Relationship between diet and acculturation among South Asian children living in Canada

Journal Pre-proof Relationship between diet and acculturation among South Asian children living in Canada Salmi Noor, Mahshid Dehghan, Scott A. Lear, Sumathi Swaminathan, Quazi Ibrahim, Sumathy Rangarajan, Zubin Punthakee PII:

S0195-6663(18)31871-3

DOI:

https://doi.org/10.1016/j.appet.2019.104524

Reference:

APPET 104524

To appear in:

Appetite

Received Date: 31 December 2018 Revised Date:

18 October 2019

Accepted Date: 18 November 2019

Please cite this article as: Noor S., Dehghan M., Lear S.A., Swaminathan S., Ibrahim Q., Rangarajan S. & Punthakee Z., Relationship between diet and acculturation among South Asian children living in Canada, Appetite (2019), doi: https://doi.org/10.1016/j.appet.2019.104524. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

1

TRACKED COPY FOLLOWED BY CLEAN COPY

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

RELATIONSHIP BETWEEN DIET AND ACCULTURATION AMONG SOUTH ASIAN CHILDREN LIVING IN CANADA Salmi Noor1, Mahshid Dehghan2, Scott A Lear3, Sumathi Swaminathan4, Quazi Ibrahim2, Sumathy Rangarajan2, Zubin Punthakee1,2 1 McMaster University, Hamilton, ON, Canada 2 Population Health Research Institute, Hamilton, ON, Canada 3 Faculty of Health Sciences, Simon Fraser University, Vancouver, BC, Canada 4 St. John’s Research Institute, Bangalore, Karnataka, India

Corresponding Author: Zubin Punthakee 1280 Main St. W., HSC 3V51 Hamilton, ON, Canada L8S 4K1 [email protected] Word count: 250 (abstract); 4170 (body). Tables: 4. Figures: 1. References: 39. Keywords: South Asian, acculturation, diet, children, youth Funding: This study was funded by joint grants from the Canadian Institutes of Health Research (FRN 227861) (SAL ZP) and the Indian Council for Medical Research www.icmr.nic.in (SS). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Contributions: SN, MD, SAL, SS, ZP contributed to study design. SAL, SS, SR, ZP contributed to data collection. QI conducted data analysis. SN, MD, SL, ZP contributed to data interpretation. SN and ZP wrote the first draft of the manuscript and MD, SAL, SS, QI, SR provided critical review. All authors approved the final manuscript. Declarations: The authors have declared that no competing interests exist. Abbreviations: CVD cardiovascular disease, FFQ food frequency questionnaire, LoR length of residency, SA South Asian(s)

1

40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Abstract INTRODUCTION Diet and South Asian ethnicity are both associated with early onset of cardiovascular risk factors. Among youth of South Asian origin, little is known about the role of culture in determining healthy dietary patterns. We aimed to assess dietary patterns and their relationships with acculturation to Western and traditional lifestyles among South Asian youth in Canada. METHODS The Research in Cardiovascular Health - Lifestyles, Environments and Genetic Attributes in Children and Youth (RICH LEGACY) study targeted South Asian children and adolescents aged 7-8 and 14-15 years in two Canadian cities. In this cross-sectional study, acculturation questionnaires and food frequency questionnaires were administered to assess the correlations between Western and traditional culture scores, immigration status (generation and length of residency) in Canada and intake frequency of various foods. RESULTS Among 759 youth, those who ate fruits and vegetables more often consumed dairy and whole grains more often (all r=0.17-0.22, all p<0.001), while those who ate fast food more often consumed meat, sweets and sugared drinks more often (all r 0.24-0.38, all p<0.001). Traditional culture scores were weakly positively correlated with whole grain intake frequency (r=0.12, p=0.001), and negatively with meat intake frequency (r=-0.14, p<0.001). Western culture scores positively correlated with high intake frequency of meat (r=0.23, p<0.001), fast food (r=0.14, p<0.001) and sweets (r=0.14, p<0.001). DISCUSSION Children who are more acculturated with Western lifestyle consumed foods associated with increased metabolic risk. However, whether this eating pattern translates into increased risk of obesity and cardiovascular diseases needs to be further explored.

2

66

INTRODUCTION

67

The Western diet (high in animal products, refined oils, refined grains, added sugars, and

68

processed foods), has been linked to increased cardiovascular disease (CVD) risk factors such

69

as blood lipoprotein, glucose levels and obesity in adults (Cordain et al., 2005) and children

70

(Anderson & Butcher, 2006; "Facts and figures on childhood obesity," 2016) which in turn

71

increase the risk of early onset of type 2 diabetes and CVD (Yach, Stuckler, & Brownell, 2006).

72

For people who immigrate to Western countries, diet is substantially impacted due to exposure

73

to a new food environment, thereby potentially leading to changes in food procurement and

74

preparation (Pan, Dixon, Himburg, & Huffman, 1999; Satia-Abouta, Patterson, Neuhouser, &

75

Elder, 2002). Furthermore, Western acculturation (the incorporation by immigrants of their host

76

country’s culture into their own lifestyle) has been linked to unhealthy eating patterns such as an

77

increase in fast food consumption among youth in several ethnic groups whose traditional diets

78

may be considered more ‘healthy’ (Unger et al., 2004).

79 80

On the contrary, South Asians (SA) have a traditional diet that is high in fat and carbohydrates

81

and is associated with CVD risk factors (Misra, Khurana, Isharwal, & Bhardwaj, 2009). Even SA

82

vegetarian diets which are prevalent for religious reasons do not seem to confer as much

83

benefit on CVD risk factors as vegetarian diets in the US (Jaacks et al., 2016). These are

84

important as CVD starts at younger ages in SA than other ethnic groups (Gupta, Singh, &

85

Verma, 2006). Children and adolescents in India (the largest South Asian country) consume

86

high amounts of cereals and pulses, fats/oils, fast foods and sugared drinks, and inadequate

87

amounts of fruits, vegetables and dairy (Rani & Sathiyasekaran, 2013; Vecchio et al., 2014).

88

Perhaps because of this traditional dietary pattern, data regarding the effects of acculturation on

89

SA adults’ dietary patterns are conflicted (Khan, Jackson, & Momen, 2016; Lesser, Gasevic, &

90

Lear, 2014; Talegawkar, Kandula, Gadgil, Desai, & Kanaya, 2016; Wandel, Raberg, Kumar, &

91

Holmboe-Ottesen, 2008), and data regarding SA children are quite minimal. One study of SA 3

92

children from the United States found excessive fat intake with low intake of fruits, vegetables,

93

meat, fish, beans and eggs overall; however, lower intake of sweets was associated with

94

greater length of residency (Martyn-Nemeth et al., 2017) suggesting some benefits and some

95

harms. In this context, immigration-related changes in the diet quality of youth may depend on

96

the difference in quality between a particular traditional diet and the Western diet, the amount of

97

passive exposure to traditional versus Western food environments and culture, and active

98

participation in traditional versus Western culture. Hence there is a need to understand more

99

about the unique eating patterns of SA youth in other Western countries, and how time in a host

100

Western country and acculturation may improve or worsen dietary patterns, in order to target

101

appropriate public health messaging to this high CVD risk group at an early age. This study

102

aimed to confirm the hypotheses that SA children and adolescents living in Canada have

103

patterns of either predominantly healthy or predominantly unhealthy eating, and that Traditional

104

or Western cultural preferences will distinguish those with greater intake frequency of healthy or

105

unhealthy food items, respectively.

106 107

METHODS

108

Study Design

109

This study was approved by the Hamilton Integrated Research Ethics Board and the Simon

110

Fraser University Research Ethics Board, and parents and participants provided written

111

informed consent and assent, respectively, after having a chance to read study information and

112

ask questions of the researchers in English, Hindi, Punjabi and Urdu. The RICH LEGACY

113

(Research in Cardiovascular Health-Lifestyles, Environments and Genetic Attributes in Children

114

and Youth) study is a cross-sectional study designed to evaluate the determinants of CVD risk

115

factors among SA children in Canada. The study aims to identify differences in body

116

composition, metabolic consequences of adiposity as well as associated behavioural, attitudinal,

4

117

social and environmental determinants of CVD risk factors. The current report focuses on the

118

relationship between acculturation and immigration status and patterns in dietary intake.

119 120

Participants

121

Recruitment targeted SA children (those with at least 3 of 4 grandparents who had ancestral

122

origins in India, Pakistan, Bangladesh or Sri Lanka (Canada, 2008)) aged 7-8 years in

123

elementary schools (elementary school age group) and 14-15 years in high schools (high school

124

age group) from two distinct urban centres in Canada with the largest SA populations (Canada,

125

2008): Surrey, British Columbia and Brampton, Ontario from 2012 to 2016. These age groups

126

were chosen to represent social environments and psycho-behavioural patterns primarily

127

influenced by parents versus peers and self (Davison & Jago, 2009). For practical purposes, the

128

7-8 year old group was also considered to be the youngest group that could reliably contribute

129

information about their preferences and health behaviours (e.g., dietary reporting by children as

130

young as 8 years is as accurate as parent reports (Burrows et al., 2013; Livingstone, Robson, &

131

Wallace, 2004; Lytle et al., 1993)). There were no other inclusion/exclusion criteria. Individuals

132

were enrolled from a convenience sample of public schools, temples and community centres.

133

Recruitment material and information sheets described the search for relationships of lifestyles

134

and environments with cardiovascular risk factors among youth in general terms without any

135

mention of ethnicity (due ethical concerns by the school boards) or specific dietary hypotheses.

136

Information packages with invitations to contact the study team were dropped off at the schools

137

and distributed at the discretion of principals and teachers to students in grades 2, 3, 9 and 10,

138

and for privacy reasons, investigators were not able to determine the number of eligible

139

candidates to whom invitations were given. Participants from temples and community centres

140

approached research staff at booths during health fairs and public events. As part of the main

141

study, questionnaires were administered to collect demographics, parental education and

142

acculturation and dietary information. The elementary school age children completed the 5

143

questionnaires by interview with a researcher in the presence of a parent who could assist if the

144

child seemed unsure or requested help. High school age youth completed questionnaires by

145

interview, or self-administration in a classroom setting as per individual school permissions.

146 147

Acculturation

148

The Acculturation Rating Scale for Mexican Americans (ARSMA-II) modified to include South

149

Asian references and examples was administered due to its previous use and validity among

150

Indian adolescents (Stigler et al., 2010), and its ease of understanding for children (i.e. it does

151

not include concepts such as humour and values as some other acculturation scales do), and

152

has been validated in children as young as 9 years old (Lopez, 2009). It independently

153

measures the degree to which individuals prefer Western cultural practices, and cultural

154

practices of their native country. Three items in each of four domains (language spoken with:

155

parents, siblings, and friends; media: TV shows, movies, and music; food: lunch/dinner, snacks,

156

and dessert; and consumer goods: clothing, restaurants, and stores/markets) were rated twice,

157

once for Western preferences (Western score) and once for traditional preferences (Traditional

158

score). For each item, participants answered “never, “sometimes”, ‘often” or “very often/always”

159

corresponding to scores of 1, 2, 3, or 4 respectively(Stigler et al., 2010). For participants with no

160

siblings, the question about language spoken with siblings was imputed as the mean score of

161

language spoken with parents and with friends. Western and Traditional scales each had a

162

minimum score of 12 and maximum score of 48, with higher Western and Traditional scores

163

indicating greater degrees of Western and Traditional cultural practices, respectively. Two

164

measures of immigration status were assessed: generation status and for 1st generation

165

participants, length of residency (LoR) in Canada. For generation status, participants were

166

categorized based on Statistics Canada definitions (1st generation - born outside Canada; 2nd

167

generation - born in Canada with at least 1 parent born outside Canada; 3rd generation or more -

168

born and both parents born in Canada) (S. Canada, 2016). In addition, LoR for 1st generation 6

169

participants was calculated by subtracting age at immigration from current age (Koya & Egede,

170

2007).

171 172

Diet

173

A short food frequency questionnaire (FFQ) was used to ask how many times participants

174

habitually ate foods from a list of 25 items per day, per week or per month, or if less than once a

175

month, with no specified time frame for the recall (See supplementary appendix). The

176

questionnaire was adapted (by adding foods and drinks more commonly consumed by children)

177

from the INTERHEART FFQ which, among adults, was able to identify dietary patterns

178

associated with myocardial infarction in multiple countries including India and Canada (Iqbal et

179

al., 2008). Its validity has not been assessed in children or adolescents. Frequency of intake per

180

day of each food item was calculated. We excluded FFQs if responses to ≥5 food items were

181

incomplete or missing, or all 25 food groups were marked with identical frequency, or mean

182

intake per day for any item was >10, or estimated total daily intake (sum of all 25 mean daily

183

intakes) was ≤2. A total of three FFQs were excluded.

184 185

Foods were grouped as whole grains (whole grains/breads/cereals); refined grains (refined

186

breads/rice/pasta and cereals); meats (meat, organ meats, poultry, eggs, fish and seafood);

187

dairy (dairy products); fruits and vegetables (fruits, leafy green vegetables, raw vegetables,

188

other cooked vegetables); fast food (pizza, fast foods, pickled foods, deep fried foods, salty

189

snacks); sweets (ice cream and pudding, dessert/sweet snacks, confectionery, sugars and

190

syrups); and sugared drinks (fruit juice and sugared drinks, non-diet carbonated beverages).

191

Legumes/pulses/nuts/seeds, potatoes (boiled and mashed), and diet carbonated beverages

192

were additional items captured in the FFQ. For each food item, both Western and South Asian

193

examples were provided as prompts (e.g. dairy: milk, yogurt, lassi, raita, …; fast food: chicken

194

nuggets, French fries, pakoras, chaat, …; desserts: cakes, pies, burfi/ladoo, rasgulla/gulab 7

195

jamun, …). The above groupings of food items were determined by using Canada’s Food Guide

196

as a reference (H. Canada, 2016). Food items that were not included in the food guide (e.g.

197

pizza and other fast foods) were grouped by consensus among the authors. Participants who

198

had not eaten any meat, organ meat, poultry or fish in the last month were classified as

199

vegetarians. Lifelong vegetarians were not distinguished from those abstaining from meat for

200

only a month.

201 202

Statistical Analysis

203

Participant demographic characteristics, acculturation and immigration status, and dietary intake

204

frequency of specific food groups (calculated as number of times per day) were described (n

205

(%) or mean ± SD) for the overall study population and across tertiles of Traditional culture

206

scores and Western culture scores. Dietary intake frequency was also presented by 1st and 2nd

207

or 3rd generation participants and across LoR tertiles.

208 209

For descriptive purposes, demographic characteristics were compared across tertiles of

210

Traditional and Western culture scores using the Cochran-Armitage test for trend. Relationships

211

among culture scores and immigration measures were described using partial Pearson

212

correlations with t-statistics for significance (Kleinbaum, Kupper, Nizam, & Muller, 2008; "Partial

213

Correlation," 2013), and where generation status was included, using partial Spearman

214

correlations (Kleinbaum et al., 2008; "Partial Correlation," 2013), all adjusted for age, sex, site of

215

enrolment (Surrey or Brampton) and maternal and paternal education (post-secondary or not).

216

In order to test for relationships between frequency of intakes of the pre-specified food groups, a

217

Pearson correlation matrix was constructed. Post hoc, Fisher r to z transformation analysis

218

(Kenny, 1979) was done to identify differences between vegetarians and non-vegetarians in

219

correlation coefficients among the non-meat containing food groups.

220 8

221

In order to test for relationships between the acculturation or immigration measures and

222

frequency of intake of the pre-specified food groups, partial Pearson correlation, or where

223

generation status was included, partial Spearman correlation were used, all adjusted for age,

224

sex, site of enrolment, and maternal and paternal education. Post hoc, differences between the

225

age groups in the relationship between culture scores and frequency of food intake were

226

assessed by an interaction term for age group and Traditional or Western culture score in a

227

linear regression model adjusted for the covariates described above.

228 229

All analyses were conducted using SAS software version 9.4 (SAS institute). For each set of

230

correlations, the Bonferroni correction was used to account for multiple testing such that each

231

correlation was considered statistically significant at p<0.007 in table 2, p<0.002 in tables 3 and

232

4, and p<0.002 in figure 1. Interaction analyses in table 3 were post hoc and exploratory; no

233

correction for multiple testing was considered.

234 235

RESULTS

236

762 participants were enrolled, with a participation rate of approximately 32% among individuals

237

with confirmed contact about the study. Reported South Asian countries of origin were India

238

(n=669, 88%), Pakistan (n=72, 9%), Sri Lanka (n=36, 5%), and Bangladesh (n=6, 1%)

239

(participants could report more than one). Culture questionnaires and valid FFQs were

240

completed by 759 participants, and 715 of them (94%) did so during the school year from

241

September to June. Participant characteristics are shown in Table 1. A high proportion of

242

participants had parents with post-secondary education. Certain demographic factors were

243

progressively associated with cultural preferences. Participants from Surrey and participants

244

whose fathers did not attain post-secondary education were more likely to have higher

245

Traditional culture scores (p=0.035 and p=0.011, respectively). Participants who were older and

246

participants whose parents attained post-secondary education were more likely to have higher 9

247

Western culture scores (p<0.001 for all), Vegetarians were more likely to have higher Traditional

248

culture scores (p<0.001) and lower Western culture scores (p<0.001) than non-vegetarians.

249 250

Culture preference patterns are shown in Table 2. Participants had higher Western culture

251

scores than Traditional culture scores, and the majority were 2nd or 3rd generation Canadians

252

(only 5 participants were 3rd generation). After adjusting for age, sex, site of enrolment and

253

maternal and paternal education, Traditional and Western culture scores were inversely

254

associated with each other (partial r=-0.23, p<0.001). After adjusting for the same variables,

255

Traditional culture scores were inversely associated with generation status (partial r=-0.12,

256

p=0.002), and LoR among 1st generation Canadians (partial r=-0.22, p=0.005), while Western

257

culture scores were positively associated with generation status (partial r=0.14, p<0.001).

258 259

Fast foods, sweets and sugared drinks were each consumed once daily, on average (Table 3).

260

The frequency of dietary intake correlation matrix is presented in Figure 1 and shows two

261

groupings of correlated foods. Intake of fruits and vegetables, dairy and whole grains were all

262

positively correlated with one another (p<0.001 for all). The second grouping consisting of fast

263

foods, sweets and sugared drinks were also all positively correlated with one another (p<0.001

264

for all). We found positive correlations between meat intake and fast food intake frequencies

265

(r=0.24, p<0.001), and meat intake and refined grains intake frequencies (r=0.14, P<0.001).

266

There was an inverse correlation between refined grain and whole grain intake frequencies (r=-

267

0.23, p<0.001). We did not find differences between vegetarians and non-vegetarians in these

268

significant correlations (all differences p>0.05 by Fisher r to z transformation), but we found one

269

marginal difference in the non-significant correlations between fruits and vegetables and sweets

270

(r=-0.15, p=0.04 for vegetarians, r=0.04, p=0.29 for non-vegetarians; p for difference=0.02).

271

10

272

The relationships between acculturation and dietary patterns are shown in Table 3. Participants

273

with higher Traditional culture scores consumed whole grains more frequently (partial r=0.12,

274

p=0.001), and had meat less frequently (partial r=-0.14, p<0.001), and there were no differences

275

among the age groups. Those with higher Western culture scores consumed sweets (partial

276

r=0.14, p<0.001), fast foods (partial r=0.14, p<0.001) and meats (partial r=0.23, p<0.001) more

277

often. The high school age group demonstrated stronger relationships than the elementary

278

school age group between Western culture scores and intake frequency of sweets (p-

279

interaction=0.003) and fast foods (p-interaction<0.001), but not meats (p=0.29). We did not find

280

significant relationships between generation status or LoR and dietary patterns (Table 4).

281 282

11

283

DISCUSSION

284

We found that among SA children and youth living in Canada there are significant correlations

285

between intake frequencies of distinct food groups that are differentially associated with cultural

286

preferences. Overall, children and youth in our study consumed fast foods, sweets and sugared

287

drinks on a daily basis. We observed two distinct groupings of correlated food intake frequency:

288

1) whole grains, dairy, and fruits and vegetables, and 2) meat, fast foods, sweets, and sugared

289

drinks. Traditional cultural preferences were associated with the first grouping and Western

290

cultural preferences with the second grouping.

291 292

In the population studied, there were correlations among intake frequency of foods associated

293

with low levels of cardiometabolic risk (“healthy foods”: whole grains, dairy and fruits and

294

vegetables) (Funtikova, Navarro, Bawaked, Fito, & Schroder, 2015) and among intake

295

frequency of foods associated with high levels of cardiometabolic risk (“unhealthy foods”: meat,

296

fast foods, sweets and sugared drinks) (Funtikova et al., 2015; Karatzi et al., 2014). Though

297

these healthy and unhealthy patterns were not mutually exclusive (demonstrated by a paucity of

298

inversely correlated foods), they suggest that those who make healthy or unhealthy food

299

choices tend to do so in multiple domains within that grouping, possibly in the form of meals

300

composed of predominantly either healthy or unhealthy foods (Evans, Cleghorn, Greenwood, &

301

Cade, 2010). This study was not able to determine the extent to which this grouping of food

302

types is driven by individual preferences and behaviours or the availability of grouped foods in

303

the food environment, but both have been shown to play a role in healthy and unhealthy eating

304

(Trapp et al., 2015).

305 306

While correlations did occur within healthy or unhealthy groupings, it remains to be investigated

307

whether promoting individual healthy foods may lead to an overall healthier dietary pattern, or

308

discouraging individual unhealthy foods may reduce overall unhealthy eating patterns. For 12

309

example, among vegetarians in our study, positive correlations between fast foods, sweets and

310

sugared drinks were noted. Other studies have shown that adult SA vegetarians living in their

311

native country eat more fried food and dessert than non-vegetarians (Jaacks et al., 2016) (both

312

of which have been associated with higher levels of cardiometabolic risk), and our study

313

extends these findings to SA youth living in a Western country. These observations highlight

314

that restrictive diets such as vegetarianism may not eliminate the challenges of unhealthy

315

eating.

316 317

Except for whole grains and refined grains there were no significant inverse correlations

318

between food groups (Figure 1). These findings suggest SA youth may substitute healthy foods

319

for unhealthy foods or vice versa within a category (i.e. grains), but not between categories (e.g.

320

fruits and vegetables for fast food). Therefore, public health interventions that focus on a single

321

category (e.g. eat more fruits and vegetables) may not be sufficient if they do not also suggest

322

from which categories to substitute (e.g. eat fruits and vegetables instead of fast foods), and

323

may additionally benefit from emphasizing healthy choices within food categories, taking into

324

consideration the community’s food preferences.

325 326

Acculturation had a significant association with frequency of food group intakes. SA youth with

327

greater Western culture preferences more frequently consumed foods associated with

328

increased cardiometabolic risk, such as meats, fast foods and sweets, while those identifying

329

more strongly with traditional culture consumed meat less often and whole grains more often.

330

This may be due to greater adherence to vegetarianism in those who retain traditionalism, and

331

greater use of raw grains than processed carbohydrates in traditional food preparation. This

332

suggests that the traditional diet of SA youth is not inherently unhealthy (especially when

333

compared to a Western diet), and dietary patterns experience a shift towards unhealthy food

334

items associated with greater Western acculturation. This is consistent with research on Asian 13

335

American and Hispanic adolescents showing that those who identify more with Western culture

336

eat more non-traditional and high fat foods (Unger et al., 2004). Our results are contrary to a

337

small study of SA adolescents showing more time in the US was associated with lower intake of

338

sweets, a potentially healthy dietary effect of acculturation (Martyn-Nemeth et al., 2017). While

339

this heterogeneity is difficult to explain, our study adds support to a consistent effect of

340

acculturation on diet among youth regardless of ethnicity. Furthermore, in our study, time

341

measures (LoR and generation status) were not associated with diet patterns, suggesting that

342

children and youth do not experience the dietary impacts of culture through passive exposure,

343

but by taking on certain beliefs, attitudes and behaviours which may happen non-linearly with

344

time.

345 346

Diets of SA often change after migration (Lesser et al., 2014), especially those of younger

347

adults, compared to older adult SA immigrants who are less likely to change their dietary

348

patterns (Wandel et al., 2008). Whether this age trend extends into youth is unknown, though

349

evidence in other settings suggests migrant youth may undergo dietary acculturation to a

350

greater extent than their parents (Wilson & Renzaho, 2015). Between elementary school age

351

children and high school adolescents, we found a greater preference for Western culture among

352

adolescents, and stronger relationships between Western culture and intake frequencies of fast

353

foods and sweets among adolescents. This, coupled with adolescents being highly susceptible

354

to the availability of fast food and convenience food in the environment (He et al., 2012), may

355

make adolescence a particularly vulnerable period for unhealthy eating as SA grow up in

356

Western countries.

357 358

Western food and Western diet are not synonymous. In Western countries, there is an

359

abundance of food with wide-ranging nutritional value. Our findings, and those of others (Sanou

360

et al., 2014) indicate that Westernization of the diet is due to acceptance of and participation in 14

361

a culture that values easily available, accessible, affordable and convenient foods that happen

362

to be energy dense and nutrient poor. Thus, for immigrants who maintain their preference for

363

traditional food types, their diet may nevertheless be influenced towards more restaurant foods,

364

packaged foods and larger portions because of overall acculturation, and not specifically dietary

365

acculturation.

366 367

This study had several notable strengths. It included a large number of participants with broad

368

variation in age, generational status and acculturation enrolled from several schools and

369

temples in diverse neighbourhoods in two cities, although it was not a random sample. Thus,

370

the results are likely generalizable to a broad spectrum of South Asian youth from elementary to

371

high school living in Western countries with similar immigration policies and in urban centres

372

where the South Asian population is large enough to maintain its own culture in the community.

373

In addition, stringent criteria for statistical significance were used to minimize spurious

374

inferences. This study also had some limitations. A short FFQ to estimate habitual dietary intake

375

of participants was used in this study. The short FFQ provided detailed information on intake

376

frequency of different foods, but not enough detail to estimate daily energy, macro- and

377

micronutrient intakes. Furthermore, it did not allow for identification of traditional cultural or non-

378

traditional foods or methods of food preparation, and would not have been able to determine if

379

Western acculturation was associated with eating more pakoras and chaat, or hot dogs and

380

French fries. However, FFQ is the most common method used by epidemiological studies. In

381

addition, the Traditional scale of the acculturation questionnaire referred to the individual’s

382

native country and may not have adequately captured traditional culture of second and third

383

generation SA Canadians. This is confounded by the fact that many South Asian countries have

384

a colonial past which makes their native country’s recent culture somewhat Westernized to

385

begin with, and this could not be measured. In addition, we aggregated acculturation across

386

many domains and did not specifically assess dietary acculturation in isolation. Furthermore, 15

387

this study is limited by providing a cross-sectional measure of acculturation which is a constantly

388

evolving process that varies across time (Stigler et al., 2010) in a non-linear way. Finally, the

389

post-hoc age subgroup analyses had less than 80% power to detect Pearson correlations

390

smaller than 0.12 in the elementary school age group and 0.22 in the high school age group.

391 392 393

CONCLUSION

394

With the increase in SA immigrants in Canada, it is important to ensure they are able to lead a

395

healthy lifestyle in their new host country starting from a young age. Among SA children and

396

youth, intake frequency of fruits and vegetables, dairy and whole grains correlated, as did intake

397

frequency of fast foods, sweets, sugared drinks, and among non-vegetarians, meat, indicating a

398

healthy grouping and an unhealthy grouping of food preferences. Those identifying more with

399

Western culture resorted to unhealthier eating practices such as more frequent intake of fast

400

foods and sweets whereas individuals identifying with traditional culture ate whole grains more

401

often and meat less often. These findings largely suggest a link between Western acculturation

402

and dietary practices associated with childhood obesity. Post hoc analyses also suggest these

403

patterns exist for both non-vegetarians and vegetarians, and are more of a concern for

404

adolescents than younger children of SA origin living in Canada. As such, interventions should

405

focus on educating parents, young children and increasingly independent adolescents (both

406

vegetarian and non-vegetarian) to increase intake of unprocessed foods and reduce intake of

407

energy dense foods. One approach to achieving this may be for community leaders such as

408

religious leaders and community resources such as school boards and public health offices in

409

regions with large SA populations to use both scientific knowledge and traditional knowledge to

410

promote traditional cultural practices among immigrants of SA origin. Future research should

411

focus on the best way of achieving this behaviour change and determining whether it impacts

412

cardiovascular risk factors. 16

413

Table 1. Characteristics of participants Traditional culture score

Western culture score

Characteristics

All

T1 16-25

T2 25.5-28

T3 28.5-44

n

759

299

218

242

Gender

p for trend*

T1 20-32

T2 32.5-36

T3 37-48

251

266

242

0.13

0.23

Females

401 (53)

146 (49)

121 (56)

134 (55)

121 (48)

150 (56)

130 (54)

Males

358 (47)

153 (51)

97 (44)

108 (45)

130 (52)

116 (44)

112 (46)

Age group

0.073

<0.001

Elementary

588 (78)

234 (78)

181 (83)

173 (72)

218 (87)

222 (84)

148 (61)

High school

171 (22)

65 (22)

37 (17)

69 (28)

33 (13)

44 (16)

94 (39)

Site

p for trend*

0.035

0.39

Brampton

479 (63)

197 (66)

144 (66)

138 (57)

141 (56)

193 (73)

145 (60)

Surrey

280 (37)

102 (34)

74 (34)

104 (43)

110 (44)

73 (27)

97 (40)

509 (70)

205 (72)

141 (67)

163 (71)

0.65

152 (62)

183 (72)

174 (78)

<0.001

464 (66)

195 (72)

131 (63)

138 (61)

0.011

142 (58)

166 (66)

156 (74)

<0.001

Yes

179 (24)

47 (16)

61 (28)

71 (29)

79 (32)

59 (22)

41 (17)

No

580 (76)

252 (84)

157 (72)

171 (71)

172 (68)

207 (78)

201 (83)

Mothers with postsecondary education Fathers with postsecondary education Vegetarian

414 415 416

<0.001

Values are n (%). * p for trend by Cochran Armitage test.

417

17

<0.001

418 419

Table 2. Relationship between acculturation and immigration measures Traditional culture score All

n Traditional culture score Western culture score

Western culture score

T1 16-25

T2 25.5-28

T3 28.5-44

759

299

218

242

26.7 ±4.4

22.5 ±2.2

27.0 ±0.8

31.7 ±3.0

-

34.8 ±4.7

36.2 ±4.3

33.5 ±3.8

34.2 ±5.4

Generation*

T1 20-32

T2 32.5-36

T3 37-48

251

266

242

-

28.0 ±3.5

26.4 ±3.5

-0.23

<0.001

29.9 ±2.2

34.4 ±1.2

-0.12

0.002

r**

p**

r**

p**

25.7 ±5.7

-0.23

<0.001

40.2 ±2.9

-

-

0.14

<0.001

0.14

0.09

1st

180 (24)

62 (21)

45 (21)

73 (30)

67 (27)

59 (22)

54 (22)

2nd & 3rd

579 (76)

237 (79)

173 (79)

169 (70)

184 (73)

207 (78)

188 (78)

5.8 ±3.5

6.3 ±3.5

5.8 ±3.5

5.3 ±3.4

5.2 ±3.4

5.4 ±2.9

6.9 ±3.9

LoR for 1st generation (y)

420 421 422 423 424 425 426

-0.22

0.005

Values are n (%), mean ±SD, Pearson or *Spearman partial r correlation coefficients or partial correlation p-values. **Adjusted for age, sex, site of enrolment (Surrey or Brampton) and maternal and paternal education (post-secondary or not). Bold text indicates significant correlations (Bonferroni p<0.007).

427

18

428 429 430

Table 3: Relationship between food intake frequency and acculturation, overall and by age group All Age group

All n

Traditional culture score T1 16-25

759

299

T2 25.5-28 218

T3 28.5-44

Western culture score

r*

p*

p* for age group difference

242

T1 20-32 251

T2 32.5-36 266

T3 37-48

r*

p*

p* for age group difference

242

Elem.

588

234

181

173

218

222

148

High

171

65

37

69

33

44

94

1.3±1.1

1.1±0.9

1.4±1.0

1.4±1.2

1.4±1.0

1.3±1.0

1.3±1.1

Daily frequency: All Whole grains

1.3±1.0

1.1±0.9

1.4±1.0

1.3±1.0

0.10

0.02

High

1.5±1.3

1.2±1.0

1.6±1.1

1.7±1.5

0.15

0.10

2.0±1.1

2.0±1.0

2.2±1.1

1.9±1.1

-0.06

2.0±1.0

2.1±1.0

2.1±1.0

1.8±1.0

-0.09

0.03

High

2.0±1.5

1.9±1.2

2.5±1.7

1.9±1.5

0.04

0.67

3.2±1.9

3.1±2.1

3.2±1.4

3.3±1.9

0.10

3.0±1.4

2.9±1.5

3.0±1.3

3.0±1.4

0.08

0.06

High

3.8±2.8

3.8±3.4

3.8±1.8

3.8±2.7

0.03

0.72

All 1.0±0.9 1.0±0.9 1.1±0.8 1.0±0.9 0.02 0.49 Sugared Elem. 1.1±0.8 1.1±0.9 1.1±0.8 1.0±0.7 -0.02 0.60 drinks High 1.0±1.1 0.8±0.9 1.1±1.1 1.1±1.3 0.11 0.22

Sweets

431 432 433 434 435 436

1.2±1.1

-0.04

1.2±1.0

1.2±0.8

1.1±0.8

-0.06

0.17

1.3±1.8

1.3±2.3

0.9±0.9

1.5±1.6

-0.02

0.85

0.9±0.8

0.9±0.6

0.8±0.5

1.0±1.1

0.07

0.9±0.6

0.9±0.5

0.8±0.6

0.9±0.6

0.01

0.87

High

1.0±1.3

0.9±0.8

0.8±0.5

1.3±1.8

0.14

0.13

0.9±1.0

0.6±0.6

0.7±1.0

Elem.

0.7±0.7

0.9±0.7

0.6±0.6

0.6±0.6

High

1.0±1.4

1.2±1.6

0.7±0.8

1.0±1.5

0.7±0.7

0.7±0.7

0.6±0.6

0.7±0.7

-0.14

0.00 -0.05

0.10

0.97

0.96

0.6±0.6

0.7±0.7

0.6±0.6

0.6±0.6

-0.07

0.10

High

0.7±0.7

0.7±0.7

0.6±0.7

0.8±0.8

0.10

0.30

0.01

1.6±1.1

1.4±1.4

1.5±1.2

-0.12

0.20

0.03

0.06

2.0±1.0

0.02

0.63

1.8±1.3

2.0±1.5

0.14

0.12

3.2±1.7

3.4±2.3

0.02

3.2±1.5

3.0±1.5

0.03

0.48

3.7±2.3

3.4±2.6

4.0±3.1

0.03

0.78

1.0±0.9

0.04

1.1±0.8

1.1±0.8

0.01

0.79

1.0±1.0

1.1±1.2

1.0±1.1

0.04

0.63

1.3±1.8

0.14

1.2±0.7

1.3±1.3

0.09

0.04

1.0±0.9

1.2±1.0

1.4±2.3

0.22

0.02

1.0±1.1

0.14

0.9±0.6

0.9±0.5

0.04

0.33

0.7±0.4

1.0±1.0

1.2±1.5

0.29

0.001

1.0±1.2

0.23

0.8±0.7

0.9±0.7

0.22 <0.001

0.5±0.8

0.8±0.8

1.2±1.7

0.26

0.8±0.8

0.07

0.005

0.29

0.07

0.6±0.6

0.6±0.6

0.7±0.8

0.08

0.06

0.6±0.7

0.6±0.6

0.8±0.8

0.08

0.36

Elem.- elementary school age, High - high school age. Values are mean ±SD, Pearson partial r and partial correlation p-values. *Adjusted for age, sex, site of enrolment (Surrey or Brampton) and maternal and paternal education (post-secondary or not). Bolded text indicates significant correlations for main effects (Bonferroni p<0.002).

19

<0.001

<0.001

0.6±0.6

0.6±0.6

0.003

<0.001

0.9±0.6

0.8±0.7

0.58

<0.001

1.1±0.7

0.9±0.7

0.80

0.37

1.0±0.7

1.2±0.8

0.11

0.57

2.7±1.3

1.1±0.9

0.96

0.14

2.1±0.9

0.6±0.6 0.14

2.0±1.2

2.4±1.6

0.5±0.6 0.12

2.0±1.0

2.0±1.0

0.8±0.5

0.23

Elem.

-0.10

1.1±0.7

<0.001

-0.19 <0.001

1.1±1.0

1.0±0.8

0.07

Elem.

0.9±0.9

0.87

0.008

1.3±0.9

2.9±1.5

0.31

1.2±0.9

All Refined grains

1.1±0.8

High

All Meats

1.2±1.4

Elem.

All Fast foods

1.2±1.2

0.12

-0.10

1.3±1.0

2.0±1.1

0.008

Elem.

All

0.72

0.15

Elem.

All Fruit & veg

0.001

Elem.

All Dairy

0.12

0.99

437 438

Table 4: Relationship between food intake frequency and immigration status st

Generation 1 n

st

2

nd

&3

rd

Length of Residency (y) among 1 generation Spearman r*

p*

T1 0.17-3.81

T2 3.82-6.79

T3 6.80-14.97

55

55

56

Pearson r*

p*

180

579

Whole grains

1.5 ± 1.0

1.3 ± 1.1

-0.09

0.02

1.6 ± 1.0

1.3 ± 0.9

1.5 ± 1.1

-0.09

0.29

Dairy

1.8 ± 1.1

2.1 ± 1.1

0.11

0.006

1.6 ± 1.1

2.2 ± 1.0

1.7 ± 1.2

0.02

0.86

Fruits and veg

3.4 ± 2.3

3.1 ± 1.7

-0.02

0.59

3.2 ± 1.5

3.4 ± 2.0

3.3 ± 2.1

-0.05

0.58

Sugared drinks

0.9 ± 0.9

1.1 ± 0.8

0.07

0.06

0.8 ± 0.6

1.2 ± 1.3

0.7 ± 0.7

-0.05

0.52

Sweets

1.2 ± 1.5

1.2 ± 1.0

0.01

0.79

1.2 ± 0.8

1.2 ± 0.8

1.3 ± 2.5

0.03

0.72

Fast foods

0.9 ± 1.2

0.9 ± 0.6

0.07

0.06

0.8 ± 0.5

1.1 ± 1.7

0.8 ± 1.0

-0.05

0.56

Meats

0.8 ± 1.0

0.8 ± 0.9

-0.07

0.07

0.7 ± 0.7

1.0 ± 1.4

0.7 ± 0.8

-0.08

0.33

Refined grains

0.7 ± 0.7

0.6 ± 0.7

0.05

0.18

0.5 ± 0.5

0.9 ± 0.6

0.8 ± 0.8

0.20

0.01

Daily frequency:

439 440 441 442 443 444

Values are mean ±SD, Spearman or Pearson partial r correlation coefficients, and partial correlation p-values. *Adjusted for age, sex, site of enrolment (Surrey or Brampton) and maternal and paternal education (post-secondary or not). Bolded text indicates significant correlations (Bonferroni p<0.002)

20

445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468

r= -0.3 -0.2 -0.1

0

0.1 0.2

0.3 0.4

Figure 1. Relationship between intake frequency from different food categories. Bolded text indicates significant correlations (Bonferroni p<0.002)

469 470

21

471 472

Supplementary Appendix – Food Frequency Questionnaire

473

22

474 23

475

24

476

References

477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524

Anderson, P. M., & Butcher, K. E. (2006). Childhood obesity: trends and potential causes. Future. Child, 16(1), 19-45. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/16532657 Burrows, T. L., Truby, H., Morgan, P. J., Callister, R., Davies, P. S., & Collins, C. E. (2013). A comparison and validation of child versus parent reporting of children's energy intake using food frequency questionnaires versus food records: who's an accurate reporter? Clin Nutr, 32(4), 613-618. doi:10.1016/j.clnu.2012.11.006 Canada, H. (2016, 1/4/2016). Canada's Food Guide. Retrieved from http://www12.statcan.gc.ca/nhs-enm/2011/ref/dict/pop036-eng.cfm Canada, S. (2008). Canada's Ethnocultural Mosaic, 2006 Census: National picture (97-562XIE2006001). Retrieved from http://www12.statcan.gc.ca/english/census06/analysis/ethnicorigin/pdf/97-562XIE2006001.pdf Canada, S. (2016, 9/16/2016). Generation Status: Canadian-born children of immigrants. Retrieved from http://www23.statcan.gc.ca/imdb/p3VD.pl?Function=getVD&TVD=117200&CVD=11720 0&CLV=0&MLV=1&D=1 Cordain, L., Eaton, S. B., Sebastian, A., Mann, N., Lindeberg, S., Watkins, B. A., . . . BrandMiller, J. (2005). Origins and evolution of the Western diet: health implications for the 21st century. Am. J. Clin. Nutr, 81(2), 341-354. doi:81/2/341 [pii];10.1093/ajcn.81.2.341 [doi] Davison, K. K., & Jago, R. (2009). Change in parent and peer support across ages 9 to 15 yr and adolescent girls' physical activity. Med Sci Sports Exerc, 41(9), 1816-1825. Retrieved from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation &list_uids=19657287 Evans, C. E., Cleghorn, C. L., Greenwood, D. C., & Cade, J. E. (2010). A comparison of British school meals and packed lunches from 1990 to 2007: meta-analysis by lunch type. Br. J. Nutr, 104(4), 474-487. doi:S0007114510001601 [pii];10.1017/S0007114510001601 [doi] Facts and figures on childhood obesity. (2016, 2016). Retrieved from http://www.who.int/endchildhood-obesity/facts/en/ Funtikova, A. N., Navarro, E., Bawaked, R. A., Fito, M., & Schroder, H. (2015). Impact of diet on cardiometabolic health in children and adolescents. Nutr. J, 14, 118. doi:10.1186/s12937-015-0107-z [doi];10.1186/s12937-015-0107-z [pii] He, M., Tucker, P., Gilliland, J., Irwin, J. D., Larsen, K., & Hess, P. (2012). The influence of local food environments on adolescents' food purchasing behaviors. Int J. Environ. Res. Public Health, 9(4), 1458-1471. doi:10.3390/ijerph9041458 [doi];ijerph9041458 [pii] Iqbal, R., Anand, S., Ounpuu, S., Islam, S., Zhang, X., Rangarajan, S., . . . Yusuf, S. (2008). Dietary patterns and the risk of acute myocardial infarction in 52 countries: results of the INTERHEART study. Circulation, 118(19), 1929-1937. doi:CIRCULATIONAHA.107.738716 [pii];10.1161/CIRCULATIONAHA.107.738716 [doi] Jaacks, L. M., Kapoor, D., Singh, K., Narayan, K. M., Ali, M. K., Kadir, M. M., . . . Prabhakaran, D. (2016). Vegetarianism and cardiometabolic disease risk factors: Differences between South Asian and US adults. Nutrition, 32(9), 975-984. doi:S0899-9007(16)00113-1 [pii];10.1016/j.nut.2016.02.011 [doi] Karatzi, K., Moschonis, G., Barouti, A. A., Lionis, C., Chrousos, G. P., & Manios, Y. (2014). Dietary patterns and breakfast consumption in relation to insulin resistance in children. 25

525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573

The Healthy Growth Study. Public Health Nutr, 17(12), 2790-2797. doi:S1368980013003327 [pii];10.1017/S1368980013003327 [doi] Kenny, D. (1979). Correlation and Causality: John Wiley & Sons Inc. Khan, S. A., Jackson, R. T., & Momen, B. (2016). The Relationship between Diet Quality and Acculturation of Immigrated South Asian American Adults and Their Association with Metabolic Syndrome. PLoS. One, 11(6), e0156851. doi:10.1371/journal.pone.0156851 [doi];PONE-D-15-46601 [pii] Kleinbaum, D. G., Kupper, L. L., Nizam, A., & Muller, K. E. (2008). Applied regression analysis and other multivariable methods (4th ed.). Belmont, CA: Duxbury Thomson Brooks/Cole. Koya, D. L., & Egede, L. E. (2007). Association between length of residence and cardiovascular disease risk factors among an ethnically diverse group of United States immigrants. J. Gen. Intern. Med, 22(6), 841-846. doi:10.1007/s11606-007-0163-y [doi] 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 [doi];PONE-D-13-40554 [pii] Livingstone, M. B., Robson, P. J., & Wallace, J. M. (2004). Issues in dietary intake assessment of children and adolescents. Br J Nutr, 92 Suppl 2, S213-222. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/15522159 Lopez, F. A. (2009). Developmental Considerations and Acculturation of Children Measures and Issues. Hispanic Journal of Behavioral Sciences, 31(1), 57-72. doi:10.1177/0739986308327958 Lytle, L. A., Nichaman, M. Z., Obarzanek, E., Glovsky, E., Montgomery, D., Nicklas, T., . . . Feldman, H. (1993). Validation of 24-hour recalls assisted by food records in third-grade children. The CATCH Collaborative Group. J Am Diet Assoc, 93(12), 1431-1436. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/8245378 Martyn-Nemeth, P., Quinn, L., Menon, U., Shrestha, S., Patel, C., & Shah, G. (2017). Dietary Profiles of First-Generation South Asian Indian Adolescents in the United States. J. Immigr. Minor. Health, 19(2), 309-317. doi:10.1007/s10903-016-0382-6 [doi];10.1007/s10903-016-0382-6 [pii] Misra, A., Khurana, L., Isharwal, S., & Bhardwaj, S. (2009). South Asian diets and insulin resistance. Br. J. Nutr, 101(4), 465-473. doi:S0007114508073649 [pii];10.1017/S0007114508073649 [doi] Pan, Y. L., Dixon, Z., Himburg, S., & Huffman, F. (1999). Asian students change their eating patterns after living in the United States. J. Am. Diet. Assoc, 99(1), 54-57. doi:S00028223(99)00016-4 [pii];10.1016/S0002-8223(99)00016-4 [doi] . Partial Correlation. (2013). In Base SAS(R) 9.4 Procedures Guide: Statistical Procedures (2nd ed.). Cary, NC: SAS Institute Inc. (Reprinted from: Not in File). Rani, M. A., & Sathiyasekaran, B. W. (2013). Behavioural determinants for obesity: a crosssectional study among urban adolescents in India. J. Prev. Med. Public Health, 46(4), 192-200. doi:10.3961/jpmph.2013.46.4.192 [doi] 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. J. Immigr. Minor. Health, 16(1), 24-34. doi:10.1007/s10903-013-9823-7 [doi] Satia-Abouta, J., Patterson, R. E., Neuhouser, M. L., & Elder, J. (2002). Dietary acculturation: applications to nutrition research and dietetics. J. Am. Diet. Assoc, 102(8), 1105-1118. doi:S0002-8223(02)80079-7 [pii] Stigler, M., Dhavan, P., Van, D. D., Arora, M., Reddy, K. S., & Perry, C. L. (2010). Westernization and tobacco use among young people in Delhi, India. Soc. Sci. Med, 71(5), 891-897. doi:S0277-9536(10)00464-8 [pii];10.1016/j.socscimed.2010.06.002 [doi]

26

574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599

Talegawkar, S. A., Kandula, N. R., Gadgil, M. D., Desai, D., & Kanaya, A. M. (2016). Dietary intakes among South Asian adults differ by length of residence in the USA. Public Health Nutr, 19(2), 348-355. doi:S1368980015001512 [pii];10.1017/S1368980015001512 [doi] Trapp, G. S., Hickling, S., Christian, H. E., Bull, F., Timperio, A. F., Boruff, B., . . . Giles-Corti, B. (2015). Individual, Social, and Environmental Correlates of Healthy and Unhealthy Eating. Health Educ. Behav, 42(6), 759-768. doi:1090198115578750 [pii];10.1177/1090198115578750 [doi] Unger, J. B., Reynolds, K., Shakib, S., Spruijt-Metz, D., Sun, P., & Johnson, C. A. (2004). Acculturation, physical activity, and fast-food consumption among Asian-American and Hispanic adolescents. J. Community Health, 29(6), 467-481. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/15587346 Vecchio, M. G., Paramesh, E. C., Paramesh, H., Loganes, C., Ballali, S., Gafare, C. E., . . . Gulati, A. (2014). Types of food and nutrient intake in India: a literature review. Indian J. Pediatr, 81 Suppl 1, 17-22. doi:10.1007/s12098-014-1465-9 [doi] Wandel, M., Raberg, M., Kumar, B., & Holmboe-Ottesen, G. (2008). Changes in food habits after migration among South Asians settled in Oslo: the effect of demographic, socioeconomic and integration factors. Appetite, 50(2-3), 376-385. doi:S0195-6663(07)003601 [pii];10.1016/j.appet.2007.09.003 [doi] Wilson, A., & Renzaho, A. (2015). Intergenerational differences in acculturation experiences, food beliefs and perceived health risks among refugees from the Horn of Africa in Melbourne, Australia. Public Health Nutr, 18(1), 176-188. doi:S1368980013003467 [pii];10.1017/S1368980013003467 [doi] Yach, D., Stuckler, D., & Brownell, K. D. (2006). Epidemiologic and economic consequences of the global epidemics of obesity and diabetes. Nat. Med, 12(1), 62-66. doi:nm0106-62 [pii];10.1038/nm0106-62 [doi]

600

27

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

RELATIONSHIP BETWEEN DIET AND ACCULTURATION AMONG SOUTH ASIAN CHILDREN LIVING IN CANADA Salmi Noor1, Mahshid Dehghan2, Scott A Lear3, Sumathi Swaminathan4, Quazi Ibrahim2, Sumathy Rangarajan2, Zubin Punthakee1,2 1 McMaster University, Hamilton, ON, Canada 2 Population Health Research Institute, Hamilton, ON, Canada 3 Faculty of Health Sciences, Simon Fraser University, Vancouver, BC, Canada 4 St. John’s Research Institute, Bangalore, Karnataka, India

Corresponding Author: Zubin Punthakee 1280 Main St. W., HSC 3V51 Hamilton, ON, Canada L8S 4K1 [email protected] Word count: 250 (abstract); 4170 (body). Tables: 4. Figures: 1. References: 39. Keywords: South Asian, acculturation, diet, children, youth Funding: This study was funded by joint grants from the Canadian Institutes of Health Research (FRN 227861) (SAL ZP) and the Indian Council for Medical Research www.icmr.nic.in (SS). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Contributions: SN, MD, SAL, SS, ZP contributed to study design. SAL, SS, SR, ZP contributed to data collection. QI conducted data analysis. SN, MD, SL, ZP contributed to data interpretation. SN and ZP wrote the first draft of the manuscript and MD, SAL, SS, QI, SR provided critical review. All authors approved the final manuscript. Declarations: The authors have declared that no competing interests exist. Abbreviations: CVD cardiovascular disease, FFQ food frequency questionnaire, LoR length of residency, SA South Asian(s)

1

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63

Abstract INTRODUCTION Diet and South Asian ethnicity are both associated with early onset of cardiovascular risk factors. Among youth of South Asian origin, little is known about the role of culture in determining healthy dietary patterns. We aimed to assess dietary patterns and their relationships with acculturation to Western and traditional lifestyles among South Asian youth in Canada. METHODS The Research in Cardiovascular Health - Lifestyles, Environments and Genetic Attributes in Children and Youth (RICH LEGACY) study targeted South Asian children and adolescents aged 7-8 and 14-15 years in two Canadian cities. In this cross-sectional study, acculturation questionnaires and food frequency questionnaires were administered to assess the correlations between Western and traditional culture scores, immigration status (generation and length of residency) in Canada and intake frequency of various foods. RESULTS Among 759 youth, those who ate fruits and vegetables more often consumed dairy and whole grains more often (all r=0.17-0.22, all p<0.001), while those who ate fast food more often consumed meat, sweets and sugared drinks more often (all r 0.24-0.38, all p<0.001). Traditional culture scores were weakly positively correlated with whole grain intake frequency (r=0.12, p=0.001), and negatively with meat intake frequency (r=-0.14, p<0.001). Western culture scores positively correlated with high intake frequency of meat (r=0.23, p<0.001), fast food (r=0.14, p<0.001) and sweets (r=0.14, p<0.001). DISCUSSION Children who are more acculturated with Western lifestyle consumed foods associated with increased metabolic risk. However, whether this eating pattern translates into increased risk of obesity and cardiovascular diseases needs to be further explored.

2

64

INTRODUCTION

65

The Western diet (high in animal products, refined oils, refined grains, added sugars, and

66

processed foods), has been linked to increased cardiovascular disease (CVD) risk factors such

67

as blood lipoprotein, glucose levels and obesity in adults (Cordain et al., 2005) and children

68

(Anderson & Butcher, 2006; "Facts and figures on childhood obesity," 2016) which in turn

69

increase the risk of early onset of type 2 diabetes and CVD (Yach et al., 2006). For people who

70

immigrate to Western countries, diet is substantially impacted due to exposure to a new food

71

environment, thereby potentially leading to changes in food procurement and preparation (Pan

72

et al., 1999; Satia-Abouta et al., 2002). Furthermore, Western acculturation (the incorporation by

73

immigrants of their host country’s culture into their own lifestyle) has been linked to unhealthy

74

eating patterns such as an increase in fast food consumption among youth in several ethnic

75

groups whose traditional diets may be considered more ‘healthy’ (Unger et al., 2004).

76 77

On the contrary, South Asians (SA) have a traditional diet that is high in fat and carbohydrates

78

and is associated with CVD risk factors (Misra et al., 2009). Even SA vegetarian diets which are

79

prevalent for religious reasons do not seem to confer as much benefit on CVD risk factors as

80

vegetarian diets in the US (Jaacks et al., 2016). These are important as CVD starts at younger

81

ages in SA than other ethnic groups (Gupta, Singh, & Verma, 2006). Children and adolescents

82

in India (the largest South Asian country) consume high amounts of cereals and pulses,

83

fats/oils, fast foods and sugared drinks, and inadequate amounts of fruits, vegetables and dairy

84

(Rani & Sathiyasekaran, 2013; Vecchio et al., 2014). Perhaps because of this traditional dietary

85

pattern, data regarding the effects of acculturation on SA adults’ dietary patterns are conflicted

86

(Khan et al., 2016; Lesser et al., 2014; Talegawkar et al., 2016; Wandel et al., 2008), and data

87

regarding SA children are quite minimal. One study of SA children from the United States found

88

excessive fat intake with low intake of fruits, vegetables, meat, fish, beans and eggs overall;

89

however, lower intake of sweets was associated with greater length of residency (Martyn3

90

Nemeth et al., 2017) suggesting some benefits and some harms. In this context, immigration-

91

related changes in the diet quality of youth may depend on the difference in quality between a

92

particular traditional diet and the Western diet, the amount of passive exposure to traditional

93

versus Western food environments and culture, and active participation in traditional versus

94

Western culture. Hence there is a need to understand more about the unique eating patterns of

95

SA youth in other Western countries, and how time in a host Western country and acculturation

96

may improve or worsen dietary patterns, in order to target appropriate public health messaging

97

to this high CVD risk group at an early age. This study aimed to confirm the hypotheses that SA

98

children and adolescents living in Canada have patterns of either predominantly healthy or

99

predominantly unhealthy eating, and that Traditional or Western cultural preferences will

100

distinguish those with greater intake frequency of healthy or unhealthy food items, respectively.

101 102

METHODS

103

Study Design

104

This study was approved by the Hamilton Integrated Research Ethics Board and the Simon

105

Fraser University Research Ethics Board, and parents and participants provided written

106

informed consent and assent, respectively, after having a chance to read study information and

107

ask questions of the researchers in English, Hindi, Punjabi and Urdu. The RICH LEGACY

108

(Research in Cardiovascular Health-Lifestyles, Environments and Genetic Attributes in Children

109

and Youth) study is a cross-sectional study designed to evaluate the determinants of CVD risk

110

factors among SA children in Canada. The study aims to identify differences in body

111

composition, metabolic consequences of adiposity as well as associated behavioural, attitudinal,

112

social and environmental determinants of CVD risk factors. The current report focuses on the

113

relationship between acculturation and immigration status and patterns in dietary intake.

114 115

Participants

4

116

Recruitment targeted SA children (those with at least 3 of 4 grandparents who had ancestral

117

origins in India, Pakistan, Bangladesh or Sri Lanka (Canada, 2008)) aged 7-8 years in

118

elementary schools (elementary school age group) and 14-15 years in high schools (high school

119

age group) from two distinct urban centres in Canada with the largest SA populations (Canada,

120

2008): Surrey, British Columbia and Brampton, Ontario from 2012 to 2016. These age groups

121

were chosen to represent social environments and psycho-behavioural patterns primarily

122

influenced by parents versus peers and self (Davison & Jago, 2009). For practical purposes, the

123

7-8 year old group was also considered to be the youngest group that could reliably contribute

124

information about their preferences and health behaviours (e.g., dietary reporting by children as

125

young as 8 years is as accurate as parent reports (Burrows et al., 2013; Livingstone et al., 2004;

126

Lytle et al., 1993)). There were no other inclusion/exclusion criteria. Individuals were enrolled

127

from a convenience sample of public schools, temples and community centres. Recruitment

128

material and information sheets described the search for relationships of lifestyles and

129

environments with cardiovascular risk factors among youth in general terms without any

130

mention of ethnicity (due ethical concerns by the school boards) or specific dietary hypotheses.

131

Information packages with invitations to contact the study team were dropped off at the schools

132

and distributed at the discretion of principals and teachers to students in grades 2, 3, 9 and 10,

133

and for privacy reasons, investigators were not able to determine the number of eligible

134

candidates to whom invitations were given. Participants from temples and community centres

135

approached research staff at booths during health fairs and public events. As part of the main

136

study, questionnaires were administered to collect demographics, parental education and

137

acculturation and dietary information. The elementary school age children completed the

138

questionnaires by interview with a researcher in the presence of a parent who could assist if the

139

child seemed unsure or requested help. High school age youth completed questionnaires by

140

interview, or self-administration in a classroom setting as per individual school permissions.

141 5

142

Acculturation

143

The Acculturation Rating Scale for Mexican Americans (ARSMA-II) modified to include South

144

Asian references and examples was administered due to its previous use and validity among

145

Indian adolescents (Stigler et al., 2010), and its ease of understanding for children (i.e. it does

146

not include concepts such as humour and values as some other acculturation scales do), and

147

has been validated in children as young as 9 years old (Lopez, 2009). It independently

148

measures the degree to which individuals prefer Western cultural practices, and cultural

149

practices of their native country. Three items in each of four domains (language spoken with:

150

parents, siblings, and friends; media: TV shows, movies, and music; food: lunch/dinner, snacks,

151

and dessert; and consumer goods: clothing, restaurants, and stores/markets) were rated twice,

152

once for Western preferences (Western score) and once for traditional preferences (Traditional

153

score). For each item, participants answered “never, “sometimes”, ‘often” or “very often/always”

154

corresponding to scores of 1, 2, 3, or 4 respectively(Stigler et al., 2010). For participants with no

155

siblings, the question about language spoken with siblings was imputed as the mean score of

156

language spoken with parents and with friends. Western and Traditional scales each had a

157

minimum score of 12 and maximum score of 48, with higher Western and Traditional scores

158

indicating greater degrees of Western and Traditional cultural practices, respectively. Two

159

measures of immigration status were assessed: generation status and for 1st generation

160

participants, length of residency (LoR) in Canada. For generation status, participants were

161

categorized based on Statistics Canada definitions (1st generation - born outside Canada; 2nd

162

generation - born in Canada with at least 1 parent born outside Canada; 3rd generation or more -

163

born and both parents born in Canada) (S. Canada, 2016). In addition, LoR for 1st generation

164

participants was calculated by subtracting age at immigration from current age (Koya & Egede,

165

2007).

166 167

Diet

6

168

A short food frequency questionnaire (FFQ) was used to ask how many times participants

169

habitually ate foods from a list of 25 items per day, per week or per month, or if less than once a

170

month, with no specified time frame for the recall (See supplementary appendix). The

171

questionnaire was adapted (by adding foods and drinks more commonly consumed by children)

172

from the INTERHEART FFQ which, among adults, was able to identify dietary patterns

173

associated with myocardial infarction in multiple countries including India and Canada (Iqbal et

174

al., 2008). Its validity has not been assessed in children or adolescents. Frequency of intake per

175

day of each food item was calculated. We excluded FFQs if responses to ≥5 food items were

176

incomplete or missing, or all 25 food groups were marked with identical frequency, or mean

177

intake per day for any item was >10, or estimated total daily intake (sum of all 25 mean daily

178

intakes) was ≤2. A total of three FFQs were excluded.

179 180

Foods were grouped as whole grains (whole grains/breads/cereals); refined grains (refined

181

breads/rice/pasta and cereals); meats (meat, organ meats, poultry, eggs, fish and seafood);

182

dairy (dairy products); fruits and vegetables (fruits, leafy green vegetables, raw vegetables,

183

other cooked vegetables); fast food (pizza, fast foods, pickled foods, deep fried foods, salty

184

snacks); sweets (ice cream and pudding, dessert/sweet snacks, confectionery, sugars and

185

syrups); and sugared drinks (fruit juice and sugared drinks, non-diet carbonated beverages).

186

Legumes/pulses/nuts/seeds, potatoes (boiled and mashed), and diet carbonated beverages

187

were additional items captured in the FFQ. For each food item, both Western and South Asian

188

examples were provided as prompts (e.g. dairy: milk, yogurt, lassi, raita, …; fast food: chicken

189

nuggets, French fries, pakoras, chaat, …; desserts: cakes, pies, burfi/ladoo, rasgulla/gulab

190

jamun, …). The above groupings of food items were determined by using Canada’s Food Guide

191

as a reference (H. Canada, 2016). Food items that were not included in the food guide (e.g.

192

pizza and other fast foods) were grouped by consensus among the authors. Participants who

193

had not eaten any meat, organ meat, poultry or fish in the last month were classified as 7

194

vegetarians. Lifelong vegetarians were not distinguished from those abstaining from meat for

195

only a month.

196 197

Statistical Analysis

198

Participant demographic characteristics, acculturation and immigration status, and dietary intake

199

frequency of specific food groups (calculated as number of times per day) were described (n

200

(%) or mean ± SD) for the overall study population and across tertiles of Traditional culture

201

scores and Western culture scores. Dietary intake frequency was also presented by 1st and 2nd

202

or 3rd generation participants and across LoR tertiles.

203 204

For descriptive purposes, demographic characteristics were compared across tertiles of

205

Traditional and Western culture scores using the Cochran-Armitage test for trend. Relationships

206

among culture scores and immigration measures were described using partial Pearson

207

correlations with t-statistics for significance (Kleinbaum et al., 2008; "Partial Correlation," 2013),

208

and where generation status was included, using partial Spearman correlations (Kleinbaum et

209

al., 2008; "Partial Correlation," 2013), all adjusted for age, sex, site of enrolment (Surrey or

210

Brampton) and maternal and paternal education (post-secondary or not).

211

In order to test for relationships between frequency of intakes of the pre-specified food groups, a

212

Pearson correlation matrix was constructed. Post hoc, Fisher r to z transformation analysis

213

(Kenny, 1979) was done to identify differences between vegetarians and non-vegetarians in

214

correlation coefficients among the non-meat containing food groups.

215 216

In order to test for relationships between the acculturation or immigration measures and

217

frequency of intake of the pre-specified food groups, partial Pearson correlation, or where

218

generation status was included, partial Spearman correlation were used, all adjusted for age,

219

sex, site of enrolment, and maternal and paternal education. Post hoc, differences between the 8

220

age groups in the relationship between culture scores and frequency of food intake were

221

assessed by an interaction term for age group and Traditional or Western culture score in a

222

linear regression model adjusted for the covariates described above.

223 224

All analyses were conducted using SAS software version 9.4 (SAS institute). For each set of

225

correlations, the Bonferroni correction was used to account for multiple testing such that each

226

correlation was considered statistically significant at p<0.007 in table 2, p<0.002 in tables 3 and

227

4, and p<0.002 in figure 1. Interaction analyses in table 3 were post hoc and exploratory; no

228

correction for multiple testing was considered.

229 230

RESULTS

231

762 participants were enrolled, with a participation rate of approximately 32% among individuals

232

with confirmed contact about the study. Reported South Asian countries of origin were India

233

(n=669, 88%), Pakistan (n=72, 9%), Sri Lanka (n=36, 5%), and Bangladesh (n=6, 1%)

234

(participants could report more than one). Culture questionnaires and valid FFQs were

235

completed by 759 participants, and 715 of them (94%) did so during the school year from

236

September to June. Participant characteristics are shown in Table 1. A high proportion of

237

participants had parents with post-secondary education. Certain demographic factors were

238

progressively associated with cultural preferences. Participants from Surrey and participants

239

whose fathers did not attain post-secondary education were more likely to have higher

240

Traditional culture scores (p=0.035 and p=0.011, respectively). Participants who were older and

241

participants whose parents attained post-secondary education were more likely to have higher

242

Western culture scores (p<0.001 for all), Vegetarians were more likely to have higher Traditional

243

culture scores (p<0.001) and lower Western culture scores (p<0.001) than non-vegetarians.

244

9

245

Culture preference patterns are shown in Table 2. Participants had higher Western culture

246

scores than Traditional culture scores, and the majority were 2nd or 3rd generation Canadians

247

(only 5 participants were 3rd generation). After adjusting for age, sex, site of enrolment and

248

maternal and paternal education, Traditional and Western culture scores were inversely

249

associated with each other (partial r=-0.23, p<0.001). After adjusting for the same variables,

250

Traditional culture scores were inversely associated with generation status (partial r=-0.12,

251

p=0.002), and LoR among 1st generation Canadians (partial r=-0.22, p=0.005), while Western

252

culture scores were positively associated with generation status (partial r=0.14, p<0.001).

253 254

Fast foods, sweets and sugared drinks were each consumed once daily, on average (Table 3).

255

The frequency of dietary intake correlation matrix is presented in Figure 1 and shows two

256

groupings of correlated foods. Intake of fruits and vegetables, dairy and whole grains were all

257

positively correlated with one another (p<0.001 for all). The second grouping consisting of fast

258

foods, sweets and sugared drinks were also all positively correlated with one another (p<0.001

259

for all). We found positive correlations between meat intake and fast food intake frequencies

260

(r=0.24, p<0.001), and meat intake and refined grains intake frequencies (r=0.14, P<0.001).

261

There was an inverse correlation between refined grain and whole grain intake frequencies (r=-

262

0.23, p<0.001). We did not find differences between vegetarians and non-vegetarians in these

263

significant correlations (all differences p>0.05 by Fisher r to z transformation), but we found one

264

marginal difference in the non-significant correlations between fruits and vegetables and sweets

265

(r=-0.15, p=0.04 for vegetarians, r=0.04, p=0.29 for non-vegetarians; p for difference=0.02).

266 267

The relationships between acculturation and dietary patterns are shown in Table 3. Participants

268

with higher Traditional culture scores consumed whole grains more frequently (partial r=0.12,

269

p=0.001), and had meat less frequently (partial r=-0.14, p<0.001), and there were no differences

270

among the age groups. Those with higher Western culture scores consumed sweets (partial 10

271

r=0.14, p<0.001), fast foods (partial r=0.14, p<0.001) and meats (partial r=0.23, p<0.001) more

272

often. The high school age group demonstrated stronger relationships than the elementary

273

school age group between Western culture scores and intake frequency of sweets (p-

274

interaction=0.003) and fast foods (p-interaction<0.001), but not meats (p=0.29). We did not find

275

significant relationships between generation status or LoR and dietary patterns (Table 4).

276 277

11

278

DISCUSSION

279

We found that among SA children and youth living in Canada there are significant correlations

280

between intake frequencies of distinct food groups that are differentially associated with cultural

281

preferences. Overall, children and youth in our study consumed fast foods, sweets and sugared

282

drinks on a daily basis. We observed two distinct groupings of correlated food intake frequency:

283

1) whole grains, dairy, and fruits and vegetables, and 2) meat, fast foods, sweets, and sugared

284

drinks. Traditional cultural preferences were associated with the first grouping and Western

285

cultural preferences with the second grouping.

286 287

In the population studied, there were correlations among intake frequency of foods associated

288

with low levels of cardiometabolic risk (“healthy foods”: whole grains, dairy and fruits and

289

vegetables) (Funtikova et al., 2015) and among intake frequency of foods associated with high

290

levels of cardiometabolic risk (“unhealthy foods”: meat, fast foods, sweets and sugared drinks)

291

(Funtikova et al., 2015; Karatzi et al., 2014). Though these healthy and unhealthy patterns were

292

not mutually exclusive (demonstrated by a paucity of inversely correlated foods), they suggest

293

that those who make healthy or unhealthy food choices tend to do so in multiple domains within

294

that grouping, possibly in the form of meals composed of predominantly either healthy or

295

unhealthy foods (Evans et al., 2010). This study was not able to determine the extent to which

296

this grouping of food types is driven by individual preferences and behaviours or the availability

297

of grouped foods in the food environment, but both have been shown to play a role in healthy

298

and unhealthy eating (Trapp et al., 2015).

299 300

While correlations did occur within healthy or unhealthy groupings, it remains to be investigated

301

whether promoting individual healthy foods may lead to an overall healthier dietary pattern, or

302

discouraging individual unhealthy foods may reduce overall unhealthy eating patterns. For

303

example, among vegetarians in our study, positive correlations between fast foods, sweets and 12

304

sugared drinks were noted. Other studies have shown that adult SA vegetarians living in their

305

native country eat more fried food and dessert than non-vegetarians (Jaacks et al., 2016) (both

306

of which have been associated with higher levels of cardiometabolic risk), and our study

307

extends these findings to SA youth living in a Western country. These observations highlight

308

that restrictive diets such as vegetarianism may not eliminate the challenges of unhealthy

309

eating.

310 311

Except for whole grains and refined grains there were no significant inverse correlations

312

between food groups (Figure 1). These findings suggest SA youth may substitute healthy foods

313

for unhealthy foods or vice versa within a category (i.e. grains), but not between categories (e.g.

314

fruits and vegetables for fast food). Therefore, public health interventions that focus on a single

315

category (e.g. eat more fruits and vegetables) may not be sufficient if they do not also suggest

316

from which categories to substitute (e.g. eat fruits and vegetables instead of fast foods), and

317

may additionally benefit from emphasizing healthy choices within food categories, taking into

318

consideration the community’s food preferences.

319 320

Acculturation had a significant association with frequency of food group intakes. SA youth with

321

greater Western culture preferences more frequently consumed foods associated with

322

increased cardiometabolic risk, such as meats, fast foods and sweets, while those identifying

323

more strongly with traditional culture consumed meat less often and whole grains more often.

324

This may be due to greater adherence to vegetarianism in those who retain traditionalism, and

325

greater use of raw grains than processed carbohydrates in traditional food preparation. This

326

suggests that the traditional diet of SA youth is not inherently unhealthy (especially when

327

compared to a Western diet), and dietary patterns experience a shift towards unhealthy food

328

items associated with greater Western acculturation. This is consistent with research on Asian

329

American and Hispanic adolescents showing that those who identify more with Western culture 13

330

eat more non-traditional and high fat foods (Unger et al., 2004). Our results are contrary to a

331

small study of SA adolescents showing more time in the US was associated with lower intake of

332

sweets, a potentially healthy dietary effect of acculturation (Martyn-Nemeth et al., 2017). While

333

this heterogeneity is difficult to explain, our study adds support to a consistent effect of

334

acculturation on diet among youth regardless of ethnicity. Furthermore, in our study, time

335

measures (LoR and generation status) were not associated with diet patterns, suggesting that

336

children and youth do not experience the dietary impacts of culture through passive exposure,

337

but by taking on certain beliefs, attitudes and behaviours which may happen non-linearly with

338

time.

339 340

Diets of SA often change after migration (Lesser et al., 2014), especially those of younger

341

adults, compared to older adult SA immigrants who are less likely to change their dietary

342

patterns (Wandel et al., 2008). Whether this age trend extends into youth is unknown, though

343

evidence in other settings suggests migrant youth may undergo dietary acculturation to a

344

greater extent than their parents (Wilson & Renzaho, 2015). Between elementary school age

345

children and high school adolescents, we found a greater preference for Western culture among

346

adolescents, and stronger relationships between Western culture and intake frequencies of fast

347

foods and sweets among adolescents. This, coupled with adolescents being highly susceptible

348

to the availability of fast food and convenience food in the environment (He et al., 2012), may

349

make adolescence a particularly vulnerable period for unhealthy eating as SA grow up in

350

Western countries.

351 352

Western food and Western diet are not synonymous. In Western countries, there is an

353

abundance of food with wide-ranging nutritional value. Our findings, and those of others (Sanou

354

et al., 2014) indicate that Westernization of the diet is due to acceptance of and participation in

355

a culture that values easily available, accessible, affordable and convenient foods that happen 14

356

to be energy dense and nutrient poor. Thus, for immigrants who maintain their preference for

357

traditional food types, their diet may nevertheless be influenced towards more restaurant foods,

358

packaged foods and larger portions because of overall acculturation, and not specifically dietary

359

acculturation.

360 361

This study had several notable strengths. It included a large number of participants with broad

362

variation in age, generational status and acculturation enrolled from several schools and

363

temples in diverse neighbourhoods in two cities, although it was not a random sample. Thus,

364

the results are likely generalizable to a broad spectrum of South Asian youth from elementary to

365

high school living in Western countries with similar immigration policies and in urban centres

366

where the South Asian population is large enough to maintain its own culture in the community.

367

In addition, stringent criteria for statistical significance were used to minimize spurious

368

inferences. This study also had some limitations. A short FFQ to estimate habitual dietary intake

369

of participants was used in this study. The short FFQ provided detailed information on intake

370

frequency of different foods, but not enough detail to estimate daily energy, macro- and

371

micronutrient intakes. Furthermore, it did not allow for identification of traditional cultural or non-

372

traditional foods or methods of food preparation, and would not have been able to determine if

373

Western acculturation was associated with eating more pakoras and chaat, or hot dogs and

374

French fries. However, FFQ is the most common method used by epidemiological studies. In

375

addition, the Traditional scale of the acculturation questionnaire referred to the individual’s

376

native country and may not have adequately captured traditional culture of second and third

377

generation SA Canadians. This is confounded by the fact that many South Asian countries have

378

a colonial past which makes their native country’s recent culture somewhat Westernized to

379

begin with, and this could not be measured. In addition, we aggregated acculturation across

380

many domains and did not specifically assess dietary acculturation in isolation. Furthermore,

381

this study is limited by providing a cross-sectional measure of acculturation which is a constantly 15

382

evolving process that varies across time (Stigler et al., 2010) in a non-linear way. Finally, the

383

post-hoc age subgroup analyses had less than 80% power to detect Pearson correlations

384

smaller than 0.12 in the elementary school age group and 0.22 in the high school age group.

385 386 387

CONCLUSION

388

With the increase in SA immigrants in Canada, it is important to ensure they are able to lead a

389

healthy lifestyle in their new host country starting from a young age. Among SA children and

390

youth, intake frequency of fruits and vegetables, dairy and whole grains correlated, as did intake

391

frequency of fast foods, sweets, sugared drinks, and among non-vegetarians, meat, indicating a

392

healthy grouping and an unhealthy grouping of food preferences. Those identifying more with

393

Western culture resorted to unhealthier eating practices such as more frequent intake of fast

394

foods and sweets whereas individuals identifying with traditional culture ate whole grains more

395

often and meat less often. These findings largely suggest a link between Western acculturation

396

and dietary practices associated with childhood obesity. Post hoc analyses also suggest these

397

patterns exist for both non-vegetarians and vegetarians, and are more of a concern for

398

adolescents than younger children of SA origin living in Canada. As such, interventions should

399

focus on educating parents, young children and increasingly independent adolescents (both

400

vegetarian and non-vegetarian) to increase intake of unprocessed foods and reduce intake of

401

energy dense foods. One approach to achieving this may be for community leaders such as

402

religious leaders and community resources such as school boards and public health offices in

403

regions with large SA populations to use both scientific knowledge and traditional knowledge to

404

promote traditional cultural practices among immigrants of SA origin. Future research should

405

focus on the best way of achieving this behaviour change and determining whether it impacts

406

cardiovascular risk factors.

16

407

Table 1. Characteristics of participants Traditional culture score

Western culture score

Characteristics

All

T1 16-25

T2 25.5-28

T3 28.5-44

n

759

299

218

242

Gender

p for trend*

T1 20-32

T2 32.5-36

T3 37-48

251

266

242

0.13

0.23

Females

401 (53)

146 (49)

121 (56)

134 (55)

121 (48)

150 (56)

130 (54)

Males

358 (47)

153 (51)

97 (44)

108 (45)

130 (52)

116 (44)

112 (46)

Age group

0.073

<0.001

Elementary

588 (78)

234 (78)

181 (83)

173 (72)

218 (87)

222 (84)

148 (61)

High school

171 (22)

65 (22)

37 (17)

69 (28)

33 (13)

44 (16)

94 (39)

Site

p for trend*

0.035

0.39

Brampton

479 (63)

197 (66)

144 (66)

138 (57)

141 (56)

193 (73)

145 (60)

Surrey

280 (37)

102 (34)

74 (34)

104 (43)

110 (44)

73 (27)

97 (40)

509 (70)

205 (72)

141 (67)

163 (71)

0.65

152 (62)

183 (72)

174 (78)

<0.001

464 (66)

195 (72)

131 (63)

138 (61)

0.011

142 (58)

166 (66)

156 (74)

<0.001

Yes

179 (24)

47 (16)

61 (28)

71 (29)

79 (32)

59 (22)

41 (17)

No

580 (76)

252 (84)

157 (72)

171 (71)

172 (68)

207 (78)

201 (83)

Mothers with postsecondary education Fathers with postsecondary education Vegetarian

408 409 410

<0.001

Values are n (%). * p for trend by Cochran Armitage test.

411

17

<0.001

412 413

Table 2. Relationship between acculturation and immigration measures Traditional culture score All

n Traditional culture score Western culture score

Western culture score

T1 16-25

T2 25.5-28

T3 28.5-44

759

299

218

242

26.7 ±4.4

22.5 ±2.2

27.0 ±0.8

31.7 ±3.0

-

34.8 ±4.7

36.2 ±4.3

33.5 ±3.8

34.2 ±5.4

Generation*

T1 20-32

T2 32.5-36

T3 37-48

251

266

242

-

28.0 ±3.5

26.4 ±3.5

-0.23

<0.001

29.9 ±2.2

34.4 ±1.2

-0.12

0.002

r**

p**

r**

p**

25.7 ±5.7

-0.23

<0.001

40.2 ±2.9

-

-

0.14

<0.001

0.14

0.09

1st

180 (24)

62 (21)

45 (21)

73 (30)

67 (27)

59 (22)

54 (22)

2nd & 3rd

579 (76)

237 (79)

173 (79)

169 (70)

184 (73)

207 (78)

188 (78)

5.8 ±3.5

6.3 ±3.5

5.8 ±3.5

5.3 ±3.4

5.2 ±3.4

5.4 ±2.9

6.9 ±3.9

LoR for 1st generation (y)

414 415 416 417 418 419 420

-0.22

0.005

Values are n (%), mean ±SD, Pearson or *Spearman partial r correlation coefficients or partial correlation p-values. **Adjusted for age, sex, site of enrolment (Surrey or Brampton) and maternal and paternal education (post-secondary or not). Bold text indicates significant correlations (Bonferroni p<0.007).

421

18

422 423 424

Table 3: Relationship between food intake frequency and acculturation, overall and by age group All Age group

All n

Traditional culture score T1 16-25

759

299

T2 25.5-28 218

T3 28.5-44

Western culture score

r*

p*

p* for age group difference

242

T1 20-32 251

T2 32.5-36 266

T3 37-48

r*

p*

p* for age group difference

242

Elem.

588

234

181

173

218

222

148

High

171

65

37

69

33

44

94

1.3±1.1

1.1±0.9

1.4±1.0

1.4±1.2

1.4±1.0

1.3±1.0

1.3±1.1

Daily frequency: All Whole grains

1.3±1.0

1.1±0.9

1.4±1.0

1.3±1.0

0.10

0.02

High

1.5±1.3

1.2±1.0

1.6±1.1

1.7±1.5

0.15

0.10

2.0±1.1

2.0±1.0

2.2±1.1

1.9±1.1

-0.06

2.0±1.0

2.1±1.0

2.1±1.0

1.8±1.0

-0.09

0.03

High

2.0±1.5

1.9±1.2

2.5±1.7

1.9±1.5

0.04

0.67

3.2±1.9

3.1±2.1

3.2±1.4

3.3±1.9

0.10

3.0±1.4

2.9±1.5

3.0±1.3

3.0±1.4

0.08

0.06

High

3.8±2.8

3.8±3.4

3.8±1.8

3.8±2.7

0.03

0.72

All 1.0±0.9 1.0±0.9 1.1±0.8 1.0±0.9 0.02 0.49 Sugared Elem. 1.1±0.8 1.1±0.9 1.1±0.8 1.0±0.7 -0.02 0.60 drinks High 1.0±1.1 0.8±0.9 1.1±1.1 1.1±1.3 0.11 0.22

Sweets

425 426 427 428 429 430

1.2±1.1

-0.04

1.2±1.0

1.2±0.8

1.1±0.8

-0.06

0.17

1.3±1.8

1.3±2.3

0.9±0.9

1.5±1.6

-0.02

0.85

0.9±0.8

0.9±0.6

0.8±0.5

1.0±1.1

0.07

0.9±0.6

0.9±0.5

0.8±0.6

0.9±0.6

0.01

0.87

High

1.0±1.3

0.9±0.8

0.8±0.5

1.3±1.8

0.14

0.13

0.9±1.0

0.6±0.6

0.7±1.0

Elem.

0.7±0.7

0.9±0.7

0.6±0.6

0.6±0.6

High

1.0±1.4

1.2±1.6

0.7±0.8

1.0±1.5

0.7±0.7

0.7±0.7

0.6±0.6

0.7±0.7

-0.14

0.00 -0.05

0.10

0.97

0.96

0.6±0.6

0.7±0.7

0.6±0.6

0.6±0.6

-0.07

0.10

High

0.7±0.7

0.7±0.7

0.6±0.7

0.8±0.8

0.10

0.30

0.01

1.6±1.1

1.4±1.4

1.5±1.2

-0.12

0.20

0.03

0.06

2.0±1.0

0.02

0.63

1.8±1.3

2.0±1.5

0.14

0.12

3.2±1.7

3.4±2.3

0.02

3.2±1.5

3.0±1.5

0.03

0.48

3.7±2.3

3.4±2.6

4.0±3.1

0.03

0.78

1.0±0.9

0.04

1.1±0.8

1.1±0.8

0.01

0.79

1.0±1.0

1.1±1.2

1.0±1.1

0.04

0.63

1.3±1.8

0.14

1.2±0.7

1.3±1.3

0.09

0.04

1.0±0.9

1.2±1.0

1.4±2.3

0.22

0.02

1.0±1.1

0.14

0.9±0.6

0.9±0.5

0.04

0.33

0.7±0.4

1.0±1.0

1.2±1.5

0.29

0.001

1.0±1.2

0.23

0.8±0.7

0.9±0.7

0.22 <0.001

0.5±0.8

0.8±0.8

1.2±1.7

0.26

0.8±0.8

0.07

0.005

0.29

0.07

0.6±0.6

0.6±0.6

0.7±0.8

0.08

0.06

0.6±0.7

0.6±0.6

0.8±0.8

0.08

0.36

Elem.- elementary school age, High - high school age. Values are mean ±SD, Pearson partial r and partial correlation p-values. *Adjusted for age, sex, site of enrolment (Surrey or Brampton) and maternal and paternal education (post-secondary or not). Bolded text indicates significant correlations for main effects (Bonferroni p<0.002).

19

<0.001

<0.001

0.6±0.6

0.6±0.6

0.003

<0.001

0.9±0.6

0.8±0.7

0.58

<0.001

1.1±0.7

0.9±0.7

0.80

0.37

1.0±0.7

1.2±0.8

0.11

0.57

2.7±1.3

1.1±0.9

0.96

0.14

2.1±0.9

0.6±0.6 0.14

2.0±1.2

2.4±1.6

0.5±0.6 0.12

2.0±1.0

2.0±1.0

0.8±0.5

0.23

Elem.

-0.10

1.1±0.7

<0.001

-0.19 <0.001

1.1±1.0

1.0±0.8

0.07

Elem.

0.9±0.9

0.87

0.008

1.3±0.9

2.9±1.5

0.31

1.2±0.9

All Refined grains

1.1±0.8

High

All Meats

1.2±1.4

Elem.

All Fast foods

1.2±1.2

0.12

-0.10

1.3±1.0

2.0±1.1

0.008

Elem.

All

0.72

0.15

Elem.

All Fruit & veg

0.001

Elem.

All Dairy

0.12

0.99

431 432

Table 4: Relationship between food intake frequency and immigration status st

Generation 1 n

st

2

nd

&3

rd

Length of Residency (y) among 1 generation Spearman r*

p*

T1 0.17-3.81

T2 3.82-6.79

T3 6.80-14.97

55

55

56

Pearson r*

p*

180

579

Whole grains

1.5 ± 1.0

1.3 ± 1.1

-0.09

0.02

1.6 ± 1.0

1.3 ± 0.9

1.5 ± 1.1

-0.09

0.29

Dairy

1.8 ± 1.1

2.1 ± 1.1

0.11

0.006

1.6 ± 1.1

2.2 ± 1.0

1.7 ± 1.2

0.02

0.86

Fruits and veg

3.4 ± 2.3

3.1 ± 1.7

-0.02

0.59

3.2 ± 1.5

3.4 ± 2.0

3.3 ± 2.1

-0.05

0.58

Sugared drinks

0.9 ± 0.9

1.1 ± 0.8

0.07

0.06

0.8 ± 0.6

1.2 ± 1.3

0.7 ± 0.7

-0.05

0.52

Sweets

1.2 ± 1.5

1.2 ± 1.0

0.01

0.79

1.2 ± 0.8

1.2 ± 0.8

1.3 ± 2.5

0.03

0.72

Fast foods

0.9 ± 1.2

0.9 ± 0.6

0.07

0.06

0.8 ± 0.5

1.1 ± 1.7

0.8 ± 1.0

-0.05

0.56

Meats

0.8 ± 1.0

0.8 ± 0.9

-0.07

0.07

0.7 ± 0.7

1.0 ± 1.4

0.7 ± 0.8

-0.08

0.33

Refined grains

0.7 ± 0.7

0.6 ± 0.7

0.05

0.18

0.5 ± 0.5

0.9 ± 0.6

0.8 ± 0.8

0.20

0.01

Daily frequency:

433 434 435 436 437 438

Values are mean ±SD, Spearman or Pearson partial r correlation coefficients, and partial correlation p-values. *Adjusted for age, sex, site of enrolment (Surrey or Brampton) and maternal and paternal education (post-secondary or not). Bolded text indicates significant correlations (Bonferroni p<0.002)

20

439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462

r= -0.3 -0.2 -0.1

0

0.1 0.2

0.3 0.4

Figure 1. Relationship between intake frequency from different food categories. Bolded text indicates significant correlations (Bonferroni p<0.002)

463 464

21

465 466

Supplementary Appendix – Food Frequency Questionnaire

467

22

468 23

469

24

470

References

471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518

Anderson, P. M., & Butcher, K. E. (2006). Childhood obesity: trends and potential causes. Future. Child, 16(1), 19-45. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/16532657 Burrows, T. L., Truby, H., Morgan, P. J., Callister, R., Davies, P. S., & Collins, C. E. (2013). A comparison and validation of child versus parent reporting of children's energy intake using food frequency questionnaires versus food records: who's an accurate reporter? Clin Nutr, 32(4), 613-618. doi:10.1016/j.clnu.2012.11.006 Canada, H. (2016, 1/4/2016). Canada's Food Guide. Retrieved from http://www12.statcan.gc.ca/nhs-enm/2011/ref/dict/pop036-eng.cfm Canada, S. (2008). Canada's Ethnocultural Mosaic, 2006 Census: National picture (97-562XIE2006001). Retrieved from http://www12.statcan.gc.ca/english/census06/analysis/ethnicorigin/pdf/97-562XIE2006001.pdf Canada, S. (2016, 9/16/2016). Generation Status: Canadian-born children of immigrants. Retrieved from http://www23.statcan.gc.ca/imdb/p3VD.pl?Function=getVD&TVD=117200&CVD=11720 0&CLV=0&MLV=1&D=1 Cordain, L., Eaton, S. B., Sebastian, A., Mann, N., Lindeberg, S., Watkins, B. A., . . . BrandMiller, J. (2005). Origins and evolution of the Western diet: health implications for the 21st century. Am. J. Clin. Nutr, 81(2), 341-354. doi:81/2/341 [pii];10.1093/ajcn.81.2.341 [doi] Davison, K. K., & Jago, R. (2009). Change in parent and peer support across ages 9 to 15 yr and adolescent girls' physical activity. Med Sci Sports Exerc, 41(9), 1816-1825. Retrieved from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation &list_uids=19657287 Evans, C. E., Cleghorn, C. L., Greenwood, D. C., & Cade, J. E. (2010). A comparison of British school meals and packed lunches from 1990 to 2007: meta-analysis by lunch type. Br. J. Nutr, 104(4), 474-487. doi:S0007114510001601 [pii];10.1017/S0007114510001601 [doi] Facts and figures on childhood obesity. (2016, 2016). Retrieved from http://www.who.int/endchildhood-obesity/facts/en/ Funtikova, A. N., Navarro, E., Bawaked, R. A., Fito, M., & Schroder, H. (2015). Impact of diet on cardiometabolic health in children and adolescents. Nutr. J, 14, 118. doi:10.1186/s12937-015-0107-z [doi];10.1186/s12937-015-0107-z [pii] He, M., Tucker, P., Gilliland, J., Irwin, J. D., Larsen, K., & Hess, P. (2012). The influence of local food environments on adolescents' food purchasing behaviors. Int J. Environ. Res. Public Health, 9(4), 1458-1471. doi:10.3390/ijerph9041458 [doi];ijerph9041458 [pii] Iqbal, R., Anand, S., Ounpuu, S., Islam, S., Zhang, X., Rangarajan, S., . . . Yusuf, S. (2008). Dietary patterns and the risk of acute myocardial infarction in 52 countries: results of the INTERHEART study. Circulation, 118(19), 1929-1937. doi:CIRCULATIONAHA.107.738716 [pii];10.1161/CIRCULATIONAHA.107.738716 [doi] Jaacks, L. M., Kapoor, D., Singh, K., Narayan, K. M., Ali, M. K., Kadir, M. M., . . . Prabhakaran, D. (2016). Vegetarianism and cardiometabolic disease risk factors: Differences between South Asian and US adults. Nutrition, 32(9), 975-984. doi:S0899-9007(16)00113-1 [pii];10.1016/j.nut.2016.02.011 [doi] Karatzi, K., Moschonis, G., Barouti, A. A., Lionis, C., Chrousos, G. P., & Manios, Y. (2014). Dietary patterns and breakfast consumption in relation to insulin resistance in children. 25

519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567

The Healthy Growth Study. Public Health Nutr, 17(12), 2790-2797. doi:S1368980013003327 [pii];10.1017/S1368980013003327 [doi] Kenny, D. (1979). Correlation and Causality: John Wiley & Sons Inc. Khan, S. A., Jackson, R. T., & Momen, B. (2016). The Relationship between Diet Quality and Acculturation of Immigrated South Asian American Adults and Their Association with Metabolic Syndrome. PLoS. One, 11(6), e0156851. doi:10.1371/journal.pone.0156851 [doi];PONE-D-15-46601 [pii] Kleinbaum, D. G., Kupper, L. L., Nizam, A., & Muller, K. E. (2008). Applied regression analysis and other multivariable methods (4th ed.). Belmont, CA: Duxbury Thomson Brooks/Cole. Koya, D. L., & Egede, L. E. (2007). Association between length of residence and cardiovascular disease risk factors among an ethnically diverse group of United States immigrants. J. Gen. Intern. Med, 22(6), 841-846. doi:10.1007/s11606-007-0163-y [doi] 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 [doi];PONE-D-13-40554 [pii] Livingstone, M. B., Robson, P. J., & Wallace, J. M. (2004). Issues in dietary intake assessment of children and adolescents. Br J Nutr, 92 Suppl 2, S213-222. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/15522159 Lopez, F. A. (2009). Developmental Considerations and Acculturation of Children Measures and Issues. Hispanic Journal of Behavioral Sciences, 31(1), 57-72. doi:10.1177/0739986308327958 Lytle, L. A., Nichaman, M. Z., Obarzanek, E., Glovsky, E., Montgomery, D., Nicklas, T., . . . Feldman, H. (1993). Validation of 24-hour recalls assisted by food records in third-grade children. The CATCH Collaborative Group. J Am Diet Assoc, 93(12), 1431-1436. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/8245378 Martyn-Nemeth, P., Quinn, L., Menon, U., Shrestha, S., Patel, C., & Shah, G. (2017). Dietary Profiles of First-Generation South Asian Indian Adolescents in the United States. J. Immigr. Minor. Health, 19(2), 309-317. doi:10.1007/s10903-016-0382-6 [doi];10.1007/s10903-016-0382-6 [pii] Misra, A., Khurana, L., Isharwal, S., & Bhardwaj, S. (2009). South Asian diets and insulin resistance. Br. J. Nutr, 101(4), 465-473. doi:S0007114508073649 [pii];10.1017/S0007114508073649 [doi] Pan, Y. L., Dixon, Z., Himburg, S., & Huffman, F. (1999). Asian students change their eating patterns after living in the United States. J. Am. Diet. Assoc, 99(1), 54-57. doi:S00028223(99)00016-4 [pii];10.1016/S0002-8223(99)00016-4 [doi] . Partial Correlation. (2013). In Base SAS(R) 9.4 Procedures Guide: Statistical Procedures (2nd ed.). Cary, NC: SAS Institute Inc. (Reprinted from: Not in File). Rani, M. A., & Sathiyasekaran, B. W. (2013). Behavioural determinants for obesity: a crosssectional study among urban adolescents in India. J. Prev. Med. Public Health, 46(4), 192-200. doi:10.3961/jpmph.2013.46.4.192 [doi] 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. J. Immigr. Minor. Health, 16(1), 24-34. doi:10.1007/s10903-013-9823-7 [doi] Satia-Abouta, J., Patterson, R. E., Neuhouser, M. L., & Elder, J. (2002). Dietary acculturation: applications to nutrition research and dietetics. J. Am. Diet. Assoc, 102(8), 1105-1118. doi:S0002-8223(02)80079-7 [pii] Stigler, M., Dhavan, P., Van, D. D., Arora, M., Reddy, K. S., & Perry, C. L. (2010). Westernization and tobacco use among young people in Delhi, India. Soc. Sci. Med, 71(5), 891-897. doi:S0277-9536(10)00464-8 [pii];10.1016/j.socscimed.2010.06.002 [doi]

26

568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593

Talegawkar, S. A., Kandula, N. R., Gadgil, M. D., Desai, D., & Kanaya, A. M. (2016). Dietary intakes among South Asian adults differ by length of residence in the USA. Public Health Nutr, 19(2), 348-355. doi:S1368980015001512 [pii];10.1017/S1368980015001512 [doi] Trapp, G. S., Hickling, S., Christian, H. E., Bull, F., Timperio, A. F., Boruff, B., . . . Giles-Corti, B. (2015). Individual, Social, and Environmental Correlates of Healthy and Unhealthy Eating. Health Educ. Behav, 42(6), 759-768. doi:1090198115578750 [pii];10.1177/1090198115578750 [doi] Unger, J. B., Reynolds, K., Shakib, S., Spruijt-Metz, D., Sun, P., & Johnson, C. A. (2004). Acculturation, physical activity, and fast-food consumption among Asian-American and Hispanic adolescents. J. Community Health, 29(6), 467-481. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/15587346 Vecchio, M. G., Paramesh, E. C., Paramesh, H., Loganes, C., Ballali, S., Gafare, C. E., . . . Gulati, A. (2014). Types of food and nutrient intake in India: a literature review. Indian J. Pediatr, 81 Suppl 1, 17-22. doi:10.1007/s12098-014-1465-9 [doi] Wandel, M., Raberg, M., Kumar, B., & Holmboe-Ottesen, G. (2008). Changes in food habits after migration among South Asians settled in Oslo: the effect of demographic, socioeconomic and integration factors. Appetite, 50(2-3), 376-385. doi:S0195-6663(07)003601 [pii];10.1016/j.appet.2007.09.003 [doi] Wilson, A., & Renzaho, A. (2015). Intergenerational differences in acculturation experiences, food beliefs and perceived health risks among refugees from the Horn of Africa in Melbourne, Australia. Public Health Nutr, 18(1), 176-188. doi:S1368980013003467 [pii];10.1017/S1368980013003467 [doi] Yach, D., Stuckler, D., & Brownell, K. D. (2006). Epidemiologic and economic consequences of the global epidemics of obesity and diabetes. Nat. Med, 12(1), 62-66. doi:nm0106-62 [pii];10.1038/nm0106-62 [doi]

27