Source Of Nutrition Information In Relation To Weight Loss Behaviours .

1y ago
47 Views
2 Downloads
1.25 MB
155 Pages
Last View : 1m ago
Last Download : 2m ago
Upload by : Maxine Vice
Transcription

Source of Nutrition Information in Relation to Weight Loss Behaviours Among Young CanadianAdults Trying to Lose WeightbyMiriam PriceA thesispresented to the University Of Waterlooin fulfilment of thethesis requirement for the degree ofMaster of ScienceinPublic Health and Health SystemsWaterloo, Ontario, Canada, 2018 Miriam Price 2018

Author’s DeclarationI hereby declare that I am the sole author of this thesis. This is a true copy of the thesis, includingany required final revisions, as accepted by my examiners.I understand that my thesis may be made electronically available to the public.ii

AbstractWeight loss efforts are pervasive among young adults and worrisome as they areassociated with poor mental health and development of eating disorders. Data on weight lossbehaviours are limited among Canadian youth, as is knowledge of environmental variables (e.g.,information sources) that are associated with such behaviours. Current literature predominantlyconsiders weight loss behaviours individually, despite evidence that health behaviours co-occur.This simplistic method of conceptualizing weight loss behaviours may have implications forresearch examining correlates and implications of strategies used to lose weight. The purpose ofthis study was to examine patterns of weight loss behaviours and nutrition information sourcesutilized among young Canadians (16-30 years of age) who reported trying to lose weight overthe past year, and to examine associations between the sources consulted and weight lossmethods utilized. Cross-sectional data were drawn from the first wave of the Canada FoodStudy, a cohort study of young adults from five urban areas. Factor analysis was used to identifypatterns of weight loss behaviours. Four factors, or patterns, were identified: Dietary Changes,Purging and Restrictive Behaviours, Non-Prescribed Supplements and Formulas, and HealthPromoting Behaviours. Factor analysis was also used to examine covariation among the sourcesof nutrition information reported, again identifying four factors: Government and HealthAssociation Materials, Health and Weight Loss Specialists, Commercial Sources, and EasilyAccessible Sources. Building on insights from the factor analyses to operationalize variables,Poisson regression modelling was used to examine associations between information sources andweight loss behaviours. Associations were found between the nutrition information source usedand weight loss behaviours. The findings of this study challenge others to re-examine the waysiii

in which weight loss behaviours are conceptualized, and provide insights into the possibleimplications of relying on certain types of sources of nutrition information for weight lossbehaviours.iv

AcknowledgmentsFirst, I would like to thank my thesis supervisor, Dr. Sharon Kirkpatrick. Her leadershipand support have been exemplary and I could not have completed this thesis if not for hermentoring and kindness. I would also like to thank my committee members, Dr. Rhona Hanning,Dr. Samantha Meyer, and Dr. Mark Ferro, for their guidance and patience throughout thisjourney. Furthermore, I would like to thank Jane Russworm from the UW Writing andCommunication Center for her writing advice, and encouragement. Lastly, I would also like tothank my friends, lab mates, and family for their support.v

Table of ContentsList of Figures .ixList of Tables . xiiChapter 1- Introduction and Overview .1Chapter 2- Literature Review .32.1 Weight Loss Behaviours .32.2 Health Information Seeking Behaviours .82.3 Summary . 23Chapter 3- Study Rationale, Research Questions, and Hypotheses . 253.1 Study Rationale . 253.2 Research Questions . 263.3 Hypotheses . 26Chapter 4- Data and Variables . 284.1 Research Design. 284.2 Canada Food Study . 284.3 Measures . 304.4. Description of the Sample and Key Variables . 36vi

Chapter 5- Clustering of Weight Loss Behaviours and Use of Nutrition Information Sources. 455.1 Analytic Methods . 455.2. Results- Weight Loss Behaviours . 505.3 Results- Sources of Nutrition Information . 61Chapter 6- Associations Between Information Sources Used and Weight Loss Behaviours . 746.1 Analytic Methods . 746.1 Associations between Dietary Weight Loss Methods and Nutrition Information Sources . 766.2 Purging and Restrictive Weight Loss Methods and Nutrition Information Sources Used . 806.3 Non-Prescribed Supplements and Formulas and Sources of Nutrition Information . 836.4 Health-Promoting Weight Loss Behaviours and Sources of Nutrition Information . 87Chapter 7- Discussion . 917.1 Co-Occurrence of Health-Compromising and -Promoting Weight Loss Behaviours . 917.2 Co-Occurrence of Credible and Less-Credible Sources for Nutrition Information . 937.3 Association Between Sources of Nutrition Information and Weight Loss Behaviours . 967.4 Limitations. 1017.5 Implications for Future Research . 1037.6 Implications for Public Health Policy and Practice. 105vii

Chapter 8- Conclusion. 107References . 108Appendices . 125Appendix A- Supplementary tables and figures. 126Appendix B- Relevant questions from wave 1 CFS . 135viii

List of FiguresFigure 1. Model by Anker, Reinhart, and Feeley (2011) demonstrating health informationseeking mechanisms. 10Figure 2. Weight loss behaviours reported among weighted analytic sample (N 1,452.15) . 43Figure 3. Sources of nutrition information reported by weighted analytic sample (N 1,452.15)44Figure 4. Structural diagram and factor loadings for Dietary Changes . 54Figure 5. Number of weight loss behaviours within the factor Dietary Changes reported amongyoung adults, Canada Food Study (N 1,452.15) . 55Figure 6. Structural diagram and factor loadings for Purging and Restrictive Behaviours . 56Figure 7. Number of weight loss behaviours within the factor Purging and Restrictive Behavioursreported among young adults, Canada Food Study (N 1,452.15) . 56Figure 8. Structural diagram and factor loadings for Non-Prescribed Supplements and Formulas. 57Figure 9. Number of behaviours within the factor Non-Prescribed Supplements and Formulasreported by young adults, Canada Food Study, (N 1,452.15) . 58Figure 10. Structural diagram and factor loadings for Health-Promoting Behaviours . 59ix

Figure 11. Number of behaviours within the Factor Health-Promoting Behaviours reportedamong young adults, Canada Food Study (N 1,452.15) . 60Figure 12. Simplified structural diagram of associations among weight loss behaviour patterns 61Figure 13. Structural diagram and factor loadings for Government and Health AssociationMaterials . 65Figure 14. Number of Government and Health Association Materials used for nutritioninformation by analytic sample (N 1,452.15) . 66Figure 15. Structural diagram and factor loadings for Health and Weight Loss Specialists . 67Figure 16. Number of Health and Weight Loss Specialists used for nutrition information byanalytic sample (N 1,452.15) . 68Figure 17. Structural diagram and factor loadings for Commercial Sources . 69Figure 18. Number of Commercial Sources used for nutrition information by analytic sample(N 1,452.15) . 70Figure 19. Structural diagram and factor loadings for Easily Accessible Sources . 71Figure 20. Number of Easily Accessible Sources used for nutrition information by analyticsample (N 1,452.15) . 72Figure 21. Simplified structural diagram of associations among sources of nutrition information. 73x

Figure 22. Structural model of weight loss behaviour factors . 128Figure 23. Structural model of sources of information factors . 129xi

List of TablesTable 1. Socio-Demographic characteristics of analytic weighted sample (N 1,452.15) . 37Table 2. Independent t-test comparing socio-demographic variables between analytic sample(N 1538) and rest of sample (N 1325) . 39Table 3. Chi-square tests comparing socio-demographic variables between analytic sample(N 1538) and rest of sample (N 1325) . 39Table 4. Characteristic differences between analytic sample (N 1538) and others (N 1325) . 41Table 5. Summary of the 5-step Factor Analysis Protocol by Williams, Onsman & Brown(2010), detailing methods and steps used in this analysis . 47Table 6. Factor loadings for four-factor confirmatory factor model for weight loss behaviours 53Table 7. Factor loadings for four-factor confirmatory factor model for sources of nutritioninformation . 64Table 8. Poisson multiple regression investigating relationship between sources used fornutrition information (proportion) and number of Dietary Changes made to lose weight in thepast year (N 1452.15) . 77Table 9. Poisson multiple regression investigating variables associated with the number ofDietary Changes made to lose weight in the past year (N 1452.15). 79xii

Table 10. Poisson multiple regression investigating relationship between sources used fornutrition information (proportion) and number of Purging and Restrictive Behaviours used tolose weight in the past year (N 1452.15) . 81Table 11. Poisson multiple regression investigating variables associated with the number ofPurging and Restrictive Behaviours used to lose weight in the past year (N 1452.15) . 82Table 12. Poisson multiple regression investigating relationship between sources used fornutrition information (proportion) and number of Non-Prescribed Supplement and Formulas usedto lose weight in the past year (N 1452.15) . 84Table 13. Poisson multiple regression investigating variables associated with the number ofNon-Prescribed Supplements and Formulas used to lose weight in the past year (N 1452.15) . 86Table 14. Poisson multiple regression investigating relationship between sources used fornutrition information (proportion) and number of Health-Promoting Behaviours used to loseweight in the past year (N 1452.15) . 88Table 15. Poisson multiple regression investigating variables associated with the number ofHealth-Promoting Behaviours used to lose weight in the past year (N 1452.15). 89Table 16. Factorability table of weight loss behaviour responses demonstrating correlationsamong response categories . 126Table 17. Factorability table of sources of nutrition information responses demonstratingcorrelations among response categories. 127xiii

Table 18. Zero-inflated Poisson regression for Purging and Restrictive Behaviours . 130Table 19. Zero-inflated Poisson regression for Purging and Restrictive Behaviours withcovariates . 130Table 20. Zero-inflated Poisson regression for Non-Prescribed Supplements or Formulas . 132Table 21. Zero-inflated Poisson regression for Non-Prescribed Supplements or Formulas withcovariates . 132xiv

Chapter 1- Introduction and OverviewAttempting to lose weight is an extremely pervasive behaviour among young Canadianadults; recent online surveys of students and staff (mean age, 28 years) at a Canadian universityfound that more than one in three women and one in five men reported currently trying to loseweight (1). This is of concern since health behaviours developed during this period of time maypersist over the lifetime (2). Over the short- and long-term, weight loss attempts are associatedwith negative implications, such as body dissatisfaction, risk of developing an eating disorder,and depression (3). Additionally, those who engage in dieting tend to regain more weight thanwas lost (4).Young Canadians engage in a variety of weight loss behaviours, which include dieting(dietary restraint and/or dietary change), exercising, supplement use, among others (1), withmultiple behaviours often used in combination (5). These behaviours may include those thatcould be conceptualized as more healthy (e.g., eating more fruits and vegetables, relatedbehaviours such as exercising) or less healthy (e.g., fasting, using diet pills, smoking), althoughsuch conceptualizations are arbitrary since the degree to which individuals restrict certain foodsor beverages or exercise is an important consideration in terms of the health-promoting or –compromising potential (5,6).The behaviours that individuals engage in to try to lose weight may be influenced by theinformation sources they seek or are exposed to in relation to nutrition, diet, and weight. Forexample, surveys of American women ages 16-24 years have shown that those who reportedseeking weight loss information on the internet were more likely than those who did not use the1

internet for this purpose to exercise, use laxatives, use diet pills, vomit after eating, skip meals,smoke more cigarettes, and avoid eating carbohydrates to lose weight (7). Credible informationsources can be thought of those that provide unbiased, up-to-date, evidence-based claims from anexpert in the field or a trusted institution (8,9). However, it is not known whether individuals usemultiple sources in combination (credible along with less credible), and how this relates to theweight loss behaviours in which they engage.The following literature review provides an overview of research related to weight lossbehaviours and weight loss information sources. This is followed by Chapter 3, which describesthe research questions and hypotheses. In Chapter 4, the data source and variables are described,followed by an overview of the sample in terms of the key variables. In chapter 5, the methodsand results of factor analyses to examine patterns of weight loss behaviours and nutritioninformation sources consulted is presented. Chapter 6 presents the regression analyses that wereused to examine associations between weight loss behaviours and nutrition information sourcesused. Chapter 7 then provides an integrated discussion of the findings and their implications, aswell as highlighting directions for future research.2

Chapter 2- Literature Review2.1 Weight Loss BehavioursThe following provides an overview of weight loss, or dieting, behaviours used by bothadults and adolescents, as little research has examined the period of transition betweenadolescence and adulthood, the population of interest for the study described here.There are a variety of weight loss behaviours that are currently being used by both adultand adolescents. Among American adults, prevalent weight loss behaviours include eating fewercalories, eating less fat, exercising (10), switching to lower calorie foods, and drinking morewater (11). Meal skipping, enrolling in weight loss programs, and the use of diet pills, diureticsand supplements are less common but nonetheless used (10,11). Among American collegewomen aged 18-24, the most common weight loss behaviours reported included exercising,eating low-fat or fat-free foods, eating less than they wanted to, and consuming fewer sugarydrinks and foods (12). It is estimated that there are over 1000 different commercial weight lossdiets (e.g., Atkins, Zone, Weight Watchers, Mediterranean Diet) available, with more continuingto appear (13); the most commonly followed commercial diet used among college women in2006 were Atkins or South Beach and Weight Watchers (12).Despite the number available, diets (which typically aim to achieve the goal of calorierestricting) have been shown to be largely ineffective as weight loss is rarely maintained, andweight regain is often greater than the amount lost (4). In fact, among adolescents with obesity,those who reported dieting (e.g., fasting, using a food substitute) at baseline were 50% as likely3

to achieve a lower body weight (transition from obesity to overweight or non-overweight BMIclassification) 10 years later compared to those who had not dieted (14).A range of factors have been found to be associated with engagement in weight lossbehaviours. These include body mass index (BMI), gender, age, race/ethnicity, socioeconomicstatus, body image and satisfaction, and health literacy. Individuals with overweight and obesityare more likely to engage in dieting than individuals defined as normal weight (10,15). However,research among American college women aged 18-24 found that, although slightly more womenwith overweight (91%) and obesity (86%) reported ever having dieted, 80% of womenconsidered normal weight also reported dieting, indicating that regardless of weight class, dietingis very prevalent (15). Specific behaviours that are more prevalent among individuals withoverweight or obesity when compared to their normal weight peers include reducing calorieconsumption, eating less than they wanted to, and the use of unhealthful weight managementbehaviours (e.g., laxatives, diuretics and diet pills) (12,16,17,18). As well, engagement in healthpromoting weight loss behaviours such as consumption of fruits and vegetables and exercisingare lower among young Americans in grades 7, 9 and 11 with overweight or obesity compared totheir non-overweight peers (10,16). On the other hand, among adult American women aged 1640 years, those with overweight or obesity were shown to be more likely to engage in healthyweight loss behaviours (6) when compared to those defined as normal weight.Some studies suggest that women are more likely to reduce calorie consumption, eat lessfat, join a weight loss program, and use diet pills whereas men are more likely to skip meals andexercise to lose weight (10,11). Age is also relevant. A study of Canadian females found that theprevalence of unhealthy weight loss behaviours (e.g., binge eating, vomiting and diet pills)4

increased through adolescence (age 12 to 18), and that early adulthood is a period during whichweight concerns and poor dieting behaviours may increase (2,19,20).A relationship also exists between race/ethnicity and weight loss behaviours; a crosssectional analysis of 2007-2012 NHANES data of American adults age 20-65 found that nonHispanic Blacks were less likely to attempt weight loss than non-Hispanic Whites and also lesslikely to change diet (e.g., ate less food, ate less fat, consuming low-calorie foods) or exercise tolose weight and less likely to consult with professionals regarding weight loss (21). Racial/ethnicdifferences in weight loss attempts and dieting behaviours may reflect differences in what isconsidered the ideal body according to identified culture (22). Further, cross-sectional researchwith middle-aged Australian adults found that those of low-income and low-education weremore likely to have a higher BMI (ratio of body weight to height used to classify by body status)and less likely to engage in weight-control behaviours when compared to their more-advantagedpeers (23). Similar weight-control strategies were used among all income-classes, except thatlow-income adults were less likely to engage in exercising to control weight, and more likely tominimize their sitting time (23).Among adolescents, body image and dissatisfaction are also strong predictors of dieting(changing eating behaviours to lose weight) (15). For example, poor body image has been foundto predict engagement in dieting and unhealthy weight loss strategies at a 5 year follow upamong American adolescents (18). Although body dissatisfaction is most common in women, itis also high among men (24). Body dissatisfaction has also been found to be associated withrace; among adolescents from Project EAT 2010 (a study of the eating, exercising, and weightrelated behaviours of Minnesota middle and high school students), Asian American girls and5

boys reported the highest rates of body dissatisfaction. Additionally, among boys in this sample,the relationship between body dissatisfaction and unhealthy weight control behaviours wasmoderated by race (25).Health literacy, one’s ability to understand health information and factors that maypromote or hinder health (26), may also impact the types of weight loss strategies used. A mixedmethods study of African American women found that those with adequate health literacy were 2times more likely to exercise to lose weight than those of low health literacy (27). This studyincluded nine possible weight loss behaviours (e.g., fasting, reducing fried foods, laxatives,consuming less sweets); however, exercise was the only variable that differed significantly byhealth literacy (27).2.1.1 Conceptualizing the Health-Promoting or Health-Compromising Potentialof Weight Loss BehavioursWeight loss behaviours are often categorized as “healthy” or “unhealthy” (6,28). Thesecategories are subjective; for example, the use of food substitutes (powder or drink) has beencategorized as both unhealthy and healthy (6,29). One reason that these behaviours cannot besimply categorized as healthy or not is because the degree/dose can affect whether and the extentto which each is health promoting or compromising. For example, exercising is generallyconsidered an important aspect of a healthy lifestyle (30). However, certain patterns ofexercising behaviours are associated with disordered eating behaviours (31). Indeed, exercise canbe health compromising when used compulsively (e.g., to reduce distress regarding weight or6

engaged in despite injury or illness) or when used as a compensatory technique (e.g., tocompensate for food intake). A cohort study of university men and women found thatcompulsive and compensatory exercise predicted disordered eating and eating disorder diagnosis(31). Therefore, the occurrence of this behaviour, which has been considered a “healthy”behaviour, may actually be a health-compromising behaviour in particular circumstances.A more nuanced way of looking at weight behaviours may be consistent with ourunderstanding of the complexity of health behaviours. For instance, research demonstrates thathealth behaviours cluster among adolescents and adults (32,33). A systematic review of researchexamining the clustering of four health risk behaviours (smoking, excess alcohol consumption,poor nutrition, and physical inactivity) found that these behaviours co-occur (34); for example,individuals engaging in smoking are more likely to engage in other risky behaviours as well.Despite evidence that health behaviours co-occur, weight loss behaviours continue to beconsidered in isolation.Although, to the authors knowledge, a formal study of the associations among weightloss behaviours has not been conducted until now, studies have utilized cluster analysis whichmay give insight into which behaviours are used simultaneously. Cluster analysis has been usedto group individuals based on similarity among individual characteristics related to weightcontrol behaviours (35,36). Cluster analysis of Australian women aged 22-27 years identified 4unique cluster groups based on common behaviours and characteristics among these women:Dieters, Healthy Living, Do Nothing, and Perpetual Dieters (36). The largest cluster was Dieters,90% of whom tried to lose weight in the past year, the majority controlled their weight throughhealthy weight management behaviours (e.g., reduced meal sizes, reduced fat and sugar intake7

and vigorous physical activity). Women in the Healthy Living cluster seldom reported engagingin weight loss attempts, weight control strategies included vigorous exercise, cutting down onmeal sizes, reducing fat and sugar intake. The Do Nothing group reported not actively trying tolose weight, did not report using any of the weight control beh

Weight loss efforts are pervasive among young adults and worrisome as they are associated with poor mental health and development of eating disorders. Data on weight loss . use laxatives, use diet pills, vomit after eating, skip meals, smoke more cigarettes, and avoid eating carbohydrates to lose weight (7). Credible information

Related Documents:

The Nutrition Care Process is defined in four steps: 1. Nutrition Assessment 2. Nutrition Diagnosis 3. Nutrition Intervention 4. Nutrition Monitoring & Evaluation The first component of the “Nutrition Assessment” is a screening of residents for those at risk for nutrition problems and is a candidate for further intervention. One of the

Nutrition during a woman's life From: ACC/SCN and IFPRI. 4th Report on the World Nutrition Situation: Nutrition Throughout the Life Cycle. Geneva: WHO, 2000. Nutrition during a woman's life From: ACC/SCN and IFPRI. 4th Report on the World Nutrition Situation: Nutrition Throughout the Life Cycle. Geneva: WHO, 2000.

Nutrition Evaluation: the systematic comparison of current findings with the previous status, nutrition intervention goals, effectiveness of overall nutrition care, or a reference standard Nutrition Care Outcomes: the results of nutrition care that are directly related to the nutrition diagnosis and the goals of the intervention plan

the Source 1 power source until the Source 2 power source does appear. Conversely, if connected to the Source 2 power source and the Source 2 power source fails while the Source 1 power source is still unavailable, the ATS remains connected to the Source 2 power source. ATSs automatically perform the transfer function and include three basic .

The Updated Nutrition Facts Label: Information for Health Educators Author: Center for Food Safety and Applied Nutrition Subject: 2019 Nutrition Facts Labeling Keywords: The Updated Nutrition Facts Label: Information for Health Educators,2019 Nutrition Facts Labeling, US FDA Created Date: 4/12/2021 2:26:13 PM

3. Regular community nutrition education (e.g. cooking demos, health talks and discussions) 4. Nutrition education in food and nutrition security projects 5. School nutrition education (curricular or extra-curricular) 6. School gardens (gloss) focusing on learning about nutrition 7. Information, education and communication (IEC) materials .

1. Nutrition inclusion in EAS 8 1.1 Different approaches integrating nutrition into EAS 8 1.2 Summary and lessons learned 15 2. Nutrition training of extension agents 16 2.1 The history of nutrition training in EAS 16 2.2 Examples of training and capacity development 16 2.3 Nutrition messages and information by EAS 17

Manual for Country-Level Nutrition Advocacy . Using PROFILES and Nutrition Costing. FANTA III. FOOD AND NUTRITION TECHNICAL A SSISTANCE. Food and Nutrition Technical Assistance III Project (FANTA) 1825 Connecticut Avenue, NW Washington, DC 20009 T: 202-884-8000 fantamail@fhi360.org www.fantaproject.org. 1. 3. Nutrition . 1 Costing. Multi-Sectoral