Weight and Body Measurement Changes In College Freshmen 
 
by 
 
Mahendran Balasubramanian 
 
 
 
 
A thesis submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Master of Science 
 
Auburn, Alabama 
December 18, 2009 
 
 
 
 
Keywords: Body Measurement change, freshman15,  
      3D body scanning, symmetric/asymmetric measurement change 
 
 
 
 
Approved by 
 
Pamela V. Ulrich, Co-chair, Professor of Consumer Affairs 
Karla P. Simmons, Co-chair Assistant Professor of Consumer Affairs 
Lenda Jo Connell, Professor of Consumer Affairs  
 
  
Abstract 
A limited number of research studies suggest that new college students gain some 
weight on average during their freshman year. Most studies have had relatively small 
samples; some have used self-reported rather than measured weight. No known studies 
have followed body circumference measurements to explore where the effects of weight 
change may be observed. The purpose of this research was to quantitatively assess links 
between weight changes and body measurement alterations in entering freshmen through 
their first year.  At the beginning of fall semester 2007, a convenience sample of 240 
subjects were weighed, had body fat measured, completed a questionnaire, and had body 
measurements extracted through 3D body scanning; 205 subjects returned at the end of 
spring semester 2008. Of these, 187 (67 males and 120 females) had no missing 
measurements and were used for the analysis.  
Of the 187, 100 (43 males and 57 females) gained weight, respectively averaging 
6.93 lb and 7.21 lb; 39 (10 males and 29 females) lost weight (respective means of 6.34 
lb and 3.72 lb). The remaining subjects remained within ?2 lb of their fall weight and 
were considered a no change group.  Six trunk and two limb (biceps and thigh) 
measurements were compared. Among males and females who gained weight, seven 
measurements increased significantly; males? chest and females? thigh measurements 
increased but not significantly. Significant trunk increases ranged 1.31-1.94 in. in males 
and 0.89-1.21 in females; males? biceps and thigh increases were 0.61 and 0.79 in., and 
ii 
 
  
females? biceps increased 0.46 in.  In the weight lost groups, only seat and thigh 
measurements decreased significantly for males (0.83 and 0.73 in.), and only thigh and 
biceps (0.59 and 0.38 in.) for females. Weight gain data were clustered into three or four 
segments for correlation analysis. Just eight significant relationships were found, two in 
clusters of ?2-6.6 lb and the others in clusters of ?4.2-20.4 lb. The limited results from 
the correlation analysis may have been influenced by small cluster sizes.  
Use of 3D body scanning enabled assessment of the degree of symmetry or 
asymmetry in circumferential changes. Planar cross-sections of the chest/bust, hip, and 
waist were derived and overlaid for the two time points. Each overlay was visually 
categorized as symmetrical or asymmetrical in four possible ways (more in the back and 
front, most in the back, most in the front, or more on the sides).  Males and females who 
gained and lost weight showed symmetric and asymmetric changes; males had 
proportionally more cases of symmetry than females.  
Findings showed that actual measurement increases that occurred with weight 
gained or lost varied across trunk locations and were not necessarily symmetric around 
the body. This has implications for the divergent fields of health care and the apparel 
industry. Health professionals are interested in where weight may be deposited in relation 
to obesity. Apparel product developers make assumptions about how bodies change for 
garment sizing. More longitudinal study is needed to expand understanding of the links 
between and patterns of weight and body measurement change. 
iii 
 
  
Acknowledgements 
 
The author wishes to express sincere appreciation to Dr. Pamela V. Ulrich and Dr. 
Karla P. Simmons, who served as thesis advisors, for their time, helpful suggestions, 
guidance, criticisms, and inspiration, which they have so willingly contributed. To the 
other member of the thesis committee, Dr. Lenda Jo Connell, the writer is deeply grateful 
for her guidance and interest. Gratitude is also expressed to Dr. Carol Warfield and Dr. 
Sareen Gropper who have been very supportive during the completion of the study. 
The author would also like to give special recognition to his brother and the 
family. The author wishes to express deep gratitude to his parents, T. Balasubramanian 
and K. Santha, and his brother, B. Karthikeyan for their constant love, support, patience, 
and understanding. 
Last but not least, the supreme energy that drives planets to platelets and galaxies 
to graviton is humbly praised. 
  
iv 
 
  
v 
 
Table of Contents 
 
Abstract??? ................................................................................................................... ii?
Acknowledgements ............................................................................................................ iv?
List of Tables .................................................................................................................... vii?
List of Figures .................................................................................................................... ix?
Chapter I. Introduction ........................................................................................................ 1?
Statement of Purpose .............................................................................................. 4?
Research Questions ................................................................................................. 4?
Chapter II. Review of Literature ......................................................................................... 6?
Changes in Weight Among Freshmen .................................................................... 6?
Fat Distribution and Physical Growth .................................................................... 9?
Fat Distribution ........................................................................................... 9?
Fat Distribution and Sexual Differences ................................................... 11?
Fat Distribution and Weight Changes ....................................................... 14?
Body Measurements for Sizing Systems .............................................................. 15?
Women?s Sizing ........................................................................................ 15?
Men?s Sizing ............................................................................................. 17?
 shown in Table 7. ..................................................................................... 17?
Body Shape ........................................................................................................... 22?
Female Figure Identification Technique (FFIT) for Apparel ................... 22?
Body Shape Assessment Scale (BSAS )
?
 ................................................. 25?
Male Body Shape Analysis ....................................................................... 29?
Body Shape and Symmetries .................................................................... 31?
Three Dimensional Body Scanning ...................................................................... 31?
Textile and Clothing Technology Corporation 3D Body Scanner ............ 32?
Anthropometric Studies Using 3D Body Scanning .................................. 34
  
vi 
 
?
Summary ............................................................................................................... 35?
Chapter III. Methodology ................................................................................................. 36?
Research Design .................................................................................................... 36?
Sampling Procedure .............................................................................................. 36?
Data Collection ..................................................................................................... 38?
Weight and Body Measurement Technology ........................................... 38?
Body Measurements.................................................................................. 38?
Data Analysis ........................................................................................................ 43?
Chapter IV. Presentation and Data Analysis..................................................................... 54?
Subjects and Data Collection ................................................................................ 54?
Data Analysis and Results .................................................................................... 55?
Subjects? Physical Activities..................................................................... 55?
Weight Change.......................................................................................... 56?
Research Question 1 ................................................................................. 57?
Research Question 2 ................................................................................. 62?
Research Question 3 ................................................................................. 75?
Research Question 4 ................................................................................. 77?
Chapter V.  Summary, Conclusions And Recommendations ........................................... 84?
Summary and Discussion ...................................................................................... 85?
Changes in Weight and Body Circumference Measurements .............................. 85?
Conclusions ........................................................................................................... 92?
Limitations ............................................................................................................ 93?
Implications and Recommendations for Future Research .................................... 94?
References ......................................................................................................................... 97?
Appendix  ........................................................................................................................ 108
  
vii 
 
List of Tables 
Table 1. ASTM Misses Standard Table ? D5585-01 ........................................................ 17?
 
Table 2 Measurement Increase on women?s sizing chart. ................................................ 17 
 
Table 3. ASTM Standard Tables of Body Measurements for Men Sizes Thirty-Four to              
Sixty (34 to 60) Regular- D6240-98 ................................................................................. 19?
 
Table 4. Chest and Hip measurement Increase in Men?s Sizing Chart ............................ 21?
 
Table 5. Waist measurement Increase in Men?s Sizing Chart .......................................... 21?
 
Table 6. Upper Arm Measurement Increase in Men?s Sizing Chart ................................. 21?
 
Table 7. Thigh  Measurement Increase in Men?s Sizing Chart ........................................ 20?
  
Table 8. Men?s Shape Designations for SizeUSA ............................................................ 30?
 
Table 9.  Weight Difference Spreadsheet Example .......................................................... 45?
 
Table 10. Weight Increased Group Body Measurement Spreadsheet Example ............... 46?
 
Table 11. Weight Decreased Group Body Measurement Spreadsheet Example .............. 46?
 
Table 12. Weight and Body Measurement Spreadsheet Example .................................... 48?
 
Table 13. Weight Difference and Body Measurement Difference Spreadsheet Example 49?
 
Table 14.  No Change in Weight and Body Measurement Spreadsheet Example ............ 51?
 
Table 15. Physical activities ............................................................................................. 56?
 
Table 16. Proportional Subject Population in Weight Lost, Gained and No Weight 
Change Groups.................................................................................................................. 57
  
viii 
 
Table 17. Mean Weight Change ....................................................................................... 57?
 
Table18.  Measurement Changes in Male and Female Subjects Belonging to the Weight 
Change Groups.................................................................................................................. 59?
 
Table 19. VMR Index ....................................................................................................... 62?
 
Table 20. Weight Range and Centroid  for Male and Female Subjects Belonging to the 
Weight Change Groups ..................................................................................................... 65?
 
Table 21. Correlation Coefficient for Weight Gained Group Males ................................ 67?
 
Table 22. Correlation Coefficient for Weight Gained Group Females ............................. 69?
 
Table 23. Correlation Coefficient for Weight Lost Group Males ..................................... 71?
 
Table 24. Correlation Coefficient for Weight Lost Group Females ................................. 73?
 
Table 25. Weight No Change Group ................................................................................ 76?
 
Table 26. Number of Subjects in Each Shape Classification for Weight Gained Males .. 79?
 
Table 27. Number of Subjects in Each Shape Classification for Weight Gained Females
........................................................................................................................................... 80?
 
Table 28. Number of Subjects in Each Shape Classification for Weight Lost Males ...... 81?
 
Table 29. Number of subjects in Each Shape Classification for Weight Lost females .... 82?
 
  
  
List of Figures 
Figure 1. Types of fat. ....................................................................................................... 11?
Figure 2. Residual fat areas in women. ............................................................................... 9?
Figure 3. Fat storage areas in men and women. ................................................................ 10?
Figure 4. FFIT
? 
Body shape.. ........................................................................................... 25?
Figure 5. BSAS
? 
Body Build. ........................................................................................... 26?
Figure 6. BSAS
? 
Body Shape.. ......................................................................................... 26?
Figure 7. BSAS
? 
Hip shape.. ............................................................................................ 26?
Figure 8. BSAS
? 
Bust shape.. ........................................................................................... 27?
Figure 9. BSAS
? 
Buttocks shape. ..................................................................................... 27?
Figure 10. BSAS
? 
Front Torso Shape. .............................................................................. 29?
Figure 11. Illustration of Robinette?s anthrpometric sizing system.. ................................ 31?
Figure 12. Six individual views of 3D data points............................................................ 33?
Figure 13. Bust circumference. ......................................................................................... 39?
Figure 14. Waist circumference. ....................................................................................... 40?
Figure 15. Hip circumference. .......................................................................................... 40?
Figure 16. Thigh circumference. ....................................................................................... 39?
Figure 17. Stomach. .......................................................................................................... 39?
Figure 18. Abdomen. ........................................................................................................ 42?
Figure 19. Seat. ................................................................................................................. 42?
Figure 20. Biceps. ............................................................................................................. 43?
Figure 21. Planar slices of bust. ........................................................................................ 53
ix 
 
  
  Figure 22. Scatter plot of weight changes in male and female subjects????.?.?..62 
Figure 23. Subdivisions formulated for correlation analysis??????????.....62                         
Figure 24. Examples of Symmetrical and Asymmetrical shapes of waist measurement..76 
 
  
x 
 
  
CHAPTER I. INTRODUCTION 
In the United States during the past 30 years, there has been a dramatic increase in 
overweight and obesity. According to the report, ?F as in Fat: How Obesity Policies Are 
Failing in America? (Trust for America?s Health Reports, 2008), the obesity rate for 
adults has been rising in 37 states. The National Health and Nutrition Examination 
Survey (NHANES) reported that obesity in adult men increased from 31.1% (2003-2004) 
to 33.3 % (2005-2006); for women, obesity increased from 33.2% (2003-2004) to 35.3% 
(2005-2006) (Centers for Disease Control and Prevention [CDC], 2007). Overall, 66 % of 
U.S. adults are categorized as overweight or obese (CDC, 2006). One-third of children 
are affected by overweight and obesity (Hedley, Ogden, Johnson, Carroll, Curtin, & 
Flegal, 2004; Ogden, Carroll, Curtin, McDowell, Tabak, & Flegal, 2006). There are 
regional differences; Mississippi ranks first, West Virginia ranks second and Alabama 
ranks third in the U.S. for obesity (Trust for America?s Health Reports, 2008). 
The years between ages 18 and 29 are marked as a critical period for elevated 
weight gain or obesity (Mokdad, Serdula, Dietz, Bowman, Marks, and Koplan, 1999). 
The ?freshman 15? is a hypothetical phenomenon concerning students coming to college 
communities that relates to the beginning of that age period. It assumes that freshman 
students, particularly women, gain weight averaging fifteen pounds by the time they 
become sophomores (Hodge, Jackson, & Sullivan, 1993). Anderson, Shapiro, and 
Lundgren (2003) reported that the transition period between high school and college 
1 
 
  
marks a drastic weight change period for young adults. Freshman weight gain has been 
attributed to junk food consumption, altered sleeping patterns (Graham & Jones, 2002), 
alcohol consumption (Wilson, Pritchard & Schaffer, 2004), cigarette smoking (Patterson, 
Lerman, Kaufmann, Neuner, & Audrain-McGovern, 2004), and disordered easting 
practices (Gutzwlller, Oliver, & Katz, 2003). Some studies have found that there is a 
significant weight increase in college freshmen (Hoffman, Policastro, Quick & Lee, 
2006; Holm-Denoma, Joiner, Vohs, & Heatherton, 2008; Hovell, Mewborn, Randle, & 
Fowler-Johnson, 1985; Levitsky, Halbmaier, & Mrdjenovic, 2004; Matvienko, Lewis, & 
Schafer, 2001). One study (Hodge et al, 1993) reported that there was no significant 
weight gain among freshman subjects. 
An approximate increase of 18.3 lb/year occurred in girls during the peak rates of 
weight gain (around 12.5 years); peak weight gain for boys was 19.8 lb in one year and 
that occurred at 14 years of age (Barnes, 1975). This weight gain is related to the fat 
distribution which occurs during the adolescence period. During the adolescence period 
for girls, growth is more in fat than muscle. Girls? fat accumulates particularly in the 
region of the trunk and upper thigh. For boys until adolescence, fat shows a greater 
increase than the growth of bones and muscles. However, the trend gets reversed during 
adolescence. Fat accumulates in body areas like the buttocks and flanks, abdomen, 
anterior and lateral surfaces of the upper thigh, breasts, back of neck and upper arm for 
girls, and for boys, fat accumulates in the trunk, chest, back and abdomen. Between ages 
6 and 18 a redistribution of fat from the extremities to the trunk region occurs (Croney, 
1971). 
2 
 
  
Weight gain can have implications for health. Some research in the last 25 years 
has explored the possible relationship between health risks and body size and where fat is 
placed, which can affect body shape. The focus has been on the central area of the torso, 
particularly waist circumference (WC) and waist-to-hip (circumferences) ratio (WHR). 
Rexrode, Carey, Hennekens, Walters, Colditz, Stampfer, Willett, and Manson (1998) 
compared the WHR and WC to determine risk of coronary heart disease (CHD) in 
women and found that WHR and WC were independently associated with coronary heart 
disease risk in women. Since WHR and WC measurements involve fat, these 
anthropometric measurements are often used in research related to weight gain. 
In relation to apparel, weight has been noted to be the best predictor of female 
body shape for girth measurements (O?Brien & Shelton, 1941). Thus weight change can 
affect body measurements. Different changes in girth measurements can affect body 
shape. An understanding of body measurements and description of body shape guide 
apparel sizing specifications.  The rising incidence of obesity in children, adolescents, 
and adults has implications for apparel sizing. 
Present sizing standards were developed with old anthropometrics and with 
assumptions about body shapes that may be invalid. Men and women have different body 
shapes, and it varies from person to person. Schofield and Labat (2005) found that pattern 
grading was not based on any anthropometric data. Pattern grading is defined as ?the 
process of increasing or decreasing the sample size pattern according to a set of body 
measurements and proportional relationships to develop a range of sizes for production? 
(Bye, 1999, p.10). Bye, LaBat, McKinney and Kim (2007) found that the existing 
grading practices in industry do not provide the necessary apparel fit for consumers. A 
3 
 
  
study by Kurt Salmon Associates reported that 50% of women and 62% of men could not 
find proper fitting apparel (Kurt Salmon Associates, 1999). Recent increases in the 
population?s weight may make the problem worse. 
 Exploring where body measurements change in relation to weight change has 
implications for the two different areas of health and apparel sizing. No research has been 
reported equating weight changes and changes in body measurements. Established 
interest in following weight change in college students and the availability of 3D body 
scanning offers the opportunity to follow body measurement change in adolescents. The 
proposed research will explore weight and body measurement changes among college 
students.  
Statement of Purpose 
The purpose of this study was to investigate possible relationships between 
changes in weight and body measurements using a sample of freshman males and 
females. The research sought to build evidence as to whether weight change (increases or 
decreases) appears to be reflected at particular girth locations on the body. 
Research Questions 
1. Did the body measurements of freshman males and females who gained or lost 
weight change from the beginning of fall 2007 to the end of spring 2008? 
2. Were weight increases or weight decreases of freshman males and females related 
to changes in body measurements? 
3. Did the body measurements change for the freshman males and females whose 
weight was stable? 
4 
 
  
5 
 
4. Do changes in body measurements appear symmetrical or asymmetrical around 
the front, back, and sides of the body when viewed as planar slices? 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
CHAPTER II. REVIEW OF LITERATURE 
The purpose of this study was to explore whether locational patterns exist that 
relate changes in weight and body measurements for freshman males and females. The 
research sought to build objective evidence to increase understanding of whether weight 
change (increases or decreases) appears more likely to occur at particular points on the 
body, as assessed by body girth measurements. It is believed that increases or decreases 
in fat and changes in fat distribution have direct impact on weight changes in the human 
body and body measurements. This review of literature discusses changes in weight 
among freshmen, fat distribution and physical growth, body measurements for apparel 
sizing systems, body shape, three dimensional body scanning, and anthropometric studies 
using 3D body scanning. These literature reviews explore research studies done in the 
past and relate them to current studies. 
Changes in Weight Among Freshmen 
 To explore whether the ?Freshman 15? is fact or fiction, several research teams 
have collected data at different institutions since 2000. Research conducted at the State 
University of New Jersey by Hoffman et al. (2006) assessed weight gain among freshman 
subjects. An initial sample of 217 volunteer participants in September ended with 67 (32 
male and 35 female) participants in the April that followed. Weight and percentage of fat, 
measured using a digital scale and Bio Electrical Impedance (BIA), were used for 
statistical analysis. Their results showed that a weight gain of approximately 6.82 lb 
6 
 
  
occurred during the freshman year. Another study (Mihalopoulus, Auinger, & 
Klein, 2008) surveyed 125 unmarried freshmen living on-campus. They collected data on 
weight, height, sex, and ethnicity through a self-report survey and analyzed the weight 
gain and Body Mass Index (BMI) for a period of seven months. Their results showed a 
significant weight gain for these college freshmen. However, the mean weights gained 
were less than those found by Hoffman et al. (2006); for men the gain was approximately 
3.7 lb and for women 1.7 lb. 
Anderson et al. (2003) had previously studied weight gain among freshman 
subjects. Weight gains over a period of four months (September to December) and nine 
months (September to May) were assessed using 145 and 46 participants, respectively. 
From the ANOVA test, they found an overall significant weight gain of approximately 
2.9 lb for the period of September to December. Of the 145 participants, 74 % (100) 
participants gained weight; 20% (27) lost, and the remaining participants showed no 
change in weight. During the timeline of nine months, a mean gain of 3.7 lb was found. 
Holm-Denoma et al. (2008) examined weight fluctuation and its predictors among 
freshmen. They conducted a survey among freshmen (266 male and 341 female) who 
were asked to self report their weight and height, and completed measures of self-esteem, 
eating habits, interpersonal relationships, exercise patterns, and disordered eating 
behaviors. Results showed that there was a significant weight change for both men and 
women. While men gained a weight of 3.5 lb, women gained 4.0 lb.  
In order to measure weight gain among freshmen over a period of 12 weeks, 
Levitsky et al. (2004) from Cornell University executed a study with 68 freshmen. 
Weights and heights were measured using a Healthometer scale at two time periods 
7 
 
  
(September and end of December). They found that the mean weight gain was 4.3 lb and 
that the increase in BMI was from 20.872.1 to 21.572.3 kg/m
2
. Delinsky and Wilson 
(2008) measured weight gain, relation between dietary restraint and weight gain, 
disordered eating, and concern about the ?freshman 15? among women. They tracked the 
students from September to April. There were 337 participants in September and 149 in 
April. They used several scales, including the Dutch Restrained Eating Scale, Eating 
Disorder Examination Questionnaire, and Rosenberg Self-Esteem Scale, and Body Shape 
Questionnaire, to collect data. From the self-reported weight, they found a mean weight 
gain of 3 lb over the studied timeline.  
Morrow, Heesch, Dinger, Hull, Kneehans, and Fields (2006) conducted research 
among freshman women. They tracked 137 students from fall, 2004 to spring, 2005. A 
balance beam scale and Stadiometer were used to measure weight and height. 
Additionally, they used the DXA model Lunar DPX-IQ to quantify body fat percent, total 
fat mass, and total fat free mass. They found that there was a significant weight gain 
averaging 2.4 lb among these college women during their freshman year. The first 
published freshman 15 research, conducted by Graham and Jones (2002), collected 
weight and body fat data from 49 (39 female and 10 male) freshman students at the 
beginning of fall and end of spring. Of the 49 participants, 59% gained weight averaging 
4.6 lb; 26 % of the participants lost weight during the academic year. Graham and Jones 
found no significant change in body fat among the participants. Though their results show 
no changes in body fat, several other researchers have studied and presented results 
pertaining to changes in body fat, fat distribution, body growth and related weight 
changes, as addressed in the next section. 
8 
 
  
A pilot study was conducted at Auburn University to examine changes in body 
weight, fat, shape and size in a group of 35 college freshmen (25 females, 10 males). 
Weight and height were measured at the beginning and end of the fall semester 2006 and 
at the end of spring 2007. By the end of fall, female subjects gained an average weight of 
1.98 lb and at the end of spring it was around 3.02 lb. Average weight gain by male 
subjects was 2.2 lb by the end of fall and 5.35 lb by the end of spring (Drawdy, Zizza, & 
Gropper, 2007).  A year later, another study on weight change and body composition was 
begun at Auburn University. It?s 216 freshman students generated the data for the 
planned research to be reported in chapter IV for September 2007 to April 2008. Analysis 
of the weight data collected at the beginning and end of fall semester of 2007 showed an 
average increase of 1.7 lb for female students and 2.8 lb for male students over the period 
of one semester (Saunders, 2008).  
Fat Distribution and Physical Growth 
Fat Distribution 
   The fat distribution pattern in an adult human body is launched during the early 
adolescence period (Ramirez, 1993). The developing pattern of fat distribution is related 
to the sexual and skeletal maturity of adolescents (Baumgartner, Roche, Guo, Lohman, 
Boileau, & Slaughter, 1986). For normal adults of both sexes, subcutaneous fat (fat 
found just beneath the skin) (Ref. Figure 1) has a preference to accumulate in specific 
body areas (Ref. Figures 2 and 3) which include the buttocks and flanks, abdomen, 
anterior and lateral surfaces of the upper thigh, breasts, back of neck and upper arm. Fat 
9 
 
  
in men is mostly stored on the trunk, chest, back and abdomen. Changes in fat 
distribution are reflected in shape and weight changes (Croney, 1971; Thibodeau, 1990). 
During the long period of adulthood, shape and body proportions are not stable 
(Hooten & Dupertis, 1951). In adulthood, body growth stops except in the central part, 
including the lower torso. However, fat continually tends to increase in the trunk region 
with increasing age;  fat in the extremities may not change or may even decrease in both 
males and females (Brozek, 1952). In the trunk region, visceral fat (fat deposited between 
the internal organs in the torso) increases more when compared to subcutaneous fat 
during adulthood (Borkan & Norris, 1977) (Ref. Figure 1). Two individuals with similar 
body weight and sum of skin fold thickness or percent body fat may have different 
distributions of subcutaneous fat (Mueller, 1982).  
 
Figure 1. Types of fat (How to get rid of belly fat!, http://www.nutrition-health-
supplements.com/belly-fat.html) 
10 
 
  
 
Figure 2. Residual fat areas in women (Croney, 1971). 
      
Figure 3.Fat storage areas in men and women (Thibodeau, 1990). 
Fat Distribution and Sexual Differences 
Mueller?s (1982) study on the changes of anatomical distribution of fat with age, 
with 810 participants aged 6-15 and 715 participants aged 25-80, found that there was a 
general increase in fat among females aged 12-30. Fat continued to increase up to 40 
years of age. Girls aged 12-17 had more fat than boys in that age range. During ages 6-18 
and into young adulthood, there is a redistribution of fat from the extremities to the trunk 
11 
 
  
for both sexes (Mueller, 1982; Ramirez & Mueller, 1980). Weight gained during late 
adolescence tends to deposit centrally. Reynolds (1951) found that there was a decrease 
in extremity fat for both sexes within the age range of 6.5-18.5 years. In adolescence, the 
redistribution of fat from leg to trunk accelerates, and this acceleration is higher in males 
than females (Baker, et al., 1951). The distribution of fat in the central region occurs 
more with obese adolescent girls, with fat being deposited especially in the upper trunk 
(Deutsch, Mueller, & Malina, 1985; Hattori, Becque, Katch, Rocchini, Boileau, 
Slaughter, & Lohman, 1987).  
A few of the longitudinal studies about skin fold thickness have reported the 
development of centripetal fat (fat in trunk) distribution after puberty for both sexes         
(Cronk, Mukherjee, & Roche, 1983; Cronk, Roche, Chumlea, & Kent, 1982). When there 
is a loss of fat in the anterior leg region (refers to the front left compartment of a normal 
leg cross section), there is a gain of fat in the mid- trunk area. This pattern continues up to 
the fourth decade of life, suggesting that an absolute decrease in fat thickness in one part 
of the body leads to an absolute gain in another part (Garn & Young, 1956).  
During the adolescence period for girls, growth is more in fat than muscle. Fat 
accumulates particularly in the region of the trunk and upper thighs for girls. It is reported 
that, in general, girls have more subcutaneous fat compared to boys. There is a 
physiologically disproportionate deposition of subcutaneous fat at central sites in girls 
during late adolescence. This deposition rate tends to increase as the girls approach their 
adult size (Hediger, Scholl, Schall & Cronk, 1995). For boys until adolescence, fat shows 
a greater increase than the growth of bone and muscle. However, that trend gets reversed 
during adolescence (Croney, 1971). 
12 
 
  
While it was a known phenomenon that girls accumulate both centripetal (fat in 
stomach) and peripheral fat (fat in arms, legs and subscapular), Ramirez (1993) found 
that many boys in his study also accumulated centripetal and peripheral fat. His subjects 
were adolescents aged between 12 and 20 whose arms and legs were measured to 
determine fat accumulation. In boys, the trunk skinfolds were found to increase until they 
reached 13 years of age, and then found to decline for a brief period through 14 years of 
age. Following that, the trunk skinfold was found to increase through late adolescence. 
(Malina, Bouchard and Bar-or, 2005). For adult men, subcutaneous fat tends to 
redistribute from peripheral to central regions (Schwartz, Shuman, Bradbury, Cain, 
Fellingham, Beard, Kahn, Stratton, Cerqueria, & Abrass, 1990) and to internal organs 
(Borkan, Hults, Gerzof, Robbins, & Silbert, 1983).  
Sinha, Kapoor and Kapoor (2008) conducted a study to track the subcutaneous fat 
accumulation for two generations of males. Data on body weight, height and skin fold 
thickness at the biceps, triceps, subscapular, suprailiac and medial calf of 414 adolescent 
boys and their fathers were collected. The boys belonged to the age group 11 to 17. 
Results showed a continuous increase in body weights for age groups of 11 to 17 and 35 
to 55. Patterns of subcutaneous fat distribution in adolescent boys and their fathers were 
assessed by their skin fold thickness. Fatness consistently increased with age, and fat 
distribution patterns were found to be established during ages 16 to 17. Another study 
reported that among a sample of boys aged 13 to 18, early-maturing boys had more 
subcutaneous fat on the trunk compared to late-maturing boys (Beunen, Malina, Lefevre, 
Claessens, Renson, Simons, Maes, Vanreusel, & Lysens, 1994). Poehlman, Toth, 
Bunyard, Gardner, Donaldson, Colman, Fonong, and Ades (1995) conducted a study to 
13 
 
  
estimate changes in waist circumference due to subcutaneous fat for a group of men and 
women aged between 18 and 88 years. He reported an average increase in waist 
circumference of 0.28 cm per year for women and 0.18 cm per year for men.  
Fat Distribution and Weight Changes 
Fat distribution and deposition results in weight change with increasing age. After 
puberty, fat accounts for 16 % of the total weight for boys and 27 % of total weight for 
girls (National Center for Health Statistics, 2001). Peak weight gain for boys was 19.8 lb 
in one year and that occurred at 14 years of age (Barnes, 1975).  Studies have reported 
that an approximate increase of 18.3 lb/year occurred in girls during the peak rates of 
weight gain (around 12.5 years). Weight gain slows during the period of menarche, but 
extends into late adolescence. From about one year after menarche to age 18 years, girls 
on average gained a total weight of 6.5 kg (14.3 lb) (Hediger, Scholl, Schall, & Cronk, 
1995). Average weight gain among females during puberty (older than ages 13.4) was 
found to be within 15-55 lb of weight and for males it was 15-65 lb. Normally, during the 
latter half of adolescence, females gained 14 lb of weight (Barnes, 1975; Wong, Wilson 
& Whaley, 1995).  
Siervogel,Wisemandle, S. Guo, Roche, Chumlea, and Towne (1998) calculated 
the rate of increase in total body fat as .57 kg/years males between ages 18 and 45, and 
.37 kg/year for men between ages 45 and 65 years. The most rapid increase in visceral 
fat, resulting in an increase in waist size, occurred in late adolescence (Ref. Figure 3). 
Kemper (1985) found a linear weight increase in boys from 38 to 63 kg as their age 
moved from 12 to 17 years, and found that the skinfold thickness of the trunk increased 
14 
 
  
with a simultaneous decrease in the upper arm. On the contrary, there was an increase in 
both trunk and upper arm thickness for girls of same age group. 
Body Measurements for Sizing Systems 
Women?s Sizing 
Women?s apparel sizing systems follow an arithmetic progression of size numbers 
that do not denote body measurements. In 1939, the U.S. Department of Agriculture 
sponsored a sizing study among U.S. women. O?Brien and Shelton (1941) manually 
collected 59 body measurements from 14,698 women. They found that women with the 
same bust measurements differed in their hip measurements by 5 to 6 inches. 
Additionally, they had calculated the correlation between weight and 19 girth 
measurements of 10,000 women. The correlation between weight and chest, bust, waist, 
abdominal extension, hip, and maximum thigh were found to be 0.8950, 0.8981, 0.8771, 
0.8990, 0.9190, and 0.8725, respectively. These statistics suggested a strong relationship 
between weight and girth measurements. The measurements were used as the foundation 
for today?s women?s sizing.  
Since all the participants were white, this study was biased from the point of data 
collection. Thus, the collected data were not sufficiently appropriate to standardize the 
sizing chart (O?Brien & Shelton, 1941). Nonetheless, these data were used to develop a 
?voluntary standard? published in 1958 by the National Bureau of Standards (NBS) as 
CS215-58, titled Body Measurements for the Sizing of Women?s Patterns and Apparel. 
This standard was based on the 1939 study. It classified women into ?Misses?, 
?Women?s?, ?Half-Sizes? and ?Juniors? categories. Furthermore, it had classifications of 
15 
 
  
height (Tall, Regular, and Short), and bust and hip measurement (Slender, Average, and 
Full) (U.S. Department of Commerce, 1958). 
In 1971, the U.S. Department of Commerce released a new voluntary standard, 
known as PS 42-70. This standard was a revision of CS 215-58 with modifications based 
on the health surveys made in 1960-62 by the U.S. Public Health Service. In this 
standard, the bust girth was increased by one grade interval per size code; ?Slender? and 
?Full? hip options were eliminated from the ?Women?s? category and ?Tall from the 
?Junior? category. (U.S. Department of Commerce, 1970). 
The current standard sizing chart was developed by the American Society of 
Testing and Materials (ASTM). In 1995, the ?Misses? standard ASTM D5585-95 was 
released with sizes ranging from 2 to 20. A sizing standard for women aged above 55, 
ASTM D5586-95, was also released by ASTM in 1995. Both the standards were based on 
the PS 42-70 chart with updated information (ASTM, 1995). Later, in 2002, the ?Juniors? 
standard ASTM D6829-02 was released with sizes ranging from 0 to 19 (ASTM, 2002). 
In the ?Misses? standard D5585-01(see Table 1), the sizes include 2, 4, 6, 8, 10, 12, 14, 
16, 18, and 20. The bust, waist and hip measurements from the chart keep increasing for 
every size from 2 to 10 with a difference of 1 in., from 10 to 16 with an increase of 1.5 in. 
and from 16 to 20 with a difference of 2 in. Similarly, the upper arm size increases by 
0.25 in. for sizes between 2 and 10, by 0.375 in. between 10 and 16, and by 0.675 in. 
between 16 and 20. The thigh max measurement follows an increase of 0.75 in. for sizes 
between 2 and 10, 1 in. between 10 and 16, and 1.25 in. between 16 and 20. (Table 2) 
Likewise, all the other measurements follow an arithmetic progression at different 
sections of intervals. 
16 
 
  
Table 1 
  ASTM Misses Standard Table ? D5585-01 (ASTM, 2001) 
 
 
Table 2 Measurement Increase on Women?s Sizing Chart 
Size 2-10 10-16 16-20
Bust 1? 1 ?? 2?
Waist 1? 1 ?? 2?
Hip 1? 1 ?? 2?
Thigh max ?? 1? 1 ?? 
Upper arm ?? 3/8? 5/8?
 
Men?s Sizing  
 In general, men?s apparel sizes are denoted by direct measurements.  American 
producers were the first to apply standard sizing techniques to make quality garments. 
During the American Civil War, more than a million military men were measured for 
their chest sizes and height. In 1890, a sizing chart for juveniles and youth was developed 
covering chest sizes from 21 in. to 36 in. In 1921, the first report of an American 
anthropometric survey including clothing sizing was published (Ashdown & Na, 2007). 
In 1988, an anthropometric study called ANSUR was conducted among 1774 army men 
17 
 
  
and 2208 women to develop a sizing chart for military clothing (Fan & Yu, 2004). In 
1998, ASTM published its first sizing chart, ?Standard Tables of Body Measurements for 
Men Sizes Thirty-Four to Sixty (34 to 60) Regular? (see Table 3), which has since 
undergone revisions (ASTM, 2004). Measurements of chest and hip monotonically 
increase by 1 in with each size increase (ref. Table 4). For the waist measurement, sizes 
34-38 have a 1 in. increment; sizes 38-52 have a 1? in. increment and sizes 54-60 have a 
2 in. increment (ref. Table 5). An increasing trend followed by a decreasing trend is seen 
in the upper arm measurement. Table 6 tabulates the measurement pattern for upper arm. 
Thigh measurements exhibit a repeating pattern of fractions, which is shown in Table 7. 
 
 
 
 
 
 
 
 
 
 
 
 
18 
 
  
Table 3  
ASTM Standard Tables of Body Measurements for Men Sizes Thirty-Four to Sixty (34 to 
60) Regular- D6240-98 (ASTM, 2004) 
 
 
19 
 
  
Table 3 (Continued) 
 
 
 
20 
 
  
Table 4  
Chest and Hip Measurement Increase in Men?s Sizing Chart 
Size 34-60 
Chest 1" 
Hip 
 
Table 5  
Waist Measurement Increase in Men?s Sizing Chart 
Size 34-38 38-52 52-54 54-60 
Waist 1" 1  1/4" 1  1/2" 2" 
 
Table 6  
Upper Arm Measurement Increase in Men?s Sizing Chart 
 
Size 34-37 37-38 38-60 
Upper arm  1/4  1/2  1/4 
 
 
Table 7  
Thigh Measurement Increase in Men?s Sizing Chart 
  
 
Size 34-36 36-37 37-38 38-44 44-45 45-46 46-52 52-53 53-54 54-60 
Thigh  5/8  2/3  4/7  5/8  2/3  4/7  5/8  2/3  4/7  5/8 
 
 
 
 
 
21 
 
  
Body Shape 
One of the most important features used to classify biological specimens is shape. 
In 1638, Galileo did a shape analysis and found that bones in small animals were not 
scaled down from bigger animals (Dryden & Mardia, 1998). Initially shape analysis was 
performed from a biological point of view. Later, several research studies were done to 
identify different types of human body shapes. In the 19
th
 century, researchers used a grid 
system to study biological size and shape variation. In this system, body shapes were 
placed on a grid and compared for size variations (Thompson, 1917). In 1930, a 
mechanism to classify body types was initiated by William Sheldon. He used nude 
photographs of 4,000 undergraduate male students aged 16-20 to arrive at a classification 
of body types (Sheldon, Stevens, & Tucker, 1940). Basically size is not related to shape. 
Mossiman (1988) elucidated that there is no relationship between size and shape. Saying 
that shape should be used for classification and could be used to classify organisms 
within species as small or large, he concluded that shape does not depend on size. 
Recently, 3D body scanned images have been used in body shape analysis, including in 
the apparel industry.  The purpose of studying human shape in relation to the apparel 
industry has been to improve sizing. 
Female Figure Identification Technique (FFIT) for Apparel  
 Research on body shape analysis using 3D body scanning by Simmons (2002) 
identified different types of body shapes in females. A convenience sample of 222 
females aged 18 and older was selected. Subjects were scanned using the Textile and 
Clothing Technology corporation [TC]
2
 NX-12 scanner. Their body measurements, 
including bust, waist, hip, stomach and abdomen circumferences were extracted and used 
22 
 
  
to identify the body shapes of the participants. Software was developed to identify 
shapes; the nine shapes  identified were hourglass, bottom hourglass, top hourglass, 
triangle, inverted triangle, spoon, oval, diamond and rectangle. These are shown in Figure 
4 and descriptions are as follows: 
? Hourglass.  Hourglass acted as a basis for many shape categorizations. 
Bust, waist and hip were the measurements used to define this shape. If 
the ratio of bust-to-waist and hip-to-waist are equal, then it is defined as 
hourglass.                                        
? Bottom hourglass. It is a subset of hourglass shape. In this shape, hip 
measurement is larger than bust measurement and the ratio of bust-to-
waist and hip-to- waist are significantly large enough to produce a definite 
waistline. 
? Top hourglass. It is also a subset of hourglass shape, but is exactly the 
opposite of the bottom hourglass shape. In this shape, bust measurement is 
larger than hip measurement and the ratio of bust-to-waist and hips-to 
waist are significant enough to produce a definite waistline. 
? Spoon. Bust, waist, hip and high hip were used to determine this shape. If 
a person has large difference between her hip and bust measurement and 
has a low ratio of bust-to-waist and high ratio of hip-to-waist when 
compared to the hourglass shape, it is defined as spoon shaped. 
? Rectangle. This shape is determined by waist, bust and hip circumference. 
The conditions for this shape are the bust and hip measurement should be 
equal and bust-to-waist and hip-to-waist ratio should be low.  
23 
 
  
? Diamond. The measurements used to define this shape are bust, waist, hip, 
stomach and abdomen. The person will fall into this category when the 
averages of the subject?s abdomen, stomach and waist are larger than the 
bust measurement. 
? Oval. Measurements used to determine this shape are the same for the 
diamond shape. The person will fall into this category when the averages 
of the subject?s abdomen, stomach and waist are smaller than bust 
measurement.                                                                     
? Triangle. This shape is determined by bust, waist and hips measurements. 
If a person has a larger hip measurement than bust measurement and low 
hip-to-waist ratio, the person has the triangular shape. 
? Inverted triangle. The measurements used to determine the inverted 
triangle shape are the same as the triangle shape. In this shape, the bust 
measurement is larger than the hip measurement and there is a smaller 
bust-to- hip ratio. 
 Of the 222 subjects, 40 % were bottom hourglass, 21.6% were hourglass, 17.1% 
were spoon, 15.8% were rectangle, 3.6% were oval, and 1.8% subjects were triangle. 
Simmons (2002) didn?t find any subject reflecting the shape of inverted triangle, diamond 
and top hourglass. This study demonstrated that female body shapes differ from person to 
person. However, the existing systems of sizing charts do not seem to take into 
consideration these variants in shape, resulting in inappropriate or compromised fit of the 
resultant garment. 
24 
 
  
                              
Hourglass         Bottom Hourglass                
         Spoon                Rectangle               Diamond 
     Oval                 Triangle         Inverted Triangle 
Figure 4.  FFIT Body shape (Simmons, 2002). 
 
Body Shape Assessment Scale (BSAS )
?
  
Connell, Ulrich, Brannon, Alexander, and Presley (2006) developed another scale 
to assess the body shape of female figures. They used 42 body scan printouts of women 
aged 20 to 55 and 14 female shape scales from the literature to classify the scans 
according to their shapes. Based on the usefulness of the 14 scales in classifying figures, 
a new scale called the Body Shape Assessment Scale (BSAS)
?
  was developed with nine 
components. They are body build, body shape, posture, hip shape, shoulder shape, front 
torso shape, bust shape, buttocks shape, back shape, and posture as its element scales. 
This scale was validated using experts. 
25 
 
  
Of the nine components, body build (see Figure 5), body shape (see Figure 6), hip 
shape (see Figure 7), bust prominence (see Figure 8), buttocks shape (see Figure 9) and 
front torso (see Figure 10) scales may be related to fat accumulation and weight increase. 
Fat accumulation is a factor that influences body build ranging from slender to heavy, 
and weight increase could be reflected in body shape changes such as front torso shape. 
Definitions of these BSAS
?
 shapes are offered in the following pages.  
  Body build. The frontal view of the overall appearance of body was used to assess    
the body build. Figures range from slender to heavy. 
 
 
Figure 5.  BSAS
? 
Body Build (Connell et al., 2006).
Body shape. This component was used to classify the body shape into hourglass, 
pear, rectangle or inverted triangle. The relationship between waist, shoulder and hip 
according to their width from a frontal view was used. 
 
26 
 
  
 
 
Hip shape. Hip shape was categorized as straight, high-hip, middle-hip, or low-
hip when viewed from the front. High, middle, and low describe the placement of the 
widest location. 
Figure 6. BSAS
? 
Body Shape  (Alexander, 2004). 
 
 
Figure 7.  BSAS
? 
Hip shape (Alexander, 2004). 
Bust shape.  The extension of the bust from flat to prominent as viewed in a side 
profile was used to classify the bust shape. Flat, average, and prominent are the three 
stages. 
 
27 
 
  
 
 
Figure 8. BSAS
? 
Bust shape (Alexander, 2004). 
Buttocks shape. Similar to bust, degree of prominence through side view was used 
to assess buttocks shape as flat, average and prominent. 
 
 
 
Figure 9. BSAS
? 
Buttocks shape (Alexander, 2004).
Front torso shape.  A side profile of the body scans was used for the 
assessment of front torso shape. The shape of the torso from bust to top of leg was 
used to categorize the shape into b, B and D categories. 
28 
 
  
 
 
Figure 10. BSAS
?  
Front Torso Shape (Connell et al., 2006). 
Alexander (2004) conducted research using the BSAS
? 
to identify female body 
shape and exploring the relationship between body shape characteristics, fit problems, 
body cathexis, and clothing benefits sought. A convenience sample of 529 women 
between ages 19-35 were studied. Some relationships between body shape and age, fit 
problems, body cathexis, clothing benefits were identified. Body shape, body build, and 
some of the body component parts were also linked. 
Male Body Shape Analysis 
As a part of the SizeUSA project, [TC]
2
 conducted a male shape analysis study. A 
set of 3691 men grouped under the age headings of 18-25, 26-35, 36-45, 46-55, 56-65 
29 
 
  
and 66 plus were scanned for their body shapes. Of the 3691 men, 47% were Non-
Hispanic White, 19% were Non-Hispanic Black, 9% were Hispanic, and 8% were other 
ethnicities. In the total population, approximately 27% were within the age range of 18-
25. Shapes of this subset were analyzed and are represented in Table 8. From the above 
study, it has become evident that different types of body shapes are prevalent even among 
men (SizeUSA, 2004). 
Table 8  
Men?s Shape Designations for SizeUSA ([TC]
2
, 2004)  
 Black Hispanic White Others 
Number of subjects 146 190 484 175 
Square Shoulders % 24.66 46.32 38.84 40 
Slope Shoulders % 50 26.32 36.36 28.57 
Right Low Shoulder % 28.08 26.32 36.57 37.71 
Left Low Shoulder % 20.27 6.84 5.37 7.43 
Erect Posture % 34.93 34.74 26.24 34.86 
Stooped Posture % 19.18 15.26 22.73 15.43 
Head Forward Posture% 30.82 30 42.15 27.43 
Lower Front Waist % 12.33 13.68 4.55 4 
High Left Hip% 2.05 7.89 4.13 4 
High Right Hip % 8.22 13.68 9.5 9.14 
Shifted Hips % 0 0.53 0.62 0 
Forward Shoulders % 41.1 41.58 45.87 41.71 
Raise Back Waist % 15.75 5.79 6.4 4.57 
Flat Seat -% 1.37 3.16 1.24 4 
Prominent Seat % 36.3 16.84 18.8 10.86 
Chest Dart % 19.18 8.95 11.36 10.29 
Prominent Blades % 1.37 1.05 0.83 0 
Prominent Biceps % 18.49 8.42 9,09 9.14 
Prominent Calves % 4.79 13.16 21.07 14.86 
Bow Legs % 11.64 7.89 8.06 6.86 
Portly% 5.48 9.47 9.3 8.57 
Sway Back. % 0 1.58 0.62 1.14 
 
 
30 
 
  
Body Shape and Symmetries 
In 1986, Robinette derived an anthropometric sizing system based on height and 
weight measurement. Three dimensional blocks in Figure 11 were used to demonstrate 
this system. Size X would fit block A and block C, Size Y would fit Block E and Block 
F, and Size Z would fit block B and D (Barker & Coletta, 1986). This system elucidated 
that even with same block volume, shapes of blocks differed from each other. Similarly, 
there is a chance for difference in the symmetry of body measurements.  
 
 
Figure 11. Illustration of Robinette?s anthropometric sizing system (Barker et al., 1986). 
 
Three Dimensional Body Scanning 
Traditionally, anthropometric measurements were taken using a tape measure, 
weight scale, anthropometer, caliper, sliding compass, and head spanner (Simmons & 
Istook, 2003). ?The precision and accuracy of measurements taken with a tape measure 
31 
 
  
are often not sufficient to detect small, but important differences in body dimensions? 
(Heisey, 1984, p. 9). For an anthropometric survey of U.S. Army personnel in 1988, four 
hours were invested to locate the physical landmarks and measure the body of a single 
subject (Paquette, 1986). The development of 3D body scanning reduces hours to 
seconds. Now, body scanning technology used for measuring body measurement is 
cheaper, faster and provides more detail than traditional anthropometry (Petrova & 
Ashdown, 2008). 
Simmons and Istook (2003) compared 3D body scanning measurement methods 
and evaluated them against anthropometric methods for apparel sizing. They chose 21 
body measurements which were used for designing a well-fitted garment. The 
measurements were mid neck/neck base, chest/bust, waist by natural indentation/waist by 
navel, hip/seat, sleeve length, arm length, inseam, outseam, shoulder length, thigh, bicep, 
and wrist. Three different types of scanners produced by [TC]
2
, SYMCAD, and 
CYBERWARE were used for the study. They concluded that the [TC]
2
 scanner identified  
the greatest number of body sites needed for the study and also had the capacity of 
producing many more measurements with specific application for apparel. 
Textile and Clothing Technology Corporation 3D Body Scanner  
 [TC]
2 
developed its 3D body scanner particularly for the apparel industry. [TC]
2 
uses the technique called Phase Measuring Profilometry (PMP) based on moir? technique 
(Paquette, 1996). ?This method involves shifting the rating preset distances in the 
direction of the varying phase and capturing images at each position? (Fan et al., 2004). 
The white light used in this scanner projects a sinusoidal fringe on the object. Due to the 
irregularities in the object, the projected grating are distorted, and from the resulting 
32 
 
  
fringe pattern, the contour of the object can be determined (Demers, Hurley, Wulpern, & 
Grindon, 1997). The latest version of this scanner, the NX16, has 16 cameras and 16 
sensors ([TC]
2
, 2008). Each sensor takes four images, each with same degree of phase 
shift of the projected sinusoidal patterns. The four captured images are used to determine 
the phase of each pixel. Finally, the phase is used to calculate the three dimensional data 
points. The data clouds for each of the six views, as shown in Figure 12, are the output of 
the PMP process. By knowing the exact orientation of each view with respect to one 
other, the individual views are combined (Demers et al., 1997). The point accuracy of this 
scanner is < 1mm and circumferential accuracy is < 3mm. ([TC]
2
, 2008). 
 
 
 
Figure 12. Six Individual views of 3D data points (Demers et al., 1997). 
 
 
 
33 
 
  
Anthropometric Studies Using 3D Body Scanning 
 CAESAR (Civilian American and European Surface Anthropometry Resource) 
was an anthropometric study which used body scanning to obtain body measurements. 
This study was conducted in the years 1997-2001 in Europe and the U.S. with 
approximately 4,500 subjects (men and women). Out of the 99 measurements collected, 
59 measurements were obtained through body scanning. Governments and commercial 
industries sponsored the CAESAR study (CAESAR, 2000).  
Another study called SizeUSA was conducted in 2004 in the U.S. This study was 
funded by the U.S. Department of Commerce and conducted by [TC]
2
 using their 3D 
body scanner. More than 10,000 subjects were scanned in different locations and with 
different demographics. The subjects were classified by age, ethnicity, weight, annual 
household income, marital status, lifestyle, education, employment status, and apparel 
shopping preferences (SizeUSA, 2004). Similar, studies have been conducted in the 
United Kingdom and Korea (SizeUK and SizeKOREA) using 3D body scanning systems.  
 SizeUSA data helped companies to understand their customer base better. JC 
Penney collaborated with [TC]
2
 to analyze body shapes of their customers. 
Approximately 67 female participants wearing sizes from 4 to 16 were scanned and 
measurements extracted. Results showed that 43% of its customer base was ?pear 
shaped?, 33% was ?apple shaped?, and 19% was ?rectangle shaped?. This study helped 
them to provide incremental alterations to their specifications for bust, waist and hip 
measurements (Sizing up garment fit issues, 2007). 
 
 
34 
 
  
35 
 
Summary 
Based on these literature surveys, it can be observed that there are studies 
showing significant increases in weight among freshmen. Also, it is obvious that fat 
distribution is intertwined with physical growth which results in changes in body 
measurements. This study explored the possible relationship between body measurement 
changes and weight variations among freshman. It is understood from studies that body 
shapes are different and that anthropometric data does not always reflect these different 
body shapes. It has been suggested that grading rules are not in coherence with 
anthropometric data and hence influence improper fit. This makes the current study 
significant wherein the changes in body shape according to weight changes can be 
assessed. 
 
 
 
 
 
 
 
 
 
 
 
  
CHAPTER III. METHODOLOGY 
Research Design 
This study was designed to explore the relationship between weight change and 
body measurement change for a sample of freshman males and females. It sought to 
increase understanding of whether there appear to be patterns as to where on the body 
measurements change with weight gain or loss and what amounts of body measurement 
change may be associated with weight change.  
Sampling Procedure  
 As part of a longitudinal study of change in the body composition and size of 
college students, a sample of freshmen (ages 17 to 19) at Auburn University was 
recruited through a convenience sampling method, at the beginning of fall semester, 
2007.  Potential participants were contacted via email, posters, and classroom contact in a 
variety of introductory courses typically taken by freshmen. Volunteers were told that 
they would be asked to participate in data collection sessions at the beginning of the 
academic year, at the end of fall semester, and at the end of spring semester. Subjects 
who were outside the age range, pregnant, married, had children or had a diagnosed 
eating disorder were not allowed to participate in this study. An incentive of $75 was 
given for participants who completed all three sessions. The number of participants in 
August 2007 was 240 freshmen (156 females, 84 males).  Among the 240, 214 freshmen 
36 
 
  
(89% retention rate) participated in data collection at time point #2 in December, 2007, 
and 205 freshmen (85 % retention rate) participated in data collection at time point #3 in 
April, 2008.
37 
 
  
Data Collection 
In order to disconnect the data from being traceable to a subject, each participant 
was allotted a specific code. Subjects completed a questionnaire which included 
demographic information, were weighted, had their body fat measured, and then were 
body scanned. Data collected at two time periods, namely beginning of fall and end of 
spring, were extracted from the 3D whole body scanner and used for this study. Subjects 
were weighed but not body scanned at the end of the fall semester time period. 
Weight and Body Measurement Technology 
 A digital weighing balance was used to measure the subjects? weight.  Care was 
taken to keep the balance level with the ground. Measured readings were noted to the 
nearest 0.1 pound. The NX-16 3D body scanner system, developed by [TC]
2
, was used to 
capture scan images and extract body measurements. The system calculates body 
measurements using the digital imaging. Subjects were scanned in a standard standing 
position. Each scan took eight seconds, and a body model with extracted body 
measurements was developed in less than one minute. The [TC]
2 
scanner uses white light 
technology, which is, ?considered to be the safest body scanning technology? ([TC]
2
, 
2007). Measurements were extracted at any time and as many times as needed following 
the scan. Point accuracy of the scanner is <1mm (0.0394 in.) and circumferential 
accuracy is <3mm (0.1181 in.) 
Body Measurements 
Normal adults of both sexes tend to accumulate subcutaneous fat in specific body 
areas. The accumulation of fat occurs in buttocks and flanks, abdomen, anterior and 
38 
 
  
lateral surfaces of the upper thigh, breasts, back of neck and upper arm (Croney, 1971). 
Studies on female fit preferences have found that most fit problems occurred at the waist, 
hip and bust. Tightness is a common fitting problem that occurs at the bust, waist, hip, 
thigh, upper arm and back width (Pisut & Connell, 2007). For these reasons, the bust, 
waist, hip, stomach, seat, thigh, biceps and abdomen are expected to be the major change-
prone measurements with weight change. Hence, the current study was focused on these 
measurements in combination with weight change. 
Body measurements (bust, chest, waist, hip, thigh, biceps, stomach, abdomen and 
seat) were extracted from the scanner using the embedded software. The definitions of 
measurement extraction rules and illustrative figure are as follows. 
? Bust: The bust measurement was taken at the largest protrusion of the 
bust. The full circumference was taken parallel to floor. (Figure 13) 
                                          
Figure 13. Bust circumference. 
? Waist: The waist was the smallest circumference between the bust/chest 
and hips determined by locating the small of the back and then going up 
and down a predetermined amount to find the waist. (Figure 14) 
 
39 
 
  
 
Figure 14. Waist circumference. 
? Hip: The largest circumference between the waist and crotch was 
identified as the hip. (Figure 15) 
 
Figure 15. Hip circumference. 
? Thigh: The largest circumference of the right/left thigh was taken between 
the knee and two inches below the crotch. (Figure 16) 
 
 
40 
 
  
 
Figure 16. Thigh circumference. 
? Stomach: The circumference at the largest protrusion of the front torso 
was taken between the front waist point and 1/3 of the distance from waist 
to shoulder. (Figure 17) 
 
 
Figure 17. Stomach. 
? Abdomen: The circumference was measured from the forward-most 
protruding point below the waist and above the crotch when viewing the 
body from the side. (Figure 18) 
  
41 
 
  
 
Figure 18. Abdomen. 
? Seat: The circumference was measured from the rear-most protruding 
point below the waist and above the crotch when viewed from the side. 
(Figure 19)  
 
Figure 19. Seat. 
? Biceps:  The circumference of the arm was taken two inches below the 
armpit. (Figure 20) 
 
42 
 
  
 
  Figure 20. Biceps. 
To extract the body measurements a separate Measurement Extraction Profile (MEP) file 
was created based on the ASTM standard with the above mentioned body measurement. 
 
Data Analysis 
1. Research Question 1: Did the body measurements of freshman males and females 
who gained or lost weight change from the beginning of fall 2007 to the end of spring 
2008? 
Weight change was considered the independent variable, and body measurement 
was the dependent variable. The independent variable was grouped into the following 
three categories: (a) weight increased group, (b) weight decreased group, and (c) no 
change group. 
The weight data was initially categorized as ?male? and ?female? data. Then, the 
male data at the two time points was entered in a spreadsheet (as shown in Table 9) to be 
used for the assessment. The differences in weight were calculated by subtracting the 
weight at the beginning of fall 2007 from the end of spring 2008 for each individual. 
43 
 
  
44 
 
Once the differences were calculated, weight differences greater than 2 lb were 
segregated as the weight gained group, and weight differences less than -2 lb were 
grouped as the weight lost group. The weight difference within the range -2 to 2 lb was 
considered as the no change group. This procedure was followed for the female group as 
well. 
After segregating the male group into the weight increased group and weight 
decreased group, the body measurements of each were tabulated in separate spreadsheets. 
An example is shown in Tables 10 and 11. A dependent t-test for paired samples was 
calculated separately for each body measurement according to the weight change group. 
The same procedure was carried out for the female weight increased and decreased 
groups.
  Table 9   Weight Difference Spreadsheet Example 
Code 
Gender 
Weight- 
beginning of
 fall 
2007 (lb) 
Weight- end 
of spring 2008 (lb) 
Weight 
difference 
     
 end of spring - 
beginning of
 
f
a
ll (lb
)
 
Chest -
beginning of
 
fall 2007 
(in.) 
Chest - end 
of spring 
2008 (in.) 
1001 
M 
142 
144 
2 
42 
42.5 
1002 
M 
186 
192 
6 
36 
37 
1003 
M 
176 
172 
-4 
32 
31 
1004 
M 
189 
192 
3 
29 
34 
1005 
M 
133 
144 
11 
40 
43 
45 
      
  
Table 10  
Weight Increased Group Body Measurement Spreadsheet Example 
Code Gender 
Weight difference 
 end of spring - 
beginning of fall 
(lb) 
Chest -beginning of 
fall 2007 (in.) 
Chest - end of 
spring 2008 (in.) 
1002 M 6 36 37 
1004 M 3 29 34 
1005 M 11 40 43 
1008 M 5 42 43 
 
Table 11  
Weight Decreased Group Body Measurement Spreadsheet Example 
Code Gender 
Weight 
difference  
end of spring 
- beginning 
of fall (lb) 
Chest -beginning of 
fall 2007 (in.) 
Chest - end of spring 
2008 (in.) 
1041 M -4 42 43 
1052 M -6 33 32 
1089 M -8 35 33 
1056 M -3 36 36 
 
 
 46  
 
  
 47  
 
2. Research Question 2: Were weight increases or weight decreases of freshman 
males and females related to changes in body measurements? 
To explore the relationship between weight increases or decreases and body 
measurement changes, the weight increased and decreased group weight difference data 
and body measurement data from the beginning of fall 2007 and the end of spring 2008 
were tabulated in a spreadsheet, as shown in Table 12. The difference between each body 
measurement was calculated by subtracting the beginning of fall 2007 measurement from 
the end of spring 2008 measurement. Following this, the differences in weight, bust, 
waist, thigh, biceps, stomach, abdomen and seat were entered in a separate spreadsheet as 
shown in Table 13. The data were tested for its dispersion using the Variance-to-Mean 
Ratio (VMR) test. This test revealed that the data were clumped or clustered in nature (as 
the VMR values were greater than 1). The weight data were then plotted to visually 
identify the number of clusters present. Using the standard k-means clustering algorithm, 
the weight data were clustered into four (or three) groups in accordance with the 
identified number of clusters. Pearson?s correlation was calculated between weight 
difference and each measurement difference for each cluster. This procedure was 
followed separately for clusters within the weight increased and decreased groups for 
males and females.
 Table 12  Weight and Body Measurement Spreadsheet Example 
Code 
Gender 
Weight- 
beginning of
 
fall 2007 (lb) 
Weight- 
end of spring 
2008 (lb) 
Weight 
di
fference end of 
spring - 
beginning of fall (lb) 
Chest -
beginning 
of fall 
2007 (in.) 
Chest - end 
of spring 
2008 (in..) 
Chest 
difference  
end of 
spring - 
beginning 
of fall (in.) 
1001 
M 
142 
144 
2 
42 
42.5 
0.5 
1002 
M 
186 
192 
6 
36 
37 
1 
1003 
M 
176 
172 
-4 
32 
31 
-1 
1004 
M 
189 
192 
3 
29 
34 
5 
1005 
M 
133 
144 
11 
40 
43 
3 
1006 
M 
132 
129 
-3 
34 
35 
1 
48 
 
 
 
 
 
 
 
49 
 
Table 13  Weight Difference and B
ody Measurem
ent Difference Spreadsheet Example 
Code 
Gender 
Weight difference 
end of spring - 
beginning of
 fall (lb)
Bust difference  
end of spring - 
beginning of
 fall 
(in.) 
W
a
ist 
difference end 
of spring - 
beginning of
 
fal
l
 (i
n.) 
Hip difference end of spring - 
beginning of
 
fal
l
 (i
n.) 
1002 
M 
6 
1 
2.5 
1 
1004 
M 
3 
5 
-2 
0.5 
1005 
M 
11 
3 
3 
1.8 
   
 
3. Research Question3: Did the body measurements change for the freshman males 
and females whose weight was stable? 
To identify any changes in the body measurements of participants in the no 
change group, their body measurements were tabulated in a spreadsheet, as shown in 
Table 14. The body measurement differences and a dependent t- test for paired sample 
were calculated. This procedure was followed for males and females.
50 
 
  Table 14   No Change in Weight and Body Measurement Spreadsheet E
x
ample 
Code 
Weight difference end of
 
spring - beginning of fall (lb) 
Chest -beginning of fall 2007 
(in.) 
Chest - end of spring 2008 
(in.) 
1001 
2 
42 
42.5 
1007 
-2 
36 
37 
1100 
0 
32 
31 
1210 
-1 
29 
34 
1123 
1.5 
40 
43 
51 
 
   
 
?
A survey form (ref. Appendix A) was developed which displayed the overlaid 
images with user selectable choices of analysis decisions. Each image pair was 
segmented into four quadrants to aid evaluation. The analysis choices were (a) 
symmetric, where the measurement changes were relatively even around the 
circumference, or (b) asymmetric, where the measurement changes were not even around 
the circumference. The asymmetric category was further divided into four more specific 
choices, viz., (a) more in the back and front than on the sides, (b) most in the back, (c) 
most in the front and (d) most on the sides. The primary researcher first analyzed all 
image sets. Then, two subject experts evaluated the cross sectional slices and verified or 
adjusted the researcher?s choices. 
4. Research Question 4: Do changes in body measurements appear symmetrical or 
asymmetrical around the front, back, and sides of the body when viewed as planar 
slices? 
To visually analyze the symmetry of changes in selected body measurements, 
e.g., bust measurement, planar slices (cross sections) of measurements were extracted for 
both periods from the 3D body scanner, and were overlapped to visually analyze the 
changes in the body measurements. Figure 20 presents an example of the separate and 
overlaid slices. The three major torso measurements of bust/chest, waist, and hip were 
selected for analysis.  
 
 
 
52 
 
 
53 
 
 
Illustration: 
?
 Cross section of Bust ? fall Cross section of Bust ? spring 
 
 
Figure 21. Planar slices of bust. 
Overlapped slices
 
 
 
 
 
 
 
 
 
CHAPTER IV. PRESENTATION AND DATA ANALYSIS 
 
The purpose of this study was to explore possible connections between the weight 
changes and circumferential body measurement changes of adolescents aged 17-19. Data 
were collected for a convenience sample of freshman subjects at two time periods, (a) 
beginning of fall 2007 and (b) end of spring 2008. The collected data were grouped into 
weight gained, weight lost and no change (in weight) groups. This classification was done 
separately for both males and females, and was used for the analysis. 
Subjects and Data Collection 
?
A sample of freshmen (ages 17 to 19) at Auburn University was recruited through 
a convenience sampling method at the beginning of fall semester, 2007.  Potential 
participants were contacted via email, posters, and classroom contact in a variety of 
introductory courses typically taken by freshmen. The participants at the beginning of fall 
2007 were contacted again through e-mail at the end of spring 2008. An incentive of $50 
($25 in the fall and $25 in the spring) was given to the participants. Data for a total of 240 
freshmen were collected at the beginning of fall semester (August and September), 2007. 
Of the 240, 205 participants took part in the study again at the end of spring semester 
(April and May), 2008. Eighteen subjects were dropped from further analysis, and the 
final sample consisted of 187 participants? data (male = 67 and female = 120). Those 18 
subjects who were excluded from further analyses had at least one of the eight 
54 
 
 
measurements missing in the 3D body scanned data due to a digitizing error. Although 
male and female freshmen were recruited in the same manner through contacting courses 
taken by freshman students, more women than men volunteered. 
Data Analysis and Results 
Subjects? Physical Activities 
?
Table 15 shows the mean numbers of days per week and mean hours per day that 
the subjects exercised. They were asked on the questionnaire they completed how 
frequently and how long they participated in vigorous, moderate, and strengthening 
exercise. Male subjects reported doing vigorous physical activities for an average of three 
days per week; their daily average was 53 minutes. For moderate physical activities like 
walking or bicycling, they averaged five days per week at 45 minutes per day. In addition 
to that, they were undertaking strengthening exercises like push-ups, sit-ups or weight 
lifting for an average two days per week for 35 minutes each day. 
Female subjects said that they participated in vigorous physical activities for two 
days per week, with an average of 35 minutes per day. They reported doing moderate 
physical activities like walking or bicycling for five days per week, spending 42 minutes 
per day. Strengthening exercises, like push-ups, sit-ups or weight lifting, were also 
carried out for an average of one day per week for 15 minutes per day.  
 
 
 
55 
 
 
Table 15                                                                                                                         
Physical Activities 
Gender Mean number of days per week of 
physical activity 
Mean minutes per day of physical 
activity 
Vigorous Moderate Strengthening Vigorous Moderate Strengthening 
Male 3 5 2 53 45 35 
Female 2 5 1 35 42 15 
 
Weight Change 
?
For analysis, data of male and female subjects were divided into three groups, 
viz., (a) set of subjects with an absolute weight gain of >2 lbs, called weight gained 
group, (b) set of subjects with an absolute weight loss of >2 lbs, called weight lost group, 
and (c) set of subjects with a weight change between [-2, 2] lbs, called no weight change 
group. Table 16 shows the proportional population in male and female groups who fell 
under the three categories.  The mean weight changes in these three groups were 
determined and tabulated in Table 17. 
More males and females gained than lost weight. Approximately two-thirds of the 
males gained 6.93 lbs. on average. Just under half of the females gained slightly more, 
7.21 lbs. on average. Among males and females, a few more had stable weight than lost 
weight. The ten males who lost weight average 6.34 lbs. The females who lost weight 
averaged less, 3.72 lbs. 
 
56 
 
 
Table 16                                                                                                               
Proportional Subject Population in Weight Lost, Gained and No Weight Change Groups 
 
Group 
 
Subjects 
 
Male Female 
Weight lost 10 (14.9%) 29 (24.2%) 
Weight gained 43 (64.2%) 57 (47.5%) 
No weight Change 14 (20.9%) 34 (28.3%) 
Total 67 (100%) 120 (100%) 
 
 
Table 17  
?
Mean Weight Change    
Group Mean change - Male (lb) Mean change - Female (lb) 
Weight Gained +6.93 +7.21 
Weight lost -6.34 -3.72 
No weight change ?1 ?0.705 
 
 
Research Question 1: Did the body measurements change for freshman males and 
females who gained or lost weight from the beginning of fall 2007 to the end of spring 
2008? 
To answer this question, it had to be determined whether the changes in body 
measurements taken at the two time periods differed significantly in the groups of males 
and females who gained and lost weight. Weight change was considered the independent 
variable, and eight selective body measurements were taken as the dependent variable. 
57 
 
 
58 
 
?
The independent variable was grouped according to the three groups categorized earlier 
as: (a) weight gained group, (b) weight lost group, and (c) no weight change group. Since 
the research question was relevant to only those groups that had weight changes, the two 
groups weight gained and weight lost group were chosen for further analyses.  
Before the data were grouped into weight change categories, male and female 
subjects? data were separated so that an independent analysis based on gender was 
possible. Body measurement data of male subjects from the two time periods (beginning 
of fall 2007 and end of spring 2008) were entered in a spreadsheet according to their 
weight change group. Similarly, the body measurement differences were calculated by 
finding the difference in measurements taken at the two time periods, with end of spring 
2008 measurements as subtrahend. The entire procedure was repeated for the female data 
set and the results were noted.  
Dependent paired t-tests were conducted on the dependent variables for the two 
groups chosen. The alpha value was chosen as 0.05, and the tests were conducted on the 
fall-spring paired data sets. Table 18 shows the comprehensive results for all the groups.
 Table18 
  
Measurement Changes in Male and Female Subj
ects Belonging to the 
Weight Change Groups 
 
       
59
       
M
eas
urem
ent
# ( 
I
n
s)
 
Chest/Bust 
Waist 
Stomach
 
Abdomen
 
Hips 
Seat 
Thigh Ma
x 
Biceps
 
Weight Gained Group 
Ma
le 
M
ean F
a
l
l 
2007 
37.63
 
32.08
 
32.02
 
35.13
 
38.93
 
38.04
 
21.86
 
11.8
 
M
ean 
Spring 2008 
38.89
 
34.02
 
33.77
 
36.68
 
40.45
 
39.35
 
22.65
 
12.41
 
Fa
ll-Spring 
Change 
1.26
 
1.94
 
1.75
 
1.55
 
1.52
 
1.31
 
0.79
 
0.61
 
p-value 
0.2
 
0.004**
 
0.002**
 
0.021*
 
0.004**
 
0.011*
 
0.016*
 
0.001***
 
Fe
ma
le 
M
ean F
a
l
l 
2007 
35.58
 
29.89
 
29.53
 
33.34
 
39.49
 
38.54
 
22.51
 
10.54
 
M
ean 
Spring 2008 
36.49
 
31.1
 
30.42
 
34.
52
 
40.42
 
39.43
 
22.73
 
11 
Fa
ll-Spring 
Change 
0.91
 
1.21
 
0.89
 
1.
18
 
0.93
 
0.89
 
0.22
 
0.46
 
p-value 
0.000***
 
0.000***
 
0.000***
 
0.
002**
 
0.000***
 
0.000***
 
0.364
 
0.004**
 
Weight Lost Group 
Ma
le 
M
ean F
a
l
l 
2007 
38 
32.67
 
32.83
 
36.2
 
39.75
 
38.66
 
22.1
 
11.86
 
M
ean 
Spring 2008 
37.74
 
32.15
 
32.1
 
34.
64
 
39.2
 
37.83
 
21.37
 
12.1
 
Fa
ll-Spring 
Change 
-0.26
 
-0.52
 
-0.73
 
-1.56
 
-0.55
 
-0.83
 
-0.73
 
0.24
 
p-value 
0.452
 
0.231
 
0.149
 
0.
152
 
0.137
 
0.011*
 
0.043*
 
0.366
 
Fe
ma
le 
M
ean F
a
l
l 
2007 
35.37
 
29.44
 
28.93
 
32.83
 
38.93
 
38.02
 
22.18
 
10.65
 
M
ean 
Spring 2008 
35.36
 
28.96
 
28.69
 
33.
24
 
38.67
 
37.69
 
21.59
 
10.27
 
Fa
ll-Spring 
Change 
-0.01
 
-0.48
 
-0.24
 
0.41
 
-0.26
 
-0.33
 
-0.59
 
-0.38
 
p-value 
0.966
 
0.104
 
0.475
 
0.
326
 
0.229
 
0.11
 
0.012*
 
0.001***
 
Note. Significant p values are noted by: *** p 
?
 0.001, ** p 
?
 0.01, *p 
?
 0.05 
  
 
 
Weight gained group: male subjects. Forty-three male participants gained at least 
two pounds. The calculated results show that the mean circumferential change in all eight 
body measurements was an increase. With p-values ranging between 0.001 and 0.021, the 
increases were significant for all measurements except the chest. Torso changes ranged 
from 1.26? at the chest to 1.94? at the waist. The magnitude of change concentrated 
between the hips and stomach was one and a half to nearly two inches. Limb (biceps and 
thigh) circumferential changes were between one-half and one inch. 
?
Weight gained group: female subjects. Fifty-seven female participants gained two 
or more pounds. The calculated results show that except for the Thigh Max, all other p-
values were less than 0.01, making them significant. Hence, measurement changes in the 
bust, waist, stomach, abdomen, hips, seat and biceps showed significant change in the end 
of spring 2008 when compared to the beginning of fall 2007. For females, the significant 
changes were increases in torso circumference measurements ranging from nearly one 
inch to nearly one and a quarter inch. The biceps increase was almost one-half inch. 
Weight lost group: male subjects. Ten male participants came under the weight 
lost group. The calculated p-values for the seat and thigh max measurements were 0.011 
and 0.043, which was less than the set alpha value. Thus, these measurements changed 
significantly (each dropping by approximately three-quarters of an inch) by the end of 
spring 2008 as compared to the beginning of fall 2007. Chest, waist, stomach, abdomen 
and hip measurements decreased between fall and spring, but calculated p-values were 
greater than 0.05, meaning that the changes were not statistically significant. The biceps 
change was actually an increased measurement, but was also not significant. 
60 
 
 
Weight lost group: female subjects. The total number of female participants in the 
weight lost group was 29. Except for the measurements of thigh max and biceps, all 
other measurements showed changes that were not statistically significant when the end 
of spring 2008 was compared to the beginning of fall 2007. The significant drops in 
measurements were approximately three-eighths (biceps) and nine-sixteenths (thigh) 
inch. Except for the abdomen, other measurement changes were decreases, none more 
than one-half inch. 
?
Summary for Research Question 1.For males and females who gained weight, all 
measurements except one in each group (chest for males and thigh for females) increased 
significantly. However, the same was not true for males and females who lost weight. In 
each of those groups, only two measurements decreased significantly (thigh for males 
and females, seat for males, and biceps for females). In the case of females, it seems 
likely that different results for the weight gained and weight lost groups could be 
attributed to the fact that the average weight gain was 7.21 lb while the average weight 
loss was only 3.72 lb. These changes showed up as significant increases in all but one 
body measurement but significant decreases in only two locations. However, no such 
reasoning was possible in the case of male subjects, as their mean weight gain and loss 
numbers differed only marginally, 6.93 and 6.34 lb, but they, too, had significant 
increases in inches in all but one location and significant decreases in only two locations. 
The magnitude of the males? decrease in inches was slightly higher than the females. In 
sum, the research question can be answered mostly in the affirmative for the weight 
gained groups but results were less definitive for the weight lost groups.   
61 
 
 
Research Question 2: Were weight gains or weight losses of freshman males and females 
related to changes in body measurements? 
?
Changes in weights and body measurements for the male and female groups were 
tabulated previously to answer Research Question 1. In order to explore any possible 
statistical relationships between weight gains or losses and body measurement changes, 
correlation analysis was done. Pearson?s correlation coefficients were calculated for the 
data set with one column being the weight difference data and the other column having 
the measurement difference data. The coefficients were calculated for the weight gained 
and weight lost groups, with male and female subjects assessed separately. However, the 
correlation coefficients were less than 0.5 in most of the cases. This poor correlation 
could have been attributed to the fact that the data was over-dispersed, meaning that the 
data set had the characteristics of clumps or clusters. The degree of dispersion was 
estimated using the variance-to-mean ratio (VMR) test. 
VMRs were calculated for the male and female data sets, and the results are 
tabulated in Table 19. From the VMR results, it was concluded that the data were 
clumped. Hence, the data in each group needed to be clustered before evaluating the 
correlation coefficients. Standard k-means clustering algorithm was used to cluster the 
data sets. 
Table 19  
VMR Index 
 
 
 
 
 
Measurements VMR value for weight change 
Weight gain 
Male 3.301382 
Female 3.180168 
Weight loss 
Male 3.1238 
Female 1.555196 
62 
 
 
 
?
Cluster analysis.  The male and female weight gained and lost subdivisions were 
subjected for cluster analysis. The data were arrayed in scatter plots (see Figure 22) to be 
used for a visual examination to determine the number of clusters ?n? within each group. 
The parameter ?n? was used as an argument for the k-means clustering. Then the k-means 
clustering algorithm was used to determine the actual clusters. Figure 23 shows the 
subdivisions, resulting from the k-means clustering algorithm, that were formulated for 
the correlation analysis. Table 20 lists the results of clustering, including the range of 
pounds changed in each cluster and the cluster centroid. In some cases, the cluster with 
the greatest weight change had only one subject in it and therefore could not be used for 
correlation analysis. This was true for the fourth cluster in the female weight gained 
group; because it was significantly deviant from its preceding cluster (centroid of cluster 
3 was at 17.07 and centroid for cluster 4 was 29.4) the subject was considered an outlier.  
The fourth cluster for the male and female weight lost groups also had only one subject. 
 
 
 
 
 
63 
 
 
 
Weight Gained Group (a) Male Subjects  (b) Female Subjects 
 
? Weight Lost Group (a) Male Subjects  (b) Female Subjects 
 
Figure 22. Scatter plot of weight changes in male and female subjects. 
 
 
                             
Male Data (n = 67) Female Data (n = 120) 
Weight 
Gained
n = 43 
Weight 
Lost      
n = 10 
No 
Weight 
Change  
n = 14 
Weight 
Gained 
n = 57 
Weight 
Lost       
n = 29 
No 
Weight 
Change
n = 34 
Clusters 
= 4 
Clusters 
= 3 
-na- 
Clusters 
= 3 
Clusters 
= 3 
-na- 
-na- refers to the data set not included for the current analyses 
Data Set (n = 187) 
 
Figure 23. Subdivisions formulated for correlation analysis. 
 
 
64 
 
 
 
 
Table 20  
Weight Range and Centroid for Male and Female Subjects Belonging to the Weight 
Change Groups 
 
 
?
Cluster Number 1 2 3 4 
Weight 
Gained 
Group 
Male 
Cluster Span in 
terms of weight 
difference (lb) 
2.2-5 5.6-10.6 12.8-17.6 
19.4-
21.6 
Cluster Centroid 3.49 7.61 15.25 20.4 
Number of 
Subjects 
21 16 4 2 
Female 
Cluster Span in 
terms of weight 
difference (lb) 
2-6.6 6.8-13 15.4-20.4 -na- 
Cluster Centroid 4.165 9.1818 17.0667 -na- 
Number of 
Subjects 
31 22 3 -na- 
Weight 
Lost 
Group 
Male 
Cluster Span in 
terms of weight 
difference (lb) 
2.2-3.6 4.8-7 12.5-15.2 -na- 
Cluster Centroid 2.6 6.3 12.6 -na- 
Number of 
Subjects 
4 4 2 -na- 
Female 
Cluster Span in 
terms of weight 
difference (lb) 
2-3.4 4.2-5.4 5.8-8.6 -na- 
Cluster Centroid 2.41 5.14 7.433 -na- 
Number of 
Subjects 
12 9 7 -na- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
65 
 
 
66 
 
?
Weight gained group: male subjects.  Table 21 shows the correlation analyses 
results for the male subjects who were in the weight gained group. The correlation results 
showed that in Cluster 3 (containing four subjects who gained from 12.8-17.6 lb) there 
was a very strong and significant positive correlation between weight changes and 
changes in abdomen and hip measurements (respective p-values of 0.025 (r =0.951) and 
0.044 (r = 0.912)). Although there was a strong association between weight change and 
seat and waist measurement changes, it was not statistically significant (a calculated p-
value of 0.145 (r = 0.855) and 0.186 (r = 0.814), respectively). Moderate negative 
correlations were seen in the cases on chest and biceps changes; they were not 
statistically significant. Clusters 1 and 2 containing those subjects who gained 2-10.6 lb, 
showed weak correlations in all cases. While the results showed that the correlation was 
perfect for all the measurements in Cluster 4, the data size of two was too limited for 
meaningful interpretation. 
Correlation analysis.  Weight change in the clusters of each group was correlated 
against the eight measurement changes, and the Pearson?s correlation coefficients were 
determined. The following guidelines ("The strength and significance of the coefficient," 
2001) were used to interpret the strength of relationships: 
0.0 to 0.2 Very weak to negligible correlation 
0.2 to 0.4 Weak, low correlation (not very significant) 
0.4 to 0.7 Moderate correlation 
0.7 to 0.9 Strong, high correlation 
0.9 to 1.0 Very strong correlation 
 
 Table 21  Correlation Coefficient for Weigh
t
 Gained Group Males
 
 Cluster
 
?
 
Measure
ment
#
 
?
 
Chest
 
Waist
 
Stomach
 
Abdom
en
 
Hip
 
Seat
 
Thigh max
 
Biceps
 
1
 
Range
 
(-5.93 ) - 7.28?
(-4.53) - 7.57?
(-4.84) - 7.57?
(-5.57) - 8.81?
(-3.17) - 8.10?
 
(-2.97) - 7.00?
(-3.18)  - 6.21?
(-1.29)  - 3.01?
r
 
0.05
4
-0.06
0.15
1
0.05
0.03
5
 
0.07
5
-0.023
-0.034
p-value
 
0.81
5
0.98
0.51
3
0.82
9
0.88
2
 
0.74
6
0.92
2
0.88
3
2
 
Range
 
(-6.02) - 5.45?
(-2.41) - 21.0
6?
(-4.65) - 15.7
9?
(-4.06) - 20.9
4?
0.26 - 17.8
7?
 
0.47 - 18.4
0?
(-0.67) - 8.03?
(-.19) - 5.51?
r
 
-0.215
-0.36
-0.332
-0.273
-0.264
 
-0.332
-0.429
-0.226
p-value
 
0.42
4
0.17
1
0.20
9
0.30
6
0.32
3
 
0.20
9
0.09
7
0.4
3
 
Range
 
1.27 - 2.70?
0.99- 2.72?
1.59- 2.33?
1.04-3
.
36?
0.61- 2.59?
 
0.94- 2.89?
0.80- 1.42?
0.37- 0.91?
r
 
-0.699
0.81
4
0.01
2
0.95
1
0.91
2
 
0.85
5
0.46
1
-0.681
p-value
 
0.30
1
0.18
6
0.98
8
.025
*
.044
*
 
0.14
5
0.53
9
0.31
9
4
 
Range
 
2.28- 7.83?
1.70- 1.83?
2.54-2
 
.62?
2.88-8
.
23?
2.30- 2.32?
 
2.04- 2.36?
1.29- 1.30?
0.59- 1.28?
r
 
1
 
1
-1
1
1
 
1
-1
-1
p-value
 
0.00
0**
*
0.00
0**
*
0.00
0**
*
0.00
0**
*
0.00
0**
*
0.00
0**
*
0.00
0**
*
0.00
0**
*
67
 
Note. Significant p values are noted by: *** p 
?
 0.001, ** p 
?
 0.01, *p 
?
 0.05 
    
 
 
Weight gained group: female subjects. Similar to the male data, female subjects? 
weight change data were subdivided into four clusters using the k-means clustering 
algorithm. Three usable clusters emerged (see Table 22). In the cases of Clusters 1 and 2 
(respectively, gains of 2-6.6 and 6.8-13 lb), the associations between weight and 
measurement change were weak for all the measurements and were not statistically 
significant. For Cluster 3, in which the three subjects gained 15.4-20.4 lb, the associations 
between weight change and changes in abdomen and biceps were strong and positive, 
whereas the association between weight change and changes in bust was strong and 
negative. All other relationships between weight change and changes in waist, hip, seat 
and thigh max were moderately strong and positive. Although this cluster exhibited some 
strong associations, none of them were statistically significant. 
68 
  Table 22  Correlation Coefficient for Weigh
t
 Gained Group females
 
 Cluster
 
?
 
Measure
ment
#
 
?
 
Chest
 
Waist
 
Stomach
 
Abdom
en
 
Hip
 
Seat
 
Thigh max
 
Biceps
 
1
 
Range
 
(-0.88) -2.85?
(-1.81) -3.20?
(-2.56)- 4.65?
(-8.19) - 6.29?
(-2.60) -6.23?
 
(-2.14) -3.93?
(-5.69)- 6.72?
(-0.92)- 1.23?
r
 
0.14
2
0.29
2
0.21
2
0.13
6
-0.024
 
0.01
7
0.26
0.03
4
p-value
 
0.44
6
0.11
1
0.25
3
0.46
6
0.89
8
 
0.92
9
0.15
7
0.85
4
2
 
Range
 
(-3.48) -2.95?
(-2.16) -2.76?
(-2.81) - 2.33?
(-7.31) - 6.74?
(-3.30) -2.24?
 
(-2.67) -2.65?
(-3.43) -1.76?
(-1.65) - 1.27?
r
 
0.14
8
0.28
5
0.22
5
0.02
6
0.15
1
 
0.17
3
-0.08
0.01
6
p-value
 
0.51
1
0.19
9
0.31
4
0.90
8
502
 
0.44
2
0.72
2
0.94
5
3
 
Range
 
1.12- 3.19?
0.10- 5.01?
0.05- 5.77?
2.71-5
.
30?
1.46- 2.48?
 
1.80- 2.51?
0.16- 2.15?
0.21- 1.05?
r
 
-0.868
0.45
1
-0.143
0.94
7
0.62
6
 
0.57
6
0.58
5
0.76
1
p-value
 
0.33
0.70
2
0.90
9
0.20
9
0.56
9
 
0.60
9
0.60
3
0.44
9
69
 
Note. Significant p values are noted by: *** p 
?
 0.001, ** p 
?
 0.01, *p 
?
 0.05 
      
 
  
Weight lost group: male subjects. For males who fell into the weight lost group, 
three usable clusters were formed using the k-means clustering. In Cluster 1 (see Table 
23) with four subjects who lost 2.2-3.6 lb, a very strong negative relationship existed 
between weight loss and change in stomach circumference and a very strong positive 
relationship between weight loss and change in biceps circumference. The p-value for 
stomach was 0.004 and for biceps 0.007, making them statistically significant. A strong 
positive relationship existed between losses in weight and waist circumference, and a 
moderate positive relationship existed between weight loss and decreases at the abdomen, 
hip and seat. These moderate to strong relationships were not significant.  
In the case of Cluster 2, which was composed of four subjects who lost 4.8-7 lb, 
changes in waist and biceps had a very strong positive relationship with weight change. 
Their p- values of 0.015 and 0.020 (respectively) were statistically significant. No other 
relationships in this cluster were significant. However, changes in abdomen, hip and seat 
measurements had strong positive relationships with weight change. A strong negative 
association existed between weight and chest measurement changes. Cluster 3 had only 
two subjects.  
 
 
 
 
 
 
 
 
 
 
 
 
 
70 
 
 Table 23  Correlation
 Coefficient for Weight L
ost Group Males
 
 Cluster
 
?
 
Measure
ment
#
 
?
 
Chest
 
Waist
 
Stomach
 
Abdom
en
 
Hip
 
Seat
 
Thigh max
 
Biceps
 
1
 
Range
 
(-0.75) - 0.40?
 
(-0.82)  -
  
(-.18)?
(-0.21) - 1.30?
(-6.78) -
 
   
(-0.21)?
0.04-  
  
0.98?
(-1.31)- 0.79?
(-0.51)-
  
  
(-0.12)?
(-0.66)- 0.33?
r
 
0.04
5
 
0.81
-0.996
0.56
4
0.42
2
0.61
9
-0.213
0.99
3
p-value
 
0.95
5
 
0.19
.004
**
0.43
6
0.57
8
0.38
1
0.78
7
.007
**
2
 
Range
 
(-2.36) - 0.28?
 
(-2.16)- 0.40?
(-3.09) - 0.02?
(-1.63)-
  
   
(-0.55)?
(-2.19)-
  
  
(-0.22)?
(-1.84)-
  
  
(-0.19)?
(-1.42) - 0.91?
(-0.27)- 1.02?
r
 
-0.784
 
0.98
5
0.62
2
0.73
1
0.72
1
0.88
2
-0.059
0.98
p-value
 
0.21
6
 
.015
*
0.37
8
0.26
9
0.27
9
0.11
8
0.94
1
.020
*
3
 
Range
 
(-0.93)- 1.59?
 
(-0.92)- 2.44?
(-1.70)- 0.47?
(-1.94)- 4.49?
(-1.48)-(- 0.54)?
(-1.12)-(- 0.69)?
(-2.77)-(- 0.22)?
(-0.28)- 3.12?
r
 
-1
 
-1
-1
-1
-1
-1
 
-1
-1
p-value
 
0.00
0**
*
 
0.00
0**
*
0.00
0**
*
0.00
0**
*
0.00
0**
*
 
0.00
0**
*
0.00
0**
*
0.00
0**
*
71
Note. Significant p values are noted by: *** p 
?
 0.001, ** p 
?
 0.01, *p 
?
 0.05 
 
 
Weight lost group: female subjects. Table 24 shows the Pearson correlation 
coefficients for the three usable clusters in the females weight lost group. Both positive 
and negative relationships between weight and measurement change were weak for the 
12 subjects who lost 2-3.4 lb. in Cluster 1. For Cluster 2 (nine subjects who lost 4.2-5.4 
lb), positive and moderate associations existed between weight and abdomen, hip and 
seat changes. The p-value for the seat association was 0.046, thereby making the 
relationship between weight change and seat statistically significant. The relationship 
between weight and waist change was moderate and negative. It and all other 
relationships were not significant. Finally, for Cluster 3 (seven subjects who lost 5.8-8.6 
lb), a moderate negative relationship was seen between weight and all measurement 
changes except thigh max, which had a strong and significant (p=0.012) negative 
relationship with weight change.  
72 
 Table 24  Correlation Coefficient for Weigh
t
 Lost Group Females
 
 
 Cluster
 
?
 
Measure
ment
#
 
?
 
Chest
 
Waist
 
Stomach
 
Abdom
en
 
Hip
 
Seat
 
Thigh max
 
Biceps
 
1
 
Range
 
(-1.02) - 0.93?
 
(-1.52)- 1.62?
 
(-1.22)- 1.24?
 
(-2.11)- 3.69?
 
(-0.71)- 0.53?
 
(-0.55)- 0.02?
 
(-0.80)- 0.54?
 
(-0.91)- 0.06?
 
r
 
0.28
6
0.16
1
0.01
6
-0.015
-0.075
 
-0.126
-0.097
-0.179
p-value
 
0.36
8
0.61
8
0.96
1
0.96
3
0.81
7
 
0.69
5
0.76
4
0.57
8
2
 
Range
 
(-1.06)- 2.85?
 
(-1.18)- 0.52?
 
(-0.91)- 1.72?
 
(-2.01)- 4.30?
 
(-1.04)- 2.96?
 
(-1.21)- 2.26?
 
(-5.40)-(- 0.50)?
 
(-1.70) - 0.25?
 
r
 
-0.266
-0.319
-0.3
0.45
4
0.65
4
 
0.67
4
-0.417
-0.062
p-value
 
0.48
8
0.40
3
0.43
2
0.22
0.05
6
 
.046
*
0.26
5
0.87
3
3
 
Range
 
(-1.36 )- (-0.10)?
 
(-4.70)- (-0.52)?
 
(-4.87)- (-0.51)?
 
(-3.54)- (-0.26)?
 
(-1.37)- (-0.27)?
 
(-1.28)- (-0.18)?
 
(-0.81)- (-0.37)?
 
(-1.36)- (-0.65)?
 
r
 
-0.605
-0.421
0.51
3
-0.531
-0.682
 
-0.671
-0.786
-0.505
p-value
 
0.15
0.34
7
0.23
9
0.22
0.09
2
 
0.09
9
.012
*
0.24
8
73 
Note. Significant p values are noted by: *** p 
?
 0.001, ** p 
?
 0.01, *p 
?
 0.05 
 
 
 
Summary for Research Question 2. Among male and female weight gained and 
weight lost groups, correlation results were both positive and negative. Thus, those who 
gained and lost weight had increased and decreased measurements, although the changes 
could be small. There were a limited number of significant relationships, and these 
occurred primarily among the limited number of subjects who gained or lost more than 
the smallest cluster of weight. Significant positive correlations were found in the 
following cases:  
? For increases at the abdomen and hip for males who gained 12.8-17.6 lb. 
? For a decrease at the biceps for males who lost 2.2-3.6 lb. 
? For decreases at the chest and biceps for males who lost 4.8-7 lb. 
? For a decrease at the seat for females who lost 4.2-5.4 lb. 
Significant negative correlations were found in the following cases: 
? For an increase at the stomach for males who lost 2.2-3.6 lb. 
? For a slight decrease at the thigh for females who lost 5.8-8.6 lb. 
The mixed results for the data analyses do not allow a clear affirmative or 
negative response to the research question. They suggest that changes in circumference 
measurements can be related changes in weight, but that may not always be in the same 
direction. 
 
 
 
 
 
 
 
 
 
74 
 
 
 
 
Research Question 3: Did the body measurements change for the freshman males and 
females whose weight was stable? 
 
To identify any changes in the body measurements of participants in the no 
weight change group, their body measurements were tabulated in a spreadsheet. Weight 
losses and gains of up to two pounds were considered as no change. A dependent t-test 
was done to determine whether or not any significant changes in body measurements 
taken at beginning of fall 2007 and end of spring 2008 occurred. This procedure was 
followed for male and female groups. The results for male and female groups are 
tabulated in Table 25. 
The results show that the p-values for all the measurements were greater than the 
set alpha value (0.05). There were no significant changes in the measurements of male 
participants having stable weight. Similar to the male group, females with stable weight 
also had no significant changes in their body measurements. Based on these results, the 
response to the research question was negative.  
 
 
 
 
 
 
 
 
75 
 
 
Table 25  
Weight No Change Group 
Weight 
no 
change 
male 
N= 14 
Measurement  Chest/Bust Waist Stomach Abdomen Hip Seat Thigh Biceps 
Mean - 
Beginning of 
fall 2007 
36.88 31.95 32.06 35.65 39.39 38.43 22.13 12.03 
Mean- End 
of spring 
2008 
38.2 32.39 32.91 36.4 39.74 38.71 22.17 12.75 
Magnitude of 
change 
between the 
two time 
points 
1.32 0.44 0.85 0.75 0.35 0.28 0.04 0.72 
p - value 0.078 0.236 0.11 0.456 0.264 0.329 0.806 0.321 
Weight 
no 
change 
female 
N = 34 
Mean - 
Beginning of 
fall 2007 
35.59 29.61 29.37 33.3 39.08 38.11 22.58 10.42 
Mean- End 
of spring 
2008 
35.62 29.59 29.18 33.41 39.4 38.3 22.19 10.4 
Magnitude of 
change 
between the 
two time 
points 
0.03 -0.02 -0.19 0.11 0.32 0.19 -0.39 -0.02 
p - value 0.078 0.236 0.11 0.456 0.264 0.329 0.806 0.321 
 
 
 
Note. Significant p values are noted by: *** p ? 0.001, ** p ? 0.01, *p ? 0.05 
 
 
 
 
 
 
 
 
 
 
 
76 
 
 
Research Question 4: Do changes in body measurements appear symmetrical or 
asymmetrical around the front, back, and sides of the body when viewed as planar slices? 
To visually analyze the symmetry of changes in body measurements, e.g., to see if 
an individual?s original and new circumferences were shaped the same, indicating an 
equality of change around the body, planar slices (cross-sections) of three key 
circumferences (bust/chest, hip and waist) were extracted for fall and spring scans and 
overlaid.  To analyze the changes, a form for recording answers relative to each slice was 
developed. It contained the overlapped planar images and the following user-selectable 
(multiple choice) responses for each slice: 
Symmetrical: Relatively even change around body circumference  
Asymmetrical:      
1. More in the back and front than on the sides 
2. Most in the back 
3. Most in the front 
4. Most on the sides 
The analysis was done for males and females who either gained or lost weight. 
The no change group was not included for analyses. Each pair of overlaid slices was 
visually assessed for symmetry or type of asymmetry by the researcher. Subsequently, his 
major professors reviewed each assessment to affirm or change it. When the researcher?s 
assessment was changed, it was a consensus discussion and decision by the major 
professors. A symmetrical designation did not imply perfect symmetry; rather, it implied 
that the overlaid slices were more symmetrical than they were asymmetrical. Examples of 
each designation are shown in figure 24. 
 
 
 
77 
 
 
 
                                        
                      Symmetry                         More in the back and front than on the sides 
 
                                        
               Most in the front                          Most in the back 
 
 
 
            Most on the sides 
 
 
 
 
Figure 24. Examples of symmetrical and asymmetrical shapes of waist measurement. 
Weight gained group: male subjects. Table 26 shows the number of subjects 
classified as symmetrical and asymmetrical, with the later being broken out into the 
78 
 
 
options for asymmetry; the incidence percentage was included to ease within and across 
group comparisons. At the chest, 65% of the subjects were symmetrical. At the hip and 
waist, more were asymmetrical (55.8% and 53.5% respectively) than symmetrical. The 
most common type of asymmetry differed at each circumference location. At the chest, 
16.3% showed the most change on the sides. At the At the hip, 20.9% showed the most 
change at the front and back rather than on the sides. At the waist, 25.6% showed the 
most change in the front. 
Table 26  
Number of Subjects in Each Shape Classification for Weight Gained Males 
Subjects in symmetric and asymmetric classifications 
N = 43 
Relatively 
even change 
around body 
circumference 
 
 
 
Asymmetrical
 
More in 
the  
back and 
front 
than on  
the sides 
Most in 
the back 
Most in 
the front 
Most on 
the sides 
Chest 
28 (65.1%) 15(34.9%) 1(2.3%) 3(7%) 4(9.3%) 7 (16.3%) 
Hip 
19 (44.2%) 26(55.8%) 9 (20.9%) 5(11.6%) 3(7%) 7(16.3%) 
Waist 
20 (46.5%) 23(53.5%) 5(11.6%) 5(11.6%) 11 
(25.6%) 
2(4.7%) 
 
Weight gained group: female subjects. Table 27 shows the number of subjects in 
symmetrical and asymmetrical categories for weight gained females. About half (50.9%) 
of the subjects were symmetrical at the bust. Hip and waist measurements showed 
asymmetric changes in 70.2% and 63.2% of the subjects. In the cases of asymmetric 
changes, no pattern emerged as to the most common type of asymmetry. At the bust, 
24.6% showed the most change in the front. At the hip, 26.3% showed the most change at 
79 
 
 
the front and back than on the sides. In the waist changes, 29.8% showed the most in the 
back. 
Table 27 
 Number of Subjects in Each Shape Classification for Weight Gained Females 
Subjects in symmetric and asymmetric classifications 
N=57 
Relatively 
even change 
around body 
circumference 
 
 
Asymmetrical
 
More in 
the  
back and  
front than 
on  
the sides 
Most in 
the back 
Most in 
the front 
Most on 
the sides 
Bust 
29 (50.9%) 28(49.1%) 7(12.3%) 4(7%) 14 
(24.6%) 
3(5.3%) 
Hip 
17 (29.8%) 40(70.2%) 15 (26.3%) 8(14%) 5(8.8%) 12 
(21.1%) 
Waist 
21 (36.8%) 36(63.2%) 3(5.3%) 17 
(29.8%) 
9(15.8%) 7(12.3%) 
 
Weight lost group: male subjects. Table 28 shows the number of weight lost male 
subjects in different classifications of symmetry or asymmetry. About 70% and 50% of 
the subjects were symmetrical at the chest and hip. At the waist, more were asymmetrical 
(60%) than symmetrical. In this group of subjects too, the most common type of 
asymmetry varied at each circumferential measurement. At the hip, most changes 
occurred on the sides (30% of the subjects) and at waist, 30% showed most changes in 
the front.  
 
 
80 
 
 
Table 28  
Number of Subjects in Each Shape Classification for Weight Lost Males 
 
Subjects in symmetric and asymmetric classifications 
N=10 
Relatively 
even change 
around body 
circumference 
 
 
 
Asymmetrical
More in the 
back and  
front than 
on  
the sides 
Most in 
the 
back 
Most in 
the 
front 
Most on 
the sides
Chest 
7 (70%) 3(30%) 0 1(10%) 1(10%) 1(10%) 
Hip 
5 (50%) 5(50%) 1(10%) 0 1(10%) 3 (30%) 
Waist 
4 (40%) 6(60%) 1(10%) 1(10%) 3 (30%) 1(10%) 
Weight lost group: female subjects. For the weight lost female group, symmetrical 
and asymmetrical changes at the bust were almost equally distributed among the subject 
population (48.3% and 51.7%, respectively; ref. Table 29). The most common 
asymmetric changes were at the front in the cases of bust and waist. A notable 72.4% 
were asymmetrical at the hip, with front and back asymmetry being the most common. 
 
 
 
 
 
 
81 
 
 
Table 29  
Number of Subjects in Each Shape Classification for Weight Lost Females 
 Subjects in symmetric and asymmetric classification 
N= 29 
Relatively 
even change 
around body 
circumference 
Asymmetrical
More in 
the back 
and front 
than on 
the sides 
Most in 
the back 
Most in 
the front 
Most 
on the 
sides 
Bust 
14 (48.3%) 15(51.7%)  4(13.8%) 3(10.3%) 8 (27.6%) 0 
Hip 
8 (27.6%) 21(72.4%)  11 
(37.9%) 
2(6.9%) 6(20.7%) 2(6.9%) 
Waist 
15 (51.7%) 14(48.3%)  1(3.4%) 5(17.2%) 8 (27.6%) 0 
 
Summary for research question 4. Among males who gained and lost weight, 65-
70% showed symmetrical change at the chest; approximately one-half of the females had 
symmetrical change at the bust. At the hip, 44-50% of the males had symmetrical change, 
whereas slightly less than 30% of the females did. At the waist, 40-46.5% of the males 
had symmetrical change. Approximately half of the females who lost weight showed 
symmetrical change at the waist, but fewer who gained weight, 37%, had symmetrical 
change. Thus, symmetrical change appears to have been more common at the chest or 
bust than at the hip or waist.  
The placement of asymmetry was not consistently patterned across locations for 
males or females. In the only case where a single asymmetry category was more 
commonly found than symmetry, it was at the front and back of the hip for females who 
lost weight. This placement was also more common for males and females who gained 
weight, but not the few males who lost weight. At the waist, asymmetry placement was 
82 
 
 
83 
 
more often in the front except for females who gained weight, where placement was more 
often in the back. Just three of the ten males who lost weight showed asymmetry at the 
chest, each at a different area. Males who gained weight showed more change at the sides 
of the chest, whereas females who gained and lost weight showed the most asymmetry in 
the front of the bust.  
Thus, the results for this research question were mixed. It appears that either 
symmetrical or asymmetrical change can be more or equally common. Additionally, 
placement of asymmetry appears to vary.  
 
 
 
 
 
 
 
 
 
 
 
CHAPTER V.  SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 
This research studied the impact of weight change on body circumference 
measurements in freshman students. One impetus was the ?Freshman 15?, the 
hypothetical phenomenon that freshmen gain 15 lb of weight in their first year of college. 
A limited number of other researchers have studied the phenomenon. While there has 
been no conclusive evidence that this degree of weight gain actually occurs across a 
significant number of students, most of the studies (Anderson et al., 2003; Delinsky & 
Wilson, 2008; Drawdy, 2007; Hoffman et al., 2006; Holm-Denomo, et al., 2008; 
Levitsky et al., 2004; Mihalopoulus et al., 2008; Morrow et al., 2006; Saunders, 2008) 
have found significant weight increases (ranging from 1.7 lb to 6.82 lb) among freshman 
students. The research reported here was part of a larger study of weight change in 
college students that incorporated additional physical, psychological, and environmental 
measures. The purpose of the segment discussed here was to increase understanding of 
whether there appear to be patterns as to where on the body measurements change with 
weight gain or loss and what amounts of body measurement change may be associated 
with weight change.  
 
 
 
84 
 
 
Summary and Discussion 
Data Collection and Sample 
Data collection was done at Auburn University, Alabama, at two time periods: (a) 
the beginning of fall semester 2007 and (b) the end of spring semester 2008, from male 
and female subjects; the results were analyzed separately. All subjects were entering 
freshmen who were 17-19 years old. A total of 240 subjects (156 females, 84 males) 
volunteered for the study at the beginning of fall 2007 semester. At the end of spring 
semester, 205 subjects returned for the follow-up assessments. Of the 205 subjects, 18 
subjects were dropped from the analysis due to missing body measurements. Thus, 187 
subjects (120 females, 67 males) were used in the data analysis. 
Data collected on both sexes were categorized into the following three groups for 
analysis:  (a) weight gained group, (b) weight lost group, and (c) no weight change group. 
The no weight change group was defined as those whose weight was within a 2 lb 
increase or decrease range. 
Changes in Weight and Body Circumference Measurements 
Fall to spring weight gain and loss.  Between the beginning of fall 2007 and the 
end of spring 2008, 64.2% of male subjects (43) and 47.5% of female subjects (57) 
gained weight. Their respective weight gains, on average, were 6.93 and 7.21 lb. A 
relatively smaller proportion of the participants [14.9% of the male subjects (10) and 
24.2% of the female subjects (29)] showed weight loss. The mean weight loss in male 
subjects was 6.34 lb and that of female subjects was 3.72 lb. After the data for those who 
gained and lost weight were evaluated using the VMR test, they were clustered using K-
85 
 
 
means clustering algorithm. Among those who changed weight, this resulted in three to 
four segments within each of the weight gained and weight lost groups; these were as 
follows: 
1. Among males who gained weight:  21 who gained 2-2.5 lb; 16 who gained 5.6-
10.6 lb; four who gained 12.8-17.6 lb, and two who gained 19.4-21.6 lb. 
2. Among females who gained weight: 31 who gained 2-6.6 lb; 22 who gained 
6.8-13 lb; three who gained 15.4-20.4 lb. 
3. Among males who lost weight: four who lost 2.2-3.6 lb; four who lost 4.8-7 lb; 
two who lost 12.5-15.2 lb. 
4. Among females who lost weight: 12 who lost 2-3.4 lb; nine who lost 4.2-5.4 lb; 
seven who lost 5.8-8.6 lb. 
Fall to spring body measurement differences.  Eight body circumferences were 
analyzed for change. Two were limbs (thigh and biceps), and the other six were torso 
locations (chest/bust, waist, stomach, abdomen, hips, and seat). Among the subjects 
whose weight was stable, mean changes in body circumferences were under an inch in all 
but one case (males? chest increased 1.32 in.). Among males and females who gained 
weight, seven of eight measurements increased significantly between fall and spring. The 
chest for males and the thigh for females increased but not significantly. Thus, mean 
weight gains of approximately 7 lb seemed to appear as recognizable body measurement 
increases at nearly all locations identified for possible change.  
86 
 
 
Among male and female subjects who lost weight, two of the eight measurements 
(seat and thigh max for males, and biceps and thigh max for females) showed significant 
decreases. Slight increases of approximately 0.24 in. at the biceps for male subjects and 
0.41 in. at the abdomen for female subjects were also seen. Male subjects reported doing 
strengthening exercises, such as push-ups, sit-ups and weight lifting, for 2 days per week 
with an average of 35 minutes per days. Since these forms of exercise can simultaneously 
increase muscular measurements and influence weight loss, they may explain the biceps 
measurement increase. In weight lost females, there was an increase in the abdomen 
measurement. This could be an example of shifting fat distribution patterns mentioned by 
Croney (1971).  
Association between weight and body measurement changes.  In addition to 
assessing the significance of measurement differences over time, the association of 
measurement and weight changes was analyzed. Correlation analysis on the clustered 
segments within the weight gained and weight lost groups showed mixed results. Both 
male and female weight gained and lost groups showed positive and negative 
relationships, and most of those were not statistically significant. The mixture of results 
could be attributed to the fact that the sample sizes in each of the clusters were small, and 
some of the measurement changes were minimal. Among the clusters with the least 
weight gain or loss (?2-6.6 lb.), there were only two significant relationships. The other 
six cases of significant correlations were found in the second or third clusters of weight 
change (?4.2-20.4 lb.).  
Positive correlations were seen for three cases of male subjects: (a) between 
weight gain and measurement changes at the abdomen and hip (Cluster 3, 12.8-17.6 lb); 
87 
 
 
(b) between waist and biceps changes for weight loss (Cluster 2, 4.8-7 lb); and (c) 
between biceps changes and weight loss (Cluster 1, 2.2-3.6 lb).  In the case of female 
subjects, no significant correlations were found between weight gain and measurement 
changes. However, weight lost females (Cluster 2, 4.2-5.4 lb) showed a positive 
correlation with seat measurement changes. Significant negative correlations were seen in 
the case of weight lost (Cluster 1, 2.2-3.6 lb) males for stomach measurements and in the 
case of weight lost females (Cluster 3, 5.8-8.6 lb) at the thigh max.  
Implications of body measurement changes.  Poehlman et al. (1995) reported an 
average increase in waist circumference of 0.28 cm (about 1/8 in.) per year for women 
and 0.18 cm (about 1/16 in.) per year for men. Among males and females in this study 
who gained weight, the largest incremental change was at the waist (+1.92 in. and +1.21 
in. respectively). The stomach (+1.75 in., measured between waist and chest) was second 
for males, and the abdomen (+1.18 in., measured at the fullest frontal location between 
the waist and the crotch) was second for females. Among males who lost weight, the 
largest decrease (-1.56 in.) was at the abdomen, and for females who lost weight, at the 
thigh (-0.59 in.). Past research on fat distribution patterns has suggested that fat is 
redistributed from the limbs to the trunk during adolescence, and fat can continue to build 
in that area during the adult years (Baumgartner et al., 1986; Brozek, 1952; Croney, 
1971; Mueller, 1982; Ramirez, 1993; Ramirez & Mueller, 1980; Thibodeau, 1990). The 
freshmen in this study were older adolescents moving towards adult maturation. 
The American College Health Association and U.S. Department of Health and 
Human Services formulated 178 health-related objectives specifically for college students 
(?Healthy Campus 2010: Making it Happen?, 2002). Of the ten major health issues 
88 
 
 
quoted by these organizations, ?Physical Activity? and ?Overweight and Obesity? were at 
the top. While growth is a natural phenomenon, abnormal changes in body fat and fat 
accumulated measurement changes are considered health disparities. For example, an 
increase in abdominal circumference was found to be one of the indicators for mild 
cardiac risks (Atsumi, Tetsuro & Hiroyuki, 2005). Researchers (Santos, Cintra & Martini, 
2008) found a correlation between accumulated trunk fat and insulin resistance, a 
metabolic disorder. Wells, Treleaven, and Cole (2007) studied the relationship between 
various body measurements in the context of health and disease risk factors. They used 
the 3D body scanner to extract body circumference measurements. They found that chest 
and waist measurements were significantly associated with BMI in the male subjects. 
BMI of female subjects was shown to have a significant association with hip and bust 
measurements. The research results reported here showed that circumference 
measurements from the stomach to the seat increased significantly in the male and female 
weight gain groups; these findings are related to health researchers? focus on 
measurement changes in the trunk region.  
 When measurements changed, the amounts were not consistently similar at 
different circumference locations. An average of a 0.91 in. increase in the bust, 1.21 in. 
increase in the waist, 1.18 in. in the abdomen, and 0.93 in. at the hips for weight gained 
females would mean that a woman with those changes could find her apparel size 
changing by one or two increments at different locations, according to the ASTM Misses 
Standard Table (ASTM, 2001). Similarly, a man with the average increases of  1.26 in. at 
the chest, 1.94 in. at the waist,  1.55 in. at the abdomen, and 1.52 in. at the hips ,would 
have uneven changes in his garment size (ASTM, 2004). Among females who on average 
89 
 
 
lost 3.72 lb, changes in bust, waist, abdomen and hips were less than ? in.; this would not 
alter their size.  Weight lost males (6.34 lb) had an average decrease of 0.26 in. at the 
chest, 0.52 in. at the waist, 1.56 in. at the abdomen, and 0.55 in. at the hips. According to 
the ASTM standard tables (?D6240-98?, ASTM, 2004), a 1.56 in. decrease at the 
abdomen could change apparel size by one or two increments, but the waist measurement 
decrease 0.52 in. could obviate the need for a size change. 
Relating measurement differences and associations.  Weight gain occurred in 
both male and female subjects (6.93 lb, and 7.21 lb, respectively). These amounts showed 
an impact on body measurement changes. In males all trunk measurements increased by 
at least one inch; chest measurement, the only change that was not statistically significant 
(p = 0.2). increased by 1.26 in. The two limb measurements showed at least a ? in. 
incremental change. Despite largely significant measurement differences, the abdomen 
and hip changes were the only places where positive correlations with weight gain were 
found.  In the case of female subjects with weight gain, all the measurements except thigh 
max showed significant increases. The entire trunk region increased somewhere between 
? and 1? in.  However, no significant correlations were found between weight and 
measurement changes.  
Although their weight loss was only approximately ? lb less than their weight 
gain, males? weight loss was reflected as significant measurement decreases (between ? 
and 7/8 in.) only at the seat and thigh max. The greatest actual measurement decrement 
was found at the abdomen (M = -1.56), but it was not a significant change. Significant 
positive correlations between weight and measurement loss were found at still different 
locations, the waist and biceps. The females lost less weight, showing significant 
90 
 
 
measurement decreases just above and below ? in. at thigh max and biceps.  However, 
the only significant correlation between weight and measurement change was a negative 
one at the thigh for the females who lost the most weight (5.8-8.6 lb). 
 Symmetry of changes around body circumferences.  This study found some 
significant changes in measurements and some significant correlations between weight 
and body circumference changes. Because of the use of 3D body scanned measurements, 
it was also possible to study changes in shape. To explore whether measurement changes 
were even around major circumference locations, planar slices of the bust or chest, hip 
and waist were visually analyzed for males and females who gained and lost weight. No 
other known study has done this. 
Changes in circumferential measurement at the chest among the male subjects 
who gained and lost weight were mostly symmetric (65-70% of the cases). At the same 
(bust) location among females who gained or lost weight, the distribution of symmetric 
and asymmetric changes was almost equal. Among those females who showed 
asymmetric changes, greater change at the front of the bust location was most common. 
At the hip, weight gained or lost males? changes were evenly or nearly evenly divided 
between symmetric and asymmetric patterns. Among women who gained or lost weight, 
hip changes were more often asymmetric than symmetric, but the locations of asymmetry 
differed somewhat between the gained and lost groups. Changes at the waist among men 
who gained or lost weight were slightly skewed towards the asymmetric side (53.5-60% 
of the cases) with the change most commonly showing more in the front. Females who 
gained weight showed more asymmetry at the waist than those who lost weight; among 
91 
 
 
those who gained, the asymmetry appeared more at the back, but it was more at front for 
those who lost.  
Conclusions 
 This was the first known study that used 3D body scanning to explore if and 
where body measurements change in connection with college freshman year weight 
change. Of the sample of 187 students, 100 gained weight, and 39 lost weight. Significant 
findings on circumference measurement changes (six around the trunk, plus the thigh and 
biceps) appeared more when assessed as fall to spring differences than when correlated 
with weight changes. Seven of eight measurements increased significantly for males and 
females who gained weight, but only two of eight decreased significantly for those who 
lost weight. However, the male subject numbers were small in the weight loss group, and 
the mean weight lost by females was relatively small. 
For average weight gains of approximately 7 lb, the range of measurement 
changes were approximately 5/8 in. (biceps) to nearly 2 in. (waist) for males, and 
approximately ? in. (thigh) to 1? in. (waist) for females. Male weight loss was 
approximately 6 lb; the largest measurement loss was 1? in. around the abdomen, and 
their biceps actually increased by about ? in. Neither were significant, but there were 
only 10 subjects, and the two significant changes were approximately ? in. Twenty-nine 
females lost under 4 lb; the greatest incremental change (significant) they had was 
approximately 5/8 in. in the thigh. 
Findings on measurement changes were not always parallel across trunk 
locations, and although there were a few significant positive correlations between weight 
92 
 
 
and body measurement changes, there were also some negative ones. Thus, weight 
increases and decreases did not always show clear patterns of body measurement 
increases and decreases. A larger sample size might be needed to better explore 
correlations. Although students reported how often and for how long they exercised, 
further probing of types of exercise and muscles worked might help better explain 
measurement change ranges.  
Along with easily deriving body measurements, the 3D body scanner software 
allowed for planar slices at the chest/bust, waist, and hip to be viewed. Qualitative visual 
analysis suggested that although both males and females who gained and lost weight 
displayed symmetric and asymmetric changes, males more frequently showed symmetric 
patterns than women. Evidence of asymmetric change (more in the front, back, front and 
back, or at the sides) was found at all three locations, just more often for women.  
Limitations 
While this study produced results that contribute to building understanding of 
body measurement change, its findings are limited in several ways. First, the sample was 
not random. Subjects were a convenience sample of the freshman population of Auburn 
University, a campus of approximately 25,000 students in the Southeast U.S. The 
findings cannot be broadly generalized. Although the sample size was larger than many 
of the freshman 15 weight change studies, it was still relatively small in terms of 
interpretability, particularly when the data were subdivided for statistical analysis. 
Results showed some relationships between weight changes and body 
measurement changes, but the small sample sizes of the clusters were one possible reason 
93 
 
 
that more or stronger relationships were not identified. The process for selecting the 
clusters also could have been a limitation. Scatter plots of the weight data were visually 
examined to identify the clusters. Particularly with a larger sample size, in future research 
an automated algorithm could be devised to detect the number of possible clusters.  
The determination of symmetry or asymmetry of circumference slices for the 
same individual from the two time periods was made through visual analysis. With a 
digital image processing procedure, this would be more consistent and quantitatively 
objective in nature (discussed in next section). As the evaluation proceeded, it could be 
observed that bodies might have slightly different stances in each scan; this had to be 
considered in making the evaluation. It also appeared in some comparisons of female bust 
slices that the subject may have worn a different bra in each scan, and this could affect 
assessment of the overlaid circumference shapes. 
Implications and Recommendations for Future Research 
The research results supported the findings of other studies showing freshman 
weight gain of less than 15 lb on average, but more importantly, they offered preliminary 
evidence where and how much these changes may be reflected in body measurement 
changes. Findings from this research can be applied to a broader spectrum ranging from 
possible impacts on garment design and possible implications for health studies. 
One possible methodological change that could be tried could relate to either 
apparel or health issues. To assess the symmetry of circumference changes, analysis 
which was done visually in this study might be automated using digital-image processing. 
The cross-sectional grey scale image of each planar slice can be considered a two-
94 
 
 
dimensional matrix which could be converted in a binary image so that the resultant 
image would have a black background with a white cross-sectional silhouette.  Images 
from two time periods for a particular cross section could be overlapped using any 
standard image processing software. Prior to overlapping, the individual images could be 
segmented into four quadrants, as they were in this study, and the area occupied by each 
quadrant calculated. The relative changes in area by quadrant could be deduced to 
determine a quantitative value indicating symmetric or asymmetric changes. 
Garment Design Implications 
Apparel fit is an important factor in consumer garment acceptance. Fit is impacted 
by decisions on the body measurements used in product development; size variants across 
size ranges are governed by pattern grading practices. Pattern grading systems are often 
static, meaning that the measurement increments are increased or decreased by fixed 
magnitudes for a given size range. The actual incremental amounts often used today are 
based on old anthropometric data and established assumptions about how bodies change. 
This study found that the changes in several trunk circumference measurements are not 
necessarily even, and that even within a particular circumference, the change around the 
body circumference may not be symmetrical. Further exploration could lead to better 
understanding of how body measurements and shapes change with weight gain or loss 
(size change). This could alter and improve grading rules through their reformulation, 
and, thus, potentially enhance fit for consumers and businesses? success in meeting 
consumer demand. Expanding research to a larger sample, and, perhaps to a wider age 
range would be beneficial. 
 
95 
 
 
96 
 
Health Research Implications 
Trunk fat accumulation, perhaps especially at specific locations such as the waist, 
has been found to be a risk factor for several syndromes. Researchers who are interested 
in fat-related diseases can use further research along the lines of this study to generate 
more data to better understand patterns of trunk fat accumulation among older 
adolescents as they approach adulthood and set patterns of eating and exercise. This and 
further research with more or larger samples could provide insights into educating 
consumers on the potential body areas that may be prone to fat accumulation and body 
measurement changes. Additionally, since this research was part of a larger study which 
also collected information such as body fat content and eating patterns, the body 
measurement data can be further analyzed in relation to these factors. 
 
REFERENCES 
 
Alexander, M. (2004). Applying three-dimensional body scanning technologies to body 
shape analysis. Unpublished doctoral dissertation, Auburn University, Auburn, 
AL. 
American College Health Association (2002). Healthy Campus 2010: Making it Happen. 
Retrieved July 30, 2009, from http://www.acha.org/Info_resources/hc2010.cfm  
American Society for Testing and Materials. (1995). Standard table of body 
measurements for adult female misses figure type, sizes 2-20. Philadelphia, PA.: 
American Society for Testing and Materials. 
American Society for Testing and Materials. (2001). Standard table of body 
measurements for adult female misses figure type, sizes 2-20. Philadelphia, PA.: 
American Society for Testing and Materials. 
American Society for Testing and Materials. (2002). Standard table of body 
measurements for juniors, sizes 1-19. Philadelphia, PA.: American Society for 
Testing and Materials. 
American Society for Testing and Materials. (2004). Standard tables of body 
measurements for men sizes thirty-four to sixty (34 to 60) regular. Philadelphia, 
PA.: American Society for Testing and Materials.
97 
 
 
 
Anderson, D. A., Shapiro, J. R., & Lundgren, J. D. (2003). The freshman year of college 
as a critical period for weight gain: An initial evaluation. Eating Behaviors, 4(4), 
363-367. 
Ashdown, S. P., & Na, H. (2007). Comparison of 3D body scan data to quantify upper-
body postural variation in older and younger women. Clothing and Textiles 
Research Journal, 26(4), 292-307. 
Atsumi , K., Tetsuro , M., & Hiroyuki , D. (2005). Visceral fat accumulation and 
Ischemic heart disease. Clinic, 54(4), 1360-1364. 
Baker, P. T., Hunt Jr, E. E., & Sen, T. (1958). The growth and interrelations of skinfolds 
and brachial tissues in man. American Journal of Physical Anthropology, 16(1), 
39-58. 
Barker, R. L., & Coletta, G. C. (1986). Performance of protective clothing: A symposium 
sponsored by ASTM Committee F-23 on Protective Clothing, Raleigh, NC, 16-20 
July 1984. ASTM special technical publication, 900. Philadelphia, PA: American 
Society for Testing and Materials.  
Barnes, H. V. (1975). Physical growth and development during puberty. Medical Clinics 
of North America, 59(6), 1305-1317. 
Baumgartner, R. N., Roche, A. F., Guo, S., Lohman, T., Boileau, R. A., & Slaughter, M. 
H. (1986). Adipose tissue distribution: The stability of principal components by 
sex, ethnicity and maturation stage. Human Biology, 58(5), 719-735. 
98 
 
 
Beunen, G., Malina, R. M., Lefevre, J., Claessens, A. L., Renson, R., Simons, J., Maes, 
H., Vanreusel, B., & Lysens, R. (1994). Size, fatness and relative fat distribution 
of males of contrasting maturity status during  
adolescence and as adults. International Journal of Obesity, 18(10), 670-678. 
Borkan, G. A., Hults, D. E., Gerzof, S. G., Robbins, A. H., & Silbert, C. K. (1983). Age 
changes  
in body composition revealed by computed tomography. Journal of Gerontology, 
38(6), 673-677. 
Borkan, G. A., & Norris, A. H. (1977). Fat redistribution and the changing body 
dimensions of the adult male. Human Biology, 49(3), 495-513. 
Brozek, J. (1952). Changes of body composition in man during maturity and their 
nutritional implications. Federation Proceedings, 11(3), 784-93.  
Bye, E. K. (1990). A visual sensory evaluation of two pattern grading methods. 
Unpublished doctoral dissertation, University of Minnesota, St. Paul. 
Bye, E., LaBat, K., McKinney, E., & Kim, D. E. (2007). Optimized pattern grading. 
International Journal of Clothing Science and Technology, 20(2), 79-92. 
CAESAR. (2000). The most comprehensive source for body measurement data.   
Retrieved December, 10, 2008, from http://store.sae.org/caesar/ 
Centers for Disease Control and Prevention. (2006). Prevalence of overweight among 
adults: United States, 2003-2004. Retrieved September 12, 2008, from 
http://www.cdc.gov/nchs/products/pubs/pubd/hestats/overweight/overwght_adult
_03.htm 
99 
 
 
Centers for Disease Control and Prevention. (2007). Overweight and obesity.   Retrieved 
September 12, 2008, from http://www.cdc.gov/nccdphp/dnpa/obesity/index.htm 
Connell, L. J., Ulrich, P. V., Brannon, E. L., Alexander, M., & Presley, A. B. (2006). 
Body Shape Assessment Scale: Instrument development for analyzing female 
figures. Clothing and Textiles Research Journal, 24(2), 80-95                                                
Croney, J. (1971). Anthropometry for designers. New York: Van Nostrand Reinhold. 
Cronk, C. E., Mukherjee, D., & Roche, A. F. (1983). Changes in triceps and subscapular 
skinfold thickness during adolescence. Human Biology: An International Record 
of Research, 55(3), 707-721. 
 Cronk, C. E., Roche, A. F., Chumlea, W. C., & Kent, R. (1982). Longitudinal trends of 
weight/stature in childhood in relationship to adulthood body fat measures. 
Human Biology, 54(4), 751-764. 
Delinsky, S. S., & Wilson, G. T. (2008). Weight gain, dietary restraint, and disordered 
eating in the freshman year of college. Eating Behaviors, 9(1), 82-90. 
Demers, M. H., Hurley, J. D., Wulpern, R. C., & Grindon, J. R. (1997). Three-
dimensional surface capture for body measurement using projected sinusoidal 
patterns. Proceedings- SPIE The International Society For Optical Engineering. 
(3023), 13-25.  
Deutsch, M. I., Mueller, W. H., & Malina, R. M. (1985). Androgyny in fat patterning is 
associated with obesity in adolescents and young adults. Annals of Human 
Biology, 12(3), 275-286. 
Drawdy, K. M. (2007). Weight Change Among College Freshmen the Freshman 4. 
Unpublished master?s thesis, Auburn University, Auburn, AL. 
100 
 
 
Dryden, I. L., & Mardia, K. V. (1998). Statistical shape analysis. Wiley series in 
probability and statistics. Chichester: John Wiley and Sons. 
Fan, J., Yu, W., & Hunter, L. (2004). Clothing appearance and fit: Science and 
technology. Cambridge; Roca Raton, Fla.: Woodhead : Textile Institute ; CRC. 
Garn, S. M., & Young, R. W. (1956). Concurrent fat loss and fat gain. American Journal 
of Physical Anthropology, 14(3), 497-504. 
Graham, M. A., & Jones, A. L. (2002). Freshman 15: Valid theory or harmful myth? 
Journal of American College Health, 50(4), 171-173. 
Gutzwiller, J., Oliver, J. M., & Katz, B. M. (2003). Eating dysfunctions in college 
women: The roles of depression and attachment to fathers. Journal of American 
College Health, 52(1), 27-32. 
Hattori, K., Becque, M. D., Katch, V. L., Rocchini, A. P., Boileau, R. A., Slaughter, M. 
H., & Lohman, T. (1987). Fat patterning of adolescents. Annals of Human 
Biology, 14(1), 23-28. 
Hediger, M. L., Scholl, T. O., Schall, J. I., & Cronk, C. E. (1995). One-year changes in 
weight and fatness in girls during late adolescence. Pediatrics, 96(2), 253-258. 
Hedley, A. A., Ogden, C. L., Johnson, C. L., Carroll, M. D., Curtin, L. R., & Flegal, K. 
M. (2004). Prevalence of overweight and obesity among US children, adolescents, 
and adults, 1999-2002.  American Medical Association, 291, 2847-2850 
Heisey, F. L. (1984). A quantitative methodology for generating specifically fitted 
garment patterns. Unpublished master?s thesis, University of Minnesota, St. Paul 
Minnesota. 
101 
 
 
Hodge, C. N., Jackson, L. A., & Sullivan, L. A. (1993). ?THE FRESHMAN 15? Facts 
and fantasies about weight gain in college women. Psychology of Women 
Quarterly, 17(1), 119-126. 
Hoffman, D. J., Policastro, P., Quick, V., & Lee, S. K. (2006). Changes in body weight 
and fat mass of men and women in the first year of college: A study of the" 
Freshman 15". Journal of American College Health, 55(1), 41-46. 
Holm-Denoma, J. M., Joiner, T. E., Vohs, K. D., & Heatherton, T. F. (2008). The" 
freshman fifteen" (the   " freshman five" actually): Predictors and possible 
explanations. Health Psychology, 27(1), S3-SS9. 
Hooton, E. A., & Dupertuis, C. W. (1951). Age Changes and Selective Survival in Irish 
Males. American Association of Physical Anthropologists. Ann Arbor; Michigan: 
Edwards Brothers. 
Hovell, M. F., Mewborn, C. R., Randle, Y., & Fowler-Johnson, S. (1985). Risk of excess 
weight gain in university women: A three-year community controlled analysis. 
Addictive Behaviors, 10(1), 15-28. 
How to get rid of belly fat! (n.d.).  Retrieved February 24, 2009, from 
http://www.nutrition-health-supplements.com/belly-fat.html 
Kemper, H. C. G. (1985). Growth, health, and fitness of teenagers: Longitudinal 
research international perspective. Basel; New York: Karger. 
Kurt Salmon Associates. (2000). KSA's annual consumer outlook survey results. New 
York, NY: Author. 
102 
 
 
Levitsky, D. A., Halbmaier, C. A., & Mrdjenovic, G. (2004). The freshman weight gain: 
A model for the study of the epidemic of obesity. International Journal of 
Obesity, 28, 1435-1442. 
Malina, R. M., Bouchard, C., & Bar-Or, O. (2004). Growth, maturation, and physical 
activity. Champaign, Ill: Human Kinetics. 
Matvienko, O., Lewis, D. S., & Schafer, E. (2001). A college nutrition science course as 
an intervention to prevent weight gain in female college freshmen. Journal of 
Nutrition Education and Behavior, 33(2), 95-101. 
Mihalopoulos, N. L., Auinger, P., & Klein, J. D. (2008). The Freshman 15: Is it real? 
Journal of American College Health, 56(5), 531-534. 
Mokdad, A. H., Serdula, M. K., Dietz, W. H., Bowman, B. A., Marks, J. S., & Koplan, J. 
P. (1999). The spread of the obesity epidemic in the United States, 1991-1998. 
Journal of the American Medical Association, 282(16), 1519-1522. 
Mossiman, J. E. (1988). Size and shape analysis. In S. Kotz & N. L. Johnson (Eds.), 
Encyclopedia of Statistical Sciences, 8, (497-507). New York: John Wiley and 
Sons. 
Morrow, M. L., Heesch, K. C., Dinger, M. K., Hull, H. R., Kneehans, A. W., & Fields, D. 
A. (2006). Freshman 15: Fact or fiction? Obesity Research, 14(8), 1438. 
Mueller, W. H. (1982). The changes with age of the anatomical distribution of fat. Social 
Science and Medicine, 16(2), 191-196. 
National Center for Health Statistics (2001). Prevalence of overweight among children 
and adolescents: United States, 1999. Centers for Disease Control and 
Prevention, Health E-Stats. 
103 
 
 
O?Brien, R., & Shelton, W. C. (1941). Women?s measurements for garment and pattern 
construction. Publication No.454. Washington, D.C.: U.S. Department of 
Agriculture.  
Ogden, C. L., Carroll, M. D., Curtin, L. R., Mcdowell, M. A., Tabak, C. J., & Flegal, K. 
M. (2006). Prevalence of overweight and obesity in the United States, 1999-2004. 
Journal of the American Medical Association, 295(13), 1549-1555. 
Paquette, S. (1996). 3D scanning in apparel design and human engineering. IEEE 
Computer Graphics and Applications, 16(5) 11-15. 
 Patterson, F., Lerman, C., Kaufmann, V. G., Neuner, G. A., & Audrain-McGovern, J. 
(2004). Cigarette smoking practices among American college students: Review 
and future directions. Journal of American College Health, 52(5), 203-212. 
Petrova, A., & Ashdown, S. (2008). Three-dimensional body scan data analysis: Body 
size and shape dependence of ease values for pants' fit. Clothing and Textiles 
Research Journal, 26(3), 227-253. 
Pisut, G., & Connell, L. (2007). Fit preferences of female consumers in the USA. Journal 
of Fashion Marketing and Management, 11(3), 366-379. 
Poehlman, E. T., Toth, M. J., Bunyard, L. B., Gardner, A. W., Donaldson, K. E., Colman, 
E., Fonong, T., &  
Ades, P. A. (1995). Physiological predictors of increasing total and central 
adiposity in aging men  
and women. Archives of Internal Medicine, 155(22), 2443-2448. 
Ramirez, M. E. (1993). Subcutaneous fat distribution in adolescents. Human Biology, 
65(5), 771-782. 
104 
 
 
Ramirez, M. E., & Mueller, W. H. (1980). The development of obesity and fat patterning 
on Tokelau children. Human Biology, 52(4), 675-687. 
Reynolds, E. L. (1951). The distribution of subcutaneous fat in childhood and 
adolescence. Fels monograph series, no. 1. [Evanston, Ill.]: Child Development 
Publications.  
Rexrode, K. M., Carey, V. J., Hennekens, C. H., Walters, E. E., Colditz, G. A., Stampfer, 
M. J., Willett, W.C., & Manson, J. E. (1998). Abdominal Adiposity and Coronary 
Heart Disease in Women. Journal of the American Medical Association, 280(21), 
1843-1848.  
Santos, L., Cintra, I., Fisberg, M., & Martini, L. (2008). Body trunk fat and insulin 
resistance in post-pubertal obese adolescents. Sao Paulo Medical Journal, 126, 
82-86. 
Saunders, D. (2008). Weight and body composition changes in first semester college 
freshmen. Unpublished master?s thesis, Auburn University, Auburn, AL. 
Schofield, N. A., & LaBat, K. L. (2005). Exploring the relationships of grading, sizing, 
and anthropometric data. Clothing and Textiles Research Journal, 23(1), 13-27. 
Schwartz, R. S., Shuman, W. P., Bradbury, V. L., Cain, K. C., Fellingham, G. W., Beard, 
J. C., kahn, S. E., Stratton, J. R., Cerqueria, M. D., & Abrass, I. B.  (1990). Body 
fat distribution in healthy young and older men. Journal of Gerontology,45(6), 
181-185. 
Sheldon, W. H., Stevens, S. S., & Tucker, W. B. (1940). The varieties of human 
physique: An introduction to constitutional psychology. New York: Harper. 
105 
 
 
Siervogel, R. M., Wisemandle, W., Guo, S. S., Chumlea, W. C., Towne, B., & Roche, A. 
F. (1998). Adult changes in body composition are associated with changes in 
cholesterol levels: The fels longitudinal study. Applied Radiation and Isotopes, 
49(5-6), 727-729. 
Simmons, K. P. (2002). Body shape analysis using three-dimensional body scanning 
technology. Unpublished doctoral dissertation, North Carolina State University, 
Raleigh, NC. 
Simmons, K. P., & Istook, C. L. (2003). Body measurement techniques: Comparing 3D 
body-scanning and anthropometric methods for apparel applications. Journal of 
Fashion Marketing and Management, 7(3), 306-332. 
Sinha, R., Kapoor, S., & Kapoor, A. K. (2008). Tracing the response of subcutaneous fat 
accumulation in two generations of males. HOMO - Journal of Comparative 
Human Biology, 59(6), 429-438. 
SizeUSA. (2004). THE US NATIONAL SIZE SURVEY. Retrieved December 10, 2008, 
from http://www.tc2.com/what/sizeusa/index.html 
Sizing up garment fit issues. (2007). Retrieved 29 September, 2008, from 
http://www.tc2.com/what/sizeusa/index.html#siz  
[TC]
2
. (2004). Body Measurements and Data Analysis Reports on the U.S. Population. 
Cary, N. C. 
 
 [TC]
2
. (2008). 3D BODY SCANNER SPECIFICATIONS.   Retrieved 12/1/2008, 2008, 
from http://www.tc2.com/products/body_scanner.html 
The strength and significance of the coefficient. (2001).   Retrieved 25 May, 2009, from 
http://www.bized.co.uk/timeweb/crunching/crunch_relate_expl.htm 
106 
 
 
107 
 
 Thibodeau, G. A., & Anthony, C. P. (1990). Anthony's Textbook of anatomy and 
physiology. St. Louis: Times Mirror/Mosby College Publication. 
Thompson, D. W. (1917). On growth and form. Cambridge, UK: University Press. 
Trust for America?s Health Report. (2008). F as in fat: How obesity policies are failing in 
America, 2008. Retrieved September 12, 2008, 
http://healthyamericans.org/reports/obesity2008/Obesity2008Report.pdf Hodge 
U.S. Department of Commerce. (1958). Commercial standard: Voluntary product 
standard, CS 215-58:Body measurements for the sizing of women?s patterns and 
apparel. (National Bureau of standards). Washington, D.C.: U.S. Government 
Printing Office. 
U.S. Department of Commerce. (1970). PS 42-70: Body measurements for the sizing of 
women?s patterns and apparel. (National Bureau of standards). Washington, 
D.C.: U.S. Government Printing Office. 
 Wells, J., Treleaven, P., & Cole, T. (2007). BMI compared with 3-dimensional body 
shape: The UK National Sizing Survey. American Journal of Clinical Nutrition, 
85(2), 419-425. 
Wilson, G. S., Pritchard, M. E., & Schaffer, J. (2004). Athletic status and drinking 
behavior in college students: The influence of gender and coping styles. Journal 
of American College Health, 52(6), 269-273 
Wong, D. L. (1999). Whaley & Wong's nursing care of infants and children. Mosby; St. 
Louis.
 
APPENDIX 
SAMPLE SLICE SYMMETRY/ASYMMETRY SURVEY FORM
108 
 
 
 
CODE-
GENDER 
 
 
Bust 
 
  Symmetrical: 
Relatively even change 
around body 
circumference  
 
Asymmetrical 
       
  More in the back and 
front than on the    
        sides 
 
  Most in the back 
 
  Most in the front 
 
  Most on the sides 
 
 
 
 
 
Hip 
 
 
  Symmetrical: 
Relatively even change 
around body 
circumference  
 
Asymmetrical 
       
  More in the back and 
front than on the    
        sides 
 
  Most in the back 
 
  Most in the front 
 
  Most on the sides 
 
 
 
 
 
Waist 
  Symmetrical: 
Relatively even change 
around body 
circumference  
 
Asymmetrical 
       
  More in the back and 
front than on the    
        sides 
 
  Most in the back 
 
  Most in the front 
 
  Most on the sides 
 
109