@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix dc: . @prefix skos: . vivo:departmentOrSchool "Education, Faculty of"@en, "Kinesiology, School of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Schwartz, Adam"@en ; dcterms:issued "2009-10-28T23:03:31Z"@en, "2003"@en ; vivo:relatedDegree "Master of Science - MSc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description "Background: The relationship between diet and human body composition has become a popular topic in recent years. Despite the popularity of vegetarianism little research has been conducted in the area of vegetarianism and body composition, and the focus of the minimal research to date has been on children and women. There is recent evidence that suggests that among older men, the ability to gain skeletal muscle with resistance training may be substantially diminished for individuals consuming a vegetarian versus an omnivorous diet. This issue has yet to be examined in a group of younger males. There is also some evidence to suggest that vegetarians may have lower levels of body fat in comparison to omnivores. Objective: The objective of this study was to identify body composition differences (muscle and fat) between young men consuming either a vegetarian or omnivorous diet. Design: 54 young males between 18-30 y participated in this study. Half of these subjects were vegetarians. Body fat and muscle mass were estimated from anthropometry. In addition, subjects completed the Three-Factor Eating Questionnaire (TFEQ) and the Godin Leisure-Time Exercise Questionnaire (GLTEQ), which were used as indicators of eating habits and activity levels. All subjects completed 3-day diet records. Results: Vegetarians had significantly lower predicted muscle mass (30.9kg vs. 32.7, p=0.049 1-tailed), and a lower sum of 12 corrected muscle girths compared to omnivores (2.5x10⁶cm³ vs. 2.7x10⁶cm³, p=0.033 1-tailed). Vegetarians had higher dietary intakes of fiber (139.8 vs. 92.3, p=0.006) and polyunsaturated fats (67.5 vs. 42.7, p=0.001) and lower intakes of saturated fats (93.8 vs. 129.1, p=0.031) than omnivores. There were no significant differences between groups with regard to body fat, dietary restraint (TFEQ), activity levels (GLTEQ and reported hours of weekly activity), or other dietary intake variables. Conclusion: Vegetarian men were found to have significantly lower muscle mass than omnivores, and these differences could not be accounted for by dietary restraint or activity levels."@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/14273?expand=metadata"@en ; dcterms:extent "6088232 bytes"@en ; dc:format "application/pdf"@en ; skos:note "BODY COMPOSITION OF VEGETARIAN AND OMNIVOROUS MEN by A D A M SCHWARTZ B.A., The Unversity of Western Ontario, 1999 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES The Faculty of Education; School of Human Kinetics; Exercise Physiology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 2003 © Adam Schwartz, 2003 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract Background: The relationship between diet and human body composition has become a popular topic in recent years. Despite the popularity of vegetarianism little research has been conducted in the area of vegetarianism and body composition, and the focus of the minimal research to date has been on children and women. There is recent evidence that suggests that among older men, the ability to gain skeletal muscle with resistance training may be substantially diminished for individuals consuming a vegetarian versus an omnivorous diet. This issue has yet to be examined in a group of younger males. There is also some evidence to suggest that vegetarians may have lower levels of body fat in comparison to omnivores. Objective: The objective of this study was to identify body composition differences (muscle and fat) between young men consuming either a vegetarian or omnivorous diet. Design: 54 young males between 18-30 y participated in this study. Half of these subjects were vegetarians. Body fat and muscle mass were estimated from anthropometry. In addition, subjects completed the Three-Factor Eating Questionnaire (TFEQ) and the Godin Leisure-Time Exercise Questionnaire (GLTEQ), which were used as indicators of eating habits and activity levels. Al l subjects completed 3-day diet records. Results: Vegetarians had significantly lower predicted muscle mass (30.9kg vs. 32.7, p=0.049 1-tailed), and a lower sum of 12 corrected muscle girths compared to omnivores (2.5xl06cm3 vs. 2.7xl06cm3, p=0.033 1-tailed). Vegetarians had higher dietary intakes of fiber (139.8 vs. 92.3, p=0.006) and polyunsaturated fats (67.5 vs. 42.7, p=0.001) and lower intakes of saturated fats (93.8 vs. 129.1, p=0.031) than omnivores. There were no significant differences between groups with regard to body fat, dietary restraint (TFEQ), activity levels (GLTEQ and reported hours of weekly activity), or other dietary intake variables. Conclusion: Vegetarian men were found to have significantly lower muscle mass than omnivores, and these differences could not be accounted for by dietary restraint or activity levels. Table Of Contents Abstract ii List of Tables • vi List of Figures vii 1. Introduction 1 1.1 Vegetarianism: a popular diet 1 1.2 Subclasses of vegetarian diets 1 1.3 Vegetarianism & body composition: possible repercussions 2 1.4 Gaps in the literature 3 1.5 Purpose of the study 4 2. Literature Review 6 2.1 Overview 6 2.2 Body composition of vegetarians vs. omnivores: studies in children 6 2.3 Body compositon of vegetarians vs. omnivores: adult populations 8 2.4 Body composition of vegetarians vs. omnivores: elderly populations 9 2.5 Body composition of vegetarians vs. omnivores: summary 10 2.6 Estimating skeletal muscle mass 10 2.7 Anthropometric prediction of body fat 15 2.8 Skinfold measurement of subcutaneous adipose tissue 16 2.9 Possible explanations for vegetarian vs. omnivore differences 16 2.10 Vegetarian vs. omnivore differences in protein, fat, fiber and possible influences on body composition 18 2.11 Conclusions 19 2.12 Research questions & hypotheses 20 2.12.1 Primary questions and hypotheses 20 2.12.2 Secondary research questions 20 3. Methods 22 3.1 Subjects 22 3.1.1 Vegetarian inclusion criteria 22 3.1.2 Vegetarian and omnivore inclusion criteria 22 3.2 Study design 23 3.3 Testing procedure 23 3.4 Questionnaires 24 3.4.1 Screening questionnaire 24 3.4.2 Three-day diet record 24 3.4.3 The three-factor eating questionnaire (TFEQ) 24 3.4.4 Godin leisure-time exercise questionnaire (GLTEQ) 25 3.5 Measurement of height and body mass 26 3.6 Anthropometric procedures 26 3.7 Variables of interest 26 3.7.1 Independent variable: diet 26 3.7.2 Dependent variables 27 3.7.2.1 Total body skeletal muscle mass 27 3.7.2.2 Sum of 12 skinfold-corrected muscle girths 27 3.7.2.3 Sum of 6 skinfolds 28 3.7.2.4 Percent fat predicted from anthropometry 28 3.7.2.5 Nutrient intakes from 3-day diet records 29 3.7.2.6 TFEQ restraint score 29 3.7.2.7 GLTEQ score 29 3.8 Statistical analysis 29 3.9 Statistical power & sample size determination 30 3.10 Ethical approval 30 4. Results 31 4.1 Body composition 31 4.2 Questionnaire scores 34 4.3 Correlations between dependent variables 35 4.4 Dietary intakes 36 5. Discussion 38 5.1 Body composition comparisons between dietary groups 38 5.2 Lifestyle comparisons between dietary groups 40 5.3 Correlations between dependent variables 41 5.4 Dietary intake comparisons 42 5.5 Summary & recommendations for future research 42 iv 6. References 44 7. Appendices 48 Appendix I: Informed consent 49 Appendix II: Ethics approval 51 Appendix III: Questionnaires 52 Screening questionnaire 53 The three-factor eating questionnaire (TFEQ) 54 Godin leisure-time exercise questionnaire (GLTEQ) 58 Appendix IV: Anthropometric proforma 59 Appendix V: Distribution of dependent variables 60 Appendix VI: Instructions for 3-day diet records 66 Appendix VII: One-way ANOVA comparison of means 67 Appendix VIII: Mann-Whitney U-Tests 70 Appendix IX: Spearman correlation coefficients 72 Appendix X: ANCOVA comparison of muscle mass 73 v List of Tables Number and Title Table 4.1.1: Physical characteristics and body composition measurements for vegetarian and omnivorous groups 31 Table 4.1.2: Skewness and kurtosis values for body composition results 33 Table 4.2.1: Skewness and kurtosis values for questionnaire results 34 Table 4.2.2: Questionnaire scores and reported hours of moderate activity for vegetarian and omnivorous groups 35 Table 4.3.1: Spearman correlation coefficients of questionnaire scores with body composition variables 36 Table 4.4.1: Skewness and kurtosis values for dietary intake variables by group 37 Table 4.4.2: Three-day dietary intakes compared between dietary groups 37 vi List of Figures Number and Title 4.1.1: Boxplots of linearly corrected muscle mass of vegetarians and omnivore 32 4.1.2: Boxplots of sum of 12 muscle girths of vegetarians and omnivores 32 4.1.3: Boxplots of sum of 6 skinfolds of vegetarians and omnivores 33 4.1.4: Boxplots of percent fat of vegetarians and omnivores 33 4.3.1: Scatterplot of relationship between GLTEQ score and percent fat 36 vn 1. INTRODUCTION /./ Vegetarianism: a popular diet The vegetarian diet is a popular dietary choice for people of all ages. In Canada 1997, the National Nutrition Survey showed that 4% of people define themselves as vegetarian (1). Reasons for selecting this diet include religious beliefs, philosophical, ecological and health concerns (2). To date, very little is actually known about the impact of vegetarian diets on some important aspects of human body composition. Understanding the relationship between a vegetarian diet and the specific components of body composition would be useful for assessing the adequacy of this diet and issues related to health and sport training. If vegetarianism has a relationship with muscle, fat or bone, this would have important implications for dieters, children, the elderly and athletes. 1.2 Subclasses of vegetarian diets The term \"vegetarian diet\" refers to a broad classification of plant based diets that are usually further divided into subclasses. The common vegetarian subclasses include: lacto-ovo-vegetarian (LOV), lacto-vegetarian (LV), and vegan. The LOV consumes milk products and eggs and excludes any meat, poultry and fish (2). The LV avoids eggs, meat, poultry and fish (2). The vegan excludes milk and any milk product in addition to meat and eggs, and in some cases any product of animal origin (2). The macrobiotic diet is a diet that doesn't necessarily exclude, but strictly limits animal foods including dairy products, and from nutritional profiles is often considered amongst vegetarian diets (2). Some people who consume fish and or poultry might consider themselves to be \"vegetarians\", however these individuals would more appropriately be classified as semi or partial vegetarians (2). 1 1.3 Vegetarianism & body composition: possible repercussions Body composition is a broad term that generally speaks collectively of the gross tissues which include adipose tissue, muscle, bone, and skin (3). The present consideration is to determine whether the vegetarian diet has a relationship with one or more of the components of body composition. The LOV diet is the least extreme of the vegetarian diets, allowing for the consumption of dairy and eggs. The LOV diet has recently been called into question regarding a reported relationship between the consumption of this diet and the muscle mass response to strength training interventions. In a study of older adults, researchers found that consumption of a LOV diet during a 12 week resistance training (RT) program did not result in the expected muscle hypertrophy (4). Subjects in this study were randomly assigned to consume either a higher protein LOV diet or a lower protein LOV diet. After 12 weeks of RT there was no significant increase in either whole body muscle mass or in mid-thigh muscle in either group. The researchers hypothesized that meat in the diet may have had a direct influence on the ability of older adults to gain muscle with RT. This research group decided to further investigate by comparing skeletal muscle gains with 12 weeks of RT between a group of older men who were omnivores and a group of men who changed from their typical meat containing diet to a LOV diet (5). After 12 weeks of training, omnivores showed gains in whole body muscle mass, while the LOV group showed losses with respect to this variable. No significant differences were found in mean dietary energy or macronutrient intakes between the groups. These results indicated that the absence of meat in the diet negatively influenced muscle mass response to RT in older men. Body fat is another topic of interest with regard to vegetarianism. When comparing vegetarian and omnivorous groups, vegetarians are found to be the leaner of the two (6-10). 2 An inherent weakness of these studies is their use of few skinfold sites, which failed to account for differences in fat patterning between individuals. The possibility of a vegetarian diet impacting bone is an important consideration for populations such as growing children and at risk elderly populations who suffer from reduced bone mass and density. In a study of 195 Dutch adolescents; 93 had followed a macrobiotic diet for on average 6-years during childhood. Whole body bone mineral content (measured by dual x-ray absorptiometry) was significantly lower among those consuming macrobiotic diets compared with controls (11). The between-group differences could not be accounted for by current calcium intakes or physical activity levels, and researchers proposed that different diets during childhood provided the best explanation. This study raises some important concerns about the possible repercussions of a childhood diet on bone development in later years. A general overview of childhood growth can be examined through the dietary effects on height and weight. Nathan and colleagues examined 50 vegetarian and 50 age, sex, and ethnic-group matched omnivorous children (age 7-11) for a one year period, and found no differences in height or weight at baseline or after one year (12). 1.4 Gaps in the literature Currently, very few studies have investigated the relationship between vegetarianism and body composition in men. As a result of past difficulties in affordable and accurate methods for quantifying skeletal muscle, little is presently known about the relationship between vegetarianism and skeletal muscle mass. A recent model has been developed in an attempt to remedy this (13). Lee et al. (13) have developed an anthropometric prediction equation (using skinfolds and muscle girths) for predicting total body skeletal muscle mass 3 (SMM), derived using multiple magnetic resonance imaging (MRI) scans. This model has not previously been used in a study of vegetarianism and body composition. In studies that looked at the relationship between vegetarianism and some indicator of fat or muscle, most have relied on crude measurements of these components. Studies on this topic that measured skinfolds, often relied on 3 or fewer subcutaneous sites to draw conclusions about body fat (6-8, 10, 12, 14, 15). Researchers have also made attempts to assess muscle using a single skinfold-corrected girth (6, 7,12). These studies provide only a site-specific analysis, and should be interpreted with great caution. To our knowledge there are no studies that have compared fat mass and muscle mass between a vegetarian and omnivorous group of young men. 1.5 Purpose of the study This study will compare the differences in body composition between young adult male vegetarians and young adult male omnivores. This study will assess whether there are differences in fat mass and skeletal muscle mass between these two groups. By observing groups of younger males (rather than older males) with relatively large amounts of skeletal muscle mass, if a difference exists between vegetarians and omnivores it would be magnified in this population. To our knowledge this is the first study on this topic to use the recent Lee et al. (13) prediction equation for quantifying skeletal muscle mass using simple anthropometric measurements. There is limited research on the topic of vegetarianism and body composition, and this study is the first to investigate a population of young adult males. A significant difference between dietary groups with regard to the components studied (skeletal muscle, body fat) would serve as the foundation on which to base future experimental research. Little is known about the relationship between a vegetarian diet^ aod 4 selected body composition outcomes, therefore it is important to address the adequacy of these diets. This study will add to the small body of knowledge in this area. 5 2. LITERATURE REVIEW 2.1 Overview There are few studies to choose from when examining the topic of vegetarianism and body composition, therefore this review will include any study which examined a vegetable-based (no meat) diet and looked at one or more body composition variables. The populations of focus for these studies have largely been children and women, but they reveal some important trends that may also be applicable to a group of adult males. In an effort to gain a greater understanding of the health benefits or problems associated with a vegetarian diet, a good place to start is the impact of this diet on the body composition of children. This will be followed by an examination of studies comparing body composition differences between vegetarians and omnivores among adult and finally elderly populations. Important questions to consider for all these populations of vegetarians include: are vegetarians leaner, and do they have less muscle mass than omnivorous controls? It should be noted that many of the studies that address these questions are cross-sectional in nature, making it impossible to attribute diet as the only cause of observed differences. 2.2 Body composition of vegetarians vs. omnivores: studies in children If vegetarian children are leaner than omnivores, then this may suggest a potential strategy for dealing with the increasing numbers of obese children. However, it is also important that there be no detrimental impact on muscle. A study by Tayter and Stanek (6) compared the triceps skinfold and the skinfold-corrected mid-arm circumference in 17 LOV and 22 omnivorous boys and girls aged 10-12. The triceps skinfold (used as an indicator of body fatness) was significantly higher among omnivorous boys but not girls, and no 6 differences were found in skinfold-corrected mid-arm circumference (used as an indicator of muscle) between vegetarian and omnivorous boys and girls. Unfortunately, this study only measured a single skinfold and skinfo ld-corrected girth. Drawing conclusions about total body muscle and fat differences based on these single measures is difficult. In an effort to understand the effect of diet on the body composition of children over time, Nathan and colleagues (12) examined anthropometric variables over the course of one year in 50 vegetarian and 50 age, sex, and ethnic group matched 7-11 year olds. Investigators measured height, weight, biceps and triceps skinfolds, and upper arm circumference, at baseline and after one year. No significant differences were found between groups at baseline or after 1 year in any of the measured variables. The longitudinal nature of this study makes it more methodologically sound, but its inherent weakness is the lack of measures taken (skinfolds and muscle girths) that would have provided a much clearer picture and perhaps identified important group differences. In support of these findings, Dwyer et al. (14) found no significant differences in the triceps skinfold in a cross-sectional study of 39 pre-school aged children consuming various vegetarian diets, compared to age matched norms. Other evidence suggests that significant differences in fat and muscle do exist between vegetarian and omnivorous children. In an earlier study by Dwyer et al. (7) comparing 142 pre-school vegetarian children (77 macrobiotic, 65 other vegetarian groups) to norms, significant differences were observed between vegetarian children and normative data. Triceps and subscapular skinfolds were measured in addition to a skinfo ld-corrected arm muscle circumference. Macrobiotic and other vegetarian children who were more than a year old had significantly smaller subscapular skinfolds compared to the normative data. Arm muscle circumference in both vegetarian groups was significantly smaller than norms for children 3 years or older. Smaller skinfold measurements were also observed among female 7 vegetarian children in a study by Hebbelinck et al. (8). Flemish female LOV children had lower triceps and suprailiac but not calf skinfolds, than reference data for age matched Flemish children. A criticism of this study is the small number of vegetarian children included (5 males, 5 females), which may have resulted in an erroneous null finding in the males, due to a lack of statistical power for detecting a difference. This study also examined some anthropometric variables in vegetarian adolescents (n=19), finding lower skinfolds in both males (triceps, suprailiac) and females (calf, suprailiac). Energy intakes of vegetarians were significantly lower than reference data, which may explain the subcutaneous fat differences. 2.3 Body composition of vegetarians vs. omnivores: adult populations Body composition research undertaken on adult populations also suffers from a lack of study. The greater leanness (lower skinfolds) of adult vegetarians is often assumed, but this is not a universal finding. Howie and Schultz (15) measured the chest, abdomen, and thigh skinfolds in 12 vegetarian and 18 non-vegetarian men between the ages of49-62, and found no significant differences between groups. Hebbelinck et al. (8) considered vegetarianism and body fat differences, taking skinfold measurements (triceps, calf, suprailiac) on young adult Flemish male (n=8) and female (n=l 1) vegetarians, but found no significant differences in any of the skinfolds, between vegetarians and reference values. However, both of these studies used small sample sizes, and as a result may have missed an important difference. Barr et al. (9) measured skinfolds (triceps, abdominal, suprailiac, and thigh) in 23 healthy vegetarian women and 22 omnivorous women and found predicted body fat percent was lower in vegetarians than controls (24.0 ± 5.5% vs. 27.4 ± 5.1%, p<0.05). 8 2.4 Body composition of vegetarians vs. omnivores: elderly populations Just as there are concerns for children, there are some important issues concerning the practice of a vegetarian diet among elderly populations, and the impact on body composition. With increasing age, which is often accompanied by a decrease in activity, elderly individuals are at risk of having lowered levels of muscle mass, and the adequacy of a vegetarian diet in optimally maintaining these components of the body must be assessed. One advantage of conducting diet-related studies in the elderly is that often these individuals have been adhering to a particular diet for a long time. This can help provide insight into the long-term effects of a vegetarian diet on body composition. The issue of fatness has not been well studied in elderly vegetarians. Anthropometric variables were measured in one study of elderly females (10). Three skinfolds (triceps, suprailiac, and thigh) and % fat estimated by quadratic regression were examined in 12 postmenopausal Caucasian vegetarian (mixed types) women and 12 postmenopausal Caucasian non-vegetarian women. The non-vegetarians had significantly higher thigh skinfolds (39.9mm vs. 30.8mm, p<0.01) and sum of three skinfolds than the vegetarian group (85.0mm vs. 67.2mm, p<0.05). These results reveal substantial differences in body fat between elderly vegetarians and non-vegetarians, but certainly the replication of these results is necessary before any definitive statements can be made. The individuals in this study were all Seventh Day Adventists, and the likelihood is that the lifestyle of these individuals is different from other vegetarians. Thus, these results may not be generalizeable to a wider population of vegetarians. Some interesting studies regarding the influence of a vegetarian diet on muscle mass during resistance training have been observed in elderly populations. The two studies by Campbell et al (previously mentioned in section 1.3) revealed that consumption of a meat-free 9 diet during a 12 week RT program does not lead to the expected gains in muscle hypertrophy or muscle mass (4, 5). Cross-sectional comparisons of muscle between vegetarians and omnivores have been neglected in an elderly population. 2.5 Body composition of vegetarians vs. omnivores: summary After investigating the small number of studies that have compared body composition variables (muscle and fat) among vegetarians and omnivores, it is difficult to draw any definitive conclusions. It seems likely that vegetarians are leaner than omnivores (6, 7, 9, 10) but many of the studies to date comparing fat differences have used only a few skinfold (3 or less) sites to assess body fatness (6-8, 10, 12, 14, 15). More comprehensive research accounting for all major body fat stores is necessary. Assessing whether or not there are muscle differences between the two dietary groups is the most difficult task of all. The cross-sectional studies that considered a muscle comparison relied on a single skinfold-corrected (7) or uncorrected (12) muscle girth of the arm as their sole measurement. Although corrected arm girth has been found to have a good correlation with criterion measures of total muscle mass (13), squaring this corrected girth and adjusting for stature to yield a three-dimensional measure shows an improved correlation (13). Surprisingly, cross-sectional studies done on adults and the elderly have ignored a comparison of muscle all together. 2.6 Estimating skeletal muscle mass A n important variable of body composition that has been almost completely neglected on the topic of vegetarianism is skeletal muscle mass. The reason has long been an absence of affordable and accurate techniques to take this measurement. Skeletal muscle masses obtained through dissection are known for only 25 men (3), meaning that any current methods 10 that have been developed and validated rely on a physical or chemical property of muscle to obtain their measurement (17), and therefore are subject to some degree of error. Computed tomography (CT) and magnetic resonance imaging (MRI) are widely regarded as the most accurate techniques to measure muscle mass in vivo. However, each of these methods have limitations. CT relies on the passage of X-rays through the body, which are attenuated. The degree of attenuation of the x-rays is related to the density of the tissues it has passed through. A computer then constructs a visual image (cross-sectional image) using the attenuated values of the x-rays. Based on different attenuation values for components of body composition such as adipose tissue, bone, and muscle, these components can be assessed independently (17). The tissue of interest is traced on the computer yielding a cross-sectional area. The thickness of the image (slice) is a known value, and with this value tissue volumes can be calculated. However, even with a set of slices taken at intervals covering most of the body, a calculation of total body muscle mass requires filling in the blank unmeasured sections with predicted values. Another problem encountered with using CT to calculate muscle mass is that the attenuation of normal muscle widely varies, depending on which muscle group is measured (17). A further drawback of this technique is that it requires x-ray exposure, and the use of radiation in body composition studies involving healthy individuals would be considered unethical. Finally, this method is very costly which serves to further narrow its scope of use in body composition research. Similar to CT, MRI can be used to assess regional muscle and to predict total body muscle by filling in the blank unmeasured sections. This method uses a magnetic field to align the nuclei in a body segment. A radio frequency electromagnetic wave is then directed through the body, and some of the nuclei absorb energy. The wave is then shut off, and the nuclei that have absorbed its energy emit the radio signal. A computer then transforms this 11 signal to an image of the chemical composition of the tissue (17). This method also yields a cross-sectional image with a known thickness, making volume calculations possible. In contrast to CT, MRI does not involve the use of radiation, however this technique finds little use in body composition studies due to its considerable expense, but might serve as a criterion method. Techniques that have been more readily used in body composition research to measure muscle include the measurement of muscle metabolites. One of the popularly measured metabolites is creatinine. Creatinine is a waste product of creatine which resides primarily in skeletal muscle (98%) (18) in (17). To obtain a measurement of muscle mass, first 24-hour urinary creatinine is measured, then assuming a constant ratio of urinary creatinine to skeletal muscle tissue, a total muscle mass can be calculated. However, there is some debate as to the ratio of urinary creatine to muscle mass (17), and there is variability in daily urinary creatinine excretion within an individual who consumes a self-selected diet (17). These variables leave this method of muscle estimation open to criticism. In an attempt to provide a simple method to measure skeletal muscle mass, the Brussels cadaver study (3) was the first study to dissect cadavers for tissue masses in addition to taking anthropometric measurements of the same cadavers. Regression equations were then developed for the estimation of whole-body muscle mass based on the anthropometric measures (19). The authors determined that regional anthropometric measurements were very good indicators of total dissected muscle mass, and the following equation was derived for predicting muscle mass in men, M M = STAT (0.0546 C T G 2 = 0.119 F G 2 + 0.0256 CCG 2 ) -2980. In this equation M M is muscle mass in kg, STAT is equal to stature in cm, CTG is corrected thigh girth, F G is uncorrected forearm girth, and C C G is corrected calf girth. This 12 equation had an R 2 =0.93 and SEE =1.58kg. However, the limitation of this equation is that it was generated on a small number of older cadavers, and as a result is very sample specific. Recently, researchers have sought to build on this foundation and provide a simple answer to the problem concerning muscle measurement. In a manner similar to the earlier anthropometric prediction equation based on cadavers (19), Lee et al. (13) developed a simple anthropometric prediction model for estimating total body skeletal mass, based on MRI measured muscle mass. The instruments required to take anthropometric measurements are both inexpensive and portable making this an attractive approach for skeletal muscle mass prediction. The model was developed in a group of non-obese adults (n=244, 139 men and 109 women) with a wide age range (20-81 years old). In this study, equations were developed taking into account height, age, sex, and race parameters in addition to skinfold-corrected muscle girths from the arm, thigh, and calf. To obtain a skinfo ld-corrected muscle girth, first a limb circumference measurement is taken, a skinfold measurement is then used to correct the cross-sectional area for adipose tissue. The result is a cross-sectional area that represents muscle. This method of estimation assumes that the measured limb segments are circular and that the portion of the cross-section that is bone is negligible. Slcinfold-corrected muscle girths were calculated using the following equation where muscle circumference (Cm) = measured limb circumference (Cumb) - n x skinfold (SF) (19). The best equation generated for predicting total body skeletal muscle mass by Lee's study is: S M (kg) = Ht x (0.00744 x C A G 2 + 0.00088 x C T G 2 + 0.00441 x CCG 2 ) + 2.4 x sex - 0.048 x age + race + 7.8, where S M is skeletal muscle, Ht is height in meters, C A G is corrected arm girth in cm, CTG is corrected thigh girth, and C C G is corrected calf girth. This equation has a high R 2 value (0.91), and a low standard error of 2.2 kg. It must be noted that the skinfo ld-corrected muscle girths cannot account for intramuscular fat, and the authors did not make adjustments for bone 13 that passes through the extremities, which makes up 5-10% of the cross-sectional area (20). A further limitation of Lee's method for predicting SMM is that the model is population specific. The sample did include a wide age range and was racially diverse. However, the non-obese model developed would not be appropriate for use in highly trained athletes, body builders, obese individuals, or patients with anorexia nervosa. Another simple measure of skeletal muscle can be obtained through a sum or set of skinfo ld-corrected girths. Previous studies have shown that certain skinfo ld-corrected limb girths show a strong correlation with dissected measures of muscle mass (19), and MRI measured muscle mass (13). In a study of male cadavers (aged 50-94) complete anthropometry, dissection, and weighing of body segments, showed a high correlation coefficient between skinfo ld-corrected girths and total muscle mass from dissection (19). In 6 unembalmed male cadavers, corrected girths had correlation coefficients of 0.896 in the arm, 0.998 in the forearm, 0.990 in the thigh, and 0.911 in the calf. In a larger sample of younger non-obese men and women (n=244), high correlation coefficients were observed between MRI measured total body skeletal muscle mass and skinfo ld-corrected girths (adjusted for height) in the arm (r=0.90, SEE 3.19kg2), thigh (r=0.83, SEE 4.18 kg2), and calf (r=0.87, SEE 3.74 kg ) (13). Some authors believe that a central weakness of skinfo ld-corrected muscle girths is that often only one or two measurements are used as indices of muscle mass (19). The result is a failure to account for individual differences in muscle distribution. It has been shown that despite some limitations simple anthropometric measurements have been validated as good predictors of skeletal muscle, and researchers can now afford to study this important component of body composition. 2.7 Anthropometric prediction of body fat Skinfold measurement has become a common practice for estimation of body fat (21). A common method for the development of a skinfold prediction equation requires the measurement of subcutaneous fat sites (skinfolds) in a group of subjects, in addition to a criterion measure of body density such as underwater weighing, and then by regression an equation is established to predict body density by skinfold measures alone. Body density is then converted to percent fat, most commonly by using the Siri equation (22), where % body fat = 495/body density - 450. However, in the process of deriving body fat from initial skinfold measurements, several assumptions are made. These assumptions include: the compressibility of adipose tissue is constant, skin thickness is negligible, there is a fixed adipose tissue pattern, adipose tissue maintains a constant fat fraction, there is a fixed proportion of internal to external fat, and the density of fat and fat free masses are constants. A thorough discussion of these assumptions and the potential for error has previously been presented (23). Often variables other than skinfolds (such as age and girth measurements) are included in prediction equations in an effort to improve the relationship with the criterion measure. Jackson and Pollock (16) have established a set of body density prediction equations in this manner using a sample of 308 adult men (mean age 32.6yrs ± 10.8) and cross-validated their equations in a sample of 95 men (33.3y ±11.5 y), using underwater weighing as the criterion method for body density measurement. Eight body density prediction equations were developed in total, using a variety of different variables, which included age, waist girth, forearm girth, and seven different skinfolds. Equation number 4 from this study was selected because it included the maximum number of skinfolds (7 in total) in addition to age. In this equation body density (g/ml) = 1.11200000 - 0.00043499 ( X , ) + 0.00000055 ( X , ) 2 -0.00028826 (X3) where X i is the sum of 7 skinfolds taken at the chest, axilla, triceps, 15 subscapula, abdomen, suprailium and front thigh, and X 3 is age (years). Body density predicted with this equation showed a strong correlation with body density from underwater weighing in both the original sample (R=0.902, SE 0.0078 g/ml) and the cross validation group (R=0.915, SE 0.0078g/ml). These results indicate that this is a very useful body density prediction equation for adult men. 2.8 Skinfold measurement of subcutaneous adipose tissue To measure and compare subcutaneous adipose tissue, a simple sum of skinfolds taken from major adipose storage sites is a common technique. Failure to account for all the major adipose storage sites will result in an inability to account for fat patterning differences between individuals. Differences in skinfold compressibility and possibly skin thickness between individuals are limiting factors to this technique (23). Overall, skinfold measurements provide a direct measure of subcutaneous fat and as a result have good face validity (21). 2.9 Possible explanations for vegetarian vs. omnivore differences The possibility exists that vegetarians and omnivores are different with respect to more than just the consumption of meat in their diets. Lifestyle differences, such as eating behaviours and physical activity levels, should also be considered, as differences between dietary groups could contribute to differences in body composition. If health issues are a primary motivator for dietary choice among vegetarians, then it might also be hypothesized that these individuals would likely engage in more frequent bouts of physical activity than omnivorous controls. Janelle and Barr (24) considered eating behaviour differences and hours of exercise per week among 45 weight stable, regularly menstruating health conscious 16 vegetarian (n=23) and nonvegetarian (n=22) women between 20-40 years of age. Sixteen of the 22 vegetarian women (70%) in this study cited health reasons when asked to indicate their motivation for becoming a vegetarian. Investigators used the Three-Factor Eating Questionnaire (TFEQ) to assess eating behaviours. This 51-item questionnaire has subscales for dietary restraint (conscious limitation of food intake), disinhibition (a tendency to eat more than usual when control over intake is lost), and hunger (25). Results indicated that vegetarians had significantly lower dietary restraint scores, and no significant differences were observed between disinhibition and hunger scores. These results lend support to the argument that health conscious vegetarians may have lower levels of dietary restraint than health conscious omnivorous controls. It should be noted that subjects in this study were selected to exclude individuals who might have eating disorders. Vegetarians in this study exercised 4.0 hours per week (± 2.3) and nonvegetarians 3.1 hours (± 1.80) but this difference was not significant. In contrast to these findings, Martins et al. (26) found that there was a link between vegetarians and higher restraint scores measured by the TFEQ. This study examined the relationships between dietary style (from meat eating to veganism) cognitive restraint (TFEQ), and feminist values in a group of 227 men and women. Among males, it was found that individuals who scored high in cognitive restraint were more likely to be vegetarians than individuals low in restraint. Only females who were high in feminist values showed the same trend. In response to their findings, the authors of this study suggested that selection of a vegetarian diet might be a way of masking dieting behaviours. 17 2.10 Vegetarian vs. omnivore differences in protein, fat, andfiber &possible influences on body composition It has previously been mentioned that there are different sub-classifications of the broad term \"vegetarian\". Although each of these diets is distinct, individuals consuming vegetarian diets share some common differences in macronutrient intakes from those consuming omnivorous diets. In general, the vegetarian diet has been found to contain considerably more dietary fiber (10, 15, 24, 27-30), less protein (6, 24, 27-30), and saturated (10, 27, 28) and polyunsaturated fat (27). Total fat has consistently shown a trend to be lower in vegetarians than omnivores (6, 10, 24, 27-29, 31), but achieved statistical significance in only two of these studies (28, 31). Unfortunately only two of the mentioned studies were carried out in samples of adult men (27, 28), making it difficult to draw conclusions about dietary intake trends in this population. It has been speculated that dietary differences may have an influence on male sex hormones (testosterone and free-testosterone) and possibly contribute to differences in skeletal muscle between vegetarian and omnivorous populations (5). Testosterone is the male sex hormone most frequently measured, and from a body composition perspective this hormone plays an important role in muscle development (32). Specifically, it is the unbound or free-testosterone that is believed to be the most active and have the greatest androgenic effect. Thus, i f a vegetarian diet were to lead to lower levels of testosterone and free testosterone then this might serve as a possible mechanism for muscle differences. Sex hormone binding globulin (SHBG) in the serum is another important quantity, because it has an inverse relationship with free testosterone (33) in (34). Results from cross-sectional studies on this issue have shown a trend towards equivalence between vegetarians and omnivores for testosterone and calculated free testosterone levels (27, 28). However, an 18 experimental study utilizing a crossover design to examine the relationship between diet, serum sex hormones and endurance performance did find a that a vegetarian diet had a significant impact on testosterone levels (35). Eight male endurance athletes consumed two isocaloric diets in succession (6 weeks for each) with a two-week washout period. One diet was LOV (57% CHO, 15% protein, 28% fat), and the other was a meat-rich omnivorous diet (58% CHO, 13%o protein, 29% fat). After 6 weeks on the LOV diet, a significant decrease in total testosterone was identified (33). Although the sample of this study was small, the results of this well designed study suggest that a vegetarian diet might lead to reduced levels of total testosterone. 2.11 Conclusions The possible impact of a vegetarian diet on major components of body composition (muscle, fat) needs more study. Preliminary research indicates that there may be significant body composition differences between vegetarians and omnivores. There is a need to investigate the body composition differences between vegetarians and omnivores in a young adult male population, a group that has been neglected to this point. With recent innovations in the field of skeletal muscle mass prediction, and simple anthropometric methods for fat prediction, there is a considerable amount of insight to be gained into the relationship between vegetarianism and body composition from a cross-sectional study. There is also the need to study eating behaviors and physical activity levels as these may be contributing factors to body composition differences. The primary aim of this study is to gain a clearer picture of the relationship between vegetarianism and body composition particularly with regards to the effect of this diet on skeletal muscle. In addition to providing a better understanding of the 19 influences of diet on body composition, this study may serve as a foundation on which to base future experimental research. 2.12 Research questions & hypotheses 2.12.1 Primary questions and hypotheses Question 1: Are there differences in whole body skeletal muscle mass from anthropometric prediction (13) between vegetarians and omnivores? Hypothesis 1: Whole body skeletal muscle mass from anthropometric prediction will be greater in omnivores than vegetarians. Question 2: Are there group differences in a sum of skinfo ld-corrected muscle girths taken from 6 upper limb and 6 lower limb sites? Hypothesis 2: Omnivores will have a higher sum of skinfo ld-corrected muscle girths than vegetarians. Question 3: Are there subcutaneous fat differences between dietary groups calculated from a sum of skinfolds? Hypothesis 3: Vegetarians will have a lower sum of skinfolds than omnivores. Question 4: Are there differences in percent body fat between groups calculated from Jackson and Pollock's (16) anthropometric prediction equation? Hypothesis 4: Vegetarians will have a lower percent body fat than omnivores, calculated from anthropometric prediction. 2.12.2 Secondary research questions Question 5: Are scores of dietary restraint (TFEQ) different between vegetarians and omnivores? 20 Question 6: Are scores for activity level measured by the Godin Leisure-Time Exercise Questionnaire (GLTEQ) (36) different between vegetarians and omnivores? Question 7: What is the relationship between TFEQ scores of restraint and measures of fat and muscle? Question 8: What is the relationship between scores from the GLTEQ and fat and muscle measures? Question 9: Are intakes of protein, dietary fat, fiber, and calories, measured by three-day food records different between dietary groups? 21 3. METHODS 3.1 Subjects The study sample consisted of 27 men who were vegetarians (Veg) and 27 omnivorous men (Omni) between the ages of 20-34 years. This population had not previously been studied with regard to vegetarianism and body composition. Subjects were recruited through flyers and posters, distributed at restaurants, bus stops, fitness centers, community centers, local colleges, and the University of British Columbia. Upon seeing a poster or flyer advertising the study, subjects phoned the investigator and were informed about the study and asked a series of screening questions (Appendix III). Approximately half of all subjects interested in participation did not meet the entry criteria and were excluded. 3.1.1 Vegetarian inclusion criteria 1) Vegetarian subjects permitted entry to the study must not have consumed meat products (red or white meat, fish) more than once a month, for a minimum of two years prior to recruitment. 3.1.2 Vegetarian and omnivore inclusion criteria 1) Subjects must have been between 20-35 years of age. 2) Subjects must not have been involved in regular resistance training (more than once a week) in the last year, as this training has an obvious influence on body composition. 3) Subjects must not have been involved in more than 7 hours of moderate intensity physical activity per week (recreational jogging, swimming, cycling, climbing, volleyball, tennis etc.). 22 4) Subjects were required to be non-smokers, to limit any confounding influences on weight. 5) Subjects must not have taken any medication in the previous year, known to effect body composition. This information (initially obtained on the phone) was confirmed with a short screening questionnaire (Appendix III). 3.2 Study design This is a cross-sectional study. Initially, subjects participated in a testing session where each subject underwent a set of anthropometric measurements (height, weight, skinfolds, muscle girths), and completed questionnaires concerning dietary behaviours and physical activity levels. Following this session, subjects completed a three-day diet record. Dietary groups (Veg and Omni) were then compared on the basis of anthropometric measurements of muscle and fat, questionnaire measures of dietary behaviours and physical activity, and nutrient intakes from diet records. 3.3 Testing procedure Testing sessions required approximately one hour. Upon arrival subjects were greeted, the study was explained and informed consent was obtained (Appendix I). Subjects then rotated through three stations. At the first station subjects completed the Godin Leisure-Time Exercise Questionnaire (GLTEQ) and the TFEQ questionnaires. Subjects also formally filled out a copy of the screening questionnaire (note: any questions that would have excluded the subject from participating were previously asked on the phone). Total time for this station was between fifteen and twenty minutes. The second station was for anthropometric measurements, and took approximately twenty minutes. The final station was used to explain 23 how to complete a three-day diet record, and subjects received a booklet to keep this record along with a set of measuring cups and spoons. This station took approximately twenty minutes. One researcher conducted each testing session. The same researcher oversaw and answered any questions regarding study procedures and performed all anthropometric measurements. All testing sessions were carried out at University of British Columbia's Buchanan Exercise Science Lab. 3.4 Questionnaires 3.4.1 Screening questionnaire The information provided ensured that subjects met all inclusion criteria prior to participation (Appendix III). 3.4.2 Three-day diet record Three-day diet records were completed by participants on three consecutive days including one weekend day, following the testing session, and returned to the instructor in person or by mail. Participants were instructed on how to complete the diet record following the anthropometric assessment stage of the testing session. An open form was used, where subjects were asked to describe all of the foods and beverages eaten for each day (37). Subjects were provided with a set of measuring cups and spoons and asked to estimate serving sizes with the given utensils when practical to do so. 3.4.3 The three-factor eating questionnaire (TFEQ) This 51-item eating behavior questionnaire (Appendix III) has subscales to measure dietary restraint, disinhibition, and hunger (25). A copy of the TFEQ was distributed to 24 subjects and used to assess whether men who were Veg or Omni differ with respect to dietary behaviours, with a particular focus on dietary restraint. There are 21 questions related to the dietary restraint subscale (factor 1 items), and they have proven to be reliable among both free eaters (a= 0.92) and dieters (a= 0.79) (25). In addition, factor 1 items have been able to discriminate between dieters and free eaters (25). Scores for dietary restraint were calculated according to the instructions of the author of the questionnaire (25). 3.4.4 Godin leisure-time exercise questionnaire (GLTEQ) The GLTEQ is a short questionnaire that assesses the number and type of activity bouts (minimum 15 minutes to qualify) regularly performed in a week (Appendix III). The first part of the questionnaire asks how many times in a week a subject performs activities in each of three listed exercise categories: strenuous exercise where the heart beats rapidly (e.g. running, jogging, hockey, football), moderate exercise which is not exhausting (e.g. fast walking, easy bicycling, volleyball), and mild exercise requiring minimal effort (e.g. yoga, archery, bowling). The second part of the questionnaire asks respondents to assign a frequency, by checking one of 3 boxes (often, sometimes, never/rarely), to the regularity that exercise bouts are sustained long enough to work up a sweat or where the heart beats rapidly. In reliability studies of the GLTEQ (36, 38, 39) strenuous exercise scores are the most reliable and scores for light exercise are least reliable (36, 38). Among 53 adults, test and 2 week retest scores showed significant correlations for all types of exercise, strenuous (0.94), moderate (0.46), light (0.48), total (0.74). The score of this questionnaire was determined using the instructions provided by the authors (36). 25 3.5 Measurement of height and body mass Barefoot height was measured in duplicate with a wall-mounted stadiometer to the nearest 0.1cm. Body mass was measured in duplicate (in shorts and T-shirt), and recorded at the nearest 0.1 kg using a digital scale. 3.6 Anthropometric procedures All skinfold measurements were taken in duplicate and measured to the nearest 0.1 mm using a Harpenden skinfold caliper. If duplicate measures differed by 0.5 mm or more, a third measurement was taken. Limb circumference measurements were made in the plane orthogonal to the long axis of the body segment being measured. A flexible standard measuring tape was used for all circumference measurements, which were taken in duplicate (unless values differed by more than 4 mm) and recorded to the nearest 1mm. The mean value (obtained from either 2 or 3 measurements) for each skinfold and circumference measurement was calculated, and this value was used in any prediction equation requiring a single value. A highly trained individual made all measurements, and intraclass correlations between measurement 1 and measurement 2 were calculated to check accuracy. 3.7 Variables of interest 3.7.I Independent variable: diet There were two groups in this study. One group was composed of individuals who had been consuming a vegetarian diet for 2 years or more (Veg), and a second group who had consumed an omnivorous diet for 2 years or more (Omni). 26 3.7.2 Dependent variables 3.7.2.1 Total body skeletal muscle mass Skeletal muscle mass was measured using Lee's anthropometric prediction equation (13) where SM (kg) = Ht x (0.00744 x CAG 2 + 0.00088 x CTG 2 + 0.00441 x CCG 2) + 2.4 x sex (1 for males) - 0.048 x age + race + 7.8. This equation has an R 2 value (compared to MRI measured muscle mass) of 0.91 and a standard error of 2.2 kg. The equation requires the measurement of both a skinfold and limb circumference from the arm (CAG), thigh (CTG), and medial calf (CCG). The precise location of these sites has been previously defined (40). Use of this prediction equation requires an assigned value for race. For Asian subjects a race value of-2 was assigned, 1.1 for African Canadians, 0 for white or Hispanics. For subjects belonging to a race other than what has been specified, a value of 0 was used. Identification of a subject's race was obtained through the screening questionnaire. The obtained value for total body skeletal muscle mass accounts for body size differences between individuals, as the prediction equation includes a height variable. Height differences between groups were linearly corrected prior to the analysis of this variable. 3.7.2.2 Sum of 12 skinfold-corrected muscle girths In addition to the anthropometric prediction equation, a sum of skinfold-corrected girths taken from six upper limb and six lower limb sites was measured as an index of muscle. Limb girths and skinfolds were taken from the upper arm, thigh, and calf at sites that have been previously defined (40). These sites have a strong correlation with total body skeletal muscle (13). In addition, a forearm girth was taken at its maximum circumference and a skinfold from the posterior aspect of the forearm at the same level. A cadaver study has shown forearm girth to be a very good predictor of total body skeletal muscle (19). To 27 account for muscle shape, sites located 2 cm above and below each of the mentioned landmarks were also measured. Each corrected girth was then squared and multiplied by stature for dimensional consistency. The sum of these values was calculated. 3.7.2.3 Sum of 6 skinfolds During the second part of the testing session (anthropometry section) skinfolds were measured. Sites for measurement included the triceps, subscapula, abdominal, front thigh, and suprailium, using landmarks previously defined (40). A medial calf skinfold was also taken, using a landmark described elsewhere (41). A sum of skinfolds (E 6) was calculated based on the mean value of each skinfold. Skinfold sites were selected to account for major regions of adiposity. 3.7.2.4 Percent fat predictedfrom anthropometry Jackson and Pollock's equation (16) for predicting body fat from anthropometry was selected for use in this study. This equation requires the measurement of skinfolds from the chest, axilla, triceps, subscapula, abdomen, suprailium, and front thigh. The procedures and location for taking each of these measurements has been previously described (40). This equation was chosen for a variety of reasons. It has a large number of sites accounting for individual fat distribution differences. It was carried out in a large sample (n = 308) of healthy adult men with a mean age of 33 y, and then cross-validated in another sizable group (n = 95). With few comprehensive fat prediction equations having been generated and cross-validated on young sedentary and healthy men this equation seemed to fit closely with the current sample. As previously mentioned, strong correlations and relatively low standard errors were observed in both the original sample and the cross validation group (16). Values 28 obtained for body density were converted to percent fat using the Siri equation, where % fat = 495/body density - 450 (22). 3.7.2.5 Nutrient intakes from three-day diet records Three-day intakes of dietary fat, protein, fiber, and total calories were calculated from three-day diet records using The Food Processor 7 software (42). 3.7.2.6 Three-factor eating questionnaire restraint score Scores of dietary restraint were calculated. See section 3.43 for more detail. 3.7.2.7 Godin leisure-time exercise questionnaire score See section 3.44 for more detail. 3.8 Statistical analysis One-way ANOVAs were used to compare group means with respect to the dependent variables. In cases where data were nonparametric, the Mann-Whitney U-test was used as a replacement test to examine group differences. ANCOVA was used to compare muscle differences between groups while controlling for group differences in height. Spearman correlation coefficients were calculated to examine the relationship between TFEQ scores, restraint scores (TFEQ) and GLTEQ scores with each of the fat and muscle variables. Significance was set at the 0.05 level for all statistical tests. 29 3.9 Statistical power & sample size determination Limited data is available regarding skeletal muscle mass in young men. For the purpose of the sample size determination, a value of 36kg ± 4kg was used for the Omni group. This was an estimation based on values that have been derived for total skeletal muscle mass in men (13, 43, 44). A mean value of 34kg ± 4kg SMM was assumed for the Veg group. No data have been collected to date that could help with this estimation, so this value was used for sample size determination. A two kilogram difference between the two groups would be an important finding and therefore we will test whether this difference reaches statistical significance. By employing these numbers (36 kg and 34 kg) we can then determine the sample size required for achieving statistical significance. Power was calculated via the UCLA stats page power calculator using mean values for SMM of 36kg ± 4kg for the Omni group, and 34kg ± 4kg for the Veg group. After inserting these values in a one tailed test with a significance level of .05 and a power of .65, the necessary sample size will be n = 30 for the Omni group, and n = 30 for the Veg group. A one tailed test is justified in this case because research has indicated that any between group differences that exist in muscle are unidirectional (4, 5), with omnivores having more muscle than vegetarians. Therefore, by selecting a one tailed test, the statistical power for detecting a difference is enhanced. 3.10 Ethical approval Ethical approval for all of the procedures previously mentioned was obtained from the University of British Columbia's clinical research ethics board (Appendix II). 30 4. Results 4.1 Body composition The results of this study were based on 54 subjects, aged between 20-34. General sample characteristics and body composition results for the two dietary groups are shown in table 4.4.1. Reliability of anthropometric measurements was assessed by calculating the intraclass correlation between the first and second measurements taken for each skinfold and girth measurement. The average intraclass correlations for skinfolds and girths were alpha = 0.9990 and 0.9995 respectively. Table 4.1.1 Physical characteristics and body composition measurements for vegetarian and omnivorous groups Variable Vegetarians (n = 27) Omnivores (n = 27) P Value 1 tailed test 2 tailed test Age 26.7 ± 3 . 2 25.8 ± 4 . 8 0.411 Height (m) 1.768+ 0.071 1.795 ± 0 . 0 6 4 0.145 Body Mass (kg) 73.0+10.2 77.0 ± 9.9 0.143 Muscle Mass (kg) 30.9 + 3.5 32.7 ± 4 . 1 0.049* 0.097 A N C O V A Muscle (kg) 30.7 32.6 0.040* 0.079 Sum of 12 C G (cm3) 2471207 ± 4 1 5 9 3 1 2683571 +413584 0.033* 0.066 % Body Fat 10.5 ± 3 . 6 12.0 ± 4 . 5 0.173 Sum of 6 Skinfolds (mm) 70.7 ± 2 2 . 7 80.7 + 29.2 0.168 Years Vegetarian 6.4 ± 5.6 N A N A N A Values are group means + SD, n = number of subjects. *P value is significant at the 0.05 level Muscle Mass = Prior to muscle mass prediction from anthropometry, vegetarians were linearly heightened 2.7cm to account for the between group height differences. A N C O V A Muscle = Mean muscle mass values generated by A N C O V A controlling for height differences between groups. Sum of 12 C G = Sum of 12 skinfold-corrected muscle girths One way ANOVA (1-tailed test) revealed significant differences between Veg and Omni in muscle mass adjusted linearly for group differences in height (Appendix VII). 31 ANCOVA results (with height as the covariate), corroborated these findings (Appendix X). Similarly, significant group differences were observed between Veg and Omni in a sum of 12 corrected muscle girths. ANOVA results (2-tailed tests) showed no significant group differences in body fat variables (S6 skinfolds, % fat). Boxplots of body composition results are pictured in Figures 4.1.1-4.1.4. Figure 4.1.1 illustrates that the middle 50 th percentile for muscle mass has a similar range between groups, but the Veg group has a significantly lower median value, as a result of a higher frequency of scores at the low end of this range. Figure 4.1.1 Boxplots of linearly Figure 4.1.2 Boxplots of sum of 12 corrected muscle mass of muscle girths of vegetarians and vegetarians and omnivores omnivores E o 1 o O 5> x o E cn iooooog_ Vegetarians Vegetarians Omnivores Diet Group Diet Group 32 Figure 4.1.3 Boxplots of sum of 6 skinfolds of vegetarians and omnivores Figure 4.1.4 Boxplots of percent fat of vegetarians and omnivores Omnivores Diet Group Vegetarians Omnivores Diet Group Comparing groups on the basis of a sum of 12 corrected muscle girths, yields a more obvious difference (figure 4.1.2). The range of values for Omni is higher than the range observed for Veg. Body composition variables showed no major departures from normality (table 4.1.2). Histograms for each body composition variable are provided in Appendix V. Table 4.1.2 Skewness and kurtosis values for body composition variables Vegetarians Skewness Kurtosis Omnivores Skewness Kurtosis Muscle Mass (kg) 0.45 -0.80 0.33 0.08 Sum of 12CG (cm3) 0.30 -0.42 -0.11 -0.75 Muscle Mass = Predicted by anthropometric equation Sum of 12 C G = Sum of 12 skinfold-corrected muscle girths Sum of 6 SF = Sum of 6 skinfolds Sum of 6 SF (mm) 0.40 0.20 -0.87 0.21 % Fat 0.61 0.61 -1.22 0.014 33 4.2 Questionnaire scores Questionnaire scores (GLTEQ, TFEQ, Restraint Subscale) violated ANOVA test assumptions for normality. Table 4.2.1 shows values for skewness and kurtosis. Acceptable values for skewness and kurtosis were between +1 and -1. For a graphic illustration of distributions on questionnaire variables, histograms are pictured in Appendix V. As a result of normality violations, the Mann-Whitney U-test (nonparametric test) was used to compare dietary group differences with respect to each of these variables. Table 4.2.2 lists the mean values, test statistics and significance levels for questionnaire scores and reported hours of weekly activity. No significant differences between Veg and Omni men were observed on any of these measures. It should be mentioned that reported number of hours of weekly activity was not in violation of ANOVA test assumptions, but the p value was identical for both the ANOVA test and the Mann-Whitney U-Test so the latter result is shown. Table 4.2.1 Skewness and kurtosis values for questionnaire results (GLTEQ, TFEQ, TFEQ restraint subscale, reported hours of activity) G L T E Q T F E Q Restraint Subscale Activity (hr/wk) Vegetarians Skewness 1.11 1.23 1.75 1.00 Kurtosis 3.34 2.55 5.34 -1.01 Omnivores Skewness 0.97 0.91 1.12 -0.83 Kurtosis 0.57 0.22 0.84 -0.33 G L T E Q = Godin Leisure-Time Exercise Questionnaire Moderate Activity = self-reported hours per week engaged in moderate intensity activities T F E Q = Three-Factor Eating Questionnaire Restraint Subscale = dietary restraint subscale score of T F E Q 34 Table 4.2.2 Questionnaire scores and reported hours of moderate activity for vegetarian and omnivorous groups Variable Vegetarians (n = 27) Omnivores (n = 27) U-Test Sig. (2 tailed) G L T E Q Score 5 1 . 8 ± 2 7 . 0 46.8 + 23.3 307.50 0.324 Moderate Activity (hrs/wk) 4.6 ± 1.8 3.8 ± 1.8 270.00 0.097 T F E Q Score 14.9 + 5.1 16.7 + 7.9 354.00 0.855 Dietary Restraint (TFEQ) 5.6 + 3.5 5.5 ± 4 . 3 337.50 0.639 Values are group means ± SD, n = number of subjects G L T E Q = Godin Leisure-Time Exercise Questionnaire Moderate Activity = self-reported hours per week engaged in moderate intensity activities T F E Q = Three-Factor Eating Questionnaire Dietary Restraint (TFEQ) = dietary restraint subscale score of T F E Q U-Test = Mann-Whitney U-Test statistic 4.3 Correlations between dependent variables The data were further analyzed to examine relationships between any of the body composition variables and the questionnaire scores (TFEQ Score, Dietary Restraint Score from TFEQ, GLTEQ Score). In this analysis the entire study population was considered as a single group. Spearman correlations were used to account for violations to normality. Table 4.3.1 shows the Spearman correlation coefficients for these relationships. The only significant correlation was between % fat and the GLTEQ score, and this relationship is provided (Figure 4.3.1). 35 Table 4.3.1 Spearman correlation coefficients of questionnaire scores with body composition variables. Variable Correlation Coefficient With Dietary Restraint (TFEQ) T F E Q Score G L T E Q Score Muscle Mass (kg) -0.051 -0.027 -0.172 Sum of 12 C G (cm3) -0.057 -0.052 -0.148 % Fat -0.199 -0.235 -0.292* Sum of 6 Skinfolds (mm) -0.182 -0.226 -0.255 1 : I •Correlation is significant at the 0.05 level (2-tailed) Sum of 12 C G = Sum of 12 skinfold-corrected muscle girths T F E Q = Three-Factor Eating Questionnaire Dietary Restraint (TFEQ) = Dietary restraint subscale score of the T F E Q G L T E Q = Godin Leisure-Time Exercise Questionnaire Figure 4.3.1 Scatterplot of relationship between GLTEQ score and percent fat 160r GLTEQ 4.4 Dietary intakes Just as there were violations of normality for questionnaire scores, 3-day intakes of dietary variables were in violation of ANOVA assumptions for normality. As a result of ANOVA violations, the Mann-Whitney U-test was used to compare dietary variables between groups (Appendix VIII). Values for skewness and kurtosis can be seen in Table 4.4.1. Histograms can be viewed in Appendix V. 36 Table 4.4.1 Skewness and kurtosis values for dietary intake variables by group. Calories Protein Fiber Total fat Sat Fat Mon fat Poly fat P : S Vegetarians Skewness 0.58 0.68 2.08 0.90 1.48 1.12 0.88 0.90 Kurtosis -0.15 0.26 5.73 1.39 1.89 1.41 0.15 0.56 Omnivores Skewness 1.44 1.47 0.87 1.80 2.11 1.33 2.26 2.01 Kurtosis 2.50 2.65 0.54 4.78 7.40 2.22 5.83 5.07 Sat Fat = Saturated fat Mon Fat = Monounsaturated fat Poly fat = Polyunsaturated fat P : S = Ratio of polyunsaturated fat to saturated fat Results of the Mann-Whitney U-Test (Table 4.4.2) showed that the Veg group consumed significantly more dietary fiber, polyunsaturated fats and a higher ratio of polyunsaturated to saturated fats (P:S), and less saturated fat. No between group differences were observed for 3-day intakes of calories, protein, total fat, or monounsaturated fat. Table 4.4.2 Three-day dietary intakes compared between dietary groups Variable Vegetarians (n = 27) Omnivores (n = 27) U-Test Sig. (2 tailed) Total Calories 1 0 7 8 1 ± 3 6 2 6 10608 + 3915 350.0 0.802 Protein (g) 354.5 + 123.7 452.7 + 210.0 264.0 0.082 Dietary Fiber (g) 139.8 ± 7 6 . 6 92.3 ± 4 3 . 3 207.0 0.006** Total Fat (g) 331.3 + 136.3 365.8+ 177.6 333.0 0.586 Saturated Fat (g) 93.8 + 62.1 129.1+71.4 240.0 0.031* Monounsaturated Fat (g) 89.5 + 44.8 101.2 + 55.0 322.0 0.462 Polyunsaturated Fat (g) 67.5 + 32.7 42.7 + 27.5 171.5 0.001** Poly Fat: Sat Fat 0.97 ± 0 . 6 3 0.36 ± 0 . 1 9 132.0 0.000** Values are group means ± SD, n = number of subjects *P value is significant at the 0.05 level **P value is significant at the 0.01 level U-Test = Mann-Whitney U-Test statistic 37 5. Discussion A review of the current literature indicated that there was a minimal amount of research on the topic of body composition as it relates to vegetarianism. The study of muscle mass has long suffered from the absence of a practical method for measurement. The intent of this study was to explore this topic in a population that has yet to receive any attention, young men. 5.1 Body composition comparisons between dietary groups There is limited research on the topic of vegetarianism and its relationship with muscle mass. The first two questions proposed in this study (see section 3.1), concerned whether or not a group of young vegetarian males would differ with respect to muscle mass from a group of young omnivorous males. The present study showed that there was a significant difference between groups in muscle quantity. The Veg group was found to have significantly lower values for predicted skeletal muscle mass (Veg 30.9kg ±3 .4 vs. Omni 32.7kg ± 4.1) after vegetarian subjects were linearly heightened 2.7 cm. In the present study, although height was not statistically different between groups, the omnivores were 2.7 cm taller. To account for this small size difference, the heights of vegetarians were linearly scaled to match the omnivores, as previous research has shown that muscle mass increases linearly with height (43). Height was selected as the variable to correct for size differences because it does not interfere with the variable of interest, muscle mass. On the contrary consider correcting muscle mass differences by the more obvious choice of body mass. If correcting by body mass, then any differences in body fat between the groups would obscure the potential for finding any differences in muscle as a percent of body mass. In the present 38 study, because omnivores were slightly fatter (10.5 % vs. 12.0%, although the difference was not statistically significant) any chance of finding omnivores to have a higher percent muscle is confounded by their increased adiposity. Therefore, it was decided that in a between group comparison of muscularity correction for body mass would be misleading. Additional statistical support for this finding of decreased muscle mass among vegetarians was obtained through ANCOVA (with height as the covariate), revealing slightly larger between group differences (Veg 30.7 kg vs. Omni 32.6 kg). Significant group differences in muscle quantity were also observed in comparisons of the sum of 12 corrected muscle girths (Veg 2.5xl06 cm3± 4.2xl05 vs. Omni 2.7x106cm3 ± 4.1xl05). There has been no previous cross-sectional research on this topic to corroborate our findings. It was hypothesized that differences in muscle would exist between dietary groups, based on the earlier research done in older male populations (4, 5). Previous experimental research has shown that in an elderly male population, consumption of a meat free diet has been inferior to a meat containing diet for gaining muscle and fat free mass with resistance training (4, 5). Although these earlier studies explored diet and the response to training, the central question posed by these studies was similar. Are there muscle differences between a group of men who consume a vegetarian (or meat free) diet compared to a group who consumes an omnivorous one? The answer so far appears to be yes. As the present study is cross-sectional in nature, no causative conclusions can be drawn. However, if a vegetarian diet does have an adverse affect on an individual's muscle building potential, this has important applications for a variety of populations including athletes, dieters, and the elderly. Discussion of a mechanism that might explain the observed muscle differences is also a matter limited to speculation. Previous authors have suggested the possibility of a diet related hormonal effect to account for muscle differences (5). The present study made no 39 attempt to monitor hormones, and as a result can offer no evidence to support or refute this proposition. In addition to considering muscle differences, body fatness was also a variable of interest in the present study. Two variables were examined, a sum of 6 skinfolds (accounting for the major areas of fat deposition), and a percent fat derived by predictive equation. It was initially hypothesized that vegetarians would have lower values for each of these variables. Although the Veg group had lower scores for both a sum of 6 skinfolds (Veg 70.7mm ± 22.7 vs. Omni 80.7 ± 29.2) and percent fat (Veg 10.5% ± 3.6 vs. Omni 12.0% ± 4.5) neither achieved statistical significance. It appears as though the small sample size of the present study and the high standard deviations on these measures were limiting factors to detecting a possible difference. This finding agrees with the body of literature which has found either no differences between dietary groups with respect to fatness, or in the cases where there is a difference it is the vegetarians that are leaner (6-10). 5.2 Lifestyle comparisons between dietary groups Vegetarianism is an unconventional diet, and perhaps an unconventional diet might be related to other unconventional lifestyle factors. A potential explanation for the lower muscle among vegetarians might be that vegetarianism is simply a socially acceptable way for men to diet. If this were true, body composition differences would certainly be expected between a group of vegetarians and omnivores. To test whether eating behaviours and attitudes differed between groups, all participants completed the 51-item Three-Factor Eating Questionnaire (TFEQ) (25). Of the three subscales measured in the TFEQ (restraint, dishinibition, and hunger), dietary restraint (the conscious limitation of food to control weight) was the focus of the present study. A higher score is indicative of greater dietary restraint. Results showed 40 dietary restraint scores to be nearly identical between Veg (5.6 ± 3.5) and Omni (5.5 ± 4.3) groups in this study. Total TFEQ score was also considered with Veg scoring lower than Omni (14.9 ±5.1 vs. 16.7 ± 7.9), although it was not a significant difference. In past research, a relationship has been identified between male vegetarians and increased dietary restraint (26). These findings are far from unanimous, as research in health conscious women has shown vegetarians to have less dietary restraint than omnivores (24). Another potential explanation for body composition differences between dietary groups would be different activity levels. If one group is significantly more active, then this could explain differences in muscularity. To test whether groups differed with respect to exercise habits, participants completed the Godin Leisure-Time Exercise Questionnaire. In addition, subjects made an estimation of their participation hours (per week) in moderate intensity physical activities (i.e. recreational jogging, swimming, cycling, rock climbing, volleyball, tennis etc.). The higher the GLTEQ score the higher the activity level of an individual. Results indicated no significant difference between groups on either the GLTEQ (Veg 51.8 ± 27.0 vs. Omni 46.8 ± 23.3) or their hours of estimated activity (Veg 4.6 ± 1.8 vs. Omni 3.8 ± 1.8). However, the possibility still exists that one group was prone to involvement in more muscle promoting activities. These findings, combined with the eating behaviour findings suggest that there is no real difference in exercise or eating behaviours in the sample population involved in this study. Therefore, lifestyle factors are not able to explain the observed differences in muscle. 5.3 Correlations between dependent variables There was only one significant correlation identified in the present study, existing between GLTEQ score and % fat (r2 = -0.292). The weak relationship that exists, is not a 41 surprise as GLTEQ is a measure of the amount of activity an individual participates in. It would be expected that as activity level increases body fat would decrease. 5.4 Dietary intake comparisons The final hypothesis posed in this study concerned whether or not vegetarians and omnivores differed with respect to 3-day dietary intake values. The Veg group were found to consume significantly more dietary fiber (139.8g vs. 92.3g), polyunsaturated fats (67.5g vs. 42.7g) and a higher ratio of polyunsaturated to saturated fats (0.97 vs. 0.36), but less saturated fat (93.8g vs. 129.Ig). These findings are in agreement with the limited dietary research that has been conducted on male vegetarians, and found increased fiber (27, 28) and decreased saturated fat (27, 28) levels in comparison to omnivores. No group differences were observed for 3-day intakes of calories, protein, total fats, saturated or monounsaturated fats. Although there was no significant difference observed in protein, it might be suggested that protein differences between Veg (354.5g ± 123.7) and Omni (452.7 ± 210.0) could explain muscularity differences. However, the average daily protein consumption of vegetarians in this study (118.2g) is nearly double the recommended daily allowance of 58-63g for men in their age range (45), and therefore was not believed to have a causal influence. 5.5 Summary and recommendations for future research Has the present study contributed to a very new body of literature on vegetarianism and body composition? Utilizing a cross-sectional design, the present study has shown significant differences between vegetarians and omnivores with respect to muscle and dietary intakes of fiber and polyunsaturated and saturated fats, but no differences in activity levels, dietary restraint, or body fat levels. These were positive results, as the aim of the present 42 study was to serve as a building block for future research on the topic of body composition and vegetarianism. Before one attempts to answer the question \"Why do vegetarians differ with respect to muscle mass?\", we must first question whether they are actually different. With this goal being accomplished in the present study, it should be the aim of future studies to answer the question of why they might be different. A recommendation for a future study would be one utilizing a design similar to the earlier Campbell study (5), but performed in a population of younger men. Like the Campbell study (5), a group composed entirely of sedentary omnivores would be randomized into 2 dietary groups. One group continues to consume a meat containing diet, while the other group is selected to consume a LOV diet. Baseline measurements of muscle are taken (ideally using MRI), in addition to fat measurements, and hormone levels (testosterone, free testosterone SHBG). Both groups would then engage in an identical 6-month training program aimed at increasing muscle hypertrophy. Al l dependent variables are re-measured monthly. In addition to Campbell's framework, the proposed study adds a more accurate means of muscle evaluation, plus hormone measurements and a longer training period. It is felt that a study with these additions might uncover a causal link between vegetarianism and body muscle. Until such a study is completed, we can only speculate about the possibility of a causal link between a vegetarian diet and decreased muscle mass. 43 6. References 1. National Institute of Nutrition. The Many Faces of Vegetarianism. News Release: Jan 1999. 2. Dwyer JT. Nutritional consequences of vegetarianism. Annu Rev Nutr 1991; 11:61 -91. 3. Clarys JP, Martin A D , Drinkwater DT. Gross tissue weights in the human body by cadaver dissection. Hum Biol 1984;56:459-473. 4. Campbell WW, Crim M C , Young VR, Joseph LJ , Evans WJ. Effects of resistance training and dietary protein intake on protein metabolism in older adults. A m J Physiol (Endocrinol Metab 31) 1995;268:E1143-E1153. 5. Campbell WW, Barton Jr. M L , Cyr-Campbell D, Davey SL, Beard JL, Parise G, Evans WJ. Effects of an omnivorous diet compared with a lactoovovegetarian diet on resistance-training-induced changes in body composition and skeletal muscle in older men. A m J Clin Nutr 1999;70:1032-1039. 6. Tayter M , Stanek K L . Anthropometric and dietary assessment of omnivore and lacto-ovo-vegetarian children. J A m Diet Assoc 1989;89:1661-1663. 7. Dwyer Jt, Andrew E M , Valadian I, Reed RB. Size, obesity, and leaness in vegetarian preschool children. J A m Diet Assoc 1980;77:434-437. 8. Hebbelinck M , Clarys P, Malsche A. Growth, development, and physical fitness of Flemish vegetarian children, adolescents, and adults. A m J Clin Nutr 1999;70 (suppl):579S-585S. 9. Barr SI, Prior JC, Janelle K C , Lentle B C . Spinal bone mineral density in premenopausal vegetarian and nonvegetarian women: cross-sectional and prospective comparisons. J A m Diet Assoc 1998;98:760-765. 10. Barbosa JC, Shultz TD, Filley SJ, Nieman DC. The relationship among adiposity, diet, and hormone concentrations in vegetarian and nonvegetarian postmenopausal women. A m J Clin Nutr 1990;51:798-803. 11. Parsons TJ, Van Dusseldorp M , Van Der Vliet M , Van De Werken K, Schaafsma G, Van Staveren WA. Reduced bone mass in dutch adolescents fed a macrobiotic diet in early life. J Bone Min Res 1997;12:1486-1494. 12. Nathan I, Hackett AF, Kirby S. A longitudinal study of the growth of matched pairs of vegetarian and omnivorous children, aged 7-11 years, in the North-West of England. Eur J Clin Nutr 1997;51:20-25. 44 13. Lee RC, Wang Z, Heo M, Ross R, Janssen I, Heymsfield SB. Total-body skeletal muscle mass: development and cross-validation of anthropometric prediction models. Am J ClinNutr 2000;72:796-803. 14. Dwyer JT, Dietz Jr WH, Andrews EM, Suskind RM. Nutritional status of vegetarian children. Am J Clin Nutr 1982;35:204-216. 15. Howie BJ, Shultz TD. Dietary and hormonal interrelationships among vegetarian Seventh-Day Adventists and nonvegetarian men. Am J ClinNutr 1985;42:127-134. 16. Jackson AS, Pollock ML. Generalized equations for predicting body density of men. Brit J Nutr 1978;40:497-504. 17. Lukaski H. Estimation of muscle mass. In: A. Roche, S Heymsfield, T. Lohman, ed. Human Body Composition. Champaign 111: Human Kinetics Publishing, 1996:109-128. 18. Borsook H, Dubnoff JW. The hyrolysis of phosphocreatine and the origin of urinary creatinine. J Biol Chem 1947;168:493-510. 19. Martin AD, Spenst LF, Drinkwater DT, Clarys JP. Anthropometric estimation of muscle mass in men. Med Sci Sports Exerc 1990;22:729-733. 20. Heymsfield SB, McManus C, Smith J, Stevens V, Nixon DW. Anthropometric measurement of muscle mass: revised equations for calculating bone-free arm muscle area. Am J Clin Nutr 1982;36:680-690. 21. Norton K. Anthropometric estimation of body fat. In: Norton K, ed. Anthropometrica. Sydney: University of New South Wales Press, 1996:171-198. 22. Siri WE. Body volume measurement by gas dilution. In: Behnke A, ed. Techniques for measuring body composition. Washington DC: National Academy of Sciences, 1961:108-117. 23. Martin AD, Ross WD, Drinkwater DT, Clarys JP. Prediction of body fat by skinfold caliper: assumptions and cadaver evidence. Int J Obes 1985;9 (Suppl. l):31-39. 24. Janelle KC, Barr SI. Nutrient Intakes and eating behaviour scores of vegetarian and nonvegetarian women. J Am Diet Assoc 1995;95:180-189. 25. Stunkard AJ, Messick S. The Three-Factor Eating Questionnaire to measure dietary restraint, disinhiition and hunger. J Psychosom Res 1985;29:71-83. 26. Martins Y, Pliner P, O'Connor K. Restrained eating among vegetarians: does a vegetarian eating style mask concerns about weight. Appetite 1999;32:145-154. 45 27. Key TJA, Roe L, M . Thorogood M , Moore JW, Clark GMG, Wang DY. Testosterone, sex hormone-binding globulin, calculated free testosterone, and estradiol in male begans and omnivores. Brit J Nutr 1990;64:111-119. 28. Belanger A, Locong A, Noel C, Cusan L, Dupont A, Prevost J, Caron S, Sevigny J. Influence of diet on plasma steroid and sex plasma binding globulin levels in adult men. J Steroid Biochem 1989;32:829-833. 29. Tesar R, Notelovitz M , Shim E, Kauwell G, Brown J. Axial and peripheral bone density and nutrient intakes of postmenopausal vegetarian and omnivorous women. Am J Clin Nutr 1992;56:699-704. 30. Lloyd T, Schaeffer JM, Walker MA, Demers L M . Urinary hormonal concentrations and spinal bone densities of premenopausal vegetarian and nonvegetarian women. Am J Clin Nutr 1991;54:1005-1010. 31. Woo J, Kwok T, Ho SC, Sham A, Lau E. Nutritional staus of elderly Chinese vegetarians. Age and Ageing 1998;27:455-461. 32. Loebel CC, Kraemer WJ. A brief review: testosterone and resistance exercise in men. J Strength Cond Res 1998;12:57-63. 33. Vermeulen A. Physiology of the testosterone-binding globulin in man. Ann NY Acad Sci 1988;538:103-111. 34. Longcope C, Geldman HA, Mckinlay JB, Araujo AB. Diet and sex hormone-binding globulin. J Clin Endocrinol Metab 2000;85:293-296. 35. Raben A, Kiens B, Richter EA, Rasmussen LB, Svenstrup B, Micic S, Bennett P. Serum sex hormones and endurance performance after a lacto-ovo vegetarian and mixed diet. Med Sci Sports Exerc 1992;24:1290-1297. 36. Godin G, Shephard RJ. A simple method to assess exercise behaviour in the community. Can J Appl Sport Sci 1985;10:141-146. 37. Nelson M . Methods for data collection at an individual level. In: Staveren WAV, ed. Manual on methodology for food consumption studies. New York: Oxford University Press:64-74. 38. Jacobs DRJ, Ainsworth BE, Hartman TJ, Leon AS. A simultaneous evaluation of 10 community used physical activity questionnaires. Med Sci Sports Exerc 1993;25:81-91. 39. Sallis JF, Buono MJ, Roby JJ, Micale FG, Nelson JA. Seven-day recall and other physical activity self-reports in children and adolescents. Med Sci Sports Exerc 1993;25:99-108. 46 40. Lohman TC, Roche AT, Martorell R. Anthropometric standaridization reference manual. Champaing. IL: Human kinetics, 1998. 41. Behnke AR, Wilmore JH. Evaluation and regulation of body build and composition. Englewood Cliffs: Prentice Hall, 1974. 42. Esha Research. The food processor nutrition and fitness software. 7.6 ed. Salem OR, 2001. 43. Janssen I, Heymsfield SB, Wang Z, Ross R. Skeletal muscle mass and distribution in 468 men and women aged 18-88 yr. J Appl Physiol 2000;89:81-88. 44. Wang SM, Baumgartner R, Burastero S, Pierson Jr. RN, Heymsfield SB. Total body skeletal muscle mass measured directly by computerized axial tomography: results in healthy men. Clin Res 1992;40:A642. 45. National Academy of Sciences. Recommended Dietary Allowances 10th ed. Advanced Nutrition and Human Metabolism 3rd ed. Belmont Ca: Wadsworth/Thomson Learning, 1989:Inside front cover. 47 7. Appendices 48 Appendix III: Questionnaires Screening Questionnaire Name: Age (years): 1. In the last two years, have you consumed meat (red meat, white meat, fish) more than once a month? Y or N b) If you answered no to question 1, how many years have you been consuming a vegetarian diet? c) If you answered yes to question 1, how many times a week do you consume meat? . How many years have you been consuming a meat containing diet? . 2. Do you participate in resistance training exercise? Y or N b) If so, how many times per week on average? 3. On average, how many hours of moderate intensity physical activity do you perform in a week? (i.e. recreational jogging, swimming, cycling, rock climbing, volleyball, tennis, etc) b) In what activities is the majority of this time spent? 4. Do you smoke? Y or N 5. Are you presently on, or have you been on any medication in the past year that may have an affect on your muscle or fat levels? Y or N 6. Place a checkmark beside one of the following. Are you African Canadian , Asian , Caucasian , Hispanic , Other . 53 The Three-Factor Eating Questionnaire Part 1 Instructions: Please circle the response that best applies to you. la. (For Meat eaters) When I smell a sizzling steak or see a juicy piece of meat, I find it very difficult to keep from eating, even if I have just finished eating a meal True False lb. (For Vegetarians) When I smell freshly baked bread I find it very difficult to keep from eating, even if I have just finished a meal True False 2.1 usually eat too much at social occasions, like parties and picnics True False 3.1 am usually so hungry that I eat more than three times a day True False 4. When I have eaten my quota of calories, I am usually good about not eating any more True False 5. Dieting is so hard for me because I just get too hungry True False 6.1 deliberately take small helpings as a means of controlling my weight True False 7. Sometimes things just taste so good that 1 keep on eating even when I am no longer hungry True False 8. Since I am often hungry, I sometimes wish that while I am eating, an expert would tell me that I have had enough or that I can have something more to eat True False 9. When I feel anxious, I find myself eating True False 10. Life is too short to worry about dieting True False 11. Since my weight goes up and down, I have gone on reducing diets more than once True False 12.1 often feel sp hungry that I just have to eat something True False 13. When I am with someone who is overeating, I usually overeat too True False 14.1 have a pretty good idea of the number of calories in common food True False 15. Sometimes when I start eating, I just can't seem to stop True False 16. It is not difficult for me to leave something on my plate True False 17. At certain times of the day, I get hungry because I have gotten used to eating then True False 18. While on a diet, if I eat food that is not allowed, I consciously eat less for a period of time to make up for it. True False 19. Being with someone who is eating often makes me hungry enough to eat also True False 20. When I feel blue, 1 often overeat True False 54 21.1 enjoy eating too much to spoil it by counting calories or watching my weight True False 22. When 1 see a real delicacy, I often get so hungry that I have to eat right away True False 23.1 often stop eating when I am not really full as a conscious means of limiting the amount that I eat. True False 24.1 get so hungry that my stomach often seems like a bottomless pit True False 25. My weight has hardly changed at all in the last ten years True False 26.1 am always hungry so it is hard for me to stop eating before I finish the food on my plate True False 27. When I feel lonely, I console myself by eating True False 28.1 consciously hold back at meals in order not to gain weight True False 29.1 sometimes get very hungry late in the evening or at night True False 30.1 eat anything I want, any time I want True False 31. Without even thinking about it, I take a long time to eat True False 32.1 count calories as a conscious means of controlling my weight True False 33.1 do not eat some foods because they make me fat True False 34.1 am always hungry enough to eat at any time True False 35.1 pay a great deal of attention to changes in my figure True False 36. While on a diet, if 1 eat a food that is not allowed, I often then splurge and eat other high calorie foods. True False Part 2 Instructions: Please answer the following questions by circling the number above the response that is appropriate to you. 37. How often are you dieting in a conscious effort to control your weight? 1 2 3 4 rarely sometimes usually always 38. Would a weight fluctuation of 51bs affect the way you live your life? 1 2 3 4 not at all slightly moderately very much 39. How ofen do you feel hungry? 1 2 3 4 only at mealtimes sometimes between often between meals almost always meals 55 40. Do your feelings of guilt about overeating help you control your food intake? 1 2 3 never rarely often 4 always 41. How difficult would it be for you to stop eating halfway through dinner and not eat for the next four hours? 1 2 3 4 easy slightly difficult moderately difficult very difficult 42. How conscious are you of what you are eating? 1 2 not at all slightly moderately extremely 43. How frequently do you avoid \"stocking up\" on tempting foods? 1 2 3 almost never seldom usually almost always 44. How likely are you to shop for low calorie foods? 1 2 unlikely slightly unlikely moderately likely very likely 45. Do you eat sensibly in front of others and splurge alone? 1 2 3 never rarely often 4 always 46. How likely are you to consciously eat slowly in order to cut down on how much you eat? 1 2 3 4 unlikely slightly likely moderately likely very likely 47. How frequently do you skip dessert because you are no longer hungry? 1 2 3 almost never seldom at least once a week almost every day 48. How likely are you to consciously eat less than you want? 1 2 3 unlikely slightly likely moderately likely very likely 49. Do you go on eating binges though you are not hungry? 1 2 3 never rarely sometimes at least once a week 50. On a scale of 0 to 5, where 0 means no restraint in eating (eating whatever you want, whenever you want it) and 5 means total restraint (constantly limiting food intake and never \"giving in\"), what number would you give yourself. 0 eat whatever you want, whenever you want it 1 usually eat whatever you want, whenever you want it 2 often eat whatever you want, whenever you want it 3 often limit food intake, but often \"give in\" 4 usually limit food intake, rarely \"give in\" 5 constantly limiting food intake, never \"giving in\" 51. To what extent does this statement describe your eating behavior? \"I start dieting in the morning, but because of any number of things that happen during the day, by evening I have given up and eat what I want, promising myself to start dieting again tomorrow. 1 2 3 4 not like me little like me pretty good describes me description of me perfectly 57 Godin Leisure-Time Exercise Questionnaire Considering a 7-Day period (a week), how many times on the average do you do the following kinds of exercise for more than 15 minutes during your free time (write on each line the appropriate number). Include the average duration of your exercise bouts. a) STRENUOUS EXERCISE (HEART BEATS RAPIDLY) (i.e. running, jogging, hockey, football, soccer, squash, basketball, cross country skiing, judo, roller skating, vigorous swimming, vigorous long distance cycling) b) MODERATE EXERCISE (NOT EXHAUSTING) (i.e. fast walking, baseball, tennis, easy bicycling, volleyball, badminton, easy swimming, alpine skiing, popular and folk dancing) c) MDLD EXERCISE (MINIMAL EFFORT) (i.e. yoga, archery, fishing from river band, bowling, horseshoes, golf, snow-> mobiling, easy walking) 2. Considering a 7-Day period (a week), during your leisure-time, how often do you engage in any regular activity long enough to work up a sweat (heart beats rapidly)? Times per week (more than 15 mins) Avg Duration O F T E N SOMETIMES N E V E R / R A R E L Y 1. • 2. • 3. • 58 Appendix IV: Anthropometric Proforma Name (last, given) Date Birthdate Subject # Height & Weight 1) Stature (cm) 2) Body Mass (kg) Skinfolds 3) Biceps 2 cm above biceps 2 cm below biceps 4) Triceps 2 cm above triceps 2 cm below triceps 5) Forearm 2 cm above forearm 2 cm below forearm 6) Subscapular 7) Supra-ilium 8) Chest 9) Axilla 10) Front thigh 2 cm above thigh 2 cm below thigh 11) Medial calf 2 cm above calf 2 cm below calf 12) Abdomen Girths 13) Biceps 2 cm above biceps 2 cm below biceps 14) forearm 2 cm above forearm 2 cm below forearm 15) Mid-thigh 2 cm above mid-thigh 2 cm below mid-thigh 16) Calf 2 cm above calf 2 cm below calf Appendix V: Distribution of Dependent Variables Figure 5: Vegetarian % Fat Figure 6: Omnivore % Fat 4 6 8 10 12 14 16 18 20 4 6 8 10 12 14 16 18 20 Figure 7: Vegetarian 26 Skinfolds (mm) Figure 8: Omnivore £6 Skinfolds (mm) 30 50 70 90 110 30 50 70 90 110 130 Figure 9: Vegetarian GLTEQ Score 7 r — 1 : n 0 20 40 60 80 100 120 140 Figure 10: Omnivore GLTEQ Score 7' 6' 5' Figure 11: Vegetarian Dietary Restraint 10| 1 0 5 10 15 Figure 12: Omnivore Dietary Restraint 61 Figure 15: Vegetarian Activity (hr/week) Figure 16: Omnivore Activity (hr/week) 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 Figure 17: Vegetarian Calories Figure 18: Omnivore Calories 4000 8000 12000 16000 20000 4000 8000 12000 16000 20000 62 63 Figure 25: Vegetarian Sat. Fat (g) Figure 26: Omnivore Sat. Fat (g) 25 75 125 175 225 275 50 150 250 Figure 27: Vegetarian Mono. Fat (g) Figure 28: Omnivore Mono. Fat (g) Figure 29: Vegetarian Poly. Fat Intake (g) ST 30 50 70 90 110 130 150 Figure 30: Omnivore Poly. Fat Intake (g) 10 30 50 70 90 110 130 64 Figure 31: Vegetarian P : S Ratio Figure 32: Omnivore P:S Ratio 65 AppendixVI: Instructions For 3-Day Diet Records An accurate dietary intake record can provide valuable information about the nutritional content of an individual's usual diet. Please try and maintain your normal eating patterns in terms of content and quantity of foods eaten during this 3-day period. Please keep a record of everything you eat or drink on the attached forms for 3 days in a row (two weekdays and one weekend day). Please be as specific as possible. *To ensure accuracy please try to record immediately after eating. *The more accurate you record; the more meaningful the analysis. Be sure to include: 1. A L L F O O D S A N D D R I N K S consumed including snacks, soft drinks, alcohol, cream and sugar in coffee/tea, butter/sauces on vegetables, jams, relishes, candies, butter/margarine/mayonnaise on sandwiches, salad dressing. Break combination foods down into their constituents (e.g. ham and cheese omelette= 3 eggs + 1 oz. Cheddar cheese + 1 slice Oscar Meyer Packaged ham Slices + 1 tsp butter in pan). 2. T H E A M O U N T O F F O O D that was consumed. It is extremely important that accurate measurements be recorded. Please use the measuring cups and spoons provided to measure the volumes of foods consumed whenever possible. * Use the V O L U M E measures such as cups, tablespoons (Tbs.), teaspoons (Tsp.) or millilitres (ml) for soups, pasta, cereals, rice, other grains, small or cut vegetables, cut fruit, tinned foods, drinks, sauces, salad dressings, butter, mayonnaise, margarine, jams, peanut butter etc. Please be as accurate as possible. For example record whether a tablespoon is 'heaping' or 'level'. 66 Appendix VII: One Way ANOVA comparison of means Descriptives 95% Confidence Interval for Mean N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum AGE 1.00 27 26.7037 3.20834 .61745 25.4345 27.9729 20.00 32.00 2.00 27 25.7778 4.83841 .93115 23.8638 27.6918 20.00 34.00 Total 54 26.2407 4.09295 .55698 25.1236 27.3579 20.00 34.00 HEIGHT 1.00 27 1.76822 .070671 .013601 1.74027 1.79618 1.655 1.955 2.00 27 1.79537 .064081 .012332 1.77002 1.82072 1.631 1.917 Total 54 1.78180 .068208 .009282 1.76318 1.80041 1.631 1.955 WEIGHT 1.00 27 72.9926 10.15022 1.95341 68.9773 77.0079 59.30 96.60 2.00 27 77.0444 9.86923 1.89934 73.1403 80.9486 59.60 102.60 Total 54 75.0185 10.12449 1.37777 72.2551 77.7820 59.30 102.60 MUSCMASS 1.00 27 30.9330 3.45494 .66490 29.5662 32.2997 25.23 37.39 2.00 27 32.6689 4.06025 .78140 31.0627 34.2751 24.15 41.90 Total 54 31.8009 3.83543 .52194 30.7541 32.8478 24.15 41.90 SUM12CGB 1.00 27 2471207 415931.28969 80046.01 2306669.715 2635743.591 1809113 3434090 2.00 27 2683571 413584.70636 79594.41 2519962.268 2847179.590 1923126 3520751 Total 54 2577389 424578.83087 57777.86 2461501.078 2693276.504 1809113 3520751 PERCFAT 1.00 27 10.4974 3.58599 .69012 9.0788 11.9160 4.50 19.27 2.00 27 12.0207 4.46554 .85939 10.2542 13.7873 4.09 19.21 Total 54 11.2591 4.08434 .55581 10.1443 12.3739 4.09 19.27 SUM6SF 1.00 27 70.7333 22.66359 4.36161 61.7679 79.6988 32.00 123.20 2.00 27 80.6852 29.19867 5.61929 69.1346 92.2358 33.80 144.70 Total 54 75.7093 26.37117 3.58866 68.5113 82.9072 32.00 144.70 ANOVA Sum of Squares df Mean Square F Sig. AGE Between Groups 11.574 1 11.574 .687 .411 Within Groups 876.296 52 16.852 Total 887.870 53 HEIGHT Between Groups .010 1 .010 2.187 .145 Within Groups .237 52 .005 Total .247 53 WEIGHT Between Groups 221.636 1 221.636 2.212 .143 Within Groups 5211.145 52 100.214 Total 5432.781 53 MUSCMASS Between Groups 40.681 1 40.681 2.863 .097 Within Groups 738.978 52 14.211 Total 779.659 53 SUM12CGB Between Groups 6.09E+11 1 6.088E+11 3.539 .066 Within Groups 8.95E+12 52 1.720E+11 Total 9.55E+12 53 PERCFAT Between Groups 31.327 1 31.327 1.910 .173 Within Groups 852.810 52 16.400 Total 884.137 53 SUM6SF Between Groups 1337.031 1 1337.031 1.957 .168 Within Groups 35521.214 52 683.100 Total 36858.245 53 67 Descriptives 95% Confidence Interval for Mean N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum GLTEQ 1.00 27 51.7778 27.02183 5.20035 41.0883 62.4673 3.00 140.00 2.00 27 46.8333 23.31804 4.48756 37.6090 56.0576 14.00 104.00 Total 54 49.3056 25.12298 3.41881 42.4483 56.1628 3.00 140.00 RESTRAIN 1.00 27 5.5556 3.45669 .66524 4.1881 6.9230 1.00 18.00 2.00 27 5.5185 4.33563 .83439 3.8034 7.2336 .00 17.00 Total 54 5.5370 3.88375 .52851 4.4770 6.5971 .00 18.00 TFEQ 1.00 27 14.8519 5.09678 .98088 12.8356 16.8681 5.00 30.00 2.00 27 16.6667 7.92270 1.52472 13.5325 19.8008 8.00 38.00 Total 54 15.7593 6.66145 .90651 13.9410 17.5775 5.00 38.00 ACTIVITY 1.00 27 4.6296 1.82711 .35163 3.9069 5.3524 .00 7.00 2.00 27 3.8085 1.75114 .33701 3.1158 4.5012 .00 6.00 Total 54 4.2191 1.82036 .24772 3.7222 4.7159 .00 7.00 CALORIES 1.00 27 10780.85 3626.15352 697.85357 9346.3948 12215.3119 4304.18 19210.17 2.00 27 10608.12 3914.82386 753.40820 9059.4691 12156.7746 4692.61 22133.05 Total 54 10694.49 3738.50061 508.74549 9674.0734 11714.9018 4304.18 22133.05 PROTEIN 1.00 27 354.5389 123.71721 23.80939 305.5980 403.4798 183.62 642.61 2.00 27 452.6711 209.95578 40.40601 369.6154 535.7268 190.18 1106.86 Total 54 403.6050 177.72546 24.18537 355.0953 452.1147 183.62 1106.86 FIBER 1.00 27 139.8137 76.63467 14.74835 109.4980 170.1294 60.00 418.46 2.00 27 92.3381 43.28275 8.32977 75.2161 109.4602 26.93 206.05 Total 54 116.0759 66.13751 9.00018 98.0239 134.1280 26.93 418.46 TOTFAT 1.00 27 331.2996 136.24622 26.22060 277.4024 385.1968 106.15 726.71 2.00 27 365.7841 177.58103 34.17548 295.5354 436.0328 120.79 984.28 Total 54 348.5419 157.73176 21.46457 305.4894 391.5943 106.15 984.28 SATFAT 1.00 27 93.7493 62.10149 11.95144 69.1827 118.3158 24.09 275.03 2.00 27 129.0722 71.41882 13.74456 100.8199 157.3246 28.64 400.75 Total 54 111.4107 68.64352 9.34120 92.6747 130.1468 24.09 400.75 MONOFAT 1.00 27 89.4463 44.79619 8.62103 71.7255 107.1671 25.20 217.00 2.00 27 101.1489 54.96715 10.57843 79.4046 122.8932 34.08 271.94 Total 54 95.2976 50.01492 6.80617 81.6461 108.9490 25.20 271.94 POLYFAT 1.00 27 67.5107 32.68529 6.29029 54.5809 80.4406 27.98 149.26 2.00 27 42.7419 27.50983 5.29427 31.8593 53.6244 13.22 138.20 Total 54 55.1263 32.42853 4.41296 46.2750 63.9776 13.22 149.26 PUFTOSAT 1.00 27 .9744 .62948 .12114 .7254 1.2234 .17 2.62 2.00 27 .3628 .19139 .03683 .2871 .4385 .16 1.02 Total 54 .6686 .55464 .07548 .5172 .8200 .16 2.62 68 ANOVA Sum of Squares df Mean Square F Sig. GLTEQ Between Groups 330.042 1 330.042 .518 .475 Within Groups 33121.667 52 636.955 Total 33451.708 53 RESTRAIN Between Groups .019 1 .019 .001 .972 Within Groups 799.407 52 15.373 Total 799.426 53 TFEQ Between Groups 44.463 1 44.463 1.002 .321 Within Groups 2307.407 52 44.373 Total 2351.870 53 ACTIVITY Between Groups 9.102 1 9.102 2.842 .098 Within Groups 166.525 52 3.202 Total 175.627 53 CALORIES Between Groups 402788.2 1 402788.223 .028 .867 Within Groups 7.40E+08 52 14237417.58 Total 7.41 E+08 53 PROTEIN Between Groups 130004.1 1 130004.096 4.378 .041 Within Groups 1544072 52 29693.689 Total 1674076 53 FIBER Between Groups 30428.033 1 30428.033 7.856 .007 Within Groups 201403.0 52 3873.135 Total 231831.0 53 TOTFAT Between Groups 16053.888 1 16053.888 .641 .427 Within Groups 1302549 52 25049.026 Total 1318603 53 SATFAT Between Groups 16844.108 1 16844.108 3.761 .058 Within Groups 232888.3 52 4478.621 Total 249732.4 53 MONOFAT Between Groups 1848.834 1 1848.834 .735 .395 Within Groups 130730.2 52 2514.043 Total 132579.1 53 POLYFAT Between Groups 8282.221 1 8282.221 9.076 .004 Within Groups 47453.094 52 912.559 Total 55735.315 53 PUFTOSAT Between Groups 5.050 1 5.050 23.331 .000 Within Groups 11.255 52 .216 Total 16.304 53 69 Appendix VIII: Mann-Whitney U-Tests Ranks DIETGRP N Mean Rank Sum of Ranks GLTEQ 1.00 27 29.61 799.50 2.00 27 25.39 685.50 Total 54 ACTIVITY 1.00 27 31.00 837.00 2.00 27 24.00 648.00 Total 54 TFEQ 1.00 27 27.11 732.00 2.00 27 27.89 753.00 Total 54 RESTRAIN 1.00 27 28.50 769.50 2.00 27 26.50 715.50 Total 54 Test Statistics3 GLTEQ ACTIVITY TFEQ RESTRAIN Mann-Whitney U 307.500 270.000 354.000 337.500 Wilcoxon W 685.500 648.000 732.000 715.500 Z -.986 -1.657 -.182 -.470 Asymp. Sig. (2-tailed) .324 .097 .855 .639 a- Grouping Variable: DIETGRP Ranks DIETGRP N Mean Rank Sum of Ranks CALORIES 1.00 27 28.04 757.00 2.00 27 26.96 728.00 Total 54 PROTEIN 1.00 27 23.78 642.00 2.00 27 31.22 843.00 Total 54 FIBER 1.00 27 33.33 900.00 2.00 27 21.67 585.00 Total 54 TOTFAT 1.00 27 26.33 711.00 2.00 27 28.67 774.00 Total 54 SATFAT 1.00 27 22.89 618.00 2.00 27 32.11 867.00 Total 54 MONOFAT 1.00 27 25.93 700.00 2.00 27 29.07 785.00 Total 54 POLYFAT 1.00 27 34.65 935.50 2.00 27 20.35 549.50 Total 54 PUFTOSAT 1.00 27 36.11 975.00 2.00 27 18.89 510.00 Total 54 Test Statistics9 CALORIES PROTEIN FIBER TOTFAT SATFAT MONOFAT POLYFAT PUFTOSAT Mann-Whitney U 350.000 264.000 207.000 333.000 240.000 322.000 171.500 132.000 Wilcoxon W 728.000 642.000 585.000 711.000 618.000 700.000 549.500 510.000 Z -.251 -1.739 -2.725 -.545 -2.154 -.735 -3.339 -4.022 Asymp. Sig. (2-tailed) .802 .082 .006 .586 .031 .462 .001 .000 a - Grouping Variable: DIETGRP Appendix IX: Spearman correlation coefficients Correlations VIUSCMASS SUM12CGB PERCFAT SUM6SF GLTEQ RESTRAIN TFEQ Spearman's rho MUSCMASS Correlation Coefficie 1.000 .960*' -.100 -.076 -.172 -.051 -.027 Sig. (2-tailed) .000 .473 .585 .214 .716 .846 N 54 54 54 54 54 54 54 SUM12CGB Correlation Coefficie .960* 1.000 .006 .032 -.148 -.057 -.052 Sig. (2-tailed) .000 .964 .818 .286 .680 .710 N 54 54 54 54 54 54 54 PERCFAT Correlation Coefficie -.100 .006 1.000 .984* -.292* -.199 -.235 Sig. (2-tailed) .473 .964 .000 .032 .150 .087 N 54 54 54 54 54 54 54 SUM6SF Correlation Coefficie -.076 .032 .984* 1.000 -.255 -.182 -.226 Sig. (2-tailed) .585 .818 .000 .063 .187 .100 N 54 54 54 54 54 54 54 GLTEQ Correlation Coefficie -.172 -.148 -.292* -.255 1.000 .048 .024 Sig. (2-tailed) .214 .286 .032 .063 .728 .864 N 54 54 54 54 54 54 54 RESTRAIN Correlation Coefficie -.051 -.057 -.199 -.182 .048 1.000 .610* Sig. (2-tailed) .716 .680 .150 .187 .728 .000 N 54 54 54 54 54 54 54 TFEQ Correlation Coefficie -.027 -.052 -.235 -.226 .024 .610*1 1.000 Sig. (2-tailed) .846 .710 .087 .100 .864 .000 N 54 54 54 54 54 54 54 **• Correlation is significant at the .01 level (2-tailed). *• Correlation is significant at the .05 level (2-tailed). 72 Appendix X: ANCOVA comparison of muscle mass Univariate Analysis of Variance Between-Subjects Factors Value Label N DIETGRP 1.00 2.00 Vegetarian Omnivores 27 27 Descriptive Statistics Dependent Variable: MUSCLE DIETGRP Mean Std. Deviation N Vegetarian 30.5985 3.41132 27 Omnivores 32.6689 4.06025 27 Total 31.6337 3.85848 54 Tests of Between-Subjects Effects Dependent Variable: MUSCLE Source Type III Sum of Squares df Mean Square F Siq. Corrected Model 70.8753 2 35.438 2.517 .091 Intercept 25.260 1 25.260 1.794 .186 HEIGHT 13.009 1 13.009 .924 .341 DIETGRP 45.258 1 45.258 3.214 .079 Error 718.183 51 14.082 Total 54826.384 54 Corrected Total 789.059 53 a - R Squared = .090 (Adjusted R Squared = .054) Estimated Marginal Means 1. Grand Mean Dependent Variable: MUSCLE Mean Std. Error 95% Confidence Interval Lower Bound Upper Bound 31.6343 .511 30.609 32.659 a - Evaluated at covariates appeared in the model: HEIGHT = 1.7818. 2. DIETGRP Estimates Dependent Variable: MUSCLE 95% Confidence Interval DIETGRP Mean Std. Error Lower Bound Upper Bound Vegetarian 30.6993 .730 29.234 32.164 Omnivores 32.568a .730 31.103 34.033 a - Evaluated at covariates appeared in the model: HEIGHT = 1.7818. Pain/vise Comparisons Dependent Variable: MUSCLE (I) DIETGRP (J) DIETGRP Mean Difference (l-J) Std. Error Sig. a 95% Confidence Interval for Difference3 Lower Bound Upper Bound Vegetarian Omnivores -1.869 1.043 .079 -3.962 .224 Omnivores Vegetarian 1.869 1.043 .079 -.224 3.962 Based on estimated marginal means a- Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments). Univariate Tests Dependent Variable: MUSCLE Sum of Squares df Mean Square F Sig. Contrast 45.258 1 45.258 3.214 .079 Error 718.183 51 14.082 The F tests the effect of DIETGRP. This test is based on the linearly independent pairwise comparisons among the estimated marginal means. 74 "@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2003-05"@en ; edm:isShownAt "10.14288/1.0077237"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Human Kinetics"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Body composition of vegetarian and omnivorous men"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/14273"@en .