The increasing prevalence of overweight and obesity in childhood and adolescence is a global health issue because of the associated health problems in childhood and the conjecture that these conditions may be antecedents of adult diseases. Conditions previously associated with adulthood such as hypertension, heart disease, diabetes, respiratory disorders, and psychological and social complications are now presenting in childhood (24). Whereas a predisposition to obesity may be genetic, environmental factors leading to a positive energy balance are likely contributors, especially during the pubertal years. Two lifestyle behaviors suggested to be associated with an excess positive energy balance leading to overweight and/or obesity are low levels of physical activity (10,21,22) and high levels of sugar-sweetened drink consumption (16,20).
Accurately assessing the independent impact of lifestyle behaviors on fat development in adolescence is problematic because normal growth-related body composition changes may mask or be greater than environmental or lifestyle effects (28). To understand the causes of excessive fat mass (FM) accumulation in adolescence, an individual's pattern of fat development over the growing years must be known. This pattern can only be identified using longitudinally gathered data, where individual growth trajectories are assessed. Longitudinal studies of human somatic growth have demonstrated clearly that the pattern of linear growth is consistent but that the timing and magnitude of growth is highly variable among individuals. Girls, on average, mature 2 yr earlier than boys; thus, biological maturity age rather than chronological age must be used when comparing growth-related factors between genders. This same principle holds for comparisons within gender, as there also is a wide variation in biological maturity within genders of the same chronological age (26). The most commonly used biological maturity milestone in growth studies is age at peak height velocity (PHV) (19,26).
This study constructed growth curves for FM by aligning all subjects on a comparable biological maturity age index (PHV). We have demonstrated that this approach significantly alters the interpretation of both physiological (4) and behavioral parameters (27).
The aim of the study was to examine the relationship of physical activity and sugar-sweetened drink intake on the development of total body fat mass. We hypothesized that when size, biological maturity age, and their interactions were accounted for, both physical activity and sugar-sweetened drink intake would significantly predict fat accumulation. The uniqueness of the present study is that serial measures of physical activity, sugar-sweetened drink consumption and body fat deposition were observed in a group of children for seven consecutive years during the adolescent growth period.
Participants were part of the University of Saskatchewan's Pediatric Bone Mineral Accrual Study (PBMAS), which has been described in detail elsewhere (1). In brief, the study utilized a mixed longitudinal design and incorporated eight age cohorts. The cohorts were aged between 8 and 15 yr at baseline. During the 7 yr of annual data collection, the composition of these clusters remained the same. Because there were overlaps in ages between the clusters, it was possible to estimate a consecutive 11-yr developmental pattern (8-19 yr) over the shorter period of 7 yr.
Eligible children had no history of chronic disease or long-term medication use. The study was approved by the university and hospital advisory committee on ethics in human experimentation. In 1991, written informed consent was obtained from 228 parents and their children (113 boys and 115 girls). Of the 94% of participants whose race was known, 95% were white.
A subsample of the PBMAS data set was selected using inclusion criteria that included complete measures of physical activity and sugar-sweetened drink consumption, as well as body composition assessment at each measurement occasion. Age at PHV, calculated from velocity of stature growth, was also required to align individuals on a biological age parameter. There were 105 males and 112 females who met these requirements and were included in the present analysis (Table 1). The median number of visits was five.
Chronological age was calculated by subtracting the date of birth from the date of measurement. Age groups were formed using 1-yr intervals from the midpoints of the age; for example, the 10-yr age group included subjects from 9.50 to 10.49 yr of age.
Standing height was measured biannually as stretch stature, using a wall-mounted stadiometer, and recorded to the nearest 0.1 cm.
Repeated measures of height over the 7-yr span of the study allowed determination of PHV. PHV was identified from individual growth curves using a cubic spline procedure (GraphPad Prism version 3.00 for Windows, GraphPad Software, San Diego, CA) (19).
Biological maturity age.
Biological age was determined by subtracting the chronological age of an individual at the time of measurement from the chronological age at PHV. Biological maturity age categories were based on 1-yr intervals; for example, −1 yr from PHV maturity ages included observations between −1.49 and −0.50 yr from PHV (2).
Participants were scanned annually during October or November by dual energy x-ray absorptiometry (DXA) (Hologic 2000 QDR, Hologic, Waltham, MA) in the array mode, employing enhanced global software version 7.10. Whole-body scans were analyzed using software version 5.67A. All scans were analyzed by one of two qualified individuals to minimize operator-related variability. Standard procedures and positioning were adopted and strictly followed to maximize consistency (14). Participants removed metal objects (e.g., jewelry, glasses) and shoes and wore loose-fitting shorts and a T-shirt during the scanning procedure (1). Data on fat-free mass (FFM) (lean plus bone mineral content) and FM were recorded in grams. Relative FFM (%) and relative FM (%) were also determined for descriptive purposes to provide a general sense of the relationship during growth. Short-term precision in vivo for FM, expressed as a coefficient of variation (%), was 2.95% for the whole body.
Nutrition information was gathered via a self-reported, 24-h recall, administered three times per year in each of the first 3 yr and then twice yearly each year thereafter. Yearly averages were then calculated and used for the analysis. A 20-min training session on food portion sizes was given to each participant prior to performing the initial data gathering. At each subsequent recall session, display boards with life-size pictures of food and portion sizes were present for participants' reference. Nutrient composition based on the 1988 Canadian nutrient file was used to analyze the 24-h recalls. The recalls were coded and the same individual checked all forms to ensure consistent interpretation (31). Average total energy intake per day was calculated (kcal·d−1). As intakes varied considerably due to activity and body size, as well as errors in reporting intakes, calories (kcal·d−1) from sugar-sweetened drink intake were removed from the total energy intake value. The purpose of this adjustment was to isolate the influence of sugar-sweetened drink intake in the analysis and also control for participant's energy consumption apart from sweetened drinks.
Consumption of sugar-sweetened beverages was assessed at each measurement occasion (31). Of interest were beverages that contained added sugar such as soft drinks, sport drinks, drinks made from crystals or flavored syrup, punches less than 50% real juice, milkshakes, liquid yogurt, hot chocolate, and iced tea. Focus was on sugar-sweetened drinks outside of 100% fruit juice because such drinks are regarded as a source of "empty calories" lacking significant amounts of other beneficial nutrients. Sugar-sweetened drinks were assumed to contain 12 kcal·oz−1 (15). Using this value, the energy contained in the sugar-sweetened drink total was estimated and subtracted from the recorded total energy intake.
Physical activity level was assessed a minimum of three times per year for the first 2 yr of the study and twice yearly thereafter using the Physical Activity Questionnaire for Children (PAQ-C) and/or Adolescents (PAQ-A) (2). This instrument scores nine items on a 5-point scale, with a higher value indicating higher levels of physical activity. A mean composite score of activity was calculated from all measurement occasions. The PAQ-C questionnaire has consistently demonstrated good internal consistency and validity with moderate relationships to teacher evaluations of activity, Caltrac motion sensors, 7-d activity recalls, step tests of fitness, and leisure-time activity scales (7). One of the limitations of the questionnaire is that energy expenditure values cannot be derived from it.
Descriptive results are expressed as mean ± SEM (SPSS version 11.5, SPSS Inc., Chicago, IL). Age group comparisons were made with t-tests (P < 0.05), and Bonferroni adjustments were made for multiple comparisons. The hypotheses were tested using hierarchical (multilevel) linear modeling using random effects models (MlwiN version 1.0, Multilevel Models Project, Institute of Education, University of London, London, UK). This procedure has been described previously (11).
Additive and gender-specific multilevel regression models were developed to describe the developmental changes in FM (kg) as follows:
where y is the FM on the measurement occasion i in the jth individual, αj is the constant for the jth individual, βjxij is the slope of fat mass with biological age (years from PHV) for the jth individual, and k1 to kn are the coefficients of various explanatory variables (e.g., physical activity (PAQ score 1-5), sugar-sweetened drink (oz·d−1), adjusted total energy expenditure (kcal·d−1)) at assessment occasion i in the jth individual, and εij is the level 1 residual (within individual variance) for the ith assessment of FM in the jth individual.
Models were built in a stepwise procedure; that is, predictor variables (κ-fixed effects) were added one at a time. Likelihood ratio statistics were used to determine whether the effects of independent variables were significant contributors to the model. The difference in likelihoods between two models follows a chi-square distribution; this difference was compared against the degrees of freedom lost to determine whether one model was a significant improvement over the other. In this way, predictor variables were added to the models and retained if deviance improved and/or if the variances at levels 1 and 2 were reduced. Predictor variables (κ) were accepted as significant if the estimated mean coefficient was greater than twice the SEE (P < 0.05). If the retention criteria were not met, the predictor variable was discarded. Biological age was added as both a fixed and random coefficient. To allow for the nonlinearity of growth, age power functions (biological age2 and biological age3) were added to the models as fixed effects. Once age, FFM, and total energy intake were modeled, physical activity and sugar-sweetened drink consumption and their interaction were incorporated into the models and their independent effects were tested.
Height and body mass were within normal reference standards ranges for all chronological ages in both genders (Figs. 1 and 2) (29). Males were older (P < 0.05), taller (P < 0.05), and heavier (P < 0.05) than females when compared across biological maturity categories. Figure 3 shows the gender-specific developmental patterns of FFM [(a) absolute (kg) and (b) percentage of total body mass (%)] and FM [(c) absolute (kg) and (d) percentage of total body mass (%)] of males and females by maturity age. Absolute FFM and FM increase with increasing age in both genders; however, the percentage values of FM and FFM relative to total body mass show gender differences. At PHV, males increase the percentage of FFM and decrease the percentage of FM; the reverse pattern is seen in females. Males had greater absolute and percentage of FFM than females at each biological maturity category (P < 0.05). Males had lower percentages of FM than females at all maturity ages (Fig. 3d); however, no significant gender difference (P > 0.05) in absolute FM was observed until after PHV had been attained (Fig. 3c).
Average total energy intake, physical activity, and sugar-sweetened drink consumption values are shown for males and females at each biological maturity category (relative to PHV) in Figure 4. No significant gender differences in total energy intake were found prior to PHV (P > 0.05). Post-PHV, total energy intake differences became significant (P < 0.05), as intake decreased in females and increased in males. No significant gender differences were found at any biological maturity category for physical activity (P > 0.05). Sugar-sweetened drink consumption appears to increase with increasing biological maturity; however, significant (P < 0.05) gender differences were only found at 2 and 3 yr post-PHV.
Cross-sectional data analyses are problematic because they do not account for the individual time-dependent confounding variables or the individual differences in tempo and timing of size and maturity between males and females. Because gender differences are particularly prominent at the ages under study, gender-specific models were built.
Tables 2 and 3 summarize the results from the multilevel models for FM development in females and males, respectively. The random effects coefficients describe the two levels of variance (within individuals (level 1 of the hierarchy) and between individuals (level 2 of the hierarchy)). For all four gender specific models, the significant variances at level 1 of the models indicate that FM was increasing significantly at each measurement occasion within individuals (E > 2*SEE; P < 0.05). The between-individuals variance matrix (level 2) for each model indicates that individuals had significantly different FM curves both in terms of their intercepts (constant/constant, P < 0.05) and the slopes of their lines (biological age/biological age, P < 0.05). The variance of these intercepts and slopes was positively and significantly correlated (constant/biological age, P < 0.05). The variance among individuals was therefore related to their biological age.
The fixed effects (Tables 2 and 3) give the estimates for the coefficients of the various explanatory variables. Once biological age (years from PHV) and FFM effects were controlled, there was no significant (P > 0.05) independent physical activity effect in females (Table 2, PA model), but in males, there was a significant physical activity effect (P < 0.05) (Table 3, PA model). This physical activity effect was at the individual level. For example, at PHV, males who had a physical activity score of 5 had 2.18 kg less fat than males of the same biological age and FFM, who had a physical activity score of 1. Once biological age, FFM, and total energy intake were controlled, there was no significant effect (P > 0.05) of sugar-sweetened drink consumption in either females or males (Tables 2 and 3, SD models). In addition, no significant (P > 0.05) physical activity by sugar-sweetened drink interaction effects was observed (Tables 2 and 3, interaction models). To illustrate the significant effects of physical activity, the interaction models (Tables 2 and 3) were used to predict the average growth curve for a male/female (A) (Fig. 5) who had an activity score of 4.5 and a sugar-drink intake of 0 oz·d−1 and who had an average total energy intake and FFM at each biological age (Figs. 3 and 4); compared with a male/female (B) who had an activity score of 0.5 and a sugar-drink intake of 96 oz·d−1, and who had average total energy intake and FFM at each biological age (Figs. 3 and 4). Both scenarios reflect the extremes of the observed distribution of these factors and serve to illustrate their influences on FM growth in adolescence. The significant difference between the two males is due to differences in physical activity levels; no significant differences are shown between the two females.
The purpose of this study was to investigate the relationship of physical activity and sugar-sweetened beverage consumption with FM accumulation over the growing years, in both males and females. The results indicated that physical activity had a significant negative relationship on FM development in males, but not females. Sugar-sweetened drink intake was not significantly related to FM development in either males or females. There were no interactions found between physical activity and sweetened drink consumption in either gender.
The uniqueness of the study lies in its longitudinal design. With longitudinal data sets, a broad picture of development can be provided and time-dependent relationships can be investigated. Longitudinal data also allow researchers to control for individuals' unique patterns of growth and maturation (26). If individuals are not aligned on a biological maturity age, the individual effects of the parameters of interest (physical activity and sugar-sweetened drink intake) may be masked by the maturity-related effects.
Male and female average heights and weights were similar to reference norms, suggesting that the participants had growth trajectories compatible with normal healthy growing children. Although patterns for total body mass appear normal, the interpretation of body composition data is problematic. At present, there are no reference data available that capture the range of variability in body compartments relative to gender, age, and maturation. Typically, the percentage of body fat is used as a single indicator of body composition, with the assumption that it adequately adjusts for body size. This approach assumes that the relationship of FM to body weight is linear; however, as indicated in Figures 2 and 3, this is not the case in growing children. The use of the percentage of body fat can result in an overestimation or underestimation of normal ranges; thus, regression techniques using absolute values, as used in the present study, are a more appropriate statistical approach. We found in our sample that average FM (kg) development curves reflected the description of FM growth by previous researchers: steadily increasing throughout adolescence in females, and in males demonstrating a plateau in late adolescence (30).
Physical activity was significantly and negatively related to FM in males. This result supports the theory that male adolescents can reduce their FM, or risk of excess fat accumulation, by increasing physical activity (10,21,22). Physical activity level is the single largest modifiable behavior for individuals to increase their energy expenditure. The result for females is consistent with the research of Garaulet et al. (10), who found physical activity to be negatively correlated with overweight in males 14-18 yr old, but not related in females. The lack of an inverse relationship between physical activity and FM in females may be a function of natural female maturation (30) and of normal levels of physical activity not being adequate enough to affect their FM development. This finding is unlikely to be a function of the fact that females were less physically active than the males because the analysis revealed that when the genders were aligned by maturational age, no significant difference existed in their activity levels (Fig. 4b) (27).
Both genders had similar increases in the intake of sugar-sweetened drinks with age and maturity, although the males' intake was significantly greater at 2 and 3 yr post-PHV. No significant relationship between sugar-sweetened drink consumption and FM was found after controlling for dietary energy intake. This finding is contrary to earlier papers that have suggested that an increasing level of sugar-sweetened drink consumption is, in part, responsible for increasing levels of excess FM in adolescents (16,18,20). There are two main rationalizations for this implication. One suggestion is that youths have poor compensatory reduction in caloric intake after sugar drink consumption, leading to positive caloric balance and FM gain over time. Poor compensatory reduction would be due to the easily digestible nature of the drinks, which are both fluid and composed of simple carbohydrates. The second idea is that fat absorption and accumulation is facilitated via the insulin response induced by sugar-sweetened drink intake. Beverages are frequently consumed with meals, but their influence on meal energy intake is poorly understood. It has been suggested, however, that they add to total energy intake without significantly affecting satiety ratings (17). For this reason, we included in our models total energy intake adjusted for sugar-sweetened drinks so that we could investigate the independent effects of sugar-added beverages rather than total sugar intake. The findings of the current study did not support either proposition and are in agreement with conclusions of one previous cross-sectional study that found no significant relationship between carbonated and fruit beverage intake and the body mass index of males and females 12-16 yr old (9).
There are several possible explanations for not finding a significant relationship between sugar-sweetened drink consumption and FM accumulation in the present sample. It might be that there is a threshold level of energy intake and/or sugar-sweetened drink consumption required to influence FM, but to our knowledge this has not yet been investigated. Perhaps an inadequate number of participants in the current sample had a high enough level of energy or soft drink consumption to reach such a threshold. The energy intakes of this sample were lower in all age groups and both genders when compared with Canadian averages. For example, the 15-yr old males and females had average caloric intakes of 2220 and 1540 kcal, respectively, whereas the Canadian average requirements for this age group are estimated at 3013 and 2362 kcal, respectively, for active males and females (15). This suggests that the current sample's energy intake was below a threshold that would influence FM, even when including sugar-sweetened drink intake. It is even possible that in the current sample, the sugar-sweetened drink intake was beneficial in resisting FM accumulation by contributing to energy balance and maintenance of metabolic body mass.
It has been suggested that energy requirements of the growth process account for up to 4% of energy requirements in adolescents (6). This requirement could translate into approximately 100-120 kcal·d−1 during adolescence and therefore be large enough to ward off increases in FM due to minor excesses in energy intake. Coincidentally, an average 12-oz serving of sugar-sweetened drink contains approximately the equivalent amount of energy (120 kcal) (8). Perhaps only following completion of maturation and attainment of adult growth status would daily consumption of sugar-sweetened drinks adversely affect body composition.
The results of the study are also limited by some aspects of the methodology. DXA has been determined to be a valid and reliable instrument for body composition assessment (13); however, its ability to assess body composition in children and adolescents has been questioned (5,23). DXA employs assumptions that mineral and water content of the fat-free body are constant, but such assumptions may not hold true in growing individuals (5,23). There is not complete agreement on this issue, however; for instance, Svendsen et al. (25) demonstrated that DXA was an accurate method for measuring soft tissue in pigs varying in weight from 35 to 95 kg, a range in which the majority of PBMAS participants would fall.
The PAQ-C/A questionnaire used to assess physical activity has demonstrated good internal consistency and validity with several other evaluations of activity level (2,7); however, it is a self-reported assessment and therefore has the associated limitations. The food recall procedure is thought to offer the best method of obtaining a dietary record (12), although it is also a self-reported measure and has associated limitations. Self-reported measures are susceptible to underreporting, especially among those who are overweight or obese. A recent longitudinal study also revealed that as 10- to 15-yr-old females aged, they reported their energy intake less accurately (no data on males) (3). The limited frequency of dietary assessment might have served to exaggerate the weaknesses of the 24-h recall if underreporting occurred on many or all occasions. It also may, in part, account for the low reported energy intake (see above) while growth was still occurring.
The present study provides support for the theory that physical activity during adolescence has a negative effect on FM accumulation in males. Speculation that higher intake levels of sugar-sweetened drinks also affect fat accumulation was not supported in the present study. Similar to Forshee et al. (9), we conclude that there is no substantiated benefit to body composition in restricting normal weight youths' access to sugar-sweetened drinks.
Future studies should focus on identifying the level of physical activity required for benefits in body composition and identifying whether such a threshold exists in female adolescents. Studies that generate reference standards for levels of physical activity most optimal for healthy adolescent FM development also are required. Also, future studies pertaining to sugar-sweetened drinks could approach this topic differently, perhaps by examining groups of extremely high and extremely low sugar-sweetened drink intake, analyzing the impact of specific sweetening agents (i.e., sucrose vs fructose) in beverages, or including beverages with naturally occurring sugars in analysis. Intervention studies that directly examine both sides of energy balance (intake and expenditure) are also required to better understand the influence of consumed foods and drinks on fat acquisition.
The original PBMAS study received ethical approval from the university and hospital advisory committee on ethics in human experimentation. The Canadian National Health and Research Development Program (NHRDP), the Canadian Institute of Health Research (CIHR), and the Saskatchewan Health Research Foundation (SHRF) provided funding for this study. PBMAS Group members include D. A. Bailey, A. D. G. Baxter-Jones, P. E. Crocker, K. S. Davison, D. T. Drinkwater, E. Dudzic, R. A. Faulkner, K. Kowalski, H. A. McKay, R. L. Mirwald, W. M. Wallace, and S. J. Whiting.
1. Bailey, D. The Saskatchewan Pediatric Bone Mineral Accrual Study: bone mineral acquisition during the growing years. Int. J. Sports Med.
2. Bailey, D., H. McKay, R. L. Mirwald, P. Crocker, and R. Faulkner. A six-year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children: The University of Saskatchewan Bone Mineral Accrual Study. J. Bone Miner. Res.
3. Bandini, L. G., A. Must, H. Cyr, S. Anderson, J. L. Spadano, and W. H. Dietz. Longitudinal changes in the accuracy of reported energy intake in girls 10-15 y of age. Am. J. Clin. Nutr.
4. Baxter-Jones, A. D. G., R. L. Mirwald, H. McKay, and D. Bailey. A longitudinal analysis of sex differences in bone mineral accrual in healthy 8-19 year old boys and girls. Ann. Hum. Biol.
5. Boileau, R., T. Lohman, and M. Slaughter. Review: exercise
and body composition of children and youth. Scand. J. Sports Sci.
6. Butte, N. F. Fat intake in children in relation to energy requirements. Am. J. Clin. Nutr.
7. Crocker, P., D. Bailey, R. Faulkner, K. Kowalski, and R. McGrath. Measuring general levels of physical activity: preliminary evidence for the Physical Activity Questionnaire for older children. Med. Sci. Sports Exerc.
8. Ebbeling, C., and D. Ludwig. Treating obesity in youth: should dietary glycemic load be a consideration. Adv. Pediatr
9. Forshee, R., P. Anderson, and M. Storey. The role of beverage consumption, physical activity, sedentary behavior, and demographics on body mass index of adolescents. Int. J. Food Sci. Nutr.
10. Garaulet, M., A. Martinez, F. Victoria, F. Perez-Llamas, R. Ortega, and S. Zamora. Differences in dietary intake and activity level between normal weight and overweight or obese adolescents. J. Pediatr. Gastroenterol. Nutr.
11. Goldstein, H., J. Rasbash, I. Plewis, et al. A User's Guide to MLwiN
, London: Institute of Education, University of London, pp. 1-140, 1998.
12. Goran, M., B. Gower, T. Nagy, and R. Johnson. Developmental changes in energy expenditure and physical activity in children: evidence for a decline in physical activity in girls before puberty. Pediatrics
13. Gutin, B., M. Litaker, T. Manos, C. Smith, and F. Treiber. Body composition measurement in 9-11-y-old children by dual energy X-ray absorptiometry, skinfold thickness measurements, and bioimpedance analysis. Am. J. Clin. Nutr.
14. Hologic. Hologic QDR 2000 Supine Lateral Bone Densitometer; Operator's Manual and User's Guide
. Waltham, MA: Hologic Inc., pp. 5.1-5.28, 1992.
15. Institute of Medicine. Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients)
, Washington, DC: National Academy Press, 2002, pp. 5.50-5.52.
16. Ludwig, D., K. Peterson, and S. Gortmaker. Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis. Lancet
17. Mattes, R. Dietary compensation by humans for supplemental energy provided as ethanol or carbohydrate in fluids. Physiol. Behav.
18. McGartland, C., P. Robson, L. Murray, et al. Carbonated soft drink consumption and bone mineral density in adolescence
: The Northern Ireland Younger Hearts Project. J. Bone Miner. Res.
19. Mirwald, R., and D. Bailey. Seasonal height velocity differences in boys and girls age 8.5-18. Am. J. Hum. Biol.
20. Mrdjenovic, G., and D. Levitsky. Nutritional and energetic consequences of sweetened drink consumption in 6-13-year-old children. J. Pediatr.
21. Proctor, M., L. Moore, D. Gao, et al. Television viewing and change in body fat from preschool to early adolescence
: the Framingham Children's Study. Int. J. Obes.
22. Robinson, T. Reducing children's television viewing to prevent obesity; a randomized controlled trial. JAMA
. 282:1561-1567, 1999.
23. Roemmich, J., P. Clark, A. Weltman, and A. Rogol. Alterations in growth and body composition during puberty: Comparing multicompartment body composition models. J. Appl. Physiol.
24. Styne, D. Childhood and adolescent obesity: prevalence and significance. Pediatr. Clin. North Am.
25. Svendsen, O., J. Haarbo, C. Hassager, and C. Christiansen. Accuracy of measurements of body composition by dual-energy x-ray absorptiometry in vivo. Am. J. Clin. Nutr.
26. Tanner, J. Growth at Adolescence
. Oxford: Blackwell Scientific Publications, 1962, pp. 1-27, 55-93.
27. Thompson, A., A. Baxter-Jones, R. Mirwald, and D. Bailey. Comparison of physical activity in male and female children: does maturation matter? Med. Sci. Sports Exerc.
28. Ulijaszek, S., F. Johnston, and M. Preece. The Cambridge Encyclopedia of Human Growth and Development
. Cambridge: Cambridge University Press, 1998, 34-38, 51-53, 212-219.
29. U.S. Department of Health and Human Services, Center for Disease Control and Prevention. 2000 CDC Growth Charts for the United States: Methods and Development
. Washington, DC: Department of Health and Human Services, Center for Disease Control and Prevention, Report 246, 2002.
30. van Lenthe, F., W. van Mechelen, H. Kemper, and G. Post. Behavioral variables and development of a central pattern of body fat from adolescence
into adulthood in normal-weight whites: The Amsterdam Longitudinal Growth and Health Study. Am. J. Clin. Nutr.
31. Whiting, S., A. Healey, S. Psiuk, R. Mirwald, K. Kowalski, and D. Bailey. Relationship between carbonated and other low nutrient dense beverages and bone mineral content of adolescents. Nutr. Res.
Keywords:©2006The American College of Sports Medicine
SOFT DRINKS; EXERCISE; ADOLESCENCE; ADIPOSITY; MULTILEVEL MODELING