Changes in Body Composition and Cardiometabolic Health Observed in Normal-Weight College Students Across Each College Year

CHANGES IN BODY COMPOSITION AND CARDIOMETABOLIC HEALTH OBSERVED IN NORMAL-WEIGHT COLLEGE STUDENTS ACROSS EACH COLLEGE YEAR

Austin Michael Peterjohn

A thesis submitted to the faculty of the University of North Carolina at Chapel Hill in partial fulfillment of the requirements for the degree of Master of Arts in the Department of Exercise and Sport Science (Exercise Physiology).

Chapel Hill 2019

Approved by:

Abbie Smith-Ryan

Kristin Ondrak

Katie Hirsch

ii ABSTRACT

Austin Peterjohn: Changes in Body Composition and Cardiometabolic Health Observed in Normal-Weight College Students Across Each College Year (Under the direction of Abbie Smith-Ryan)

The present study described changes in body composition and cardiometabolic health during one academic year in male and female college students of each academic class (n=143).

Students completed body composition and cardiometabolic health analyses during a single visit on three separate occasions: early Fall semester (August-October) 2017, late spring semester

(March-May) 2018, and early Fall semester (August-October) 2018. There was a main effect for time for body mass (p=0.035), BMI (p=0.025), lean mass (p=0.007), and fasting glucose

(p=0.046). Females saw significant gains in body mass (p=0.015), BMI (p=0.015), and lean mass

(p=0.002). Lean mass of seniors significantly increased (p=0.010) from Fall 2017 to Spring

2018, but no other differences between classes were found with respect to body composition and cardiometabolic health. The results of this study suggest minimal changes in body composition and cardiometabolic health occur during one year of college, regardless of class.

iii ACKNOWLEDGEMENTS

I would like to thank my committee for their contribution to this project and their guidance throughout this experience. The quality of this project stems from the contributions and advice my committee gave to me, and I can’t thank them enough.

My lab team deserves a huge “thank you,” for their effort in this project. Not only were they helpful and supportive, but I also had a blast working with such amazing individuals, and I value their friendship greatly.

I also have my friends and family to thank for their support over the course of this project and my time studying at UNC. It was a marathon I may not have finished without their continued support and encouragement.

Lastly, I’d like to thank my advisor, Dr. Smith-Ryan, for her continued encouragement and belief in me as a student and researcher. She was always in my corner and pushed me to succeed when I needed it most.

TABLE OF CONTENTS

LIST OF TABLES ...... vii LIST OF FIGURES ...... viii LIST OF ABBREVIATIONS ...... ix CHAPTER I: INTRODUCTION ...... 1 Purpose ...... 3 Research Questions: ...... 4 Research Hypotheses: ...... 5 Limitations: ...... 5 Delimitations: ...... 6 Significance: ...... 6 CHAPTER II: REVIEW OF LITERATURE ...... 7 Introduction ...... 7 Body Composition ...... 8 College Weight Gain ...... 11 Normal Weight Obesity ...... 13 Lifestyle Factors ...... 15 Physical Activity ...... 16 Alcohol Consumption ...... 17 Dietary Habits ...... 17 Stress ...... 18 Sleep ...... 19 Questionnaires and Surveys ...... 20 Conclusions ...... 21 CHAPTER III: METHODS ...... 23 Participants ...... 23 Experimental Design ...... 24 Body Composition ...... 24 Cardiometabolic Health ...... 25

Questionnaires ...... 25 Statistical analysis ...... 26 CHAPTER IV: MANUSCRIPT ...... 28 Introduction ...... 28 Participants and Methods ...... 30 Experimental Design ...... 31 Anthropometric and Body Composition Analysis ...... 31 Cardiometabolic Health ...... 32 Lifestyle Questionnaires ...... 33 Statistical Analysis ...... 33 Results ...... 34 Anthropometrics ...... 34 Body Composition ...... 35 Cardiometabolic Health ...... 35 Males vs. Females ...... 36 Lifestyle Factors ...... 37 Discussion ...... 38 Body Mass and BMI ...... 39 Body Composition ...... 40 Cardiometabolic Health ...... 41 Lifestyle Factors ...... 43 Limitations ...... 43 Future Considerations ...... 44 CHAPTER V: CONCLUSION ...... 45 TABLES ...... 46 FIGURES ...... 52 REFERENCES ...... 56

LIST OF TABLES

Table 1 – Descriptive statistics for the total sample and males vs. females…………………… 46

Table 2 – Change in weight, BMI, and body composition from Fall 2017-

Spring 2018 for all subjects and each class……………………………………………... 47

Table 3 – Change in weight, BMI, and body composition from Fall 2017-

Spring 2018; male vs. female changes………………………………………………….. 48

Table 4 – Change in Total cholesterol, HDL, nHDL, and glucose from

Fall 2017-Fall 2018; male vs female changes…………………………………………… 49

Table 5 – Change in Total cholesterol, HDL, nHDL, and glucose from

Fall 2017-Fall 2018 for all subjects and each class with data at both time points………. 50

Table 6 – Standardized Beta-Coefficients and p-values of regression

analyses conducted investigating nine predictor variables for FM, LM, and VAT…….. 51

vii

LIST OF FIGURES

Figure 1 – Consort diagram detailing enrollment and retention of the sample………………… 52

Figure 2 – Distribution in the changes in body weight over the 2017-2018 academic year…… 53

Figure 3 – Changes in body weight over the 2017-2018 academic year in males……………... 54

Figure 4 – Changes in body weight over the 2017-2018 academic year in females…………… 55

viii

LIST OF ABBREVIATIONS

%BF Body fat percentage

2C Two-compartment model

3C Three-compartment model

4C Four-compartment model

ANOVA Analysis of variance

AUDIT Alcohol use disorders identification test

BMD Bone mineral density

BMI Body mass index

CAPS-r College alcohol problems scale - revised

CHD Coronary heart disease

DASS Depression, anxiety and stress scale

DEXA Dual energy X-ray absorptiometry

FFDM Fat-free dry mass

FFM Fat-free mass

FM Fat mass

GLU Glucose

HDL High-density lipoprotein

ICC Intra-class correlation coefficient

IPAQ International physical activity questionnaire

LM Lean mass

Mo Bone mineral content

NCHA National collegiate health assessment nHDL Non-high-density lipoprotein

PSQI Pittsburgh sleep quality index

PSS Perceived stress scale rMEQ Reduced morningness-eveningness questionnaire

RMR Resting metabolic rate

SAT Subcutaneous adipose tissue

SBQ Sedentary behavior questionnaire

SEM Standard error of mean

SM Skeletal muscle

TC Total cholesterol

TBW Total body water

VAT Visceral adipose tissue

VAT DEXA Visceral adipose tissue from dual energy X-ray absorptiometry

CHAPTER I: INTRODUCTION

According to the World Health Organization more than 1.9 billion adults are overweight, with 650 million of these individuals considered obese 1. Across the United States, it is the young adult age group (18 to 29 years) that has experienced the greatest increases in overweight and obesity 3. Paralleled with this increase in overweight and obesity, full-time college enrollment increased 15% between 2005 and 2015 4. The college years may be a critical time period during which lifestyle and health behaviors are established. The college setting introduces many obesogenic factors including unlimited access to dining halls, long hours spent studying, initial exposure to alcohol, and high stress, to name a few. The phenomenon of weight gain in the first year of college, commonly referred to as the “freshman 15,” has been investigated and ultimately dismissed in research, with an average weight gain of 3.85 lbs occurring during the first year of college 2. This nontrivial weight gain is generally attributed to significant increases in body fat percentage (%BF), fat mass (FM) and waist circumference during the freshman year 5,6 . Less is known about these changes in body composition among other classes, with some initial evidence suggesting a similar trajectory, although conflicting evidence exists 5,7-8. Data from 2012 reported

70% of students gain weight over the course of the first three years of college, and a 1.4 ± 1.2 kg/m 2 increase in body mass index (BMI), a 3.5 ± 3 kg increase in FM, and a 4.0 ± 3.1% increase in %BF were seen in these weight-gainers 5. On the other hand, Hull et al. (2007) observed significant decreases in %BF, FM, and a significant increase in fat-free mass (FFM) among female college sophomores 8.

Beyond weight gain, the concept of normal weight obesity (NWO) has gained attention as a condition that raises risk of cardiometabolic dysfunction. Obesity is most commonly measured by assessing BMI. While BMI is a convenient relative index of body mass, it fails to indicate body composition measures such as fat free mass (FFM), FM, and %BF. An individual is considered NWO with a calculated BMI within the normal range (18.5-24.9 kg/m 2), while also carrying an unhealthy amount of body fat (generally >20-25% in men and 30-37% in women) 9-

10 . NWO individuals are at higher risk for certain diseases when compared to leaner individuals with normal BMI values 10 . An estimated 30 million adults are considered NWO in the United

States 10 .

Prevalence of NWO has primarily been studied in middle-aged and older adults (35-75 yrs) and has even been identified in adolescents 102 . However, NWO has only once been exclusively studied in a college-aged population 29 . Certain lifestyle factors during this phase in life may be considered attributable to the development of NWO, making young adulthood a potentially critical time to improve health and lifestyle behaviors. Furthermore, college students are exposed to many new environments and create new habits, making this an ideal time to promote healthy lifestyle changes 13 . According to the 2017 National Collegiate Health

Assessment (NCHA) survey, 33.6% of undergraduate students are considered overweight or obese according to self-reported BMI values 14 . However, the remaining 66.3% of individuals should not all be assumed healthy, and some may even have similar health risk compared to their overweight peers due to the hidden prevalence of NWO 21 . The NCHA survey also found 50.1% of college students do not meet the recommended minimum amounts physical activity, with

37.8% students reporting no moderate and vigorous exercise on a weekly basis. Identification of

2 high risk time periods during college will allow for better understanding of contributors to weight gain and for development of effective prevention efforts.

Change in body weight is always a primary outcome in most “freshman 15” and other college weight gain studies. However, by focusing solely on body weight, many studies are unable to determine changes in lean mass (LM) or FM, which are both closely associated with health and functionality 20 . Measuring body composition changes, not simply body mass, will provide descriptive and meaningful data to fully evaluate how health may change throughout one’s college career.

Purpose

1. The primary purpose of this study was to quantify changes in body composition that

occur over one college year (Fall 2017- Spring 2018) in normal-weight college students.

a. To identify class-specific changes (i.e. freshman, sophomore, junior, or senior) in

body composition including FM, LM and %BF, in addition to weight and BMI.

2. The secondary purpose of this study was to examine changes in cardiometabolic health

that occur over one calendar year (Fall 2017- Fall 2018) in normal-weight college

students.

a. Fasting blood glucose (GLU), total cholesterol (TC), high-density lipoprotein

(HDL), and non-high-density lipoprotein (nHDL).

3. The tertiary purpose of this study was to evaluate the relationship between specific

lifestyle habits and body composition and cardiometabolic health in normal-weight

college students.

a. Determine the relationship between body composition outcomes (LM, FM, %BF

VAT) with moderate physical activity, vigorous physical activity, and sedentary

behavior.

b. Determine the relationship between cardiometabolic health outcomes (TC, HDL,

nHDL, GLU) with moderate physical activity, vigorous physical activity, and

sedentary behavior.

c. Determine whether FM, LM, %BF or VAT had a relationship with a variety of

predictor independent variables.

i. Minutes of vigorous and moderate physical activity

ii. Minutes of sedentary behavior

iii. Perceived stress scale (PSS) score

iv. Pittsburgh sleep quality index (PSQI) component score

v. Depression, anxiety, and stress scale (DASS-21) - anxiety score

vi. Depression, anxiety, and stress scale (DASS-21) - stress score

vii. Reduced morningness-eveningness questionnaire (rMEQ) score

viii. Alcohol use disorders identification test (AUDIT) score

4. The quaternary purpose of this study was to characterize sex-specific changes in body

composition and cardiometabolic profile observed over one college year.

a. Identical outcomes of interest as purpose statements 1 and 2

Research Questions:

1. Do class-specific differences exist in the observed changes in body composition over

one academic year?

2. Do cardiometabolic health differences exist in the observed changes in body

composition over one calendar year?

3. What lifestyle factors are associated with changes in body composition and

cardiometabolic health?

4. Do males and females see similar changes in body composition and cardiometabolic

health over one year?

Research Hypotheses:

1. Freshmen will gain more weight and FM than any other class over the course of one

year.

2. Total daily sedentary behavior, evening-type sleep chronotype, and stress will be

positively associated with weight gain.

3. Males will gain more weight, FM, LM, and %BF than females.

4. Cardiometabolic health will significantly change in only the freshman class over the

course of one calendar year.

Limitations:

1. Subjects were all be enrolled at the same southeastern university, making the sample

not completely random.

2. There is a possibility of subjects dropping out of school, graduating, studying abroad,

transferring schools, or moving away from the area, making secondary data difficult

to obtain.

3. Although exercise was assessed through the use of the IPAQ, type of exercise was not

assessed (resistance training, aerobic training, sport, etc.)

4. Many students were recruited from the Exercise and Sports Science Department at

UNC. We believe many of these students are more health-conscious than the general

population at the university.

Delimitations:

1. The study consisted of one visit per semester.

2. Participants were excluded if they had gained or lost ≥ 10 lbs in the last three months.

3. Premenopausal women and healthy men between the ages of 18-25 yrs will be chosen

to participate.

4. BMI of subjects will be between 18.5-24.9 kg/m 2.

5. Body composition will be estimated dual energy x-ray absorptiometry.

6. Glucose, high-density lipoprotein, and non-high-density lipoprotein will be measured

from a single blood draw.

Significance:

This study was the first longitudinal study observing body composition and cardiometabolic changes that included students representing all four college classes. With data from all four classes, it was possible to observe class-specific differences in lifestyle factors and how they relate to changes in body composition and cardiometabolic health, potentially identifying specific years during college in which body composition and cardiometabolic health are more likely to change.

CHAPTER II: REVIEW OF LITERATURE

Introduction

Obesity has reached epidemic proportions according to the World Health Organization, with over 650 million individuals considered obese in a 2016 survey 1. Recently, young adults

(18-29 years) have experienced the greatest rise in overweight and obesity, and many of this group are enrolled in college 3. The college experience is typically accompanied by academic pressure, social stress, and it is usually the first occasion in which individuals live away from home. It has been claimed that, on average, college students gain 15 pounds during the first year of school, which is commonly referred to as the “freshman 15.” No study has observed a mean

15 lbs weight gain in a college population, although individual cases do likely exist 2,6 . A recent meta-analysis which reviewed 32 research trials reported anywhere from a weight loss of 1.50 lbs up to a weight gain of 6.8 lbs. The average weight gain during the first year of college across all 32 studies was 3.85 lbs, far from the claimed 15 lbs that is generally perceived by the

American public to be a realistic expectancy 2. Literature surrounding college weight gain is mostly aimed at freshmen; very little data exists describing weight gain in college sophomores, juniors, and seniors. Not only does existing research lack diversity in the subset of the college population it focuses on, it lacks depth in what it measures. Most studies currently published have utilized a rudimentary primary measure of body mass to gauge a college student’s health status. More valuable and applied conclusions could be drawn from descriptive body

composition measures, specifically LM, FM, and %BF. Cardiometabolic and functional health outcomes can be better assessed by including a measure of body composition in a college weight gain study, specifically a multi-compartment criterion model. This review aims to evaluate and discuss literature surrounding cardiometabolic health and body composition as it relates to normal-weight college students and the academic environment.

Body Composition

There are many reasons to collect body composition estimates in the college student population, such as establishing total body and regional distributions of fat and lean mass, determining skeletal integrity, developing a reference base for energy expenditure, and defining relative hydration 53 . Estimating body composition can be done using multiple methods which divide the body into different compartments comprised of unique tissues. The simplest compartment model features two compartments (2C) in which the body is subdivided into FM and FFM. Two commonly used methods to estimate body composition using the 2C model are hydrodensitometry and air displacement plethysmography. The 2C model is the primary method in which the majority of our knowledge of body composition is based upon. While the 2C model is widely used, validity and reliability are lacking. An accepted gold standard involves a multicompartment model, referred to as the four-compartment model (4C), which determines four separate compartments of the body, including FM, LM, total body water (TBW), and total body bone mineral content (Mo).

Baumgartner et al. 19 have suggested the 4C model be considered the “gold standard” measure of body composition in adults due to the multicompartment nature and its ability to estimate bone mineral density (BMD) and TBW rather than assuming a constant FFM (aqueous

8 and bone) density of 1.100 g/cm 3 and hydration status of 73.2%, which was the common assumption made prior to the development of the 4C model 36 . The classical 2C model is limited by assumptions regarding the consistency of FFM, namely water and mineral content 54 . The 3C model, which is able to estimate FM, fat-free dry mass (FFDM), and water content, can overcome the boundaries surrounding FFM hydration, but this model has limitations of its own.

The 3C model is not able to distinguish between protein and mineral content of FFDM, so the existence of a common ratio between individuals must be assumed 36 . The 4C model is free of these assumptions due to its ability to distinguish between fat, protein, water and mineral content.

Beyond basic body composition parameters of FM, LM, %BF, regional distributions of body fat, such as subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT), may be more advantageous for detecting chronic disease risk. VAT is commonly considered to have the most pernicious effects on the body’s cardiovascular system, endocrine system, and metabolism 24 . Pouliot et al. 23 observed a positive correlation between VAT area and glucose intolerance 23 . Further evidence has correlated glucose intolerance with VAT area independent of total adiposity and SAT 23,25 . VAT has been associated with decreased insulin sensitivity measured using a euglycemic hyperinsulinemic glucose clamp in young, healthy males 26-27 . In addition to altering glucose-insulin homeostasis, increased intra-abdominal fat has been associated with abnormal circulating levels of plasma lipoprotein and lipid levels, specifically suppressed high-density lipoprotein (HDL) and increased plasma triglycerides 28-29 . SAT, although attributed to less cardiometabolic risk than VAT, has been associated with insulin resistance, especially truncal SAT which includes adipose tissue of the thorax and abdomen 30 .

A component of body composition that is commonly overlooked as an indicator of metabolic health, especially in context of weight status in which reduction of FM is heavily emphasized, is LM. LM represents the collective mass of an individual’s skeletal muscle (SM), tendons, ligaments, and internal organs, with the latter three components having generally fixed masses. Muscle weakness and insufficient amounts of LM have been consistently linked to increased rates of mortality amongst older adults 31 . Reduced LM and strength have been shown to directly impact functionality, quality of life, and the ability to perform activities of daily living 20,31-32 . SM serves as a site at which insulin binds in order to augment glucose disposal from the bloodstream into the SM tissue where it will either enter glycolysis and be utilized as fuel, or be stored as glycogen for later use. Rate and efficiency of glucose metabolism and glycogen synthesis could be compromised by the absence of ample amount or quality of SM. It has been hypothesized that insulin resistance in SM may promote atherogenic dyslipidemia by inhibiting glycogen storage and promoting hepatic de novo lipogenesis, resulting in an increase in plasma triglyceride concentration and a reduction in plasma HDL concentration 33 . Loss of, or low amounts of LM reduces the total number of insulin binding sites, potentially exacerbating metabolic disorders such as type 2 diabetes mellitus, insulin resistance, and metabolic syndrome 32 . Resting metabolic rate (RMR) is also heavily dependent on an individual’s LM, with more LM leading to higher RMR, on average, as a result of continual protein synthesis and breakdown within SM tissue. Developing and maintaining adequate amounts of SM through young adulthood and preserving this tissue with age can improve one’s quality of life, functionality and overall metabolic health.

College Weight Gain

Weight gain during the college years has been considered a normal occurrence in the view of the public and is supported by the majority of scientific literature. The college years are a critical and vulnerable period for body weight changes and adoption of unhealthy lifestyle behaviors 16,21 . College students experience high stress brought about by academic pressure, social factors, and newfound independence. The “freshman fifteen” describes the alleged fifteen- pound weight gain that occurs during the first year of college. However, every study tracking weight change in college freshmen reports findings far less than fifteen pounds on average. A

2009 meta-analysis conducted by Vella-Zarb et al. (2009) examined 24 studies and found the average amount of weight gained during the first year of college was 3.86 lbs 21 . A second meta- analysis was conducted by Vabeboncoeur et al. (2015) that came to similar conclusions, reporting an average of 2.99 pounds gained over the course of the freshman year 2.

Change in body weight over the course of the freshman year is not consistent within the literature, which can be seen in the prior-mentioned meta analyses 2,21 . Of the 22 studies analyzed by Vabeboncoeur et al. (2015), mean weight gain ranged from -1.5 to 6.8 lbs. A substantial discrepancy exists between average weight gain in weight-gainers and the entire sample, with the sample gaining 2.99 lbs. on average and high-gainers gaining 7.5 lbs. It should be mentioned the average study length of the included studies ranged from 1.38-8 months, which could explain some of the variance observed.

Weight gain occurring in college is not limited to the freshman year 7,34 . The literature, although lacking, suggests weight gain may continue through the entirety of the undergraduate academic career. Racette et al. (2008) reported an average weight gain of 3.75 lbs and 9.26 lbs between the beginning of freshman year and the end of senior year for females and males,

11 respectively 7. Prevalence of overweight and obesity was 15% when measured freshman year and increased to 23% by the end of senior year 7. Pope et al. (2016) also tracked college students across all four years of their undergraduate careers and observed an average weight gain of 9.64 lbs. Initially, 23% of freshmen were classified as overweight or obese based on BMI, and this increased to 41% by the end of senior year 34 . To date, no data exists over the four-year duration of college reporting the composition of weight gain, such as FM and LM, and to our knowledge only five studies have evaluated change in weight beyond the freshman class.

Not all literature surrounding college weight gain neglects to measure body composition.

Few studies have looked further than BMI, using DEXA and BIS to estimate body composition.

Morrow et al. (2006) used DEXA to estimate body composition of freshman women at the beginning and end of the academic year. They observed a significant increase in body mass

(BM), FM, FFM, and %BF. However, the women in this study gained 2.4 lbs. over the course of the year, which, although statistically significant, is far less weight gain compared to previous research 7,35 . FM accumulated comprised of 1.75 lbs, or 73% of mean weight gain observed 6. Hull et al. (2007) characterized body composition changes occurring across the sophomore year of college using DEXA 8. The results of this study contradict the general trend of most college weight gain studies; although body weight did not significantly change (60.4 kg vs. 60.6 kg), females during their sophomore year significantly lost weight (-0.2 kg) decreased %BF (-1.0%) and decreased FM (-0.6 kg). Interestingly, these favorable changes were repeated when subjects were analyzed according to living situation (on-campus vs. off campus), with the off-campus students significantly decreasing FM (-1.2 kg) and %BF (-2.0%) while increasing LM (1.1 kg) 8.

On-campus students did not change significantly in terms of body weight nor body composition.

Gropper et al. (2012) observed significant loss of FFM in females (-0.5 kg ± 2.2) between the

12 beginning of freshman year and the end of junior year. This same group concluded the majority of weight gain (2.1 ± 4.8 kg) observed in their three-year study was comprised of FM (2.1 ±

4.8kg body mass gain vs. 2.3 ± 3.5 kg FM gain). Interestingly, males in this study gained significantly more weight, %BF, and FM than females (3.6 ± 4.8 vs. 1.3 ± 4.5 kg; 3.7 ± 2.9 vs.

2.2 ± 3.3%; 3.3 ± 3.0 vs. 1.9 ± 3.5 kg FM, respectively) 5.

The majority of studies surrounding college weight gain only compare weight changes over the course of one academic year, failing to take tissue-specific or regional distribution changes into account 2,21 . Future research in this area should utilize criterion methods for body composition measurement, in order to track changes in LM, FM, VAT, and regional distribution of body composition across each academic year. The stages of the collegiate career in which weight gain is more commonly observed would help illuminate time points at which health behavior interventions could be stressed.

Normal Weight Obesity

Normal weight obesity (NWO) describes individuals with normal body weight and BMI

(18.5-24.9 kg/m 2) with increased %BF, a value which lacks definitive universal criteria in the literature surrounding NWO. Proposed %BF cutoff points, values which individuals who exceed are deemed NWO, range between 20-25% for men and 30-37% for women across the current literature 9. Based on the 2005 United States consensus, Romero et al. (2008) estimated 30 million Americans are considered NWO 10 . Multiple studies exist which attempt to estimate the prevalence of NWO in specific populations. Pertaining to young adults, Olafsdottir et al. (2016) studied a population of eighteen-year-old high school students (n=182) in the capital area of

Iceland 102 . This group used %BF cutoffs of >17.6% for males and >31.6% to assign NWO status

13 and used DEXA to estimate body composition. NWO prevalence was 42% in this population

(n=76), 61% of which were male. This study is one of very few investigating NWO in young adults. Marques-Vidal et al. (2008) aimed to assess the prevalence of NWO in Switzerland using different %BF cutoff points 11 . %BF was estimated in 3213 women and 2912 men aged 35-75 using bioelectrical impedance analysis. Thereafter, subjects were deemed NWO or NWL based off four different %BF cutoff points. Cutoff points ranged from 26-32.6% in men and 30-44.4% in women. In men, prevalence of NWO was <1% irrespective of the definition used. Conversely, prevalence of NWO ranged from 1.4 to 27.8% in women. These results suggest age and gender- specific %BF cutoff points should be used when assessing NWO status.

Not dissimilar to obesity, individuals with NWO tend to develop characteristic health conditions such as insulin resistance, proinflammatory status, oxidative stress, and dyslipidemia, all which may lead to increased risk of CVD, metabolic syndrome, or another form of cardiovascular-related death 12 . Romero-Corral et al. (2008) reported that individuals >20 years old with NWO had lower concentrations of HDL, higher concentrations of LDL and triglycerides and altered ratios of apolipoprotein B to apolipoprotein A1 when compared to NWL individuals.

On the other hand, De Lorenzo et al. (2006) found no differences in lipid profiles between NWO and NWL Italian women between the ages of 20-35 89 . Kim et al. (2013) found NWO Korean individuals with an average age of 53.4 years old had a higher risk of developing one or more cardiovascular risk factor (dyslipidemia, hypertension and/or hypoglycemia) than individuals of normal %BF 88 . Romero et al. (2008) concluded NWO is associated with increased cardiometabolic dysfunction, metabolic syndrome, and cardiovascular risk factors after analyzing 6171 subjects >20 years old from the Third National Health and Nutrition

Examination Survey (NHANES) and the NHANES mortality study. Olafsdottir et al. (2016)

14 found NWO to be positively associated with having one or more risk factors for metabolic syndrome when adjusted for sex in adolescent Northern Europeans. In addition, cardiorespiratory fitness was assessed in this study. Aerobic fitness was 5.1% lower in the NWO group compared to the normal weight lean (NWL) group. These findings strongly support the need to promote a healthy lifestyle with respect to diet and physical activity in young adulthood, regardless of BMI, as metabolic syndrome is typically diagnosed later in life.

Lifestyle Factors

Approaches to combat the obesity epidemic regularly involve lifestyle alterations to both diet and exercise habits. Weight gain is linked to increased energy consumption 42 and decreased energy expenditure 43 . The college years, most notably the freshman year, have been identified as a period of high risk for weight gain due to the long-term development of health and lifestyle behaviors 37 . One in four people between the ages of 18 and 24 years in the United States is a current full-time or part-time undergraduate student 44-45 , and 50% of all persons between the ages of 20 and 24 years has attended college 46 . Many studies have demonstrated first-year college students in the U.S. gain more weight on average than aged-matched individuals who do not attend college 38 . A 2005 survey indicates 3 of every 10 college students is either overweight or obese, and nearly 6 of every 10 participate in fewer than 3 days per week of vigorous intensity or moderate intensity physical activity 87 . Certain psychosocial and behavioral factors contribute to the innate tendency for college students to change preexisting exercise and dietary habits often leading to acute weight gain and an increased risk to adopt long-lasting poor health behaviors 38-

39 . It is vital to identify key variables contributing to regular weight gain seen in college students.

Physical Activity

Physical inactivity among college students is a driving force of weight gain observed in college students. According to a meta-analysis conducted by Keating et al. (2010), 40-50% of college students are physically inactive. Along with this, only 50% of students meet the

American College of Sports Medicine’s minimum physical activity recommendations 47 .

Sedentary behavior is highly prevalent in this population and has been positively correlated to weight gain 48-49 . A meta-analysis identified low levels of physical activity among college students as a predictor of weight gain when 17 studies were compiled 21 . Intensity of exercise is seldom reported in the existing literature, but exercise type has been characterized in a college student population 21 . Racette et al. (2005) report 59% of college freshmen participated in aerobic exercise, 45% in strengthening exercise, 31% in stretching exercise, and 29% of students did not participate in exercise regularly in a 2008 study. The only observed significant change that occurred from freshman year to senior year was a significant increase in stretching exercise

(p=0.013) which directly would not affect body composition 7. Most investigators report more physical inactivity in college-aged women than in college-aged men, with 10-37% of men and

22-48% of women reporting no physical activity in the past month 96 . Butler et al. (2004) observed decreased physical activity across all forms measured, including leisure, sport, occupational, and total activity 98 . Another study revealed the percentage of students participating in moderate- and high-intensity strength training decreased from 62.5% to 45.9% in men but increased from 37.8% to 42.1% in women over the course of freshman year 97 . Overall, college students fail to meet exercise guidelines, despite having free access to recreational facilities and organized intramural sports and leisure activities.

Alcohol Consumption

Regular and excessive alcohol consumption is considered a cultural norm in the college environment, which may contribute to increased alcohol-related activity on college campuses 40 .

De Vos et al. (2015) surveyed 1095 freshman students and determined 30.7% of the sample increased alcohol consumption over the course of their first semester (Sept. to Dec.) 18 . This increase in alcohol consumption was found to be a significant determinant of weight gain in

Dutch freshmen who gained weight across one semester. A study conducted by Bergen-Cico et al. attributed alcohol abuse to 79% of emergency care visits amongst first-year college students 40 . Alcohol itself is very calorically dense, providing 7 calories per gram consumed.

Additionally, alcohol consumption may induce alcohol-related eating, which was associated with a 25% increase in overweight status in a 2009 study conducted by Nelson et al.83 .

Dietary Habits

Most colleges offer “all-you -can-eat” style meal plans to students who are, in most cases, recently removed from parental supervision and are nutritionally autonomous for the first time 35,41 . This new-found independence may increase one’s susceptibility to making poor food choices or overconsumption of food in individuals with minimal experience making independent food decisions. College students do not meet recommended intake of whole grains, vegetables and calcium and consume excessive sugar-sweetened beverages 60 . Researchers found that students consume excess amounts of total fat, saturated fat, cholesterol and sodium 87 . According to students, their poor dietary decisions are made due to minimal healthy foods available on campus, in restaurants, and in their dorm rooms 82-85 . All of these serve as reasons in which many

17 describe the college environment as “obesogenic,” leaving college students in a disadvantaged position to maintain a healthy weight or improve body composition.

Stress

Although poor diet and exercise habits act as main drivers toward obesity, other factors are known to play smaller, yet significant roles in weight gain. Psychosocial stress has been associated with high blood pressure, cardiovascular disease, cancer, and recent evidence suggests stress can lead to long-term weight gain 77-78 . Stress is thought to lead to weight gain through neuroendocrine and inflammatory pathways which directly influence adiposity. Stress is also thought to lead to obesity by affecting health behaviors such as diet and exercise. College has proven to be a stressful time for many students. Past studies reported 75% to 80% of college students are moderately stressed and 10% to 12% are severely stressed 79-80 . Clinically, it is estimated that 11.9% of college students have a diagnosed anxiety disorder, and 7-9% of college students are diagnosed with depression 90 . Stress may lead to weight gain by reducing time available to prepare food and increase desire for high-fat energy-dense meals 78 . Additionally, stress has been shown to decrease participation in leisure-time physical activity 78 .

Epidemiological evidence has weakly linked stress with weight gain 78 . One key conclusion of this 2011 meta-analysis was the need to identify moderating variables of the relationship between stress and adiposity. Prior research has suggested differences in sex, initial BMI, and cortisol reactivity may each cause some people to gain more weight during extended stressful circumstances, while others may gain less weight or even loss weight when stressed 100 .

Sleep

Several studies show a relationship between overweight and obesity with sleep, specifically with shorter sleep duration, poor sleep quality, and variability in day-to-day sleep times 67-69 . Insufficient sleep has been associated with obesity and adverse changes in metabolic health, as well as poor impulse control, risk-taking behaviors and other psychological and cognitive defects which could contribute to weight gain 70-74 . Poor sleep duration has also been identified as a risk factor for coronary heart disease, cerebrovascular disease, hypertension on the basis of epidemiological data 94 . Interestingly, risk of the mentioned conditions follows a U- shaped curve, with sleep deficit and excess sleep both being associated with increased risk for cardiovascular disorders 94 . Similarly, both short and long duration of sleep have been associated with increased risk of all-cause mortality. Short sleep has been defined by numerous maximum sleep hour values (≤ 5 hours per night, ≤ 6 hours, ≤7 hours) in previous studies, and long sleep has been defined as ≥8 hours, ≥9 hours, ≥10 hours, ≥12 hours 95 . Sleep restriction is proposed to induce weight gain via endocrine dysfunction, specifically via modulation of leptin and ghrelin 75-

76 . Some studies report a later bedtime could be associated with obesity due to more time available to consume food and less access to healthy foods. Interestingly, it has been suggested that males are more susceptible to weight gain due to poor sleep patterns 65 . Roane et al. (2015) investigated the association between sleep duration and variability with weight gain in college freshmen and found students who gained weight in the study experienced significantly greater variability in sleep duration compared to students who lost weight 65 . Culnan et al. (2013) evaluated the relationship between weight gain and sleep chronotype using the reduced Horne-

Östeberg Morningness Eveningness questionnaire (rMEQ) 66 . The rMEQ defines the times of day which a person prefers be active or asleep, categorizing individuals as either morning-type,

19 evening-type, or neutral. Morning-types prefer to wake earlier and fall asleep earlier in the evening relative to evening-types. Culnan and colleagues concluded being considered an evening type according to the rMEQ was associated with a 0.50 increase in BMI, or a 2.35 lb. weight gain in college students over the course of the eight-week study 66 .

Questionnaires and Surveys

In order to adequately assess the factors contributing to weight gain observed in college students, valid and reliable questionnaires that exist in the literature surrounding these topics will be administered. Questionnaires are practical and economical means to measure outcomes. As mentioned, physical inactivity has been shown to be a major contributor to weight gain, which warrants the use of a physical activity questionnaire in this study. One of the most commonly applied physical activity questionnaires is the International Physical Activity Questionnaire

(IPAQ), developed in 1996 as an instrument to be used in diverse countries and populations 50 , with the reliability and validity having been tested in more than 12 countries 50-59 . Most importantly, the IPAQ has been deemed reliable and valid in college-aged populations 58,60 . In addition to the IPAQ’s assessment of inactivity, the Sedentary Behavior Questionnaire (SBQ) will be used to evaluate individuals’ time spent in a sedentary state during nine specific sedentary behaviors. The IPAQ and SBQ correlate well, with intraclass correlation coefficients acceptable for all items assessed (range=0.51-0.93)64 . Diet is another crucial area of lifestyle to be considered. A 3-day diet log has been used in research to evaluate an individual’s dietary intake. Yang et al. (2010) concluded that a 3-day food record to be an acceptable dietary assessment tool 61 . Stress can have an impact on overall health and weight via stimulation of the hypothalamic-pituitary-adrenal cortex axis 62 . The Perceived Stress Scale (PSS) is the most

20 widely used instrument to evaluate the perception of stress 63 . Use of the PSS will allow for parallels to be made between students’ perceived stress levels and changes in body composition and cardiometabolic health. Additionally, the Depression, Anxiety, and Stress Scale-21 items

(DASS-21) will be administered. DASS-21 is a set of three self-report scales designed to examine the emotional states of depression, anxiety and stress within individuals. Sleep quality and duration will be assessed using the Pittsburgh Sleep Quality Index (PSQI), and individuals’ diurnal preferences, sleep-wake pattern for activity, and alertness in the morning and evening will be assessed using the Reduced Horne-Östeberg Morningness Eveningness questionnaire

(rMEQ). The College Alcohol Problems Scale (CAPS-r) and the Alcohol Use Disorders

Identification Test (AUDIT) will be used to evaluate subjects’ alcohol consumption and related behaviors. Subjects will also be asked about substance use (i.e. caffeine, recreational drugs). In addition, subjects will be asked to report living status (on- or off-campus).

Conclusions

Young adults have experienced the greatest overall rise in overweight and obesity in recent years when compared to all other age groups 3. Since college students make up a large proportion of this group, body composition and cardiometabolic health should be evaluated and described in this population to elucidate dietary, physical activity, or any other health behaviors that could play a role in the commonly observed weight gain in this population, especially since many behaviors established during college are solidified through early adulthood and beyond 37 .

Prior research has focused solely on weight gain in college freshmen across one year, neglecting to adequately investigate changes in body composition and cardiometabolic health in this population. Additionally, few studies include sophomores, juniors and seniors. Likewise, few

21 studies in this area have utilized DEXA to assess body composition in addition to weight gain. A complete evaluation of body composition via DEXA, cardiometabolic health, and health behavior in college students representing all four classes could illuminate and potentially discourage behaviors leading to negative health consequences and instead promote behaviors correlated with positive health standing to improve health during college and thereafter.

CHAPTER III: METHODS

Participants

Male and female subjects currently participating in an ongoing study characterizing normal weight obesity in college students (Study IRB#: 17-0952), which has identical methods and outcomes, were included in this study. Based off previous literature 5,29 a sample size of 54 subjects was calculated. Significant changes in %BF were detected between freshmen and juniors within each individual (18.3 ± 7.5% vs. 21.0 ± 7.4%) giving an effect size of 0.405, and significant changes were detected between freshmen and juniors within-individuals in FM (12.0

± 8.1 kg) vs. 14.3 ± 9.1 kg) giving and effect size of 0.297 5. These effect sizes were combined and averaged which gave a sample size of 54 subjects. Power was set a priori to 0.80, correlation among repeated measures to 0.6, and two-tailed analysis was employed. Sample size was calculated using G*Power version 3.1.

In order to be included in this study, subjects were required to have a BMI between 18.5-

24.9 kg/m 2. Subjects were required to be healthy and have no medical issues that would potentially influence the results of this study, including renal, hepatic, musculoskeletal, or cardiometabolic disorders. Subjects were excluded if they had gained or lost ten pounds three months prior to enrollment or had a self-diagnosed eating disorder. Subjects were instructed to complete a three-day diet log and sign an approved informed consent from prior to enrollment.

Subjects were included in the study if they participated in ≤ 5 days of exercise per week and were not NCAA student-athletes.

Experimental Design

Before each visit, subjects were asked to arrive to the lab following a fast ≥8 hours.

Anthropometric measurements were assessed, including weight (kg), height (cm), and BMI.

Subjects completed a comprehensive assessment of body composition using DEXA. Lastly, a blood sample was taken to evaluate measures of lipids and glucose, and questionnaires were completed to assess various lifestyle factors including alcohol consumption and related behaviors/emotional states, physical activity sleep duration and quality, and perceived stress. In a subsequent cross-sectional approach (Fall 2018), newly-enrolled subjects (n=43) were recruited to account for attrition and graduation from the previous year, as well as to further explore more detailed aspects of lifestyle habits. Questionnaires used to examine sedentary behavior, sleep quality and preferences, alcohol consumption, and mental health were included in addition to variables previously measured in the Fall and Spring assessment periods.

Body Composition

Dual-energy X-Ray absorptiometry (GE Lunar iDEXA, GE Medical Systems Ultrasound and Primary Care Diagnostics, Madison, WI, USA) was used to estimate LM, FM, %BF and

VAT. All scans were performed by a trained DEXA technician. Before each scan, the trained technician entered the necessary subject information including height, weight, ethnicity, age, sex, and identification code. After ensuring the subject had removed all metal jewelry and any other

24 objects from their pockets, the technician instructed the subject to lie supine on the DEXA scanner. Once on the DEXA table, the technician adjusted the subject’s head, hips, shoulders, and limbs to center the subject within the DEXA’s measurement area. The subject’s hands lay pronated on the DEXA table next to their legs. Once in the correct position, a Velcro strap was placed around their ankles to hold the correct position. The subject was instructed to limit movement and to not alter breathing during the scan. The scan lasted between 7-13 minutes based on body size.

Cardiometabolic Health

A single blood sample was obtained during both Fall visits (Fall 2017 and Fall 2018) to assess cardiometabolic health. Blood samples were not collected in the spring of 2018. The blood sample was taken from the antecubital fossa, specifically from the median cubital, cephalic, or basilic vein. All samples were analyzed immediately post-draw using an Alere Cholestech

LDX® Analyzer (Alere Inc., Waltham, MA, USA) to determine total cholesterol (TC), high- density lipoprotein (HDL), non-high-density lipoprotein (nHDL), and glucose (GLU). 40 µL of blood were aliquoted from the sample into a test cassette and placed in the device for analysis.

Questionnaires

During each visit, study participants completed multiple validated self-administered questionnaires regarding their lifestyle. The Shortened International Physical Activity

Questionnaire (IPAQ) is a 7-item survey that was used to gather data on physical activity over 7 days prior to the visit. The perceived stress scale (PSS) is a survey consisting of 14 items which was used to evaluate stress levels. Additionally, the Depression, Anxiety, and Stress Scale-21 items (DASS-21) was administered. The DASS-21 is a set of three self-report scales designed to

25 examine the emotional states of depression, anxiety and stress within individuals. The Pittsburgh

Sleep Quality Index (PSQI) was given to provide an indication of subjects’ perceived sleep quality, sleep efficiency, and daily disturbances. The Morningness-Eveningness Questionnaire

(rMEQ) was used to determine subjects’ diurnal preferences, sleep-wake pattern for activity, and alertness in the morning and evening. The revised College Alcohol Problems Scale (CAPS-r) and the Alcohol Use Disorders Identification Test (AUDIT) were both used to evaluate alcohol consumption and related behaviors. Sedentary behavior was assessed using the Sedentary

Behavior Questionnaire (SBQ). Additionally, participants were asked about their living arrangements, including if they lived on campus or off, and how many roommates they had.

Statistical analysis

Analyses were performed using SPSS software (IBM Corp. Released 2016. IBM SPSS

Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp.). Repeated measures analysis of variance (ANOVA) [group (freshman vs. sophomore vs. junior vs. senior) × time (fall 2017 vs. spring 2018)] followed by Bonferroni post-hoc analysis were employed to examine changes in body composition and anthropometric outcome measures. Repeated measures analysis of variance (ANOVA) [group (freshman vs. sophomore vs. junior vs. senior) × time (fall 2017 vs. fall 2018)] followed by Bonferroni post-hoc analysis were employed to examine changes in cardiometabolic health. An additional sex (male vs. female) × group (freshman vs. sophomore vs. junior vs. senior) ANOVA was employed to assess sex-specific changes observed across class year. Pearson correlations were used to assess the direction and magnitude of relationships between nominal lifestyle factors and body composition changes observed across each academic year. Similar Pearson correlations were employed to assess the direction and magnitude of the

26 relationships between nominal lifestyle factors and observed cardiometabolic changes. A multiple regression was used to examine the relationship between categorical lifestyle factors with body composition and cardiometabolic changes observed. An alpha-level of 0.05 was employed in all statistical analyses. All data are presented as mean ± standard deviation unless noted otherwise.

CHAPTER IV: MANUSCRIPT

Introduction

According to the World Health Organization more than 1.9 billion adults are overweight, with 650 million of these individuals considered obese 1. Across the United States, it is the young adult age group (18 to 29 years) that has experienced the greatest increases in overweight and obesity 3. Paralleled with this increase in overweight and obesity, full-time college enrollment increased 15% between 2005 and 2015 4. The college years may be a critical time period during which lifestyle and health behaviors are established. The college setting introduces many obesogenic factors including unlimited access to dining halls, extended periods of sedentarism, initial exposure to alcohol, and high amounts of stress. The common belief that college students gain 15 lbs during their first year of college, commonly referred to as the “Freshman 15,” has been investigated and ultimately dismissed in research, with an average weight gain of 3.85 lbs occurring during the first year of college 2. This nontrivial weight gain is generally attributed to significant increases in body fat percentage (%BF), fat mass (FM) and waist circumference during the freshman year 5,6 . Less is known about these changes in body composition among other classes, with some initial evidence suggesting a similar trajectory, although conflicting evidence exists 5,7-8. Data from 2012 reported 70% of students gain weight over the course of the first three years of college, and a 1.4 ± 1.2 kg/m 2 increase in body mass index (BMI), a 3.5 ± 3 kg increase in FM, and a 4.0 ± 3.1% increase in %BF were all seen in these weight-gainers 5.

Contradictory to these findings, Hull et al. (2007) observed significant decreases in %BF, FM, and a significant increase in fat-free mass (FFM) among female college sophomores.

Change in body weight is typically a primary outcome in any “freshman 15” or other college weight-gain studies. However, by focusing solely on body weight, many studies are unable to determine changes in lean mass (LM) or FM, which are both closely associated with health and functionality 20 . Measuring body composition changes, not simply body mass, will provide descriptive and meaningful data to fully evaluate how health may change over the course of an academic year in the college setting. Specifically, utilizing DEXA to estimate body composition provides the ability to establish which tissues make up the body, including FM and

LM, so weight gain or loss can be further defined.

In addition to body composition, cardiometabolic health of this young adult population is not typically presented in college weight gain studies. More than one half of young adults (18-24 years) have at least 1 coronary heart disease (CHD) risk factor 106 . However, due to the majority of these students’ healthy outward appearances, many do not regularly assess their risk for CHD.

Students are at critical point in life to set diet and physical behavior standards for themselves, and it has been suggested that up to 80% of heart disease is preventable through diet and exercise 106 .

The purpose of this study was to add to the existing “freshman 15” literature, but also to contribute to literature surrounding body composition changes (including body fat percentage, fat mass and lean mass) seen across one college year in male and female students of all classes. We hypothesized freshmen would see significantly greater increases in weight, FM, and %BF compared to the other three classes. We also predicted males would gain significantly more weight than females, as this has been observed in previous studies 5,7,104 . Additionally, this study

29 aimed to describe cardiometabolic health changes (including total cholesterol, high density lipoproteins (HDL), non-high density lipoproteins (nHDL), and fasting glucose) that occur across one college year in these students. To our knowledge, this has not been previously investigated across all undergraduate class years. Lastly, lifestyle and behavioral factors were investigated, and the relationship with body composition and cardiometabolic health.

Participants and Methods

Male and female subjects currently participating in an ongoing study characterizing normal weight obesity in college students (Study IRB#: 17-0952) were included in this study.

Data collection took place at three separate time points: beginning of Fall semester (Aug-Oct.

2017; n=143); end of spring semester (April-May 2018; n=59 returned); and beginning of Fall semester (Aug-Oct. 2018; n=85, including returners and new participants). A total of 143 students (mean ± SD: Age: 20.3 ± 1.5 years; Height: 169.7 ± 9.8 cm; Weight: 63.7 ± 9.5 kg;

BMI: 22.0 ± 1.7 kg/m 2) (54 males and 89 females) (Table 1) were included in analyses across all three time points (Figure 1). Participant dropout occurred for reasons including graduation, busy schedules, general loss of interest, and lost to follow-up. Participants were restricted to those admitted as full-time students at the university and were 18-25 years old, a BMI between 18.5-

24.9 kg/m 2, healthy and free from medical issues that would potentially influence the results of this study, including renal, hepatic, musculoskeletal, or cardiometabolic disorders. Subjects were excluded if they had gained or lost ten pounds during the three months prior to enrollment or had a self-diagnosed eating disorder. Subjects were included if they exercised ≤ 5 days per week and were not NCAA student-athletes. All subjects signed a written Informed Consent document approved by the University’s Biomedical Institutional Review Board.

Experimental Design

During each of the three visits, subjects were asked to arrive to the lab following a fast lasting ≥8 hours. Anthropometric measurements were assessed, including weight (kg), height

(cm), and BMI. Subjects completed a comprehensive assessment of body composition using dual energy x-ray absorptiometry (DEXA). Lastly, a blood sample was taken to evaluate measures of lipids and glucose, and questionnaires were completed to assess various lifestyle factors including alcohol consumption and related behaviors/emotional states, physical activity, sleep duration and quality, and perceived stress. In a subsequent cross-sectional approach (Fall 2018), newly-enrolled subjects were recruited to account for attrition and graduation from the previous year, as well as to further explore more detailed aspects of lifestyle habits. Figure 1 visually depicts the sequence of the three visits, including subject participation and dropout rates.

Questionnaires used to examine sedentary behavior, sleep quality and preferences, alcohol consumption, and mental health were included in addition to variables previously measured in the Fall and Spring assessment periods.

Anthropometric and Body Composition Analysis

Weight was measured using a calibrated digital scale (Health o meter, Countryside, IL,

USA) and height measured using a stadiometer (Perspective Enterprises, Portage, MI, USA).

Subjects removed all heavy clothing and shoes before weight and height were measured. Body mass index was calculated using the standard formula, dividing weight by height squared

(kg/m 2).

Body composition was measured using DEXA (GE Lunar iDEXA, GE Medical Systems

Ultrasound and Primary Care Diagnostics, Madison, WI, USA). All scans were performed by a

31 trained DEXA technician. Before each scan, the trained technician entered the necessary subject information including height (in), weight (lbs), ethnicity, age, sex, and identification code. After ensuring the subject had removed all metal jewelry and any other objects from their pockets, the technician instructed the subject to lie supine on the DEXA scanner. Once on the DEXA table, the technician adjusted the subject’s head, hips, shoulders, and limbs to center the subject within the DEXA’s measurement area. The subject’s hands rested in a pronated position on the DEXA table beside their upper legs. Once in the correct position, a Velcro strap was placed around their ankles to hold the correct position. The subject was instructed to limit movement and to not alter breathing during the scan. The full-body scan lasted 7 minutes. The test-retest reliability for the

DEXA for FM is as follows: intra-class correlation coefficient (ICC)= 0.98 and standard error of mean (SEM)= 0.85 kg. For LM, ICC= 0.99 and SEM= 1.97 kg. For %BF, ICC= 0.96 and SEM=

1.279% 108 . Changes in body composition were measured between Fall 2017 and Spring 2018 visits, encompassing the entire academic year.

Cardiometabolic Health

A single blood sample was obtained during both Fall visits (Fall 2017 and Fall 2018) to assess cardiometabolic health. Only students who participated during both Fall 2017 and Fall

2018 visits were included in this analysis (males: n=4; females n=27) Blood samples were not collected in the spring of 2018. The blood sample was taken from the antecubital fossa, specifically from the median cubital, cephalic, or basilic vein. All samples were analyzed immediately post-draw using an Alere Cholestech LDX® Analyzer (Alere Inc., Waltham, MA,

USA) to determine total cholesterol TC, HDL, nHDL, and GLU. 40 µL of blood were aliquoted from the sample into a test cassette and placed in the device for analysis. Changes in

32 cardiometabolic health were measured between Fall 2017 and Fall 2018, encompassing an entire calendar year rather than the academic year.

Lifestyle Questionnaires

During each visit, study participants completed multiple validated self-administered questionnaires regarding their lifestyle. The Shortened International Physical Activity

Sleep Quality Index (PSQI) was given to provide an indication of subjects’ perceived sleep quality, sleep efficiency, and daily disturbances. The Morningness-Eveningness Questionnaire

(rMEQ) was used to determine subjects’ diurnal preferences, sleep-wake pattern for activity, and alertness in the morning and evening. The revised College Alcohol Problems Scale (CAPS-r) and the Alcohol Use Disorders Identification Test (AUDIT) were both used to evaluate alcohol consumption and related behaviors. Additionally, participants were asked about their living arrangements, including if they lived on campus or off, and how many roommates they had.

Lifestyle

Statistical Analysis

Analyses were performed using SPSS software (IBM Corp. Released 2016. IBM SPSS

Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp.). Repeated measures analysis of

33 variance (ANOVA) [group (freshman vs. sophomore vs. junior vs. senior) × time (fall 2017 vs. spring 2018)] followed by Bonferroni post-hoc analysis were employed to examine changes in body composition, cardiometabolic, and anthropometric outcome measures. An additional sex

(male vs. female) × group (freshman vs. sophomore vs. junior vs. senior) ANOVA was employed to assess sex-specific changes observed across class year. Pearson correlations were used to assess the direction and magnitude of relationships between nominal lifestyle factors and body composition changes observed across each academic year. Similar Pearson correlations were employed to assess the direction and magnitude of the relationships between nominal lifestyle factors and observed cardiometabolic changes. A multiple regression was used to examine the relationship between lifestyle factors and body composition and cardiometabolic changes observed. An alpha-level of 0.05 was employed in all statistical analyses. All data are presented as mean ± standard deviation unless noted otherwise.

Results

Anthropometrics

For body mass, there was no significant interaction (p=0.621) and no main effect for group (p=0.614), there was a main effect for time (p=0.035) with a significant increase in body mass over time (FALL17: 61.3 ± 9.5 kg, SPRING18: 62.2 ± 9.3 kg) in the whole sample. BMI also demonstrated no significant interaction (p=0.532) and no main effect for group (p=0.856), but a significant main effect for time (p=0.025). BMI increased significantly from FALL17 (21.8

± 1.7 kg) to SPRING18 (22.1 ± 2.1 kg) in the whole sample. Figure 2 displays the distribution in the observed changes in body weight over the course of the 2017 to 2018 academic year. One male student (28.8 lbs) and one female student (16.0 lbs) gained ≥15lbs over their freshman year.

Body Composition

For FM there was no significant interaction (p=0.125), no main effect for time (p=0.291), and no main effect for group (p=0.839). For LM, there was no significant interaction (p=0.862) and no main effect for group (0.391). There was a significant main effect for time (p=0.007) with a significant increase in LM from FALL17 (42.7 ± 8.1 kg) to SPRING18 (43.4 ± 8.0 kg). Seniors gained a significant amount LM (0.8 ± 1.1 kg, p=0.010) from FALL17 to SPRING18. For %BF there was no significant interaction (p=0.097), no main effect for time (p=0.948), and no main effect for group (p=0.527). Table 2 includes differences in body composition and anthropometric changes observed in students of each class. The only significant change was seen in LM of seniors, which increased by 0.8 ± 1.1 kg (p=0.01).

Cardiometabolic Health

No significant interaction (p=0.572) and no main effect for group (p=0.755) was observed for glucose. Glucose resulted in a significant main effect for time (p=0.046), decreasing from FALL17 (88.1 ± 6.7 mg/dL) to FALL18 (86.7 ± 7.4 mg/dL). Glucose significantly decreased between FALL17 to FALL18 in males (-7.5 ± 2.5mg/dL). For TC, no significant interaction (p=0.521), no main effect for time (p=0.437), and no main effect for group (p=0.132) was observed. HDL yielded no significant interaction (p=0.930), no significant main effect for time, nor was there a main effect for group (p=0.417). For nHDL, there was no significant interaction (p=0.595), no main effect for time (p=0.908), and no main effect for group (p=0.355).

No significant differences were detected between any class and each of the cardiometabolic health outcomes (Table 5).

Males vs. Females

For body mass, there was no significant interaction (p=0.983). There was a significant main effect for time (p=0.038) and main effect for group (p<0.001). Females weighed significantly less during FALL17 (58.3 ± 6.2 kg) compared to SPRING18 visit (59.2 ± 6.3 kg)

(p=0.015). Males did not present a significant change in body mass (p=0.394) over this one-year period. Figures 3 and 4 display the distribution in the observed changes in body weight over the course of the 2017 to 2018 academic year in males and females, respectively. For BMI, there was a significant main effect for time (p=0.037), but no significant main effect for group

(p=0.816) and no interaction (p=0.448). BMI significantly increased from FALL17 (21.7 ± 1.6 kg/m 2) to SPRING18 (22.0 ± 1.9 kg/m 2) in females (p=0.015). No significant interaction was observed between sex and FM (p=0.983). A significant main effect for time (p=0.038) and main effect for group (p<0.001) were both observed; when decomposing the model, there were no significant changes in FM observed in males or females (p=0.567; p=0.627, respectively).

Similar to FM, LM yielded no significant interaction (p=0.707), but there was a significant main effect for time (p=0.006) and main effect for group (p<0.001). LM of males did not significantly change over one academic year [FALL17: 53.7 ± 7.2 kg; SPRING18: 54.2 ± 7.6 kg (p=0.221)], but LM of females significantly increased from FALL17: 53.7 ± 7.2 kg to SPRING18: 54.2 ± 7.6 kg (p=0.002). For %BF, there was no significant interaction (p=0.496) and no main effect for time (p=0.946). There was a significant main effect for group (p<0.001), with males reporting lower %BF than females. Table 3 compares all changes in body composition and anthropometrics observed in males vs. females.

No significant interaction (p=0.325) was observed between sex and TC. Main effects for time and group were not statistically significant (p=0.088 and p=0.076, respectively). There was

36 no significant interaction between HDL and sex (p=0.762), and no significant main effect for time (p=0.307). There was a significant main effect for group (p=0.035). For nHDL, there was no significant interaction (p=0.197), no main effect for time (p=0.197) and no main effect for group (p=0.517). Lastly, there was no significant interaction between fasting glucose and sex

(p=0.126), and there was not a significant main effect for group (p=0.836). There was a significant main effect for time (p=0.019). A significant decrease in fasting glucose from

FALL17 (91.3 ± 5.0 mg/dL) to FALL18 (83.8 ± 7.1 mg/dL) was observed in males (Table 4). No significant change in fasting glucose was observed in females (p=0.243). Table 3 compares all changes in cardiometabolic health observed in males vs. females.

Lifestyle Factors

Total FM was significantly negatively correlated with total LM (r=-0.429, p<0.001) and total minutes of vigorous physical activity measured via IPAQ survey (r=-0.387, p<0.001). In addition to FM, total LM was significantly positively associated with VAT (r=0.314, p=0.003) and total minutes of vigorous physical activity (r=0.283, p=0.010). In terms of cardiometabolic health, TC was positively associated with total minutes of moderate physical activity measured via IPAQ survey (r=0.236, p=0.049). There were no other significantly correlated cardiometabolic factors with physical activity (p>0.05).

Regression analyses were conducted to examine the relationship between body composition variables (FM, LM, VAT) and various potential predictors. These 9 selected predictors (Table 6) explained 28.1% of the variance in the data (R 2=0.28, F(9, 61)=2.65, p=0.012) in regard to FM. It was found that minutes of vigorous activity significantly predicted

FM (β=-0.426, p=0.001), as did the subject’s score on the AUDIT questionnaire (β=-0.254,

37 p=0.039). For LM, these same predictors significantly explained 25.2% of the variance in the data (R 2=0.25, F(9,61)=2.284, p=0.028). Again, minutes of vigorous activity per week significantly predicted LM (β=0.416, p=0.001). In addition, rMEQ score significantly predicted

LM (β=0.669, p=0.031). In terms of VAT, the 9 predictors did not significantly predict this outcome (R 2=0.04, F(9,61)=0.286, p=0.976).

Discussion

While an extensive body of evidence exists suggesting weight gain commonly occurs in college, especially during the freshman year, very few studies have assessed changes in body composition, and even fewer have examined changes in cardiometabolic health in combination.

The majority of studies published on this matter have almost exclusively focused on changes in body mass observed in college students and have neglected to consider changes in body composition measures such as FM, LM or %BF, information which could prove to be more insightful 7, 16, 18, 21, 35 . The current study was the first to longitudinally track changes in body composition and cardiometabolic health in undergraduate students of all four classes across one academic year. In our sample, females gained significant amounts of weight (0.9 ± 2.4 kg), increased BMI (0.4 ± 0.9 kg/m 2), and LM (0.7 ± 1.4 kg). However, males did not see significant changes in weight nor body composition. Opposing results were observed between sophomores and juniors in a few body composition variables. Sophomores saw the largest weight gain, which consisted of mostly FM (1.8 of 2.3 kg). Juniors, on the other hand, saw no weight change, while

FM moved in a negative direction (-1.0 ± 2.2 kg) and LM in a positive direction (0.9 ± 1.5 kg).

In general, existing literature suggests freshmen experience the most weight gain compared to the other classes, and some classes have even shown improvements in body composition

2,8,18,21,104,107 . However, many of the results vary between studies, and the majority of studies only include freshmen. We believe our results differ from existing literature, and not many significant findings were discovered due to the healthy nature of our sample. Not only were all of the students normal-weight, but many were recruited from the Exercise and Sport Science

Department, which leads us to believe our sample is more health-conscious than more general samples used in other studies in this area.

Body Mass and BMI

With respect to body mass, changes were modest and widely variable, ranging from -4.27 to 13.1 kg (-9.4 to 28.8 lbs). Across the literature, similar changes in weight gain have reported across one academic year (-5.0 to 20 lbs) 104 , whereas another showed a larger range of weight change (-25.8 to 46.8 lbs) 5. In the present study, average weight gain across all years was 0.9 ±

2.9 kg (2.0 ± 6.4 lbs), with both females and males gaining 0.9 kg (2.0 lbs) on average. Weight gain observed in our sample is slightly smaller in magnitude compared to existing literature which reports an average 3-4 lbs increase over a college year 18, 21, 104, 107 . Specific to freshmen, there was an average 1.1 kg gained in the current sample; Mihalopolous et al. 104 observed a similar 1.2 kg (2.7 lbs) weight gain in a sample of 125 college freshmen. In contrast, there was a significantly greater increase in weight (1.7 kg) for male freshmen compared to a 0.8 kg gain in female freshmen after one academic year 104 . Across the literature, males have often been found to gain more weight than females after the freshman year of college 16, 107 .

Freshmen in our sample gained 1.1 kg, far from the “freshman 15.” One male and one female freshman in our sample gained 15 lbs. Morrow et al. (2006) also only saw one student gain 15 lbs. over the course of the freshman year with a sample of (n=137) 6. The increase in

39 body mass in the male student in our sample showed was largely LM (10.8 lbs.); similar results were seen for the female, with 8.4 lbs being LM.

Body Composition

The only significant finding when comparing body composition changes between classes was a 0.8 kg (1.8 lbs) increase in LM within the senior class of students over the course of the year. However, since this value falls within the SEM (1.97 kg) of the DEXA, it is likely this change is within the error of the DEXA technology 108 . Sophomore and junior findings were contrasting; sophomores saw the most weight gain of any class (2.3 kg), with the majority of that weight consisting of FM accumulation (1.8 kg). This resulted in a 1.7% increase in BF%. Juniors saw no change in weight (0.0 ± 2.0 kg) or BMI (0.0 ± 0.7 kg/m 2), with a decrease in FM and increase in LM. BF% decreased by 1.7% in this group. Hull et al. 8 saw a similar discrepancy between freshmen and sophomores in their study conducted in 2007, with sophomores improving in body composition and freshmen seeing negative changes. In their sample, %BF and

FM increased during the freshman year (0.7% %BF and 0.8 kg FM) while they decreased during the sophomore year (-1.0 %BF and -0.6 kg FM). Hull and colleagues ultimately attributed this discrepancy to where students lived in relation to campus (on vs. off-campus living) 8. Students who lived off-campus saw more favorable outcomes when compared to their on-campus counterparts.

Gropper et al. 5 tracked weight and body composition of students for the first 3 years of college and saw significant gains in weight, FM, and %BF during the freshman and junior years, while the sophomore year saw no changes or significant improvements in body composition. The degree of worsened body composition observed in freshmen was about twice in magnitude

40 compared to juniors, with freshmen gaining 1.2 ± 3.1 kg, 1.6 ± 2.5 %BF, 1.4 ± 2.3 kg of FM, and losing -0.2 ± 2.1 kg of FFM, and juniors gaining 0.6 ± 3.8 kg, 0.8 ± 2.9 %BF, 0.7 ± 3.0 kg FM, and losing -0.2 ± 2.0 kg of FFM. When comparing sophomore year vs. freshman year, Hull et al. 8 saw opposing findings between the two classes, with sophomores changing body composition positively and freshmen changing negatively. They also noted students living off campus saw a decrease in %BF (33.0 vs. 31.0%) and FM (19.4 vs. 18.2 kg) and increased FFM

(36.1 vs 37.2 kg) with no difference in students living on campus 8. Subjects in a Hajhosseini et al. 107 study of 27 college freshmen saw a 1.4 kg increase in weight with significant increase in

FM and decrease in FFM. Morrow et al. 6 studied 137 freshmen and saw significant increases in weight (58.6 vs 59.6 kg), BMI (21.9 vs. 22.3 kg/m 2), FM (16.9 vs. 17.7 kg), FFM (38.1 vs 38.4 kg), %BF (28.9 vs. 29.7 %), waist circumference (69.4 vs. 70.3 cm) and hip circumference (97.4 vs. 98.6 cm) from early Fall semester to late Spring semester. While males and females both gained 0.9 kg over the course of one academic year in this study, only females saw significant changes in body composition. Females reported significant increases in weight (0.9 ± 2.4 kg),

BMI (0.4 ± 0.9 kg/m 2), and LM (0.7 ± 1.4 kg). Interestingly, a majority of the weight gained by females consisted of lean mass (0.7 of 0.9 kg), signifying an overall improvement in body composition in this group. This result is similar to the results of Hull et al. 8, where female sophomores gained significant weight (0.2 kg), but also lost significant amounts of %BF (-1.0%) and FM (-0.6 kg).

Cardiometabolic Health

This study was the first to assess changes in cardiometabolic health over the course of an academic year in college students. No differences were observed in cardiometabolic health

41 changes between each of the three classes in which were analyzed (freshman, sophomore, and junior). Freshmen experienced a small decrease in TC, with the majority of that decrease consisting of HDL. Sophomores experienced a net increase in TC, particularly with a lower

HDL:nHDL ratio, a predictor used by physicians to estimate one’s risk of cardiovascular disease.

Juniors saw a decrease in TC, with HDL and nHDL decreasing. The only significant finding was a decrease in fasting glucose seen in males (-7.5 mg/dL). However, due to the small sample size of male Fall 2018 returners (n=4), strong conclusions cannot be drawn from this result. The lack of significant findings in this area of the study could be due to the slow rate of change in blood lipids over time, or possibly due to the young, healthy nature of this cohort, and even the small sample size.

Due to unhealthy dietary choices and eating behaviors observed in college students, we believe significant detrimental changes in cardiometabolic health could likely occur as a result, but may take more than one calendar year to do so. College students have been shown to gain weight up to 11 times faster than age-matched non-college students, and this has been linked with “all you can eat” style college meal plans 106 . Rapid and substantial increases in body weight combined with poor dietary choices can lead to dyslipidemia and cardiovascular disease risk in this young population. The most prevalent cardiovascular risk factors seen in one sample of college students were elevated triglycerides (12% of sample) and low HDL (14% of sample).

Another study observed elevated triglycerides in 18% of their sample and 20% of students had low HDL levels 106 . Our sample may not have seen results similar to this previous literature since participants of this study all exhibited normal BMI values and were generally healthy.

Examining four-year results from these subjects may be necessary in order to detect significant changes in cardiometabolic profile.

Lifestyle Factors

Total LM and vigorous activity were significantly positively correlated (r=0.283, p=0.01); total FM and vigorous activity had the opposite relationship (r=-0.429, p<0.001).

Additionally, vigorous physical activity was found to be a significant predictor for both FM and

LM. These findings support the recommendation for college students to engage in physical activity, especially high-intensity exercise. College campuses offer many opportunities for students to engage in this activity through recreational facilities, group fitness classes, intramural leagues, and sport clubs offered year-round.

Limitations

The results of this study provide valuable information detailing changes in body composition and cardiometabolic health that occur across each of the four academic years of college. However, this study only included normal-weight students, and this can be considered a limitation. Future studies should include students who are classified as underweight and overweight/obese in order to encompass a more generalizable sample. All students involved in this study attended a large public university in the Southeast. Based upon this, results may differ from students attending private or more rural universities. Self-selection bias is likely in this sample since students interested in learning about changes in their body composition and cardiometabolic health may have been more likely to participate than students who were not.

Due to this potential bias, our sample may not be entirely representative of the undergraduate community at this university.

Future Considerations

While no class gained a statistically significant amount of weight, it is speculative, yet entirely likely to propose the possibility that cumulative weight gain in small increments, such as those seen in this study, could lead to significant weight gain after four years of college. For example, the additive mean weight gained by the four classes in our sample was 4.4 kg (1.1 kg +

2.3 kg + 0.0 + 1.0 kg), which could have clinical implications, especially if students’ dietary and physical activity habits remain the same after college. Racette et al.7 measured weight and BMI of male and female students (n=204) at the beginning of freshman year, and four years later, at the end of those students’ senior year. This group saw a significant 2.5 kg increase in body weight and a 0.7 kg/m 2 increase in BMI over the course of four years. This study did not measure weight at any point between the beginning and end of college, but by using existing literature on rapid weight gain seen in freshmen, they concluded weight gain occurs quickly initially and the rate seen during the freshman year does not persist through the rest of college. Future studies should longitudinally track students across a four-year collegiate career to expose trends in weight gain/loss and changes in body composition and cardiometabolic profile.

Future studies should also consider implementing both beginning and end of semester visits to establish magnitude of weight change and change in body composition per semester.

Including a visit at the end of the Fall semester and at the beginning of the Spring semester would also allow the opportunity to investigate weight and body composition changes over the holiday season (mid-December to mid-January) and the summer months, which would add to the few studies which have investigated weight changes during these time periods 105 . Additionally, in order to address self-selection bias, future studies should consider not disclosing DEXA results to study participants to attract a more random sample.

CHAPTER V: CONCLUSION

The results of this study indicate that weight and body composition do not significantly change across one college year, with the exception of a significant increase in LM seen in seniors. No class lost weight on average, which suggests further changes can be made to improve the body composition trajectory of these students across their careers. Only females gained significant body mass, LM, and BMI, with LM making up the majority of weight gain. Vigorous physical activity seemed to play an important role in optimizing body composition in this sample. This result reinforces the value of vigorous exercise on body composition outcomes. No class saw significant changes in their cardiometabolic profile, which could be due to the relatively short-term timeframe of roughly nine months. Freshmen involved in this study were far from undergoing the phenomenon commonly known as the “freshman 15,” with only two students gaining 15 lbs. From this result, establishing healthy diet, exercise, and lifestyle habits should be implemented as early-on in a student’s academic career as possible, but it is equally important to maintain these healthy behaviors throughout all four years of school, since weight gain is commonly seen in the subsequent years.

TABLES

Table 1. Descriptive statistics for the total sample and males vs. females. Data presented as mean ± standard deviation.

Total Sample (n=143) Male (n=54) Female (n=89) Age (years) 20.3 ± 1.5 20.2 ± 1.5 20.4 ± 1.6

Weight (kg) 63.7 ± 9.5 71.0 ± 9.1 59.2 ± 6.5

Height (cm) 169.7 ± 9.8 178.4 ± 8.3 164.4 ± 6.3

BMI (kg/m 2) 22.0 ± 1.7 22.2 ± 1.8 21.9 ± 1.7

Table 2. Change in weight, BMI, and body composition from Fall 2017-Spring 2018 for all subjects and each class (freshman, sophomore, junior, senior). (*) indicates a significant change over time. Data presented as mean ± standard deviation.

Total Freshman Sophomore Junior Senior Sample (n=43) (n=9) (n=21) (n=21) Weight (kg) 0.9 ± 2.9 1.1 ± 3.6 2.3 ± 1.9 0.0 ± 2.0 1.0 ± 1.9

BMI 0.3 ± 1.0 0.4 ± 1.3 0.8 ± 0.6 0.0 ± 0.7 0.4 ± 0.7 (kg/m 2) Fat Mass 0.2 ± 2.2 0.4 ± 2.3 1.8 ± 1.3 -1.0 ± 2.2 0.3 ± 1.9 (kg) Lean Mass 0.7 ± 1.5 0.5 ± 1.8 0.5 ± 0.5 0.9 ± 1.5 0.8 ± 1.1* (kg) Body Fat -0.2 ± 2.5 0.1 ± 2.4 1.7 ± 1.9 -1.7 ± 2.9 -0.2 ± 2.3 (percentage)

Table 3: Mean change in weight, BMI, and body composition from Fall 2017-Spring 2018; male vs. female changes. * denotes a statistically significant change (p<0.05).

Male (n=15) Female (n=44)

Weight (kg) 0.9 ± 4.0 0.9 ± 2.4*

BMI (kg/m 2) 0.3 ± 1.2 0.4 ± 0.9*

Fat Mass (kg) 0.4 ± 2.5 0.2 ± 2.1

Lean Mass (kg) 0.6 ± 1.7 0.7 ± 1.4*

Body Fat 0.2 ± 2.4 0.3 ± 2.6 (percentage)

Table 4: Mean change in Total cholesterol, HDL, nHDL, and glucose from Fall 2017-Fall 2018; male vs female changes. * denotes a statistically significant change (p<0.05).

Male (n=4) Female (n=27) Total Cholesterol -16.0 ± 13.1 -4.4 ± 22.3 (mg/dL) HDL (mg/dL) -2.5 ± 2.6 -4.6 ± 13.4

nHDL (mg/dL) -13.5 ± 14.1 0.0 ± 19.5

Glucose (mg/dL) -7.5 ± 2.5* -1.7 ± 7.2

Table. 5: Mean change in total cholesterol, HDL, nHDL, and glucose from Fall 2017-Fall 2018 for all subjects with data at both time points (freshman, sophomore, junior).

Total Sample Freshman Sophomore Junior (n=10) (n=31) (n=18) (n=3) Total -6.0 ± 21.5 -4.8 ± 20.5 4.3 ± 36.1 -11.7 ± 19.5 Cholesterol (mg/dL) HDL (mg/dL) -4.3 ± 12.5 -4.4 ± 5.8 -1.7 ± 40.1 -4.9 ± 9.4

nHDL (mg/dL) -1.8 ± 19.2 -0.6 ± 20.8 5.7 ± 16.0 -6.7 ± 17.5

GLU (mg/dL) -2.5 ± 7.0 -1.4 ± 8.3 -5.7 ± 5.1 -3.4 ± 4.2

Table 6. Standardized Beta-Coefficients and p-values of regression analyses conducted investigating nine predictor variables for FM, LM, and VAT. Bolded text values are significant.

FM LM VAT β p-value β p-value β p-value Vigorous -0.43 .001 0.42 .001 -0.08 .576 Minutes Moderate -0.79 .480 0.08 .496 0.08 .550 Minutes Sitting -0.19 .865 -0.01 .924 -0.10 .441 Minutes PSS 0.16 .228 0.03 .825 0.04 .773

PSQI -0.42 .735 0.18 .150 0.04 .768

DASS -0.01 .931 0.06 .677 0.13 .376 (anxiety score) DASS -0.01 .961 -0.20 .128 -0.11 .461 (stress score) rMEQ -0.13 .283 0.26 .031 0.01 .973

AUDIT -0.25 .039 -0.19 .134 0.07 .634

FIGURES

Fall 2017 N=94 (Fr=43; Soph=9;94 Jr=21; Sr=21)

N=35 did not return; lost to follow-up, winter graduation, etc. Spring 2018 N=59 Fr=28; Soph=4; Jr=11; Sr=16

N=27 did not return; 16 graduating seniors

Fall 2018 N=85 (Fr=19; Soph=29; Jr=17; Sr=20)

4 non-SP18 49 new FA17 returners enrollments

Figure 1. Consort diagram detailing enrollment and retention of the sample at each data collection point and recruitment/dropout statistics. Freshman = Fr, Sophomore = Soph, Junior = Jr, Senior = Sr.

Figure 2: Distribution in the changes in body weight over the course of the 2017 to 2018 academic year.

Figure 3: Distribution in the changes in body weight over the course of the 2017 to 2018 academic year in males.

Figure 4: Distribution in the changes in body weight over the course of the 2017 to 2018 academic year in females.

REFERENCES

1. Obesity and Overweight (Factsheet) [http://www.who.int/mediacentre/factsheets/fs311/en/]

2. Vadeboncoeur, Claudia et al. “A meta-analysis of weight gain in first year university students: is freshman 15 a myth?” BMC obesity (2015).

3. Mokdad, Ali & K. Serdula, Mary & H. Dietz, William & A. Bowman, Barbara & S. Marks, James & P. Koplan, Jeffrey. (1999). The Spread of the Obesity Epidemic in the United States, 1991-1998. JAMA : the journal of the American Medical Association. 282. 1519-22. 10.1001/jama.282.16.1519.

4. U.S. Department of Education, National Center for Education Statistics. (2018). Digest of Education Statistics, 2016 (NCES 2017-094), Chapter 3.

5. Gropper, S. S., Simmons, K. P., Connell, L. J., & Ulrich, P. V. (2012). Weight and Body Composition Changes during the First Three Years of College. Journal of Obesity, 2012, 634048.

6. 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, 14(8), 1438–1443.

7. Racette, S. B., Deusinger, S. S., Strube, M. J., Highstein, G. R., & Deusinger, R. H. (2008). Weight Changes, Exercise, and Dietary Patterns During Freshman and Sophomore Years of College.

8. Hull, H. R., Morrow, M. L., Dinger, M. K., Han, J. L., & Fields, D. A. (2007). Characterization of body weight and composition changes during the sophomore year of college.

9. Oliveros, E., Somers, V. K., Sochor, O., Goel, K., & Lopez-Jimenez, F. (2014). The Concept of Normal Weight Obesity. Progress in Cardiovascular Diseases, 56, 426–433.

10. Romero-Corral, A., Somers, V. K., Sierra-Johnson, J., Korenfeld, Y., Boarin, S., Korinek, J., … Lopez-Jimenez, F. (2010). Normal weight obesity: a risk factor for cardiometabolic dysregulation and cardiovascular mortality. European Heart Journal, 31(6), 737–46.

11. Marques-Vidal, P., Pécoud, A., Hayoz, D., Paccaud, F., Mooser, V., Waeber, G., & Vollenweider, P. (2008). Prevalence of normal weight obesity in Switzerland: effect of various definitions. European Journal of Nutrition, 47(5), 251–257. 12. Franco, L. P., Morais, C. C., & Cominetti, C. (2016). Normal-weight obesity syndrome: diagnosis, prevalence, and clinical implications. Nutrition Reviews, 74(9), 558–570.

13. McAnarney ER. Textbook of Adolescent Medicine. Saunders; 1992.

14. American College Health Association. American College Health Association-National College Health Assessment II: Reference Group Undergraduate Executive Summary Spring 2017. Hanover, MD: American College Health Association; 2017.

15. Marques-Vidal, P., Pécoud, A., Hayoz, D., Paccaud, F., Mooser, V., Waeber, G., & Vollenweider, P. (2010). Normal weight obesity: Relationship with lipids, glycaemic status, liver enzymes and inflammation. Nutrition, Metabolism and Cardiovascular Diseases, 20, 669–675.

16. Anderson, D. A., Shapiro, J. R., & Lundgren, J. D. (n.d.). The freshman year of college as a critical period for weight gain: An initial evaluation.

17. Cluskey, M., & Grobe, D. (2009). College Weight Gain and Behavior Transitions: Male and Female Differences. YJADA, 109, 325–329. https://doi.org/10.1016/j.jada.2008.10.045

18. De Vos, P., Hanck, C., Neisingh, M., Prak, D., Groen, H., & Faas, M. M. (2015). Weight gain in freshman college students and perceived health. Preventive Medicine Reports, 2, 229–34.

19. Baumgartner RN, Heymsfield SB, Lichtman S..Body composition in elderly people: effect of criterion estimates on predictive equations..AM J Clin Nutr 53: 1345-1353, 1991.

20. McLeod, M., Breen, L., Hamilton, D. L., & Philp, A. (2016). Live strong and prosper: the importance of skeletal muscle strength for healthy ageing. Biogerontology, 17(3), 497– 510.

21. Vella-Zarb, R. A., & Elgar, F. J. (n.d.). The “Freshman 5”: A Meta-Analysis of Weight Gain in the Freshman Year of College. JOURNAL OF AMERICAN COLLEGE HEALTH, 58(2). Retrieved from

22. Crombie, A. P., Ilich, J. Z., Dutton, G. R., Panton, L. B., & Abood, D. A. (n.d.). The freshman weight gain phenomenon revisited.

23. Pouliot, M.-C., Despres, J.-P., Nadeau, A., Moorjani, S., Prud ’homme, D., Lupien, P. J., … Bouchard, C. (n.d.). Visceral Obesity in Men Associations With Glucose Tolerance, Plasma Insulin, and Lipoprotein Levels. Retrieved from

24. Wajchenberg, B. L. (2000). Subcutaneous and Visceral Adipose Tissue: Their Relation to the Metabolic Syndrome. Endocrine Reviews, 21(6), 697–738. https://doi.org/10.1210/edrv.21.6.0415

25. Despres, J.-P., Nadeau, A., Tremblay, A., Ferland, M., Moorjani, S., Lupien, P. J., … Bouchard, C. (n.d.). Role of Deep Abdominal Fat in the Association Between Regional Adipose Tissue Distribution and Glucose Tolerance in Obese Women. Retrieved from

26. Park, K., Rhee, B., Lee, K., Kim, S., Lee, H., Koh, C., & Min, H. (n.d.). Intra-abdominal Fat Is Associated With Decreased Insulin Sensitivity in Healthy Young Men. Retrieved from

27. Marin, P., Andersson, B., Ottosson, M., Olbe, L., Chowdhury, B., Kvist, H., … Bjijrntorp, P. (n.d.). The Morphology and Metabolism of Intraabdominal Adipose Tissue in Men.

28. Howard, B. V. (n.d.). Lipoprotein metabolism in diabetes mellitus. Retrieved from https://pdfs.semanticscholar.org/86aa/ac117ef4745223ec03f3eb2fcb86de285ebd.pdf

29. Anderson, K.C., Hirsch, K.R., Peterjohn, A.M., Blue, M.N.M., Pihoker, A.A., Ward, D.S., Ondrak, K.S., Smith-Ryan, A.E., The Characterization of Normal Weight Obesity in College Students. In Review, May 2018.

30. Laakso, M., Sarlund, H., & Mykkanen, L. (n.d.). Insulin Resistance Is Associated with Lipid and Lipoprotein Abnormalities in Subjects with Varying Degrees of Glucose Tolerance. Retrieved from

31. Abate, N., Garg, A., Peshock, R. M., Stray-Gundersen, J., & Grundy, S. M. (n.d.). Relationships of Generalized and Regional Adiposity to Insulin Sensitivity in Men. Retrieved from

32. Goodpaster, B. H., Park, S. W., Harris, T. B., Kritchevsky, S. B., Nevitt, M., Schwartz, A. V, … Newman, A. B. (n.d.). The Loss of Skeletal Muscle Strength, Mass, and Quality in Older Adults: The Health, Aging and Body Composition Study.

33. Robert R Wolfe; The underappreciated role of muscle in health and disease, The American Journal of Clinical Nutrition, Volume 84, Issue 3, 1 December 2006, Pages 475–482,

34. Falk Petersen, K., Dufour, S., Savage, D. B., Bilz, S., Solomon, G., Yonemitsu, S., … Shulman, G. I. (2007). The role of skeletal muscle insulin resistance in the pathogenesis of the metabolic syndrome. Retrieved from

35. Pope, L., Hansen, D., & Harvey, J. (2017). Examining the Weight Trajectory of College Students. J Nutr Educ Behav, 49, 137–141. https://doi.org/10.1016/j.jneb.2016.10.014

36. Levitsky, D., Halbmaier, C., & Mrdjenovic, G. (2004). PAPER The freshman weight gain: a model for the study of the epidemic of obesity. International Journal of Obesity, 28, 1435–1442.

37. Fuller, N. J., Jebb, S. A., Laskey ’, M. A., Coward, W. A., & Ella, M. (1992). Four- component model for the assessment of body composition in humans: comparison with

alternative methods, and evaluation of the density and hydration of fat-free mass. Clinical Science, 82, 687–693.

38. de Vos, P., Hanck, C., Neisingh, M., Prak, D., Groen, H., & Faas, M. M. (2015). Weight gain in freshman college students and perceived health. Preventive Medicine Reports, 2, 229–234.

39. Graham, M. A., & Jones, A. L. (n.d.). Freshman 15: Valid Theory or Harmful Myth? JOURNAL OF AMERICAN COLLEGE HEALTH, 50(4).

40. Lowry, R., Galuska, D. A., Fulton, J. E., Wechsler, H., Kann, L., & Collins, J. L. (2000). Physical activity, food choice, and weight management goals and practices among U.S. college students. American Journal of Preventive Medicine, 18(1), 18–27.

41. Bergen-Cico, Dessa. Patterns of Substance Abuse and Attrition Among First-Year Students. Journal of the First- Year Experience & Students in Transition, 2000, 12 (No. 1) pp. 61-75., 2000, 14 (NO. 3) pp. 61-75.

42. Levitsky, D. A., Garay, J., Nausbaum, M., Neighbors, L., & DellaValle, D. M. (2006). Monitoring weight daily blocks the freshman weight gain: a model for combating the epidemic of obesity. International Journal of Obesity, 30(6), 1003–1010. https://doi.org/10.1038/sj.ijo.0803221

43. Gross LS, Li L, Ford ES, Liu S. Increased consumption of refined carbohydrates and the epidemic of type 2 diabetes in the United States: an ecologic assessment. Am J Clin Nutr 2004; 79 (5): 774–779.

44. Schmitz KH, Jacobs Jr DR, Leon AS, Schreiner PJ, Sternfeld B. Physical activity and body weight: associations over ten years in the CARDIA study. Coronary Artery Risk Development in Young Adults. Int J Obes Relat Metab Disord 2000; 24 (11): 1475–1487.

45. National Center for Education Statistics. Digest of education statistics, 1996. Washington, DC: US Dept of Education, Office of Educational Research and Improvement, 1996. National Center for Education Statistics pub no 96–133.

46. Bureau of the Census. Statistical abstract of the United States, 1996 (116th ed). Washington, DC: US Dept of Commerce, 1996.

47. Kominski R, Adams A. Educational attainment in the United States: March 1993 and 1992. Current Population Reports P20–476.Washington, DC: US Bureau of the Census, 1994.

48. Dinger M. Physical activity and dietary intake among college students. Am J Health Stud. 1999;15:139–148.

49. Boone JE, Gordon-Larsen P, Adair LS, Popkin BM. Screen time and physical activity during adolescence: longitudinal effects on obesity in young adulthood. Int J Behav Nutr Phys Act. 2007;4:26.

50. De Cocker KA, van Uffelen JGZ, Brown WJ. Associations between sitting time and weight in young adult Australian women. Prev Med. 2010;51:361–7

51. Craig CL, Marshall AL, Sjostrom M, Bauman AE, Booth ML, Ainsworth BE, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35(8):1381–95.

52. Ekelund U, Sepp H, Brage S, Becker W, Jakes R, Hennings M, et al. Criterion-related validity of the last 7-day, short form of the International Physical Activity Questionnaire in Swedish adults. Public Health Nutr. 2006;9(2):258–65.

53. Wolin KY, Heil DP, Askew S, Matthews CE, Bennett GG. Validation of the International Physical Activity Questionnaire-Short among Blacks. J Phys Act Health. 2008;5(5):746– 60.

54. Papathanasiou G, Georgoudis G, Papandreou M, Spyropoulos P, Georgakopoulos D, Kalfakakou V, et al. Reliability measures of the short International Physical Activity Questionnaire (IPAQ) in Greek young adults. Hellenic J Cardiol. 2009;50(4):283–94.

55. Lee PH, Yu YY, McDowell I, Leung GM, Lam TH, Stewart SM. Performance of the international physical activity questionnaire (short form) in subgroups of the Hong Kong chinese population. Int J Behav Nutr Phys Act. 2011;8:81.

56. Macfarlane DJ, Lee CC, Ho EY, Chan KL, Chan DT. Reliability and validity of the Chinese version of IPAQ (short, last 7 days). J Sci Med Sport. 2007;10(1):45–51.

57. Macfarlane DJ, Lee CC, Ho EY, Chan KL, Chan D. Convergent validity of six methods to assess physical activity in daily life. J Appl Physiol. 2006;101(5):1328–34.

58. Kolbe-Alexander TL, Lambert EV, Harkins JB, Ekelund U. Comparison of two methods of measuring physical activity in South African older adults. J Aging Phys Act. 2006;14(1):98– 114.

59. Dinger MK, Beherns TK, Han JL. Validity and Reliability of the International Physical Activity Questionnaire in college students. Am J Health Ed. 2006;37:337–43.

60. Timperio A, Salmon J, Rosenberg M, Bull FC. Do logbooks influence recall of physical activity in validation studies? Med Sci Sports Exerc. 2004;36(7):1181–6.

61. Alomari MA, Keewan EF, Qhatan R, et al. Blood pressure and circulatory relationships with physical activity level in young normotensive individuals: IPAQ validity and reliability considerations. Clin Exp Hypertens. 2011;33(5):345-353.

62. Yang, Y. J., Kim, M. K., Hwang, S. H., Ahn, Y., Shim, J. E., & Kim, D. H. (2010). Relative validities of 3-day food records and the food frequency questionnaire. Nutrition Research and PracticeNutr Res Pract), 4(2), 142–148. https://doi.org/10.4162/nrp.2010.4.2.142

63. Marniemi J, Kronholm E, Aunola S, et al. Visceral fat and psychosocial stress in identical twins discordant for obesity. J Intern Med. 2002;251(1):35-43.

64. Cohen, S. and Williamson, G. Perceived Stress in a Probability Sample of the United States. Spacapan, S. and Oskamp, S. (Eds.) The Social Psychology of Health. Newbury Park, CA: Sage, 1988.

65. Rosenberg, D. E., Norman, G. J., Wagner, N., Patrick, K., Calfas, K. J., & Sallis, J. F. (2010). Reliability and Validity of the Sedentary Behavior Questionnaire (SBQ) for Adults. Journal of Physical Activity and Health, 7, 697–705. Retrieved from https://pdfs.semanticscholar.org/ba46/83487c3898953e237ea7d4b9112844f69293.pdf

66. Roane, B. M., Seifer, R., Sharkey, K. M., Van Reen, E., Bond, T. L. Y., Raffray, T., & Carskadon, M. A. (2015). What Role Does Sleep Play in Weight Gain in the First Semester of University? Behavioral Sleep Medicine, 13(6), 491–505. https://doi.org/10.1080/15402002.2014.940109

67. Culnan, E., Kloss, J. D., & Grandner, M. (2013). A prospective study of weight gain associated with chronotype among college freshmen. Chronobiology International, 30(5), 682–90. https://doi.org/10.3109/07420528.2013.782311

68. Cappuccio FP, D'Elia L, Strazzullo P, Miller MA. Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis. Diabetes Care. 2010; 33(2):414–420. doi: 10.2337/ dc09-1124. [PubMed: 19910503]

69. Chaput JP, Sjodin AM, Astrup A, Despres JP, Bouchard C, Tremblay A. Risk factors for adult overweight and obesity: the importance of looking beyond the 'big two'. Obes Facts. 2010; 3(5): 320–327. doi: 10.1159/000321398. [PubMed: 20975298]

70. Hart CN, Larose JG, Fava JL, James BL, Wing RR. The Association between Time in Bed and Obesity Risk in Young Adults. Behav Sleep Med. 2012 doi: 10.1080/15402002.2012.700289.

71. Garaulet M, Ortega FB, Ruiz JR, Rey-Lopez JP, Beghin L, Manios Y, Moreno LA. Short sleep duration is associated with increased obesity markers in European adolescents: effect of physical activity and dietary habits. The HELENA study. Int J Obes (Lond). 2011; 35(10):1308–1317. doi: 10.1038/ijo.2011.149. [PubMed: 21792170]

72. Koren D, Levitt Katz LE, Brar PC, Gallagher PR, Berkowitz RI, Brooks LJ. Sleep architecture and glucose and insulin homeostasis in obese adolescents. Diabetes Care. 2011; 34(11):2442–2447. doi: 10.2337/dc11-1093. [PubMed: 21933909]

73. Beebe DW. Cognitive, behavioral, and functional consequences of inadequate sleep in children and adolescents. Pediatr Clin North Am. 2011; 58(3):649–665. doi: 10.1016/j.pcl.2011.03.002. [PubMed: 21600347]

74. Carskadon MA, Acebo C, Jenni OG. Regulation of adolescent sleep: implications for behavior. Ann N Y Acad Sci. 2004; 1021:276–291. doi: 10.1196/annals.1308.032. [PubMed: 15251897]

75. Roberts RE, Roberts CR, Xing Y. Restricted sleep among adolescents: prevalence, incidence, persistence, and associated factors. Behav Sleep Med. 2011; 9(1):18–30. doi: 10.1080/15402002.2011.533991. [PubMed: 21218291]

76. Spiegel K, Tasali E, Penev P, Cauter EV. Brief communication: sleep curtailment in healthy young men is associated with decreased leptin levels, elevated ghrelin levels, and increased hunger and appetite. Ann Intern Med. 2004; 141:846–850. [PubMed: 15583226]

77. Taheri S, Lin L, Austin D, et al. Short sleep duration is associated with reduced leptin, elevated ghrelin, and increased body mass index. PLoS Med. 2004; 1:210–217.

78. Block, J. P., He, Y., Zaslavsky, A. M., Ding, L., & Ayanian, J. Z. (2009). Original Contribution Psychosocial Stress and Change in Weight Among US Adults. American Journal of Epidemiology a The Author, 170(2). https://doi.org/10.1093/aje/kwp104

79. Harding, J. L., Backholer, K., Williams, E. D., Peeters, A., Cameron, A. J., Hare, M. J., … Magliano, D. J. (2014). Psychosocial stress is positively associated with body mass index gain over 5 years: Evidence from the longitudinal AusDiab study. Obesity, 22(1), 277–286. https://doi.org/10.1002/oby.20423

80. Abouserie, R. (1994). Sources and levels of stress in relation to locus of control and self- esteem in university students. Educational Psychology, 14(3), 323–330.

81. Pierceall, E. A., & Keim, M. C. (2007). Stress and coping strategies among community college students. Community College Journal of Research and Practice, 31(9), 703–712.

82. Martinez, Y. T. S., Harmon, B. E., Nigg, C. R., Bantum, E. O., & Strayhorn, S. (2016). Diet and Physical Activity Intervention Strategies for College Students. Health Behavior and Policy Review, 3(4), 336–347. https://doi.org/10.14485/HBPR.3.4.5

83. Nelson MC, Story M. Food environments in university dorms: 20,000 calories per dorm room and counting. Am J Prev Med. 2009;36(6):523–526. [PubMed]

84. Larson N, Neumark-Sztainer D, Laska MN, Story M. Young adults and eating away from home: associations with dietary intake patterns and weight status differ by choice of restaurant. J Am Diet Assoc. 2011;111(11):1696–1703.

85. Nelson MC, Kocos R, Lytle LA, Perry CL. Understanding the perceived determinants of weight-related behaviors in late adolescence: a qualitative analysis among college youth. J Nutr Educ Behav. 2009;41(4):287–292.

86. Larson N, Neumark-Sztainer D, Story M, et al. Identifying correlates of young adults' weight behavior: survey development. Am J Health Behav. 2011;35(6):712–725.

87. Brevard PB, Ricketts CD. Residence of college students affects dietary intake, physical activity, and serum lipid levels. J Am Diet ASSOC. 1996;96( 1):35-38.

88. American College Health Association American College Health Association National College Health Assessment (ACHA-NCHA): Spring 2005 reference group data report (Abridged) J Am Coll Health. 2006;55(1):5–16.

89. Kim JY, Han SH, Yang BM. Implication of high-body-fat percentage on cardiometabolic risk in middle-aged, healthy, normal-weight adults. Obesity. 2013;21:1571–1577.

90. De Lorenzo A, Martinoli R, Vaia F, et al. Normal weight obese (NWO) women: an evaluation of a candidate new syndrome. Nutr Metab Cardiovasc Dis. 2006;16:513–523

91. Pedrelli, P., Nyer, M., Yeung, A., Zulauf, C., & Wilens, T. (2015). College Students: Mental Health Problems and Treatment Considerations. Academic Psychiatry : The Journal of the American Association of Directors of Psychiatric Residency Training and the Association for Academic Psychiatry, 39(5), 503–11.

92. Bridge JA, Goldstein TR, Brent DA. Adolescent suicide and suicidal behavior. J Child Psychol Psychiatry. 2006; 47(3–4):372–94. [PubMed: 16492264]

93. Hudson JI, et al. The prevalence and correlates of eating disorders in the National Comorbidity Survey Replication. Biol Psychiatry. 2007; 61(3):348–58. [PubMed: 16815322]

94. Eisenberg D, et al. Eating disorder symptoms among college students: Prevalence, persistence, correlates, and treatment-seeking. J Am Coll Health. 2011; 59(8):700–7. [PubMed: 21950250]

95. Covassin, N., & Singh, P. (2016). Sleep Duration and Cardiovascular Disease Risk: Epidemiologic and Experimental Evidence. Sleep Medicine Clinics, 11(1), 81–9.

96. Cappuccio, F. P., D’Elia, L., Strazzullo, P., & Miller, M. A. (2010). Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies. Sleep, 33(5), 585–92.

97. Linderman, J. K., Board, R., Astorino, T., Baker, J., Boone, T., Birnbaum, L., … Ferrara, C. M. (2009). THE COLLEGE EXPERIENCE: PHYSICAL ACTIVITY, NUTRITION, AND IMPLICATIONS FOR INTERVENTION AND FUTURE RESEARCH, 12. Retrieved from https://www.asep.org/asep/asep/Ferrara12_1_23-35.pdf

98. Kasparek DG, Corwin SJ, Valois RF, Sargent RG, Morris RL. Selected health behaviors that influence college freshman weight change. J Am Coll Health. 2008;56:437– 444.

99. Butler SM, Black DR, Blue CL, Gretebeck RJ. Change in diet, physical activity, and body weight in female college freshman. Am J Health Behav. 2004;28:24–32.

100. Nelson MC, Lust K, Story M, Ehlinger E. Alcohol use, eating patterns, and weight behaviors in a university population. Am J Health Behav. 2009; 33:227–237. [PubMed: 19063644]

101. Block, J.P., He, Y., Zaslavsky, A.M., Ding, L., & Ayanian, J.Z. (2009). Psychosocial stress and change in weight among US adults. American Journal of Epidemiology, 170(2), 181-92.

102. Olafsdottir, A.S., Torfadottir, J.E., & Arngrimsson, S.A. (2016). Health Behavior and Metabolic Risk Factors Associated with Normal Weight Obesity in Adolescents. PLOS ONE, 11(8), e0161451.

103. Wang, Z., Pi-Sunyer, F.X., Kotler, D.P., Wielopolski, L., Withers, R. T., Pierson, R. N., & Heymsfield, S.B. (2002). Multicomponent methods: evaluation of new and traditional soft tissue mineral models by in vivo neutron activation analysis. The American journal of clinical nutrition , 76(5), 968-974.

104. Mihalopoulos, N L, Auinger, P and Klein, J D. 2008. The Freshman 15: is it real?. J AM Coll Health, 56:531-533

105. Hull, H.R., Hester, C. N., & Fields, D. A. (2006). The effects of the holiday season on body weight and composition in college students. Nutrition & metabolism, 3, 44. Doi: 10:1186/1743-7075-3-44

106. Arts, Jennifer & Fernandez, Maria & Lofgren, Ingrid. (2014). Coronary Heart Disease Risk Factors in College Students. Advances in nutrition (Bethesda, Md.). 5. 177-87. 10.3945/an.113.005447.

107. Hajhosseini L, Holmes T, Mohamadi P, Goudarzi V, McProud L, Hollenbeck CB. Changes in body weight, body composition and resting metabolic rate (RMR) in first- year university freshmen students. J Am Coll Nutr. 2006;25:123–7.

108. Smith-Ryan, A.E., et al., Validity and reliability of a 4-compartment body composition model using dual energy x-ray absorptiometry-derived body volume. Clin Nutr, 2016.