. A systematic review by Garcia-Fernandez et al., (2014) provided strong evidence that Mediterranean diet pattern is associated with reduced risk of cardio-diabetic morbidity and mortality. The Mediterranean diet has plenty of physical and mental health benefits, and regular exercise is an essential component of a Mediterranean life style (Henrique-Sanchez et al., 2012). In a study performed in the United States (USA), there was also strong evidence that linked the adherence to a Mediterranean diet with reduction in risk of death from all causes (Mitrou et al., 2007). The beneficial protective effects of Mediterranean diet extend to other non-communicable diseases including cancer, Parkinson’s disease, Alzheimer’s disease, and gout disease (Perona et al. 2010; Whayne, 2014). 13 DIETARY PATTERNS OF ETHNIC POPULATIONS Figure 2. Mediterranean diet pyramid 14 DIETARY PATTERNS OF ETHNIC POPULATIONS My review of the literature also revealed an interesting finding regarding the positive effect of Mediterranean diet on health amongst specific migrant groups in France despite their limited access to primary health care services (Darmon & Khlat, 2001). The authors attributed the paradoxical lower rates of chronic diseases and higher life expectancy amongst North African and South European migrants when compared to the local French born, to the Mediterranean dietary pattern of these groups. Their food habits consisted mainly of high consumption of fruits, vegetables, and starchy products and lower consumption of alcohol. However, such positive effects are hypothesized to be lost within the second generation of migrants due to dietary acculturation and adoption of different dietary pattern similar to the local population (Darmon & Khlat, 2001). 2.2.1.1 Review of Studies: Mediterranean Diet and Cardiovascular Disease In a multi-center, randomized controlled trial (RCT) known as the PREDIMED trial; participants with high cardio-vascular risk were assigned to either one of two intervention groups, Mediterranean diet supplemented with extra-virgin olive oil group, and Mediterranean diet supplemented with mixed nuts group, or a low fat diet control group. The results of the study suggested a causal association between a Mediterranean-type diet supplemented with extra-virgin olive oil or mixed nuts and the reduction in independent cardiovascular risk factors, which include blood pressure, plasma lipid levels, and carotid atherosclerosis (Buil-cosiales et al., 2008; Estruch et al., 2006; Sala-Villa et al., 2014 & Toledo et al., 2013). There has also been a similar relationship between Mediterranean diet and the prevention of major cardio-vascular events (i.e., stroke, myocardial infarction, and cardiovascular mortality) by reducing the incidence of such events by approximately 30% (Estruch et al., 2013; Panagiotakos, 2014; Ros et al., 2014). Furthermore, Damasceno et al. 15 DIETARY PATTERNS OF ETHNIC POPULATIONS (2013) asserted that consumption of a Mediterranean diet enriched with nuts plays a cardiovascular protective role through shifting lipoprotein sub-fractions to a less atherogenic pattern. Tresserra-Rimbau et al. (2014) revealed an association between greater intake of polyphenols, rich in vegetables and fruits in Mediterranean diet, and decreased cardiovascular risk. Martínez-González et al. (2014) suggested that the anti-oxidant and antiinflammatory effect of extra-virgin olive oil, in the context of Mediterranean dietary pattern, may play a significant role in reducing the risk of atrial fibrillation. Furthermore, a casecontrol study by Mattioli et al., (2013) revealed that patients with atrial fibrillation who had better adherence to Mediterranean diet were more likely to experience a better outcome through spontaneous conversion of atrial fibrillation. 2.2.1.2 Review of Studies: Mediterranean diet and Type 2 Diabetes The literature review revealed inconsistent results regarding the role of a Mediterranean diet in reducing the incidence of type 2 diabetes (T2D). Some researchers assert that Mediterranean diet plays a significant role in the primary prevention of type 2 diabetes (T2D). An extensive meta-analysis that included 136,846 participants revealed that adherence to the Mediterranean diet was associated with 23% reduced risk of developing type 2 Diabetes (T2D) (Koloverou, Esposito, Giugliano, & Panagiotakos, 2014). Furthermore, a secondary analysis in a subgroup of the PREDIMED trial revealed that Mediterranean diet supplemented with extra virgin olive significantly reduced the risk of type 2 diabetes (T2D) in individuals with high cardiovascular risk (Salas-Salvadó et al., 2014). However, the Multi-Ethnic Study of Atherosclerosis (MESA) found no association 16 DIETARY PATTERNS OF ETHNIC POPULATIONS between adherence to the Mediterranean diet and incidence of diabetes despite the reported lower insulin levels among non-diabetics (Abiemo et al., 2013). In patients with type 2 diabetes, consumption of Mediterranean diet has a significant role in management and control of the disease, with better prognosis and less complications. Esposito et al. (2014) revealed an association between adherence to Mediterranean diet and reduced HbA1c level, lower rate of diabetes remissions, and lower need for diabetic medications. Furthermore, Ceriello et al. (2014) concluded that a Mediterranean diet, using olive oil, prevents the acute hyperglycemia effect on the endothelial functions in T2D patients, which reduces the risk of CVD in such patients. 2.2.1.3 Review of Studies: Mediterranean diet and Metabolic Syndrome Several epidemiological studies investigated the association between adherence to Mediterranean diet and the reduced risk of incidence of MetS. Babio, Bullo, & SalasSalvado, (2009) asserted that consumption of Mediterranean diet is associated with reduced risk of the individual components of MetS. The ATTICA study utilized a cross-sectional survey to collect data about the dietary patterns of 2282 subjects in Greece. The results showed that adoption of Mediterranean diet was associated with 20% lower odds of having metabolic syndrome (Panagiotakos et al., 2004). In the SUN prospective cohort study, 3,497 participants were followed up after six years for adherence to Mediterranean diet, waist circumference, blood pressure, plasma glucose level, HDL cholesterol level, and triglycerides (Tortosa et al., 2007). The authors concluded the presence of an inverse relationship between adherence to Mediterranean diet and the cumulative incidence of the MetS. The dietary intervention study by Esposito et al., (2004) revealed that following a Mediterranean diet for patients with MetS resulted in 51% net risk reduction of having the 17 DIETARY PATTERNS OF ETHNIC POPULATIONS disease, when compared to patients with MetS who followed a prudent diet. Many other recent clinical trials provided clear evidence about the beneficial role of Mediterranean-style diet in the management and reversion of metabolic syndrome (Babio et al., 2014; SalasSalvado et al., 2008). Those findings have been explained by the reduction in oxidative damage of lipids and DNA, and the anti-inflammatory property of this dietary pattern (Esposito et al., 2004; Mitjavila et al., 2013). 2.2.2 South Asian Diet Traditional South Asian diet is characterized by high intake of vegetables and fruits, and is also rich in cereals, legumes, and spices. It consists of low intake of meat, and low to moderate intake of fat with relatively high intake of saturated fat due to the widespread use of hydrogenated oils in South Asian cooking (i.e., Vanaspati). Some of the traditional patterns of food preparation for traditional South Asian diet include unhealthy food preparation practices such as deep frying. Although traditional South Asian diet is considered healthier diet when compared to western diet style, South Asian populations are at high risk of dyslipidemia, type 2 diabetes, cardiovascular disease, and metabolic syndrome (Vuksan et al., 2012) due to high intake of saturated fat and salt (Lesser et al., 2014). Misra, Khurana, Isharwal, and Bhardwaj (2009) attributed the higher susceptibility to develop insulin resistance and MetS amongst Asian Indians and South Asians to dietary nutrients rather than a genetic role; Low intake of multi-unsaturated fatty acids (MUFA), and high intake of saturated fats, trans-fatty acids, and carbohydrate in traditional Indian diets result in several nutritional imbalances that are associated with higher prevalence of dyslipidemia and insulin resistance in South Asian ethnicity. Furthermore, the risk of such diseases increases significantly for South Asian immigrants when adopting mixed food habits 18 DIETARY PATTERNS OF ETHNIC POPULATIONS with higher intake of sugar and fat (Gilbert & Khokhar, 2008). According to the Southhall study by McKeigue & Shah (1991), the prevalence of type 2 diabetes among South Asian immigrants was four times that of their Caucasian counterparts in London, UK (20% vs. 5%) despite matching for age and BMI. Furthermore, UK Indian Asians were at higher risk of CVD and Metabolic Syndrome MetS when compared to the mainstream Caucasian population; including higher glucose and insulin levels, lower HDL cholesterol level, higher systolic blood pressure, and higher waist to hip ratio (Cleland & Sattar, 2004). 2.2.3 Chinese Diet The traditional Chinese diet pattern is characterized by high consumption of rice, noodles, coarse grains, fresh vegetables, fruit, fish, shrimp, and beans. The ingredients and cooking style of traditional Chinese diet vary from one region to another (Gilbert & Khokhar, 2008). To illustrate, dietary intake in northern China is high in wheat, wheat products, and maize known as the “yellow earth” or “traditional northern” pattern, whereas dietary intake in southern China is high in rice, vegetables, and seafood known as the “green water” or “traditional southern” pattern (Batis et al., 2014). Dietary intake of Chinese populations is greatly influenced by food beliefs and the yin/yang belief system of traditional Chinese Medicine (TCM). According to this belief system, health is dependent on a balance of intake between cold foods (yin) such as honey and yogurt, and hot foods (Yang) such as ginger (Lee & Shen, 2008). The literature shows that traditional Chinese diet pattern is associated with various health benefits. A study by Shu et al. (2015) revealed that traditional Chinese diet pattern is inversely associated with BMI and waist circumference in both genders. Furthermore, a recent cross sectional study investigated the association between traditional Chinese diet pattern and impaired fasting glucose (i.e., prediabetes) concluded that increased 19 DIETARY PATTERNS OF ETHNIC POPULATIONS consumption of vegetables, fruit, and cereals in traditional Chinese diet reduces the risk of development of type 2 diabetes (Zhang et al., 2015). 20 DIETARY PATTERNS OF ETHNIC POPULATIONS Chapter 3: Design and Methods A cross-sectional survey design was used as quantitative method for data collection. The study received UNBC Research Ethics Board approval for a period of 12 months on March 1, 2016 (Appendix D), and then later extended for another year. A renewal approval for another 12 months period was received on February 28, 2017. The study collected primary data on food consumption patterns for ethnic populations in northern BC through administration of a paper-and-pencil survey instrument (i.e., food frequency questionnaire). This approach intended to allow the researcher to gain knowledge about the diet behavior & food habits, life style, and health status of multi-ethnic adult population in northern BC, and also to explore factors that influence their dietary patterns. Multivariate statistical techniques were employed to explore dietary patterns using principal component analysis (PCA). The researcher was also interested in identifying relationship between dietary components and healthy body weight status, as assessed by body mass index (BMI) [BMI=weight(kg)/height(m2)], using multiple regression analysis technique. 3.1 Recruitment Strategy Participants from different non-Indigenous and non-Caucasian ethnic minorities in Northern BC were recruited to participate in the study. Partnerships were established with potential community stakeholders including Immigrant & Multi-Cultural Service Society (IMSS), Prince George Islamic Centre (PGIC), Prince George Tennis Club (PGTC), Northern Undergraduate Student Society (NUGSS), Northern British Columbia Graduate Student Society (NBCGSS), Gurugobind Singh Temple Association, and Gurudwara Sahib Guru Nanak Darbar Sikh Temple, Local Hindu Community. Research posters were attached to bulletin boards at UNBC, IMSS, and PGIC to recruit potential participants. 21 DIETARY PATTERNS OF ETHNIC POPULATIONS The researcher conducted a series of research recruitment events in collaboration with stakeholders between March 2016 and March 2017. UNBC research recruitment dinner event was held on March 19, 2016. Electronic recruitment invitations were sent out by email to UNBC graduate students and other community members from different ethnic minorities for this event. Social media was also used to recruit participants from UNBC and the local community (i.e., facebook event). Also, the researcher held a daily research booth for data collection at UNBC winter garden over a three-week period extending between end of March 2016 and mid April 2016. The Immigrant Multicultural Service Society (IMSS) is a non-profit organization that provides a multitude of programs and services to new comers to Canada, recent immigrants & temporary foreign workers, who reside in northern BC. These programs and services include free English classes, settlement programs, and specific programs for youth and women. The overall goal of the organization is to create more welcoming and inclusive communities for immigrants through successful settlement and integration (IMSS, 2015). The researcher recruited current and previous IMSS clients through collaboration with the organization. Free information session for immigrants regarding healthy eating habits was organized on April 26, 2016 parallel to a research recruitment event. The information session was specifically conducted after data collection to avoid introducing any bias into participant responses. The researcher carried out a research recruitment event during Holi celebration of light organized by local Hindu community in Prince George on April 2, 2016. The researcher provided a brief verbal recruitment message to all attendees to facilitate participation. All attendees were given an option to send the completed survey though pre- 22 DIETARY PATTERNS OF ETHNIC POPULATIONS paid stamped envelope to facilitate recruitment. The researcher conducted another research recruitment event during the Vaisakhi Celebration organized by Gurudwara Sahib Guru Nanak Darbar Sikh Temple on May 21, 2016. A similar event was conducted at the Gurugobind Singh Temple on March 26, 2017. Multiple small recruitment events were conducted at the Prince George Islamic Centre (PGIC) during the summer of 2016. The researcher directly recruited some members of the Prince George tennis club during the summer season of 2016. Lastly, two organized research recruitment events were carried out at Lheidli T'enneh Memorial Park on May 7 and May 14, 2016. Both event included participants with Arabic, Korean, and Caribbean heritage. Participants were entered into a draw to win a monetary prize (i.e., iPad tablet and two fitbit trackers) to facilitate recruitment. Funding for those monetary prizes was secured through a graduate UNBC Research Project Award (RPA). A draw for prizes was conducted on March 29, 2017 to select three recipients, amongst all research participants. The researcher privately informed the winners to maintain their privacy and confidentiality. The initial three drawn participants claimed their prizes within the pre-specified 10 days period in the research proposal. No subsequent draws were required for this purpose. 3.1.1 Sample Size and Sampling Method The survey instrument was administrated via paper and pencil method to 444 participants above 19 years of age or older from both genders. The estimated sample size of this study was determined a priori to need to be somewhere between 400-500 participants. The rationale behind sample size is to have optimal statistical power to identify statistically and clinically significant effects, especially for Principal Component Analysis of dietary intake variables. Further, Suresh & Chandrashekara (2012) indicates that “Descriptive 23 DIETARY PATTERNS OF ETHNIC POPULATIONS studies need hundreds of subjects to give acceptable confidence interval for small effects” p. 8. Participants were recruited to this study using convenience sampling technique through approaching ethnic groups in northern BC. A snowball sampling technique was utilized as a sub-method of convenience sampling. Snowball sampling was very effective in recruiting large number of participants to the study. The initial group of participants acted as the “seeds” to recruit other eligible participants in subsequent waves. This has been seen largely during the research booth event held at UNBC. Some of the initial participants in the UNBC research dinner event were advocates for this study and recruited potential participants in their social cycle. Key individuals and community leaders in different ethnic groups were also identified as “seeds” to recruit other members within their ethnic communities. Multiple waves of snowball sampling were very effective in recruiting the targeted sample size (U.S. Geological Survey, 2015). Respondent-driven sampling, a modified form of snowball sampling was used to weight the sample and target specific population demographic. The intent for using this technique was to partially compensate for lack of random sampling and to recruit a more diversified sample (Abdul-Quader et al., 2008). Peers are considered more effective than researchers in locating and recruiting other individuals from a “less accessible” population such as ethnic population. Snowball sampling is an excellent method in recruiting individuals from unknown populations that have not been previously identified and are more difficult to contact. It provides a great opportunity for better understanding and complete characterization of unknown populations as well as being quick and cost-effective sampling method (U.S. Geological Survey, 2015). In a pilot study for recruiting drug users in New York City, Abdul-Quader et al. (2008) managed to recruit 24 DIETARY PATTERNS OF ETHNIC POPULATIONS 618 participants in 13 weeks, beginning with only eight seeds. In my study, I managed to recruit 277 participants in 15 weeks using convenience and snowball sampling techniques. The rest of the sample was mainly recruited following respondent-driven sampling and later waves of snowball sampling until targeted sample size was reached. In this study, inclusion criteria included ethnic, non-Indigenous, and non-Caucasian adults living in northern BC who are 19 years of age and older. Exclusion criteria included participants being on special diet due to health or medical reasons (i.e., kidney failure andceliac disease), or participants residing outside northern BC. 3.1.2. Response Rate There were 745 delivered paper-based surveys with 444 being returned. The number of participants recruited from different events is displayed in table 1. Despite survey length, there was approximately 59% response rate. Factors that might have helped to maximize response rate include providing participants with information about the purpose/importance of the study and the scarce research addressing health implications of dietary transitions in ethnic minorities in specific, and health status of immigrants and ethnic groups in general, especially in northern BC context. Also, the researcher offered collecting completed surveys through different venues including offering pre-paid stamped mail envelops, collecting completed surveys at the researcher’s UNBC office, collecting surveys at the UNBC research booth, and personally collecting completed surveys. Offering financial incentives for a prize could be another factor in increasing response rate. Further, development of an effective recruitment plan was a key for high level of participation in this study. 25 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 1 Recruitment Event Date and Number of Participants Event UNBC Research Recruitment Dinner Event UNBC Research Recruitment Booth Holi Celebration Research Recruitment Event IMSS Research Recruitment Luncheon Event Lheidli T'enneh Memorial Park First Research Recruitment Event Lheidli T'enneh Memorial Park Second Research Recruitment Event Vaisakhi Celebration Research Recruitment Event Prince George Islamic Centre Multiple Research Recruitment Events Gurugobind Singh Temple Research Recruitment Event Date Number of Participants March 19, 2016 85 From March 22, 2016 To April 11, 2016 April 2, 2016 103 April 26, 2016 42 May 7, 2016 12 May 15, 2016 27 May 21, 2016 14 Summer 2016 24 March 26, 2017 15 14 3.2. Data Collection Data were collected on food consumption patterns (i.e., semi-quantitative FFQ), measures of dietary acculturation (i.e., length of stay in host country, change in diet following relocation), information on self-reported health measures such as body weight status (i.e., BMI), level of physical activity, risk factors for chronic disease chronic disease (e.g., blood cholesterol level, diabetes status, and blood pressure level), and sociodemographic variables (e.g., education level, income level). A food frequency questionnaire (FFQ) was administered to participants in person. The researcher received an internal award from UNBC for this research study (i.e., Research Project Award – spring 2016 competition). 26 DIETARY PATTERNS OF ETHNIC POPULATIONS The funding secured for this research supported hiring research assistants, printing research surveys, organizing recruitment events, and purchasing prizes to facilitate recruitment. The research team included three undergraduate research assistants and nine volunteers. All volunteers and research assistants received relevant training and signed a confidentiality and non-disclosure agreement prior to their involvement in research recruitment, event organization, or data collection. During the data collection phase, the researcher explained the purpose of the research to potential participants and provided answers and clarifications related to any questions or concerns. All participants were required to provide an informed consent prior to participation in the study and filling the survey. In the information letter and consent form, participants were informed about the ability to withdraw from the study at any point. All participants were assured that their identities will be protected in any future reports or publications, and that all collected data will remain confidential and stored in a secured cabinet in the researcher’s office. Participants who were interested in knowing the results of this research were informed that a copy of the research thesis will be available at the UNBC library once the study is completed, and a summary report of the research results will be emailed to interested participants. 3.2.1. Survey Questionnaire Development Quantitative data were collected via administering a cross-sectional survey to participants. The survey instrument includes information and consent sheet, demographic questions, FFQ instrument based on modification of the semi-quantitative FFQ by Hu et al. (2000) food groupings (Table 2), and dietary acculturation assessment component adapted from Rosenmoller et al., (2011). The semi-quantitative FFQ was used in this study to 27 DIETARY PATTERNS OF ETHNIC POPULATIONS provide an objective assessment to explore the dietary patterns of participants (Appendix A). The reproducibility and validity of this instrument has been previously reported in other studies (Rimm et al., 1992; Feskanich et al., 1993). An Additional demographic question was added to the questionnaire to measure length of residence in the host country, as it is one of the markers for dietary acculturation documented in the literature. Regarding dietary assessment items, some questions were modified based on a literature review and informal communication with ethnic individuals from different heritages in order to make the survey more sensitive to different ethnic diets, and to improve the reliability of the answers without affecting the reported validity and reproducibility of the FFQ. For example, question 40 was modified to include “Ghee” other than butter and margarine to increase the reliability of answers particularly from south Asian participants. Modifications and new variables added to the adapted survey instrument were pilot tested to ensure clarity of wording and proper comprehension. The survey instrument included sociodemographic variables (i.e., age, gender, education level, annual household income, current marital status, ethnicity, place of birth, and number of years lived in Canada or any western country, and self-reported measures of weight and height to derive BMI. The dietary acculturation assessment component of the survey has been adapted fully from Rosenmoller et al. (2011) with the addition of two sub-items. This segment of the survey was only applicable to immigrants who were born in a foreign country, and then relocated to Canada in a later stage of their life. The data were used to explore dietary acculturation within first generation immigrants, in addition to studying perceptions of participants regarding diet related issues following immigration to Canada. These issues include change in amount of food intake, awareness of healthy and unhealthy diet, change in 28 DIETARY PATTERNS OF ETHNIC POPULATIONS dietary pattern, change in method of food preparation, frequency of dining out, and availability of ethnic food ingredients & healthier diet choices. The survey instrument was provided in English language. The researcher provided answers and clarifications to participants during all research data collection events. The researcher applied various methods to overcome language in this study. During the Vaisakhi research recruitment event, one research assistant who is proficient in English and Punjabi assisted some participants who were not proficient in the English language and communicated inquiries and answers between participants and the researcher whenever needed. Prior to the IMSS research event, it was known to the researcher that some of the potential participants, especially those who are more recent immigrants, have language barrier to some degree, which could potentially limit their ability to identify different food items in the survey questions. The researcher utilized a projector screen during survey administration to show pictures of dietary items in each question of the FFQ. This was done to help assist participants with language barrier to mitigate missing and unreliable responses. In addition, some participants had assistance from their family members who were proficient in English. 29 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 2 Food Groupings Adapted from Hu et al. (2000) 30 DIETARY PATTERNS OF ETHNIC POPULATIONS 31 DIETARY PATTERNS OF ETHNIC POPULATIONS 3.2.2 Modifications to the Cross-Sectional Survey Instrument On one hand, modifications to the survey instrument were meant to include new relevant variables related to dietary acculturation. On the other hand, some of these modification were intended to make the semi-quantitative FFQ more culturally sensitive in order to increase the validity and reliability of the responses. Modifications to the population demographics portion and the semi-quantitative FFQ a) q65 was added to the demographic questions as follows: q65. How many years have you lived in Canada or any other western country? <1> Since birth <2> Since the year of 19…/20… b) q6e was modified through adding passion fruit to the list of exotic fruits as follows: q6e. EXCLUDING juices, over the past month, how often on average did you eat exotic fruits such as papaya, mango, avocado, passion fruit or pineapple including canned exotic fruits such as canned papaya, canned mango, and canned pineapple? c) q11 was modified through adding paratha, roti, and chapati to the list of products made of white flour as follows: q11. Over the past month, how often on average did you eat products made of white flour, such as white bread, pita, naan, paratha, roti, chapati muffins, bagels, buns, rolls, muffins, biscuits, pancakes, waffles, or plain crackers? d) q12 was modified through adding whole wheat paratha, whole wheat roti, and whole wheat chapati to the list of whole wheat foods as follows: q12. Over the past month, how often on average did you eat whole wheat foods, such as whole wheat bread, whole wheat pita, whole wheat naan, whole wheat paratha, whole 32 DIETARY PATTERNS OF ETHNIC POPULATIONS wheat roti, and whole wheat chapati, whole wheat muffins, whole wheat bagels, whole wheat buns, whole wheat rolls, whole wheat biscuits, whole wheat pancakes, whole wheat waffles, or whole wheat crackers? Whole wheat means either 60% or 100%? e) q13 was modified through adding rice pudding or kheer, and rice noodles as follows: q16. Over the past month, how often on average did you eat white rice including rice pudding or kheer, pasta, spaghetti, macaroni, noodles, rice noodles, or polenta? f) q27 will be modified through adding sesame butter as follows: q27. Over the past month, how often on average did you eat peanut butter, or sesame butter? g) q29 will be modified through adding papadum as follows: q29. EXCLUDING NUTS, over the past month how often on average did you eat snacks such as potato chips, corn chips, popcorn with butter, nachos, tortilla chips, or papadum? h) q34 was modified through adding paneer as follows: q34. Over the past month, how often on average did you eat paneer, cottage cheese, or ricotta cheese? i) q36 was modified through adding paneer to the excluded items as follows: q36. EXCLUDING paneer, cottage cheese, ricotta cheese, and cream cheese, over the past month how often on average did you eat cheese such as cheddar cheese, mozzarella cheese, brie, feta cheese or goat cheese? j) q40 was modified through adding some south asian and Chinese spices such as curry, turmeric, cardamon, fennel seeds, cumin, coriander, and Szechuan peppercorn as follows: 33 DIETARY PATTERNS OF ETHNIC POPULATIONS q40. Over the past month, how often on average did you add spicy condiments such as chilli sauce, salsa, mustard, pepper, curry, turmeric, cardamon, fennel seeds, cumin, coriander, Szechuan peppercorns, chutney, steak sauce, soy sauce, Worcestershire sauce, or wasabi to your food? k) q44 was modified through adding ghee as follows: q44. Do you prefer to use ghee, butter or margarine in your food? Modifications to the dietary acculturation assessment component of the survey instrument. a) q66 was added as follows: q66. Has your general portion size (how much you eat in one meal) changed since coming to Canada? (1- Eat much less, 3- No change, 5- Eat much more) b) q67 was added as follows: q67. How has your method of food preparation changed since coming to Canada? - Stir-frying/BBQ (1-Much less often, 3- No change, 5-Much more often) - Baking/grilling food (1-Much less often, 3- No change, 5-Much more often) - Boiling food (1-Much less often, 3- No change, 5-Much more often) - Deep frying food (1-Much less often, 3- No change, 5-Much more often) - Microwaving food (1-Much less often, 3- No change, 5-Much more often) - Eating vegetables raw (1-Much less often, 3- No change, 5-Much more often) - Home-made foods (1-Much less often, 3- No change, 5-Much more often) c) q68 was modified to include a new sub-item "Home-made foods" as follows q68. How have the foods you generally eat changed since coming to Canada? - Vegetables (1- Eat much less, 3- No change, 5- Eat much more) - Potatoes, rice (1- Eat much less, 3- No change, 5- Eat much more) - Convenience foods - Fruit (1- Eat much less, 3- No change, 5- Eat much more) - Soft drinks/pop/soda (1- Eat much less, 3- No change, 5- Eat much more) 34 DIETARY PATTERNS OF ETHNIC POPULATIONS - Dairy products (1- Eat much less, 3- No change, 5- Eat much more) - High fat/fried foods (1- Eat much less, 3- No change, 5- Eat much more) - Deserts/candy/sweets (1- Eat much less, 3- No change, 5- Eat much more) - White meat (1- Eat much less, 3- No change, 5- Eat much more) - Red meat (1- Eat much less, 3- No change, 5- Eat much more) - Restaurant meals/dining out (1- Eat much less, 3- No change, 5- Eat much more) d) q70 was modified to include a new sub-item “Finding ethnic food ingredients” as follows: q70. How do you feel about these issues in Canada compared to your home country? - Finding fresh fruit and vegetables: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada) - Finding low fat food options: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada) - Choosing healthy food when dining out: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada - Finding ethnic food ingredients: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada) 3.2.3 Pilot Testing of the Survey Instrument Data collection started with survey pilot testing phase in mid-March 2016. The survey instrument was pilot tested through administering it to eight participants to identify issues related to wording or clarity of questions. Pilot testing revealed that q58 was unclear “On the days that you are physically active, on average how many minutes do you usually spend being physically active in a typical week?” Participants were not sure whether the question requested duration of exercise per day or week. Upon consultation with the study supervisor, the question has been modified to avoid any misinterpretation “On average how 35 DIETARY PATTERNS OF ETHNIC POPULATIONS many minutes/hours do you usually spend being physically active in a typical week?”. No other changes or modification were applied. 3.3 Data Entry Upon receiving completed surveys, the researcher assigned a numerical order to each survey that generally reflects the order upon which it was received. For example, the first completed survey was assigned number “001”, while the last completed survey was assigned number “444”. Data were entered into SPSS® for Windows statistical software package version 24.0 (SPSS Inc.). During data entry, each completed survey was assigned a respondent ID number that consisted of seven numbers. The first four numbers represent “year of data collection”, and the last three numbers represent the numerical order. For example, the first received survey had the following respondent ID number “2016001”, while the last received survey had the following respondent ID number “2017444”. Assigning respondent ID number was introduced to avoid using personally identifiable information when referring to a specific survey. Research assistants were trained to use “respondent ID number” when manually logging data errors/issues for further examination by the researcher. Three undergraduate research assistants were hired to assist with data entry into SP SS 24.0 for Windows software program on the researcher laboratory office desktop. The researcher provided individual training to each research assistant. This training consisted of providing an overview on the research study, information on protecting privacy and confidentiality of participants, familiarizing trainees with the SPSS software program, practical training on data entry, and guidelines on dealing with data errors/issues e.g. manually logging data errors in a laboratory notebook with both respondent ID and question number. Research assistants were responsible for entering dietary intake responses for each 36 DIETARY PATTERNS OF ETHNIC POPULATIONS survey. The researcher entered socio-demographic & health status variables, and dietary acculturation responses for each survey, in addition to reviewing data entry errors/issues documented by research assistants. The researcher had two separate meetings with the study supervisor to come into agreement regarding resolving some of the data errors/issues. Data entry commenced midway during data collection phase, and was completed by the end of data collection phase. 3.4 Data Recoding and Analysis Following data entry, data was visually inspected as a check for data entry errors such as data points outside the normal range for each variable. Any detected errors were verified through referring to the original data. 3.4.1 Recoding of Dietary Variables For the purpose of convenience, the questionnaire was constructed in a way where answers were arranged from highest frequency of consumption to lowest frequency of consumption. Recording of all dietary variables intended to make interpretation easier, and data values more meaningful. Responses to all dietary intake variable were numerically recoded to make higher values represent higher frequency consumption and lower values represent lower frequency of consumption as follows: - A response <1> More than once a day was originally entered as data value “1”, and later recoded as value “6” - A response <2> Everyday (7days per week) was originally entered as a value “2”, and later recoded as value “5” - A response <3> 5-6 days per week was originally entered as a value “3”, and later recoded as value “4” 37 DIETARY PATTERNS OF ETHNIC POPULATIONS - A response <4> 2-4 days per week was originally entered as a value “4”, and later recoded as value “3” - A response <5> 1 day per week was originally entered as a value “5”, and later recoded as value “2” - A response <6> Less than 1 day per week (e.g., once or twice a month) was originally entered as a value“6”, and later recoded as value “1” - A response <7> Never was originally entered as a value “7”, and later recoded as value “0” 3.4.2 Deriving Body Mass Index Body Mass Index (kg/m2) is derived through dividing body weight in kilograms by height in meters squared (Poustie, Smyth, & Cole, 2005; Wallner-Liebmann et al., 2013). Height in meters was calculated using the following formula: Height in Meters = Height in Centimeters ÷ 100 Or Height in Meters = (Feet × 0.3048) + (Inches × 0.0254) Weight in Kilograms was derived from either stated weight in Kilograms or weight in Pounds using the following formula: Weight in Kilograms = weight in Pounds × 0.453592 3.4.3 Deriving Age in Years Age was mainly derived from “Year of Birth” variable rather than “Age in Years” variable. The following formula was used to derive age from “Year of Birth” variable: Age = Year of data collection – Year of Birth 38 DIETARY PATTERNS OF ETHNIC POPULATIONS For example, when a conflict between an age derived from “Year of Birth” variable, and an age derived from “Age in Years” variable for the same participant, the former would take precedence over the latter as it is thought to be a more reliable response due to less chance of memory bias. 3.4.4 Deriving Length of Stay/Residence in Years Length of residence for first generation immigrants was derived using the following formula: Length of residence = Year of data collection – Year of relocation Following a priori exclusion criteria, 30 cases with mainly Caucasian ethnicity were excluded from our study. Another three cases were excluded due to being on special diet regimen. Further, there was a consensus between the researcher and study supervisor to exclude another case due to consistent lack of reliability in answers i.e., multiple responses to each survey question. As a result, there were 410 cases that were deemed appropriate to enter the final analysis stage for this study 3.4.5 Data Analysis Techniques All data were analyzed using the statistical analysis systems (SPSS®) for Windows (version 24.0). Descriptive frequency analysis was carried out on all data (continuous and categorical variables). Descriptive statistics were used to describe the study sample. An exploratory type of factor analysis known as principal components analysis (PCA) was used to derive the factorial structure of the qualitative FFQ items. This exploratory data reduction technique can be used to analyze and explain the interrelationship among a set of variables (e.g., food items). Principal Component Analysis (PCA) was used as a reduction technique for dietary intake variable leading to few principal components accounting for most 39 DIETARY PATTERNS OF ETHNIC POPULATIONS significant amount of variability amongst dietary variables. The association between principal components scores and BMI was tested using multiple regression analysis after adjusting for socio-demographic variables which include age, gender, marital status, income, and education. Bivariate analysis of BMI by gender was also performed. 40 DIETARY PATTERNS OF ETHNIC POPULATIONS Chapter 4: Results The data were screened for missing responses. The amount of missing responses for 63 dietary items represents only 2.2%. Missing data across dietary items and cases had a relatively random pattern (Appendix B). Since missing data of dietary items accounted for quite a small percentage in a large dataset, using any method to address missing data would yield similar results (Tabachnick & Fidell, 2013). Deleting cases with any missing data seems impractical as it will lead to a much smaller sample size insufficient for analysis. For example, deleting a case with only one missing response, which represents less than 2% of all responses for this particular case, is not justifiable. Missing data could be estimated by a regression method or replaced by mean value; and using the regression method to replace missing data runs the risk of data overfitting and inflating correlation between variables (Tabachnick & Fidell, 2013). This could affect reliability of factors in PCA analysis and increase the risk of type I error in the subsequent multiple regression analysis. Also, estimated values through the regression method could fall outside the range of values for variables (i.e., 0-6). Thus, missing data was replaced using the mean. 4.1 Descriptive Statistical Analysis Following data screening step, descriptive statistics including distributional properties of all quantitative and qualitative variables were examined for univariate normality and linearity of data. Large sample size (n = 410) is considered important factor in normalizing distributions. 4.1.1 Socio-demographic Characteristics Approximately, 58% of study participants were males and 42% were females. The average age for participants was 30.9 years old (n = 397, SD = 12.21). The average age for 41 DIETARY PATTERNS OF ETHNIC POPULATIONS males participants was 30.1 (n = 228, SD = 11.54), and average age of females participants was 31.5 (n = 228, SD = 11.54). In terms of marital status, 41.7 % of participants were either married, in a common law relationship, or living with a partner, whereas 57.3% of participants were not in relationship with a partner (i.e., single). Further, 21.2% of participants had an education equivalent to high school diploma (i.e., grade 12 completion) or less, while 78.8% had post-secondary education. When participants were asked about their household family income for the year 2015, 54.2% of subjects had an annual household family income less than $40,000 Canadian dollars (CAD), 19.6% of participants were in the range of $40,000 CAD – 69,000 CAD, 13.0% of participants were in the range of $70,000 CAD - 99,000 CAD, and 13.3% of subjects had an annual household family income of $100,000 CAD or greater. The largest ethnic group participating in this study was South Asians (38.0%). Black, Chinese, Arab, Filipino, Korean, and West Asian participants accounted for 16.8%, 12.0%, 8.8%, 6.3%, 4.6%, and 4.4%, respectively. First generation immigrants who were born outside Canada represented 81.9% of all subjects, 17.2% were second generation immigrants, and only 1% were third generation immigrants. Average length of residence for first generation immigrants was 9.17 years (n = 316, SD = 10.45). A detailed description of socio-demographic characteristics of study sample is summarized in table 3. 42 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 3 Sociodemographic Characteristics of Study Sample variable Age in years (n = 397) Gender (n = 404) Gender*Age (n = 391) Marital Status (n = 406) Education level (n = 406) Annual household family income (n = 301) Subgroup 18-25 26-35 36-45 46-55 56-65 ≥ 66 Mean ± SD Males Females Male (18-25) Female (18-25) Male (26-35) Female (26-35) Male (36-45) Female (36-45) Male (46-55) Female (46-55) Male (56-65) Female (56-65) Males > 65 Female > 65 Male (Mean ± SD) Female (Mean ± SD) Married/Common law Living with a partner Never married Widowed Divorced Separated Uncompleted high school Frequency Percentage 176 44.3 131 33.0 39 9.8 24 6.0 18 4.5 9 2.3 30.85 ± 12.21 231 57.7 177 42.3 104 26.6 72 18.4 80 20.5 48 12.3 19 4.9 19 4.9 14 3.6 9 2.3 5 1.3 13 3.3 6 1.5 2 0.5 30.06 ± 11.54 31.51 ± 12.76 142 35.0 27 6.7 164 40.4 71 17.5 1 0.2 1 0.2 21 5.2 Completed high school Some university or college Completed undergraduate degree Post graduate education 65 122 113 16.0 30.1 27.8 85 20.9 Less than $10,000 $10,000-$19.999 $20,000-$29.999 $30,000-$39.999 58 32 34 39 19.3 10.6 11.3 13.0 43 DIETARY PATTERNS OF ETHNIC POPULATIONS Ethnic background (n = 410) Length of residence (n = 408) $40,000-$49.999 $50,000-$59.999 $60,000-$69.999 $70,000-$79.999 $80,000-$89.999 $90,000-$99.999 $100,000-$150,000 > $150,000 Chinese South Asian Black Filipino Latin American South East Asian Arab West Asian Japanese Korean Indigenous First generation (0-5 years) First generation (6-10 years) First generation (11-15 years) First generation (16-20 years) First generation (≥ 20 years) First generation (missing) First generation (Mean ± SD) Second generation Third generation 19 20 20 9 17 13 25 15 49 156 69 28 9 9 36 18 8 19 9 165 60 28 19 44 14 9.17 ± 10.45 - 6.3 6.6 6.6 3.0 5.6 4.3 8.3 5.0 12.0 38.0 16.8 6.8 2.2 2.2 8.8 4.4 2.0 4.6 2.2 40.4 14.7 6.9 4.7 10.8 4.4 17.2 1.0 4.2 Body Mass Index BMI (kg/m2) value was calculated for 367 subject who had self-reported weight and height, leading to BMI response rate equivalent to 88.8%. Participants were classified according to the Canadian Guidelines for Body Weight Classification in Adults into four categories (Health Canada, 2003). This classification identified 210 subjects in normal body weight category (M = 22.25, SD = 1.77), 21 subjects in underweight category (M = 17.16, SD = 1.51), 102 subjects in overweight category (M = 27.05, SD = 1.37), and 34 subjects in 44 DIETARY PATTERNS OF ETHNIC POPULATIONS obese category (M = 35.12, SD = 4.84). BMI characteristics across gender are further illustrated in table 4. Table 4 Mean, Standard Deviation & Categories of BMI and BMI by Gender Variable BMI (n = 367) BMI by Gender (n = 362) Class Normal weight Underweight Overweight Obese Total Frequency 210 21 102 34 367 Normal weight Underweight Overweight Obese Total 120 9 73 19 221 Normal weight Underweight Overweight Obese Total 88 12 28 13 141 Percentage 57.2 5.7 27.8 9.3 100.0 Male 54.3 4.1 33.0 8.6 100.0 Female 62.4 8.5 19.9 9.2 100.0 M 22.25 17.16 27.05 35.12 24.48 SD 1.77 1.51 1.37 4.84 4.78 24.7 4.32 23.99 5.29 Females had higher missing BMI values (20.3%), compared to males (4.3%). Chi square test for a 2x2 table showed that pattern of missing BMI values between male and females was significantly different (χ2 (1, N = 404) = 16.26, p < 0.001). Chi square test was also performed to examine the relationship between gender of subject and BMI category. The test revealed BMI categories were not equally distributed for males and females (χ2 (3, N = 362) = 9.30, p = .026). Males were more likely to be overweight, and females were more likely to be underweight. Exactly, 33.0 % of males were in the overweight BMI category, compared to 19.9% for females. Alternatively, 8.5% of females were in the underweight BMI category, compared to 4.1% for males. 45 DIETARY PATTERNS OF ETHNIC POPULATIONS 4.3 Multivariate Analyses Assumptions Data was examined for multivariate analysis assumptions including multivariate linearity and normality, factorability of R, absence of multi-collinearity and singularity, and absence of outliers. The distribution shape for variables and bivariate scatterplots was inspected to examine normality and linearity assumptions. As a result, multivariate normality and linearity were assumed. Mahalanobis test revealed the presence of 17 cases classified as potential multivariate outliers. This might be the result of mixing different groups in this sample (Tabachnick & Fidell, 2013). Influential multivariate outliers carry the risk of affecting correlations between variables and altering the fit of a model (Herman, Ryan, & Harry, 2013). To investigate the influence of outliers on multivariate analysis results (i.e., PCA, Multiple Regression). Two separate scenarios were performed to eliminate any risk related to lack of handling outliers. The first scenario included all cases with those identified as multivariate outliers (i.e., 410 cases), whereas the second scenario excluded multivariate outliers (i.e., 393 cases). The results indicated no difference between both scenarios in terms of significance and model fit. Based on these results, it was determined that outliers in this study were not influential. In this sense, keeping outliers was determined not to influence making proper inferences from study sample. There is a notion that deletion of non-significant number of outliers might make the sample a biased representation of the population (Herman et al., 2013). To examine factorability of R to justify performing PCA, Bartlett’s Test of Sphericity and the Kaiser-Meyer-Olkin Test of Sampling Adequacy (KMO) were performed. The result of Bartlett’s Test of Sphericity was significant (p < .001), indicating correlation matrix is statistically different from a singular matrix, which suggests factorability of variables 46 DIETARY PATTERNS OF ETHNIC POPULATIONS (Tabachnick & Fidell, 2013). Further, the value of Kaiser-Meyer-Olkin Measure of Sampling Adequacy was .84, which indicated shared variance between dietary items is meritorious for PCA (Beavers et al., 2013). Lack of multicollinearity and singularity assumption required for multiple regression analysis were assumed through examining tolerance and correlation between dependent variables. In terms of adequacy of sample size, there were 410 cases eligible to enter the final multivariate analysis. According to Tabachnick and Fidell (2013), sufficient sample for PCA could range between 100 to well over 500 depending on strength of factors/principal components (i.e., number of variables loading into a factor and their degree of loading). A well determined factor has large number of indicators with high loading. Tabachnick and Fidell (2013) conclude “At least 300 cases are needed with low communalities, small number of factors, and just three or four indicators for each factor” p. 618. Our PCA analysis revealed that average communality for all 63 dietary variables was 0.4, and each factor had at least four unique variables with a loading higher than 0.6 for a four-factor solution, and 0.4 for a six-factor solution. Beavers et al. (2013) indicates that a sample of at least 300 cases is required when factors are comprised of few variables with moderate to low loading. A larger sample size has an advantage of producing a more stable solution with factorial analysis (Hogarty, Hines, Kromrey, Ferron, & Mumford, 2005). According to Wilson VanVoorhis and Morgan (2007), 300 cases would be considered a good sample size. 4.4 Principal Components Analysis Principal components extraction was employed for all 63 dietary food items in the FQ using SPSS® for Windows statistical software package version 24.0. Initial extraction of components by PCA identified 17 components with eigenvalue more than one using Kaiser 47 DIETARY PATTERNS OF ETHNIC POPULATIONS Criterion for retention of components. Using the scree plot method identified that a bend in the elbow shaped slope occurring after the sixth component which suggests retaining six components in the final solution (Figure 3). For the purpose of interpretability of components and to avoid overestimation of components, subsequent PCA was done while retaining four, five and six components to evaluate the best solution in terms of interpretability and comprehensibility. Practically, this approach is accepted and commonly used practice (Beavers et al., 2013; Suhr, 2006). Figure 3. Scree plot produced by orthogonal Principal Components extraction method 48 DIETARY PATTERNS OF ETHNIC POPULATIONS The four-components solution accounted for 32.5% of variance within dietary items. Alternatively, the five-components solution accounted for an extra 3.5% of variance, compared to the four-components solution, leading to 36% of total variance being explained. six-components solution accounted for an extra 3.1% of variance, compared to the fivecomponents solution, leading to 39.1% of variance being explained. Rotation of components is an essential step in any type of factorial analysis to achieve more interpretable and meaningful solutions (Fabrigar, Wegener, MacCallum, Strahan, 1999). Orthogonal rotation of components assumes independence of rotated components; it is widely considered the most appropriate method of rotation when the purpose of the analysis is to produce component scores (Beavers et al., 2013). On the other side, non-orthogonal oblique rotation of components is most appropriate when there is a theoretical hypothesis pertinent to correlation between factors (Tabachnick & Fidell, 2013). If the relationship between components is unknown, Fabrigar et al. (1999) suggests performing oblique rotation first. If correlation between components were found to be low, opting out to orthogonal rotation would be most appropriate. To assess inter-correlation between components, initial approach included PCA with oblique rotation (i.e., oblimen rotation) while extracting six principal components; this approach identified close to zero correlation among all components as components, Subsequent PCA was done while retaining four, five and six components to shown in table 5. The largest correlation existed between component 1 and component 4 (r = -.255). Since components were largely uncorrelated and component scores were required, orthogonal rotation of components was deemed most appropriate for this analysis. Orthogonal independent rotation is required when populating component scores is needed (Tabachnick & Fidell, 2013). 49 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 5 Correlation among Six Dietary Components Variable COMP_1 COMP_2 COMP_3 COMP_4 COMP_5 COMP_6 — COMP_1 — COMP_2 0.042 — COMP_3 -0.204 -0.079 — COMP_4 -0.255 -0.224 0.185 — COMP_5 0.081 -0.016 0.057 0.025 — COMP_6 0.181 0.238 -0.039 -0.224 0.100 Note. COMP_1 = high calorie fatty food & alcohol component, COMP_2 = high fiber food “vegetables & fruits” component; COMP_3 = high protein food “meats and fish” component; COMP_4 = plant oils component; COMP_5 = low fat dairy, whole grain products & nuts component; COMP_6 = high fat dairy, refined grain products & condiments component. Varimax orthogonal rotation of components was used for all three PCA scenarios to find out the most interpretable and comprehensible solution. The results of beta component loading for all variables are illustrated in table 6, table 7, and table 8. Using a cut off equivalent to .32 for beta component loading, the six-components solution had 61 unique dietary items out of total 63 dietary items which loaded on one or more of the sixth components. Item q40 (i.e., 100% natural fruit juice) and item q17 (i.e., peas, beans, and lentils) were the only items that did not load into any of the six components. Each component was labelled based on dietary items loading on it. For example, component 3 was labelled high protein food “meats and fish”, as it had high loadings of red meat, poultry & turkey, fish, sea foods & shellfish, eggs, processed meats, organ meats such as liver and heart, deep fried fish or chicken, Veal, goat & rabbit meat, pasta & Spaghetti, and rice. Alternatively, component 1 was labelled high calorie fatty food & alcohol, component 2 was labelled high fiber food “vegetables & fruits”, component 4 was labelled plant oils, component 5 was labelled low fat dairy and whole grain products & nuts, and lastly component 6 was labelled high fat dairy, refined grain products & condiments. Results of 50 DIETARY PATTERNS OF ETHNIC POPULATIONS the beta component loadings for the six-factor component solution and percent of variance and covariance are illustrated in table 6. For the five-components solution, there were 60 unique dietary items out of total 63 dietary items which loaded on one or more of the sixth components. Item q40 (i.e., 100% natural fruit juice), item q32 (i.e., Milk 1%, 2% and skim milk), and item q14 (i.e., whole wheat pita, whole wheat bread, whole wheat bagel) did not load into any of the five components. Compared to the six-components solution, the simple structure was generally similar except for Low fat dairy and whole grain products & nuts component dropping out of the factorial structure leaving five components with relatively similar loadings in terms of item type and degree of loading. Results of beta component loadings for the five-components solution are illustrated in table 7. For the four-components solution, there were 56 unique dietary items out of total 63 dietary items which loaded on one or more of the sixth components. Item q42 (i.e., spicy condiment), item q34 (i.e., plain or sweetened yogurt), item q39 (peanut butter or sesame butter), Item q40 (i.e., 100% natural fruit juice), item q32 (i.e., Milk 1%, and skim milk), item q22b (i.e., egg white), and q35a (i.e., homogenized milk) did not load into any of the four components. Compared to the five-components solution, the simple structure was relatively similar except for high fat dairy, refined grain products & condiments component dropping out of the factorial structure leaving four components with relatively similar loadings in terms of item type and degree of loading. Another observation was related to complex dietary items loading into more than one factor compared to the other two solutions. This might be problematic when interpreting the results, as it is ideal when each dietary item loads into one component (Tabachnick & Fidell, 2013). Results of beta component loadings for the four-components solution are illustrated in table 8. 51 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 6 Six Dietary Components Extracted by PCA with Beta Component Loading and Percent of Variance and Covariance Using Varimax Orthogonal Rotation Ordinal dietary variable item Beta component loading Component 1 (High Calorie Fatty Food & Alcohol) French Fries (q11a) .65 Frozen potato product (q11b) .62 Alcohol (q55) .60 Lard/shortening (q52) .59 Gravy/cheese sauce (q44) .60 Soft drinks/fruit flavored drinks (q41a) .57 Diet soft drinks (q41b) .55 Pizza (q10b) .51 Ketchup (q10a) .53 Duck/goose (q21b) .48 Chips (q31) .46 Whipping cream/sour cream (q37b) .45 Low fat ice cream/yogurt (q33) .44 Ice cream (q35b) .40 Cottage and ricotta cheese (q36) .40 Homogenized milk (q35a) .38 Peanut butter (q29) .37 Cream (q37a) .33 Potatoes (q11c) .34 Sum of squared loadings = 6.28 Percent of variance = 9.97 Percent of covariance = 25.52 Component 2 (High Fiber Food “Vegetables & Fruits”) Other vegetables e.g., sweet peppers (q7) .72 Cabbage-type vegetables (q6) .71 Green leafy vegetables (q4) .67 Orange and dark yellow vegetables (q5) .61 Beans (q8) .58 Berries (q9d) .55 Exotic fruits (q9e) .53 Citrus fruits (q9b) .52 Non-citrus fruits (q9a) .48 Melon (q9c) .48 Soy (q39) .48 Bran/whole unrefined grains (q16) .42 Sum of squared loadings = 4.82 Percent of variance = 7.65 Percent of covariance = 19.60 Component 3 (High Protein Food “Meats and Fish”) Red meat (q26) .76 Poultry/turkey (q21) .68 52 DIETARY PATTERNS OF ETHNIC POPULATIONS Ground red meat (q23) .68 Dark fish (q19) .62 Sea foods/shellfish (q20) .60 Eggs (q22) .57 Processed meats (q25) .57 White fish (q18) .56 Organ meats (q24) .51 Fish & chicken (deep fried/breaded) (q18a) .46 Rice/pasta/spaghetti (q15) .46 Veal/goat/rabbit meat (q27) .42 Sum of squared loadings = 4.76 Percent of variance = 7.55 Percent of covariance = 19.33 Component 4 (Plant Oils) Olive oil with food (q50) .89 Corn oil and sunflower oil (q51) .89 Olive oil in salad dressing (q49) .67 Olive oil for frying or grilling (q48) .60 Sum of squared loadings = 3.10 Percent of variance = 4.91 Percent of covariance = 12.58 Component 5 (Low Fat Dairy, Whole Grain Products & Nuts) Yoghurt (q34) .59 Whole wheat pita/bread/bagel (q14) .44 Nuts (q30) .44 Milk 1%, 2% and skim milk (q32) .42 Low fat ice cream/yogurt (q33) .38 Spicy condiments (q42) .39 Garlic (q45) .38 Cereals (q12) .35 Cottage and Ricotta cheese (q36) .33 Sum of squared loadings = 2.88 Percent of variance = 4.57 Percent of covariance = 11.70 Component 6 (High Fat Dairy, Refined Grain Products & Condiments) Cheddar/mozzarella/feta cheese (q38) .62 Chocolate/desserts (q28) .53 White flour products (q13) .45 Tomato sauce/paste (q10) .40 Spicy condiments (q42) .40 Sweet condiments (q43) .39 Butter/margarine (q46) .39 Chips (q31) .37 Sum of squared loadings = 2.77 Percent of variance = 4.40 Percent of covariance = 11.26 Percent of variance (all Components) = 39.05 53 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 7 Five Dietary Components Extracted by PCA with Their Beta Component Loading and Percent of Variance Using Varimax Orthogonal Rotation Ordinal dietary variable item Beta component loading Component 1 (High Calorie Fatty Food & Alcohol) French Fries (q11a) .64 Frozen potato product (q11b) .62 Alcohol (q55) .62 Lard/shortening (q52) .58 Gravy/cheese sauce (q44) .62 Soft drinks/fruit flavored drinks (q41a) .56 Diet soft drinks (q41b) .56 Pizza (q10b) .49 Ketchup (q10a) .52 Duck/goose (q21b) .50 Chips (q31) .47 Whipping cream/sour cream (q37b) .46 Low fat ice cream and yogurt (q33) .47 Regular ice cream (q35b) .44 Cottage and ricotta cheese (q36) .42 Homogenized milk (q35a) .41 Organ meats (q24) .39 Peanut butter (q29) .37 Fish & chicken - deep fried/breaded (q18a) .37 Veal/goat/rabbit meat (q27) .36 Cream (q37a) .35 Potatoes (q11c) .35 Egg white (q22b) .33 Sum of squared loadings = 6.48 Percent of variance = 10.29 Percent of variance = 28.55 Component 2 (High Fiber Food “Vegetables & Fruits”) Other vegetables e.g., sweet peppers (q7) .69 Cabbage-type vegetables (q6) .69 Green leafy vegetables (q4) .65 Orange and dark yellow vegetables (q5) .59 Beans (q8) .57 Berries (q9d) .55 Exotic fruits (q9e) .57 Citrus fruits (q9b) .52 Non-citrus fruits (q9a) .52 Melon (q9c) .47 Soy (q39) .44 Bran/whole unrefined grains (q16) .47 54 DIETARY PATTERNS OF ETHNIC POPULATIONS Peas/beans/lentils (q17) .35 Garlic (q45) .34 Sum of squared loadings = 5.19 Percent of variance = 8.24 Percent of covariance = 22.88 Component 3 (High Protein Foods “Meats and Fish”) Red meat (q26) .77 Poultry/turkey (q21) .69 Ground red meat (q23) .69 Dark fish (q19) .59 Sea foods/shellfish (q20) .60 Eggs (q22) .58 Processed meats (q25) .63 White fish (q18) .50 Organ meats (q24) .46 Fish & chicken - deep fried/breaded (q18a) .43 Rice/Pasta/Spaghetti (q15) .47 Veal/goat/rabbit meat (q27) .35 Sum of squared loadings = 4.71 Percent of variance = 7.48 Percent of covariance = 20.76 Component 4 (Oils) foods eaten with olive oil (q50) .83 Other oils e.g., corn oil, sunflower oil (q51) .83 Salad dressing with olive oil (q49) .63 Fry/grill with olive oil (q48) .58 Nuts (q30) .35 Yoghurt (q34) .35 Sum of squared loadings = 3.40 Percent of variance = 5.39 Percent of covariance = 14.97 Component 5 (High Fat Dairy, Refined Grain Products & Condiments) Cheddar/mozzarella/feta cheese (q38) .57 Chocolate/desserts (q28) .49 White flour products (q13) .45 Spicy condiments (q42) .50 Tomato sauce/paste (q10c) .43 Sweet condiments (q43) .41 Butter/margarine (q46) .48 Chips (q31) .35 Sum of squared loadings = 2.91 Percent of variance = 4.63 Percent of covariance = 12.84 Percent of variance (all Component) = 36.03 55 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 8 Four Dietary Components Extracted by PCA with Their Beta Component Loading and Percent of Variance Using Varimax Orthogonal Rotation Ordinal dietary variable item Beta component Loading Component 1 (High Calorie Fatty Food & Alcohol) French Fries (q11a) .59 Frozen potato product (q11b) .58 Alcohol (q55) .38 Lard/shortening (q52) .37 Gravy/cheese sauce (q44) .60 Soft drinks/fruit flavored drinks (q41a) .53 Diet soft drinks (q41b) .38 Pizza (q10b) .51 Ketchup (q10a) .53 Chips (q31) .58 Whipping cream/sour cream (q37b) .52 Low fat ice cream/yogurt (q33) .34 Ice cream (q35b) .48 Cottage and ricotta cheese (q36) .49 Homogenized milk (q35a) .34 Peanut butter (q29) .30 Cream (q37a) .35 Potatoes (q11c) .39 Cheddar/mozzarella/feta cheese (q38) .49 Butter/margarine .49 Chocolate/desserts (q28) .45 Sweet condiments (q43) .43 White flour products (q13) .45 Sum of squared loadings = 5.81 Percent of variance = 9.22 Percent of covariance = 28.40 Component 2 (High Fiber Food “Vegetables & Fruits”) Other vegetables e.g., sweet peppers (q7) .69 Cabbage-type vegetables (q6) .67 Green leafy vegetables (q4) .64 Orange and dark yellow vegetables (q5) .60 Beans (q8) .57 Berries (q9d) .55 Exotic fruits (q9e) .53 Citrus fruits (q9b) .52 Non-citrus fruits (q9a) .51 Melon (q9c) .41 Soy (q39) .43 56 DIETARY PATTERNS OF ETHNIC POPULATIONS Bran/whole unrefined grains (q16) .46 Peas/lentils (q17) .39 Nuts (q30) .38 Garlic (q45) .38 Whole wheat pita/bread/bagel (q14) .34 Sum of squared loadings = 5.17 Percent of variance = 8.21 Percent of covariance = 25.30 Component 3 (High Protein Food “Meats and Fish”) Red meat (q26) .75 Poultry/turkey (q21) .62 Ground red meat (q23) .68 Dark fish (q19) .59 Sea foods/shellfish (q20) .63 Eggs (q22a) .52 Processed meats (q25) .65 White fish (q18a) .55 Organ meats (q24) .52 Fish & chicken - deep fried/breaded (q18a) .47 Rice/Pasta/Spaghetti (q15) .44 Veal/goat/rabbit meat (q27) .41 Sum of squared loadings = 5.01 Percent of variance = 8.05 Percent of covariance = 24.78 Component 4 (Plant Oils and Fatty Food) foods eaten with olive oil (q50) .79 Other oils e.g., corn oil, sunflower oil (q51) .79 Salad dressing with olive oil (q49) .61 Fry/grill with olive oil (q48) .52 Cereals (q12) .46 Low fat ice cream/yogurt (q33) .39 Duck/goose (q21b) .38 Alcohol (q55) .38 Lard/shortening (q52) .37 Diet soft drinks (q41b) .36 Sum of squared loadings = 4.40 Percent of variance = 6.99 Percent of covariance = 21.52 Percent of variance (all Components) = 32.47 57 DIETARY PATTERNS OF ETHNIC POPULATIONS Based on total amount of variance explained by the solution and beta component loading for dietary items, the six-components solution was the most fitting with better interpretability than other two solutions. Thus, it was considered as the final solution for this analysis. In the final solution, variables are well defined with large amount of their variance being explained by the final solution, and only two dietary variables had less than 20 % of their variance accounted for by the solution (Appendix C). Further, 13 dietary variables had 20-30% of their variance accounted for by the solution, 19 dietary variables had 30-40% of their variance explained, 19 dietary variables had 40-50% of their variance explained, 7 dietary variables had 50-60% of their variance explained, and three dietary variables had more than 60% of their variance explained. Altogether, 39.1% of the variance in the set of dietary variables was accounted for by the solution. Component 1 accounts for 10.0% of the variance in the set of dietary variables, and 25.5% of the variance in the solution. Component 2 accounts for 7.7% of the variance in the set of dietary variables, and 19.6% of the variance in the solution. Component 3 accounts for another 7.6% of the variance in the set of dietary variables, and 19.3% of the variance in the solution. Component 4 accounts for 4.9% of the variance in the set of dietary variables, and 12.6% of the variance in the solution. Component 5 accounts for 4.6% of the variance in the set of dietary variables, and 11.7% of the variance in the solution. Component 6 accounts for 4.4% of the variance in the set of dietary variables, and 11.3% of the variance in the solution. Together, all six components account for 39.1% of the variance in the set of dietary variables. 58 DIETARY PATTERNS OF ETHNIC POPULATIONS Figure 4. Western diet patterns identified from the food frequency questionnaire (n = 410) 59 DIETARY PATTERNS OF ETHNIC POPULATIONS Figure 5. Prudent diet patterns identified from the food frequency questionnaire (n = 410) 60 DIETARY PATTERNS OF ETHNIC POPULATIONS Component scores were produced for all subjects using SPSS program though Bartlett’s regression method. Component scores are considered variables that represent how much would a subject score on each principal component (Yong & Pearce, 2013). In this sense, each subject had six meaningful and representative component scores based on their 63 dietary responses and their weighted value on each component (e.g., S1 component Score 1, S1 component score 2, S1 component score 3, etc.). These factor scores can be subsequently used as variables in further statistical analyses such as multiple regression (Tabachnick & Fidell, 2013). 4.5 Multiple Regression Multiple regression analysis was performed via SPSS® for Windows statistical software package version 24.0. In this analysis, BMI was considered as a “criterion/outcome variable”, six dietary components scores variables were considered as “predictor variables”, and five sociodemographic variables were considered as “covariates”. Those sociodemographic variables included age, gender, marital status, education level, and income. There were 135 cases removed from the multiple regression analysis due to missing data for one or more of the abovementioned sociodemographic variables. Thus, 275 cases were entered into the multiple regression analysis. Evaluation of assumptions for multiple regression analysis revealed that number of cases to independent variables was large enough (Tabachnick & Fidell, 2013). Assuming medium size relationships, at least 110 cases are required for regression analysis with six predictor variables to test both the overall model and individual predictors (Wilson VanVoorhis & Morgan, 2007). Thus, sample size was deemed appropriate to test medium and large size effects in this analysis. Further, normal probability plot and scatterplot for residuals produced through SPSS output indicated that normality and 61 DIETARY PATTERNS OF ETHNIC POPULATIONS linearity assumption were to be assumed. Examining tolerance showed that independent variables were largely uncorrelated with lack of multicollinearity or singularity. Other multivariate assumptions including outliers were checked through examining Mahalanobis distance and Cook’s distance for independent and dependent variables as previously discussed. Table 9 Zero Order Correlations among Outcome and Covariate Variables Variable Raw_BMI AGE GENDER MARITAL EDUC INCOME Raw_BMI — AGE .076 — GENDER .009 .078 — MARITAL -.119* -.556*** -.147** — EDUC -.050 -.046 -.062 -.042 — INCOME -.020 .255*** -.058 -.201*** -.027 — *p < .05, **p < .01, ***p < .001 Note. Raw_BMI = body mass index (kg/m2); AGE = age in years at time of survey intake; GENDER = gender of participant; MARITAL = marital status; EDUC = highest grade/level of education completed; INCOME = 2015 annual household family income. regression analysis was employed through SPSS program to determine if individual dietary components (COMP_1, COMP_2, COMP_3, COMP_4, COMP_5, COMP_6) contribute a significant unique proportion of variance to body mass index (Raw_BMI) while controlling for sociodemographic variables including age (AGE), gender (GENDER), marital status (MARITAL), level of education (EDUC), and 2015 annual household family income (INCOME). Correlations among criterion/outcome variable and covariate variables are displayed in table 9. Sociodemographic variables were entered in the first block of the analysis, and the six dietary component variables were entered in the second block of the analysis. Correlations among criterion/outcome variable and predictor variables are displayed in table 10. 62 DIETARY PATTERNS OF ETHNIC POPULATIONS Step one of the analysis revealed that sociodemographic variables (i.e., control variables) were not significant predictors for body mass index with R value of .141 and R2 value of .020 (F(5, 269) =, MSE = 25.10, p = 0.362). Steps two of the analysis revealed that adding the dietary components variables to the model resulted in R2 value of .048 and a non-significant R2 change of .028 (F(6, 263) = 1.28, p = .266). Adjusted R2 value for both step one and step two was .002 and .008, respectively. The gap between R2 values and adjusted R2 value reflects a large theoretical correction that could be explained by the small effect size and the large number of predictor variables (Lewis, 2007). Table 10 Zero Order Correlations among Outcome and Predictor Variables Variable Raw_BMI COMP _1 — Raw_BMI — COMP_1 .036 COMP_2 -.064 .045 COMP_3 .120* -.029 COMP_4 -.015 .028 COMP_5 -.008 -.011 COMP_6 .024 .050 *p < .05, **p < .01, ***p < .001 COMP_ 2 COMP_ 3 COMP_ 4 COMP_ 5 COMP_ 6 — -.070 -.049 -.066 -.066 — -.029 .007 -.018 — -.026 -.010 — 0.036 — Note. COMP_1 = high calorie fatty food & alcohol component, COMP_2 = high fiber food – vegetables & fruits component; COMP_3 = high protein food – meats and fish component; COMP_4 = plant oils component; COMP_5 = low fat dairy, whole grain products & nuts component; COMP_6 = high fat dairy, refined grain products & condiments component. The overall model test was not significant (F(11, 263) = 1.20, MSE = 24.94, p = .286). Correlation among covariate and predictor variables is displayed in table 11. Significance test for individual predictors showed that one of the dietary component (i.e., component 3) contributed a significant unique proportion of variance to body mass index. With all BMI variance explained by sociodemographic variables being removed, component 63 DIETARY PATTERNS OF ETHNIC POPULATIONS 3 accounted for a significant unique amount of variance to BMI (1.7%), which is not shared with other components. None of the other components accounted for any significant unique amount of variance to BMI. Table 11 Zero Order Correlations among Covariate and Predictor Variables Variable COMP_1 COMP_2 COMP_3 COMP_4 COMP_5 COMP_6 AGE -.219*** -.023 -.105* .048 .144** -.168** GENDER -.213*** .150** -.090 -.073 -.013 .088 MARITAL .220*** .014 -.049 -.057 -.228*** .039 EDUC -.115* -.158** .170** .128* .114* .045 INCOME -.054 .082 .024 -.010 .000 -.003 *p < .05, **p < .01, ***p < .001 Note. AGE = age in years at time of survey intake; GENDER = gender of participant; MARITAL = marital status; EDUC = highest grade/level of education completed; INCOME = 2015 annual household family income, COMP_1 = high calorie fatty food & alcohol component, COMP_2 = high fiber food – vegetables & fruits component; COMP_3 = high protein food – meats and fish component; COMP_4 = plant oils component; COMP_5 = low fat dairy, whole grain products & nuts component; COMP_6 = high fat dairy, refined grain products & condiments component. Table 12 displays the unstandardized regression coefficient, standardized regression coefficient, p value, semipartial correlation, and unique variance contributions for each predictor variable. 64 DIETARY PATTERNS OF ETHNIC POPULATIONS Table 12 Coefficients, p Value, Semipartial Correlations, and Unique Variance Contribution for Each Covariate and Predictor Variable in the Regression Analysis Covariate/Predictor B p sr sr2  Variables 1 AGE 0.008 .019 .793 .016 .000 GENDER -0.159 -.017 .783 -.017 .000 MARITAL -0.278 -.124 .094 -.101 .010 EDUC -0.126 -.057 .352 -.056 .003 INCOME -0.075 -.053 .403 -.051 .003 2 COMP_1 0.351 .071 .268 .067 .004 COMP_2 -0.330 -.064 .303 -.062 .004 COMP_3 0.665 .134 .033 .129 .017 COMP_4 -0.087 -.018 .774 -.017 .000 COMP_5 -0.217 -.042 .505 -.040 .002 COMP_6 0.196 .037 .549 .036 .001 Note. AGE = age in years at time of survey intake; GENDER = gender of participant; MARITAL = marital status; EDUC = highest grade/level of education completed; INCOME = 2015 annual household family income, COMP_1 = high calorie fatty food & alcohol component, COMP_2 = high fiber food – vegetables & fruits component; COMP_3 = high protein food – meats and fish component; COMP_4 = plant oils component; COMP_5 = low fat dairy, whole grain products & nuts component; COMP_6 = high fat dairy, refined grain products & condiments component Steps 65 DIETARY PATTERNS OF ETHNIC POPULATIONS Chapter 5: Discussion & Conclusion This community-based survey study provides new information on dietary pattern and eating habits of multiethnic population in northern British Columbia in relation to their risk of obesity, and subsequently chronic disease conditions. There was a relatively high response rate (59.6%) compared to average response rate of 20-25% to mail out surveys (Sax, Gilmartin, & Bryant, 2003). Generally, a downward trend in survey response rate has been evident over the last few decades. Survey response rate is heavily influenced by several factors such as mode of administration (e.g., paper-based versus web-based), length of survey, presence of financial incentive, protection of privacy and confidentiality, and sensitivity of survey research topic (Fincham, 2008; Korkeila et al., 2001; Sahlqvist et al., 2011). The high response rate to survey intake was paralleled with low item non-response rate for diet intake variables as well (2.2%). Such a high response rate is considered a very important indicator for well representativeness of study sample, and non-selection bias, especially for small populations (Fincham, 2008). Other research has indicated that women tend to participate more actively than men in survey research studies (Korkeila et al., 2001; Sax, Gilmartin, & Bryant, 2003); In the current study however, men had a slightly larger representation compared to women (57.7% versus 44.3% respectively). Since the identity of non-respondents is unknown, it is not plausible to assume men had a higher response rate than women, given the non-random convenience and snowball sampling approach implemented in the current study. Furthermore, the study sample had higher representation from younger age groups. The results also revealed that composition of various ethnic groups in this research was not well representative of the relative ethnic population size in British Columbia (Statistics Canada, 2012). This is largely 66 DIETARY PATTERNS OF ETHNIC POPULATIONS due to using convenience sampling rather than random sampling method. Add to that, the distribution of ethnic population might vary from northern BC. 5.1 Dietary Patterns for Study Sample Diet patterns could be defined from FFQ collected data through two main statistical methods, namely factor analysis and cluster analysis (Hu, 2002). Cluster analysis classifies subjects according to their pattern of food intake into homogenous clusters (Désilets et al., 2007; Lenighan et al., 2017), whereas factor analysis classifies food groups and dietary items into dimensions based on correlations of different food items with one another (Newby et al., 2004). Factor analysis includes both common factor analysis and PCA (Tabachnick & Fidell, 2013). Most previous research conducted to define dietary patterns favored PCA over common factor analysis (Brenner et al., 2011; Panagiotakos, Pitsavos, Skoumas, & Stefanadis, 2007; Rezazadeh et al., 2016). This presented an important consideration when selecting an analytical method to define dietary patterns in this study, particularly when comparing results is a matter of interest. The results of the PCA in my study reduced 63 food items or food groupings into six main components or dietary patterns. The different dietary components were defined by 61 food items indicating a good fit for this analysis. Each component was labeled based on the defining food items leading to the following components in order of their strength high calorie fatty food & alcohol component, high fiber food “vegetables & fruits” component, high protein food “meats & fish” component, plant oils component, low fat dairy and whole grain products & nuts component, and high fat dairy, refined grain products & condiments component, respectively. 67 DIETARY PATTERNS OF ETHNIC POPULATIONS The high calorie fatty food & alcohol component was the most prominent diet pattern defined by mostly unhealthy energy dense and fatty food items that are consumed in western diet such as French fries, soft drinks, high fat cheese, gravy, whipping cream, ice cream, chips, pizza, high fat cheese, peanut butter, and alcohol. Duck and goose meat, which are mainly consumed by some ethnic sub-populations such as Chinese, were also part of this dietary component. Interestingly, duck and goose are often consumed as part of unhealthy fatty meals. This component was the largest and consequently most predictive explaining 10% of the between-subject variation in dietary intake. This high calorie fatty food and alcohol diet pattern was found to be associated with participants who have younger age (p < .001), male gender (p < .001), unmarried (p < .001), and lower level of education (p < .05). The high fiber food “vegetables & fruits” component was defined by food items rich in fiber including vegetables and fruits (e.g., zucchini, cabbage, lettuce, Orange, green beans, green peas, berries, pineapple, orange, apple, watermelon), and unrefined grains (e.g., brown rice, whole wheat pasta, and bran). Those food items are highly consumed in many ethnic diets including Arab, South Asian, and Chinese. All vegetable and fruit food items were defined by this component except for the tomato food item, which was defined by the high fat dairy, refined grain products & condiments. This might be attributed to lumping tomato with processed tomato products such as tomato sauce and tomato paste in the FFQ (q10c). This component is the second largest and consequently explains 7.7% of the between-subject variation in dietary intake. This high fiber diet pattern was found to be associated with participants with female gender (p < .01), lower level of education (p < .01), and higher level of income (p = .08). 68 DIETARY PATTERNS OF ETHNIC POPULATIONS The high protein food “meats & fish” component was defined by food items rich in protein that are largely part of western diet pattern including red meat, processed meats, chicken, poultry, turkey, fish, veal, deep fried and breaded chicken or fish, sea foods, and eggs. Further, food items that are commonly consumed with high protein food clustered on the same component such as white rice, pasta, and spaghetti. This component was the third largest component explaining 7.6% of the between-subject variation in dietary intake. It was almost a similar amount to that explained by the high fiber food “vegetables & fruits” component. This high protein diet pattern was found to be associated with participants who have younger age (p < .05), and male gender (p = .06), and higher level of education (p < .01). Lenighan et al. (2017) found total red meat intake in males is approximately 1.5 times that for females. The plant oils component was purely defined by different types of plant-based sources of fat including olive oil, corn oil, sunflower oil, and canola oil. It explained 4.9% of the between-subject variation in dietary intake. This plant oils component was found to be associated with participants who have higher level of education (p < .05). The low fat dairy and whole grain products & nuts component was defined by healthier dietary items including low fat dairy product (e.g., low fat milk, low fat ice cream & yogurt, cottage and ricotta cheese), whole wheat products (e.g., whole wheat pita, whole wheat bread), nuts, and garlic, and cereals. It explained 4.6% of the between-subject variation in dietary intake. This low fat whole grain diet pattern was found to be associated with participants who have older age (p < .01), married (p < .001), with higher level of education (p < .05). Finally, the high fat dairy, refined grain products & condiments component was defined by unhealthy dietary items that are largely part of western diet pattern including high fat dairy products (e.g., 69 DIETARY PATTERNS OF ETHNIC POPULATIONS cheddar & mozzarella cheese, and butter), refined grain products (e.g., white bread, white pita, muffins, pancakes, and waffles), chocolates & desserts, and margarine. Further, other food item clustered on this component included condiments and tomato products (e.g., tomato, tomato sauce, and tomato paste). It explained 4.4% of the between-subject variation in dietary intake. This high fat dairy refined grains diet pattern was found to be associated with participants who have younger age (p < .01). Almost all dietary items in the FFQ clustered on the identified dietary components. Natural fruit juices, and peas & lentils were the only exception as none of them clustered with other food items in any component. Additionally, very few items clustered on two dietary components such as ricotta and cottage cheese loading on both the high calorie fatty food & alcohol component and low fat dairy and whole grain products & nuts component. Conversely, most dietary items clustered uniquely on one dietary component except very few dietary items that clustered on two factors. This indicates a more reliable solution with easier interpretability for the dietary components (Tabachnick & Fidell, 2013). Together, the six main dietary patterns in this study explained a significant amount of the variability in the dietary variables accounting for 39.1% of the between-subject variation in dietary intake. Previous studies conducted to identify dietary patterns through the PCA have produced different dietary patterns that ranged between three to six main components. The difference in the number of identified diet patterns is attributed to using different food grouping classifications and different number of dietary items in the FFQ, in addition to the researcher’s preference to extract a specific number of components or dietary patterns based on interpretability. Brenner et al. (2011) identified three distinct diet patterns amongst a sample of young adults living in Toronto with diverse ethnic backgrounds, which 70 DIETARY PATTERNS OF ETHNIC POPULATIONS cumulatively accounted for only 16% of the between person variance in dietary intake. Another study by Newby et al. (2004) identified six distinct diet food patterns amongst a sample of Caucasian and ethnic men and women living in the Baltimore area, which cumulatively accounted for 27% of the variance in dietary intake. The ATTICA study identified six diet patterns amongst a random sample of 3042 participants from Greece, which accounted for 56% of the total variation in the dietary intake. More importantly, this study identified independent association between diet patterns and components of the metabolic syndrome. The Peel and Nutrition Heart Health Survey study also identified six dietary patterns amongst 759 subjects from the Peel region in Ontario, Canada (Shubair, McColl, & Hanning, 2005). The study concluded the presence of an association between Mediterranean diet pattern and reduced prevalence of overweight and obesity. Recent research produced the concept of healthy “prudent” diet, which is the antithesis of the western atherogenic diet (Hu 2002; Hu et al., 2000; Hu & Willet, 2002). Prudent diet is characterized by higher consumption of fruits, vegetables, legumes, nuts, whole grains, low fat dairy products, fish, and soy products. On the other side, western diet is characterized by higher consumption of energy dense foods rich in calories and saturated fat & trans-fat with higher glycemic index which includes increased intake of red & processed meats, high fat dairy products, French fries & potatoes, sweets & desserts, refined grains, and alcohol (Cordain et al., 2005). Prudent diet, a modification of Mediterranean diet, is believed to be applicable to populations from different ethnic and cultural backgrounds (Enas, Senthilkumar, Chennikkara, & Bjurlin, 2003). According to the study results, three dietary components were identified as western diet patterns including the high calorie fatty food & alcohol component, the high protein food “meats and fish” component, and the high 71 DIETARY PATTERNS OF ETHNIC POPULATIONS fat dairy, refined grain products & condiments component (Figure 4). The western diet patterns consisted of mainly unhealthy food items and were significantly associated with younger age men. Conversely, the other three dietary components were identified as prudent diet patterns including the high fiber food “vegetables & fruits” component, the plant oils component, and the low fat dairy and whole grain products & nuts component (Figure 5). The prudent diet patterns consisted of mainly healthy food items and were associated significantly with higher level of education and mostly women. 5.2 Healthy Body Weight Status and Association with Dietary Patterns Men had higher BMI than woman with relatively larger proportion of men being overweight (i.e., pre-obese), compared to women. This finding was echoed in a study by Brenner et al. (2011), which explored dietary patterns for ethno-culturally diverse population of young adults in Toronto. Overall, approximately 37% of participants were considered either obese or pre-obese. More importantly, the three prudent-like diet patterns identified in this study were inversely correlated with BMI, where the high fiber food “vegetables & fruits” diet pattern had the strongest negative association. On the other hand, the other three western-like diet patterns were positively correlated with BMI, where the high protein food “meats & fish” diet pattern had the strongest positive association. To test the relationship between diet patterns identified in our study and body mass index while taking into consideration sociodemographic variables, the results of a multiadjusted multiple regression analysis revealed that a diet pattern that includes meats, pasta and rice was a significant predictor of body mass index. Participants with higher consumption of red meats, processed meats, sea foods, breaded and fried chick and fish, eggs, and rice & pasta tend to have higher body mass index even after controlling for 72 DIETARY PATTERNS OF ETHNIC POPULATIONS sociodemographic variables. This is a significant finding in our study that indicates acculturation to western diet is associated with higher risk of morbid obesity and subsequently other chronic disease conditions such as hypertension, type 2 diabetes mellitus, and cardiovascular disease. A working group related to the International Agency for Research on Cancer (IARC) classified the consumption of processed red meat in humans as “probably carcinogenic” based on epidemiological evidence from several cohort studies that concluded positive associations with colorectal and other types of cancer (Bouvard et al., 2015). The other two western diet patterns were not significant predictors of body mass index. The results also indicated a significant correction for the adjusted squared multiple correlation (SMC) between the dietary component variables and body mass index while adjusting for sociodemographic variables. This could potentially explain the inability to detect some significant association with the statistical power of our study, which was geared towards medium and large size effects based on sample size. The findings were consistent with the results from Newby et al. (2004), who hypothesized that a healthy food pattern would predict small modest changes in the BMI and waist circumference than would a nonhealthy diet pattern. This is attributed to BMI being influenced by total energy intake, which minimally changes with healthier diet patterns. This theoretical hypothesis also justifies the non-significance of the three prudent diet patterns as predictors for BMI in the present study. 5.3 Study Limitations Dietary patterns were not examined separately for each gender or ethnic group due to the inadequate number of subjects in each of these groups, and an inability to make sufficiently powered statistical comparisons. Also, there was neither an equal representation 73 DIETARY PATTERNS OF ETHNIC POPULATIONS of men and women (58% males versus 42% females), nor a true ethno-cultural representation of different ethnic minorities in BC at large. According to the 2006 Canadian census, Chinese were the largest ethnic minority in BC representing 40% of the total non-Caucasian ethnic population (Statistics Canada, 2010). South Asians, Filipino, Korean, South East Asians, and Japanese were the other largest ethnic minorities in BC representing 26%, 9%, 5%, 4%, and 3%, respectively. On the other hand, the sample population in this study had 12% Chinese, 38% South Asians, 17% Black, and 9% Arabs, 7% Filipino, and 5% Koreans. The misrepresentation and lack of generalizability in the sample population might have influenced the study results. Another limitation to this study is the possibility of response bias and socially desirable responses, which is anticipated when responding to measures of healthfulness of diet and body weight status. Using PCA as an analytical technique has its own limitations regarding assuming independence between dietary dimensions (components), and the subjectivity of choosing the number of factors by the researcher. Similar to all observational research designs, causal relationship cannot be inferred from this study. 5.4 Future Research As concluded in this study, men and women have differences in their dietary habits, and also their risk of obesity. Diet patterns are likely to differ by gender as indicated by previous research by Brenner et al. (2011) and Newby et al. (2004). Further research should examine dietary patterns separately for men and women to test whether such findings are due to biologic interaction between diet and gender rather than possible methodological limitations. Also, further research is needed to investigate the relationship between dietary patterns in ethnic populations and risk of the metabolic syndrome and CVD. There is a need 74 DIETARY PATTERNS OF ETHNIC POPULATIONS to examine dietary patterns separately for each ethnic subpopulation. This might yield unique results that are possibly masked by lumping different heterogeneous subgroups together. Exploring dietary pattern specific to ethnicity and culture shall help guide a more sensitive and effective dietary interventions. In light of some findings in this study regarding socioeconomic influences on diet patterns and dietary intake, future research examining food security in northern Canadian rural communities is warranted. 5.5 Conclusion To the best of my knowledge, this is the first study to examine dietary patterns among different non-Indigenous ethnic population in a Canadian northern rural community or elsewhere in British Columbia. This study investigated the association between the dietary patterns and healthy body weight status. Most previous studies on dietary patterns of ethnic populations were conducted with residents of large urban Canadian cities including Toronto, and Montreal (Brenner et al., 2011; Désilets et al., 2007). Recent research suggests that studying diet pattern rather than individual nutrients is considered a more robust and holistic approach. This is based on research findings indicating that diet acts via different metabolic pathways through synergistic mechanisms and complex interactions where beneficial healthy food effects are augmented when certain types of food are consumed in conjunction (Jacobs & Steffen, 2003). From a nutritional epidemiology point of view, studying diseases in relation to “dietary pattern” rather than “single nutrient” is a more superior approach with less conceptual and methodological limitations (Hu, 2002). Similarly, introducing dietary modifications and interventions altering dietary pattern is far more effective than single nutrient supplementation (Enas et al., 2003). There is also a multitude of prospective research 75 DIETARY PATTERNS OF ETHNIC POPULATIONS studies that suggests examining the relationship between diet patterns and risk for different chronic diseases (Kenfield et al., 2014; Mattei et al., 2017; VanKim et al., 2017). The current study found that six dietary components might characterize the dietary intake of participants from multi-ethnic background in northern British Columbia. Further examination identified some patterns related to western diet, which typically occurs as a result of dietary transitional process following immigration to western countries referred to as “dietary acculturation”. Such dietary patterns consistent with western diet are found to place ethnic individuals at higher risk of obesity. More significantly, younger ethnic men regardless of their level of education were found to be at a disadvantage of higher adherence to western diet patterns. Men in general were also found to be at higher risk of obesity. It was found that the diet pattern structure for ethnic minorities include some prudent-like diet patterns defined by healthy dietary intake, which is typically part of many ethnic diets such as Chinese, South Asian, and Mediterranean diets. There has been emerging research around promoting prudent diet in ethnic minorities to reduce the risk of obesity, type 2 diabetes, cardiovascular disease, and the metabolic syndrome. Enas and colleagues assert that promoting prudent diet rich in vegetables, fruits, nuts, legumes and olive oil with moderate intake of fish and poultry and low intake of red meats, eggs, dairy products and wine is feasible through applying adaptations to different ethnic diets and related food preparation methods. Such prudent diet is more beneficial than low fat and low carbohydrate diet regimens commonly adopted for body weight management (Enas et al., 2003). Bedard et al. (2010) showed an evidence that promoting a Mediterranean diet intervention in healthy French-Canadian women led to significant reduction in both unhealthy fast food consumption and body weight. Dietary interventions in the form of promoting healthy eating 76 DIETARY PATTERNS OF ETHNIC POPULATIONS pattern consistent with prudent diet is a cornerstone in addressing the ever-rising epidemics of obesity, diabetes and CVD. This has a greater importance for populations getting acculturated to unhealthy eating habits including immigrant and ethnic populations. Promoting adherence to ethnic food culture and habits in these populations definitely has scientific merit. Such recommendations for tailoring ethnically appropriate dietary interventions promise to promote healthier life style modification including healthy dietary intake in vulnerable ethnic populations. 77 DIETARY PATTERNS OF ETHNIC POPULATIONS References Abdul-Quader, A. S., Heckathorn, D. D., McKnight, C., Bramson, H., Nemeth, C., Sabin, K.,… Des Jarlais, D. C. (2006). Effectiveness of respondent-driven sampling for recruiting drug users in New York City: findings from a pilot study. Journal of Urban Health, 83(3): 459–476. doi: 10.1007/s11524-006-9052-7 Abiemo, E. E., Alonso, A., Nettleton, J. A., Steffen, L. M., Bertoni, A. G., Jain, A., & Lutsey, P. L. (2013). Relationships of the Mediterranean dietary pattern with insulin resistance and diabetes incidence in the Multi-Ethnic Study of Atherosclerosis (MESA). The British Journal of Nutrition, 109(8): 1490-7. doi: 10.1017/S0007114512003339 Altomare, R., Cacciabaudo, F., Damiano, G., Palumbo, V. D.,Gioviale , M. C., Bellavia, M., Tomasello, G., & Lo Monte, A. L. (2013). The Mediterranean diet: A history of health. Iranian Journal of Public Health, 42(5): 449-457. Babio, N., Bullo, M., & Salas-Salvado, J. (2009). Mediterranean diet and metabolic syndrome: the evidence. Public Health Nutrition 12(9A), 1607–1617. doi:10.1017/S1368980009990449 Babio, N., Toledo, E., Estruch, R., Ros, E., Martínez-González, M. A., Castañer, O., … PREDIMED Study Investigators. (2014). Mediterranean diets and metabolic syndrome status in the PREDIMED randomized trial. Canadian Medical Association journal, 186(17): E649-57. doi: 10.1503/cmaj.140764 Batis, C., Sotres-Alvarez, D., Gordon-Larsen, P., Mendez, M. A., Adair, L., & Popkin, B. (2014). Longitudinal analysis of dietary patterns in Chinese adults from 1991 to 2009. British Journal of Nutrition. 111(8): 1441-51. doi: 10.1017/S0007114513003917 78 DIETARY PATTERNS OF ETHNIC POPULATIONS Beavers, A. S., Lounsbury, J.W., Richards, J. K., Huck, S. W., Skolits, G, J., & Esquivel, S. L. (2013). Practical Considerations for Using Exploratory Factor Analysis in Educational Research. Practical Assessment, Research & Evaluation, 18(6): 1-13. Bédard, A., Goulet, J., Riverin, M., Lamarche, B., & Lemieux, S. (2010). Effects of a dietary intervention promoting the adoption of a Mediterranean food pattern on fast-food consumption among healthy French-Canadian women. British Journal of Nutrition, 104(11): 1662-1665. doi:10.1017/S000711451000262X Bouvard, V., Loomis, D., Guyton, K. Z., Grosse, Y., Ghissassi, F. E., Benbrahim-Tallaa, L., . . . International Agency for Research on Cancer Monograph Working Group. (2015). Carcinogenicity of consumption of red and processed meat. The Lancet. Oncology, 16(16): 1599-1600. doi: 10.1016/S1470-2045(15)00444-1 Brenner, D. R., Boucher, B. A., Kreiger, N., Jenkins, D., & El-Sohemy, A. (2011). Dietary patterns in an ethnoculturally diverse population of young canadian adults. Canadian Journal of Dietetic Practice and Research : A Publication of Dietitians of Canada = Revue Canadienne De La Pratique Et De La Recherche En Diététique : Une Publication Des Diététistes Du Canada, 72(3): e161-e168. doi:10.3148/72.3.2011e161. Broesch, J., & Hadley, C. (2012). Putting culture back into acculturation: Identifying and overcoming gaps in the definition and measurement of acculturation. The Social Science Journal, 49(3): 375-385. doi: 10.1016/j.soscij.2012.02.004 Buil-Cosiales, P., Irimia, P., Berrade, N., Garcia-Arellano, A., Riverol, M., Murie-Fernández, M., … Serrano-Martínez, M. (2008). Carotid intima-me2wdia thickness is inversely associated with olive oil consumption. Atherosclerosis. 196(2): 742-8. 79 DIETARY PATTERNS OF ETHNIC POPULATIONS Ceriello, A., Esposito, K., La Sala, L., Pujadas, G., De Nigris, V., Testa, R.,... Genovese, S. (2014). The protective effect of the Mediterranean diet on endothelial resistance to GLP-1 in type 2 diabetes: a preliminary report. Cardiovascular Diabetology, 13: 140. doi:10.1186/s12933014-0140-9 Chui, T., Tran, K., & Maheux, H. (2006). Immigration in Canada: A Portrait of the Foreign-born Population, 2006 Census. Statistics Canada Catalogue no. 97-557-XIF. Ottawa. Retrieved from: http://www12.statcan.ca/census-recensement/2006/as-sa/97-557/index-eng.cfm. Cleland, S. J., & Sattar, N. (2004). Impact of ethnicity on metabolic disturbance, vascular dysfunction and atherothrombotic cardiovascular disease. Diabetes, Obesity and Metabolism 7(5): 436-470. Cordain, L., Eaton, S. B., Sebastian, A., Mann, N., Lindeberg, S., Watkins, B. A.,… Brand-Miller, J. (2005). Origins and evolution of the Western diet: health implications for the 21st century. American Journal of Clinical Nutrition, 81(2): 341-54. Damasceno, N. R., Sala-Vila, A., Cofán, M., Pérez-Heras, A. M., Fitó, M., Ruiz-Gutiérrez, V.,… Ros E. (2013). Mediterranean diet supplemented with nuts reduces waist circumference and shifts lipoprotein subfractions to a less atherogenic pattern in subjects at high cardiovascular risk. Atherosclerosis, 230(2): 347-53. doi: 10.1016/j.atherosclerosis.2013.08.014. Darmon, N., Khlat, M. (2001). An overview of the health status of migrants in France, in relation to their dietary practices. Public Health Nutrition, 4(2): 163-72. Davis, N. J. Schechter, C. B., Ortega, F., Rosen, R., Wylie-Rosett, J., & Walker, E. A. (2013). Dietary patterns in Blacks and Hispanics with diagnosed diabetes in New York City's South Bronx. American Journal of Clinical Nutrition, 97(4): 878-885. doi: 10.3945/ ajcn.112.051185 80 DIETARY PATTERNS OF ETHNIC POPULATIONS de Lorgeril, M., Salen, P., Martin, J. L., Monjaud, I., Delaye, J. & Mamelle, N. (1999). Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: final report of the Lyon Diet Heart study. Circulation 99(6): 779-785. Denova-Gutierrez, E., Castanon, S., Talavera, J. O., Gallegos-Carrillo, K., Flores, M., DosamantesCarrasco, D., Salmeron, J. (2010). Dietary patterns are associated with metabolic syndrome in an urban Mexican population. The Journal of Nutrition, 140(10):1855-1863. doi: 10.3945/jn.110.122671 Désilets, M., Rivard, M., Shatenstein, B., & Delisle, H. (2007). Dietary transition stages based on eating patterns and diet quality among Haitians of Montreal, Canada. Public Health Nutrition, 10(5), 454-463. doi:10.1017/S1368980007222931 Dunn, R. A., Sharkey, J. R., & Horel, S. (2012). The effect of fast-food availability on fast-food consumption and obesity among rural residents: An analysis by race/ethnicity. Economics and Human Biology, 10: 1-13. El-Sayed, A. M., Scarborough, P., & Galea, S. (2011). Ethnic inequalities in obesity among children and adults in the UK: a systematic review of the literature. Obesity Reviews, 12(5): 516-34. doi: 10.1111/j.1467-789X.2010.00829.x Enas, E. A., Senthilkumar, H., Chennikkara, H. & Bjurlin, M. A. (2003). Prudent diet and preventive nutrition from pediatrics to geriatrics: current knowledge and practical recommendations. Indian Heart Journal, 55(4): 310–338. Estruch, R., Martínez-González, M. A., Corella, D., Salas-Salvadó, J., Ruiz-Gutiérrez, V., Covas, M. I.,… PREDIMED Study Investigators. (2006). Effects of a Mediterranean-style diet on cardiovascular risk factors: a randomized trial. Annals of Internal Medicine, 145(1): 1-11. 81 DIETARY PATTERNS OF ETHNIC POPULATIONS Estruch, R., Ros, E., Salas-Salvado, J., Covas, M. I., Corella, D., Aros, F.,… Martinez-Gonzalez, M. A. (2013). Primary prevention of cardiovascular disease with a Mediterranean diet. The New England Journal of Medicine, 368(14): 1279-1290. doi: 10.1056/NEJMoa1200303 Esposito, K., Maiorino, M. I, Petrizzo, M., Bellastella, G., & Giugliano, D. (2014). The effects of a Mediterranean Diet on the need for diabetes drugs and remission of newly diagnosed type 2 diabetes: Follow-up of a randomized trial. Diabetes Care, 37(7): 1824-1830. doi: 10.2337/dc13-2899 Esposito, K., Marfella, R., Ciotola, M., Di Palo, C., Giugliano, F., Giugliano, G.,… Giugliano, D. (2004). Effect of a Mediterranean-style diet on endothelial dysfunction and markers of vascular inflammation in the metabolic syndrome: a randomized trial. Journal of the American Medical Association, 292(12): 140-6. Fabrigar, L.R., Wegener, D.T., MacCallum, R.C. & Strahan, E.J. (1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4(3): 272-299. doi:10.1037/1082-989X.4.3.272 Feskanich, D., Rimm, E. B., Giovannucci, E. L., Colditz, G. A., Stampfer, M. J., Litin, L. B., Willet, W. C. (1992). Reproducibility and validity of food intake measurements from semiquantitative food frequency questionnaire. Journal of the American Dietetic Association, 93(7): 790-796. doi:10.1016/0002-8223(93)91754-E Fincham, J. E. (2008). Response rates and responsiveness for surveys, standards, and the journal. American Journal of Pharmaceutical Education, 72(2): 1-3. García-Fernández, E., Rico-Cabanas, L., Rosgaard, N., Estruch, R., & Bach-Faig, A. (2014). Mediterranean diet and cardio-obesity: A Review. Nutrients, 6(9): 3474-3500. doi:10.3390/nu6093474 82 DIETARY PATTERNS OF ETHNIC POPULATIONS Garduño-Diaz, S. D., & Khokhar, S. (2013). South Asian dietary patterns and their association with risk factors for the metabolic syndrome. Journal of Human Nutrition and Dietetics: The Official Journal Of The British Dietetic Association, 26(2): 145-155. doi:10.1111/j.1365277X.2012.01284.x Ghaddar, S., Brown, C. J., Pagan, J. A., & Diaz, V. (2010). Acculturation and healthy lifestyle habits among Hispanics in United States-Mexico border communities. Pan American Journal of Public Health. 28(3): 190-7. Gilbert, P. A., & Khokhar, S. (2008). Changing dietary habits of ethnic groups in Europe and implications for health. Nutrition Reviews, 66(4): 203-215. doi: 10.1111/j.17534887.2008.00025.x. Health Canada. (2003). Canadian guidelines for body weight classification in adults, nutrition and healthy eating, food and nutrition. Retrieved from: https://www.canada.ca/en/healthcanada/services/food-nutrition/healthy-eating/healthy-weights/canadian-guidelines-bodyweight-classification-adults/questions-answers-public.html#a1. Henríquez-Sánchez, P., Ruano, C., de Irala, J., Ruiz-Canela, M., Martínez-González, M. A., & Sánchez-Villegas, A. (2012). Adherence to the Mediterranean diet and quality of life in the SUN Project. European Journal of Clinical Nutrition, 66(3):360-368. doi: 10.1038/ejcn.2011.146. Herman, A., Ryan. K., Harry, J. (2013). Best-Practice Recommendations for Defining, Identifying, and Handling Outliers. Organizational Research Methods, 16(2): 270 – 301. doi: 10.1177/1094428112470848 Hill, J. L., Chau, C., Luebbering, C. R., Koivras, K. K., & Zoellner, J. (2012). Does availability of physical activity and food outlets differ by race and income? Findings from an enumeration 83 DIETARY PATTERNS OF ETHNIC POPULATIONS study in a health disparate region. International Journal of Behavioral Nutrition and Physical Activity, 9(105): 1-10. doi: 10.1186/1479-5868-9-105 Hogarty, K.Y., Hines, C.V., Kromrey, J. D., Ferron, J. M & Mumford, K. R. (2005). The quality factor solutions in exploratory factor analysis: the influence of sample size, communality, and overdetermination. Educational and Psychological Measurement, 65(2): 202-226. Hu, F. B. (2002). Dietary pattern analysis: a new direction in nutritional epidemiology. Current opinion in lipidology, 13(1): 3-9. Hu, F. B., Rimm, E. B., Stampfer, M. J., Ascherio, A., Spiegelman, D., & Willett, W. C. (2000). Prospective study of major dietary patterns and risk of coronary heart disease in men. American Journal of Clinical Nutrition, 72(4): 912-921. Hu, F. B., & Willett, W. C. (2002). Optimal diets for prevention of coronary heart disease. JAMA, 288(20): 2569-2578. doi:10.1001/jama.288.20.2569 IMSS. (2015). Programs and Services. Retrieved from: http://www.imss.ca Jacobs, D. R., & Steffen, L. M. (2003). Nutrients, foods, and dietary patterns as exposures in research: A framework for food synergy. The American Journal of Clinical Nutrition, 78(3 Suppl): 508S-513S. doi: 10.1093/ajcn/78.3.508S Jemal, A., Center, M. M., DeSantis, C., & Ward, E. M. (2010). Global Patterns of Cancer Incidence and Mortality Rates and Trends. Cancer Epidemiology, Markers & Prevention, 19: 18931907. doi: 10.1158/1055-9965.EPI-10-0437 Jessri, M., Ng, A. P., & L’Abbe, M. R. (2017). Adapting the healthy eating index 2010 for the Canadian population: Evidence from the Canadian community health survey. Nutrients, 9(8), 910. doi: 10.3390/nu9080910 84 DIETARY PATTERNS OF ETHNIC POPULATIONS Jones-McLean, E. M., Shatenstein, B., Whiting, S. J. (2010). Dietary patterns research and its applications to nutrition policy for the prevention of chronic disease among diverse North American populations. Applied Physiology, Nutrition, and Metabolism, 35(2): 195-8. doi: 10.1139/H10-003 Kant, A. K. (2004). Dietary patterns and health outcomes. Journal of American Dietetic Association104(4): 615-635. Kenfield, S. A., DuPre, N., Richman, E. L., Stampfer, M. J., Chan, J. M., & Giovannucci, E. L. (2014). Mediterranean diet and prostate cancer risk and mortality in the health professionals follow-up study. European Urology, 65(5): 887-894. doi: 10.1016/j.eururo.2013.08.009 Kennedy, E. T., Ohls, J., Carlson, S. & Fleming, K. (1995). The Healthy Eating Index: design and applications. Journal of the American Dietetic Association, 95(10): 1103-1108. Khamis, T. K. (1996). South Asian foodways in Britain - Diversity and change and the implications for health promotion (Doctoral dissertation), University of London: UK. Kimokoti, R. W., & Millen, B. E. (2011). Diet, the global obesity epidemic, and prevention. Journal of the American Dietetic Association, 111(8): 1137-1140. doi: 10.1016/j.jada.2011.05.016 Koloverou, E., Esposito, K., Giugliano, D., & Panagiotakos, D. (2014). The effect of Mediterranean diet on the development of type 2 diabetes mellitus: a meta-analysis of 10 prospective studies and 136,846 participants. Metabolism, 63(7): 903–911. Korkeila, K., Suominen, S., Ahvenainen, J., Ojanlatva, A., Rautava, P., Helenius, H., & Koskenvuo, M. (2001). Non-Response and Related Factors in a Nation-Wide Health Survey. European Journal of Epidemiology, 17(11): 991-999. Retrieved from http://www.jstor.org/stable/3582579 85 DIETARY PATTERNS OF ETHNIC POPULATIONS Lai, D. W., & Chau, S. B. (2007). Predictors of health service barriers for older Chinese immigrants in Canada. Health & Social Work, 32(1): 57-65. Landman, J., & Cruickshank, J. K., (2001). A review of ethnicity, health and nutrition-related diseases in relation to migration in the United Kingdom. Public Health Nutrition, 4(2B): 64757. doi: 10.1079/PHN2001148 Lee, M. M., & Shen, J. M. (2008). Dietary patterns using Traditional Chinese Medicine principles in epidemiological studies. Asia Pacific Clinical Nutrition, 17(S1): 79-81 Lenighan, Y., Nugent, A., Li, K., Brennan, L., Walton, J., Flynn, A., . . . McNulty, B. (2017). Processed red meat contribution to dietary patterns and the associated cardio-metabolic outcomes. British Journal of Nutrition, 118(3): 222-228. doi:10.1017/S0007114517002008 Lesser, I. A., Gasevic, D., & Lear, S. A. (2014). The association between acculturation and dietary patterns of South Asian immigrants. PLoS ONE 9(2): e88495. doi: 10.1371/journal.pone.0088495 Lewis, M. (2007). Stepwise versus hierarchical regression: Pros and cons. Paper presented at the annual meeting of the Southwest Educational Research Association, San Antonio, TX. Marchioni, D. M., Latorre, M. R., Eluf-Neto, J., Wünsch-Filho, V., & Fisberg, R. M. (2005). Identification of dietary patterns using factor analysis in an epidemiological study in São Paulo. Sao Paulo Medical Journal 123(3): 124-127. Martínez-González, M. Á., Toledo, E., Arós, F., Fiol, M., Corella D., Salas-Salvadó, J.,… Alonso, A. (2014). Extravirgin olive oil consumption reduces risk of atrial fibrillation: the PREDIMED (Prevención con Dieta Mediterránea) trial. Circulation, 130(1): 18-26, doi: 10.1161/CIRCULATIONAHA.113.006921 86 DIETARY PATTERNS OF ETHNIC POPULATIONS Mattioli, A. V., Miloro, C., Pennella, S., Pedrazzi, P., & Farinetti, A. (2013). Adherence to Mediterranean diet and intake of antioxidants influence spontaneous conversion of atrial fibrillation. Nutrition Metabolism and Cardiovascular Diseases, 23(2): 115–21. doi: 10.1016/j.numecd.2011.03.005 Mattei, J., Willette, W. C., Li, Y., Hu, F. B., Bhupathiraju, S. N., Pan, A., . . . Rimm, E. B. (2017). Association of changes in diet quality with total and cause-specific mortality. The New England Journal of Medicine, 377(2): 143-153. doi:10.1056/NEJMoa1613502 McKeigue, P., & Shah, B. (1991). Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet, 337(8738): 382-386. Mitjavila, M. T., Fandos, M., Salas-Salvadó, J., Covas, M. I., Borrego, S., Estruch, R., … Sáez, G. T. (2013). The Mediterranean diet improves the systemic lipid and DNA oxidative damage in metabolic syndrome individuals. A randomized, controlled, trial. Clinical Nutrition, 32(2): 172-178. doi: 10.1016/j.clnu.2012.08.002 Misra, A., Khurana, L., Isharwal, S., & Bhardwaj, S. (2009). South Asian diets and insulin resistance British Journal of Nutrition. 101(4): 465-473. doi:10.1017/S0007114508073649 Mitrou, P. N., Kipnis, V., Thiébaut, A. C., Reedy, J., Subar, A. F., Wirfält, E,… Schatzkin, A. (2007). Mediterranean dietary pattern and prediction of all-cause mortality in a US population: results from the NIH-AARP Diet and Health Study. Archives of Internal Medicine, 167(22): 2461-2468. Nettleton, J. A., Steffen, L. M., Mayer-Davis, E. J., Jenny, N. S., Jiang, R., Herrington, D. M., & Jacobs, D. R. Jr. (2006). Dietary patterns are associated with biochemical markers of inflammation and endothelial activation in the Multi-Ethnic Study of Atherosclerosis (MESA). American Journal of Clinical Nutrition 83(6): 1369-1379. 87 DIETARY PATTERNS OF ETHNIC POPULATIONS Newby, P. K., Muller, D., Hallfrisch, J., Andres, R., & Tucker, K. L. (2004). Food patterns measured by factor analysis and anthropometric changes in adults. The American Journal of Clinical Nutrition, 80(2): 504-513. doi: 10.1093/ajcn/80.2.504 Ngo, J., Gurinovic, M., Anderson, L. F., & Majem, L. S. (2009). How dietary intake methodology is adapted for use in European immigrant population groups - a review. British Journal of Nutrition, 101(S2): S86-S94. doi: 10.1017/S0007114509990614 Ottesen, G. H., & Wandel, M. (2012). Changes in dietary habits after migration and consequences for health: a focus on South Asians in Europe. Food & Nutrition Research 56. doi: 10.3402/fnr.v56i0.18891 Panagiotakos, D. B., Pitsavos, C., Chrysohoou, C., Skoumas, J., Tousoulis, D., Toutouza, M.,... Stefanadis, C. (2004). Impact of lifestyle habits on the prevalence of the metabolic syndrome among Greek adults from the ATTICA study. American Heart Journal 147(1): 106–112. Panagiotakos, D., Pitsavos, C., Skoumas, Y., Stefanadis, C. (2007). The association between food patterns and the metabolic syndrome using principal components analysis: The ATTICA study. Journal of American Dietetic Association, 107(6): 979-987. doi:10.1016/j.jada.2007.03.006. Panagiotakos, D. B., Polystipioti, A., Papairakleous, N., & Polychronopoulos, E. (2007). Long-term adoption of a Mediterranean diet is associated with a better health status in elderly people; a cross-sectional survey in Cyprus. Asia Pacific Journal Of Clinical Nutrition, 16(2): 331-337. Panagiotakos, D. (2014). A Mediterranean diet supplemented with olive oil or nuts reduces the incidence of major cardiovascular events in high-risk patients. Evidence Based Medicine, 19(1): 10. doi:10.1136/eb-2013-101391 88 DIETARY PATTERNS OF ETHNIC POPULATIONS Perona, J. S., Covas, M. I., Fitó, M., Cabello-Moruno, R., Aros, F., Corella, D.,… Ruiz-Gutierrez, V. (2010). Reduction in systemic and VLDL triacylglycerol concentration after a 3-month Mediterranean-style diet in high-cardiovascular-risk subjects. The Journal of Nutritional Biochemistry, 21(9):892-898. doi: 10.1016/j.jnutbio.2009.07.005 Public Health Agency of Canada. (2015). Chronic disease and injury indicator framework, Quick Stats. Retrieved from: http://www.phac-aspc.gc.ca/publicat/hpcdp-pspmc/3510/assets/pdf/ar-03-eng.pdf Poustie, V. J., Smyth, R. L., & Cole, T. J. (2005). Reliability of calculating body mass index centile. European Journal Of Clinical Nutrition, 59(5): 717-719. Quatromoni, P. A., Copenhafer, D. L., Demissie, S., D'Agostino, R. B., O'Horo, C. E., Nam, B. H., Millen, B. E. (2010). The internal validity of a dietary pattern analysis. The Framingham Nutrition Studies. Journal of Epidemiology and Community Health, 56(5):381-8. Remington, P. L., & Brownson, R. C. (2011). Fifty years of progress in chronic disease epidemiology and control. MMWR Surveillance Summary, 60(S4): 70-77. Rezazadeh, A., Omidvar, N., Eini-Zinab, H., Ghazi-Tabatabaie, M., Majdzadeh, R., Ghavamzadeh, S., & Nouri-Saeidlou, S. (2016). Major dietary patterns in relation to demographic and socioeconomic status and food insecurity in two iranian ethnic groups living in urmia, iran. Public Health Nutrition, 19(18): 3337-3348. doi:10.1017/S1368980016001634 Rimm, E. B., Giovannucci, E. L., Stampfer, M. J., Colditz, G. A., Litin, L. B., & Willett, W. C. (1992). Reproducibility and validity of an expanded self-administered semi-quantitative food frequency questionnaire among male health professionals. American Journal of Epidemiology, 135(10): 1114-1126. 89 DIETARY PATTERNS OF ETHNIC POPULATIONS Rogol, A. D., Clark, P. A., & Roemmich, J. N. (2000). Growth and pubertal development in children and adolescents: Effects of diet and physical activity. The American Journal of Clinical Nutrition, 72(2 Suppl): 521S.-528S. doi: 10.1093/ajcn/72.2.521S Ros, E., Martínez-González, M. A., Estruch, R., Salas-Salvadó, J., Fitó, M., Martínez, J. A., & Corella, D. (2014). Mediterranean diet and cardiovascular health: Teachings of the PREDIMED study. Advanced Nutrition, 5(3): 330S-6S. doi: 10.3945/an.113.005389 Rosenmoller, D. L., Gasevic, D., Seidell, J., & Lear, S. A. (2011). Determinants of changes in dietary patterns among chinese immigrants: A cross-sectional analysis. International Journal of Behavioral Nutrition & Physical Activity, 8(42): 1-8. doi: 10.1186/1479-5868-8-42. Sahlqvist, S., Song, Y., Bull, F., Adams, E., Preston, J., Ogilvie, D., & the iConnect consortium. (2011). Effect of questionnaire length, personalisation and reminder type on response rate to a complex postal survey: Randomised controlled trial. BMC Medical Research Methodology, 11(1): 62-62. doi: 10.1186/1471-2288-11-62 Salas-Salvadó, J., Bulló, M., Estruch, R., Ros, E., Covas, M. I., Ibarrola-Jurado, N., … MartínezGonzález, M. A. (2014). Prevention of diabetes with Mediterranean diets: a subgroup analysis of a randomized trial. Annals of Internal Medicine, 160(1): 1-10. doi: 10.7326/M131725. Salas-Salvadó, J., Fernández-Ballart, J., Ros, E., Martínez-González, M. A., Fitó, M., Estruch, R., … PREDIMED Study Investigators. (2008). Effect of a Mediterranean diet supplemented with nuts on metabolic syndrome status: one-year results of the PREDIMED randomized trial. Archives of Internal Medicine, 168(22): 2449-58. doi: 10.1001/archinte.168.22.2449 Sala-Vila, A., Romero-Mamani, E. S., Gilabert, R., Núñez, I., de la Torre, R., Corella, D.,… Ros, E. (2014). Changes in ultrasound-assessed carotid intima-media thickness and plaque with a 90 DIETARY PATTERNS OF ETHNIC POPULATIONS Mediterranean diet: a substudy of the PREDIMED trial. Arteriosclerosis, thrombosis, and vascular biology. 34(2): 439-445. doi: 10.1161/ATVBAHA.113.302327. Sanou, D., O'Reilly, E., Ngnie-Teta, I., Batal, M., Mondain, N., Andrew, C., Bourgeault, I. L. (2014). Acculturation and nutritional health of immigrants in Canada: a scoping review. Journal of Immigrant and Minority Health, 16(1): 24-34. doi: 10.1007/s10903-013-9823-7 Sarkar, J., Lix, L. M., Bruce, S., & Young, T. K. (2010). Ethnic and regional differences in prevalence and correlates of chronic diseases and risk factors in northern Canada. Preventing Chronic Disease, 7(1): 1-11. Retrieved from: http://www.cdc.gov/pcd/issues/2010/jan/09_0001.htm Satia, J. A., Patterson, R. E., Kristal, A. R. Hislop, T. G., Yasui, Y., & Taylor, V. M. (2001). Development of scales to measure dietary acculturation among Chinese-Americans and Chinese-Canadians. Journal of the American Dietetic Association, 101(5): 548-553. Satia, J. A., Patterson, R. E., Neuhouser, M. L., & Elder. J. (2002). Dietary acculturation: applications to nutrition research and dietetics. Journal of the American Dietetic Association, 102(8): 1105-1118. Sax, L. J., Gilmartin, S. K., & Bryant, A. N. (2003). Assessing response rates and nonresponse bias in web and paper surveys. Research in Higher Education 44(4): 409-432. Shatenstein, B., & Ghadirian, P. (1998). Influences on diet, health behaviours and their outcome in select ethno-cultural and religious groups. Nutrition, 14(2): 223-230. Shu, L., Zheng, P. F., Zhang, X. Y., Si, C. J., Yu, X. L., Gao, W.,…Liao, D. (2015). Association between Dietary Patterns and the Indicators of Obesity among Chinese: A CrossSectional Study. Nutrients, 7(9): 7995-8009. doi: 10.3390/nu7095376. 91 DIETARY PATTERNS OF ETHNIC POPULATIONS Shubair, M. M., McColl, R. S., & Hanning, R. M. (2005). Mediterranean dietary components and body mass index in adults: The peel nutrition and heart health survey. Chronic Diseases in Canada, 26(2-3): 43-51. Schulze, M. B. & Hoffmann, K. (2006). Methodological approaches to study dietary patterns in relation to risk of coronary heart disease and stroke, British Journal of Nutrition, 95(5): 860869. Statistics Canada. (2008). Canadian demographics at a glance. Statistics Canada Catalogue no. 91003-X. Ottawa. pp. 1-54. Retrieved from: http://www.statcan.gc.ca/pub/91-003-x/91-003-x2007001-eng.pdf Statistics Canada. (2010). Visual census, 2006 Census. Ottawa. Retrieved from 2010. http://www12.statcan.gc.ca/census-recensement/2006/dp-pd/fsfi/index.cfm?Lang=ENG&TOPIC_ID=11&PRCODE=59 Statistics Canada. (2012). Focus on Geography Series, 2011 Census. Statistics Canada Catalogue no. 98-310-XWE2011004. Ottawa, Ontario. Analytical products, 2011 Census. Retrieved from: http://www12.statcan.gc.ca/census-recensement/2011/as-sa/fogs-spg/Facts-preng.cfm?Lang=Eng&GK=PR&GC=59 Suhr, D. (2006). Exploratory or Confirmatory Factor Analysis. SAS Users Group International Conference (pp. 1 - 17). Cary: SAS Institute, Inc. Retrieved from http://www2.sas.com/proceedings/sugi31/200-31.pdf Suresh, K., & Chandrashekara, S. (2012). Sample size estimation and power analysis for clinical research studies. Journal of Human reproductive sciences, 5(1): p. 8. doi: 10.4103/09741208.97779. 92 DIETARY PATTERNS OF ETHNIC POPULATIONS Tabachnick, B. G. & Fidell, L. S. (2013). Using Multivariate Statistics (6TH ed.). Boston: Allyn and Bacon. Toledo, E., Hu, F. B., Estruch, R., Buil-Cosiales, P., Corella, D., Salas-Salvadó, J., … MartinezGonzalez, M. A. (2013). Effect of the Mediterranean diet on blood pressure in the PREDIMED trial: results from a randomized controlled trial. BMC Medicine, 11: 207. doi: 10.1186/1741-7015-11-207. Tortosa, A., Bes-Rastrollo, M., Sanchez-Villegas, A., BasterraGortari, F. J., Nunez-Cordoba, J. M. & Martinez-Gonzalez, M. A. (2007). Mediterranean diet inversely associated with the incidence of metabolic syndrome: the SUN prospective cohort. Diabetes Care 30 (11): 2957– 2959. Tresserra-Rimbau, A., Rimm, E. B., Medina-Remón, A., Martínez-González, M. A., de la Torre, R., Corella, D., … PREDIMED Study Investigators. (2014). Inverse association between habitual polyphenol intake and incidence of cardiovascular events in the PREDIMED study. Nutrition, metabolism, and cardiovascular diseases, (6): 639-47. doi: 10.1016/j.numecd.2013.12.014. Trichopoulou, A., & Lagiou, P. (1997). Healthy traditional Mediterranean diet: an expression of culture, history, and lifestyle. Nutrition Reviews, 55(11): 383-389. Tucker, K. L. (2010). Dietary patterns, approaches, and multicultural perspective. Applied Physiology, Nutrition, and Metabolism, 35(2): 211-218. doi: 10.1139/H10-010 Twells, L. K., Gregory, D., Reddigan, J., Midodzi, W. (2014). The current prevalence and future predictions of obesity in Canada: a trend analysis. Canadian Medical Association Journal 2(1): E18-E26. doi:10.9778/cmajo.20130016 93 DIETARY PATTERNS OF ETHNIC POPULATIONS U.S. Department of Agriculture. (1992). The Food Guide Pyramid. Human Nutrition Information Service. Washington, DC. (Homes and Garden Bulletin no. 252). U.S. Geological Survey. (2015). Fort Collins science center, Snow sampling. Retrieved from: https://www.fort.usgs.gov/LandsatSurvey/SnowballSampling VanKim, N. A., Austin, S. B., Jun, H., Hu, F. B., & Corliss, H. L. (2017). Dietary patterns during adulthood among lesbian, bisexual, and heterosexual women in the nurses’ health study II. Journal of the Academy of Nutrition and Dietetics, 117(3): 386-395. 10.1016/j.jand.2016.09.028 Vargas, P., & Jurado, L. (2016). Dietary acculturation among filipino americans. International Journal of Environmental Research and Public Health, 13(1): 16. doi: 10.3390/ijerph13010016 Varghese, S., & Moore-Orr, R. (2002). Dietary acculturation and health-related issues of Indian immigrant families in Newfoundland. Canadian Journal of Dietetic Practice and Research, 63(2): 72-79. Vuksan, V., Rogovik, A., Jenkins, A., Peeva, V., Beljan-Zdravkovic, U., Stavro, M., Watson, W. (2012). Cardiovascular risk factors, diet and lifestyle among European, South Asian and Chinese adolescents in Canada. Paediatric Child Health, 17(1): e1-e6. Wahi, G., Boyle, M. H., Morrison, K. M., & Georgiades, K. (2014). Body mass index among immigrant and non-immigrant youth: evidence from the Canadian Community Health Survey. Canadian Journal of public Health, 105(4): e239-244. Wallner-Liebmann, S. J., Kruschitz, R., Hübler, K., Hamlin, M. J., Schnedl, W. J., Moser, M., & Tafeit, E. (2013). A measure of obesity: BMI versus subcutaneous fat patterns in young athletes and nonathletes. Collegium Antropologicum, 37(2): 351-357. 94 DIETARY PATTERNS OF ETHNIC POPULATIONS Wilson VanVoorhis, & C., Morgan, B.L., (2007). Understanding power and rules of thumb for determining sample sizes. Tutorials in Quantitative Methods for Psychology 3(2): 43–50. doi: 10.20982/tqmp.03.2.p043 World Health Organization. (2003). Diet, nutrition and the prevention of chronic diseases (WHO Techical Report series 916). Geneva, Switzerland. Whayne, T. F. Jr. (2014). Ischemic heart disease and the Mediterranean diet. Current cardiology reports, 16(6):491-498. doi: 10.1007/s11886-014-0491-6 Yang, E. J., Kerver, J. M., & Song, W. O. (2005). Dietary patterns of Korean Americans described by factor analysis. Journal of American College of Nutrition, 24(2): 115-121. Yong. G., Pearce, S. (2013). A Beginner’s Guide to Factor Analysis: Focusing on Exploratory Factor Analysis. Tutorials in Quantitative Methods for Psychology, 9(2): 79-94. Yusuf, S., Redyy, S., Ounpuu, S., & Anand, S. (2001). Global burden of cardiovascular diseases: Part I: general considerations, the epidemiologic transition, risk factors, and impact of urbanization. Circulation, 104(22): 2746-2753. doi:10.1161/hc4601.099487 Zhang, M., Zhu, Y., Li, P., Chang, H., Wang, X., Liu, W.,… Huang, G. (2015). Associations between Dietary Patterns and Impaired Fasting Glucose in Chinese Men: A Cross-Sectional Study. Nutrients, 7(9): 8072-8089. doi: 10.3390/nu7095382 95 DIETARY PATTERNS OF ETHNIC POPULATIONS Appendix A Dietary Pattern Survey Questionnaire Thesis Title: Dietary Patterns of Ethnic Non-Indigenous Populations in Northern British Columbia: A Cross-Sectional Survey Study Yaser Ahmed March 29, 2016 96 DIETARY PATTERNS OF ETHNIC POPULATIONS The next question is included for screening purposes: SCREEN_1. Are you on a special diet regimen because of kidney failure or Celiac disease? <1> Yes <2> No <98> Don't know SOCIO_1. What is your age? Enter age between 19 years and 130 years:_|_|_|_|_| OR AND year of birth: :_|_1|9_|_|_| ☐ Don’t know SOCIO_2. Gender: Are you a male or female? <1> Male <2> Female <3> Other <4> Do not wish to disclose <98> Don't know SOCIO_3. Weight: What is your weight? Can you please measure your body weight using a scale and indicate weight below? Enter your weight: |_|_|_| in ☐Pounds (lbs) or ☐Kilograms (kg) OR ☐ Don’t know SOCIO_4. Height: What is your height? 97 DIETARY PATTERNS OF ETHNIC POPULATIONS Enter your Height: |_| Feet (ft.) and |_|_| Inches (in) OR |_|_|_| Centimeters (cm) OR ☐ Don’t know Postal code What are the first three digits of your postal code? Enter the first three digits: |_|_|_| OR ☐ Don’t know First Section: I’d like to start by asking you three questions regarding your health status: q1. Are you presently…? <1> diabetic <2> prediabetic <3> non-diabetic q2. Do you have high blood pressure…? <1> yes <2> no <98> Don’t know <99> Refused q3. Do you have high cholesterol…? <1> yes <2> no <98> Don’t know <99> Refused 98 DIETARY PATTERNS OF ETHNIC POPULATIONS Second Section: Diet Pattern Questions: I would like to ask some questions about your eating habits for specific foods. There are nine sections to this part of the survey. Please circle one response only. Please review in your mind how often on overage have you eaten certain vegetables and fruits over the past month. You will be able to respond in the following categories: More than once per day; Everyday; Five to six days per week; Two to four days per week; One day per week; Less than one day per week; Never; or Don’t know. q4. First, over the past month, how often on average did you eat green leafy vegetables such as lettuce, spinach, bok choy? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q5. Over the past month, how often on average did you eat orange and dark yellow vegetables such as carrots, squash, or sweet potatoes, including such vegetables in soups, casseroles, or stir fries AND fresh, frozen, or canned vegetables? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q6. Over the past month, how often on average did you eat cabbage-type of vegetables such as Chinese cabbages, broccoli, coleslaw and uncooked cabbage, cooked cabbage, cauliflower, Brussels 99 DIETARY PATTERNS OF ETHNIC POPULATIONS sprouts, kale, mustard greens, chard greens, or collard greens, including fresh, frozen, or canned vegetables? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q7. Over the past month, how often on average did you eat other vegetables such as celery, mushrooms, sweet peppers, corn, mixed vegetables, eggplant, zucchini, okra, bean sprouts, or alfalfa sprouts including fresh, frozen, or canned vegetables? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q8. Over the past month, how often on average did you eat green and yellow beans, fava beans, OR green peas or snow peas? This includes fresh, frozen, or canned beans or peas? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 100 DIETARY PATTERNS OF ETHNIC POPULATIONS q9a. Over the past month, how often on average did you eat non-citrus fruits such as apples, grapes, pears, peaches, apricots, plums, or bananas? This is EXCLUDING fruit juices, but would include non-sweetened canned (e.g., apple sauce), frozen, or dried fruits (e.g., dried apricots). <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q9b. Over the past month, how often on average did you eat citrus fruits such as oranges, lemon, or grapefruit? Again, this EXCLUDES juices, but would include non-sweetened canned or jarred fruit? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q9c. EXCLUDING juices, over the past month, how often on average did you eat melons such as cantaloupes, watermelons? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 101 DIETARY PATTERNS OF ETHNIC POPULATIONS q9d. EXCLUDING juices, over the past month, how often on average did you eat berries such as strawberries, blueberries, raspberries, or blackberries including unsweetened canned or frozen berries? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q9e. EXCLUDING juices, over the past month, how often on average did you eat tropical fruits such as papaya, mango, avocado, passion fruit or pineapple including canned tropical fruits such as canned papaya, canned mango, and canned pineapple? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q10a. Over the past month, how often on average did you eat or use ketchup with your food? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 102 DIETARY PATTERNS OF ETHNIC POPULATIONS q10b. Over the past month, how often on average did you eat pizza which used tomato sauce or tomato paste as a topping? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q10c. EXCLUDING ketchup and tomato sauce or tomato paste on pizza; over the past month, how often did you eat tomatoes, tomato sauce, or tomato paste in your food or drink tomato juice? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know Third Section: The next set of questions deals with how often you have eaten grains, lentils, and starch-type foods over the past month. You will be able to respond in the following categories: More than once per day; Everyday; Five to six days per week; Two to four days per week; One day per week; Less than one day per week; or Never. q11a. Over the past month, how often on average did you eat french fries which were deep fried in oil, including fries made at home or fries eaten away from home including hash brown for example? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week 103 DIETARY PATTERNS OF ETHNIC POPULATIONS <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q11b. EXCLUDING french fries, over the past month, how often on average did you eat frozen potato products? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week 7 <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q11c. Over the past month, how often on average did you eat baked, boiled, mashed, or roasted potatoes BUT EXCLUDING french fries, potato chips, or any frozen potato products? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q12. Over the past month, how often on average did you eat cooked breakfast cereals such as porridge, oatmeal, cream of wheat, or other cooked breakfast cereal? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week 104 DIETARY PATTERNS OF ETHNIC POPULATIONS <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q13. Over the past month, how often on average did you eat products made of white flour, such as white bread, pita, naan, paratha, roti, chapatti, muffins, bagels, buns, rolls, muffins, biscuits, pancakes, waffles, or plain crackers? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q14. Over the past month, how often on average did you eat whole wheat foods, such as whole wheat bread, pita, or naan, whole wheat paratha, whole wheat roti, whole wheat chapati, whole wheat muffins, whole wheat bagels, whole wheat buns, whole wheat rolls, whole wheat biscuits, whole wheat pancakes, whole wheat waffles, or whole wheat crackers? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q15. Over the past month, how often on average did you eat white rice including rice pudding or kheer, pasta, spaghetti, macaroni, noodles, rice noodles, or polenta? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week 105 DIETARY PATTERNS OF ETHNIC POPULATIONS <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q16. Over the past month, how often on average did you eat brown rice, whole wheat pasta, other grains i.e., barley, millet, whole wheat couscous, buck wheat, kasha, bulgar, OR bran included in your food, wheat germ, roti, chapatti, or tabouli? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q17. Over the past month, how often on average did you eat split peas, chickpeas or hummus, or dried beans such as kidney beans, black beans, white beans, fava beans, pinto beans, or lentils? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know This section of the survey, we will be asking you how often on average have you eaten meats, fish, and poultry over the past month. q18a. Over the past month, how often on average did you eat deep fried or breaded fish or chicken? <1> More than once per day 106 DIETARY PATTERNS OF ETHNIC POPULATIONS <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q18b. Excluding deep fried and breaded fish products, over the past month how often on average did you eat fresh, frozen, or canned white fish, such as sole, halibut, cod, whitefish, haddock, or other white fish? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q19. EXCLUDING deep fried and breaded fish products, over the past month how often on average did you eat fresh, frozen, or canned dark fish, such as salmon, sardines, mackerel, trout, herring, tuna, bluefish, or swordfish? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 107 DIETARY PATTERNS OF ETHNIC POPULATIONS q20. EXCLUDING deep fried and breaded shellfish products, over the past month how often on average did you eat shellfish or other sea foods, such as shrimp, lobster, crab, squid, octopus, clams, oysters, mussels, scallops or other shellfish? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q21a. EXCLUDING deep fried and breaded chicken or turkey, over the past month how often on average did you eat chicken or turkey whether it’s roasted, baked, or grilled? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q21b. EXCLUDING deep fried and breaded duck or goose, over the past month how often on average did you eat duck or goose whether it’s roasted, baked, or grilled? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 108 DIETARY PATTERNS OF ETHNIC POPULATIONS q22a. Over the past month, how often on average did you eat eggs with yolks? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q22b. Over the past month, how often on average did you eat egg whites; that is to say eggs without yolks? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q23. Over the past month, how often did you eat foods made with ground meats, such as meat balls, meat loaf, ground meats in sauces such as spaghetti meat sauce, hamburgers or cheeseburgers? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q24. Over the past month, how often on average did you eat organ meats, such as liver, kidney, heart, or other organ meats? 109 DIETARY PATTERNS OF ETHNIC POPULATIONS <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q25. Over the past month, how often on average did you eat processed meats such as hot dogs, salami, corned beef, bologna, wieners or sausages, bacon, ham, luncheon meats or other processed deli-style or canned meats? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q26. EXCLUDING ground meats AND processed meats, over the past month how often on average did you eat beef, pork, or lamb such as steaks, roasts, stews, chops or other cuts? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q27. EXCLUDING ground meats AND processed meats, over the past month how often on average did you eat veal, goat, or rabbit such as steaks, roasts, stews, chops or other cuts? 110 DIETARY PATTERNS OF ETHNIC POPULATIONS <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know The following section deals with Snack-type foods. You will be able to respond in the following categories: More than once per day; Everyday; Five to six days per week; Two to four days per week; One day per week; Less than one day per week; or Never. q28. Over the past month, how often on average did you eat chocolate bars, chocolate pieces, cookies, brownies, doughnuts, cake, pie, coffee cake, sweet roll, or pastry? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q29. Over the past month, how often on average did you eat peanut butter, or sesame butter? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 111 DIETARY PATTERNS OF ETHNIC POPULATIONS q30. Over the past month, how often on average did you eat nuts such as peanuts, pistachios, almonds, hazelnuts, pecans, cashews, chestnuts, walnuts, or pine nuts? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q31. EXCLUDING NUTS, over the past month how often on average did you eat snacks such as potato chips, corn chips, popcorn with butter, nachos, tortilla chips, or papadum? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q32. EXCLUDING cream and homogenized milk, over the past month, how often on average did you drink 2% milk, 1% milk, or skim milk? This includes both fresh milk or evaporated or powdered milk. This also includes milk used in cereals, sauces, or soups? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 112 DIETARY PATTERNS OF ETHNIC POPULATIONS The next set of questions deals with dairy and other high-protein foods. q33. Over the past month, how often on average did you eat frozen low-fat dairy products, such as low-fat or non-fat ice cream, frozen yoghurt, ice milk or sherbet? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q34. Over the past month, how often on average did you eat plain or sweetened yoghurt? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q35a. Over the past month, how often on average did you drink homogenized milk? This includes milk used in cereals, sauces, or soups? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 113 DIETARY PATTERNS OF ETHNIC POPULATIONS q35b. Over the past month, how often on average did you eat regular or rich ice cream? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q36. Over the past month, how often on average did you eat paneer, cottage cheese or ricotta cheese? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q37a. Over the past month, how often on average did you use cream in your coffee or tea? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q37b. EXCLUDING cream added to your coffee or tea, over the past month how often on average did you use whipping cream, sour cream, or cream cheese in your food? <1> More than once per day 114 DIETARY PATTERNS OF ETHNIC POPULATIONS <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q38. EXCLUDING paneer, cottage cheese, ricotta cheese, and cream cheese, over the past month how often on average did you eat cheese such as cheddar cheese, mozzarella cheese, brie, feta cheese or goat cheese? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q39. Over the past month, how often on average did you eat soy products such as soybeans, tofu, miso, tempeh, soy ‘milk’, soy cheese, soy flakes, soy grits, or soy flour, but NOT soy sauce? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know Sixth Section: The next set of questions deals with various beverages you may drink. 115 DIETARY PATTERNS OF ETHNIC POPULATIONS q40. EXCLUDING tomato juice, over the past month, how often on average did you drink 100% or natural fruit juices such as apple juice, orange juice, grapefruit juice, or other 100% fruit juices? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q41a. Over the past month, how often on average did you drink regular soft drinks such as regular coke, regular pepsi, or regular gingerale, OR fruit-flavoured drinks such as Kool-Aid, Fruite, Gatoraide, Sunny Delight, lemonade, fruit punch, or iced tea? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q41b. Over the past month, how often on average did you drink diet soft drinks such as diet coke, diet pepsi, diet gingerale, or any other diet soft drink? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know 116 DIETARY PATTERNS OF ETHNIC POPULATIONS The next set of questions will ask you about foods that are commonly used as flavourings. q42. Over the past month, how often on average did you add spicy condiments such as chilli sauce, salsa, mustard, pepper, curry, turmeric, cardamon, fennel seeds, cumin, coriander, Szechuan peppercorns, chutney, steak sauce, soy sauce, Worcestershire sauce, or wasabi to your food? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q43. Over the past month, how often on average did you add sweet condiments such as jam, jelly, corn or maple syrup, honey, or molasses to your food? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q44. Over the past month, how often on average did you add gravy, hollandaise, or cheese sauce (such as Velveeta cheese sauce) to your food? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never 117 DIETARY PATTERNS OF ETHNIC POPULATIONS <98> Don’t know q45. EXCLUDING garlic supplements, minced or powdered garlic, and garlic salt over the past month how often on average did you use fresh garlic in your food? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know Eighth Section: Now I would like to ask you about Fats and Oils that you use in your food. q46. Do you prefer to use ghee, butter or margarine in your food? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q47. Over the past month, did you include olive oil in your meals? <1> Yes [go to q48] <2> No [skip to q51] – EXIT olive oil questions and go to other liquid cooking oils question <98> Don't know q48. Over the past month, how often on average did you eat food that was fried or grilled with olive oil? <1> More than once per day 118 DIETARY PATTERNS OF ETHNIC POPULATIONS <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q49. Over the past month, how often on average did you eat salad dressings made with olive oil? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q50. Over the past month, how often on average did you eat other foods with olive oil, for example olive oil added to food such as pita bread, hummus, tabouli, or olive oil on top of other foods? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q51. Over the past month, how often on average did you eat food that was fried or grilled with liquid cooking oils OTHER THAN olive oil, such as canola oil, peanut oil, corn oil, safflower oil, or soybean oil but EXCLUDING “spray-on oils” such as PAM? <1> More than once per day <2> Everyday (7 days per week) 119 DIETARY PATTERNS OF ETHNIC POPULATIONS <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know q52. Over the past month, how often on average did you eat food that was fried or grilled with lard or shortening? <1> More than once per day <2> Everyday (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know We are coming towards the end of the survey. I just have a few more questions dealing with demographics, so that we can see how our sample compares to the general Canadian population. q53. Over the past five years, have you changed your diet or eating habits? <1> Yes <2> No [go to q54] <98> Don't know If Yes, how have you changed your diet or eating habits over the past five years? _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________ (open-ended question) q54. Do you consider yourself a vegetarian or vegan? 120 DIETARY PATTERNS OF ETHNIC POPULATIONS <1> Yes <2> No <98> Don't know q55. Over the past month, how often on average did you have a drink containing alcohol? <1> More than once per day <2> Every day (7 days per week) <3> 5-6 days per week <4> 2-4 days per week <5> 1 day per week <6> Less than 1 day per week (e.g., once or twice a month) <7> Never <98> Don’t know The next three questions are about leisure-time (recreational) physical activity. q56. First, which of the following best describes your physical activity pattern? <1> I am not physically active. <2> I do light activity like washing a car, walking the dog. <3> I do moderate activity, like brisk walking, dancing, skating. <4> I do heavy activity, like aerobics, jogging, or hiking <5> I do strenuous activity, like hockey, running, squash. <98> Don’t know q57. In a typical week, how often are you physically active? Would you say daily or almost every day, three to five times per week, one or two times per week, or a few times per month? <1> Daily or almost every day <2> 3 to 5 times per week <3> 1 to 2 times per week <4> A few times per month <5> Never <98> Don't know 121 DIETARY PATTERNS OF ETHNIC POPULATIONS q58. On average how many minutes/hours do you usually spend being physically active in a typical week? |_|_|_|_| in ☐ Minutes (mins) or |_|_|_|_| in ☐Hours (hrs) OR ☐ Don’t know q59. In general, how would you describe your health? Would you say excellent, good, fair, or poor? <1> Excellent <2> Good <3> Fair <4> Poor <98> Don't know q60. Do you consider yourself to be overweight, underweight, or just about right? <1> Overweight <2> Underweight <3> Just about right <98> Not sure/ Don’t know q61. Would you like to do something to lose weight? <1> Yes <2> No <98> Undecided/ Don’t know q62. What is your current marital status? <1> Married or common law or living with a partner <2> Living with a partner <3> Living alone <4> Divorced <5> Separated <6> Never married q63. What is the highest grade or level of education you have completed? ______________ highest grade or level of education 122 DIETARY PATTERNS OF ETHNIC POPULATIONS <1> No formal schooling <2> Some elementary school <3> Completed elementary school (Grade 8) <4> Some high school/junior high <5> Completed high school (Grade 12 or 13) <6> Some community or technical college <7> Completed community or technical college <8> Some university <9> Completed bachelors or undergraduate degree (arts, science, engineering, etc.) <10> Postgraduate degree: MA, MSC, MLS, MSW, etc. <11> Postgraduate degree: PhD or Doctorate <12> Professional degree: Law, Medicine, Dentistry, Engineering etc. <13> Other (Specify) ___________________ <98> Don’t know q64. Could you please tell me how much income you and other members of your household received in the year ending December 31, 2015, before taxes? Please include income FROM ALL SOURCES such as savings, investments, pensions, rent, and unemployment insurance as well as salaries and wages. We do NOT need the exact amount; could you tell me which of these broad categories it falls into… <1> Less than $10,000 <2> Between $10,000 and $19,999 <3> Between $20,000 and $29,999 <4> Between $30,000 and $39,999 <5> Between $40,000 and $49,999 <6> Between $50,000 and $59,999 <7> Between $60,000 and $69,999 <8> Between $70,000 and $79,999 <9> Between $80,000 and $89,999 <10> Between $90,000 and $99,999 <11> Between $100,000 and $150,000 <12> More than $150,000 <98> Don’t know 123 DIETARY PATTERNS OF ETHNIC POPULATIONS q65. What is the place of your birth? ________________ q66. Where were your parents born? Place of birth (mother) _______________ Place of birth (father) ________________ q67a. People living in Canada come from many different cultural and ethnic backgrounds. Are you: 1. White? 2. Chinese? 3. South Asian (e.g., East Indian, Pakistani, Sri Lankan, etc.?) 4. Black? 5. Filipino? 6. Latin American? 7. Southeast Asian (e.g., Cambodian, Indonesian, Laotian, Vietnamese, etc.)? 8. Arab? 9. West Asian (e.g., Afghan, Iranian, etc.)? 10. Japanese? 11. Korean? 12. Aboriginal Peoples of North America (North American Indian, Métis, Inuit)? 13. Other (please specify): ________________ 14. Nothing 15. Don’t know q67b. How about your parents’ cultural or ethnic background? 1. White? 2. Chinese? 3. South Asian (e.g., East Indian, Pakistani, Sri Lankan, etc.?) 4. Black? 5. Filipino? 6. Latin American? 7. Southeast Asian (e.g., Cambodian, Indonesian, Laotian, Vietnamese, etc.)? 8. Arab? 9. West Asian (e.g., Afghan, Iranian, etc.)? 10. Japanese? 11. Korean? 12. Aboriginal Peoples of North America (North American Indian, Métis, Inuit)? 13. Other (please specify): ________________ 14. Nothing 15. Don’t know 124 DIETARY PATTERNS OF ETHNIC POPULATIONS q68. How long have you been living in Canada or any other western Country? <1> Since the year of 19____ /20____ <2> Second generation immigrant (i.e. one parent or both born outside Canada) <3> Third generation immigrant or more (i.e., both parents born in Canada but at least one grandparent or more born outside Canada <97> Nothing <98> Don’t know Ninth Section (Acculturation): At the end, I would like to ask you few questions about your eating habits before and after moving to Canada from your home country. You should not answer this set of questions if you were born in Canada, moved at a very young age, or think that this set of questions does not apply to you: q69. Has your general portion size (how much you eat in one meal) changed since coming to Canada? (1- Eat much less, 3- No change, 5- Eat much more) q70. How has your method of food preparation changed since coming to Canada? - Stir-frying/BBQ (1-Much less often, 3- No change, 5-Much more often) - Baking/grilling food (1-Much less often, 3- No change, 5-Much more often) - Boiling food (1-Much less often, 3- No change, 5-Much more often) - Deep frying food (1-Much less often, 3- No change, 5-Much more often) - Microwaving food (1-Much less often, 3- No change, 5-Much more often) - Eating vegetables raw (1-Much less often, 3- No change, 5-Much more often) - Home-made foods (1-Much less often, 3- No change, 5-Much more often) q71. How have the foods you generally eat changed since coming to Canada? - Vegetables (1- Eat much less, 3- No change, 5- Eat much more) - Potatoes, rice (1- Eat much less, 3- No change, 5- Eat much more) - Fruit (1- Eat much less, 3- No change, 5- Eat much more) - Soft drinks/pop/soda (1- Eat much less, 3- No change, 5- Eat much more) - Dairy products (1- Eat much less, 3- No change, 5- Eat much more) - High fat/fried foods (1- Eat much less, 3- No change, 5- Eat much more) 125 DIETARY PATTERNS OF ETHNIC POPULATIONS - Deserts/candy/sweets (1- Eat much less, 3- No change, 5- Eat much more) - White meat (1- Eat much less, 3- No change, 5- Eat much more) - Red meat (1- Eat much less, 3- No change, 5- Eat much more) - Restaurant meals/dining out (1- Eat much less, 3- No change, 5- Eat much more) q72. Has your interest in information about the food you eat, such as ingredients, nutrition information and taking note of food labels, changed since coming to Canada? - I look at ingredients in the food I buy: (1- Much less often, 3- No change, 5- Much more often) - I put effort into making sure the food I buy has good nutritional value: (1- Much less often, 3- No change, 5- Much more often) - I read the nutritional information table on food products: (1-Much less often, 3- No change, 5-Much more often) - I understand the nutritional information table on food products: (1-Much less often, 3- No change, 5- Much more often) - I hear about which foods are good for me through the media and advertising: (1-Much less often, 3- No change, 5-Much more often) q73. How do you feel about these issues in Canada compared to your home country? - Finding fresh fruit and vegetables: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada) - Finding low fat food options: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada) - Choosing healthy food when dining out: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada) - Finding ethnic food ingredients: (1- Much harder in Canada, 3- No change, 5- Much easier in Canada) Thank you for your participation. This is the *END* of the Survey. 126 DIETARY PATTERNS OF ETHNIC POPULATIONS Appendix B Number of Missing Responses by Dietary Item Dietary item q4 q5 q6 q7 q8 q9a q9b q9c q9d q9e q10a q10b q10c q11a q11b q11c q12 q13 q14 q15 q16 q17 q18a q18b q19 q20 q21a q21b q22a q22b q23 q24 q25 q26 q27 q28 q29 q30 “do not know” response 1 2 2 2 9 2 7 8 7 7 2 3 2 2 8 5 6 5 6 5 6 9 3 6 10 2 0 5 1 6 4 10 4 1 7 0 6 5 No response Missing responses % Missing responses 3 0 2 3 5 0 1 1 1 1 2 2 1 8 2 0 3 0 0 1 2 2 4 1 2 4 3 2 1 2 2 12 2 1 5 1 1 1 4 2 4 5 14 2 8 9 8 8 4 5 3 10 10 5 9 5 6 6 8 11 7 7 12 6 3 7 2 8 6 22 6 2 12 1 7 6 1.0% 0.5% 1.0% 1.2% 3.4% 0.5% 2.0% 2.2% 2.0% 2.0% 1.0% 1.2% 0.7% 2.4% 2.4% 1.2% 2.2% 1.2% 1.5% 1.5% 2.0% 2.7% 1.7% 1.7% 2.9% 1.5% 0.7% 1.7% 0.5% 2.0% 1.5% 5.4% 1.5% 0.5% 2.9% 0.2% 1.7% 1.5% 127 DIETARY PATTERNS OF ETHNIC POPULATIONS q31 q32 q33 q34 q35a q35b q36 q37a q37b q38 q39 q40 q41a q41b q42 q43 q44 q45 q46 q48 q49 q50 q51 q52 q55 Total 5 1 4 2 8 10 11 0 15 7 7 6 2 4 4 4 7 3 10 23 14 15 18 32 4 392 1 4 1 3 3 3 2 4 5 3 1 15 4 3 3 1 2 1 0 8 4 3 18 2 7 185 128 6 5 5 5 11 13 13 4 20 10 8 21 6 7 7 5 9 4 10 31 18 18 36 34 11 577 1.5% 1.2% 1.2% 1.2% 2.7% 3.2% 3.2% 1.0% 4.9% 2.4% 2.0% 5.1% 1.5% 1.7% 1.7% 1.2% 2.2% 1.0% 2.4% 7.6% 4.4% 4.4% 8.8% 8.3% 2.7% 2.2% DIETARY PATTERNS OF ETHNIC POPULATIONS Appendix C Communality for 63 Dietary Items following PCA with Varimax Orthogonal Rotation Dietary item q4 q5 q6 q7 q8 q9a q9b q9c q9d q9e q10a q10b q10c q11a q11b q11c q12 q13 q14 q15 q16 q17 q18a q18b q19 q20 q21a q21b q22a q22b q23 q24 q25 q26 q27 q28 q29 q30 Communality 0.47 0.39 0.53 0.54 0.41 0.31 0.30 0.41 0.36 0.45 0.31 0.38 0.21 0.51 0.45 0.27 0.37 0.23 0.24 0.26 0.29 0.24 0.40 0.47 0.45 0.44 0.52 0.42 0.38 0.22 0.56 0.45 0.54 0.61 0.40 0.35 0.22 0.33 129 DIETARY PATTERNS OF ETHNIC POPULATIONS q31 q32 q33 q34 q35a q35b q36 q37a q37b q38 q39 q40 q41a q41b q42 q43 q44 q45 q46 q48 q49 q50 q51 q52 q55 0.36 0.20 0.41 0.41 0.28 0.33 0.41 0.23 0.35 0.44 0.31 0.11 0.37 0.35 0.39 0.31 0.47 0.26 0.39 0.45 0.54 0.42 0.85 0.41 0.85 130 ',(7$5<3$77(5162)(7+1,&3238/$7,216 $SSHQGL[ ' 81%&5HVHDUFK(WKLFV%RDUG$SSURYDO /HWWHU