UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Factors associated with calcium intake in adolescent athletes Webster, Brenda L. 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_spring_webster_brenda.pdf [ 5.09MB ]
Metadata
JSON: 831-1.0086178.json
JSON-LD: 831-1.0086178-ld.json
RDF/XML (Pretty): 831-1.0086178-rdf.xml
RDF/JSON: 831-1.0086178-rdf.json
Turtle: 831-1.0086178-turtle.txt
N-Triples: 831-1.0086178-rdf-ntriples.txt
Original Record: 831-1.0086178-source.json
Full Text
831-1.0086178-fulltext.txt
Citation
831-1.0086178.ris

Full Text

FACTORS ASSOCIATED WITH CALCIUM INTAKE IN ADOLESCENT ATHLETESbyBRENDA LEA WEBSTERB.S.H.Ec., The University of Saskatchewan, 1986A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(School of Family and Nutritional Sciences)We accept this thesis as conformingto the required standardsTHE UNIVERSITY OF BRITISH COLUMBIADecember 1992©Brenda Lea Webster, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department of 1Q,44 , h i^fy i t- 00a. I  5c_^e_sThe University of British ColumbiaVancouver, CanadaDate  De 6- 2 ,4/ bee- ig;^92_DE-6 (2/88)ABSTRACTThis study investigated dietary calcium intakes, and factors potentially affectingthem, in adolescent athletes of both genders competing in either an aesthetic sport(gymnastics) or non-aesthetic sport (speed skating). For all athletes, data collectionwas conducted using a self-administered questionnaire which assessed the following:estimated calcium intake using a validated food frequency questionnaire, pubertalstatus, demographics (age, gender, racial origin, use of vitamin/mineral supplements,employment, allergies to milk or dairy products, parental socioeconomic status),training volume, competitive level, and social environmental factors (differentialassociation (ie. family, friends, health experts, media), lifestyle, dieting behaviour,modelling behaviour and social and non-social reinforcement). Body composition wasassessed using skinfold measurements.The athletes competed at a minimum of a provincial level in either gymnastics(males =25, females =32) or speed skating (males =32, females =25). The meanages of the male athletes were similar for the two sports (14.4 ± 1.8 yrs, 3i ± SD vs14.7± 1.9 yrs for skaters and gymnasts respectively); however, the skaters weretaller (170.2 ± 10.3 cm vs 158.5 ± 13.0 cm, p <0.05), heavier (60.5 ± 11.9 kg vs50.1 ± 12.5 kg, p <0.05) and had higher estimated percent body fat (%13F)(8.0 ± 3.4% vs 3.7 ± 2.1%, p <0.001) and sum of skinfold measurements (S4SF)(35.8 ± 9.8 mm vs 23.9 ± 3.6 mm, p <0.001). Similar to the males, the mean agesof the female athletes were similar for skaters and gymnasts (14.3 ± 1.6 yrs vs14.1 ±1.6 yrs); however, the skaters were taller (162.2 ±8.2 cm vs 153.6 ± 4.9 cm,p < 0.001), heavier (58.5 ±9.5 kg vs 45.2 7.9 kg, p < 0.001) and had higheriiestimated %BF (21.0 ± 7.7% vs 9.4 ± 5.7%, p <0.001) and S4SF (65.5 ± 18.9 mmvs 38.5 ± 12.3 mm, p <0.001). A 2X2 ANOVA (gender X sport) was used to assessdifferences in total dietary calcium intake. The analysis showed that there was asignificant main effect of sport (F 1100 = 6.63, p =0.01), indicating that averaged overgender, the skaters had significantly higher calcium intakes than the gymnasts. Therewas no main effect of gender (F 1 , 100 = 2.90, p =0.09) and no interaction betweengender and sport (F 1 , 100 = 0.52, p = 0.47), meaning that the difference in calcium intakebetween the two sports was similar for both genders. Additionally, all groups ofathletes exceeded the recommended nutrient intake for calcium.Among the independent variables, 2X2 ANOVA (gender by sport) revealedsignificant main effects for scores on the dieting sub-scale and for socialreinforcement score. A significant main effect of gender was detected for scores onthe dieting sub-scales (F 1 , 105 = 21.86, p <0.001), meaning that averaged over the twosports, the female athletes scored significantly higher than the male athletes. Nosignificant effect of sport was detected, nor was an interaction effect detected.Although a gender difference existed for the dieting sub-scale scores, neither themean scores of the male nor the female athletes suggested tendencies towardsdisturbed eating behaviours. ANOVA of the social reinforcement variable showed thatthere was a significant main effect of sport (F 1 , 107 = 5.78, p = 0.02), indicating thatwhile athletes in both sports disagreed that consuming milk evoked positive feelingsand a sense of belonging to a group, the gymnasts exhibited stronger disagreementto the statements. No significant effect of gender was detected, nor was there aninteraction effect.Stepwise forward entry multiple regression analysis (MRA) was used toiiidetermine the variable(s) which best predicted total dietary calcium intake. Twomodels were analyzed using MRA, a Traditional Model which included age, gender,%BF, weight and sport, and a Social Model which included dieting sub-scale score,differential association score, social and non-social reinforcement scores, modellingbehaviour score, and a lifestyle score. The analysis showed that for all athletescombined, only one variable from the Social Model (social reinforcement) and onevariable from the Traditional Model (sport) significantly explained variance in totaldietary calcium. Each variable explained 6% and 10% of the variance in thedependent variable respectively. When the athletes were divided by sport and gender,different relationships emerged. For the male athletes, only differential associationcould significantly predict total dietary calcium intake, explaining 9% of the variancein the dependent variable. For female athletes combined, only one variable from theSocial Model (modelling behaviour) and one variable from the Traditional Model (sport)entered the predictive equations, explaining only 9% and 15% of the variance incalcium intake, respectively. For the female skaters, differential association explained21 % of the variance in the dependent variable, while for the female gymnasts, non-social reinforcement explained 18% of the variance in total dietary calcium intake.The results from this study show that for all athletes combined, variables fromneither the Social nor Traditional Models were strong predictors of total dietarycalcium intake. The social variables explained more variance in the dependent variablethan the traditional variables only when the athletes were divided by gender and sport,and still approximately 80% of the variance was left unexplained Therefore, thevariables studied were not predictors of total dietary calcium intake in these athletes.ivTABLE OF CONTENTSABSTRACT ^  iiTABLE OF CONTENTS ^LIST OF TABLES  viiACKNOWLEDGMENTS ^  ixCHAPTER I ^  1INTRODUCTION ^  1CHAPTER II ^  4LITERATURE REVIEW ^  41. Introduction  ^42. Calcium Intake  ^43. Assessing Calcium Intake  ^74. Dieting Behaviour ^  105. Evaluating Eating Disturbances ^  126. Factors Influencing Eating Behaviour  147. Body Composition ^  188. Summary ^  219.^Hypotheses  22CHAPTER III  23METHODOLOGY ^  231. Subjects  232. Data Collection ^  23A) Body Composition ^  23B) Questionnaire  24i) Maturation Status ^  25ii) Employment  25iii) Supplement Usage  25iv) Racial Origin ^  26v) Allergies  26vi) Training Volume and Competitive Level ^ 27vii) Socioeconomic Status ^  27viii) Calcium Intake ^  28ix) Dieting Behaviour  31x) Differential Association ^  31xi) Lifestyle ^  34xii) Social Reinforcement  34xiii) Modelling Behaviour ^  37xiv) Non-Social Reinforcement (Taste) ^ 393.^Data Analysis ^  39CHAPTER IV ^  43RESULTS  431. Internal Consistency of the Questionnaire ^ 432. Descriptive Characteristics of Athletes  45A) Males ^  45B) Females  503. Food Frequency Questionnaire ^  554. Dieting Sub-Scale ^  655. Differential Association Variable  686. Lifestyle Variable  717. Social Reinforcement Variable ^  748. Modelling Behaviour Variable  779. Non-Social Reinforcement (Taste) Variable ^ 8010. Univariate Correlation Analyses ^  83A) All Athletes ^  83B) Males  83C)^Females  8611.^Multiple Regression Analysis ^  88A)^All Athletes ^  89i) Traditional Model  89ii) Social Model ^  89Male Athletes  91C)^Female Athletes  92i) Traditional Model ^  92ii) Social Model ^  9312. Summary of Results  94CHAPTER V ^  97DISCUSSION^971. Estimated Dietary Calcium Intake ^  972. Factors Influencing Eating Behaviour  100A) Dieting Sub-Scale Score  100B) Social Model Variables ^  102C)^Traditional Model Variables  1043. Multiple Regression Analysis  1054. Conclusion and Recommendations for Future Research  ^112REFERENCES ^ 115122APPENDICESAppendix A:Appendix B:Appendix C:Appendix D:Definition of Terms ^  122Informed Consent Form, Ethics Approval Form andQuestionnaire  124Body Composition Equations and Constants ^ 136Differential Association Sub-Scale Scores  138viLIST OF TABLESCalcium Food Frequency Questionnaire ^  29Dieting Behaviour Questions from the Dieting Sub-Scale of theEating Attitudes Test ^  32Items Used to Determine Differential Association ^ 33Items Used to Determine Lifestyle ^  35Items Used to Determine Social Reinforcement Score ^ 36Items Used to Determine Modelling Behaviour Score  38Items Used to Determine Taste Enjoyment ^  40Cronbach's Alpha Reliability Coefficient for Questionnaire Scales^  44Descriptive Physical Characteristics of Male Athletes ^ 46Use of Vitamin Supplementation in Male Athletes  47Training Volume and Competitive Level of Male Athletes ^ 49Descriptive Physical Characteristics of Female Athletes  51Use of Vitamin Supplementation in Female Athletes ^ 52Training Volume and Competitive Level of Female Athletes.Number of Servings Per Month of Non-Dairy Product Foods For 54 Male Athletes ^  56Number of Serving Per Month of Dairy Products for MaleAthletes  57Number of Servings Per Month of Non-Dairy Product Foods ForFemale Athletes ^  58Number of Serving Per Month of Dairy Products for FemaleAthletes ^  59Calcium Intake for Male and Female Athletes ^  60Type of Milk Consumed by Male Athletes  63Type of Milk Consumed by Female Athletes  64Dieting Sub-Scale Scores for Male Athletes (means and standarddeviations) ^  66Dieting Sub-Scale Scores for Female Athletes ^  67Average Scores for Differential Association Sub-Scales and OverallDifferential Association Score for Male Athletes  69Average Scores for Differential Association Sub-Scales and OverallDifferential Association Score for Female Athletes ^ 70Number of Meals and Snacks Eaten and "Eaten Out" Per Week byMale Athletes ^  72Number of Meals and Snacks Eaten and "Eaten Out" Per Week byFemale Athletes  73Scores for Social Reinforcement Questions and Overall SocialReinforcement Score for Males ^  75Scores for Social Reinforcement Questions and Overall SocialReinforcement Score for Females  76Modelling Scores for Male Athletes ^  78Modelling Scores for Female Athletes  79Scores for Non-Social Reinforcement (Taste) Questions andOverall Non-Social Reinforcement Score for Male Athletes ^ 81Table 1:Table 2:Table 3:Table 4:Table 5:Table 6:Table 7:Table 8:Table 9:Table 10:Table 11:Table 12:Table 13:Table 14:Table 15:Table 16:Table 17:Table 18:Table 19:Table 20:Table 21:Table 22:Table 23:Table 24:Table 25:Table 26:Table 27:Table 28:Table 29:Table 30:Table 31:Table 32:viiScores for Non-Social Reinforcement (Taste) Questions andOverall Non-Social Reinforcement Score for Female Athletes ..^82Univariate Correlation Coefficients for Daily Calcium Intake andthe Independent Variables for All Athletes ^  84Univariate Correlation Coefficients for Daily Calcium Intake andthe Independent Variables for Male Athletes  85Univariate Correlation Coefficients for Daily Calcium Intake andthe Independent Variables for Female Athletes ^ 87Multiple Regression Analysis Summary Table  90Age and Sex Specific Constants for Conversion of Body Density,Water, and Potassium to Percent Fat in Children and Youths .. 137Scores for Friends Sub-Scale Questions and Overall Sub-ScaleScore for MalesScores for Friends Sub-Scale Questions and Overall Sub-ScaleScore for Females ^Scores for Family Sub-Scale Questions and Overall Sub-ScaleScore for Males Scores for Family Sub-Scale Questions and Overall Sub-ScaleScore for Females ^Scores for Media Sub-Scale Questions for Males and Overall Sub-Scale Score Scores for Media Sub-Scale Questions for Females and OverallSub-Scale Score ^Scores for Experts Sub-Scale Questions for Males and OverallSub-Scale Score Scores for Experts Sub-Scale Questions for Females and OverallSub-Scale Score ^Table 33:Table 34:Table 35:Table 36:Table 37:Table 38:Table 39:Table 40:Table 41:Table 42:Table 43:Table 44:Table 45:Table 46:138139140141142143144145viiiACKNOWLEDGMENTSI would like to express my sincere appreciation to my thesis supervisor, Dr. SusanBarr, for her excellent guidance and encouragement throughout my project. I wouldalso like to acknowledge my committee members, Dr. Susan Crawford and Dr. GwenChapman, for their valuable contributions to my project at all stages. I would like toextent a special thank you to Dr. Joseph Leichter for chairing my thesis defense.I thank the athletes who participated in this study, their coaches, and also Sport B.C.and Sask Sport for facilitating their participation. I gratefully acknowledge my familyfor their support through out the project.This project was supported by a grant from the Dairy Bureau of Canada (5-56801).ixCHAPTER II. INTRODUCTION:Nutritional requirements are increased during adolescence due to the naturalcourse of growth and development (Guenther, 1986; Moffatt, 1986; Perron andEndres, 1985). In addition to this increased need for growth and development,exercise and training may further escalate nutrient requirements during this time(Perron and Endres, 1985; Rucinski, 1989; Chen et al., 1989). It is well recognizedthat poor nutrition may impede athletic performance (Moffatt, 1986; Clark et al.,1988; Benson et al., 1990). For this reason, a nutritionally adequate diet may playan important role in both the athletic performance and the future health of adolescents(Moffatt, 1986; Perron and Endres, 1985; Chen et al., 1989).Relatively few studies, to date, have examined dietary patterns or nutritionpractices in adolescent athletes. The available literature, however, describes nutrientinadequacies, including inadequate calcium intake, in adolescent gymnasts, volleyballplayers, figure skaters and swimmers (Moffatt, 1986; Perron and Endres, 1986;Rucinski, 1989; Chen et al., 1989; Benson et al., 1990). Some research alsoindicates that participants in certain types of sports appear to be at greater risk forinadequate nutrient intake. In aesthetic sports such as gymnastics and figure skating,for example, excess body weight, especially body fat, is detrimental in competition(Moffatt, 1986; Rucinski, 1989; Highet, 1989). Therefore, this emphasis on bodyimage may predispose the athlete to chronic caloric restriction, inadequate nutrientintake and potential pathogenic weight control behaviour (Moffatt, 1986; Rucinski,1989; Committee on Sports Medicine, 1989).Calcium is one of the nutrients of concern for adolescent athletes. Some1literature suggests that adequate calcium intake during childhood and adolescencemay contribute to peak bone mass (Matkovic et al., 1979; Matkovic et al., 1990;Matkovic, 1991) thereby reducing the risk of osteoporosis in later life. A recent study(Johnston et al., 1992) demonstrated that increasing calcium intake above theRecommended Dietary Allowance (RDA) significantly increased bone mineral densityof prepubertal subjects even when dietary calcium intake appeared to be adequate.This study supports earlier studies (Matkovic et al., 1979; Matkovic et al., 1990;Matkovic, 1991) that calcium intake during childhood and adolescence contributes topeak bone mass. Therefore, adequate calcium intake during their adolescent years isimportant for the future health of this population.It is well accepted that nutritional practices during adolescence are importantfor growth and development, athletic performance and future health, however, themost effective method of influencing food selection behaviour in this population hasnot been directly addressed. Krondl and Lau (1982) suggest that the main obstaclefor changing food selection behaviour in an affluent society is lack of knowledge ofthe motives behind food selection behaviour. Therefore, Lewis et al. (1989)developed a framework based on cognitive and social factors to examine behavioursand to structure the relationship between measurable factors important to thefrequency of selection of four beverages in college students and middle-aged adults.The researchers found that the social factors were better predictors of beverageconsumption than the traditional variables such as nutrition knowledge. The authorsalso reported that the factors that influenced beverage consumption varied betweenthe two groups (Lewis et al., 1989) suggesting that motives behind food selectionmay vary between populations.2Some studies suggest that participation in sports, particularly aesthetic sports,predisposes the athlete to chronic calorie restriction, nutrient inadequacies andpathogenic weight control behaviours. It has also been reported that some groups ofathletes do not meet the recommended intakes for calcium which is an importantnutrient during adolescence . Understanding themotive for food selection appears to be necessary to facilitate behaviour change. Inthe framework developed by Lewis et al. (1989), social factors were shown to beuseful in predicting beverage selection behaviours in adults. Therefore, the purposeof this study was to assess the calcium intakes of adolescent athletes and to examinefactors that may influence milk and dairy product intake, and therefore calcium intakein adolescent athletes of both genders. To assess whether sport is an influence, twosports were studied, an aesthetic sport (gymnastics) and a non-aesthetic sport (speedskating). Identifying and understanding the motives behind food selection inadolescent athletes may aid the development of effective nutrition educationstrategies to improve the health and performance of these young athletes. To clarifythe terminology used in this study, definitions of terms can be found in Appendix A.3CHAPTER IIII.^LITERATURE REVIEW1. IntroductionIn a study such as this, it is important to review the literature to gain anunderstanding of the nutrition practices of adolescent athletes, methods used toassess these practices, and the model used for predicting food selection behaviour.Therefore, this chapter first addresses the issue of calcium intake in adolescentathletes. This is followed by a review of methods for assessing calcium intaketogether with a description of the rationale for selecting a food frequencyquestionnaire for use in the study.The next section of the review deals with the issue of disturbed eatingbehaviours in athletes. This section also addresses methods of assessing disturbedeating behaviours. It is followed by a description of a framework developed by Lewiset al. (1989) which was used to predict food selection behaviours in adults. Thisframework is reviewed in terms of the researchers' findings and its potential use forpredicting food selection behaviours in adolescent athletes. Finally, a review of anappropriate field technique for assessing body composition of adolescent athletes isprovided.2. Calcium IntakeThere is increasing concern about the composition and nutrient value ofadolescent diets (Guenther, 1986). Calcium is one of the nutrients of concern.Differences of opinion among countries as to the recommended intake for calcium4during adolescence complicate interpretation of the extent of concern about the intakeof this nutrient. For example, in Canada, the recommended nutrient intake (RNI) forcalcium for females 16 years of age is 700 mg/day (Health and Welfare Canada,1990) in contrast to the recommended dietary allowance (RDA) in the United Statesof 1200 mg/day (Food and Nutrition Board, 1989). Therefore, intakes that meet therecommendations for one country may not be sufficient to meet the recommendationsof another country. There is inconclusive evidence to date as to which values aremore appropriate.Research suggests that in general, adolescents may be replacing their milkintake with soft drinks (Guenther, 1986). Guenther demonstrated that soft drinkconsumption was inversely related to milk and calcium intake in American teenagers.This was most apparent with females. Those who consumed soft drinks achieved anaverage of only 59 percent of their RDA for calcium. Females not consuming softdrinks, however, still consumed an average of only 75 percent of the RDA for calcium.Studies on adolescent gymnasts, volleyball players, figure skaters and swimmersdescribe nutrient inadequacies, including inadequate calcium intake, when comparedto the RDA (Moffatt, 1986; Perron and Endres, 1986; Rucinski, 1989; Chen et al.,1989; Benson et al, 1990).Since adequate calcium consumption during childhood and adolescence maycontribute to optimal peak bone mass (Matkovic et al., 1979; Matkovic et al., 1990;Matkovic, 1991), sub-optimal calcium intake is an important problem. Peak bonemass attained in early adulthood is one of the recognized variables contributing to therisk (or reduced risk) of osteoporosis in later life (Matkovic et al., 1979; Sandler et al,1985; Bailey et al., 1986; Kreipe and Forbes, 1990). In 1979 Matkovic et al. studied5two Yugoslavian populations that differed in calcium intake. The authors reported ahigher mean peak bone mass at age 30 in the population that consumed higheramounts of calcium, but the males were significantly heavier and the females weresignificantly taller compared to the group with lower calcium intakes. Matkovic et al.(1990) assessed the effect of calcium supplementation on calcium balance and thenused a two-year intervention trial to investigate the effects of calciumsupplementation on bone mass in adolescent females. Results from the calciumbalance studies showed that the main determinant of calcium balance was calciumintake. An increased calcium intake from 807 mg/day to 1242 mg/day resulted in anet calcium absorption with no change in urinary calcium excretion. Net calciumabsorption was highly correlated with calcium retention (r =0.93, p <0.001). Theresults of the two-year intervention study showed a general increase in the bone massmeasurements for the adolescents on the high calcium intake, however, thesedifferences were not significant (p > 0.05). The researchers felt that their smallsample size may have resulted in a Type II error. In view of the results from thecalcium balance and bone mass studies, the authors suggested that their data supportthe hypothesis that inadequate calcium intake may translate into inadequate calciumretention and decreased peak bone mass.Further evidence in support of the importance of calcium intake in young peoplewas provided by a three year, double blinded, placebo-controlled clinical trial(Johnston et al., 1992). This study was conducted to assess the effects of calciumsupplementation on bone density of adolescents. Forty-five pairs of identical twinsof both genders, initially age 10 ± 2 years completed the study. In each pair of twins,one twin served as the control. Bone density was measured using photon6absorptiometry at three sites, radius, hip and spine, at baseline and three years later.The mean dietary intake of the placebo group was 908 mg calcium per day while thesupplemented group consumed 1612 mg (894 mg from diet and 718 mg fromsupplements). After three years, those subjects who received supplements and wereprepubescent throughout the study had significantly greater bone mineral densities inthe radius and lumbar spine sites compared to their controls. However, in thosesubjects who went through puberty or were post-pubertal, no significant effects ofsupplementation were observed. The researchers concluded that in prepubertalchildren whose dietary intake of calcium approximated the RDA, calciumsupplementation enhanced the rate of increase in bone mineral density. At present,it is unknown whether the gain in bone mineral density will persist, however, thesedata lend support to a higher recommended calcium intake during childhood and earlyadolescence.3. Assessing Calcium IntakeThere is a need for a method of estimating calcium intake that is practical witha large sample size (Cummings et al., 1987; Gibson, 1990). There are two generalmethods for assessing dietary intakes. The first method is quantitative, for examplediet recalls or food records (Gibson, 1990; Angus et al., 1989). These measures,however, are not typically indicative of usual intakes unless diet records are kept foran extended period of time (Gibson, 1990; Angus et al., 1989). The second methodis qualitative and gathers retrospective information on food patterns or usual intakesof foods or nutrients (Gibson, 1990; Angus et al., 1989). This method includes diethistories and food frequency questionnaires (FFQ)(Gibson, 1990). The most7appropriate method for assessing dietary intake depends on the desired accuracy andtype of diet information required as well as the resources available to the researchers(Angus et al., 1989).There is no "gold standard" for assessing dietary intake (Cummings et al.,1987; Angus et al., 1989; Block, 1982). However, both the 7 day weighed foodrecord (Cummings et al., 1987; Angus et al., 1989) and the multiple, non-consecutivediet record have been reported to be the most accurate tools for assessing dietaryintake (Block and Hartman, 1989). Both of these methods provide information on"usual" intakes, however, they are labour intensive and not practical for large dietsurveys (Angus et al., 1989). Alternatively, a 24 hour recall is faster, but does notprovide information regarding usual dietary intake of an individual (Cummings et al.,1987; Block, 1982). The FFQ on the other hand is a useful tool designed to assessusual dietary intakes in large numbers of subjects (Angus et al., 1989). Theadvantages of the FFQ are: 1. it is easy to administer; 2. it requires minimalcoding of the data, therefore, it is less expensive and faster than other methods (ie.24 hour recall); 3. it provides a better approximation of the usual diet compared toa 24 hour recall (Margetts et al., 1989); 4. it is easier to complete (Gibson, 1990;Angus et al., 1989; Margetts et al, 1989).The FFQ can be used to assess the frequency with which certain foods areconsumed during a specified time period (ie. daily, weekly, or monthly)(Gibson, 1990)or it can be designed to act as a predictor of intakes of specific nutrients includingcalcium (Gibson, 1990; Margetts et al., 1989)). The latter method is used to assesscalcium intake. In addition, FFQ can be semi-quantitative. This is an attempt toquantify usual portions of food items in order to provide quantitative data about usual8dietary intake (Gibson, 1990).A number of studies have compared intakes determined by FFQ and weigheddiet records and have shown good agreement between the two methods for calciumintake (Cummings et al., 1987; Angus et al., 1989; Margetts et al, 1989; Willett,1987). Cummings et al. (1987) studied the intakes of free-living women over 65years of age comparing calcium estimated by FFQ to 7 day weighed food records.They reported a correlation coefficient of r = 0.76. Similar results were reported byAngus et al. (1989) in a study of women between the ages of 29 and 72 years. Theresearchers compared calcium estimated by FFQ to a 4 day weighed food record. Thecorrelation coefficient was r =0.79. These results suggest that a FFQ designed toestimate calcium is capable of assessing calcium intake.Other studies do not demonstrate the same ability of a FFQ to estimate calciumintake when compared to a food record (Willett et al., 1987; Stuff et al., 1983;Bergman et al., 1990). However, these studies all used a FFQ to assess the usualintakes for a number of nutrients simultaneously. The questionnaires were large andinvolved, containing over 100 food items. In contrast, for the assessment of aspecific nutrient, the FFQ can be simplified and still include a sufficient number ofitems to ensure reasonable accuracy. A semi-quantitative FFQ is, therefore, apractical and efficient method for assessing calcium intake (Cummings et al., 1987;Angus et al, 1989).Barr and Pare (1992) developed and validated a FFQ to assess calcium intakein adolescents. The FFQ included foods that were high in calcium content and/orconsumed frequently by adolescents. One hundred and thirty-eight studentscompleted the FFQ and provided a 24-hour recall. The validity of the FFQ was9examined in two ways.To assess the ability of the FFQ to quantify calcium intake, each participant's24-hour recall was coded onto a blank FFQ. Then, the 24-hour recall and the FFQfrom the same day's intake were analyzed for calcium intake. The analysis showedthat the mean calcium intake from the 24-hour recall was 878 mg and the meancalcium intake from the FFQ for the same day's intake was 715 mg. Therefore, onaverage, the FFQ was able to quantify 81 percent of the calcium intake from the 24-hour recall. These two values were highly correlated (r =0.98, p <0.0001).Next, the mean calcium intake analyzed from the students' completion of theFFQ was 954 mg. This value was correlated with the mean calcium intake from the24-hour recall. While these two values were significantly correlated, the correlationcoefficient was not extremely high (r =0.59, p <0.0001). These values, however, arereasonably similar since calcium intake is highly variable from day to day. Overall,these results demonstrate that the FFQ is a valid tool for estimating calcium intake inthis age group.4. Dieting BehaviourParticipation in competitive sports typically requires a certain degree of leannessfor optimal performance. The potential concern is that those athletes who restrictfood intake or "diet" to achieve leanness, will also be limiting calcium intake andtherefore possibly jeopardizing their calcium status. Some sports place an evengreater emphasis on leanness for aesthetic reasons. In addition, sports such asgymnastics and figure skating require the athletes to support their body weight whilemoving gracefully through various routines (Moffatt, 1986; Rucinski, 1989; Highet,101989). Thus, the need for leanness is also of practical importance (Moffatt, 1986;Rucinski, 1989). Studies in both male (Faria and Faria, 1989) and female (Falls andHumphrey, 1978) gymnasts have shown that, in top level athletes generally, percentbody fat is inversely related to national ranking.As a group, athletes who participate in sports which place a greater emphasison leanness have been considered to be at greater risk for developing disturbed eatinghabits (Borgen and Corbin, 1987; Rosen and Hough, 1988; Davis and Cowles, 1989).Tendencies toward eating disturbances have been reported in a number of femalecollege athletes who participate in sports with a greater emphasis on leanness suchas gymnastics, body building, ballet, synchronized swimming, diving, figure skatingand long distance running (Borgen and Corbin, 1987; Rosen and Hough, 1988; Davisand Cowles, 1989). Similar tendencies toward eating disturbances have been foundin male college wrestlers (Enns et al., 1987). A study in adolescent figure skaterssuggested that a gender difference may exist regarding tendencies toward eatingdisturbances with males feeling less pressure to strive for thinness compared tofemales (Rucinski, 1989).It has been suggested that poor nutrition and tendencies toward disturbedeating behaviours were primarily a problem for weight control sports (Borgen andCorbin, 1987). However, this was not the case in a study on pathogenic weightcontrol behaviours in female varsity athletes (Rosen et al, 1986). These femalevarsity athletes were participants in sports with varying degrees of emphasis onleanness. At both extremes a high prevalence of pathogenic weight controlbehaviours was revealed, with field hockey having the second highest prevalence nextto gymnastics (Rosen et al., 1986). In a study on female collegiate swimmers,11although tendencies toward eating disturbances were not pathogenic, there was apositive correlation (r = 0.72, p < .01) between tendencies toward eating disturbancesand being heavier for a given height (Barr, 1991). Similarly, in a recent study infemale adolescent athletes (Benson et al., 1990), although the gymnasts had a lowerpercent body fat compared to swimmers and the control group, the swimmersexhibited greater tendencies toward eating disturbances compared to both thegymnasts and the control group. This study supports the premise that participantsin sports which place less emphasis on leanness may not be exempt from eatingbehaviour disturbances, especially during adolescence.Eating disorders are primarily found in late adolescence and early adulthood(Wilmore, 1991; Wardle and Marsland, 1990). However, some data suggest that girlsas young as 11 years of age are already weight conscious (Wardle and Marsland,1990). There is increasing evidence that participation in athletic events furtheremphasizes the need for thinness and leanness, thus predisposing the adolescent toeating disorders (Wilmore, 1991). In addition to the demands to be lean for optimalperformance in a sport, adolescent athletes are subject to the same socio-culturalinfluences that affect the non-athlete (Wilmore, 1991). These influences combinedwith the psychological make up of the elite athlete, who is both goal oriented and aperfectionist, make this type of adolescent particularly vulnerable to eating disorders(Wilmore, 1991).5. Evaluating Eating DisturbancesThe two measures most frequently used to assess eating disorders are theEating Attitudes Test (EAT) and the Eating Disorders Inventory (EDI)(Wilmore, 1991).12The EDI is a 64 item questionnaire containing sub-scales of: drive for thinness,bulimia, body dissatisfaction, ineffectiveness, perfectionism, interpersonal distrust,interceptive awareness and maturity fears (Wilmore, 1991). The EAT originallyconsisted of a 40 item questionnaire that was proposed to measure symptoms ofanorexia nervosa (Garner and Garfinkel, 1979). The most recent version of the EATcontains 26 of the original 40 items (Garner et al.,1982). It contains sub-scales of:1. dieting (related to avoidance of fattening food and a preoccupation with beingthinner); 2. bulimia and food preoccupation (items reflecting thoughts about food aswell as those indicating bulimia); 3. oral control (related to self control of eating andthe perceived pressure from others to gain weight)(Garner et al., 1982) The highera subject scores on the EAT, the more suggestive it is of a disordered attitude towardeating.The EAT has been validated with anorexia nervosa patients in clinical settings.However, it should not be used as a diagnostic criterion for anorexia nervosa in non-clinical groups (Garner et al., 1982). It has been shown that in a non-clinical setting,individuals with high EAT scores do not satisfy the diagnostic criteria for anorexianervosa. However, the majority were found to experience abnormal eating patternswhich interfered with normal psychosocial functioning (Button and Whitehouse,1981). This suggests the EAT would be a suitable screening tool for abnormal eatingpatterns.The 'dieting' sub-scale (13 items) of the EAT questionnaire reflects apathogenic avoidance of fattening foods and shape preoccupation. It is not relatedto bulimia (Garner et al., 1982). Subjects who score high on this sub-scale can bedescribed as over estimators of their body size, dissatisfied with their body shape, and13having a desire to be smaller (Garner et al. 1982). Therefore, if calcium intake isbeing investigated, the dieting sub-scale would appear to be the most appropriatescale to use since restricting intake will logically restrict calcium intake. In addition,the 'dieting' sub-scale has the highest correlation with the total 26 item EAT (r = 0.93)compared to the other sub-scales. It has also been suggested as an economicalsubstitute for assessing dieting behaviours in circumstances where brevity isimportant (Garner et al.,1982). A score of 8 or greater suggests pathologicaltendencies towards shape preoccupation and avoidance of fattening foods (Garner andGarfinkel, 1982).6. Factors Influencing Eating BehaviourNutrition education can be defined as "a process that assists the public inapplying knowledge from the nutrition sciences and the relationship between diet andhealth to their food practices. It is a deliberate effort to improve the nutritional well-being of people by assessing the multiple factors that affect food choices, tailoringeducational methodologies and messages to the public being reached, and evaluatingresults." (ADA Reports, 1990).Nutrition education programs have served to provide knowledge and a changeto more positive attitudes. However, their ability to produce changes in eatingbehaviour is somewhat less established (Johnson and Johnson, 1985). The desiredoutcome of a nutrition education program is to achieve a positive change in foodpractices (Johnson and Johnson, 1985). Since nutritionists desire a behavioralchange for the success of their education programs, it is suggested that they integratethe principles of the behavioral sciences into their education strategies to facilitate this14change (Olson and Gillespie, 1981). For this reason, the nutrition education messageshould be based in the nutritional sciences, however, the foundation forcommunicating this message should be found in the social and behavioral sciences(ADA, 1990).The process of food selection is part of an intricate behaviour system influencedby many factors (Krondl and Lau, 1982). It is recognized as more than a function ofphysiological need in that both social and environmental influences play a strong role(Krondl and Lau, 1982). Inadequate knowledge of the motives behind food selectionis a major obstacle for modifying food selection behaviour (Lewis et al. 1989).Therefore, when studying food selection, there appears to be a need to include arange of cognitive and environmental influences (Lewis et al., 1989).Lewis et al. (1989) developed a framework which allowed the inclusion ofcognitive and environmental influences in an attempt to predict beverage selectionbehaviour in young and middle age adults. In addition, the researchers compared thepredictive power of their cognitive and environmental model to the predictive powerof a traditional model more commonly used in food and health behaviour studies. Inview of the relevance to the present study, the framework developed by Lewis et al.(1989) is reviewed in greater detail below.To facilitate their study of cognitive and environmental influences, theinvestigators (Lewis et al., 1989) developed and validated a number of scales whichmeasured various traditional and social factors which have been identified asinfluencing food selection behaviour and were relevant to a social model. To test theframework, the researchers used the frequency of consumption of milk and soda asthe food selection behaviour. The factors that were included in the model were:15differential association, social and non-social reinforcement, modelling behaviour,general nutrition knowledge, commitment, and attitudes relative to the frequency ofconsuming the beverages.The first factor included in the model was the social environment factor. Thisfactor was termed differential association and included five elements of the socialenvironment that the researchers identified as influencing food selection behaviour:1. Family, 2. Friends, 3. Health Experts, 4. Media and 5. Lifestyle. A scale wasdeveloped to measure the impact of these social environmental elements on foodselection. The scale contained statements reflecting a) family members' use of thebeverage and perception of their feelings about it, b) friends' use of the beverage andtheir perception of their feelings about it, c) perception of health experts'recommendations concerning the beverage, and d) perception of media advertising forthe beverage. The participants were to respond by marking the extent of theiragreement with the statement on a Likert-type scale. Lifestyle was defined as thefrequency with which breakfast, lunch and supper were eaten away from home.Participants responded to questions about how often they consumed meals away fromhome and a "Lifestyle " score was computed.The next factor was social reinforcement. To operationalize socialreinforcement, a scale was developed which contained questions reflecting positivefeelings, a sense of belonging to a group and pleasing others as a function of drinkingthe beverages. Separate scales were constructed for milk and soda. An attitudescale was also constructed. This scale required the participants to respond bymarking their extent of agreement with statements such as "It is important to drinkmilk for good nutrition." and "Drinking soda is an acceptable dietary practice.".16Behaviour modelling was evaluated since the researchers felt that eating is asocial behaviour and it has been documented that many other social behaviours canbe influenced by modelling behaviour (Bandura, 1977). To assess modellingbehaviour, a scale was constructed to measure how often the participants saw or hadseen certain models drinking milk or soda. The models were: mother, father, anotheradult in the family liked/admired by participant, husband/wife or boy/girlfriend, friendadmired by participant and favourite media star.Non-social reinforcement was defined by taste enjoyment of the beverage. Theinvestigators felt that the function of taste was a critical component of food selectionand that there is a clear association between food choice and taste. Therefore, theparticipants were asked to score how they enjoyed the taste of the beverages.Measures for the traditional model were also collected. The researchersdeveloped a true/false scale to assess general nutrition knowledge. Demographic datasuch as age, gender, education and socioeconomic index were also collected.Data were collected through the use of self-administered questionnairesdistributed by mail. Six hundred and ninety-three students (approximately half werewomen) and 422 middle-aged adults (58% women) responded to the questionnaire.The data were analyzed first by comparing a general regression analysis of the socialmodel and for the traditional model (age, sex, education, socioeconomic status,attitude, nutrition knowledge) in both groups. Since food consumption is highlyvariable, the authors decided that the social model must be able to explain 35% of thevariance in beverage consumption to be acceptable. The results indicated that allversions of the social model explained at least 35% of the variance in beverageconsumption in both groups whereas the traditional model did not meet the a priori17criteria for explained variance for any of the beverages in either of the groups. Theauthors also employed path analysis which allowed the development of models whichstructured the relationships between the measurable variables important to foodconsumption.The results of the path analysis indicated that behaviour modelling, socialreinforcement and nutrition knowledge may influence beverage consumption indirectlythrough attitudes and behaviour commitment. However, the factors that influencedbeverage consumption varied between the two groups of adults. Nutrition knowledgewas related to attitude in adult soda-drinking models but not in the student soda-drinking models. The influencing factors for consumption also varied with type ofbeverage since for the students, nutrition knowledge was related to taste enjoymentfor low-fat milk but not for whole milk. This suggests that the factors influencingfood selection may vary between populations and between the foods beingconsumed.The overall finding of the study was that the social model constructed by Lewiset al. (1989) explained more variance in food selection than the traditional model (age,gender, education, socioeconomic index, nutrition knowledge and attitude).Therefore, researchers concluded that the social model was appropriate for predictingfood selection behaviour in adults. To date, this is the only reported use of thisframework for predicting food selection.7. Body CompositionHuman body composition has been a topic of interest for decades due to itsrelevance to health and physical performance. Excess fat is associated with increased18morbidity and mortality (Durnin and Womersley, 1974; Bandini and Dietz, 1987) andthere is a high negative correlation between percent body fat and physicalperformance in activities that require vertical or horizontal movement of body weightthrough space (Wilmore, 1983). Therefore, the assessment of body composition isnecessary to provide better estimates of minimal and optimal weights for physicalperformance and health (Lohman, 1986).There are an array of methods available for assessing body composition rangingfrom simple measurements of body size, such as body weight used in conjunctionwith height and frame size to evaluate "ideal weight", to precise laboratory-basedtechniques that are not practical for field studies (Hergenroeder and Klish, 1990). Theformer type of measurement does not address the issue of dividing bodycompartments into fat or fat free body (FFB). It therefore, provides limited insight intothe actual composition of the body. In addition, it may incorrectly classify leanmuscular individuals (ie. athletes) as obese (Lohman et al., 1984). Other methods ofbody composition analysis are thus required to obtain this information in the field.Skinfold measurement is a simple and non-invasive method for assessing bodycomposition in the field. Percent body fat can be estimated from the data obtainedby measuring skinfold thicknesses at various sites on the body. Although there aremany equations published in the literature for first predicting body density, and then,body fat from skinfolds, few of these equations have been developed using anadolescent population.Determination of body composition in adolescents is a complex problem.Changes in body shape, fat proportion and fat patterning during adolescence mayinvalidate the assumptions underlying the skinfold thickness technique (SF)19(Deurenberg et al., 1990b). Furthermore, it is well accepted that the water andmineral content of fat free body (FFB) changes with maturation (Deurenberg et al.,1990a; Forbes, 1987; Lohman et al., 1984; Lohman, 1986), invalidating theassumptions of constant composition of fat and FFB in the two-compartment model.Therefore, use of adult regression equations (which assume a constant compositionof FFB) to estimate percent body fat (%BF) from SF in adolescents can result insignificant errors (Slaughter et al., 1984; Slaughter et al., 1988). Accordingly, someresearchers have developed equations based on a multi-compartment model thatattempt to adjust for the changes in body composition during maturation (Deurenberget al., 1990b; Lohman, 1986; Weststrate and Deurenberg, 1989). From the availabledata in the literature, Lohman (1986) has developed age and sex specific constantsfor the density of FFB (and its constituents) based on the multi-compartment model.Unfortunately, to date, neither these equations nor constants have been cross-validated on an athletic adolescent population.Thorland et al. (1984) conducted a cross-validation study between body densityby underwater weighing and selected anthropometric regression equations in bothmale (age range 14-19 years) and female (age range 11-19 years) adolescent athletes.In males, equations of either a linear or quadratic form demonstrated acceptableaccuracy in predicting body density. In females, only the quadratic equationsdisplayed the same degree of accuracy. For use in an adolescent athletic populationThorland et al. (1984) suggested the quadratic equations of Jackson and Pollock forfemales (1980) and the linear equation of Forsyth and Sinning for males (1973). Tocalculate %BF from body density, Lohman's formula (1986) and his age and sexspecific constants for body density appear to be the most acceptable method which20addresses the chemical immaturity of youths.8. SummaryThe review of the literature has suggested that adequate calcium intake duringchildhood and adolescence appears to make a positive contribution to bone mineraldensity. There is documentation, however, that some groups of adolescent athleteshave had inadequate calcium intakes according to their country's recommendationsfor the nutrient. The literature also provides support for the use of a food frequencyquestionnaire for estimating calcium intake in this age group. Thus, the assessmentof calcium intake in adolescent athletes using a FFQ warrants investigation.There is inconclusive evidence to date as to whether adolescent athletescompeting in aesthetic sports are more prone to disturbed eating behaviours andnutrient inadequacies than those who participate in non-aesthetic sports. The dietingsub-scale of the Eating Attitudes Test has been identified as a useful indicator ofpathogenic tendencies towards shape preoccupation and avoidance of fattening foods.Since calcium intake is affected by dieting behaviour, the dieting sub-scale of theEating Attitudes Test appears to be a good choice for investigating disturbed eatingbehaviours in this study.Factors influencing food selection behaviour seem to vary between age groups.However, it is not known whether the framework developed by Lewis et al. (1989)is useful in predicting food selection behaviour in adolescents. Understanding themotives behind food selection appears to be important for developing strategies tomodify food selection behaviour. Therefore, there is support for applying theframework developed by Lewis et al. (1989) in attempt to identify factors which may21predict calcium intake in adolescent athletes. If the framework is useful, it willfacilitate the development of nutrition intervention strategies to improve calciumintake in this population.9.^HypothesesTo study the influence of various factors on calcium intake in adolescentathletes, the following null hypotheses were investigated:1. There will be no difference in the mean calcium intake of athletes participatingin aesthetic or non-aesthetic sports.2. There will be no relationship between the mean daily calcium intake and:a) differential association score;b) lifestyle score;c) modelling behaviour score;d) social reinforcement score;e) non-social reinforcement score;f) demography;g) dieting behaviour score;h) percent body fat.22CHAPTER IIIIII.^METHODOLOGY1. SubjectsThrough contacts with two provincial gymnastic and speed skatingorganizations, athletes competing at a minimum of a provincial level in either of thesesports were asked to take part in the study. One-hundred and fourteen adolescentathletes of both genders between 12 and 18 years of age volunteered for this study.Fifty seven males (gymnasts = 25, speed skaters =32) and fifty seven females(gymnasts =32, speed skaters =25) participated in the study. Participants fromVancouver and surrounding area were registered with Sport B.C. while participantsfrom Saskatchewan were registered with Sask Sport. Written informed consent wasobtained from all athletes, and parental consent was also obtained when the athleteswere younger than 18 years of age (see Appendix B for copy of consent form). Thestudy was approved by the University of British Columbia Behavioral SciencesScreening Committee for Research and Other Studies involving Human Subjects.2. Data CollectionThere were two components to the data collection, body composition analysisand a self-administered questionnaire. All data collection was conducted at theathletes' training sites.A)^Body CompositionAll body composition determinations (height, weight, SF) were performed prior23to training to minimize the effects of fluid changes on estimation of body composition.Height was measured using a wall scale to the nearest 0.1 cm. Body weight wasmeasured using an electronic scale to the nearest 0.5 kg. Hip circumference wasmeasured in triplicate to the nearest 0.1 cm using a flexible cloth tape, and a meanvalue calculated.Triplicate skinfold measurements were taken, using Lange Calipers (CambridgeScientific Industries, Cambridge MA) on the right side of the body, by the sameindividual to reduce technical variability. Separate regression equations were used formales and females. These equations were previously cross validated in an adolescentathletic population (Thorland et al., 1984). For females, the quadratic equation ofJackson and Pollock (1980) was used to predict body density as suggested byThorland et al. (1984). This equation required the measurement of the following sites:triceps, supra-iliac, abdomen, thigh, and hip circumference (Harrison et al, 1989). Formales, the linear equation of Forsyth and Sinning (1973) was used to predict bodydensity as suggested by Thorland et al. (1984). This equation required themeasurement of the following sites: subscapular, abdomen, triceps, and midaxillary(Harrison et al, 1988). Both regression equations included age. Percent body fat wascalculated using Lohman's age and sex specific constants for density of FFB and thecorresponding formula to calculate percent body fat (Lohman, 1986). See AppendixC for the equations used to estimate %BF and the age and sex specific constants forFFB. The sum of the four skinfold sites (S4SF) was also calculated.B) QuestionnaireAll athletes completed a questionnaire that contained several parts which24assessed the following factors: pubertal status, demographics (age, sex, racial origin,use of vitamin/mineral supplements, employment, allergies to milk or dairy products),training volume, competitive level, total calcium intake, differential association (ie.family, friends, experts, media), lifestyle, dieting behaviour, modelling behaviour,social and non-social reinforcement and parental socioeconomic status (see appendixB for sample of the questionnaire completed by the participants). Each part of thequestionnaire will be addressed separately.i) Maturation StatusMaturation status was assessed differently for each gender. For females,maturation status was assessed by the self-reported presence or absence ofmenarche. For males, it was assessed by the self-reported presence or absence offacial hair and a change in the pitch of their voice. If the males responded "Yes" toeither question, they were considered to have reached puberty.ii) EmploymentThe athletes responded to the employment question by marking "Yes" or "No"indicating whether or not they were employed. If the athletes answered "Yes", theywere to mark or specify what type of work they did, fast food restaurant, retail sales,paper route or babysitting or 'other' and to specify the type of job. Next theemployed athletes were to report the number of hours they worked per week.iii)^Supplement UsageThe participants were asked to report whether or not they used supplements25by marking a response of either "Yes" or "No". If the athlete used supplements theywere to respond by marking which supplements they used from the followingsupplement list, multivitamin/mineral, vitamin C, Iron, Calcium, Other (and specify).iv) Racial OriginThe participants were asked to mark one or two categories which describestheir racial origin since race may influence consumption of milk and dairy products andtherefore calcium intake. A list of various racial origins was provided, "Caucasian"(white), "Oriental", "East Indian", "Black", "Native Canadian" (Indian or Inuit), "Other"(please specify), and "Don't Know". The athletes' responses were then divided intothree categories: 1) "Caucasian"; 2) "Oriental"; 3) "Other and Mixed". Therefore,if an athletes responded that their racial origin was only Caucasian, they were codedas Caucasian. If an athlete selected Oriental as their only response, they were codedOriental. If the athletes marked other racial origins such as Black, East Indian, NativeCanadian or Other, or if they marked more than one choice for racial origin, they werecoded as "Other or Mixed".v) AllergiesThe participants were to respond by marking either "Yes" or "No" to thequestion "Are you allergic to milk or dairy products?". The intent of this question wasthat the participant would be excluded from further analysis if they marked "Yes" tothe question as it could have a substantial impact on their daily calcium intake.However, if the athlete marked "Yes" and consumed milk, they were not excluded.26vi) Training Volume and Competitive LevelTraining volume and competitive level was also self-reported. The athletesresponded to questions about their training status, number of hours per session theytrained, number of sessions per week and number of months per year of training fortheir sport.Next, the athletes were asked to respond to questions about their competitivelevel. They were asked to respond to whether they competed in City, Provincial,National and International competitions by circling either "Yes" or "No" to eachquestion. A criteria for participating in the study was that the athletes competed ata minimum of a provincial level, however, due to the structure of the athletes'competitive season, some athletes may not have competed in a city or provincialchampionship, however had competed in a national competition and were thereforeincluded in the study.vii) Socioeconomic StatusSince socioeconomic status (SES) can influence nutrition practices, the athleteswere asked to specify the occupation of their mother and father. Socioeconomicstatus was then calculated using the 1981 socioeconomic index for occupations inCanada which reflects income and education level (Blishen et al. 1987). If only oneparent worked, that parent's occupation was used to represent the household's SES.However, if both parents worked, a mean of the two values was used to representthe SES of the household.27viii) Calcium IntakeA previously validated food frequency questionnaire (Barr and Pare, 1 992) wasused to assess daily calcium intake, the dependent variable. The athletes were askedto respond to how often they consumed the various foods, "Per Day", "Per Week","Per Month", or "Don't Eat" by recording the number of times that they wouldconsume the specified quantity of the food during that period. All values wereconverted to a per month basis. If the respondent stated that he/she consumed thefood daily, the number of times the food was eaten per day was multiplied by 30, ifweekly, the number of servings was multiplied by four and if monthly, the value wasnot adjusted. For each food, the adjusted values were recorded and used to estimatecalcium intake. The calcium content of the foods was analyzed using a computerprogram PC Nutricom (Smart Engineering LTD., Vancouver, B.C.) based on theCanadian Nutrient File (Health and Welfare Canada, 1982). The number of servingsof the food the athlete would consume per month was multiplied by the calciumcontent of the food. These values were calculated for all food items and a monthlycalcium intake computed. To calculate daily calcium intake, the monthly calciumintake was divided by thirty. Table 1 shows the instrument used to quantify calciumintake. In addition, since it has been suggested that teenagers may be replacing theirmilk intake with soda (Guenther, 1986), the frequency of consumption of regular anddiet soda was included in the FFQ.The athletes were also asked to indicate which type of milk they usually drank,"whole milk" (coded 1), "two percent milk" (coded 2), "one percent milk" (coded 3),"skim milk" (coded 4), "chocolate milk" (coded 5) or "no milk" (coded 6).28Table 1:^Calcium Food Frequency QuestionnaireFood^ Per Dav i^Per Week 2 Per Month 3 Don't EatBread or Toast, 1 slice or 1 rollwhite brownMuffin, 1 largePizza, 1 medium sliceCheeseburgerCheese - 1 slice processed OR1 piece hard cheese (plain orin sandwich)Broccoli, 1/2 cup (125 ml)Gai-lan (Chinese Broccoli),1/2 cupBok-choi (Chinese Cabbage),1/2 cupIce Cream (large scoop)Frozen Yogurt (large scoop)Fast Food MilkshakeCottage Cheese, 1/2 cupYogurt, small (175 ml) cartonor equivalent29Table 1: Calcium Food Frequency Questionnaire (continued)Food^ Per Day'^Per Week 2 Per Month 3 Don't EatCanned Salmon or Sardineswith bones 1/2 small canSoft Drink, regular, 1 can orlarge glassSoft Drink, diet, 1 can orlarge glassCoffee or Tea, 1 cupTofu, 2 oz (60 gm)Milk on CerealOrange Juice, 1 cupMilk (any type includingchocolate) 1 cupMacaroni & Cheese,1 cup (250 ml)'Values were multiplied by 30 to transform them to a per month basis.2Values were multiplied by four to transform them to a per month basis.'Values were not transformed.30ix) Dieting BehaviourThis scale contained the thirteen items of the dieting sub-scale from the EatingAttitudes Test (Garner et al., 1982) used to measure dissatisfaction with body shapeand a desire to be thinner. The athletes responded to the questions indicating howfrequently they had certain feelings about each statement by marking the responsewhich matched how often they experienced those feelings, "Always" (scored 3) "VeryOften" (scored 2), "Often" (scored 1) "Sometimes" (scored 0),"Rarely" (scored 0),"Never" (scored 0) (Garner and Garfinkel, 1979). A dieting behaviour score wascalculated by adding the scores from all 13 questions after reversing the scoring forquestion number 13. A score of eight or greater suggests pathological tendencies ofshape preoccupation and a desire to be thinner (Garner and Garfinkel, 1982). Thedieting behaviour instrument is shown in Table 2. The reliability of the instrumentwas assessed using the Cronbach's alpha reliability coefficient calculated using allthirteen responses after reversing the scoring for question 13.x) Differential AssociationA previously validated questionnaire was used to assess differential association(family, friends, health experts, media)(Lewis et al., 1989). Table 3 shows thequestions the athletes answered about their: 1) family members' use of milk andtheir perception of their feelings about it, 2) friends' use of milk and their perceptionof their feelings about it, 3) perception of health experts' recommendationsconcerning milk, and 4) perception of entertainment/persuasive quality of televisionadvertising of milk (Lewis et al., 1986). The athletes responded by marking their31Question^ Always Very^Often Some- Rarely Never^Often times1. Am terrified aboutbeing overweight2. Aware of the caloriecontent of foods that I eat3. Particularly avoid foodswith high carbohydratecontent (eg. bread, rice,potatoes, etc)4. Feel extremely guiltyafter eating5. Am preoccupied (thinka lot about) with a desireto be thinner6. Think about burning upcalories when I exercise7. Am preoccupied with(think a lot about) with thethought of having fat onmy body8. Avoid foods with sugarin them9. Eat diet foods or drinks10. Feel uncomfortableafter eating sweets11. Go on diets to loseweight12. Like my stomach to beempty13. Enjoy trying rich newfoods'Table 2:^Dieting Behaviour Questions from the Dieting Sub-Scale of the EatingAttitudes Test'Reversed scoring prior to analysis32Table 3:^Items Used to Determine Differential AssociationCategory ItemFRIENDS' 1) I have a lot of friends who drink milk2) A meal with my friends usually doesn't include milk 23) My friends seem to feel that it's important to drinkmilk4) My friends think milk is drunk only by young childrenand not by teens or adults 2EXPERTS' 1) My doctor recommends that I drink milk of some kind2) I often hear health experts recommend drinking milk3) I have heard nutritionists recommend that people ofmy age drink milk4) My doctor has shown no concern about whether Idrink milk2MEDIA' 1) Advertisements for milk catch my attention2) The advertisements I see for milk make me want todrink it3) I hardly ever pay attention to advertisements for milk 24) I think advertisements for milk and dairy products areentertainingFAMILY' 1) Most teens and adults in my family drink milk as partof a snack or with meals2) Very few adults in my family use milk on a regularbasis 23) It is unusual for adults in my family to drink milk 24) My family feels that drinking milk is an important partof the diet for teens'Athletes responded to the statements by marking their extent of agreement with thestatements on a five point scale, Strongly Disagree, Disagree, Unsure, Agree, StronglyAgree.2Scores on these statements were reversed prior to analysis due to negative wording.33response according to their feelings with the statements. They answered using a fivepoint scale, "Strongly Disagree" (scoring 1), "Disagree" (scoring 2), "Unsure" (scoring3), "Agree" (scoring 4), and "Strongly Agree" (scoring 5). After reversing the scoresfor negatively worded items, the internal consistency of the instrument was assessedusing the Cronbach's alpha reliability coefficient by first calculating coefficients foreach of the sub-scales, family, friends, experts and media and then calculating thecoefficient for the entire scale. A mean score was calculated for each question, eachsub-scale and an overall differential association score was computed by averaging theresponses to all sixteen questions.xi) LifestyleThe questionnaire also contained a section to assess lifestyle (as shown inTable 4). In this section, the athletes answered questions about the frequency ofconsumption of meals and snacks away from home. The athletes were asked torespond to the number of times per week that they would consume breakfast,morning snack, lunch, afternoon snack, dinner and an evening snack. Then, theywere to record the number of times that they would eat those meals and snacks "out"per week. Eating "out" was defined as not consuming food at home or brought fromhome. The number of meals and snacks eaten away from home per week wascalculated, and the percentage of meals and snacks eaten away from home wastermed the lifestyle variable.xii) Social ReinforcementTable 5 shows the four questions which comprised the social reinforcement34Meals and Snacks^Times Eaten^Times eatenper week^"out " per weekBreakfastMorning SnackLunchAfternoon SnackDinnerEvening SnackTable 4:^Items Used to Determine Lifestyle35Table 5:^Items Used to Determine Social Reinforcement Score1) Drinking milk makes me feel part of a special group of peopleStrongly Agree^Agree^Unsure^Disagree^Strongly Disagree2) When I drink milk I feel I get approval from the people who matter to meStrongly Agree^Agree^Unsure^Disagree^Strongly Disagree3) It gives me a nice feeling to have a glass of milk with friendsStrongly Agree^Agree^Unsure^Disagree^Strongly Disagree4) When I drink milk, I please people who are important to meStrongly Agree^Agree^Unsure^Disagree^Strongly Disagree36instrument. The athletes were asked to respond to questions about their perceptionsconcerning positive feelings, a sense of belonging to a group, and pleasing others asa function of consuming milk (Lewis 1986). The respondents were asked to indicatethe extent of their agreement with the statement by marking their answer on a fivepoint scale, "Strongly Disagree" (scoring 1), "Disagree" (scoring 2), "Unsure" (scoring3), "Agree" (scoring 4), "Strongly Agree" (scoring 5). An overall score was calculatedto represent the social reinforcement score by averaging the responses to all fourquestions. A Cronbach's alpha reliability coefficient was calculated using all four ofthe questions. In accordance with the questionnaire developed by Lewis (1986), thesocial reinforcement questions were interspersed with the differential associationquestions in the questionnaire compiled for the athletes.xiii) Modelling BehaviourTable 6 illustrates the questions which have been used as a measure ofmodelling behaviour. Modelling behaviour was assessed by determining the frequencyin which the athletes saw or had seen certain models consume milk. The participantsindicated how often they had observed the model (mother, father, another adult intheir family admired or liked by the athlete, a friend of the opposite sex and a friendthey admire) drinking milk by circling, "Never or Almost Never" (scored 1), "Not VeryOften" (scored 2), "Fairly Often" (scored 3), "Very Often" (scored 4). The originalmodelling behaviour scale developed by Lewis (1986) included 'a favourite movie star'as one of the models. However, when the scale was pre-tested using an adolescentpopulation, the youths were either unable to answer the question because they eithercommented that they never saw their favourite movie star so how would they know37Table 6:^Items Used to Determine Modelling Behaviour ScoreYour Mother...VERY^FAIRLY^NOT VERYOFTEN^OFTEN OFTENYour Father...VERY^FAIRLY^NOT VERYOFTEN^OFTEN OFTENNEVER ORALMOST NEVERNEVER ORALMOST NEVERAnother adult in your family whom you admire (for example, aunt orgrandparent)...VERY^FAIRLY^NOT VERY^NEVER OROFTEN^OFTEN OFTEN ALMOST NEVERA friend of the opposite sex...VERY^FAIRLY^NOT VERY^NEVER OROFTEN^OFTEN OFTEN ALMOST NEVERA friend you admire...VERY^FAIRLY^NOT VERY^NEVER OROFTEN^OFTEN OFTEN ALMOST NEVER38if they drank milk or not or they left the question blank. Therefore, it was decided todrop the question from the modelling behaviour scale for the adolescents.The internal consistency of the revised scale was assessed using Cronbach'salpha reliability coefficient using all five questions. A modelling behaviour score wascomputed by averaging the scores from all five questions.xiv) Non-Social Reinforcement (Taste)The instrument developed to measure non-social reinforcement is shown inTable 7. Using a six point scale, participants were asked to indicate how much theyenjoyed the taste of certain dairy products. The athletes responded by circling,"Don't Know" (scored 1), "Not At All" (scored 2), "Not Very Much" (scored 3), "JustSo-So" (scored 4), "It's O.K." (scored 5), "Very Much" (scored 6). Cronbach's alphareliability coefficient was calculated using all eight questions. A non-socialreinforcement (taste) score was calculated by averaging the scores from all eightquestions.3. Data AnalysisThe Statistical Package for the Social Sciences (SPSSX Inc., Chicago, IL) wasused for all data analysis. In this study the dependent variable was daily calciumintake. The independent variables were: 1) sport; 2) percent body fat; 3) age;4) sex; 5) weight; 6) dieting behaviour; 7) social reinforcement; 8) non-socialreinforcement (taste); 9) behaviour modelling; 10) lifestyle; 11) differentialassociation. A significance level of alpha = 0.05 was established for all statisticalanalyses except where otherwise specified.39Table 7:^Items Used to Determine Taste Enjoyment1. How much do you enjoy the taste of whole milk?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW2. How much do you enjoy the taste of low-fat (2% or 1 %) milk?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW3. How much do you enjoy the taste of skim milk?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW4. How much do you enjoy the taste of fruit-flavoured yogurt?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW5. How much do you enjoy the taste of plain yogurt?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW6. How much do you enjoy the taste of cottage cheese?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW7. How much do you enjoy the taste of hard cheese (such as cheddar)?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW8. How much do you enjoy the taste of ice cream?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K.^"SO-SO"^MUCH^ALL^KNOW40To analyze the descriptive data, the athletes were divided into 2 groupsaccording to gender and the two sports were compared. Descriptive statistics,including means, ranges and frequencies were used to describe the athletes within thesports. T-tests were used to evaluate the differences between the means of the twosports for the continuous descriptive data. Chi-Square was used to test differencesbetween the two sports when the data was categorical, for example presence ofmenarche.Analysis of variance was used to compare gender and sport differences for thefollowing variables: daily dietary calcium intake (and contributions to calcium intakefrom fluid milk, other dairy foods and non-dairy food sources), dieting behaviour score,differential association variable, modelling behaviour variable, lifestyle variable, non-social reinforcement (taste) variable and the social reinforcement variable. Thehomogeneity of the dependent variable was examined, and resulted in a significantBartlett-Box F, therefore, total dietary calcium intake was transformed by using thenatural log for all statistical tests. The Mann-Whitney U test was used to compare theconsumption of regular and diet soda between the sports.Then, the athletes who reported allergies to milk or dairy products and thereforedid not consume milk were excluded from further analysis. Next, the relationshipbetween daily calcium intake and the following independent variables: dietingbehaviour score, modelling behaviour score, differential association score, lifestylescore, non-social reinforcement score, social reinforcement score, age, weight,percent body fat and consumption of soda (regular and diet) was examined for allathletes using a Pearson Correlation Coefficient. Then, the athletes were divided bygender and sport and the correlations re-examined.41Stepwise forward entry multiple regression analysis was used to find theindependent variable(s) that best predicted daily calcium intake as the dependentvariable. Two different multiple regression tests were conducted for all athletescombined, one to test the traditional variables including age, sport, gender, weightand %BF, and the second to test the independent variables from a modified versionof the social model used by Lewis et al. (1989) (modelling behaviour variable,differential association variable, lifestyle variable, non-social reinforcement variable,and social reinforcement variable) and the dieting sub-scale score from the EatingAttitudes Test (Garner et al., 1982). Then the athletes were divided by gender andsport and the multiple regression analysis computed again. Equations to predictcalcium intake were generated where possible.42CHAPTER IVIV.^RESULTSFor all analysis, missing data have been treated as pairwise deletions unlessotherwise specified.1.^Internal Consistency of the QuestionnaireThe Cronbach's alpha reliability coefficient (1951) was used to test theinternal consistency of each scale within the questionnaire. A scale was consideredto be internally consistent if the alpha coefficient was 0.65 or greater (Lewis, 1986).Only data from internally consistent scales were used for statistical analysis. Table8 shows the Cronbach's alpha reliability coefficients for all scales.The internal consistency of each scale was assessed by using all questionswithin each scale. However, the reliability of the differential association scale wasfirst assessed by assessing the reliability of the following sub-scales: family, friends,media and health experts. This analysis showed the sub-scales of family and mediato be internally consistent, whereas the sub-scales of friends and health experts werenot internally consistent. Next, a Cronbach's alpha reliability coefficient wascalculated using all sixteen questions from the four sub-scales (family, friends, healthexperts, media). This calculation demonstrated internal consistency when the sub-scales were analyzed collectively with a coefficient of 0.75. Therefore, it was decidedthat an overall differential association score, which averaged the scores of all sixteenquestions, would be used for the differential association variable and included in thestatistical analysis. The other scales, dieting sub-scale, modelling behaviour, non-43Table 8:^Cronbach's Alpha Reliability Coefficient for Questionnaire ScalesScaleCronbach's AlphaReliability CoefficientDieting Sub-scale (EAT) 0.86Differential Association 0.75Friends 0.53Family 0.72Media 0.78Experts 0.54Modelling 0.66Non-Social Reinforcement 0.66Social Reinforcement 0.7044social reinforcement and social reinforcement were deemed reliable since their alphacoefficients were greater than 0.65 and therefore, their scores were included in thestatistical analysis.2. Descriptive Characteristics of AthletesTo examine the descriptive characteristics of the participants, the athletes weredivided into two groups by gender and then further divided by sport for comparativepurposes.A) MalesAs shown in Table 9, the two groups of athletes had similar mean ages, but thespeed skaters were significantly (p < 0.05) taller and heavier and had significantly(p <0.001) higher %BF and S4SF measurements than the gymnasts. For maturation,there was no significant difference in the number of speed skaters who had reachedpuberty (78%) compared to the gymnasts (60%), as assessed using Chi-Square.When the two sports were compared for employment status, there was nosignificant difference between the number of speed skaters who were employedcompared to the gymnasts (41 % and 24% respectively). For the athletes who wereemployed, there was no significant difference between the sports for the number ofhours worked per week (2.6±4.7 vs 2.0 ± 3.9, skaters and gymnasts respectively)or the types of positions held (for example, paper route or fast food restaurant).Table 10 shows that the percentage of athletes who took vitamin/mineralsupplements was similar for each sport. For the two sports combined, 21% tookmultivitamin/mineral supplements, 18% took vitamin C, four percent took iron, nine45Table 9:^Descriptive Physical Characteristics of Male Athletes (Means andstandard deviations)Variable Speed Skaters Gymnasts All Athletesn =32 n =25 n =57AGE (yrs) 14.4±1.8 14.7 ±1.9 14.5 ±1.8HT (cm) 170.2±10.3* 158.5±13.0 165.1 ± 12.9WT (kg) 60.5±11.9* 50.1 ±12.5 55.9±13.1%BF 8.0±3.4** 3.7±2.1 6.1 ±3.6S4SF 35.8±9.8** 23.9±3.6 30.6±9.7*p<0.05 Speed Skaters vs Gymnasts using a two-tailed t-test.**p<0.001 Speed Skaters vs Gymnasts using a two-tailed t-test.46Table 10:^Use of Vitamin Supplementation in Male Athletes (percentages)Variable Speed Skaters Gymnasts All Athletesn =32 n=25 n =57Use Supplements 21.9% 36.0% 28.1%Multivitamin 18.3% 24.0% 21.1%Vitamin C 9.4% 28.0% 17.5%Iron 0% 8% 3.5%Calcium 3.1% 16.0% 8.8%Other 0% 4% 1.8%No significant difference for supplement use between speed skaters and gymnastsusing Chi-Square test.47percent took calcium and two percent took a B-Complex supplement.Racial origin can influence calcium intake, therefore, the racial origins of theathletes in the two sports were compared. Ninety-four percent of the skaters and84% of the gymnasts were Caucasian. Other racial origins included two (6%)gymnasts who were Oriental and three (12%) gymnasts who were in the "Other orMixed" category. For the skaters, none were Oriental and two (6%) were in the"Other or Mixed" category. None of the non-Caucasian athletes reported beingallergic to milk or dairy products.A major source of calcium for an adolescent population is milk and dairyproducts, therefore, an allergy to these foods may substantially affect the athlete'sdietary calcium intake. For this reason, the athletes were asked to respond towhether or not they were allergic to milk or dairy products. One skater and onegymnast, both Caucasians, reported allergies to milk or dairy products, however,according to the FFQ data, only the skater avoided milk and dairy products. Thisathlete was therefore excluded from the Pearson Correlation and Multiple RegressionAnalysis.As shown in table 11, training volume and competitive status were assessedto provide a comparison between the quality of the male athletes who participated inthe two sports. For training volume, the analysis showed that the gymnasts trainedsignificantly (p <0.001) more hours per session than the speed skaters, while sessionsper week and months of training per year did not differ between groups. However,both groups of athletes were considered highly conditioned since on average, theytrained in excess of four sessions per week for more than nine months of the year.Since there was no difference in the competitive level of the male athletes, the48Table 11:^Training Volume and Competitive Level of Male Athletes (means,standard deviations and percentages)Variable Speed Skatersn=32Gymnastsn=25All Athletesn=57Hours/Session 1.4±0.4* * 3.4±0.6* * 2.3 ± 1.1Sessions/Week 4.4±2.2 5.9 ±1.6 5.1 ± 2.1Months/Year 9.3±2.3 11.4±1.4 10.2±2.2CityChampionships'84.4%$ 72.0% 79.9%ProvincialChampionships'90.6%$ 96.0% 93.0%NationalChampionships'53.1%$ 60.0% 56.1%InternationalChampionships'37.5%$ 16.0% 28.1%**p<0.001 Speed Skaters vs Gymnasts using two-tailed t-test.$ = No significant difference between Speed Skaters vs Gymnasts for competitive levelusing a Chi-Square test.'Values represent percent of athletes who competed in competitions at that level.49difference in training was likely due to the type of training required for each sport asopposed to a difference between quality of the athletes participating in the two sports(for example, speed skaters train more aerobically than gymnasts).B) FemalesAs shown in Table 12, the two groups of athletes had similar ages, however,the speed skaters were significantly (p <0.001) taller, heavier and had higher %BF andS4SF measurements than the gymnasts. For maturation, the Chi-Square analysisshowed that significantly (p < 0.05) more speed skaters had reached menarchecompared to the gymnasts (80% vs 53% respectively), suggesting that the speedskaters were more physiologically advanced.When the two sports were compared for employment status, there was nosignificant difference between the number of skaters who were employed comparedto the gymnasts (24% vs 31 % respectively). For the athletes who were employed,both groups of athletes spent a similar number of hours at work each week (1.2 ± 3.9hours vs 1.4 ± 2.7 hours for skaters and gymnasts respectively).As shown in Table 13, the use of vitamin and mineral supplements was similarfor both groups of female athletes. Sixty percent of the speed skaters and 38% ofthe gymnasts used at least one type of vitamin or mineral supplement. There was nosignificant difference between the sports for the type of supplement used exceptsignificantly (p <0.05) more speed skaters took Vitamin C than gymnasts.As indicated previously, racial origin and allergies to milk and dairy products caninfluence dietary calcium intake, therefore, these factors were investigated. When theathletes in the two sports were compared for racial origin, 96% of the skaters and50Table 12:^Descriptive Physical Characteristics of Female Athletes (means andstandard deviations)Variable Speed Skaters Gymnasts All Athletesn =25 n = 32 n =57AGE (yrs) 14.3±1.6 14.1±1.6 14.2±1.6HT (cm) 162.2±8.2** 153.6±4.9 157.4±7.8WT (kg) 58.5±9.5** 45.2 ± 7.9 51.0±10.8%BF 21.0±7.7** 9.4 ± 5.7 14.5 ± 8.8S4SF (mm) 65.5±18.9** 38.5±12.3 50.3±20.5**p<0.001 Speed Skaters vs Gymnasts using a two-tailed t-test.51Table 13:^Use of Vitamin Supplementation in Female Athletes (percentages)Variable Speed Skaters Gymnasts All athletesn =25 n =32 n =57Use Supplements 60% 37.5% 47.4%Multivitamin 7.0% 25.0% 26.3%Vitamin C 48.0%* 18.8% 31.6%Iron 16.0% 12.5% 14.0%Calcium 16.0% 6.3% 10.5%Other 4.0% 3.1% 3.5%*p<0.05 Speed Skaters vs Gymnasts using Chi-Square test. All other comparisonswere not significant.5284% of the gymnasts were Caucasian. Other racial origins included one skater (6%)who was in the "Other or Mixed" category. For gymnasts, two (6%) were Orientaland three (9%) were coded in the "Other or Mixed" category. Of the athletes whowere not Caucasian, none reported avoiding milk or dairy products.The athletes in the two sports were compared for reported allergies to milk ordairy products, and two skaters and no gymnasts reported allergies. However,following examination of the FFQ data, only one of the skaters avoided milk or dairyproducts. Therefore, only the athlete who avoided milk or dairy products wasexcluded from further analyses.As shown in Table 14, training volume and competitive level of the two groupsof athletes were compared. For training volume, the gymnasts had significantly(p <0.001) longer training sessions and trained for significantly (p <0.05) moremonths of the year compared to the skaters However, athletes participating in thetwo sports trained a comparable number of times per week. Similar to the males, itis not possible to suggest that the gymnasts were more highly trained from this datasince the type of training involved for each sport cannot be compared. There was nosignificant difference in the competitive level of athletes in the two sports exceptsignificantly (p <0.05) more skaters competed at an international level compared tothe gymnasts. Regardless, both groups of athletes were considered highlyconditioned since on average, they trained in excess of four sessions per week formore than ten months of the year.The final descriptive characteristic that was to be addressed for both genderswas socioeconomic status. To facilitate the calculation, the athletes were asked tospecify their parents' occupations. Unfortunately, many of the participants were53Table 14:^Training Volume and Competitive Level of Female Athletes (means,standard deviations and percentages)Variable Speed Skatersn =32Gymnastsn =25All Athletesn =57Hours/Session 1.5±0.6** 4.3 ± 0.6 3.0 ±1.5Sessions/Week 4.2 ± 1.3 4.6±0.9 4.4± 1.1Months/Year 10.0±1.5* 11.3±0.8 10.7±1.3CityChampionships'88.0%$ 96.9% 93.0%ProvincialChampionships'100.0%$ 100.0% 100.0%NationalChampionships'68.0%$ 53.1% 59.6%InternationalChampionships'72.0%* 37.5% 52.6%**p <0.001 Speed Skaters vs Gymnasts using a two tailed t-test.*p <0.05 Speed Skaters vs Gymnasts using Chi-Square for level of competition anda two-tailed t-test for training volume.$ = No significant difference between Speed Skaters vs Gymnasts for competitive levelusing a Chi-Square test.'Values represent percent of athletes who competed in competitions at that level.54unable to provide sufficient information about their parents' occupations to determinetheir socioeconomic status. Therefore, socioeconomic status could not be includedin the data analysis because over one third of the participants had missing data.3. Food Frequency QuestionnaireTables 15, 16, 17 and 18 summarize the average number of servings per monthof each food from the FFQ for male and female athletes. "Visual Inspection" ofTables 15 to 18 suggests that for male athletes, skaters and gymnasts generally atea similar amount of most foods, while for female athletes, skaters appeared to eatmore of most foods than gymnasts. However, no statistical analyses were conductedfor individual foods with two exceptions, diet soda and regular soda. Thesebeverages were included in the FFQ since it has been reported that calcium intake isinversely related to soft drink consumption in American teenagers (Guenther, 1986).In this section, these beverages were investigated in terms of sport differences inconsumption. The analyses showed that, for male athletes, there were no significantdifferences between the two sports for the consumption of diet soda, regular soda orboth types of soda combined using the Mann-Whitney U test. In contrast, for thefemale athletes, the skaters consumed significantly (p <0.001) more regular soda andsignificantly (p <0.05) less diet soda than the gymnasts. However, when both typesof soda were combined, the consumption of soda was similar for both sports.Table 19 shows the comparisons between the sports for daily calcium intake,the contribution of dairy products (fluid milk and other dairy foods) and contributionof non-dairy foods to daily calcium intake for male and female athletes. Analysis ofvariance was used to examine the difference between the groups for total dietary55Table 15:^Number of Servings Per Month of Non-Dairy Product Foods For MaleAthletes (means and standard deviations)'Food Item and ServingSizeSpeed Skatersn=32Gymnastsn=25All Athletesn=57White Bread (1 Slice) 51.3±64.8 46.7±58.6 49.3±61.7Brown Bread (1 slice) 45.4±64.7 51.0±51.1 47.8±58.8Muffin (1 large) 6.3±7.7 6.7±7.5 6.5 ± 7.3Pizza (1 medium slice) 7.9±9.0 8.2 ± 13.2 8.0 ± 10.9Cheeseburger (1 slice) 4.4±4.4 3.4±4.9 4.0 ± 4.6Broccoli (125 ml) 3.3±3.9 5.6±5.7 4.3 ± 4.8Chinese Broccoli (125ml)0.4±1.0 0.3 ± 0.9 0.4 ± 1.0Chinese Cabbage (125ml)0.7±2.9 0.7±2.0 0.7±2.5Canned Salmon (1/2small can)2.3±4.7 2.0±4.7 2.2±4.7Regular Pop (1 can) 13.0 ± 15.0 12.4±11.2 12.7±13.3Diet Pop (1 can) 4.3±11.7 2.9±8.6 3.7±10.4Tea/Coffee (1 cup) 5.1±12.4 7.6±12.7 6.2±12.5Tofu (60 gm) 0.1 ±0.4 1.8±6.4 0.8 ± 4.2Orange Juice (250 ml) 30.0±34.4 35.6±39.3 32.4±36.4Macaroni & Cheese(250 ml)4.8±5.6 4.0±4.0 4.5±5.0'Data were not analyzed statistically except for regular and diet pop. There was nosignificant difference between skaters and gymnasts for the consumption of regularor diet soda using a Mann-Whitney U test.56Table 16:^Number of Serving Per Month of Dairy Products for Male Athletes(means and standard deviations► 1Food and Serving SizeSpeed Skatersn=32Gymnastsn=25All Athletesn=57Milk (250 ml) 89.6±70.2 70.1± 67.0 81.1 ± 68.9Milk on Cereal (125 ml) 32.3±34.2 30.3±23.5 31.4±29.8Cheese (20 g) 35.0±41.4 36.4±64.0 35.6±52.0Ice Cream (125 ml) 18.6±44.0 17.8±29.9 18.2±38.3Frozen Yogurt (125 ml) 2.0±3.0 1.2±1.9 1.6±2.6Fast Food Milk Shake 2.5±2.4 1.3±2.6 2.0±2.5Cottage Cheese (125 ml) 0.9±1.7 3.0 ± 6.5 1.8 ± 4.5Yogurt (175 ml) 3.3±4.2 7.4 ± 19.0 5.0 ± 12.7'Data were not analyzed statistically.57Table 17:^Number of Servings Per Month of Non-Dairy Product Foods For FemaleAthletes'Food Item and ServingSizeSpeed Skatersn=25Gymnastsn=32All Athletesn=57White Bread (1 Slice) 49.3±46.3 31.2±32.7 39.3±40.0Brown Bread (1 slice) 23.7±36.3 19.4±29.5 21.3±32.4Muffin (1 large) 6.6±7.5 4.2±3.6 5.3±5.7Pizza (1 medium slice) 4.9±4.4 3.4±3.2 4.1 ±3.8Cheeseburger (1 slice) 3.8±5.3 1.1±1.3 2.3±3.8Broccoli (125 ml) 8.2 ± 14.4 4.7±5.7 6.3 ± 10.5Chinese Broccoli (125ml)0 0.2±0.7 0.1 ±0.6Chinese Cabbage (125ml)0.3±0.8 0.3±0.9 0.3±0.8Canned Salmon (1/2small can)0.7±.4 0.6 ± 1.3 0.7±1.3Regular Pop (1 can) 15.3±29.3** 3.4±4.3 8.6 ± 20.3Diet Pop (1 can) 1.5±1.8* 5.3±6.9 3.7±5.6Tea/Coffee (1 cup) 6.8 ± 11.9 4.2±6.9 5.4±9.4Tofu (60 gm) 0 0.2±0.8 0.1 ±0.6Orange Juice (250 ml) 30.5±28.6 26.8±34.0 28.4±31.5Macaroni & Cheese(250 ml)11.3±24.0 7.4± 14.5 9.1 ± 19.2'Data were not analyzed statistically except for comparison between sports for dietand regular soda.*p<0.05 Skaters vs Gymnasts using a Mann-Whitney U test.**p<0.001 Skaters vs Gymnasts using a Mann-Whitney U test.58Table 18:^Number of Serving Per Month of Dairy Products for Female Athletes(means and standard deviations)'Food and Serving SizeSpeed Skatersn=25Gymnastsn=32All Athletesn=57Milk (250 ml) 78.7±56.5 42.9±3.5 58.6±48.0Milk on Cereal (125 ml) 24.6±27.5 25.4±32.4 25.0±30.1Cheese (20 g) 37.3±43.7 20.5±16.5 27.9±32.3Ice Cream (125 ml) 6.8±8.9 4.8±6.9 5.7±7.8Frozen Yogurt (125 ml) 3.8±8.1 2.4±3.5 3.0±6.0Fast Food Milk Shake 1.3±1.4 0.7±1.0 1.0±1.2Cottage Cheese (125 ml) 4.4 ± 12.3 1.2±3.6 2.6±8.6Yogurt (175 ml) 12.7±19.2 5.7±7.8 8.8±4.3'Data were not analyzed statistically.59Table 19:^Calcium Intake for Male and Female Athletes (means and standarddeviations)Variable Speed Skaters Gymnasts All AthletesFluid Milk(mg/day) *Males 1107±845 871± 695 1006±789Females 953±643 582 ± 435 745 ±562Other DairyCalcium (mg\day)Males 276±241 304±523 288 ± 381Females 362±314 188±147 264± 249Non-Dairy Calcium(mg/day)Males 289±150 275±134 289 ± 143Females 308 ±280 209 ± 148 253 ±220Total CalciumIntake (mg/day)*Males 1681 ±931 1456±984 1590±951Females 1527 ± 750 1005 ± 534 1235 ± 683* 2X2 ANOVA (Gender X Sport) revealed a significant effect of sport, (p<0.05).60calcium intake. The results showed that averaged over the two sports, there was nosignificant main effect of gender on total daily calcium intake (F 1 , 100 =2.90, p =0.09).There was, however, a significant main effect for sport, meaning that averaged overgender, the skaters consumed significantly more dietary calcium per day than thegymnasts (F1,100 = 6.63, P =0.011). There was no significant interaction effect fortotal calcium intake suggesting that the differences in calcium intake between the twosports were similar for each gender (F 1 , 100 =0.52, p =0.47).Since sport differences were found for intake of total dietary calcium, it wasdecided to investigate where the differences existed. For this reason, analysis ofvariance was used to examine the differences between the groups for calcium intakeprovided from fluid milk, other dairy products and non-dairy products.The analysis showed that when calcium intake from fluid milk was averagedover the two sports, there was no significant main effect of gender (F 1 , 109 =3.63,p =0.06). There was, however, a significant main effect of sport on calcium intakefrom fluid milk (F 1 , 109 = 6.10, p =0.02). There was no significant interaction effect ofgender and sport on calcium intake (F 1 , 109 =0.62, p = 0.43), indicating that thedifferences in calcium intake from fluid milk between the two sports were similar forboth genders.When calcium intakes from other dairy sources were analyzed for differences,the results showed that there were no significant main effects of either gender(F1,108 = 0.18, p = 0.67) or sport (F 1108 = 2.23, p =0.14). Similarly, there was nosignificant interaction effect of gender and sport on calcium intake from other dairysources (F 1 , 108 =0.10, p =0.75). These results suggest that all groups of athletes hadsimilar calcium intakes from other dairy food sources.61Comparable results were seen when calcium intakes from non-dairy sourceswere analyzed. There were no significant main effects of either gender (F 1100 = 2.62,p = 0.11) or sport (F1100 = 2.66, p = 0.11) on calcium intake and there was nosignificant interaction effect of gender and sport on calcium intake (F 1100 = 2.23,p =0.14). These results indicate that all groups of athletes had similar calcium intakesfrom non-dairy food sources.To investigate whether employment status influenced total dietary calciumintake, a Mann-Whitney U test was used to compare intakes of athletes who did anddid not work. The analysis showed that both groups of athletes had similar totaldietary calcium intakes. Similarly, to investigate the influence of supplementation ontotal dietary calcium intake, a Mann-Whitney U test was used to compare intake ofathletes who took vitamin/mineral supplements to those who did not. The resultsshowed that the supplement users had similar calcium intakes to the non-supplementusers.The type of milk usually drunk by the athletes was investigated using Chi-Square. Tables 20 and 21 show that no significant difference existed between thesports for the type of milk usually drunk by the athletes (males and femalesrespectively). For male athletes, approximately 88% of the skaters and 84% of thegymnasts consumed low-fat milk (2%, 1 %, or skim milk). For female athletes,approximately 92% of the skaters and 91 % of the gymnasts reported consuming low-fat milk (2%, 1 %, or skim milk) as the milk they usually drank.Overall, the results from the analyses of variance suggest that sport differencesexisted for total calcium intake and that these differences were a result of the skatersconsuming greater amounts of calcium from fluid milk than the gymnasts. In addition,62Table 20:^Type of Milk Consumed by Male Athletes (percentages)Type of Milk Speed Skaters Gymnasts All Athletesn =32 n =25 n =57Whole Milk 6.3% 4.0% 5.3%2% Milk 53.1% 44.0% 49.1%1% Milk 28.1% 32.0% 29.8%Skim Milk 6.3% 8.0% 7.0%Chocolate Milk 3.1% 4.0% 3.5%No Milk 3.1% 8.0% 5.3%No significant difference existed in the type of milk drunk by the athletes in the twosports as assessed using Chi-Square.63Table 21:^Type of Milk Consumed by Female Athletes (percentages)Type of Milk Speed Skaters Gymnasts All Athletesn =25 n =32 n =57Whole Milk 0% 3.1% 1.8%2% Milk 52.0% 34.4% 42.1%1% Milk 20.0% 43.8% 33.3%Skim Milk 20.0% 12.5% 15.8%Chocolate Milk 4.0% 3.1% 3.5%No Milk 4.0% 0% 1.8%No significant difference existed in the type of milk drunk by the athletes in the twosports as assessed using Chi-Square.64the differences that existed between the two sports for calcium intake from fluid milk,and therefore total calcium intake, were consistent for both genders.4. Dieting Sub-ScaleThe dieting sub-scale from the Eating Attitudes Test was used to measurepathological tendencies towards dissatisfaction with body shape and avoidance offattening foods. A score of eight or greater suggests disturbed dieting behaviours.Tables 22 and 23 show the dieting sub-scale scores for individual questions andoverall scores for male and female athletes respectively. The overall scores wereassessed for differences between gender and sport using an ANOVA.The analysis showed that averaged over the two sports, there was a significantmain effect of gender on the dieting sub-scale score (F 1 , 105 = 21.86, p <0.001),meaning that the female athletes had a significantly higher score than the maleathletes. However, there was no significant main effect for sport (F 1 , 105 =0.23,p =0.64), indicating that when the scores were averaged over gender, the two sportshad similar scores. No significant effect was seen for the interaction of gender andsport on the dieting sub-scale score (F1105 =1.09, p =0.30), meaning that thedifferences in scores between gender were similar for both sports. Although genderdifferences existed for the dieting sub-scale score, the mean scores were notsuggestive of tendencies toward disturbed eating behaviour.Next, the dieting sub-scale scores were examined to determine the percentageof athletes who scored eight or higher on the scale. The Chi-Square analysis showedthat for both male and female athletes there was a similar percentage of athletes whoscored eight or higher in both sports. For females, 16% of the skaters and 25% of65Table 22:^Dieting Sub-Scale Scores for Male Athletes (means and standarddeviations)Question SpeedSkatersn=32Gymnastsn = 25AllAthletesn=571. Am terrified about beingoverweight0.0±0.0 0.1 ±0.3 0.0±0.22. Aware of the calorie content offoods that I eat0.1 ±0.4 0.1 ±0.3 0.1 ±0.43. Particularly avoid foods with highcarbohydrate content (eg. bread, rice,potatoes, etc)0.0 ±0.0 0.0 ± 0.0 0.0 ±0.04. Feel extremely guilty after eating 0.0±0.2 0.0 ± 0.0 0.0±0.05. Am preoccupied (think a lot about)with a desire to be thinner0.3±0.7 0.0 ±0.0 0.0 ± 0.16. Think about burning up calorieswhen I exercise0.0±0.2 0.2±0.6 0.2±0.77. Am preoccupied with (think a lotabout) with the thought of having faton my body0.0 ±0.2 0.0 ± 0.0 0.0 ±0.18. Avoid foods with sugar in them 0.0±0.2 0.1 ±0.4 0.1 ±0.39. Eat diet foods or drinks 0.1 ±0.2 0.1 ±0.4 0.1 ±0.310. Feel uncomfortable after eatingsweets0.2±0.4 0.1 ± 0.4 0.1 ±0.411. Go on diets to lose weight 0.0±0.0 0.0±0.0 0.0±0.012. Like my stomach to be empty 0.0±0.2 0.0±0.0 0.0±0.113. Enjoy trying rich new foods' 1.1 ± 1.0 0.9 ± 0.9 0.1 ± 0.9Dieting Sub-Scale Total Score 2 1.9±2.0 1.5±1.4 1.7±1.8'Scores were reversed.20nly variable analyzed statistically. 2X2 ANOVA (Gender X Sport) showed asignificant main effect of gender, (p <0.05). (See Table 23 for data on females).66Table 23:^Dieting Sub-Scale Scores for Female Athletes (means and standarddeviations)Question SpeedSkatersn =32Gymnastsn=25AllAthletesn =571. Am terrified about beingoverweight0.7±1.1 1.0±1.2 0.8±1.22. Aware of the calorie content offoods that I eat0.4±0.7 0.3 ±0.6 0.3 ±0.73. Particularly avoid foods withhigh carbohydrate content (eg.bread, rice, potatoes, etc)0.0±0.2 0.0±0.0 0.0±0.14. Feel extremely guilty after eating 0.1 ±0.4 0.2±0.7 0.2±0.65. Am preoccupied (think a lotabout) with a desire to be thinner0.5±1.0 0.4±0.8 0.5±0.96. Think about burning up calorieswhen I exercise0.7±1.1 0.4±0.7 0.5 ±0.97. Am preoccupied with (think a lotabout) with the thought of havingfat on my body0.6±1.2 0.6±0.8 0.6±1.08. Avoid foods with sugar in them 0.1 ±0.3 0.2±0.6 0.2 ±0.59. Eat diet foods or drinks 0.3 ± 0.6 0.7 ± 1.0 0.5 ± 0.910. Feel uncomfortable after eatingsweets0.3 ±0.7 0.5 ±1.0 0.4 ±0.911. Go on diets to lose weight 0.2±0.8 0.1 ±0.4 0.2±0.612. Like my stomach to be empty 0.0±0.0 0.1 ±0.4 0.0±0.313. Enjoy trying rich new foods' 1.1±1.1 1.3±0.8 1.2±0.9Dieting Sub-Scale Total Score 2 4.3±5.0 5.4±4.5 4.9±4.7'Scores were reversed.2Only variable analyzed statistically.^2X2 ANOVA (Gender X Sport) showed asignificant main effect of gender, (p<0.05). (See Table 22 for data on males).67the gymnasts had dieting sub-scale scores of eight or greater, while for the maleathletes, 8% of the skaters and none of the gymnasts had scores in the rangesuggestive of disturbed eating behaviours. These results indicate that in this study,there were no significant differences between the sports for the percentage ofathletes who had scores suggestive of tendencies towards disturbed eatingbehaviours.5. Differential Association VariableAs discussed under the section, Internal Consistency of the Questionnaire inthis chapter, the overall differential association score was used to represent thedifferential association variable in the statistical analysis. This variable was createdby averaging the scores from all sixteen questions in the differential association scale.Therefore, it represents the athletes' interaction and identity with different groups ofthe social environment (family, friends, health experts and media) in terms of theirperceptions, consumption, persuasive qualities and recommendations about the useof milk. Tables 24 and 25 show the differential association sub-scale scores andoverall differential association scores for the male and female athletes respectively.Mean scores for the individual questions from the sub-scales for all groups of athletescan be found in Appendix D.Analysis of variance showed that when differential association scores wereaveraged over the two sports, there was no significant main effect of gender on thedifferential association variable (F1,103=0.76, p =0.38). Neither was there asignificant main effect for sport (F 1 , 103 = 0.56, p 0.46), meaning that when bothgenders were combined, the scores were similar between the two sports. No68Table 24:^Average Scores for Differential Association Sub-Scales and OverallDifferential Association Score for Male AthletesScale Speed Skatersn =32Gymnastsn =25All Athletesn =57Friends 3.3 ± 0.6 3.3 ± 0.5 3.3 ± 0.6Family 3.8 ± 0.7 3.6 ± 0.8 3.7±0.7Media 2.9 ± 1.0 3.1 ±0.7 3.0 ± 0.9Experts 3.4±0.6 3.4 ±0.7 3.4±0.6Total Score' 3.4 ± 0.4 3.3 ± 0.5 3.4±0.4'Only variable analyzed statistically. 2X2 ANOVA (Gender X Sport) showed nosignificant main effects for gender or sport and no significant interaction effect. (SeeTable 25 for data on females).69Table 25:^Average Scores for Differential Association Sub-Scales and OverallDifferential Association Score for Female Athletes (means and standarddeviations)Scale Speed Skatersn =25Gymnastsn =32All Athletesn =57Friends 3.5 ±0.7 3.5 ±0.5 3.5 ± 0.6Family 3.8 ± 1.0 3.9 ± 0.7 3.9 ± 0.8Media 3.3 ±0.8 3.0 ± 0.7 3.1 ±0.8Experts 3.4 ± 0.7 3.3 ±0.6 3.3 ± 0.7Total Score' 3.5 ±0.6 3.4 ± 0.4 3.4 ± 0.5'Only variable analyzed statistically. 2X2 ANOVA (Gender X Sport) showed nosignificant main effects for gender or sport and no significant interaction effect. (SeeTable 24 for data on males).70significant interaction effect was seen for the differential association score(F 1 , 103 =0.05, p = 0.82). These results suggest that no significant differences existedbetween the groups for the differential association variable, therefore, the socialenvironment appeared to exert similar influences on all groups of athletes. Thus, theathletes agreed to some extent that they perceived that their social environment wassupportive of the consumption of milk. Since milk comprises approximately 65% oftotal calcium intake, the social environment exerted a similar impact on total calciumintake for these athletes.6. Lifestyle VariableThe frequency of eating away from home has been identified as a factor of thesocial environment that may influence calcium intake (Lewis et al., 1989). For thepurpose of this study, "Lifestyle" has been defined as the percent of meals and snackseaten away from home. Tables 26 and 27 show the total numbers of meals andsnacks eaten per week and the percent of meals and snacks eaten away from homeper week (lifestyle variable) for male and female athletes respectively. The analysisof variance showed that when the total number of meals and snacks eaten per weekwas averaged over the two sports, there was a significant main effect of gender(F1,109 =26.14, p <0.001), meaning that the male athletes consumed significantlymore meals and snacks per week than the female athletes. However, when bothgenders were averaged, there was no significant main effect of sport(F1,109= 0.03,P = 0.86), meaning that the total number of meals and snacks eaten perweek was similar between the two sports. There was no significant interaction effect(F1,109= 3 . 81 , p = 0.53), indicating that the differences in the number of meals and71Table 26: Number of Meals and Snacks Eaten and "Eaten Out" Per Week by MaleAthletes (means and standard deviations)Meals and Snacks Speed Skatersn =32Eaten("Out")Gymnastsn =25Eaten("Out")All Athletesn =57Eaten("Out")Breakfast 6.4 ± 1.2 6.3 ± 1.2 6.4 ± 1.2(0.1 ±0.3) (0.0±0.0) (0.0 ±0.2)Morning Snack 2.1 ±2.4 2.9 ±2.7 2.4 ± 2.6(0.3±0.8) (0.1 ±0.5) (0.2 ±0.7)Lunch 6.5 ± 0.9 6.6 ± 1.0 6.6±0.9(1.0 ± 1.4) (0.9 ± 1.1) (1.0 ± 1.3)Afternoon Snack 4.1 ± 2.5 5.0 ±2.0 4.5±2.3(0.4±0.8) (0.3±0.5) (0.3 ±0.7)Dinner 6.9 ± 0.4 7.0±0.0 7.0 ± 0.3(0.8 ±0.8) (0.9 ± 0.8) (0.8 ± 0.8)Evening Snack 5.0±2.1 5.0 ± 2.3 5.0 ± 2.2(0.3±0.8) (0.2±0.7) (0.2±0.7)Total Number of 31.0 ± 5.5 32.8±5.5 31.8±5.5Meals & Snacks' (2.8 ± 2.4) (2.4 ± 2.4) (2.6 ± 2.4)Lifestyle Variable 8.9 ± 7.5 7.2±7.2 8.2 ± 7.3(Percent of allMeals & Snackseaten out)''Only variables analyzed statistically. 2X2 ANOVA (Gender X Sport) showed asignificant (p < 0.05) main effect of gender on total number of meals and snacks butno significant main effect of sport and no significant interaction effect. 2X2 ANOVA(Gender X Sport) for the lifestyle variable showed no significant main effects ofgender or sport and no significant interaction effect. (See Table 27 for data onfemales).72Table 27: Number of Meals and Snacks Eaten and "Eaten Out" Per Week byFemale Athletes (means and standard deviations)Meals and SnacksSpeed Skatersn = 25Eaten^("Out")Gymnastsn =32Eaten^("Out)All Athletesn = 57Eaten^("Out")Breakfast 5.5 ± 2.3 5.5 ± 1.9 5.5 ± 2.0(0.2 ± 0.4) (0.1 ±0.2) (0.1 ±0.3)Morning Snack 1.6 ± 2.1 2.0 ± 2.1 1.8 ± 2.1(0.2 ± 1.0) (0.1 ±0.3) (0.1 ±0.7)Lunch 6.0 ± 1.8 6.2 ± 1.3 6.1 ±1.5(0.8 ±0.9) (0.7 ± 1.0) (0.7 ± 1.0)Afternoon Snack 3.7 ± 2.7 2.3 ± 1.9 2.9 ± 2.4(0.1 ±0.5) (0.1 ±0.2) (0.1 ±0.3)Dinner 6.8 ±0.7 6.3 ± 1.6 6.5 ± 1.3(1.0± 1.0) (0.9±0.9) (0.9± 1.0)Evening Snack 4.2 ± 2.0 3.4 ± 2.2 3.8 ± 2.1(0.1 ±0.3) (0.0 ± 0.2) (0.1 ±0.2)Total Number of 27.8 ± 5.9 25.7 ± 4.4 27.7 ±5.2Meals & Snacks' (2.4 ± 2.4) (1.8 ± 2.0) (2.1 ±2.2)Lifestyle Variable 8.5 ± 8.2% 6.9±8.1% 7.6 ± 8.1%(Percent of allMeals & Snackseaten out)"'Only variables analyzed statistically. 2X2 ANOVA (Gender X Sport) showed asignificant (p <0.05) main effect of gender on the total number of meals and snacksbut no significant main effect of sport and no significant interaction effect. 2X2ANOVA (Gender X Sport) for the lifestyle variable showed no significant main effectsof gender or sport and no significant interaction effect. (See Table 26 for data onmales).73snacks eaten per week between the genders was similar for the skaters andgymnasts.The results from the analysis of the lifestyle variable showed that there was nosignificant main effect of gender on the lifestyle variable (F1102 = 0.06, p =0.80),indicating that when averaged over the two sports, both genders consumed a similarpercent of meals and snacks away from home. Similarly, there was no significantmain effect of sport (F 1 , 102 =1.20, p = 0.28), meaning that when averaged over bothgenders, there was no significant difference between the sports for the percent ofmeals and snacks eaten away from home. No significant interaction effect was seen(F1,102 =0.00, p =0.97), indicating that the differences between the genders for thepercent of meals and snacks eaten away from home was similar for both sports.Overall, while the males consumed more meals and snacks than the females, allgroups were similar for the percent of meals and snacks that they consumed awayfrom home (lifestyle variable).7. Social Reinforcement VariableThe social reinforcement variable reflected the athletes' perceptions concerningpositive feelings, a sense of belonging to a group and pleasing others as a function ofconsuming milk (Lewis, 1986). The participants' responses to the questions werescored on a five point scale. Therefore, scores greater than 3.0 reflected agreementthat the consumption of milk evoked positive feelings and a sense of belonging to agroup. A score of less than 3.0 reflected disagreement with the above statement.Tables 28 and 29 show the average responses to all questions and the overall scoresfor male and female athletes respectively.74Table 28:^Scores for Social Reinforcement Questions and Overall SocialReinforcement Score for Males (means and standard deviations)Question SpeedSkatersn =32Gymnastsn =25All Athletesn =571. Drinking milk makes mefeel part of a special group ofpeople2.0± 1.1 1.6 ± 0.9 1.8± 1.02. When I drink milk I feel Iget approval from the peoplewho matter to me2.6 ±0.9 2.3 ±1.0 2.5 ±0.93. It gives me a nice feelingto have a glass of milk withfriends2.7 ±1.0 2.4 ±0.9 2.6 ± 1.04. When I drink milk, I pleasepeople who are important tome2.6±0.9 2.2±0.9 2.4 ± 0.9Social Reinforcement Score' 2.5 ± 0.7 2.1 ±0.6 2.3 ± 0.7'Only variable analyzed statistically (p<0.05). 2X2 ANOVA (Gender X Sport)revealed a significant main effect of sport, no significant main effect of gender andno significant interaction effect. (See Table 29 for data on females).75Table 29:^Scores for Social Reinforcement Questions and Overall SocialReinforcement Score for Females (means and standard deviations)Question SpeedSkatersn =25Gymnastsn =32All Athletesn = 571. Drinking milk makes mefeel part of a special group ofpeople2.5 ± 1.0 2.0 ± 0.8 2.2 ± 0.92. When I drink milk I feel Iget approval from the peoplewho matter to me2.6 ± 1.2 2.3 ± 0.9 2.4 ± 1.03. It gives me a nice feelingto have a glass of milk withfriends2.7 ± 1.0 2.3 ±0.9 2.5 ±0.94. When I drink milk, I pleasepeople who are important tome2.6 ± 1.2 2.8 ± 0.8 2.7 ± 1.0Social Reinforcement Score' 2.6 ± 0.9 2.3 ± 0.6 2.4 ± 0.7'Only variable analyzed statistically (p <0.05). 2X2 ANOVA (Gender X Sport) revealeda significant main effect of sport, no significant main effect of gender and nosignificant interaction effect. (See Table 28 for data on males).76The results from the ANOVA showed there was no significant main effect ofgender when the social reinforcement score was averaged over the two sports(F1107 = 1 . 67, p = 0.20), meaning that when the two sports were combined, bothgenders disagreed to a similar extent that the consumption of milk evoked positivefeelings, a sense of belonging to a group and pleasing others. There was, however,a significant main effect of sport (F 1107 = 5.78, p = 0.02), indicating that while athletesin both sports disagreed with the statements, the gymnasts exhibited strongerdisagreement to the statements that the consumption of milk evoked positive feelingsand a sense of belonging to a group than the skaters. There was no significantinteraction effect (F 1 , 107 = 0.36, p = 0.55), meaning that the differences which existedbetween the sports for the social reinforcement variable were similar for both genders.8. Modelling Behaviour VariableThe modelling behaviour variable was created by averaging the scores from allfive questions in the modelling behaviour scale. This scale reflected the frequencywith which the athletes had seen certain models consuming milk. Statistical analysiswas only conducted for the overall modelling behaviour score. A score of 2.5 orgreater suggests that the athletes saw the models, as a group, drinking milk at least"Fairly Often". A score of 2.0 or less suggests that the athletes observed the modelsconsuming milk "Not Very Often" or "Almost Never". Tables 30 and 31 show theresults from the modelling behaviour scale for male and female athletes respectively.The analysis of variance showed that there was no significant main effect ofgender on the modelling behaviour variable (F193 = 3.46, p = 0.07), indicating that forboth sports combined, both the male and female athletes observed the models77Table 30:^Modelling Scores for Male Athletes (means and standard deviations)Question Speed Skatersn=32Gymnastsn =25All Athletesn=571. Mother 2.6±0.8 2.1 ±0.9 2.4±0.92. Father 2.7±0.9 2.5 ±0.9 2.6±0.93. Adult in Familyyou admire2.6±0.9 2.3 ±0.9 2.5 ± 0.94. Friend ofopposite sex2.7±0.9 2.5±1.0 2.6 ± 1.05. Friend youadmire2.6±0.9 2.6±0.9 2.6±0.9ModellingBehaviour Score'2.6±0.6 2.3±0.8 2.5 ±0.7'Only the overall behaviour modelling score was analyzed statistically. 2X2 ANOVA(Gender X Sport) showed no significant main effects of gender or sport and nosignificant interaction effect. (See Table 31 for data on females).78Table 31:^Modelling Scores for Female Athletes (means and standard deviations)Question Speed Skatersn =25Gymnastsn=32All Athletesn =571. Mother 2.3 ±1.1 2.7 ± 1.0 2.5 ± 1.02. Father 2.9±0.9 2.8 ± 1.2 2.7 ± 1.03. Adult in Familyyou admire2.7 ±0.9 2.7 ± 0.7 2.6±0.94. Friend ofopposite sex2.8 ± 0.8 2.3 ± 0.9 2.6 ± 0.95. Friend youadmire3.1 ±0.7 2.9 ± 0.8 2.8 ± 0.9ModellingBehaviour Score'2.7 ±0.6 2.7±0.5 2.7 ±0.6'Only the overall behaviour modelling score was analyzed statistically. 2X2 ANOVA(Gender X Sport) showed no significant main effects of gender or sport and nosignificant interaction effect. (See Table 30 for data on males).79consuming milk "Fairly Often". There was no significant main effect of sport onmodelling behaviour (F193 =2.17, p =0.14), meaning that averaged over gender,athletes in the two sports indicated that they observed the models consuming milkmore often than not. There was no significant interaction effect (F 1 , 93 = 0.66,p = 0.42), meaning that the differences between genders for the modelling behaviourvariable were similar for both sports. Overall, no significant differences were detectedbetween the groups for the modelling behaviour variable.9. Non-Social Reinforcement (Taste) VariableThe non-social reinforcement variable represented how much the athletesenjoyed the taste of dairy products. The scores of the eight foods were averaged andrepresented the non-social reinforcement variable. A score of three or greater on thefive point scale suggested that the athletes enjoyed the taste of the foods. Tables 32and 33 show the taste enjoyment scores for the individual dairy foods and the overallnon-social reinforcement variable for the male and female athletes respectively.The analysis showed that when the non-social reinforcement variable wasaveraged over the two sports, there was no significant main effect of gender(F1,107 = 2 . 73 , P = 0.10), meaning that, on average, both male and female athletesreported enjoying the taste of dairy products to a similar extent. When averaged overgender, there was no significant main effect of sport (F1107 = 0.58, p = 0.45). Nor wasthere a significant interaction effect of gender and sport on the non-socialreinforcement variable (F1107 = 3.41, p = 0.07). These results indicate that on average,all groups of athletes tended to enjoy the taste of milk and dairy products to a similarextent.80Table 32:^Scores for Non-Social Reinforcement (Taste) Questions and Overall Non-Social Reinforcement Score for Male Athletes (means and standarddeviation)Question Speed Skatersn =32Gymnastsn = 25All athletesn =571. How much do you enjoythe taste of whole milk?1.3 ±1.7 2.2 ± 2.0 1.7 ± 1.92. How much do you enjoythe taste of low-fat (2% or1 %) milk?4.2 ± 1.3 4.3 ± 1.2 4.2 ± 1.33. How much do you enjoythe taste of skim milk?2.0 ± 1.6 2.6 ± 1.4 2.3 ± 1.54. How much do you enjoythe taste of fruit-flavouredyogurt?3.9 ± 1.4 4.1 ± 1.3 4.0 ± 1.45. How much do you enjoythe taste of plain yogurt?2.0 ± 1.3 2.1 ± 1.3 2.0 ± 1.36. How much do you like thetaste of cottage cheese?2.2 ± 1.5 2.4 ± 1.7 2.3 ± 1.67. How much do you like thetaste of hard cheese (such ascheddar)?4.0 ± 0.9 4.4 ± 0.7 4.2 ± 0.88. How much do you like thetaste of ice cream?4.8 ±0.5 4.8 ± 0.5 4.8 ± 0.5Overall Score' 3.1 ±0.7 3.4 ± 0.7 3.2 ± 0.7"Only variable analyzed statistically. A score of 1 revealed that the athletes liked thetaste "Not At All", while a score of 5 indicated that the taste was liked "Very Much".No significant effects were detected in a 2X2 ANOVA (Gender X Sport). (See Table33 for data on females).81Table 33:^Scores for Non-Social Reinforcement (Taste) Questions and Overall Non-Social Reinforcement Score for Female Athletes (means and standarddeviation)Question Speed Skatersn =25Gymnastsn =32All athletesn =571. How much do you enjoythe taste of whole milk?1.9 ± 1.4 1.5 ± 1.3 1.7 ± 1.42. How much do you enjoythe taste of low-fat (2% or1%) milk?3.9 ± 1.3 4.1 ± 1.1 4.0 ± 1.23. How much do you enjoythe taste of skim milk?2.9 ± 1.7 2.7 ± 1.5 2.8 ± 1.54. How much do you enjoythe taste of fruit-flavouredyogurt?4.4 ± 1.2 4.3 ± 0.9 4.3 ± 1.05. How much do you enjoythe taste of plain yogurt?2.4 ± 1.4 2.3 ± 1.3 2.3 ± 1.46. How much do you like thetaste of cottage cheese?3.4 ± 1.7 2.9 ± 1.4 3.1 ± 1.67. How much do you like thetaste of hard cheese (such ascheddar)?4.5 ±0.9 4.4± 1.0 4.4± 1.08. How much do you like thetaste of ice cream?4.4 ± 1.2 4.6±0.6 4.5 ± 0.9Overall Score' 3.5 ± 0.6 3.3 ± 0.5 3.4 ± 0.5"Only variable analyzed statistically. A score of 1 revealed that the athletes liked thetaste "Not At All", while a score of 5 indicated that the taste was liked "VeryMuch".No significant effects were detected in a 2X2 ANOVA (Gender X Sport). (SeeTable 32 for data on males).8210. Univariate Correlation AnalysesPearson Correlation Coefficients were calculated to examine the univariate linearrelationship between daily dietary calcium intake and the independent variables. Thefollowing independent variables were included in the analyses: age, soda intake,dieting sub-scale score, differential association variable, social reinforcement variable,modelling behaviour variable, lifestyle variable, weight, percent body fat and the non-social reinforcement variable. Correlation coefficients were calculated for all athletescombined and then the athletes were divided into two groups by gender and thenfurther divided by sport. Athletes who reported allergies to milk or dairy products andtherefore avoided the consumption of milk, were excluded from these analyses.A) All AthletesTable 34 shows the correlation coefficients calculated using data collected fromall athletes. These results show that there was a weak but significant positive linearrelationship (p < 0.05) between estimated dietary calcium intake and the followingindependent variables: differential association variable, social reinforcement, non-social reinforcement, modelling behaviour and weight, explaining only 7%, 7%, 4%,4% and 5% of the variance in total dietary calcium intake respectively.B) MalesTable 35 shows the univariate correlation coefficients for the independentvariables with daily dietary calcium intake for the male athletes. When both sportswere combined, the only two variables which had a significant (p <0.05) linearcorrelation with calcium intake were differential association and social reinforcement.83Table 34:^Univariate Correlation Coefficients for Daily Calcium Intake and theIndependent Variables for All Athletes'Variable Correlation Coefficient(n =112)Alpha LevelAge .02 .411Weight .22 .014*Soda Intake 2 .06 .274Dieting Sub-scaleVariable-.07 .238Differential AssociationVariable.26 .006*Social ReinforcementVariable.26 .005*Modelling BehaviourVariable.21 .022*Lifestyle Variable .17 .055Percent Body Fat .10 .156Non-SocialReinforcement Variable.19 .030*'Analysis does not include athletes who reported allergies to milk or dairy productsand who avoided milk.2Soda Intake = Diet Soda + Regular Soda.*p <0.0584Table 35:^Univariate Correlation Coefficients for Daily Calcium Intake and theIndependent Variables for Male Athletes'VariableSpeed Skaters(n =31)"r"^alphaGymnasts(n =25)"r"^alphaAll Athletes(n =56)"r"^alphaAge 0.20 .145 -0.05 .418 0.06 .336Weight 0.04 .419 -0.14 .263 0.04 .393Soda Intake 2 -0.02 .465 -0.01 .483 0.01 .460Dieting Sub-scaleScore-0.21 .130 -0.19 .193 -0.18 .105DifferentialAssociationScore0.31 .049* 0.35 .059 0.33 .009*SocialReinforcementScore0.10 .300 - 0.32 .080 0.25 .038*ModellingBehaviour Score-0.08 .337 0.41 .051 0.20 .090Lifestyle^Score -0.13 .244 0.36 .067 0.09 .263Percent Body Fat -0.02 .457 0.01 .477 0.12 .199Non-SocialReinforcementScore0.25 .091 0.28 .101 0.20 .078'Analysis does not include athletes who reported allergies to milk or dairy productsand who did not consume milk.2Soda Intake =Diet Soda + Regular Soda.*p<0.0585This analysis suggests that once the effect of gender was controlled, the associationsof the non-social reinforcement variable, the modelling behaviour variable and weightwith total dietary calcium were no longer significant. However, the linear correlationsbetween total calcium intake and the independent variables, differential associationand social reinforcement, were weak since only 11% and 6% of the variance in totaldietary calcium intake was explained by each variable respectively.Next, the athletes were divided by sport and the linear relationships re-examined. These results show that for skaters, only one weak but significant(p <0.05) linear relationship between daily dietary calcium intake and the independentvariable, differential association, remained. This indicated that once gender and sportwere controlled for, the differential association variable explained only 10% of thevariance in total dietary calcium intake. However, for gymnasts, there were nosignificant linear relationships between estimated total dietary calcium intake and theindependent variables, suggesting that once gender and sport are controlled for, noneof the independent variables were linearly related to total dietary calcium intake.C) FemalesTable 36 shows the univariate correlation coefficients for the independentvariables with daily dietary calcium intake for female athletes. For both sportscombined, these results showed that there were significant (p < 0.05) positive linearcorrelations between daily calcium intake and the following variables: socialreinforcement variable, non-social reinforcement variable, modelling behaviourvariable,body weight and percent body fat. These variables were able to explain 10%, 8%,12%, 14%, and 12% of the variance in total calcium intake respectively.86Table 36:^Univariate Correlation Coefficients for Daily Calcium Intake and theIndependent Variables for Female Athletes'Variable Speed Skaters(n =24)"r"^alphaGymnasts(n =32)"r"^alphaAll Athletes(n =56)"r"^alphaAge -0.25 0.139 0.03 0.439 -0.08 0.284Weight 0.26 0.130 0.12 0.266 0.38 0.004*Soda Intake 2 0.01 0.483 -0.07 0.371 0.07 0.312Dieting Sub-scaleScore0.33 0.082 0.02 0.456 0.08 0.314DifferentialAssociation Score0.50 0.010* -0.10 0.320 0.22 0.071SocialReinforcementScore0.46 0.017* 0.10 0.306 0.31 0.017*Modelling BehaviourScore0.30 0.104 0.38 0.029* 0.34 0.012*Lifestyle^Score 0.11 0.323 0.30 0.070 0.23 0.063Percent Body Fat 0.21 0.176 0.03 0.432 0.35 0.007*Non-SocialReinforcementScore0.11 0.324 0.47 0.006* 0.28 0.026*'Analysis does not include athletes who reported allergies to milk or dairy productsand who avoided milk.2Soda Intake =Diet Soda + Regular Soda.*p<0.0587When the sports were analyzed separately, different relationships wereobserved. For skaters, there was a significant (p <0.05) positive relationship betweendietary calcium and two of the independent variables. The differential associationvariable and the social reinforcement variable were able to explain 25% and 21 % ofthe variance in total calcium intake respectively. For gymnasts, significant (p <0.05)positive correlations were found between the dependent variable and two differentindependent variables. Modelling behaviour and non-social reinforcement explained14% and 22% of the variance in total calcium intake for the female gymnasts. Theseresults suggest that for the female athletes in this study, sport differences wereevident for the univariate linear relationships between estimated dietary calcium intakeand the independent variables.Overall, these results suggest that the groups of athletes studied were verydifferent in terms of the types of independent variables which predicted total dietarycalcium intake as the dependent variable. However, the groups of athletes were alsosimilar in the respect that only a small number of the independent variables influencedcalcium intake.11. Multiple Regression AnalysisStepwise forward entry multiple regression Analysis (MRA) was used to studythe relationship between calcium intake and the independent variables. The purposeof conducting MRA was to determine which independent variable(s) were the mostuseful for predicting calcium intake. The intent of this study was to establish howtraditional independent variables compared to social independent variables forpredicting the dependent variable, calcium intake. Therefore, the first MRA included88all athletes and two analyses were conducted, one analysis for the Traditional Modeland one analysis for the Social Model. Then, the athletes were divided by gender andby sport and additional MRA conducted. When variables from the models significantly(p <0.05) explained variance in total calcium intake, the results were summarized onTable 37 and equations were generated. Residuals from the MRA were plotted andexhibited normality and linearity when the dependent variable was transformed bynatural logs. (As indicated in the Methods, all statistical analyses related to calciumintake were performed using log transformed calcium values).A)^All Athletesi) Traditional ModelThe results of the MRA for all athletes are shown in Table 37. This analysisshowed that when predicting calcium intake using the traditional variables (age, sex,sport, weight, %BF), only sport explained a significant amount of variance in calciumintake. Based on the adjusted R squared value, only 10% of the variance in theathletes' total calcium intake was explained by the type of sport. These resultsindicated that much of the variance in total dietary calcium intake was unexplainedby this model. The regression equation generated for this model was (F 1100 = 11.78,p<0.001):In(Predicted Calcium Intake) = 7.681 - 0.381(Sport)Where the numeric values for sport were:^1 =skaters 2 = gymnasts.ii) Social ModelThe MRA for the Social Model showed that when predicting calcium intake89Table 37:^Multiple Regression Analysis Summary TableATHLETES^MODEL STEP ANDVARIABLER Adj R 2 "T"and alphaALL Traci' 1. Sport 0.32 0.10 -3.43p<0.001Socb 1. SRV 0.26 0.06 2.52p = 0.013MALES Socb 1. DAV 0.33 0.09 2.30P =0.026FEMALES Trade 1. Sport 0.41 0.15 -3.06p =0.004Socb 1. MBV 0.35 0.10 2.45p = 0.019FEMALE Socb 1. NSRV 0.47 0.18 2.54GYMNASTS p =0.018FEMALE Socb 1. DAV 0.50 0.21 2.41SKATERS p =0.028aTrad =Traditional Model (Age, Sport, %BF, Weight, Gender)bSoc =Social Model [Dieting Sub-Scale Score (DSS), Differential Association Variable(DAV), Social Reinforcement Variable (SRV), Non-Social Reinforcement Variable(NSRV), Lifestyle Variable (LV), Modelling Behaviour Variable (MBV)]90using social independent variables (dieting behaviour score, differential associationvariable, social reinforcement variable, non-social reinforcement variable, modellingbehaviour variable, and the lifestyle variable), only social reinforcement explained asignificant amount of the variance in total calcium intake. Table 37 shows that, whenusing the adjusted R squared value, the social reinforcement variable only accountedfor six percent of the variance in total calcium intake. Similar to the Traditional Model,much of the variance in total dietary calcium intake remains unexplained by thevariables in the Social Model. The predictive equation generated for this Model was(F188 = 6.37, p = 0.013):In(Predicted Total Calcium Intake) = 6.577 + 0.221(Social ReinforcementVariable).These results suggest that while the Traditional Model was able to explain morevariance in calcium intake than the Social Model, neither Model contained powerfulpredictors of total calcium intake.B) Male AthletesMRA were conducted using both the Traditional and Social Models for the maleathletes combined, the male gymnasts and then male skaters. As shown in Table 37,for male athletes combined, only one variable from the Social Model explained asignificant amount of variance in the dependent variable. The differential associationvariable explained nine percent of the variance in total dietary calcium intake for themale athletes combined. The predictive equation generated for this model was91(F143 =5.31, p =0.026):In(Predicted Total Calcium Intake) = 5.743 + 0.438(Differential AssociationVariable)None of the variables from the Traditional Model were able to explain variancein the dependent variable. Similarly, when the athletes were divided by sport, noneof the variables from either model (Traditional or Social) were able to explain variancein total calcium intake. These results suggest that for the variables studied, nonehave a significant influence on predicting total calcium intake for the male athletes inthis study. Thus, for male athletes, the variance in total dietary calcium intake wasnot explained by the variables in either the Traditional Model or the Social Model.C)^Female Athletesi)^Traditional ModelMRA was conducted using variables from the Traditional Model for all femaleathletes combined. Table 37 shows that when total calcium intake was predictedfrom the traditional variables, only sport explained a significant amount of variance inthe dependent variable. Based on the adjusted R squared value, sport was able toexplain 15% of the variance in calcium intake for the female athletes. The predictiveequation generated for this model was (F 147 = 9.36, p = 0.004):In(Predicted Calcium Intake) = 7.719 - 0.466(Sport)Where the numeric values for sport were:^1 = Skaters^2 =Gymnasts.92These results show that when the athletes were divided by gender, sport became amore powerful predictor of calcium intake for the female athletes. However, whenthe MRA was conducted using the Traditional Model for each sport separately, noneof the variables were able to significantly explain the variance in total calcium intake.These results suggest that for the Traditional variables, only sport was useful forpredicting total calcium intake.ii) Social ModelTable 37 shows the results of the MRA for the Social Model for all femaleathletes combined. The analysis showed that the modelling behaviour variable wasthe only variable that was able to explain a significant amount of variance in calciumintake. Modelling behaviour explained nine percent of the variance in total calciumintake based on the adjusted R squared value. The equation created for this modelwas (F142= 5.51, p = 0.024):In(Predicted Calcium Intake) = 6.042 + 0.350(Modelling Behaviour Variable).Table 37 shows the results from the MRA using the Social Model for femalegymnasts. This analysis showed that only one social variable entered into theequation, non-social reinforcement variable. Based on the adjusted R squared value,the non-social reinforcement variable explained 18% of the variance in total calciumintake. These results show that for the social variables included in this study, thenon-social reinforcement variable was the only variable capable of predicting calciumintake in adolescent female gymnasts. The equation generated for this MRA was93(F123 = 6.44, p = 0.018):In(Predicted Calcium Intake) = 5.108 + 0.501(Non-Social ReinforcementVariable)The MRA was also conducted for the female skaters using variables from theSocial Model. Table 37 shows that, for the skaters, the differential associationvariable explained 21 % of the variance in total calcium intake based on the adjustedR squared value. The equation generated from this MRA was (F117 =5.81, p = 0.028):In(Predicted Total Calcium Intake) = 5.723 + 0.438(Differential AssociationVariable)These results suggest that sport differences exist for the predictive power thedifferent variables in the Social Model on total calcium intake for adolescent femaleathletes in this study. However, overall, these results suggest that for the femaleathletes, much of the variance in total calcium intake remains unexplained by thevariables included in these two models.12. Summary of ResultsThe results from this study provide sufficient evidence to reject the nullhypothesis that there is no difference in the mean calcium intakes of athletesparticipating in an aesthetic sport (gymnastics) or non-aesthetic sport (speed skating).The 2x2 ANOVA (Gender X Sport) showed that averaged over gender, there was a94significant main effect of sport on total dietary calcium intake, indicating that theskaters had a significantly higher intake of calcium than the gymnasts. However,there was no significant interaction effect of gender and sport, meaning that thedifferences between the two sports for calcium intake were similar for both genders.The analyses also showed that there was a significant main effect of gender onthe dieting sub-scale scores, meaning that the female athletes scored significantlyhigher than the male athletes on the scale. Although gender differences existed,neither the male nor female athletes had mean scores suggestive of disturbed eatingbehaviours. The univariate correlational analysis and the MRA showed that there wasno relationship between the dieting sub-scale scores and total dietary calcium intakefor any of the groups of athletes.The MRA showed that for all athletes combined, sport and the socialreinforcement variable had a weak but significant influence on total dietary calciumintake. When the athletes were divided by gender and the MRA re-examined, for maleathletes combined, the differential association variable had a weak but significantinfluence on total dietary calcium intake. However, none of the variables from theTraditional Model explained a significant amount of variance in total dietary calciumintake for the male athletes. Similarly, when the male athletes were divided by sport,none of the variables from the Social or Traditional Models could significantly predictcalcium intake. For female athletes, sport and the modelling behaviour variable hada weak but significant predictive power for the dependent variable. When the femaleathletes were divided by sport, different relationships emerged. For the femaleskaters, only the differential association variable entered the equation and was ableto explain 21 % of the variance in total dietary calcium intake. For the female95gymnasts, only the non-social reinforcement variable entered the MRA equation,explaining 18% of the variance in total dietary calcium intake. Based on the MRA, theother variables from the Traditional (age, weight, gender, %BF) and Social Models(lifestyle, dieting behaviour) were not able to significantly explain variance in thedependent variable, total dietary calcium intake, in any of the athlete groups.96CHAPTER VV.^DISCUSSIONThe purpose of this study was to assess the calcium intakes of adolescentathletes and to examine factors that may influence milk and therefore calcium intakesin these youths. Adolescent athletes of both genders who participated in either anaesthetic sport (gymnastics) or a non-aesthetic sport (speed skating) took part in thestudy. In this chapter, estimated dietary calcium intakes and factors that mayinfluence calcium intake in this population are discussed separately. This is followedby a discussion of the best predictors of calcium intake comparing variables from theTraditional and Social Models. Finally, a conclusion and recommendations for futureresearch are provided.1.^Estimated Dietary Calcium IntakeInadequate calcium intakes compared to the RDA have been reported elsewhereparticularly for normally-active female adolescents (Guenther, 1986) and adolescentfemale athletes who participate in either aesthetic or non-aesthetic sports (Moffatt,1986; Perron and Endres, 1986; Rucinski, 1989; Chen et al., 1989; Benson et al.,1990). The interpretation of the severity of this problem is complicated by thedifferences of opinion among countries as to the recommended intakes for calciumduring adolescence. In the present study, the mean estimated dietary calcium intakeexceeded the RNI for gymnasts (1456 ± 984 mg and 1005 ± 534 mg) and skaters(1681 ± 931 mg and 1527 ± 750 mg) male and female athletes respectively.However, in Canada, the recommended intakes of calcium for adolescents, which97range from 700 mg to 1100 mg for various age/sex groupings, are lower than therecommended intake of calcium for American teenagers of 1200 mg per day. Thus,athletes meeting the RNI for Canada may have intakes that would be considered sub-optimal by another country's standards. A recent study by Johnston et al. (1992),has provided further evidence in support of increasing the recommended intake ofcalcium during childhood and early adolescence. The researchers found that the RDAfor calcium in the United States was insufficient to maximize bone mineral densityespecially in the normally-active prepubescent youths. However, optimal calciumintake to maximize peak bone mass has not yet been established for the adolescentpopulation since Johnston et al. (1992) examined the effects of only two levels ofcalcium intake, approximately 900 mg and 1600 mg per day, on bone mineral density.Thus, whether similar increases in bone mineral density could have been achieved atcalcium intakes below 1600 mg remains unknown, as does the converse, whetherfurther increases in bone mineral density could be achieved with calcium intakesgreater than 1600 mg per day. Finally, it is not yet known whether the increases inbone mineral density that occurred at higher calcium intakes will persist over time.It has been suggested that, as a group, athletes who participate in sports whichplace a greater emphasis on leanness may be at greater risk for developing disturbedeating habits (Borgen and Corbin, 1987; Rosen and Hough, 1988; Davis and Cowles,1989) and therefore nutrient inadequacies. In the present study, sport differences inmean calcium intakes were found with the skaters having higher intakes of totaldietary calcium than the gymnasts, which appears to support the above hypothesis.As mentioned previously, although as a group the skaters had higher dietary calciumintakes than the gymnasts, all groups of athletes exceeded the RNI for the nutrient.98Moreover, as will be discussed further, no differences in dieting sub-scale scores weredetected between sports, suggesting that the observed differences in calcium intakewere not mediated by a greater tendency toward disturbed eating habits in thegymnasts.To investigate where the differences existed between the two sports forcalcium intake, differences in estimated dietary calcium from fluid milk, other dairyproducts and non-dairy products were assessed. The results showed that thedifference in total dietary calcium intake between the sports was due to the skatersconsuming more fluid milk than the gymnasts. Otherwise, all groups of athletes hadsimilar calcium intakes provided from other dairy food sources and non-dairy foodsources.It has been suggested that American teenagers may be replacing their milkintake with soda, since a negative correlation between calcium intake and sodaconsumption was reported by Guenther (1986). This is a potential health problem fortwo reasons; first, the adolescents may be consuming inadequate amounts of calciumas a result of increased soda consumption and secondly, the high phosphorus contentof soft drinks may limit calcium absorption and alter the calcium/phosphorus ratiowhich may contribute to the increased risk of bone fractures in later life (Wyshak etal., 1989; Gong and Spear, 1988). However, in the present study on adolescentathletes, no such relationship between total dietary calcium intake and sodaconsumption was established. There were no significant correlations (positive ornegative) between total dietary calcium and soda intake when the analyses wereconducted on all athletes together, nor when the athletes were divided by gender andthen by sport. Therefore, for the athletes in this study, these results demonstrate that99the youths were not replacing their milk intake and therefore calcium intake with soda.2.^Factors Influencing Eating BehaviourA)^Dieting Sub-Scale ScoreThere is controversy in the literature as to whether athletes who participate inaesthetic sports are more prone to tendencies toward disturbed eating behaviours thanthose who participate in non-aesthetic sports (Rosen et al., 1986; Benson et al.,1990). In fact, in the study by Benson et al. (1990) on female adolescent athletes,the swimmers scored higher on the body dissatisfaction sub-scale of the EatingDisorders Inventory test than either the gymnasts or the controls. In the presentstudy, however, the results showed that in both the aesthetic and non-aestheticsports, male athletes scored lower than the female athletes. Thus, for theseadolescent athletes, the type of sport they participated in did not influence theirdieting sub-scale scores.Although gender differences were evident in this study, neither the mean scoresof the male nor female athletes were suggestive of pathogenic tendencies towardavoidance of fattening foods and shape preoccupation for the groups as a whole.However, 21 % of the female athletes and eight percent of the male athletes haddieting sub-scale scores suggestive of tendencies towards disturbed eatingbehaviours. When the dieting sub-scale scores were examined by gender and sport,all groups showed a similar proportion of athletes with scores suggestive of disturbedeating behaviour. The results from the present study suggest that the differences intendencies towards disturbed eating behaviours may be a result of gender differences,not type of sport.100In 1989, Rucinski reported gender differences in the EAT scores for figureskaters with the female skaters scoring significantly higher than the male skaters.Forty-eight percent of the female figure skaters and none of the male skaters hadscores that were suggestive of anorexic behaviour, suggesting that gender differencesexist for tendencies towards pathogenic weight control behaviours for athletescompeting in aesthetic sports. While the results of the present study support thisconcept of gender differences in tendencies towards disturbed eating behaviours inaesthetic sports, it should be noted that similar gender differences were found for theathletes participating in a non-aesthetic sport.It has been reported that dieting sub-scale scores increase significantly with agein normally active female adolescents (Wood et al., 1992). However, the reportedincrease in mean scores with age resulted in the sixteen year old females scoringapproximately one point higher than those females twelve years of age on the dietingsub-scale. There were no significant increases seen in the scores as a result of agefor the other sub-scales of the EAT (Wood et al., 1992). While the mean age of thefemale athletes in Rucinski's study (1989) was approximately three years older thanthe age of the gymnasts in the present study, the age difference is not able to explainthe differences between the two groups of athletes for tendencies towards disturbedeating behaviours (48% in Rucinski's skaters verses 21 % of gymnasts in the presentstudy). Thus, maturation could not account for the entire difference between thefemale gymnasts and the female figure skaters for tendencies towards disturbedeating behaviours.When comparing the competitive level of the athletes participating in the twostudies, the athletes in Rucinski's study (1989) were all national calibre figure skaters.101However, while some gymnasts in the present study competed at national andinternational competitions, clearly, not all athletes were national calibre athletes.Therefore, it is possible that the difference between the two studies for tendenciestowards disturbed eating behaviours may have resulted from the figure skaters inRucinski's study experiencing greater pressures to strive for leanness because ofincreased performance expectations. Thus, increased performance expectations mayhave predisposed the figure skaters to greater tendencies towards disturbed eatingbehaviours and nutrient inadequacies than the gymnasts. To investigate thishypothesis in the present study, the dieting sub-scale scores of athletes competingat a national or international level were compared to those who participated only ata provincial level. The t-test showed that the athletes who competed at the higherlevel scored significantly higher on the dieting sub-scale (6.3 ± 5.3, n =33) than theprovincial level athletes (2.4± 1.7, n =18). However, neither group had mean scoresthat were suggestive of tendencies towards disturbed eating behaviour.When the univariate relationship between total dietary calcium intake and thedieting sub-scale score was examined for all athletes combined, and then re-examinedwhen the athletes were divided by gender and then by sport, no significant linearrelationships were found. These results suggest that there were no linearrelationships between total dietary calcium intake and tendencies towards disturbedeating behaviours for any of the groups of athletes in this study.B)^Social Model VariablesLewis et al. (1989) have identified social variables which have been shown tohave an indirect and direct influence on beverage selection behaviour in adults. For102selection of milk, these variables were differential association, social and non-socialreinforcement, nutrition knowledge, attitude, commitment and behaviour modelling.In the present study, similar social variables were studied, however, overall thevariables were shown to have a weak association with total dietary calcium intake andthe strength of association varied among different groups of athletes.The Social Model variables were first examined for differences between thegroups of athletes on the basis of gender and sport. Main effects were seen in thedieting sub-scale scores and the social reinforcement scores. The influence of genderon dieting sub-scale scores has been discussed above, and therefore, will not bediscussed in this section. The results of the ANOVA for the social reinforcementvariable indicated that the gymnasts exhibited stronger disagreement to thestatements that the consumption of milk evoked positive feelings and a sense ofbelonging to a group than the skaters. These sport differences in the socialreinforcement variable were similar for both genders. When the univariate correlationbetween dietary calcium intake and the social reinforcement variable was examined,a weak but significant positive relationship was found. However, when the athleteswere divided by gender and sport, the strength of the correlation increased for femaleathletes combined and female skaters. Thus, despite the general disagreement of theathletes that the consumption of milk evoked positive feelings and a sense ofbelonging to a group, for female athletes combined and for female skaters, there wasa weak but positive relationship between these positive feelings and calcium intake.This relationship, however, was not seen in the other groups of athletes.When the other independent variables from the Social Model were examined forsport and gender differences, the results showed that, on average, all groups of103athletes had similar scores for these variables. However, it became apparent that therelationships between these social variables and calcium intake varied greatly amongthe groups of athletes. For example, there was a weak but significant correlationbetween the modelling behaviour variable and calcium intake for all athletes combined.However, when the athletes were divided by gender and sport, the relationships thatemerged were very different between the groups of athletes, with a strongerrelationship between the modelling behaviour variable and dietary calcium intake forfemale gymnasts, while no such relationship was seen for the other groups ofathletes. Other linear correlational analyses helped to confirm that despite the similarmean scores between the groups for most variables, linear univariate relationships ofthe independent variables with total dietary calcium intake varied greatly among thegroups of athletes. In addition, the relationships between the independent variablesand total dietary calcium intake were often strengthened when the athletes weredivided by gender and sport. Overall, the univariate correlational analyses suggestthat sport and gender differences existed for the relationships between theindependent variables of the Social Model and total dietary calcium intake.C) Traditional Model VariablesUnivariate correlational analyses with total dietary calcium intake could not beconducted on all variables of the Traditional Model because it is not possible toconduct correlational analysis on categorical data. Therefore, only age, weight and%BF were analyzed for linear univariate relationships with total dietary calcium intake.Weight had a significant relationship with the dependent variable for all athletescombined. When the athletes were divided by gender, weight and %BF were104significantly correlated with total dietary calcium intake for all female athletes, whilefor male athletes, no such relationships were seen for the Traditional variables. Whenthe athletes were further divided by sport, the relationships no longer existed.Therefore, weight and %BF may be highly correlated with the type of sport.3.^Multiple Regression AnalysisThe results of the MRA confirm what the univariate analyses suggested;namely, that there is great variation in the ability of the independent variables topredict total dietary calcium intake among the different groups of athletes. When allathletes were analyzed together, variables from neither the Social nor TraditionalModels were powerful predictors of total dietary calcium intake. In addition, when theathletes were divided by gender and sport, different social variables entered theregression equations for each group. For example, the social reinforcement variablewas a weak but significant predictor for all athletes; however, it was not a predictorwhen the athletes were divided by gender and sport. Although some of the socialvariables were inter-correlated, they were not highly inter-correlated, therefore, thisshould not have been the cause of the variation of predictive variables for total dietarycalcium intake among the groups of athletes. Rather, it may be that the influence ofthe social variables on the dependent variables differed among the groups of athletes.Lewis et al. (1989) set an a priori level for the Social and Traditional Models.The researchers felt that the Models must be able to explain 35% of the variance inthe frequency of beverage selection to be useful. In addition, the researchers useda different approach than the present study for data analysis. Forced entry of thevariables was used for the multiple regression analysis, and therefore, did not105investigate which independent variable(s) from each model best predicted beverageconsumption. By Lewis's standards, however, none of the variables in the presentstudy, whether Social or Traditional, were able to explain 35% of the variance in totaldietary calcium intake.The ability of the Models as a whole to explain variance in total dietary calciumintake could not be directly compared to that found by Lewis et al. (1989) since thevariables in the Models differed between the studies. For example, Lewis et al.(1989) included nutrition knowledge as a variable in the Traditional Model. Thisvariable was not included in the present study because of the difficulty of constructinga knowledge test appropriate for athletes 12 to 18 years of age. Moreover, previousresearch generally does not support a major role for nutrition knowledge as a positivefactor influencing adolescents' food intakes (Schwartz, 1975; Douglas, 1984; Whiteand Skinner, 1988; Byrd-Bredbenner et al., 1988). As previously mentioned, theability of the individual variables to predict beverage consumption using MRA was notanalyzed by Lewis et al. (1989), so the predictability of the independent variables onthe dependent variables cannot be compared between the two studies.In the present study, when the athletes were divided by gender and the MRAre-examined, different variables entered the predictive equations. For all male athletescombined, only the differential association variable from the Social Model was able topredict total dietary calcium intake, explaining only nine percent of the variance intotal dietary calcium intake. When the male athletes were further divided by sport,none of the variables from either the Social or Traditional Models were able to predictthe dependent variable. Thus, the results of this study suggest that although ingeneral male athletes enjoyed the taste of dairy products, felt that their social106environment promoted the consumption of milk and had the opportunity to witnesspeople they admire consuming milk frequently, none of these had an impact on theirtotal dietary calcium intake. Therefore, these models were not useful for predictingtotal dietary calcium intake in adolescent male athletes, suggesting gender differencesexist for factors that influence calcium intake in adolescent athletes.These findings for male athletes are not necessarily surprising. Adolescence isa time of rapid growth and development which increases energy and nutrient needs.These needs are further escalated by the demands of training. Since increasednutrient needs are followed by a natural increase in appetite, it is possible that thephysiological need for food dominates any possible social influences. It has beenreported elsewhere that the nutritional adequacy of male adolescent athletes may beattributed to a high food consumption rather than careful food selection practices onthe part of the athletes (Douglas, 1984). While both groups of male athletes in thepresent study had mean calcium intakes which exceeded the RNI, it cannot bediscerned whether these findings reflect a general pattern of intentional healthy foodchoices or rather occurred inadvertently due to a high food intake.For female athletes, sport itself was a significant predictor of total dietarycalcium intake. For example, in the Traditional Model, sport increased in predictivepower once the effects of gender were controlled. Since the same relationship wasnot seen in the male athletes, including them in the initial analysis reduced thepredictive power of sport in the female athletes. It is difficult to determine why sportwas a significant predictor of total dietary calcium intake for the female athletes. TheANOVAs showed that the skaters consumed more total dietary calcium than thegymnasts primarily in the form of fluid milk. The skaters were taller and heavier than107the gymnasts which could explain the difference in intake since larger people generallyeat more. However, sport entered the regression equations and weight did not, thus,sport had a higher partial correlation with total calcium intake than weight. When thefemale athletes were further divided by sport, thus in effect controlling for gender andsport in the MRA, none of the traditional variables enter into the predictive equation.Accordingly, for the traditional variables, only sport had weak but significant powerfor predicting total dietary calcium intake.In the Social Model, the modelling behaviour variable was the only variable thatcould significantly predict total dietary calcium intake for all female athletes combined.However, the predictive power was weak and inferior to that of sport in theTraditional Model. When the athletes were divided by sport and the analysis re-run,different predictive equations were generated. For the female skaters, the differentialassociation variable could significantly predict total dietary calcium intake. It was theonly variable that entered the equation, and although the predictive power increasedrelatively, it was still a weak predictor of total dietary calcium intake. For femalegymnasts, a different equation was generated. The non-social reinforcement variableentered the equation, explaining 18% of the variance in the dependent variable. Theresults of the MRA indicate that although only a weak association was found, factorsthat influence total dietary calcium intake varied between the two groups of femaleathletes.In 1988, Contento et al. conducted a study to examine factors that influencefood choices of adolescents and found that, in fact, great variation existed betweenadolescents in terms of influential forces on food selection behaviour. The researcherswere able to identify five sub-groups within an adolescent population (n = 355) on the108basis of their food choice motivations. For the food attributes studied, each sub-group's food selections were influenced to a different extent by the various attributeswith some attributes exerting stronger influences and other factors exerting noinfluences. This study supports the findings of the present study that different sub-groups may be influenced by different components of their social environment.It is difficult to determine why differential association and non-socialreinforcement (taste) were weak yet significant predictors of total dietary calciumintake for the female skaters and female gymnasts, respectively. All groups ofathletes had similar scores for both variables as shown by ANOVA. For differentialassociation, all athletes agreed to a similar extent that they perceived that their socialenvironment was supportive of the consumption of milk. For the female skaters,however, those who had stronger agreement with the statements had higher calciumintakes as well as the converse, those who agreed to a lesser extent or evendisagreed with the statements, had lower calcium intakes. The differential associationvariable was a mean score derived by averaging the scores from all sub-scales,friends, family, health experts and media. Thus, it is not possible to determinewhether some groups of the social environment exerted stronger influences on thedependent variable than other social environmental groups. In addition, it is possiblethat the sub-groups of the social environment exerted different influences on variousathletes within a sub-population, thereby, diluting the association of the differentialassociation variable with total dietary calcium intake. Unfortunately, for thepreviously discussed reasons, it is not possible to analyze the association of eachsocial environmental sub-scale with calcium intake.As for the female gymnasts, taste may have had the strongest association with109the dependent variable because female gymnasts likely had lower energy requirementsthan the other groups of athletes in the study because of their smaller body size andthe lower energy expenditure related to their sport. Thus, it is reasonable to speculatethat the taste of food would be a strong predictor of food consumption since if foodconsumption is more limited for this group in order to maintain energy balance, it isreasonable that the athletes would want to consume foods that tasted good to them.In the relative context of this study, variables from the Social Model were onlyable to explain 18% and 21% of the variance in total dietary calcium intake in thesub-groups of female athletes, while the Traditional variable, sport, explained 15% ofthe variance in the dependent variable for all female athletes combined. If strategieswere developed to improve calcium intake using the social variables identified in thisstudy, based on an 18% and 21% explained variance, the athletes' calcium intakescould be improved by approximately 210 mg and 320 mg of calcium per day forfemale gymnasts and skaters respectively. While such an improvement would seemsignificant, different strategies would need to be developed for each sub-group ofathletes since differential association was associated with the skaters' intakes andtaste of the dairy products was associated with the female gymnasts' intakes.Moreover, since on average the athletes exceeded the RNI for calcium, the merit ofdeveloping group strategies to improve calcium intake could be questioned.Since variables from neither the Social nor Traditional Models were shown tobe good predictors of total dietary calcium intake, the question becomes what doespredict food selection behaviour in adolescent athletes? From the literature reviewed,no specific studies have empirically examined motives behind food selection in thispopulation. A variety of assumptions/hypotheses about factors that influence food110selection in adolescents athletes have been proposed, however, they do not appearto be supported by empirical evidence. For example, there is some literature thataddresses the issue of who adolescent athletes consider as their best sources fornutrition information (Douglas, 1984), where the athletes cited parents as the bestsource and coaches to a lesser extent. However, the athletes were high schoolcalibre and the influence of the nutrition information sources on food selectionpractices was not assessed. Another study suggested that parents, friends, peers andcoaches influence female adolescent swimmers' perceptions about weight (Drummeret al., 1987), however, the association between these external influences and foodselection behaviour was not examined. The importance of role models has beenimplicated in modifying food selection behaviour (Rush, 1990), however, once againthis hypothesis was not supported by research findings. The assumption that "peershave a strong influence on food selection habits" has also been used frequently in theliterature, however, there is insubstantial evidence to support or contradict thishypothesis. Regardless, peer influence has somehow been advocated as an influentialfactor on food selection behaviour (Carruth and Skinner, 1988).In terms of nutrition education strategies, the recurrent findings of the inabilityof nutrition education programs (particularly knowledge based) to demonstrate achange in food practices (Schwartz, 1975; Douglas, 1984; White and Skinner,1988) has not deterred the recommendation of providing nutrition education forathletes to resolve the problem of inadequate nutrient intakes (Moffatt, 1984; Chenet al., 1989; Berning et al., 1991). It appears that the lack of information regardingfactors that influence food selection behaviour in adolescent athletes has perpetuatedthe use of unsubstantiated assumptions for structuring nutrition education programs111and explaining food practices.4. Conclusion and Recommendations for Future ResearchIn general, all groups of athletes met their RNI for calcium intake, suggestingthat calcium is not a nutrient of concern for the adolescent athletes in this study usingCanadian standards. The results also indicated that, based on the mean scores fromthe dieting sub-scale, none of the groups of athletes exhibited tendencies towardsdisturbed eating behaviours. Accordingly, the athletes who participated in theaesthetic sport were not more prone to disturbed eating behaviour than those athleteswho participated in the non-aesthetic sport even though they were considerablyleaner.Overall, the MRA demonstrated that variables from neither the Traditional norSocial Models were strong predictors of total dietary calcium intake. While someathletes did have low calcium intakes, none of the variables in the Social or TraditionalModels could explain these low intakes. Thus, overall, in contrast to Lewis et al.(1989), social learning variables were not useful in predicting dietary calcium intakesin adolescent athletes.The reasons why the social variables were not useful in predicting total dietarycalcium intake in adolescent athletes can only be speculated. For the male athletes,it is possible that the strongest influence on food selection behaviour is thephysiological need for food to support the rapid growth and development whichoccurs during adolescence. For female athletes, some of the social variables did havea weak but significant association with total dietary calcium intake suggesting thatfactors that influence the dependent variable vary between the athletic groups. In112addition to the previously cited possibilities why the Social variables were not goodpredictors of total dietary calcium intake; namely, diversity of social influences withinthe sub-groups of athletes, other methodological difficulties may have contributed tothe weak findings of this study. For example, the number of subjects included in theMRA for all athletes combined and when the athletes were separated by gender, wasadequate for the number of independent variables. However, because separatevariables affected the sub-groups differently, the relationships may have been maskedin the analyses that included both genders and both sports. It is also possible that byseparating the athletes into four groups by gender and sport, power was lost. Thenumber of subjects per group may have been too small to allow the independentvariable(s) to be strong predictors of total dietary calcium intake since calcium intakeitself is so highly variable. However, it should be noted that the practical utility of thesocial learning variables is likely to be very limited since they were not obviouspredictors of total dietary calcium intake for the athletes as a group.It is also possible that the instruments used to assess social influences were nottruly reflective of the athletes' social environments. For example, the modellingbehaviour scale may have gathered different results if the models were moremeaningful to an athletic population, such as including "the World Champion for yoursport" as a model. The differential association scale may have been more reflectiveof the athletes' social environment if "coach" was added to the scale. It is possiblethat the social environment of normally active adolescents differs from the socialenvironment of adolescent athletes. Therefore, the problem of identifying variableswhich predict food selection behaviour across situations remains unsolved foradolescent athletes.113Finally, there appear to be some serious gaps in the literature regarding thenutritional intakes and practices of adolescents and especially adolescent athletes.Further research is required to establish optimal calcium intakes which maximize bonehealth for both normally active and athletic adolescents. In addition, there is a lackof information on factors that influence food selection in these youths. Identifyingfactors which influence food selection is complicated by the recurrent assumptionsfound in the literature about adolescents which are not supported or contradicted byresearch (Carruth and Skinner, 1988). Thus, the challenge remains to develop a toolto identify the influences on food selection behaviour, to accept that great variationexists within this population in terms of influences on food selection behaviour andto use this information to develop effective education strategies for the appropriatesub-groups of adolescents and adolescent athletes to improve nutrition practices ifrequired.114REFERENCESAkers RL, Krohn MD, Lanza-Kaduce L, Rodosevich M. Social learning and deviantbehavior: a specific test of a general theory. Am Soc Rev 1979;44:636-655.American Dietetic Association. Position of the American Dietetic Association:Nutrition education for the public. J Am Diet Assoc 1990;90:107-1 10.Anderson JJB. Dietary calcium and bone mass through the life cycle. Nutr Today1990;25:9-14.Angus RM, Sambrook PN, Pocock NA, Eisman JA. A simple method for assessingcalcium intake in caucasian women. J Am Diet Assoc 1989;89:209-214.Bachrach LK, Guido D, Katzman D, Litt IF, Marcus R. Decreased bone density inadolescent girls with anorexia nervosa. Pediatrics 1990;86:440-447.Bailey DA, Martin AD, Houston CS, Howie JC. Physical activity, nutrition, bonedensity and osteoporosis. Austr J Sci Med Sport 1986;18:3-8.Bandini LG and Dietz WH. Assessment of body fatness in childhood and obesity:evaluation of laboratory and anthropometric techniques. J Am Diet Assoc1987;87:1344-1348.Barr SI. Relationship of eating attitudes to anthropometric variables and dietaryintakes of female collegiate swimmers. J Am Diet Assoc 1991;91:976-977.Barr SI and Pare S. Development and validation of a food frequency questionnairefor estimation of adolescents' calcium intake. (Abstract) J Can Diet Assoc1992;53:236.Benson JE, Allemann Y, Theintz GE, Howald H. Eating problems and calorie intakelevels in Swiss adolescent athletes. Int J Sports Med 1990;11:249-252.Bergman EA, Boyungs JC, Erickson ML. Comparison of a food frequencyquestionnaire and a 3 day diet record. Am Diet Assoc 1990;90:1431-1433.Berning JR, Troup JP, Van Handel PJ, Daniels J, Daniels N. The nutritional habitsof young swimmers. Int J Sport Nutr 1991;1:240-248.Block G. A review of validation of dietary assessment methods. Am J Epidemiol1982;1 15:492-505.Block G and Hartman AM. Issues in reproducibility and validity of dietary studies.Am J Clin Nutr 1989;50:1133-1138.115Borgen JS and Corbin CB. Eating disorders among female athletes. PhysSportsmed 1987;15(2):89-95.Button EJ and Whitehouse A. Subclinical anorexia nervosa. Psychological Med1981;1 1:509-516.Byrd-Bredbenner C, Shannon B, Hsu L, Smith DH. A nutrition education curriculumfor senior high home economics students: its effect an students'knowledge, attitudes, and behaviors. J Nutr Educ 1988;20:341-346.Carruth BR and Skinner JD. What do we know and what should we know aboutnutrition education for adolescents? J Nutr Educ 1988;20:271-272.Chen JD, Wang JF, Zhao YW, Wang SW, Jiao Y, Hou XY. Nutritional problemsand measures in elite and amateur athletes. Am J Clin Nutr 1989;49:1084-1089.Clark N, Nelson M, Evans W. Nutrition education for elite female runners. PhysSports Med 1988;16(2):124-136.Committee on Sports Medicine, the American Academy of Pediatrics. Amenorrheain adolescent athletes. Pediatrics 1989;84:394-395.Contento IR, Michela JL, Goldberg CJ. Food choices among adolescents:population segmentation by motivations. J Nutr Educ 1988;20:289-298.Cronbach L. Coefficient alpha and the internal structure of tests. Psychometrika1951;16:297-334.Cummings SR, Block G, McHenry K, Baron RB. Evaluation of two food frequencymethods of measuring dietary calcium intake. Am J Epidemiol1987;126:796-802.Davis C and Cowles M. A comparison of weight and diet concerns and personalityfactors among female athletes and non-athletes. J Psychosomat Res1989;33:527-536.Dawson-Hughes B, Dallal GE, Krell EA, Sadowski L, Sahyoun N, Tannenbaum S. Acontrolled trial of the effect of calcium supplementation on bone density inpostmenopausal women. N Engl J Med 1990;323:878-883.Deurenberg P, Kusters CLS, Smit HE. Assessment of body composition bybioelectrical impedance in children and young adults is strongly age-dependent. European J Clin Nutr 1990a;44:261-268.116Deurenberg P, Pieters JJ, Hautvast JG. The assessment of the body fat percentageby skinfold thickness measurements in childhood and young adolescence. BrJ Nutr 1990b; 63:293-303.Douglas PO. Nutrition knowledge and food practices of high school athletes. J AmDiet Assoc 1984;84:1198-1202.Drummer GM, Rosen LW, Heusner WW, Roberts PJ, Counsilman JE. Pathogenicweight-control behaviors of young competitive swimmers. Phys Sportsmed1987;15(5):75-84.Durnin JVGA and Womersley J. Body fat assessed from total body density and itsestimation from skinfold thickness: measurements on 481 men and womenaged from 16 to 72 years. Br J Nutr 1974;32:77-97.Enns MP, Drewnowski A, Grinker JA. Body composition, body size estimation andattitudes towards eating in male college athletes. Psychosomatic Med1987;49:56-64.Evans RA. Calcium and osteoporosis. Med J Austr 1990;152:431-433.Falls HB and Humphrey LD. Body type and composition differences betweenplacers and non-placers in an AIAW gymnastics meet. Res Q 1978;49:38-43.Faria IE and Faria EW. Relationship of the anthropometric and physicalcharacteristics of male junior gymnasts to performance. J Sports Med PhysFitness 1989;29:369-378.Food and Nutrition Board. Recommended Dietary Allowances. 10th ed.Washington, D.C.:National Academy of Sciences, 1989.Forbes GB. Human body composition growth, aging, nutrition and activity. NewYork, Springer-Verlag, 1987.Forsyth HL and Sinning WE. The anthropometric estimation of body density andlean body weight of male athletes. Med Sci Sports 1973;5:174-180.Garner DM and Garfinkel PE. The eating attitudes test: an index of the symptomsof anorexia nervosa. Psychological Med 1979;9:273-279.Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The eating attitudes test:psychometric features and clinical correlations. Psychological Med1982;12:817-878.Gibson RS. Principles of nutritional assessment. New York: Oxford UniversityPress, 1990.117Gong EJ and Spear BA. Adolescent growth and development: implications fornutritional needs. J Nutr Educ 1988;20:273-279.Guenther PM. Beverages in the diets of American teenagers. J Am Diet Assoc1986;86:493-499.Harrison GG, Buskirk ER, Carter JE, Johnston FE Lohman TG, Pollock ML, RocheAF, Wilmore J. Skinfold thickness and measurement technique. In LohmanTG, Roche AF, Martorell R. Anthropometric Standardization ReferenceManual. Champaign IL. Human Kinetics Books, 1988, pp55-68.Health and Welfare Canada. Canadian Nutrient File. Ottawa: Bureau of NutritionalSciences, 1982.Health and Welfare Canada. Nutrition Recommendations. The Report of theScientific Review Committee. Ottawa: Supply and Services Canada, 1990.Hergenroeder AC and Klish WJ. Body composition in adolescent athletes. PediatricClinics North America 1990;37:1057-1083.Highet R. Athletic amenorrhoea: an update on aetiology, complications andmanagement. Sports Med 1989;7:82-108.Jackson AS, Pollock ML, Ward A. Generalized equations for predicting bodydensity of women. Med Sci Sports Exerc 1980;12:175-182.Johnson DW and Johnson RT. Nutrition education: a model for effectiveness, asynthesis of research. J Nutr Educ 1985;17:S1-S44.Johnston CC, Miller JZ, Slemenda CW, Reister TK, Hui S, Christian JC, PeacockM. Calcium supplementation and increases in bone mineral density inchildren. N Engl J Med 1992;327:82-87.Kreipe RE and Forbes GB. Osteoporosis: A new morbidity for dieting femaleadolescents? Pediatrics 1990;86:478-480.Krondl M and Lau D. Social determinants on human food selection. In Barker LM.(ed) Psychobiology of Human Food Selection. Westport, CT:AVI PublishingCo. Inc., 1982, p139-151.Lewis CJ. Use of a social learning model to predict frequency of four foodconsumption behaviors, a thesis in nutrition. Pennsylvania State University,1986.Lewis CJ, Sims LS, Shannon B. Examination of specific nutritional health behaviorsusing a social cognitive model. J Am Diet Assoc 1989;89:194-202.118Lohman TG. Applicability of body composition techniques and constants forchildren and youths. Exerc Sport Sci Review 1986;14:325-357.Lohman TG, Boileau RA, Slaughter MH. Body composition in children and youths.In Boileau RA (ed), Advances in Pediatric Sport Sciences. Champaign, IL:Human Kinetics, 1984, p29-57.Margetts BM, Cade JE, Osmond C. Comparison of a food frequency questionnairewith a diet record. Int J Epidemiol 1989;18:868-873.Matkovic V. Diet, genetics and peak bone mass of adolescent girls. Nutr Today1991;26:21-24.Matkovic V, Fontana D, Tominac C, Goel P, Chestnut CH. Factors that influencepeak bone mass formation: a study of calcium balance and the inheritanceof bone mass in adolescent females. Am J Clin Nutr 1990;52:878-888.Matkovic V, Kostial K, Simonovic I, Buzina R, Brodarec A, Nordin BEC. Bone statusand fracture rates in two regions of Yugoslavia. Am J Clin Nutr1979;32:540-549.Moffatt RJ. Dietary status of elite female high school gymnasts: inadequacy ofvitamin and mineral intake. J Am Diet Assoc 1986;85:1361-1363.Olson CM and Gillespie AH. Proceedings of the workshop on nutrition educationresearch, applying principles from the behavioral sciences. J Nutr Educ1981;13 Suppl 1:S1.Olson CM and Leitch Kelley G. The challenge of implementing theory-basedintervention research in nutrition education. J Nutr Educ 1989;21:280-284.Perron M and Endres J. Knowledge, attitudes, and dietary practices of femaleathletes. J Am Diet Assoc 1985;85:573-576.Raisz LG. Logical and systemic factors in the pathogenesis of osteoporosis. N EnglJ Med 1988;318:797-803.Ried IR. The role of dietary calcium in the pathogenesis and treatment ofosteoporosis. NZ Med J 1989;102:532-533.Riis B, Thomsen K, Christiansen C. Does calcium supplementation preventpostmenopausal bone loss? N Engl J Med 1987;316:173-177.Rosen LW and Hough DO. Pathogenic weight-control behaviors of female collegegymnasts. Phys Sportsmed 1988;16(9):141-144.119Rosen LW, McKeag DB, Hough DO, Curley V. Pathogenic weight-control behaviorin female athletes. Phys Sportsmed 1986;14(11:79-86.Rucinski A. Relationship of body image and dietary intake of competitive iceskaters. J Am Diet Assoc 1989;89:98-100.Rush KL. The school nurse and sport nutrition for the adolescent athlete. J SchoolHealth 1990;60:425-427.Sandler R, Slemenda CW, LaPorte RE, Cauley JA, Schramm MM, Barresi ML,Kriska AM. Postmenopausal bone density and milk consumption in childhoodand adolescence. Am J Clin Nutr 1985;42:270-274.Schwartz NE. Nutrition knowledge, attitudes, and practices of high schoolgraduates. J Am Diet Assoc 1975;66:28-31.Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Stillman RJ, Van Loan MD,Bemben DA. Skinfold equations for estimation of body fatness in childrenand youth. Human Biol 1988;60:709-723.Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Wilmore JH. Influences ofmaturation of relationships of skinfolds to body density: a cross-sectionalstudy. Hum Biol 1984;56:681-689.Strickland DE. Social learning theory and deviant behavior: a specific test of ageneral theory- a comment and critique. Am Soc Rev 1982;47:162-173.Stuff JE, Garza C, O'Brian Smith E, Nichols BL, Montandon CM. A comparison ofdietary methods in nutritional studies. Am J Clin Nutr 1983;37:300-306.Thorland WG, Johnson GO, Tharp GD, Fagot TG, Hammer RW. Validity ofanthropometric equations for the estimation of body density in adolescentathletes. Med Sci Sport Exerc 1984;16:77-81.Wardle J and Marshland L. Adolescent concerns about weight and eating: a social-developmental perspective. J Psychosomatic Res 1990;34:377-391.Weststrate JA and Deurenberg P. Body composition in children: proposal for amethod for calculating body fat percent from total-thickness measurements.Am J Clin Nutr 1989;50:1104-1115.White AA and Skinner JD. Can goal setting as a component of nutrition educationeffect behaviour change among adolescents? J Nutr Educ 1988;20:327-335.120Willett WC, Reynolds RD, Cottrell-Hoehner S, Sampson L, Browne ML. Validationof a semi-quantitative food frequency questionnaire: comparison with a 1year diet record. J Am Diet Assoc 1987;87:43-47.Wilmore JH. Body composition in sport and exercise: directions for futureresearch. Med Sci Sport Exerc 1983;15:21-31.Wilmore J. Eating and weight disorders in the female athlete. Int J Sports Med1991;1:104-117.Wood A, Waller G, Miller J, Slade P. The development of eating attitude testscores in adolescence. Int J Eating Disorders 1992;11:279-282.Wyshak G, Frisch RE, Albright TE, Albright NL, Schiff I, Witschi J. Nonalchololiccarbonated beverage consumption and bone fractures among women formercollege athletes. J Orthopaedic Res 1989;7:91-99.121APPENDICESAppendix A:^Definition of Terms1. Athlete: an individual competing at a provincial level or higher in a selectedorganized sport. They are members of a provincial sporting organization.2. Adolescent: an individual within the chronological age range of 12 to 18 years.3. Nutrient Insufficiency/Inadequacy: failure to meet the RNI or RDA. Clearly, notall individuals with an intake of less than the RNI/RDA will be failing to meettheir individual requirements; however, the probability of an individual's intakebeing inadequate increases as their intake falls below the RNI/RDA.4. Aesthetic sports: Sports which include aesthetics in performance scores. Inthis study gymnastics represents the aesthetic sports.5. Non-aesthetic sports: Sports that have quantitative performance results suchas fastest times and most points scored. In this study speed skating representsthe non-aesthetic sports.6. Differential Association: a group of environmental factors which influence foodselection. These factors have been operationalized by Lewis (1986):a) Family: use of and feelings about a given food;122b) Friends: use of and feelings about a given food;c) Experts: perception of health experts' recommendations concerning a givenfood;d) Media: perceptions of entertainment/persuasive qualityof television advertising for a given food;e) Lifestyle: degree to which breakfast, lunch, and supperare eaten away from home. In the present study this factor was separatedfrom differential association and considered a separate variable since it wasdefined as the percentage of meals and snacks eaten away from home.7. Behaviour modelling: how often certain persons (models) are seen consuminga given food (Lewis, 1986).8. Reinforcement as operationalized by Lewis (1986):a) Social: perception of positive feelings as a functionof consuming a given food;b) Non-social: enjoyment of the taste of a given food.123r. R.D. .ratleyDirector, Research Servicesand Acting ChairmanTHIS CERTIFICATE OF APPROVAL IS VALID FOR THREE YEARSFROM THE ABOVE APPROVAL DATE PROVIDED THERE IS NOCHANGE IN THE EXPERIMENTAL PROCEDURESAppendix B:^Informed Consent Form, Ethics Approval Form and Questionnaire:The University of British Columbia^B91-280Office of Research ServicesBEHAVIOURAL SCIENCES SCREENING COMMITTEE FOR RESEARCHAND OTHER STUDIES INVOLVING HUMAN SUBJECTSCERTIFICATE^of APPROVALINVESTIGATOR: Barr, S.I.UBC DEPT:^Family & Nutr SciINSTITUTION:^Athletes' training ctrTITLE:^Factors associated with calcium intake inadolescent athletesNUMBER:^B91-280CO-INVEST:^Webster, B.APPROVED:^OCT it 1991The protocol describing the above-named project has beenreviewed by the Committee and the experimental procedures werefound to be acceptable on ethical grounds for researchinvolving human subjects.124THE UNIVERSITY OF BRITISH COLUMBIAFOOD HABITS OF TEENAGE ATHLETESDear Athlete:School of Family andNutritional SciencesDivision of Human Nutrition2205 East MallVancouver, B.C. Canada V6T 1W5You are invited to participate in a nutrition study to help us learn more about the food habitsof teenage athletes. Everyone knows that the teen years are important ones for buildinggood nutrition habits to last a lifetime. However, the purpose of this study is to learn moreabout what teenage athletes think about their food and nutrition and about how their busy'lives affect what they eat. We also know that body shape and size are important issue atthis time, and we need to know more about how food habits are affected by these and otherconcerns.Your participation in this study will be of great assistance in obtaining the information that isneeded about teenage athletes' nutrition. We will need you to complete a questionnaire thatwill take approximately 30 minutes, for which there are no right or wrong answers to thequestions.Measurements of your height, weight and assessment of percent body fat, takingapproximately 10 minutes, will also be required. Percent fat will be measured using skinfoldmeasurements (calipers) and a new technique called bioelectrical impedance analysis (BIA►.BIA is a non-invasive method of assessing body composition that requires the attachment ofremovable adhesive electrodes to your wrist and ankle. Both of these techniques are safeand do not cause any discomfort. All information will be kept completely confidential sinceonly the investigators will see your results. You may request a copy of your own bodycomposition results.This study is being conducted by Brenda Webster, P. Dt. and Dr. Susan Barr of U.B.C.'sSchool of Family and Nutritional Sciences. If you have any questions regarding the study,you can contact Brenda Webster at (604) 736-6133 or Dr. Susan Barr at (604) 822-2502.^ hereby voluntarily consent to participate in thenutrition study which includes completing a questionnaire and having my body fat measuredby skinfolds (calipers) and BIA. I am aware that I may withdraw from the study at any timewithout jeopardy to participation in my sport. I have read and understood the contents ofthis form and acknowledge that I have received a copy of this consent form.Signature of athlete ^Parental/Guardian consent if athlete is under 18 years of age:I (check one) ^ consent to my child's participation in this study.do not consent to my child's participation in this study.Signature of Parent/Guardian^ Date125FOOD HABITS OF TEENAGE ATHLETESThis study is being conducted by Brenda Webster a graduate student and Dr. Susan Barr ofU.B.C.'s School of Family & Nutritional Sciences (Phone 604 228-2502), to learn more aboutfood habits of teenage athletes. Everyone knows that the teen years are important ones forbuilding good nutrition habits to last a lifetime. However, we need to know more about whatteenage athletes think about their food and nutrition, and about how their busy lives andtraining affect what they eat. We also know that body size and shape are important issues foryour sport, and need to know more about how food habits are affected by these and otherconcerns.Your participation in this study will be of great assistance in obtaining the information that isneeded about teenage athletes' nutrition. There are no right or wrong answers to thequestions that follow - we simply want to know what your opinions are. Completing thequestionnaire will require about 30 minutes of your time. The results are completelyconfidential since only the investigators will see the results.If you complete the questionnaire, it will be assumed that you have agreed to participate in thestudy. If you have any questions at any time, please feel free to ask them. Participation isvoluntary, and you are free to refuse to participate or withdraw at any time without affectingyour participation in your sport.Thank you in advance for your valuable opinions.WHAT YOU EATFor each meal or snack listed below, please fill in how many times you eat the meal or snackper week. In the second column, fill in how many times per week you eat that meal or snack"out" (at a restaurant or fast food place).EXAMPLE: Eric eats breakfast on weekdays, but doesn't eat it on weekends. He never eats"out" for breakfast. He eats dinner every day, and eats this meal out with friends or familyabout twice a week.Times eaten/week^Times eaten "out"/weekBreakfast ^ 5^ 0 Dinner  7 2 YOUR MEALS AND SNACKSTimes eaten/week^Times eaten "out"/weekBreakfast ^Morning snackLunch^Afternoon snack ^Dinner^Evening Snack^Next, we'd like to know about some of the foods you eat. For each food listed on the nextpage, please fill in how often you usually eat a portion of the size stated. If you eat the food:- every day or more than once a day, fill in how many times you have it per day.- less than once a day but more than once a week, fill in the times per week.- less than once a week, but more than once a month, fill in the times per month.- less often than once a month, or never eat it, put an "X" under "do not eat".EXAMPLE: Jason has a glass of orange juice every morning, along with two slices of whitetoast. He usually has two sandwiches on brown bread at lunch, and eats french fries aboutthree times a week. He almost never eats cauliflower.Per day Per week Per month Don't eat Orange juice, 1 cup^ 1 French fries, regular serving^ 3 Cauliflower, 1/2 cup (125 ml) ^ X Bread or toast, 1 slicewhite^ 2 brown  4NUMBER OF TIMES I EAT THE FOODPer day^Per week Per month Don't eatBread or toast, 1 slice or 1 rollwhite^brown Muffin, 1 large^Pizza, 1 medium slice^Cheeseburger^Cheese - 1 slice processed OR 1 piecehard cheese (plain or in sandwich) ^Broccoli, 1/2 cup (125 ml)^Gai-lan (Chinese broccoli), 1/2 cup ^Bok-choi (Chinese cabbage), 1/2 cup ^Ice cream (large scoop) ^Frozen yogurt (large scoop) ^127Per day^Per week Per month Don't eatFast food milkshake^Cottage cheese, 1/2 cup (125 ml) ^Yogurt, small (175 ml) cartonor equivalent ^Canned salmon or sardines with bones1/2 small can^Soft drink, regular, 1 can orlarge glass^Soft drink, diet, 1 can orlarge glass^Coffee or tea, 1 cup^Tofu, 2 oz (60 gm)Milk on cereal^Orange juice, 1 cup ^Milk (any type including chocolate)1 cup^Macaroni & cheese, 1 cup (250 ml) ^I usually drink (choose one only)^ homo (whole) milk OR2% milk OR1% milk OR^ skim milk OR^ chocolate milkno milk128INFORMATION ABOUT YOUPlease fill in the following information about yourself.1. Age: ^ Years2. Sex: MaleFemaleFEMALES ONLY 3. Have you ever had a menstrual period? _noyes4. If YES, approximately how many menstrual cycleshave you had in the past 12 months? ^cyclesMALES ONLY^5. Has the pitch of your voice changed? ^ no^ yes6. Have you started to grow hair on your face?EVERYONE 7. Racial Origin: (check one or two)^Caucasian (white)Oriental^East IndianBlack^Native Canadian (Indian or Inuit)^Other (specify:^Don't know8. Are you allergic to milk or dairy products?^ no^ yes9. Do you work at a paying job?no (Go to question 10)yes - If so, where do you work?^fast food restaurantretail sales^paper route or babysitting^other (^- How many hours per week do you work?^10. What is your mother's job? ^11. What is your father's job? 12. Do you use any vitamin/mineral pills every day? ^noyesIf yes, check all that you use every day:^ multivitamin/mineral pill vitamin C pill^ iron pill calcium pill^ other ^noyes129HOW YOU FEEL ABOUT EATINGIn this section, we are interested in how you feel about eating various foods, and howyou eat. For each of the following statements, put an X in the column that bestdescribes how often you feel or act the way that is described in the statement.EXAMPLE: Erica sometimes feels guilty after eating, and usually knows the caloriecontent of the foods she eats.Very^Some-Always Often Often^times^Rarely NeverFeel extremely guilty after eating ^ X Aware of the calorie content offoods that I eat ^XVery^Some-Always Often^Often^times^Rarely NeverAm terrified about being overweight ^Aware of the calorie content offoods that I eat^Particularly avoid foods with ahigh carbohydrate content (eg.bread, rice, potatoes, etc.) ^Feel extremely guilty after eating^Am preoccupied (think a lot about) with adesire to be thinner ^Think about burning up calorieswhen I exerciseAm preoccupied (think a lot about) with thethought of having fat on my body^Avoid foods with sugar in themEat diet foods or drinks ^Feel uncomfortable after eating sweets ^Go on diets to lose weight ^Like my stomach to be emptyEnjoy trying rich new foods ^YOUR THOUGHTS ABOUT MILKBelow are some statements about how milk might be a part of your life. For each state-ment, please place an "X" under the word or phrase which best describes how much youagree or disagree with the statement. Please respond to each statement.EXAMPLE: Jennifer does not enjoy walking in the rain.STRONGLY^ STRONGLYAGREE AGREE UNSURE DISAGREE DISAGREEI enjoy walking in the rain^STRONGLY^ STRONGLYAGREE AGREE UNSURE DISAGREE DISAGREE1. Most teens and adults in my family drinkmilk as part of a snack or with meals. ^2. I have a lot of friends who drink milk^3. Advertisements for milk catch myattention. ^4. My doctor recommends that I drink milkof some kind. ^5. Drinking milk makes me feel part of aspecial group of people. ^6. A meal with my friends usually doesn'tinclude milk. ^7. The advertisements I see for milk makeme want to drink it ^8. I often hear health experts recommenddrinking milk^9. When I drink milk I feel I get approval fromthe people who matter to me^10. Very few adults in my family use milkon a regular basis^11. I hardly ever pay attention toadvertisements for milk^12. I have heard nutritionists recommendthat people of my age drink milk ^13. My family feels that drinking milk isan important part of the diet for teens^131STRONGLY^ STRONGLYAGREE AGREE UNSURE DISAGREE DISAGREE14. My friends seem to feel that it'simportant to drink milk^15. My doctor has shown no concern aboutwhether I drink milk^16. It gives me a nice feeling to havea glass of milk with friends^17. It is unusual for adults in my familyto drink milk^18. My friends think milk is drunk only by youngchildren and not by teens or adults ^19. I think advertisements for milk and dairyproducts are entertaining^20. When I drink milk, I please people who areimportant to me^Now, would you please indicate about how often you have seen certain people drink milk.Circle phrase which describes how often you see him or her drink a glass of milk ofany kind.Your mother...VERY FAIRLY NOT VERY NEVER OROFTEN OFTEN OFTEN ALMOST NEVERYour father...VERY FAIRLY NOT VERY NEVER OROFTEN OFTEN OFTEN ALMOST NEVERAnother adult in your family whom you like or admire (for example, aunt or grandparent)..VERY FAIRLY NOT VERY NEVER OROFTEN OFTEN OFTEN ALMOST NEVERA friend of the opposite sex...VERY FAIRLY NOT VERY NEVER OROFTEN OFTEN OFTEN ALMOST NEVERA friend you admire...VERY FAIRLY NOT VERY NEVER OROFTEN OFTEN OFTEN ALMOST NEVER132HOW MILK AND DAIRY PRODUCTS TASTEFinally, we are interested in how milk and dairy products taste to you. Please circlethe phrase which best describes how you feel about the taste of each of the foodslisted below. If you have never tasted the food, circle "DON'T KNOW".1. How much do you enjoy the taste of whole (homo) milk?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW2. How much do you enjoy the taste of low-fat (2% or 1%) milk?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW3. How much do you enjoy the taste of skim milk?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW4. How much do you enjoy the taste of fruit-flavoured yogurt?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW5. How much do you enjoy the taste of plain yogurt?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW6. How much do you enjoy the taste of cottage cheese?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW7. How much do you enjoy the taste of hard cheese (such as cheddar)?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW8. How much do you enjoy the taste of ice cream?VERY^IT'S^JUST^NOT VERY NOT AT^DON'TMUCH^O.K. "SO-SO" MUCH^ALL KNOW133ABOUT YOUR TRAININGPlease answer the following questions about your training. Include bothteam practices and any other training sessions.1. How many hours per session do you train?(hours).2. How many training sessions do you have per week?sessions/week.3. How many months per year do you train for your sport?^months/year.4. Have you ever participated in the following competitions? Please circle"YES" OR "NO":a) City Championships^YES^NOb) Provincial Championships^YES^NOc) National Championships YES^NOd) International Championships^YES^NO(ie. Pan-Americans, North Americans,World Championships, Olympic Games)NAME:^Would you like a copy of your results?If yes, please provide a complete address: YES^NOTHANK YOU VERY MUCH FOR YOUR HELP!134Appendix C:^Body Composition Equations and Constants1.^Skinfold Predictive Equations:The quadratic equation of Jackson and Pollock (1980) used to predict body densityof female athletes as suggested by Thorland et al.(1984) is:Body Density = 1.1454464 - 0.0006558(X, + X5 + X6 X7) 0.0000015(X, +X5^X6^X7) - 0.0000604(X9 ) - 0.0005981(X 16 )and the linear equation of Forsyth and Sinning for males (1973) is:Body Density = 1.10647 - 0.00162(X 2) - 0.00144(X 6 ) - 0.00077(X 1 ) + 0.00071 (X3 ).Where:X 1 = triceps skinfold in mmX2 = scapular skinfold in mmX3 = midaxillary skinfold in mmX5 = supra-iliac skinfold in mmX6 = abdomen skinfold in mmX, = thigh skinfold in mmX9 = age in yearsX15 = hip circumference in cm.Percent body fat is then determined using Lohman's formula (1986) for youths andhis age and sex specific constants (see Table for the constants).% Body Fat =^[ 1 (d 1 d 2 - d 2 ] X 100[ Db (d 1 d 2 ) d 1 - d 2Where^Db = body density in g/mld 1 = density of age and sex specific FFB in g/mld 2 = density of fat (0.90 g/ml)135Table 38:^Age and Sex Specific Constants for Conversion of Body Density, Water,and Potassium to percent Fat in Children and YouthsFat Free Body Composition Density FFBg/ml Water % FFB' K FFBg/kia 2 Mineral %FFB 3 Age Male Female Male Female Male Female Male Female7-9 1.081 1.079 76.8 77.6 2.40 2.32 5.1 4.99-11 1.084 1.082 76.2 77.0 2.45 2.34 5.4 5.211-13 1.087 1.086 75.4 76.6 2.52 2.36 5.7 5.513-15 1.094 1.092 74.7 75.5 2.56 2.38 6.2 5.915-17 1.096 1.094 74.2 75.0 2.61 2.4 6.5 6.117-20 1.099 1.095 74.0 74.8 2.63 2.41 6.6 6.020-25 1.100 1.096 73.8 74.5 2.66 2.42 6.8 6.21 % Water FFB-determined using isotope dilution and the Siri Water equation.2K FFB in g/kg-indirectly calculated using 4°K, body density, TBW %FFB.3Mineral %FFB-determined using single photon absorptiometry of the radius toestimate total body bone mineral.Adapted from: Lohman TG. Applicability of body composition techniques andconstants for children and youths. Exerc Sport Sci Rev 1986;14:325-357.136Question1. I have a lot of friendswho drink milk.2. A meal with my friendsusually doesn't includemilk. 23. My friends seem tofeel that it's important todrink milk.4. My friends think milk isdrunk only by youngchildren and not by teensor adults. 2Overall Sub-Scale ScoreSpeed Skatersn =32Gymnastsn =25All Athletesn =573.5 ± 0.7 3.6 ± 0.8 3.5 ± 0.83.1 ± 1.0 2.6^1.1 2.9±1.13.0 ± 0.7 2.9^1.0 2.9 ± 0.93.8 ± 1.0 4.0 ±0.8 3.9±0.93.3 ± 0.6 3.3±0.5 3.3±0.6Appendix D:^Differential Association Sub-Scale ScoresTable 39:^Scores for Friends Sub-Scale Questions and Overall Sub-Scale Score forMales (means and standard deviations► 1'Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5)2Scoring was reversed for negatively worded questions137Speed Skatersn =25Gymnastsn =32All Athletesn =573.9 ± 1.0 3.7±0.7 3.8 ±0.83.0 ± 1.3 3.3 ± 0.9 3.2 ± 1.13.0 ±1.1 3.1 ±0.7 3.1 ±0.94.0 ± 1.0 3.8 ± 0.5 3.9 ± 0.73.5 ± 0.7 3.5±0.5 3.5±0.6Question1. I have a lot of friendswho drink milk.2. A meal with my friendsusually doesn't includemilk. 23. My friends seem tofeel that it's important todrink milk.4. My friends think milk isdrunk only by youngchildren and not by teensor adults. 2Overall Sub-Scale ScoreTable 40:^Scores for Friends Sub-Scale Questions and Overall Sub-Scale Score forFemales (means and standard deviations)''Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5).2Scoring was reversed for negatively worded questions.138Speed Skaters Gymnasts All Athletesn=32 n=25 n =574.0±0.8 3.5±1.2 3.8_1.03.7±1.2 3.6±1.0 3.6±1.14.0 ± 0.9 3.8 ± 1.0 3.9±1.03.6±1.2 3.6±1.0 3.6±1.13.8 ±0.7 3.6±0.8 3.7±0.7Question1. Most teens and adultsin my family drink milk aspart or a snack of withmeals.2. Very few adults in myfamily use milk an aregular basis. 23. My family feels thatdrinking milk is animportant part of the dietfor teens.4. It is unusual for adultsin my family to drinkmilk. 2Overall Sub-Scale ScoreTable 41:^Scores for Family Sub-Scale Questions and Overall Sub-Scale Score forMales (means and standard deviations/ 1'Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5).2Scoring was reversed for negatively worded questions.139Question1. Most teens and adultsin my family drink milk aspart of a snack or withmeals.2. Very few adults in myfamily use milk an aregular basis. 23. My family feels thatdrinking milk is animportant part of the dietfor teens.4. It is unusual for adultsin my family to drinkmilk. 2Overall Sub-Scale ScoreSpeed Skatersn=25Gymnastsn =32All Athletesn=573.9±1.2 4.0 ±0.8 3.9 ± 1.03.9 ± 1.3 3.7±1.1 3.8±1.23.8 ± 0.9 4.3 ± 0.7 4.1±0.83.5±1.4 4.0 ± 1.0 3.8±1.23.8 ± 1.0 3.9±0.7 3.9±0.8Table 42:^Scores for Family Sub-Scale Questions and Overall Sub-Scale Score forFemales (means and standard deviations)''Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5).2Scoring was reversed for negatively worded questions.140Question^ Speed Skaters^Gymnasts^All Athletesn =32^n=25 n =573.0 ±1.2 3.0±1.1 3.0 ±1.12.9 ±1.1 3.1 ±1.2 3.0 ±1.12.9±1.0 3.1 ±0.7 3.0±0.91. Advertisements for^3.1 ± 1.0^3.4 ±1.0^3.2±1.0milk catch my attention.2. The advertisements I^2.7±1.1^2.6 ± 1.0^2.7±1.0see for milk make mewant to drink it.3. I hardly ever payattention toadvertisements for milk. 24. I think advertisementsfor milk and dairyproducts are entertaining.Overall Sub-Scale ScoreTable 43:^Scores for Media Sub-Scale Questions for Males and Overall Sub-ScaleScore (means and standard deviations)''Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5).2Scoring was reversed for negatively worded questions.141Speed Skatersn =25Gymnastsn =32All Athletesn =573.8 ± 1.1 3.0 ± 1.0 3.3 ± 1.13.0 ± 1.3 2.5 ± 1.0 2.7 ± 1.23.4 ± 1.1 3.2 ±0.9 3.3 ± 1.03.2 ± 1.1 3.1 ± 1.0 3.1 ± 1.13.3 ± 0.8 3.0±0.7 3.1 ±0.8Question1. Advertisements formilk catch my attention.2. The advertisements Isee for milk make mewant to drink it.3. I hardly ever payattention toadvertisements for milk. 24. I think advertisementsfor milk and dairyproducts are entertaining.Overall Sub-Scale ScoreTable 44:^Scores for Media Sub-Scale Questions for Females and Overall Sub-ScaleScore (means and standard deviations► 1'Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5).'Scoring was reversed for negatively worded questions.142Speed Skatersn =32Gymnastsn =25All Athletesn =572.9±1.1 2.9±1.1 2.9±1.13.8±0.9 3.9±0.8 3.8 ± 0.93.9 ± 1.0 3.7± 1.0 3.8 ± 1.03.0 ± 0.8 3.0±1.1 3.0±0.93.4 ± 0.6 3.4±0.7 3.4±0.6Question1. My doctorrecommends that I drinkmilk of some kind.2. I often hear healthexperts recommenddrinking milk.3. I have heardnutritionists recommendthat people of my agedrink milk.4. My doctor has shownno concern aboutwhether I drink milk. 2Overall Sub-Scale ScoreTable 45:^Scores for Experts Sub-Scale Questions for Males and Overall Sub-ScaleScore (means and standard deviations)''Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5).2Scoring was reversed for negatively worded questions.143Speed Skatersn =25Gymnastsn =32All Athletesn =572.9±1.2 2.8 ±0.9 2.8±1.03.9±0.9 3.7±1.0 3.8 ±0.93.8±1.1 3.7±0.9 3.7±1.02.9±1.2 3.1 ±0.9 3.0±1.13.4±0.7 3.3 ± 0.6 3.3±0.7Question1. My doctorrecommends that I drinkmilk of some kind.2. I often hear healthexperts recommenddrinking milk.3. I have heardnutritionists recommendthat people of my agedrink milk.4. My doctor has shownno concern aboutwhether I drink milk. 2Overall Sub-Scale ScoreTable 46:^Scores for Experts Sub-Scale Questions for Females and Overall Sub-Scale Score (means and standard deviations)''Participants answered questions using a five point scale ranging from stronglydisagree (scored 1) to strongly agree (scored 5).2Scoring was reversed for negatively worded questions.144

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086178/manifest

Comment

Related Items