PHYSICAL ACTIVITY, DAILY CORTISOL PATTERNS AND THE METABOLIC
SYNDROME IN OBESE ADOLESCENTS
By
Emily Hill Guseman

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Kinesiology
2012

ABSTRACT
PHYSICAL ACTIVITY, DAILY CORTISOL PATTERNS AND THE METABOLIC
SYNDROME IN OBESE ADOLESCENTS
By
Emily Hill Guseman
The influence of cortisol and stress in the development of obesity and metabolic syndrome has
received increased interest in recent years due to the high prevalence of overweight and obesity
in children, youth, and adults. Some studies suggest that physical activity may attenuate the
relationship between stress and metabolic syndrome, but more studies are necessary to fully
understand this relationship. Many studies have shown a relationship between stress and
metabolic syndrome using survey measures, but researchers have not linked survey measures to
HPA axis activity. Therefore, the purpose of this study was to examine the relationship between
stress (using daily cortisol and survey measures) and metabolic syndrome in obese youth, and to
examine the potential moderating effects of physical activity on that relationship. METHODS:
Obese adolescents were recruited from the Healthy Weight Center (HWC) and Academic
General Pediatrics (AGP) clinics at Helen DeVos Children’s Hospital in Grand Rapids, MI.
Height, weight, waist circumference, blood pressure, and pubertal stage were obtained at the
time of a routine clinic visit, and BMI was calculated. Participants completed the Pediatric
Symptom Checklist (PSC) to evaluate psychosocial function (stress). Physical activity
(steps/day and minutes per day spent in moderate-to-vigorous physical activity – MVPA) was
monitored using the SenseWear Pro III armband (SWA). Participants provided saliva samples
(while at home on a weekend day) that were taken at prescribed times: 1) immediately upon

waking; 2) 30-minutes after waking; 3) 3 hours after waking; 4) 6 hours after waking; 5) 9 hours
after waking; 6) 12 hours after waking. Fasting measures of HDL cholesterol, triglycerides, and
glucose were obtained from physician-ordered blood draws and used with waist circumference
and systolic blood pressure to calculate a continuous metabolic syndrome risk score (cMetS).
RESULTS: A total of 50 subjects (15 boys, 35 girls; mean age 14.8 ± 1.9 y) agreed to
participate. The mean cMetS score was 4.16 ± 4.30 and did not differ by clinic or sex. Subjects
participated in approximately 46 min of moderate-to-vigorous physical activity per day. Mean
cAUC was 1.337 ± 0.867 µg/dl (1.180 ± 0.753 µg/dl and 1.408 ± 0.922 µg/dl for males and
females, respectively) and did not differ by sex or clinic. No significant relationship was found
2

between cAUC and cMetS (R =0.113, p=0.66). Additionally, neither the interactions of
steps/day nor MVPA with cAUC significantly predicted cMetS. Partial correlations revealed a
significant inverse relationship between PSC total score and cAUC (r=-0.45, p=0.04). Significant
inverse relationships were also found between cAUC and PSC total score, internalizing score,
and externalizing score, as well as between all three PSC scores and cortisol sample 2. Finally,
no significant relationship was found between PSC score and cMetS, nor was there a significant
moderating effect of steps/day or MVPA on this relationship. DISCUSSION: The results of this
study did not support a relationship between either cortisol or PSC scores and MetS in this
sample, nor was there evidence of a moderating effect of physical activity. However, more
research is necessary, including survey measures that assess specific domains of stress and
longitudinal data to identify the timing of altered cortisol release, metabolic effects, and exposure
and reactions to stressors

Copyright by
EMILY HILL GUSEMAN
2012

ACKNOWLEDGEMENTS

Many thanks to everyone who helped with this dissertation. First I’d like to extend special
thanks to my committee for all of their guidance and support. To my friends and family, thank
you for listening, distracting me, and always being there. To the staff at the Healthy Weight
Center and Academic General Pediatrics, thanks for letting me hang around the clinic for
countless hours trying to recruit, and for always asking how many subjects were left. Very
special thanks to Eric Gurzell, Jason Wiesinger, and Elizabeth Hunt for helping me run the
cortisol assays and keeping me company those long days, and to Dr. Jennifer Fenton for letting
me borrow her lab space and staff. Lastly, I’d like to acknowledge all of the other graduate
students in Kinesiology for all of their friendship, advice, and support these last four years.

v

TABLE OF CONTENTS

LIST OF TABLES ...................................................................................................................... vii
LIST OF FIGURES ..................................................................................................................... ix
CHAPTER 1: INTRODUCTION................................................................................................ 1
BACKGROUND AND SIGNIFICANCE ............................................................................................... 1
SPECIFIC AIMS AND HYPOTHESES ................................................................................................ 6
SIGNIFICANCE OF THE DISSERTATION .......................................................................................... 7
REFERENCES ........................................................................................................................... 9
CHAPTER 2: REVIEW OF LITERATURE ........................................................................... 16
INTRODUCTION .......................................................................................................................... 16
CHILD OBESITY: PREVALENCE AND CONSEQUENCES................................................................. 16
WHAT IS METABOLIC SYNDROME? ............................................................................................ 18
PHYSICAL ACTIVITY AND METABOLIC SYNDROME/CARDIOVASCULAR DISEASE IN CHILDREN . 23
INFLUENCE OF DIET ................................................................................................................... 25
ASSESSMENT OF DIETARY PATTERNS ........................................................................................ 31
STRESS ....................................................................................................................................... 35
ASSESSMENT OF PSYCHOSOCIAL STRESS ................................................................................... 36
PSYCHOSOCIAL FUNCTION ......................................................................................................... 39
BIOMARKERS OF THE STRESS RESPONSE .................................................................................... 41
MATURITY-RELATED VARIATION IN HPA AXIS ACTIVITY ........................................................ 45
SEX-RELATED DIFFERENCES IN HPA AXIS ACTIVITY ............................................................... 47
CORTISOL AND ADIPOSITY ......................................................................................................... 53
PSYCHOSOCIAL FUNCTION AND CORTISOL ................................................................................ 56
DIETARY PATTERNS AND HYPERCORTISOLEMIA ........................................................................ 58
INFLUENCE OF PHYSICAL ACTIVITY ON STRESS AND CORTISOL ................................................ 60
PHYSICAL ACTIVITY, STRESS, AND METABOLIC SYNDROME ..................................................... 62
SUMMARY AND CONCLUSIONS .................................................................................................. 63
REFERENCES ......................................................................................................................... 65
CHAPTER 3: METHODOLOGY ............................................................................................ 77
SPECIFIC AIMS AND HYPOTHESES .............................................................................................. 77
METHODS .................................................................................................................................. 78
Study design and subjects. .................................................................................................... 78
Assessment of the main outcome – metabolic syndrome. ..................................................... 78
Daily Cortisol Measures. ...................................................................................................... 79

vi

Psychosocial Functioning ..................................................................................................... 83
Pubertal Stage....................................................................................................................... 83
REFERENCES ......................................................................................................................... 88
CHAPTER 4: RESULTS ........................................................................................................... 91
SUBJECT CHARACTERISTICS ...................................................................................................... 91
METABOLIC SYNDROME ............................................................................................................ 92
HABITUAL PHYSICAL ACTIVITY................................................................................................. 97
CORTISOL .................................................................................................................................. 97
PEDIATRIC SYMPTOM CHECKLIST .............................................................................................. 98
THREE-DAY DIET RECORD ........................................................................................................ 99
SPECIFIC AIM 1: ......................................................................................................................... 99
SPECIFIC AIM 2: ....................................................................................................................... 101
SPECIFIC AIM 3: ....................................................................................................................... 101
SPECIFIC AIM 4: ....................................................................................................................... 102
SPECIFIC AIM 5: ....................................................................................................................... 103
CHAPTER 5: DISCUSSION ................................................................................................... 104
METABOLIC SYNDROME .......................................................................................................... 104
PHYSICAL ACTIVITY ................................................................................................................ 106
DAILY CORTISOL ..................................................................................................................... 109
PSYCHOSOCIAL FUNCTIONING ................................................................................................. 112
MAIN FINDINGS ....................................................................................................................... 114
Cortisol and Metabolic Syndrome ...................................................................................... 114
Aims 1 and 2: ...................................................................................................................... 116
Aim 3: .................................................................................................................................. 118
Aims 4 and 5: ...................................................................................................................... 120
STRENGTHS AND LIMITATIONS ................................................................................................ 121
FUTURE DIRECTIONS ............................................................................................................... 126
REFERENCES ....................................................................................................................... 128
APPENDICES ........................................................................................................................... 137
APPENDIX A ......................................................................................................................... 137
APPENDIX B: ........................................................................................................................ 145
APPENDIX C: ........................................................................................................................ 150
APPENDIX D: ........................................................................................................................ 158
APPENDIX E: ........................................................................................................................ 163
APPENDIX F:......................................................................................................................... 165
APPENDIX G: ........................................................................................................................ 167
APPENDIX H: ........................................................................................................................ 171
APPENDIX I: ......................................................................................................................... 173

vii

LIST OF TABLES
Table 2.1. Summarized adult definitions of the metabolic syndrome

……………………20

Table 2.2. International Diabetes Federation (IDF) consensus definition of the metabolic
syndrome for children and adolescents (age 6-16 yrs)
……………………………………21
Table 2.3. Macronutrient intakes of adolescents compared to dietary reference intakes
(DRIs) ……………………………………………………………………………………………26
Table 2.4. Intakes of selected vitamins and nutrients related to cardiovascular health
……27
Table 2.5. Advantages and disadvantages of several common dietary assessment methods
(51)(135)
……………………………………………………………………………………32
Table 4.1. Distribution of subjects who completed data collection for study variables

……94

Table 4.2. Descriptive characteristics of the sample. Values are mean (SD) unless otherwise
indicated
……………………………………………………………………………………95
Table 4.3. Percent of subjects in each stage of pubertal development

……………………95

Table 4.4. Features of the Metabolic Syndrome within the sample. Values are mean (SD)
unless otherwise indicated ……………………………………………………………………96
Table 4.5. Physical activity and total daily energy expenditure within the sample. Values are
mean (SD) unless otherwise noted ……………………………………………………………97
Table 4.6. Pediatric Symptom Checklist. Values are mean (SD) unless otherwise noted

……99

Table 4.7. Independent and combined associations of cAUC and steps/day with cMetS

…..100

Table 4.8. Independent association of PSC total score and combined PSC total score +
steps/day with cMetS …………………………………………………………………………..103
Table F.1. Instructions for collection of saliva samples

viii

…………………………………..165

LIST OF FIGURES
Figure 3.1. Example timeline for saliva sampling
Figure 4.1. Recruitment flowchart

……………………………………………82

……………………………………………………………93

Figure 4.2. Distribution of metabolic syndrome features

……………………………………96

Figure 4.3. Mean cortisol level at each sampling time

……………………………………98

Figure 4.4. Relationship between cAUC and cMetS …………………………………………..101
Figure A.1 Example timeline for collection of saliva samples

…………………………..139

Figure B.1 Example timeline for collection of saliva samples

…………………………..146

Figure C.1 Example timeline for collection of saliva samples

…………………………..152

Figure D.1 Example timeline for collection of saliva samples

…………………………..159

Figure F.1. Saliva sampling log

….……………………………………………………….165

Figure G.1 Three-day diet record day 1

…………………………………………………..167

Figure G.2 Three-day diet record day 2

…………………………………………………..168

Figure G.3 Three-day diet record day 3

…………………………………………………..169

ix

CHAPTER 1: INTRODUCTION

Background and Significance
The past several decades have seen a dramatic increase in the prevalence of overweight
and obesity among youth in the United States. Currently, an estimated 32% of U.S. youth 6-19
years of age are overweight or obese (1). The current epidemic of childhood obesity, along with
its immediate and long-term health consequences, makes this disorder one of the most important,
if not the most important, health issue facing youth today (2–6). Beyond obesity, excess intraabdominal adipose tissue, or central adiposity, is thought to be especially problematic, given its
strong links to the metabolic syndrome (MetS)(7). The MetS is defined as a clustering of
cardiovascular risk factors including dyslipidemia, hypertension, insulin resistance/glucose
intolerance, and abdominal obesity. Specific adult diagnostic criteria, shown in Table 2.1, have
been proposed by the International Diabetes Federation (IDF)(8), World Health Organization
(WHO) (9), National Cholesterol Education Program – Third Adult Treatment Panel (NCEPATP III)(10) and the European Group for the Study of Insulin Resistance (EGIR)(11). Several
adaptations of the adult criteria have been proposed for use with children and adolescents. These
definitions have been summarized and condensed by the IDF (12). One set of criteria, defined by
Cook and colleagues (13), involves the adaptation of the NCEP-ATP III (see Table 2.2) for a
pediatric population. The child and adolescent cut points proposed by the IDF classify children
by age (12), allowing for growth-related variation in certain variables such as waist
circumference. A comparison of these two classification systems is shown in Table 2.2.
Publication of age- and sex-specific percentiles based on NHANES III data have made it

1

possible to classify children and adolescents appropriately for each factor of MetS except glucose
tolerance/insulin resistance (14). Age- and sex-specific percentiles have not been computed for
glucose because glucose levels are known to stay fairly constant throughout childhood and
adolescence (14; 15). Thus, it is necessary to use different criteria for this factor. To date, the
cut-points and percentiles provided by Cook’s group have not been used together to classify
MetS.
It is difficult to express the prevalence of MetS among US children and adolescents with
great certainty, because of the numerous classification schemes available. Current estimates
indicate that 4-7% of children and adolescents in the US possess MetS, and that among obese
adolescents the prevalence of MetS is approximately 30% (13). However, MetS has been
reported to be present in as many as 49-53% of severely obese adolescents (16; 17). Similar to
prevalence rates for overweight and obesity, there are ethnic (18) and socioeconomic status
(SES) (19) differences in the prevalence of MetS. Among ethnic groups, the highest prevalence
is found in Mexican American youth (5.6%), followed by white (4.8%) and black (2.0%) youth.
Rates are also higher among youth living below the poverty level (5.7%) as compared to those
living above the poverty level (3.7%) (13). Boys tend to have a higher prevalence of MetS than
girls (6.1% vs. 2.1%, respectively). Prevalence of MetS appears to be highest during puberty
(7.2% during pubic hair stage 3) (13), which may correspond with the transient insulin resistance
that is seen during this time (20).
It is now well established that cardiovascular risk factors cluster in childhood (21), track
over time (22–26), and increase the accumulation of fatty streaks in the coronary arteries of
adolescents and young adults (27–29). In fact, risk factors that cluster tend to track more
strongly than each individual risk factor when examined independently (26). Pooled evidence
2

from the Bogalusa Heart Study and Cardiovascular Risk in Young Finns study indicates that
pediatric MetS is associated with increased risk of adult MetS. Furthermore, the presence of
MetS during adolescence is associated with Type II Diabetes Mellitus and high carotid intimamedial thickness in adulthood (27). A recent analysis of data pooled from large longitudinal
studies indicates risk factor measurements are most predictive of adult CV risk beginning at the
age of nine years (29). Data from the Bogalusa Heart study also suggest that not only do adverse
levels of CVD risk factors (clustered as MetS) track into adulthood, individuals with favorable
risk factor profiles in childhood are less likely to display MetS in adulthood (25). Thus, the early
prevention and treatment of pediatric MetS are important areas of study.
Following an adolescent growth spurt, fat is deposited at high rates, especially around the
abdomen (30; 31). This increased fat deposition combined with lifestyle changes including
decreases in habitual physical activity (32) and the dramatic changes in the hormonal
environment that occur during adolescence (33) make adolescence a particularly vulnerable
period in the development of obesity and metabolic syndrome. Recent studies have shown that
early adolescence (ages 11-13 y) maybe be the time when CVD risk is most strongly influenced
by body fatness and cardiorespiratory fitness (33). Additionally, adolescents who are at high risk
for CVD may reap the largest benefits from increased OA and cardiorespiratory fitness, further
suggesting that adolescence is an important time to intervene in an attempt to reduce the
development of MetS (34–36).
Physical activity and diet are typically considered the most important factors in the
prevention and treatment of both obesity and MetS (37), however these two factors explain only
a portion of the total variance in either trait (38). An often overlooked but equally important
factor is the dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis and altered cortisol
3

secretion in response to stressful situations (39–41). Alterations in the normal activity of the
HPA axis and disruption of the normal daily pattern of cortisol secretion can influence both
glucose and fat metabolism. Some resulting changes, such as the activation of lipoprotein lipase,
promote storage of lipids in visceral adipocytes (39) and thus, the accumulation of excess
adipose tissue. This notion is supported by research involving individuals with Cushing’s
disease, in whom chronically elevated cortisol levels are associated with greatly increased
visceral adipose tissue (39). Other studies in apparently healthy adults and children have
examined the relationships between cortisol and obesity (42–46) including three studies from our
group (47–49). To date, the results are equivocal, which may be partially explained by
methodological limitations. The largest potential limitation of previous studies is the use of a
single cortisol sample collected immediately upon waking or at another point in the morning. A
comparison of multiple samples taken over a 12-hour period and waking samples has shown that
the waking sample alone fails to predict mean cortisol levels over the course of the entire day
(50). Thus, a single waking salivary cortisol sample is not likely to be the most accurate way to
examine possible relationships between cortisol and obesity; in fact, the immediate waking
sample may actually be more indicative of HPA axis activity during the final stages of sleep
rather than the diurnal pattern (50).
Stress can also be examined using psychosocial measures that assess a number of
different domains and types of stressors. Another variable of interest is psychosocial functioning,
which is related to stress in the sense that optimal psychosocial functioning is necessary to
adequately cope with and tolerate stressors (51). Individuals with poor psychosocial functioning
may experience a number of day-to-day difficulties, including poor self esteem and a lack of
supportive peer relationships. Certain psychosocial domains, such as depression, anxiety,

4

attention difficulties, and behavioral problems are associated with alterations in HPA axis
activity (52–54). Altered cortisol responses to stress are often seen among children and
adolescents who are victims of bullying or experience behavior problems (55–57). Whether these
experiences are associated with higher or lower cortisol levels seems to vary based on the type of
stressor or psychosocial difficulty experienced, and may also vary when comparing acute cortisol
responses to stress and basal cortisol levels (55; 56; 58).
Several surveys exist to evaluate overall psychosocial function or specific domains such
as experience with bullying, lifetime or recent exposure to stressful events, depression, and
anxiety (59). The Pediatric Symptom Checklist is a survey often used in clinical practice to
evaluate overall psychosocial functioning and identify impaired functioning in healthy children
and adolescents seen in primary care, as well as individuals with chronic illnesses (60–62). When
using this checklist, a higher score is indicative of poorer psychosocial functioning, which may
be associated with behavior problems, depression, attention difficulties, or other psychological
difficulties, as well as altered cortisol patterns (52–54; 60).
Several investigators have described the daily cortisol pattern in children (46; 54; 63)
However, few have considered the daily cortisol pattern in conjunction with obesity. In adults,
elevated daily cortisol levels predict increased body mass index (BMI) (64), and those with MetS
(65) or increased abdominal fat (66) exhibit increased HPA axis activity as determined by serum,
urinary, and salivary measures of cortisol. The relationship between HPA axis activity and MetS
in adults suggests that understanding their relationship is important to better understand the
etiology of obesity and MetS in youth, as well. To our knowledge, no studies have used both
survey and biological measures of stress (cortisol) to evaluate the relationships between stress
and MetS in adolescents. Therefore, it is difficult to compare results from studies using surveys
5

to those using cortisol, and more research is necessary to understand the role of each type of
measure in evaluating these relationships.
As previously mentioned, physical activity is a cornerstone in the prevention and
treatment of child obesity and MetS (38). However, the role of physical activity in altering the
relationship between stress and MetS is largely unknown. Holmes et al. (47) previously showed
that physical activity attenuates the relationship between survey-measured anxiety and MetS in
boys, and another study found that post-exercise cortisol was 32% lower than pre-exercise levels
in obese children (67). Therefore, it is reasonable to hypothesize that habitual physical activity
would attenuate the relationship between stress (measured by surveys and/or cortisol levels) and
MetS in obese children and adolescents.
The overall purpose of this dissertation is to examine the relationships between
habitual physical activity, stress (via survey and daily cortisol levels), and MetS in obese
adolescents. The organization of this dissertation proposal is as follows: Chapter 1 –
Introduction, Chapter 2 – Review of Literature, and Chapter 3 – Methodology, Chapter 4 –
Results, and Chapter 5 – Discussion.
Specific Aims and Hypotheses

Specific aim 1: Determine the relationship between daily cortisol levels and the continuous
metabolic syndrome score (cMetS) in obese adolescents. It was hypothesized that cortisol area
under the curve (cAUC) will be positively related to cMetS.
Specific aim 2: Determine if habitual physical activity (MVPA and/or steps/day) attenuates the
relationship between daily cortisol and cMetS in obese adolescents. It was hypothesized that

6

there will be a significant interaction of PA and cAUC, such that PA will attenuate the
relationship between cAUC and cMetS.
Specific Aim #3: Determine the relationship between psychosocial functioning score and cAUC
in obese adolescents. It was hypothesized that psychosocial functioning will be positively related
to cAUC.
Specific Aim #4: Determine the relationship between psychosocial functioning score and cMetS
in obese adolescents. It was hypothesized that psychosocial functioning will be positively
related to cMetS.
Specific aim 5: Determine if habitual physical activity (MVPA and/or steps/day) attenuates the
relationship between psychosocial functioning score and cMetS in obese adolescents. It was
hypothesized that there will be a significant interaction of PA and psychosocial functioning, such
that PA will attenuate the relationship between psychosocial functioning and cMetS.
Significance of the Dissertation

Given the complex etiology of obesity and MetS, it is expected that the outcomes of this
study will improve current understanding of the relationship between cortisol and MetS in obese
adolescents. Additionally, we hope to determine the influence of regular habitual physical
activity on this relationship. Because obese adolescents are at increased risk of MetS, it is
important to identify and reduce factors that may additionally contribute to that risk. Excess
cortisol secretion is known to have detrimental effects on glucose tolerance, insulin sensitivity,
and blood pressure (39; 40). If the increased HPA axis activity is attenuated by physical activity,
even in obese adolescents, a positive effect may be observed on metabolic risk factors even if

7

weight loss does not result. Thus, results of this study may provide important insight into the
role of both physical activity and stress reduction programs for pediatric obesity.

8

REFERENCES

9

REFERENCES
1.

Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity and trends in body
mass index among US children and adolescents, 1999-2010. JAMA. 2012
Feb;307(5):483–90.

2.

Power C, Lake JK, Cole TJ. Measurement and long-term health risks of child and
adolescent fatness. Int J Obes Relat Metab Disord. 1997 Jul;21(7):507–26.

3.

Ogden CL, Carroll MD, Flegal KM. High body mass index for age among US children
and adolescents, 2003-2006. JAMA. 2008 May;299(20):2401–5.

4.

Katzmarzyk PT, Srinivasan SR, Chen W, Malina RM, Bouchard C, Berenson GS. Body
mass index, waist circumference, and clustering of cardiovascular disease risk factors in a
biracial sample of children and adolescents. Pediatrics. 2004 Aug;114(2):e198–205.

5.

Kaplan EL, Rocchini AR. Cardiovascular disease in children. Reflections on the last half
century and expectations for the future. Minn Med. 1991 Sep;74(9):26.

6.

Smoak CG, Burke GL, Webber LS, Harsha DW, Srinivasan SR, Berenson GS. Relation of
obesity to clustering of cardiovascular disease risk factors in children and young adults.
The Bogalusa Heart Study. Am J Epidemiol. 1987 Mar;125(3):364–72

7.

Syme C, Abrahamowicz M, Leonard GT, Perron M, Pitiot A, Qiu X, Richer L, Totman J,
Veillette S, Xiao Y, Gaudet D, Paus T, Pausova Z. Intra-abdominal adiposity and
individual components of the metabolic syndrome in adolescence: sex differences and
underlying mechanisms. Arch Pediatr Adolesc Med. 2008 May;162(5):453–61.

8.

Alberti KGMM, Zimmet P, Shaw J. Metabolic syndrome--a new world-wide definition. A
Consensus Statement from the International Diabetes Federation. Diabet Med. 2006
May;23(5):469–80.

9.

Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and
its complications. Part 1: Diagnosis and classification of diabetes mellitus provisional
report of a WHO consultation. Diabetic Med. 1998 Jul;15(7):539–53.

10.

Executive Summary of The Third Report of The National Cholesterol Education Program
(NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol
In Adults (Adult Treatment Panel III). JAMA. 2001 May;285(19):2486–97.

11.

Balkau B, Charles MA. Comment on the provisional report from the WHO consultation.
European Group for the Study of Insulin Resistance (EGIR). Diabetic Med. 1999
May;16(5):442–3.

10

12.

Zimmet P, Alberti KGM, Kaufman F, Tajima N, Silink M, Arslanian S, Wong G, Bennett
P, Shaw J, Caprio S. The metabolic syndrome in children and adolescents - an IDF
consensus report. Pediatr Diabetes. 2007 Oct;8(5):299–306.

13.

Cook S, Weitzman M, Auinger P, Nguyen M, Dietz WH. Prevalence of a metabolic
syndrome phenotype in adolescents: findings from the third National Health and Nutrition
Examination Survey, 1988-1994. Arch Pediatr Adolesc Med. 2003 Aug;157(8):821–7.

14.

Cook S, Auinger P, Huang TT-K. Growth curves for cardio-metabolic risk factors in
children and adolescents. J Pediatr. 2009 Sep;155(3):S6.e15–26.

15.

Jolliffe CJ, Janssen I. Development of age-specific adolescent metabolic syndrome criteria
that are linked to the Adult Treatment Panel III and International Diabetes Federation
criteria. J Am Coll Cardiol. 2007 Feb;49(8):891–8.

16.

Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW, Allen K, Lopes
M, Savoye M, Morrison J, Sherwin RS, Caprio S. Obesity and the metabolic syndrome in
children and adolescents. NEJM. 2004 Jun;350(23):2362–74.

17.

Guseman EH, Eisenmann JC, Stratbucker W. Does physical activity modify the
relationship between psychosocial stress and metabolic syndrome in obese youth? In
Preparation. 2012.

18.

Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalence of high body mass
index in US children and adolescents, 2007-2008. JAMA. 2010 Jan;303(3):242–9.

19.

Shrewsbury V, Wardle J. Socioeconomic status and adiposity in childhood: a systematic
review of cross-sectional studies 1990-2005. Obesity (Silver Spring). 2008
Feb;16(2):275–84.

20.

Bloch C a, Clemons P, Sperling M a. Puberty decreases insulin sensitivity. J Pediatr. 1987
Mar;110(3):481–7.

21.

Lauer RM, Connor WE, Leaverton PE, Reiter M a, Clarke WR. Coronary heart disease
risk factors in school children: the Muscatine study. J Pediatr. 1975 May;86(5):697–706.

22.

Andersen LB, Harro M, Sardinha LB, Froberg K, Ekelund U, Brage S, Anderssen SA.
Physical activity and clustered cardiovascular risk in children: a cross-sectional study (The
European Youth Heart Study). Lancet. 2006 Jul;368(9532):299–304.

23.

Camhi S, Katzmarzyk P. Tracking of cardiometabolic risk factor clustering from
childhood to adulthood Int J Pediatr Obes. 2010;(5):122–129.

24.

Chen W, Bao W, Begum S, Elkasabany A, Srinivasan SR, Berenson GS. Wei Chen,
Weihang Bao, Shahana Begum, Abdalla Elkasabany, Sathanur R. Srinivasan, and Gerald
S. Berenson. Age-related patterns of the clustering of cardiovascular risk variables of

11

syndrome x from childhood to young adulthood in a population made up of black and
white subjects. Blood Press. 2000;(May 1999):1042–1048.
25.

Chen W, Srinivasan S, Li S, Xu J, Berenson GS. Metabolic syndrome variables at low
levels in childhood are beneficially associated with adulthood cardiovascular risk: The
Bogalusa Heart Study. Diabetes Care. 2005 Jan;28(1):126-31.

26.

Bao W, Srinivasan SR, Wattigney WA, Berenson GS. Persistance of multiple
cardiovascular risk clustering related to Syndrome X from childhood to young adulthood:
the Bogalusa Heart Study. Arch Intern Med. 1994;154:1842–1847.

27.

Magnussen CG, Koskinen J, Chen W, Thomson R, Schmidt MD, Srinivasan SR, Kivimäki
M, Mattsson N, Kähönen M, Laitinen T, Taittonen L, Rönnemaa T, Viikari JSA, Berenson
GS, Juonala M, Raitakari OT. Pediatric metabolic syndrome predicts adulthood metabolic
syndrome, subclinical atherosclerosis, and type 2 diabetes mellitus but is no better than
body mass index alone: the Bogalusa Heart Study and the Cardiovascular Risk in Young
Finns Study. Circulation. 2010 Oct;122(16):1604–11.

28.

McMahan CA, McGill HC, Gidding SS, Malcom GT, Newman WP, Tracy RE, Strong JP.
PDAY risk score predicts advanced coronary artery atherosclerosis in middle-aged
persons as well as youth. Atherosclerosis. 2007 Feb;190(2):370–7.

29.

Juonala M, Magnussen CG, Venn A, Dwyer T, Burns TL, Davis PH, Chen W, Srinivasan
SR, Daniels SR, Kähönen M, Laitinen T, Taittonen L, Berenson GS, Viikari JSA,
Raitakari OT. Influence of age on associations between childhood risk factors and carotid
intima-media thickness in adulthood: the Cardiovascular Risk in Young Finns Study, the
Childhood Determinants of Adult Health Study, the Bogalusa Heart Study, and the
Muscatine Study. Circulation. 2010 Dec;122(24):2514–20.

30.

Malina RM, Koziel S, Bielicki T. Variation in Subcutaneous Adipose Tissue Distribution
Associated With Age, Sex, and Maturation. Am J Hum Biol.1999;200(May 1998):189200.

31.

Baumgartner R, Roche AF. Tracking of fat pattern indices in childhood: The Melbourne
Growth Study. Human Biol. 1988;60(4):549.

32.

Eaton DK, Kann L, Kinchen S, Shanklin S, Ross J, Hawkins J, Harris WA, Lowry R,
McManus T, Chyen D, Lim C, Whittle L, Brener ND, Wechsler H. Youth risk behavior
surveillance - United States, 2009. MMWR Surveill Summ. 2010 Jun;59(5):1–142.

33.

Ondrak KS, McMurray RG, Bangdiwala SI, Harrell JS. Influence of aerobic power and
percent body fat on cardiovascular disease risk in youth. J Adolesc Health. 2007
Aug;41(2):146–52.

12

34.

Katzmarzyk P, Malina R, Bouchard C. Physical Activity, Physical Fitness, and Coronary
Heart Disease Risk Factors in Youth: The Quebec Family Study Preventive Med.
1999;562:555–562.

35.

McMurray RG, Ainsworth BE, Harrell JS, Griggs TR, Williams OD. Is physical activity
or aerobic power more influential on reducing cardiovascular disease risk factors? Med
Sci Sports Exer. 1998 Oct;30(10):1521–9.

36.

Lobelo F, Pate RR, Dowda M, Liese AD, Daniels SR. Cardiorespiratory fitness and
clustered cardiovascular disease risk in U.S. adolescents. J Adolesc Health. 2010
Oct;47(4):352–9.

37.

Keith SW, Redden DT, Katzmarzyk PT, Boggiano MM, Hanlon EC, Benca RM, Ruden
D, Pietrobelli a, Barger JL, Fontaine KR, Wang C, Aronne LJ, Wright SM, Baskin M,
Dhurandhar NV, Lijoi MC, Grilo CM, DeLuca M, Westfall a O, Allison DB. Putative
contributors to the secular increase in obesity: exploring the roads less traveled. Int J Obes
(Lond). 2006 Nov;30(11):1585–94.

38.

Barlow SE. Expert committee recommendations regarding the prevention, assessment, and
treatment of child and adolescent overweight and obesity: summary report. Pediatrics.
2007 Dec;120 Suppl:S164–92.

39.

Björntorp P. Do stress reactions cause abdominal obesity and comorbidities? Obes Rev.
2001 May;2(2):73–86.

40.

Rosmond R, Dallman MF, Björntorp P. Stress-related cortisol secretion in men:
relationships with abdominal obesity and endocrine, metabolic and hemodynamic
abnormalities. J Clin Endocrinol Metab. 1998 Jun;83(6):1853–9.

41.

Ljung T, Holm G, Friberg P, Andersson B, Bengtsson BA, Svensson J, Dallman M,
McEwen B, Björntorp P. The activity of the hypothalamic-pituitary-adrenal axis and the
sympathetic nervous system in relation to waist/hip circumference ratio in men. Obes Res.
2000 Oct;8(7):487–95.

42.

Netherton C. Salivary cortisol and dehydroepiandrosterone in relation to puberty and
gender. Psychoneuroendocrinology 2004 Feb;29(2):125–140.

43.

Barat P, Gayard-Cros M, Andrew R, Corcuff J-B, Jouret B, Barthe N, Perez P, Germain
C, Tauber M, Walker BR, Mormede P, Duclos M. Truncal distribution of fat mass,
metabolic profile and hypothalamic-pituitary adrenal axis activity in prepubertal obese
children. J Pediatr. 2007 May;150(5):535–9, 539.el.

44.

Weigensberg MJ, Toledo-Corral CM, Goran MI. Association between the metabolic
syndrome and serum cortisol in overweight Latino youth. J Clin Endocrinol Metab. 2008
Apr;93(4):1372–8.

13

45.

Rosmalen JGM, Oldehinkel a J, Ormel J, de Winter a F, Buitelaar JK, Verhulst FC.
Determinants of salivary cortisol levels in 10-12 year old children; a population-based
study of individual differences. Psychoneuroendocrinology. 2005 Jun;30(5):483–95.

46.

Kiess W, Meidert A, Dressendörfer R a, Schriever K, Kessler U, König A, Schwarz HP,
Strasburger CJ. Salivary cortisol levels throughout childhood and adolescence: relation
with age, pubertal stage, and weight. Pediatr Res. 1995 Apr;37(4 Pt 1):502–6.

47.

Holmes M, Eisenmann J, Ekkekakis P, Gentile D. Physical activity, stress, and metabolic
risk score in 8-to 18-year-old boys. J Phys Act Health. 2008;(5):294–307.

48.

Hill EE, Eisenmann JC, Holmes ME, Heelan KA. Morning cortisol is not associated with
truncal fatness or resting blood pressure in children: cross-sectional and 1-2 year followup analyses. J Pediatr Endocrinol Metab. 2010 Oct;23(10):1031–7.

49.

Hill EE, Eisenmann JC, Gentile D, Holmes ME, Walsh D. The association between
morning cortisol and adiposity in children varies by weight status. J Pediatr Endocrinol
Metab. 2011 Jan;24(9-10):709–713.

50.

Edwards S, Clow a, Evans P, Hucklebridge F. Exploration of the awakening cortisol
response in relation to diurnal cortisol secretory activity. Life Sci. 2001 Mar;68(18):2093–
103.

51.

D’Amico D, Barbarito C. Health & Physical Assessment in Nursing. Prentice Hall; 2006.

52.

Schechter JC, Brennan P a, Cunningham PB, Foster SL, Whitmore E. Stress, cortisol, and
externalizing behavior in adolescent males: an examination in the context of
multisystemic therapy. J Abnorm Child Psychol. 2012 Aug;40(6):913–22.

53.

Shirtcliff EA, Granger DA, Booth A, Johnson D. Low salivary cortisol levels and
externalizing behavior problems in youth. Dev Psychopathol. 2005 Winter;17(1)167–184.

54.

Shirtcliff E a, Essex MJ. Concurrent and longitudinal associations of basal and diurnal
cortisol with mental health symptoms in early adolescence. Dev Psychobiol. 2008
Nov;50(7):690–703.

55.

Gustafsson PE, Anckarsa H, Lichtenstein P, Nelson N, Gustafsson PA. Does quantity
have a quality all its own ? Cumulative adversity and up- and down-regulation of circadian
salivary cortisol levels in healthy children. Psychoneuroendocrinology. 2010;35:1410–
1415.

56.

Ouellet-morin I, Odgers CL, Danese A, Bowes L, Shakoor S, Papadopoulos AS, Caspi A,
Moffitt TE, Arseneault L. Blunted cortisol responses to stress signal social and behavioral
problems among maltreated/bullied 12-year-old children. Biol Psychiatry.
2011;70(11):1016-1023.

14

57.

Janicke DM, Marciel KK, Ingerski LM, Novoa W, Lowry KW, Sallinen BJ, Silverstein
JH, David M, Lisa M, Janet H. Impact of Psychosocial Factors on Quality of Life in
Overweight Youth. Obesity (Silver Spring). 2007;15(7):1799–1807.

58.

Adam EK, Doane LD, Zinbarg RE, Mineka S, Craske MG, Griffith JW. Prospective
prediction of major depressive disorder from cortisol awakening responses in adolescence.
Psychoneuroendocrinology. 2010;35(6):921–931.

59.

De Vriendt T, Moreno L a, De Henauw S. Chronic stress and obesity in adolescents:
scientific evidence and methodological issues for epidemiological research. Nutr Metab
Cardiovasc Dis. 2009 Sep;19(7):511–9.

60.

Jellinek MS, Murphy JM, Little M, Pagano ME, Comer DM, Kelleher KJ. Use of the
Pediatric Symptom Checklist to Screen for Psychosocial Problems in Pediatric Primary
Care A National Feasibility Study. Arch Pediatr Adolesc Med. 1999 Nov;153:254–260.

61.

Gardner W, Kelleher KJ, Pajer K a. Multidimensional adaptive testing for mental health
problems in primary care. Med Care. 2002 Sep;40(9):812–23.

62.

Stoppelbein L, Greening L, Jordan SS, Elkin TD, Moll G, Pullen J. Factor analysis of the
Pediatric Symptom Checklist with a chronically ill pediatric population. J Dev Behav
Pediatr. 2005;26(5)

63.

Knutsson U, Dahlgren J, Marcus C, Rosberg S, Brönnegård M, Stierna P, AlbertssonWikland K. Circadian cortisol rhythms in healthy boys and girls: relationship with age,
growth, body composition, and pubertal development. J Clin Endocrinol Metab. 1997
Feb;82(2):536–40.

64.

Roberts C, Troop N, Connan F, Treasure J, Campbell IC. The effects of stress on body
weight: biological and psychological predictors of change in BMI. Obesity (Silver
Spring). 2007 Dec;15(12):3045–55.

65.

Ward AMV, Fall CHD, Stein CE, Kumaran K, Veena SR, Wood PJ, Syddall HE, Phillips
DIW. Cortisol and the metabolic syndrome in South Asians. Clin Endocrinol (Oxf). 2003
Apr;58(4):500–5.

66.

Pasquali R, Cantobelli S, Casimirri F, Capelli M, Bortoluzzi L, Flamia R, Labate AM,
Barbara L. The hypothalamic-pituitary-adrenal axis in obese women with different
patterns of body fat distribution. J Clin Endocrinol Metab. 1993 Aug;77(2):341–6.

67.

Hershberger a M, McCammon MR, Garry JP, Mahar MT, Hickner RC. Responses of
lipolysis and salivary cortisol to food intake and physical activity in lean and obese
children. J Clin Endocrinol Metab. 2004 Sep;89(9):4701–7.

15

CHAPTER 2: REVIEW OF LITERATURE
Introduction
This review of the literature covers several areas related to this dissertation including:
child obesity, metabolic syndrome, stress and the stress response, and the relationships between
physical activity, stress and the metabolic syndrome (MetS). There are several excellent review
papers in several of these areas [e.g., child obesity (1), physical activity and obesity (2), physical
activity and MetS (or CVD risk factors) in children (3), assessment of stress (4), and recently a
seminal review on physical activity, stress, and MetS (5)]; therefore, some sections will be
abbreviated compared to others.
Child Obesity: Prevalence and Consequences
Prevalence. The past several decades have seen a dramatic increase in the prevalence of
overweight and obesity among youth in the United States. Presently, an estimated 32% of U.S.
youth are overweight or obese (6). A recent report indicated that among a diverse group of
children and adolescents, 37.1% were overweight, 19.4% were obese, and 6.4% were extremely
obese (BMI ≥ 1.2*95th percentile) (7).
Compared to their normal weight peers, overweight and obese youth are more likely to be
overweight or obese as adults. Relative risks of adult obesity reported in the literature range
from 2.0-10.0 for overweight children and increase into adolescence [RR 12.0 (male) to 15.0
(female)] (8). While most obese adults were not overweight or obese as youth, as many as 2258% of overweight adolescents become obese adults and 24-90% of obese adolescents become
obese adults (1; 8). Those youth who are obese during late adolescence seem to be especially
susceptible to adult obesity (1). Fat deposition, especially on the trunk, occurs at high rates

16

following the adolescent growth spurt in stature (9; 10); thus, adolescence may represent an
especially important opportunity for early intervention to prevent adult obesity (1).
Obesity during the childhood and adolescent years is also associated with a number of
short-term health consequences. Obese youth tend to have lower self esteem than their normal
weight peers, are often the victims of bullying and teasing, and are less likely to form strong
friendships (11; 12). Behavior problems also tend to be more frequent among obese youth (11).
In addition to these psychosocial factors, up to 70% of young girls (including those who are at a
normal weight) are actively trying to lose weight or have tried in the past (12). Rates of
psychological problems appear to be higher among clinically obese youth than the non-clinical
population (13).
Results from the Bogalusa, Muscatine, and Pathological Determinants of Atherosclerosis
in Youth (PDAY) studies have highlighted the early emergence of CVD risk factors among
overweight and obese youth (14; 15). Hypertension occurs at a higher rate among obese youth
than their normal weight peers and some research has shown that blood pressure in obese youth
may be especially sensitive to dietary sodium intake. Obese youth also exhibit high total and
low-density lipoprotein (LDL) cholesterol, high triglycerides, and low serum high-density
lipoprotein (HDL) cholesterol levels (16).
Physically, obese youth are susceptible to a number of comorbidities, some of which
occur as a result of the elevated CVD risk factors discussed here. Reported abnormalities of the
endocrine system include menstrual irregularities (amenorrhea, oligomenorrhea, early menarche,
and polycystic ovarian syndrome), insulin resistance, excess androgen production (especially in
girls) (12). Obese youth are at increased risk for gall stones and liver damage ranging from the

17

relatively mild fatty liver disease to steatosis, steatohepaptitis, or even cirrhosis, and sleep apnea
(11; 12).
Long-term medical and socioeconomic consequences are also a concern. As discussed
previously, obesity tends to track into adulthood. The same is true for other CVD risk factors
that tend to co-occur with obesity. Results from the PDAY study indicate that deposition of
plaques within the coronary arteries begins as early as 15 years of age (15). Sixty-one percent of
children participating in the Bogalusa Heart Study who were in the highest quintile of CVD risk
(indicated by multiple risk factor index) at baseline remained in the highest quintile 8-years later,
indicating clustered risk factors persist beginning in childhood (17). The persistence of obesity
and CVD risk factors from early life into adulthood increases the body’s exposure to these
factors beyond what was traditionally seen with adult onset of these traits. Thus, individuals
who are obese as youth may experience increased morbidity and mortality (cardiovascular and
all-cause) as adults. Additionally, menstrual abnormalities experienced during adolescence and
young adulthood may translate to fertility problems and increased risk of gestational
complications (12).
What is Metabolic Syndrome?
The term “metabolic syndrome” refers to the clustering of several risk factors for
cardiovascular disease including abdominal or visceral obesity, dyslipidemia, hypertension, and
insulin resistance or glucose intolerance. The presence of MetS is associated with increased risk
of disease (including cardiovascular disease, diabetes, and some cancers) and all-cause mortality
(18). Four definitions of adult MetS are widely used; these include definitions proposed by the
World Health Organization (WHO), the International Diabetes Foundation (IDF), and National
Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III), and the European
18

Group for the Study of Insulin Resistance (EGIR). These definitions have been reviewed by
Huang (19) and are summarized in Table 2.1. Additionally, researchers have proposed several
definitions for the diagnosis of MetS in children and adolescents, which the IDF has condensed
into one, universal definition of MetS for children and adolescents (Table 2.2).
Recent data indicate that 4%-7% of children in the US possess MetS (20–22). A clear
gradient exists across weight classifications, with the lowest prevalence being reported at less
than 1% in normal weight youth (22) and as high as 38.7% in moderately obese and 49.7% in
severely obese youth (21). Additionally, there is a clear increase in the risk of MetS with
increasing obesity, such that each half-unit increase in BMI z-score increases the risk of MetS
(OR = 1.55) (21). The prevalence of MetS does not appear to differ by ethnic group (22);
however sex differences do exist, with boys generally exhibiting a higher prevalence than girls
(~9% vs. ~4%, respectively) (22).
The prevalence of MetS in youth reported in the literature varies based on the
classification system used. As a result of this variation and in response to the low prevalence of
MetS among youth, several researchers have proposed the use of a continuous metabolic
syndrome score for the evaluation of risk in these age groups. Scores have been derived using
population-specific z-scores of the relevant components, principal component analysis, and
percentile rankings. These continuous scores (cMetS) are used as an index of cardiovascular
risk, indicating whether a child or adolescent is at risk for the development of MetS and
cardiovascular diseases (23).

19

Table 2.1. Summarized adult definitions of the metabolic syndrome.
Criteria
Obesity
Hyperglycemia

Dyslipidemia
Dyslipidemia
(additional)

WHO (1998)
Waist:hip ratio >0.90
(M) or >0.85 (W); or
BMI >30 kg/m2
Insulin resistance (by
IGT, IFG, T2D or other
evidence)
TG ≥150mg/dL or
HDL-C: <35mg/dL
(M) or <39mg/dL (F)
--

IDF (2005)
Waist circumference ≥
94 cm (M) or ≥ 80 cm
(F)
Fasting glucose
≥100mg/dL

NCEP-ATP III (2005) EGIR (1999)
Waist circumference > Waist circumference ≥
40 in (M) or > 35 in (F) 94 cm (M) or ≥ 80 cm
(F)
Fasting glucose
Hyperinsulinemia
≥100mg/dL or
(plasma insulin >75th
pharmacologic
percentile)
treatment
TG ≥150mg/dL or
TG ≥150mg/dL or
TG ≥177mg/dL or
pharmacologic
pharmacologic
HDL-C <39mg/dL
treatment
treatment
HDL-C: <40mg/dL
HDL-C: <40mg/dL
(M), <50mg/dL (F); or (M), <50mg/dL (F); or
-pharmacologic
pharmacologic
treatment
treatment
>130 mmHg systolic or >130 mmHg systolic or ≥140/90 mmHg or
>85 mmHg diastolic,
>85 mmHg diastolic,
pharmacologic
or pharmacologic
or pharmacologic
treatment
treatment
treatment

Hypertension

≥140/90 mmHg

Other

Urine albumin
≥20µg/min or
-albumin:creatinine
ratio ≥30mg/g
Insulin resistance or
Obesity, plus 2 of the 4
diabetes, plus 2 of the 5 criteria
criteria

Threshold for
diagnosis

--

--

Any 3 of the 5 criteria

Hyperinsulinemia, plus
2 of the 4 criteria

Criteria in bold are absolutely required for a positive MetS diagnosis; IGT = impaired glucose tolerance; IFG = impaired
fasting glucose; T2D = type 2 diabetes; TG = plasma triglyceride; HDL-C = high density lipoprotein cholesterol;
Adapted from Huang, 2009.
20

Table 2.2. International Diabetes Federation (IDF) consensus definition of the metabolic syndrome for children and adolescents (age
6-16 yrs).
Obesity
Age group
Triglycerides
HDL-C
Blood Pressure
Glucose
(WC)
th
MetS not diagnosed. Further consideration required in cases of family history of MetS,
≥ 90
6 - <10
hypertension, cardiovascular disease, dyslipidemia, and/or obesity.
percentile
th

IDF
Consensus
Definition

10 - < 16
Metabolic
Syndrome

≥ 90
percentile
or adult
cut-off if
lower

16+
Metabolic
Syndrome*

Cook
Criteria

≥ 94 cm
(M) or ≥ 80
cm (F)

--

≥ 90
percentile

≥ 150 mg/dL

< 40 mg/dL

≥ 130 mmHg
systolic or ≥ 85
mmHg

TG
≥150mg/dL or
pharmacologic
treatment

< 40 mg/dL
(M), < 50
mg/dL (F); or
pharmacologic
treatment

≥130 mmHg
systolic or ≥ 85
mmHg
diastolic, or
pharmacologic
treatment

≥ 110 mg/dL

< 40 mg/dL
(M&F)

≥ 90
percentile for
age and sex

Fasting glucose ≥ 100 mg/dL
(OGTT recommended for ≥ 100
mg/dL or known T2D)

Fasting glucose ≥ 100 mg/dL or
diagnosed T2D

th

th

Fasting glucose ≥ 110 mg/dL

* Adult MetS criteria are used for diagnosis in adolescents age 16 yrs and older
WC = waist circumference; HDL-C = high densitiy lipoprotein cholesterol; T2D = type 2 diabetes; OGTT = oral glucose tolerance
test

21

The use of a cMetS score improves upon the use of specific diagnostic criteria in a
number of ways. As previously discussed, the prevalence of MetS identified through the use of
the traditional dichotomous criteria is very low (~4%-7%) (20–22). This low prevalence results
in decreased statistical power in association studies. Additionally, the use of the dichotomous
criteria excludes youth who are at high risk of developing MetS and who may benefit from early
intervention to prevent progression of the syndrome. Although the use of a continuous score
addresses some limitations of the dichotomous MetS classification, several scores using differing
criteria have been reported in the literature (23).
The traditional method of computing cMetS scores is severely limited by the use of
sample-specific statistics. These sample-specific cMetS scores have poor external validity
because they are specific to the study sample on which they are based, and thus it is difficult to
extrapolate findings from individual studies to the population at large. For this same reason, the
use of a sample-specific score for specific study groups (i.e. obese adolescents) has limited
utility. Using NHANES III data, Cook and colleagues developed growth curves for each feature
of MetS and constructed age- and sex-specific percentiles for each variable. These data were
also used to identify specific percentile cut points that correspond to the adult NCEP-ATPIII
MetS criteria for each variable, which can be used to diagnose MetS in youth (24). The
normative data described in the study by Cook can thus be used to create a cMetS score that is
based on nationally representative data.
In summary, several definitions of MetS exist for youth, making it difficult to compare
the prevalence of MetS between studies. Recent growth curves published by Cook and
colleagues (24) allow children and adolescents from the US to be classified based on age- and
sex-specific nationally representative data and should increase the accuracy of classification of
22

MetS in this population. Additionally, these growth curves allow cMetS to be calculated based
on national data rather than population-specific data, and allows metabolic risk to be treated as a
continuous variable rather than as a dichotomous variable.
Physical Activity and Metabolic Syndrome/Cardiovascular disease in Children
Physical activity has been associated with a number of health benefits and inadequate
activity is generally regarded as one of the cornerstone risk factors for overweight and obesity
(25). Thus, it is reasonable to believe that PA may be beneficial in the prevention of treatment of
overweight, obesity, and related conditions in childhood as well as adulthood.
In general, the results of large cross-sectional studies have demonstrated inverse
relationships between habitual physical activity and MetS in children and adolescents. Study
methods have varied in terms of measuring PA and classification of MetS and/or CVD risk status
but over all, PA appears to be beneficial in reducing cardiovascular risk among youth. Results
from the European Youth Heart Study (EYHS), in which physical activity was measured by
accelerometry, indicate that time spend in PA is inversely associated with cMetS independently
of the effects of cardiorespiratory fitness on the score (26). These findings are supported by
those of other large population studies also implementing accelerometry, such as NHANES in
the US (27), AFINOS in Spain (28), and another study in Vietnam (29). When PA is measured
by self-report, less active youth are still significantly more likely to meet the criteria for MetS
than their more active counterparts (30).
Differing intensities of activity may have varying influences on MetS. In the AFINOS
study, only vigorous PA was associated with MetS (28). In the EYHS, on the other hand, low-,
moderate-, and vigorous-PA were each independently associated with cMetS (26). The differing

23

results on intensity of activity may be explained, in part, by the relationship between physical
activity and cardiorespiratory fitness (CRF). While these terms are not the same they are often
used interchangeably, and often the assumption is that increased PA means increased CRF.
Studies examining the independent relationships of PA and CRF with MetS have found differing
results. In general, those studies that have found only CRF to be independently associated with
MetS have used self-reported measures of PA (3). When using objective measures of PA,
Ekelund and colleagues found that increasing PA was associated with MetS even beyond the
influence of CRF (26). Still others have found that the effect of PA is weakened once adjusted
for CRF (28; 31). However, consideration of body fatness may further weaken these
associations. In a large study of children and adolescents, Ondrak and colleagues found that
CVD risk was associated with aerobic fitness but that percent body fat was a much stronger
predictor (32). These results suggest a need to further examine the influence of CRF on MetS
and levels of individual variables, independently of body composition. Interestingly, the
relationships reported by Ondrak and colleagues were strongest in the youngest children (ages 810yrs) and weakest in the oldest (ages 14-16 years) while CVD risk score was highest in those
ages 11-13 yrs. This finding supports the notion that adolescents are especially vulnerable to
increased levels of CVD risk factors during puberty, and the authors suggested that the dramatic
changes in the hormonal environment that occur during puberty might explain the reduced
influence of fatness on CVD risk score after this point (32).
Adolescents at highest risk for CVD may benefit more from increased PA and/or CRF
than their lower-risk peers (33–35). When comparing quintiles of CRF, Lobelo and colleagues
found decreasing CVD risk score with increasing CRF. In fact, the largest decrease in CVD risk
was found between the first and second (very low to low) CRF quintiles (35). The results of the

24

AFINOS study suggest that the relationship between CRF categories may be dependent upon the
intensity of habitual PA. Adolescents with lower fitness levels were apparently protected from
MetS by participating in more VPA. Therefore, while it is still unclear whether PA or CRF has a
greater influence on MetS, and whether the effects of these traits are independent of each other,
increasing VPA may be effective for reducing MetS risk among adolescents with low fitness
levels (28).
Influence of Diet
As shown in Table 2.3, adolescents typically consume amounts of carbohydrates
(including added sugar), total fat, and protein within the Acceptable Macronutrient Distribution
Range (AMDR) set by the IOM. However, most fail to meet the recommendations for saturated
fat, mono-unsaturated fatty acids (MUFA), poly-unsaturated fatty acids (PUFA), and fiber (36–
38). In terms of micronutrients (Table 2.4), most adolescents do not consume sufficient vitamin
A, folate, potassium, calcium, and magnesium, while also consuming much more sodium than
recommended (39–42). Overweight and obese adolescents tend to exhibit different dietary
patterns than those of their normal weight peers. Specifically, low dairy intake is associated with
increased BMI, sum of skinfolds (suprailiac and subscapular) (43), and increased waist
circumference (44) in adolescent girls and boys. Low intakes of fruits and vegetables are also
inversely related to adiposity, while children who eat more servings of meat tend to have larger
waist circumferences (44). In light of these relationships, it is not surprising that overweight and
obese adolescents tend to have more adverse CVD risk factor profiles than their normal weight
peers.

25

Table 2.3. Macronutrient intakes of adolescents compared to dietary reference intakes (DRIs)
Recommended (AMDR)

Actual Intake

Difference (AMDRActual)

Carbohydrates Total
45-65
52.6 ± 0.4
Within range
(% of kcals)
Added Sugars
< 25%
19.1 ± 0.4
Within range
(% of kcals)
Dietary Fiber
14
6.0 ± 0.1
-60.3%*
(g/1000 kcal)
Fat
Total
25-35
33.4 ± 0.34
Within range
(% of kcals)
Saturated
< 10
11.8 ± 0.16
+18%
(% of kcals)
Monounsaturated (% of
Up to 20%
12.6 ± 0.15
-37%
kcals)
Polyunsaturated
10%
6.7 ± 0.21
-33%
(% of kcals)
Protein
10-30
13.7 ± 0.2
Within range
AMDR: Acceptable Macronutrient Distribution Range. Note: AMDR values are taken from the DRI tables at http://fnic.nal.usda.gov,
adapted from the DRI reports at www.nap.edu; Recommendation for dietary fiber taken from 2010 Dietary Guidelines for Americans.
Repored intakes adapted from Deshmukh-Taskar et al.(37), and Troiano et al. (36).
* Because of differing recommendations, mean reported for both sexes combined was compared to the DRI for each sex separately.

26

Table 2.4. Intakes of selected vitamins and nutrients related to cardiovascular health.
Recommended (DRI)
Both
Males Females
Sexes

Actual Intake
Both Sexes

Males

Females

Difference (DRI-Actual)
Both
Males
Females
Sexes

Vitamins
Vitamin A
-900
700
-874
623
--2.9%
-11.0%
(µg/day)
Vitamin C
-75
65
-94.2
67.6
-+25.6%
+4.0%
(mg/day)
Vitamin E
5
--7.7
--+54%
--(mg/day)
Vitamin D
5
--6.7
--+34%
--(µg/day)
Folate
400
--231
---42.2%
--(µg/day)
Minerals
Sodium
1500
--3363
--+124.2%
--(mg/day)
Potassium
4700
--2459 ± 43.6
---47.7%
--(mg/day)
Calcium
1300
--1134
---12.8%
--(mg/day)
Magnesium
-410
360
250 ± 3.9
--39.0%+ -30.6%+
(mg/day)
Selenium
55
--81
--+47.3%
--(µg/day)
Note: Recommended Dietary Allowances (RDA) and Adequate Intakes (AI) values are taken from the DRI tables at
http://fnic.nal.usda.gov, adapted from the DRI reports at www.nap.edu. Reported intakes adapted from Lytle et al. (39); Stang et al.
(40), Sugiyama et al. (41); and Volpe et al. (42).

27

Intervention studies examining the influence of dietary patterns on MetS and CVD risk
factors in adolescents are limited; therefore much of the information presented here comes from
studies involving adults. One of the most notable dietary interventions for reducing CVD risk is
the DASH (Dietary Approaches to Stop Hypertension) eating plan. This was the first trial to test
the effects of entire patterns of eating, rather than simply reducing sodium intake or
supplementing with other nutrients (i.e. magnesium), on lowering blood pressure in hypertensive
adults. Participants randomized to the DASH group received dietary and behavior change
counseling, and were directed to eat a diet high in fruits and vegetables (approximately 9
servings per day), that included consumption of low-fat dairy products (2 servings per day), and
low in sodium (<3000mg/day) and saturated fat (<6% of total calories per day). Subjects
followed the prescribed diet for 8 weeks. By the end of the intervention period, the reduction in
systolic blood pressure in the DASH group exceeded that of the control group by 11.4mmHg and
importantly, also exceeded the reduction in the high fruit and vegetable group by 4.1mmHg. The
DASH group also experienced a greater reduction in diastolic blood pressure, though this
difference was not as large (-5.5 mmHg in DASH compared to control) (45). These results
indicate that following the DASH eating pattern allows for significant reductions in both systolic
and diastolic blood pressure even without weight loss.
The DASH diet has also been used to treat other components of MetS among overweight
and obese individuals in Tehran, Iran (46). Participants following a hypocaloric DASH diet were
compared to a weight reduction group and a control group. After the 6-month long intervention,
participants in the DASH group experienced improvements in blood pressure (systolic and
diastolic), lipids, triglycerides, and weight, and improvements in this group exceeded those in the
weight control group for each component (46). The DASH diet remained superior to the weight

28

control diet even after controlling statistically for weight loss, a finding that is supported by the
previous DASH study, in which blood pressure was improved without a significant change in
body weight (45).
Diets that are particularly high in fruit and vegetables (and thus, fiber and select
micronutrients) have also been successful in management of type 2 diabetes (T2D) (47) and
other CVD risk factors (48). A recent review (47) of the role of vegetarian diets in the treatment
of T2D summarizes several observational and intervention studies. Seventh-day Adventists, who
tend to lead generally low-risk lifestyles, the prevalence of T2D is about twice as high among
non-vegetarians as it is among vegetarians. Intervention studies in which subjects follow low-fat
vegetarian (or near-vegetarian) diets have shown improvements in insulin sensitivity, fasting
blood glucose, and hemoglobin A1c, to the point that some subjects taking insulin have been
able to discontinue it’s use. Because some of these studies have also included intense exercise
programs and have seen rather significant weight loss (7.2 kg in one study), some of the
improvement in glycemic control is probably a result of this weight loss. In addition to the
improvements in measures related to diabetes, some studies have also found that the adoption of
vegetarian diets results in lower LDL cholesterol and higher HDL cholesterol, reducing the risk
of cardiovascular events among individuals with adverse lipid profiles (47).
The CARDIA study (Coronary Artery Risk Development in Young Adults) study yielded
two papers examining the relationship between dietary factors and MetS or CVD risk factors (48;
49). Subjects in this study were 18-30 years old, and had been followed for 10 years at the time
of publication. Dietary habits were assessed using a food frequency questionnaire at study
enrollment and also during year seven. In this study, participants who ate more fiber tended to
weigh less and to gain less weight over the follow-up period. Fiber intake was also inversely
29

associated with fasting insulin in both white and black participants. Similar patterns were found
for blood pressure and triglycerides, though a large portion of this relationship was explained by
the improvements in insulin levels (48). In an additional paper from the same study, dairy intake
was also found to be associated with a more favorable insulin resistance syndrome (IRS) profile,
but only among overweight subjects (49). As might be expected from the results of the previous
paper (48), increased fiber intake was also associated with a more favorable IRS profile. The
authors also adjusted for weight gained over the 10-year follow-up period, and found that dairy
intake was still associated with lower odds of IRS, although the confidence intervals were
somewhat wider in the adjusted model than in the unadjusted model. Interestingly, both lowand high-fat dairy products were associated with reduced odds of IRS. The high content of
calcium, magnesium, and potassium in dairy products may be partially responsible for this
relationship (49), and it is important to note that those three nutrients are also emphasized in the
DASH eating plan (45).
Somewhat in contrast with the findings in adults, Latino adolescents who increased their
fiber intake across a 16-week intervention experienced significant reductions in both BMI and
visceral fat (measured by MRI), but these reductions did not lead to improvements in blood
glucose or insulin levels (50). Those adolescents who reduced their sugar intake did see
improvements in insulin secretion, although no change in BMI or visceral fat occurred in this
group. These results seem to suggest that changes in visceral fat mass may not have the same
positive effect on insulin resistance that is seen in obese adults. However, it is possible that these
adolescents experienced other positive changes as a result of this decreased fatness, perhaps in
the form of reduced blood pressure or circulating triglycerides, which were not reported in this

30

paper. More studies are necessary to determine the best dietary intervention to treat MetS during
adolescence.
Assessment of Dietary Patterns
Several different dietary assessment methods are available, and each method has a
number of strengths and weaknesses. When selecting a dietary assessment tool, it is necessary to
weigh the advantages and disadvantages associated with each method in terms of the accuracy of
the data obtained, the investment of time, money, and personnel, and subject/respondent burden.
Advantages and disadvantages of several common methods are summarized in Table 2.5.
Arguably, the most accurate method of dietary assessment is observation. In this method,
a very well trained, unobtrusive observer watches individuals (usually at meal-time) and is able
to quantify food intakes. This is often done in school lunchrooms or similarly controlled
settings, and thus it is possible to obtain detailed information about the foods consumed,
including preparation, condiments, and other details. Although the observation method is
accurate, it is very time-consuming and expensive, since trained observers are required. Also,
the observer has to take great care to be unobtrusive, so that the subjects do not alter their eating
behaviors (51). Another very accurate but expensive and time-consuming option is diet history.
In this case, a trained interviewer talks with the subject about his or her usual dietary intake
(recent and long-term). The interview format allows for the collection of very detailed data and
any unclear answers can be clarified right away, reducing confusion or misinterpretation on the
part of the interviewer. Food frequency questionnaires (FFQ) can also assess long- or short-term
dietary intakes, though these can be self-administered and require significantly less investment of
staff time. The time period for FFQs can vary from a week-by-week basis up to a year.

31

Table 2.5. Advantages and disadvantages of several common dietary assessment methods (51)(135).
Method
24-hour recall

-

Food record

-

Food Frequency
Questionnaire

-

Diet History
Questionnaire
Observation

-

Advantages
Low subject burden
Can be performed quickly
Accurate for estimating intakes at the
group level with just one recall
Doesn’t cause subjects to change their
behaviors
Specific time-period
Doesn’t rely on memory (if foods are
recorded right away)
Subjects can be trained in groups,
reducing time investment of research staff
One day provides accurate information
for group-wide estimates of dietary intake
Doesn’t require trained interviewers
Can be self-administered or interviewerassisted
Not affected by alterations in habits
Can assess the diet as a whole, or select
nutrients or foods
Inexpensive
Low subject burden
Not affected by altered habits
Can get very detailed information about
foods, preparation methods, etc.
Not affected by altered habits
Doesn’t rely on memory at all
Very low subject burden
Defined time period
Possible to measure portion sizes

32

-

Disadvantages
Requires trained interviewer
Time intensive for interviewer, can be expensive
Not appropriate for estimating individual intakes (with just
one day)
Requires good memory
Portion sizes may not be estimated accurately
Higher subject burden
Subjects may alter dietary patterns
Recording for several days may result in reduced validity
Large volume of data, expensive to code

- Heavily reliant on subject memory of dietary patterns
- Recall period is imprecise
- Quantification is imprecise – portion sizes, frequency of
consumption
- Food descriptions aren’t as detailed as with diet records or
recalls
-

Requires highly trained interviewers
Very high subject burden, interview can take a couple hours
Recall period and portions may be imprecise
Requires a lot of staff time, can be expensive
Requires highly trained observers
Expensive – requires a lot of staff time

Two other common dietary assessment methods are diet records and dietary recalls (52;
53). Diet records can include data from as little as one day, or can be kept continuously for
indefinite periods of time. Frequently, several single-day records are used to estimate overall
patterns, though several days (sometimes a month or more) are necessary to accurately quantify
intakes of some micronutrients (54). Recalls are generally performed over the phone, and
typically cover a single day of food intake (24 hours). With both diet records and recalls it is
possible to gather detailed information about foods and drinks, including preparation methods.
With these advantages and limitations in mind, the 3-day diet record method is a good
choice to reasonably estimate dietary habits among 12-18 year-old adolescents. Collecting three
days worth of records should increase the validity of the data obtained while keeping the subject
and data-entry burden relatively low. While recording intake for a week or more might provide a
more accurate assessment of intake of some nutrients (especially micronutrients) (54), it is
important to balance that with the increased subject burden and the chance that subjects might
not record data as accurately when required to do so over an extended period of time (51).
Another option would be to perform several 24-hour recalls, though researcher burden would be
increased in this case.
Several steps can be taken to improve the quality of the data obtained using this method.
First, it is important to make sure the subject is well trained in the method and understands the
importance of recording absolutely everything consumed, along with preparation methods,
condiments, and anything else added to the food. Providing a pocket-sized record the subject
could carry on their person throughout the day may increase compliance. Collecting data on two
weekdays on which the subject follows different schedules (i.e. Monday and Thursday) and one
weekend day should allow for differences that occur on a daily basis, such as evening activities
33

or sleeping in on the weekends. To aid in the reporting of portion sizes, the subjects can be
given prompts (such as pictures or descriptions) to help them remember what constitutes a
serving of different foods and estimate the amount of each item eaten. Researchers should
contact subjects to clarify any unclear entries in order to minimize error in data entry and
quantification of nutrient intakes. Additionally, the data should be analyzed using well-known
nutrition software that is generally accepted as accurate.
The Healthy Eating Index (HEI) was originally developed in 1995 by Kennedy and
colleagues (55), in an attempt to design a single index of dietary quality based on US dietary
guidelines. The HEI can be used either to assess dietary quality or to measure changes in dietary
quality over time. A total of 10 components are included in the scale; each of these 10
components is scored on a range from 0-10, so that subjects receive a diet quality score ranging
from 0-100, with a higher score indicating superior dietary quality. The ten components making
up the HEI include the five major food groups (fruits, vegetables, grains, dairy, and meat), total
dietary fat (percent of total kilocalories), saturated fat intake (percent of total kilocalories),
cholesterol intake, sodium intake, and dietary variety. Scores for the first five components, the
major food groups, are based on the age- and sex-specific total daily serving recommendations
for each individual. The two components dealing with fat intake, components 6 and 7, are
scored based on the daily recommendations for each nutrient, such that individuals consuming
less than the recommended daily intake receive a score of 10. Components 8 and 9, cholesterol
and sodium, are based on milligrams consumed per day. Finally, dietary variety is assessed by
counting foods consumed at a minimum of one-half serving in each food group and equivalents
are added to determine how many different foods each individual consumed. A perfect score of
10 is assigned to this category if an individual consumes at least 16 different foods over the 3-

34

day study period (55). In 2008, the HEI criteria were updated to reflect the 2005 USDA dietary
guidelines (56). These new guidelines, HEI-2005, reflect both total and whole fruit intake, total
vegetables and individual types of vegetables (i.e. dark green leafy vegetables), total and whole
grains, milk, meat and beans, oils, saturated fat, sodium, and calories from solid fats, alcohol,
and added sugars.
Stress
The term “stress” is generally used in a broad sense to describe any physical,
environmental, psychological, or physiological challenge requiring adaptation (57). The human
stress response was first characterized by Walter Cannon, who coined the term “fight-or-flight”
to describe the endocrine changes that occur to prepare the body to deal with acute danger (58).
Later, Hans Selye suggested that the body responds and adapts to stressors through a process
consisting of three phases, which he termed the General Adaptation Syndrome (GAS) (59).
More recently, the terms “allostasis” and “allostatic load” have been used to describe the body’s
adjustments and adaptations to challenges (57).
Selye suggested that the body responds to stressors in a non-specific way (59). Thus, the
assumption was that the body responds in similar ways to physical threats, such as a serious
injury, and perceived threats, such as a final exam or the loss of a job. More recently, however,
the importance of the individual’s subjective appraisal of stress has received more attention. As
originally suggested by Lazarus (60), the stress response can vary greatly based on the way in
which a stressor is perceived. In adults, neighborhood problems such as crime, violence, and
unemployment have been associated with increased psychological distress and poor health
behaviors, even after adjustment for neighborhood SES, age, sex, and other confounders (61).
These results highlight the importance of perception of stressors, suggesting that people from
35

lower-income areas and neighborhoods with more crime and violence interpret potential
stressors in a more threatening way.
Several instruments have been used to assess perceived stress in a number of populations.
Many of these instruments were developed for use in adults and adapted for use with children.
Additionally, studies implementing paper-and-pencil instruments for use with children have been
conducted in fairly homogenous groups and have not considered the unique experiences of an
ethnically and economically diverse group of individuals (62).
Assessment of Psychosocial Stress
Stress Inventories. Stress inventories or checklists can be generalized, asking about a
wide range of potential stressors, or more specialized and focusing on specific issues.
Inventories are often limited by individual interpretations of threats, inclusion of limited stressor
items, and a lack of detailed information about an individual’s perception of each stressor.
Perceived Stress Scale (PSS). The PSS was originally developed for use with adults by
Cohen and colleagues in 1983 (63). This inventory was designed to provide an appraisal of
stress and coping over the preceding month. A full, 10-item scale and a shorter 4-item version
are both available, however the longer version is more reliable and valid. Each item is rated on a
5-point Likert scale (never, rarely, sometimes, often, and very often). One example question is
“How often have you felt that you could not control the important things in your life?” The score
from all items are scored to create a total perceived stress score ranging from 0-40 for the long
version or 0-16 for the short version (63).
Adolescent Stress Questionnaire (ASQ). Originally developed in 1993, the ASQ was
developed in focus groups of adolescents in school years 7-12 (64). The questionnaire was

36

updated in 2006 to include stressors that are more appropriate for today’s adolescents. Ten
scales with a total of 58 items are included in the questionnaire.
Adolescent Perceived Microsystems Scales (APMS). Many inventories used in youth studies
have either been originally designed for use with adults, or have been developed using focus
groups from specific ethnic and SES groups, often limiting their generalizabilty to lower SES
groups and minorities. The APMS (62) was developed specifically for a culturally diverse,
urban population, using focus groups from three urban areas (Baltimore, MD; Washington, DC;
and New York, NY). The original sample was 26% black, 26% white, 37% Latino, and 59%
female, extending the applicability of the questionnaire to these groups. The APMS provides an
assessment of an adolescent’s perception of his or her family, peer, school, and neighborhood
microsystems. These microsystems are assessed on three different scales:
-

Social support: An adolescent’s family, peer, and school relationships.

-

Daily hassles: school performance, family responsibilities, problems in the neighborhood

-

Involvement: extracurricular school activities, leisure, recreational activities,
neighborhood involvement

Pediatric Symptoms Checklist. The Pediatric Symptoms Checklist (PSC) is a 35-item
questionnaire designed for use by pediatric practitioners to detect psychosocial problems among
their patients (65). For each item, the child (or a parent) indicates whether the statement
describes the child never, sometimes, or often. Each item is scored based on the response and
these scores are added to compute at total score. Used primarily with chronically ill children, the
PSC has been implemented to identify symptoms of internalizing and externalizing difficulties
and other psychological distress in a primary care (66) and/or chronically ill patients (67; 68).

37

Validity and reliability of the PSC has been demonstrated in children with varying degrees and
types of psychological problems (69). Based on this information, the PSC is the instrument used
by clinical staff at the Helen DeVos Children’s Hospital Healthy Weight Center (HWC), and is
the instrument that will be used to assess psychosocial stress in this study. Staff at the HWC
have communicated with the developer of the PSC and have identified subscales indicative of
depression and anxiety.
Interviews. Interviews improve upon some of the weaknesses of questionnaires in that they
are more able to distinguish context and an individual’s interpretation of a given stressor.
Life Events and Difficulties Schedule (LEDS). Originally developed for adults, the LEDS is
administered by trained interviewers and scored by separate raters who are blind to the
individual’s subjective responses. Stressfulness of specific events are rated based on the event
context and past experiences, long-term threat, timing of the event, and whether the event is
focused on the individual or on others. The LEDS also differentiates between acute stressors,
termed events, and chronic stressors, termed difficulties (70).
Standardized Event Rating System (SEPRATE). The SEPRATE is more specific than the
LEDS, and is made up of yes/no questions regarding 84 standardized stressors. Any “yes”
answers are further probed with additional questions in order to identify the magnitude of change
induced by the event, desirability of the event, disruptiveness to daily routines, threat to life, and
the degree of the individual’s control over the event. As in the LEDS, raters are blind to social
vulnerability and subjective responses.
In summary, many methods exist to assess stress using either surveys or interviews. Surveys
can evaluate a wide range of stressors or be very specific, but can be limited in their ability to

38

distinguish context and individual reactions to stressors. Interviews improve upon these
weaknesses in that they often are more able to distinguish context and interpretations of
stressors, but they may be more susceptible to interviewer bias. This bias can be reduced through
interviewer training. The method selected should be appropriate for the subject population (i.e.,
have been developed for individuals of similar age and social situation) and as free as possible of
bias. The PSC (youth report) is the instrument used by clinical staff at the Helen DeVos
Children’s Hospital Healthy Weight Center (HWC), and is the instrument used to assess
psychosocial stress in this study.
Psychosocial Function
The term “psychosocial functioning” has been used in a very broad sense, often
encompassing peer and family relationships, health-related quality of life (HRQoL), school
performance, and mental health. In most cases, psychosocial functioning describes the way
individuals interact with and relate to others and to themselves. It can be defined as “…the
ability to cope and tolerate stress, and the capacity for developing a value and belief system.”
(71). In this sense, optimal psychosocial functioning is necessary in order to adequately cope
with stressors and threats that are encountered on a daily basis. An individual with poor
psychosocial functioning might lack supportive peer relationships, have poor self esteem, or
experience depression, anxiety, or other mental health problems, especially when faced with
stressors such as difficulties in school or disruption of the family home environment.
Conversely, an individual with good psychosocial functioning would be able to successfully
cope with these stressors without experiencing a major disruption in day-to-day functioning.
Psychosocial factors that have been associated with the onset of pediatric obesity include
interpersonal problems with peers, divorce in the immediate family, personal medical concerns,
39

and a variety of other peer and family factors. Many of these are also frequently named as
barriers to healthy behaviors among obese youth, especially in terms of the ability to make
healthy lifestyle changes (72). Studies of psychosocial function among obese youth generally
consider experiences with bullying, adversity such parental divorce and family uncertainties, and
periods of major depressive disorder and other acute psychiatric difficulties (73–75). Many of
these studies in both children and adults focus on HRQoL as a marker of psychosocial function
(76; 77).
Physical activity and participation in leisure time activity are associated with improved
HRQoL among both children and adults, and these findings are consistent within both healthy
and chronically ill populations. Among adults with prediabetes, individuals who meet PA
recommendations score better on HRQoL assessments than those who do not meet
recommendations. More specifically, improved HRQoL is associated with increased
MET*min/week of PA, maintenance of a healthy BMI, and avoidance of smoking (76). Other
factors that have been associated with improved HRQoL among adults include lower blood
pressure, waist circumference, perception of better physical function, and total cortisol.
Furthermore, participation in LTPA is associated with improvement in both psychosocial and
physical measures of well-being, including higher levels of positive psychosocial states (i.e
vigor, well-being, and calm) and reduced frequency of depression and negative affect (78).
In obese adolescents, a higher BMI and increasing number of comorbidities is associated
with decreasing HRQoL (77). Participants in this study were seeking treatment for pediatric
MetS clinic, and thus experienced at least two of the components of MetS. Health related quality
of life was assessed using the Impact of Weight on Quality of Life-Kids (IWQOL-Kids)
questionnaire, and assessed the domains of body esteem, social life, family relationships,
40

physical comfort, and a total score. Scores on both the youth and parent report forms of the
IWQOL-Kids were inversely associated with BMI and BMI z-score (all subscales), indicating
that HRQoL decreased with increasing levels of obesity. Additionally, the number of
comorbidities a participant had varied with the total score and subscale scores on the IWQOLKids, such that participants with more comorbidities had higher scores on the inventory. Female
participants and non-Hispanic white participants experienced poorer quality of life, even after
consideration of BMI, indicating that females and non-Hispanics may experience more of the
negative psychosocial affects of obesity (77).
Biomarkers of the Stress Response
The stress system, made up of the sympathetic nervous system (SNS) and the HPA axis.
Stressors may be of either internal or external origin, and may be physical, psychological,
chemical, social or biological. Exposure to a stressor stimulates both the SNS and HPA axis to
regulate changes in metabolism, circulation, and respiration in order to prepare the body for
intense physical activity in emergency situations (79).
Activation of the SNS is termed the “fight or flight” response, initially described by
Walter Cannon (58). Exposure to a stressor and activation of the SNS causes increased secretion
of epinephrine and norepinephrine into the peripheral circulation, and results in increases in heart
rate (HR), blood pressure (BP), blood glucose levels, blood flow to the skeletal muscles, and
bronchodilation. SNS activation also results in suppression of growth, reproduction, and
digestion.
While SNS activation is the body’s initial response to an acute stressor, prolonged
stressors result in activation of the HPA axis. When stimulated, corticotropin-releasing

41

hormone (CRH) is released from the hypothalamus and stimulates the pituitary gland to release
adrenocorticotropic hormone (ACTH) into the blood stream. Circulating ACTH stimulates
release of cortisol from the adrenal cortex. Increases in circulating cortisol then cause metabolic
changes throughout the body, promoting glycogenolysis and lipolysis in the skeletal muscle and
liver, as well as protein catabolism and gluconeogenesis from non-carbohydrate precursors.
Integration of SNS and HPA axis activation allows the body to properly respond to both
acute and chronic stressors. However, chronic over-activation of these systems can have
detrimental effects. The work of McEwen and Bjorntorp highlights the detrimental effects of
chronic HPA and SNS activation, which include increased susceptibility to dementia and
depression (79), hypertension (80), and abdominal obesity (80). Because activation of the SNS
and HPA axis results in changes in circulating hormones, these hormones are often used as
biomarkers of the stress response. Cortisol is one of the most common biomarkers used for this
purpose.
When the HPA axis is functioning normally, cortisol is released into the circulation in a
predictable diurnal pattern. Cortisol release peaks early in the morning, approximately 30
minutes after waking and falls throughout the day. Normal HPA axis activity can be
characterized by pulsatile release of both ACTH and cortisol over the course of the day. The
cortisol peak seen early in the day is due to more frequent, stronger pulses of ACTH and cortisol
release. Activity of the HPA axis is regulated primarily through a negative feedback system, so
that high levels of circulating cortisol inhibit the release of ACTH from the pituitary, and high
levels of ACTH suppress CRH release from the hypothalamus. This negative feedback control
limits tissue exposure to cortisol, thus regulating cortisol’s effects on metabolism and other
physiological functions. Other factors that may alter HPA axis activity include other hormones
42

and cytokines (81). Acute stressors, such as a meal, physical activity, or mental stressors
stimulate the release of cortisol, such that the normal diurnal pattern can be altered at any time
throughout the day.
Once released into the blood, cortisol circulates bound to corticosteroid-binding globulin
(CBG) and a small amount circulates freely in solution (about 10%) or bound to albumin (about
15%). Plasma levels of CBG are increased in high estrogen states (i.e. pregnancy) or in the
presence of other endocrine conditions, such as hyperthyroidism or diabetes. Conversely,
plasma CBG is reduced in cases of hypothyroidism; genetics also play a role in either increased
or decreased plasma CBG concentration. At the target tissues, cortisol binds to glucocorticoid
receptors within cells, and this cortisol-receptor complex is responsible for most of the actions of
cortisol in the body. In fact, differences in glucocorticoid receptor gene expression may explain
some race and gender differences in cortisol levels (82).
Once it reaches the tissue, the action of cortisol is controlled by several different
enzymes. One of these enzymes, 11β-hydroxysteroid dehydrogenase (11β-HSD) exists in two
isoforms. The first isoform, 11β-HSD type 1 (11β-HSD1) is present in both the liver and
adipose tissue and is responsible for reactivating cortisone to its active form, cortisol. The
second isoform, 11β-HSD type 2 (11β-HSD2), is found in the kidney and converts cortisol to its
inactive form, cortisone. When functioning properly, the action of 11β-HSD types 1 and 2 help
to control the effects of cortisol at the tissue level, ensuring adequate levels within tissues
depending on cortisol for metabolic control (liver and adipose tissue) and inactivating excess
cortisol in the circulation (83). Tissue-specific alterations in the activity of 11β-HSD1 have been
documented in obese men (83) and obese pubertal adolescents (84). In these groups, reactivation
of cortisone to cortisol is increased within adipose tissue while it is simultaneously reduced
43

within the liver, resulting in increased excretion of cortisol metabolites in urine in concert with
increased concentrations of cortisol within adipose tissue. As suggested by Wiegand and
colleagues (84), this altered enzyme activity may protect individuals from increased circulating
cortisol levels and the concomitant hyperinsulinemia and insulin resistance that has been
demonstrated in individuals with Cushing’s syndrome.
Traditionally, cortisol responses to acute stressors and diurnal variation in cortisol levels
have been measured in serum either via a single venipuncture or repeated samples drawn from a
venous catheter. Because cortisol is released in response to stress, serum measures may be
subject to error. More recently, saliva sampling has been used to assess these parameters.
Cortisol levels in saliva have been found to accurately reflect serum free-cortisol levels (that not
bound to cortisol-binding globulin) (85), which is often thought to be the more biologically
active portion. Saliva collection is non-invasive and allows researchers to avoid the additional
stress of venipuncture. Additionally, saliva collection does not need to be done in a laboratory,
allowing for analysis of cortisol levels in normal daily living situations. Analysis of saliva
samples can be done via enzyme-linked immuo-absorbent assay (ELISA) techniques, avoiding
the potential radiation exposure necessary to analyzed serum cortisol values via
radioimmunoassay (RIA).
Though the availability of technologies to analyze cortisol levels in saliva have made
assessment of cortisol in free-living situations and among all age groups significantly more
feasible, several factors must be considered when designing studies.
Endocrine Changes During Puberty. The onset of puberty and maturation of the
hypothalamic-pituitary-gonadal (HPG) axis, resulting in the attainment of sexual maturation, are

44

the major characteristic of the adolescent period. At the onset of puberty, pulsatile release of
gonadotropin-releasing hormone (GnRH) from the hypothalamus stimulates release of the
gonadotropins luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from the
pituitary gland (86). The frequency of release of the gonadotropins in modified by circulating
steroid hormones – in fact, circulating cortisol is known to decrease the LH response to GnRH in
both males and females (82). Ultimately, the gonads (testes in boys, ovaries in girls) begin to
produce larger quantities of sex steroids (testosterone in boys, estradiol and progesterone in girls)
(86). These changes in sex hormone release, as well as increased release of
dehydroepiandrosterone-sulfate (DHEA-S) and growth hormone result in the rapid gain in height
and changes in body composition that are characteristic of the adolescent growth spurt (87).
Maturity-Related Variation in HPA Axis Activity
Infancy. Before the advent of technology to measure cortisol levels in saliva, assessment
of HPA axis in infants was conducted largely in animals or relied on use analysis of urinary
cortisol and its metabolites. The adoption of salivary cortisol assessment has greatly increased
the number of studies of both basal and stress-related activity of the infant HPA axis (88).
Newborn infants exhibit low levels of total cortisol in serum, and the concentration of
cortisol increases over the first year parallel to increases in the concentration of CBG. Cortisol
in saliva, however, shows a different pattern, with similar levels in the newborn compared to
later in the first year (88) – this difference is because salivary levels of cortisol reflect the amount
of free cortisol in serum, which is largely unaffected by CBG levels as discussed previously.
The early morning peak in cortisol levels is evident as early as seven weeks of age in about 30%
of infants, and in almost all infants by 12 weeks (89). Additionally, cortisol levels increase after
feeding (post-prandial increase) in infants as young as 9 months (88; 90). Because of this, it is
45

important to control for the time of cortisol sampling beginning with very young infants,
especially if sampling is to occur in the morning when cortisol levels are changing quickly (88).
Other factors that should be controlled in studies include maternal use of medications and/or
caffeine (for breast-fed infants), the timing of naps and car rides (90), and gestational age (88).
Between the ages of 12- and 18-months, correlations have been found between waking time and
baseline cortisol levels, as well as with percent change (with respect to baseline) in cortisol levels
in response to stress though baseline levels themselves may not be correlated with the stress
response (90), a finding that differs from those in older children (91).
Childhood and Adolescence. Ease of sampling and the non-invasive nature of saliva
collection have made this method very popular for cortisol research in young children. However,
some characteristics of cortisol secretion, and of children’s behavior, need to be considered when
designing studies in this age group (92).
In order to collect useful cortisol data in all age groups, it is necessary to standardize the
collection time, use consistent procedures and materials, and control for food, drinks,
medications, and other external factors that may affect cortisol levels. The use of a salivette, or
cotton swab-like collection device, may not be appropriate for young children. Because of this,
other methods have been developed for use in this age group. The “passive drool” technique
involves asking a child to spit into a collection tube through a short straw. Plastic syringes have
also been used to suction saliva from the child’s mouth, and other investigators have used cotton
swabs that an adult can hold firmly in order to avoid any potential choking hazard. Occasionally,
researchers have used powdered drink mix, a small amount of sweetener, or sugarless gum to
stimulate saliva flow. Because cortisol assays may be affected by the pH of saliva, it is
important to use stimulants that have been identified by the specific assay’s manufacturer (92).
46

Many of the studies examining HPA axis changes during childhood and adolescence are
designed to examine the relationships between either basal or stress-related cortisol secretion and
mental health concerns, such as depression and anxiety (91; 93–95) . These studies vary in
cortisol sampling methods (i.e. time of day), assessment of chronic and acute stressors, and
evaluation of pubertal status. As a result, it is difficult to draw definitive conclusions regarding
any changes in HPA axis activity in these age groups.
Studies examining normal cortisol levels and responses to stress in children have been
largely cross-sectional in nature and vary in methodology and the ages of children included in
each study. The majority of studies aim to determine whether a gender difference exists prior to
the onset of puberty.
Sex-Related Differences in HPA Axis Activity
In a study of 5 year-old children (91), cortisol was assessed upon wakening and 10-, 20-,
and 30-minutes post-awakening indicated that girls exhibited higher levels of cortisol at all time
points. In addition, girls were more likely to be high-secretors of cortisol and had higher stressinduced levels of cortisol, suggesting that children with higher baseline HPA activity are also
more responsive to stress. Interestingly, no awakening response was found in the boys
participating in this study, which is in contrast to other studies involving children of varying ages
(96). These results are supported by a study of older children (10-12 yrs), in which girls
exhibited higher cortisol levels than boys upon wakening and 30-minutes later, beginning at
pubertal stage 1 (96). An additional study by Tornhage and colleagues (97) is more difficult to
compare to the others because saliva samples were collected between 8:00 and 9:00 AM while
the children were in school, thus missing the waking value and response to waking. Regardless,
these researchers also found a sex difference in basal cortisol levels, though in this case 7 year47

old boys exhibited higher cortisol levels than girls of the same age. The authors suggested that
the higher levels seen in these boys might be related to adrenarche (97), at which time increased
ACTH secretion stimulates release of both cortisol and the adrenal androgens DHEA and
androstenedione (86). Two other studies comparing basal cortisol levels between pre-pubertal
boys and girls have not found any differences between the two sexes (98; 99), though Knutsson
et al. did find a slight sex difference at pubertal stage 2 (girls > boys); however, cortisol in their
study was measured in serum, limiting the ability to compare these results to those of other
investigators. Due to the methodological differences in timing of sampling and the use of
salivary vs. serum cortisol, it is not yet clear whether a gender difference in basal cortisol levels
exists prior to the onset of puberty.
It has been suggested that a sexual dimorphism in HPA axis activity may emerge around
the time of puberty. Most studies of older children and adolescents do support the notion of
gender differences within this age group, though some have not found any differences in basal
cortisol levels (95; 99). Studies by Tornhage (97), Netherton (98), Kelly (100), and Bouma (101)
have found differences between older boys and girls though again, the ability to compare
between studies is limited by methodological differences. In the study by Tornhage et al. (97), in
which saliva samples were collected after the start of the school day, 11- and 14-year-old girls
exhibited greater cortisol levels than boys of the same age. Samples were also collected an
average of 2 hours after waking in the study by Kelly and colleagues (100) and again, girls in
this study exhibited higher cortisol levels compared to boys, though in this case the sex
difference disappeared when samples were collected 30-minutes later. Samples were also
collected at 8:00 AM in Netherton’s study (98), though in this case the timing of sampling with
respect to awakening is unclear. Again, mid-to post-pubertal girls exhibited greater cortisol

48

levels than mid- to post-pubertal boys, such that levels in girls were approximately 20-30%
greater than those in boys (98). The only study to sample cortisol immediately upon waking in
adolescents was that by Bouma and colleagues (101). In this case, girls were again found to
have greater cortisol levels than boys, though the awakening response (determined from samples
collected 30-minutes later) did not differ by gender. It has been suggested that the sex difference
emerging around puberty may be related to interactions between cortisol and sex hormones
secreted by the HPG axis (98), and this is supported by the results of animal studies discussed in
the following section.
Differences by Stages of Pubertal Development. Several of the studies that compared
basal cortisol levels between sexes also compared children and adolescents classified by stages
of pubertal development. Unfortunately, not all studies include an assessment of pubertal
staging, and those that did include staging did not always clearly define the classification scheme
used. The studies by Rosmalen (96), and Netherton (98) used sketches of secondary sex
characteristics described by Tanner and Whitehouse (102) for either self- or parent-assessment of
development, though neither study specified whether criteria were based on pubic hair, breast, or
genital development. Pubertal development was assessed according to Tanner’s criteria for
pubic hair and breast development and by Prader testicular volume in the study by Knutsson et
al. (99), though it is unclear whether this was done by physician- or self-assessment. The studies
by Tornhage et al. (97), Gunnar et al. (95), and Matchock et al. (103) do not provide any detail
regarding their criteria for pubertal staging.
Similarly to the study by Hatzinger discussed previously (91), Rosmalen’s study involved
the collection of a saliva sample immediately upon waking and a second sample 30-minutes
later; a third sample was collected at 8:00 PM (96). In this sample, approximately 85% of
49

children were classified as either pubertal stage 1 or 2 and very few children were sexually
mature. No differences were found according to pubertal stage, such that no change in HPA axis
activity was evident as children became more sexually mature (96). It is important to remember,
however, that the cross-sectional study design limits the ability to draw conclusions on the
changes in HPA axis activity across pubertal development. The majority of children
participating in this study (71%) did exhibit a cortisol awakening response and values tended to
be higher in the summer months than the other months of the year, a difference the authors
suggested might be related to participation in different activities (96).
In addition to comparing subjects by age, Tornhage and colleagues also classified
subjects according to pubertal stage and compared cortisol levels across the stages. The criteria
for pubertal staging are not given in the paper; however, a peak in basal cortisol levels was noted
for both sexes in stages 2 and 3, with levels in stages 4 and 5 declining to prepubertal levels in
boys but remaining elevated in girls (97). Without more information about pubertal staging,
however, it is difficult to compare these results to those of other studies.
An additional study by Netherton and colleagues (98) also included children and
adolescents across the transition to adolescents (8-16 y, mean age 12.8 y). Participants were
classified into groups according to pubertal status similarly to the study by Rosmalen (96). Two
groups were then created, grouping children in stage 1 and stage 2 into a pre- to early-pubertal
group (PEP) and those in or beyond stage 3 in a mid- to post-pubertal group (MPP); the majority
of the sample (approximately 70%) were in the MPP group. Although a significant positive
association was found between age and mean cortisol in girls, when PEP and MPP girls were
compared no differences were found, and the effect of age disappeared. No association was
found between mean cortisol and either age or pubertal status among boys in this study (98).
50

Studies by Matchock et al. (103) and Gunnar et al. (95) also suggest that the sexual
dimorphism evident during adolescence may be dependent upon pubertal status. In the study by
Gunnar and colleagues, 13 year-old girls exhibited a greater cortisol response to a laboratory
stressor than boys of the same age, however, the girls were more pubertally advanced at that age
than the boys (95). Similarly, the study by Matchock and colleagues indicates a later peak in
boys (and an earlier peak in girls) in the later stages of puberty (103); this difference in peak time
was the only sex difference evident in the Matchock study. Unfortunately, neither of these
papers included information regarding classification of pubertal status.
In a study of adolescent girls, Oskis and colleagues (104) used menarcheal status as an
indicator of maturity and compared daily cortisol patterns during the different phases of the
menstrual cycle. Post-menarcheal girls peaked later in the morning (45-min post awakening vs.
30-min post awakening) and had higher cortisol levels over the course of the day; however, this
relationship disappeared after the analyses were controlled for BMI and age. The authors
concluded that pubertal status, age, and BMI are all interrelated variables that should be
statistically controlled in studies of adolescents, highlighting a weakness in other studies cited
here (104). These reported differences in the timing of the morning peak between boys and girls
in later puberty (103; 104) may be one reason for the sexual dimorphism that seems to be
apparent around puberty. It is possible that cortisol sampling may capture the morning peak in
either boys or girls while missing the peak in the other gender; this is further complicated by the
lack of coherence in sampling methods between studies.
All of the studies here are limited by the fact that they include cross-sectional data only.
This makes it extremely difficult to compare across age or pubertal stage, as it is not possible to
know whether any apparent differences come from puberty itself or from individual variation.
51

One study, performed by Knutsson et al. (99) did include longitudinal data. These data help to
clarify the patterns of cortisol secretion across childhood and adolescence; however, cortisol was
measured in serum, making it more difficult to compare to other studies. As discussed
previously, salivary cortisol reflects only free cortisol concentrations. While this is generally
seen as an advantage because it allows for assessment of the most biologically active form of
cortisol, serum measurements capture cortisol circulating free in serum and also that bound to
CBG and albumin (85). Of a total of 235 children and adolescents participating in this study, 28
were followed longitudinally (2-7 repeated measurements over 0.5-8.0 y). Children spent the
night in a hospital and cortisol concentration was measured in serum, first at 1400h and then at
1800h, 2200h, 0200h, 0400h, 0600h, and 1000h, providing a profile of the full daily pattern of
circulating cortisol. The longitudinal analysis revealed stable cortisol patterns across childhood
and adolescence with wide inter-individual variability. Importantly, no differences were found
by pubertal stage in the longitudinal analysis (99).
Animal Studies. Studies of rats indicate a change in the basal activity of the HPA axis
and in the stress response around the time of puberty. As in humans, increased levels of GnRH
are released by the HPG axis, ultimately increasing production and release of testosterone in
males and estradiol and progestin in females. In males, this increase in testosterone decreases
HPA axis stress reactivity while in females, increased release of the female sex hormones
increases stress reactivity (105). Similarly, the work of Carey and colleagues (106) indicates that
exposure to exogenous estrogens increases basal HPA axis activity of adult female rats. This
work was furthered by a 2009 study by Evuarherhe and colleagues (107) in which estrogen
levels were altered in both pre-pubertal and ovariectomized rats. Pre-pubertal rats that received
an injection of estradiol did not differ from those given an injection of cholesterol in terms of

52

basal corticosterone concentrations; however, the pre-pubertal rats exposed to estradiol had
reduced stress reactivity than did those exposed only to cholesterol. Interestingly, when female
rats were ovariectomized prior to puberty and given an injection of estradiol in adulthood, they
had a greater stress response and greater basal activity than did rats that had only a sham surgery.
These results suggest a shift in the response of the HPA axis to estrogens around puberty that is
caused by something other than sex steroids themselves (107). Researchers have suggested that
this reversal may be in response to other changes occurring around the time of puberty, such as
maturation of the HPA axis negative feedback system through changes in glucocorticoid receptor
mRNA expression in the brain (105).
Cortisol and Adiposity
Stress is thought to influence weight gain through several mechanisms. Some of these
are behavioral and emotional effects like “comfort eating” and decreased physical activity, while
others may be related to hormonal changes in energy homeostasis and appetite regulation.
Studies have frequently used socioeconomic status, allostatic load, and cortisol as indicators of
chronic stress.
To date, studies in youth examining relationships between stress and cortisol, using
socioeconomic status as an indicator of chronic stress, have shown varied results. Some of these
studies suggest a positive relationship between stress and cortisol, while others do not.
Methodological variation may explain some of these differences. Additionally, age-related
differences may make it more difficult to reach a clear conclusion (108).
Recent studies have associated health disparities, including increased rates of childhood
overweight and obesity, with socioeconomic status (SES) and indicated that differences in SES

53

may contribute some of the differences seen among ethnicities (109). Children from low SES
families are at a greater risk of becoming overweight (110). Possible contributing factors may
include unsafe neighborhoods and resultant physical inactivity, reduced access to safe places to
play such as community centers and safe parks, grocery stores (111), and reduced school
resources (112). In addition to potentially decreased PA, children from a lower SES background
may experience increased stress (113) and may be more vigilant than their higher-SES peers in
response to potential threats (114). This suggestion is supported by studies in adults indicating
that increased stress is associated with increased adiposity, especially around the abdomen (80;
115).
Physical, psychological, and physiological stressors stimulate cortisol production and
release. Cortisol is a glucocorticoid hormone regulated through negative feedback of the HPA
axis. In response to an appropriate stressor, such as a drop in blood glucose, CRH is secreted
from the hypothalamus, which then causes the release of ACTH from the anterior pituitary.
Increased circulating levels of ACTH then stimulate the release of cortisol from the adrenal
cortex. Cortisol then travels to the body tissues and mediates several physiological effects, such
as the breakdown of triglycerides in adipose tissue to free fatty acids and glycerol. In addition,
cortisol activates lipoprotein lipase release from adipose tissue, which promotes storage of lipids
in adipocytes (80). Circulating levels of cortisol in the blood follow a circadian pattern, reaching
their peak early in the morning, falling in the evening, and rising again towards midnight (116).
Food ingestion, exercise, and psychological stressors additionally modify cortisol release.
Disruption of normal HPA axis activity can lead to changes in the diurnal rhythms
mentioned above, and increased HPA axis activity can inhibit the gonadal and growth hormone
axes, further disposing individuals to visceral fat accumulation (80). This has been supported by
54

research of Cushing’s syndrome, in which it has been shown that the elevated levels of cortisol
associated with Cushing’s syndrome are associated with increased amounts of visceral adipose
tissue (80).
Several studies have examined cortisol levels upon waking, both in blood and saliva,
when cortisol levels are typically near their peak. Recently, more attention has been turned to
associations between cortisol and adiposity in children and adolescents. Two recent studies have
examined the relationships between cortisol levels and adiposity (117; 118). Both studies found
significant positive relationships between serum cortisol and abdominal fat. Additionally, Barat
et al. found whole body fat mass was not correlated with cortisol levels, however, trunk fat was
positively correlated with morning serum cortisol in the total study population (118). Serum
cortisol was associated with insulin resistance in both studies. Weigensberg and colleagues also
found a positive relationship between morning cortisol levels and the number of features of MetS
exhibited (117). Interestingly, when cortisol levels were adjusted for intra-abdominal adipose
tissue (IAAT), the relationships were no longer significant, and cortisol was positively related to
IAAT, indicating that cortisol is related to abdominal adiposity in overweight Latino youth.
Salivary cortisol measures used in the paper by Barat et al. represent free cortisol, and
were not associated with fat distribution (118). In fact, the salivary response to lunch was
negatively correlated with trunk fat, especially in girls. This is in contrast to previous research
by the same group which saw increased or similar salivary responses in obese compared with
normal weight premenopausal women and obese compared to non-obese children (118).
These two studies indicate that there is likely a relationship between circulating cortisol
levels and adiposity, particularly visceral (abdominal) adiposity. The relationship of elevated

55

cortisol to insulin resistance indicated in both studies suggests that overweight and obese
children with elevated cortisol may be at increased risk for T2D, and are already known to be at
greater risk for MetS as demonstrated in the paper by Weigensberg.
In a 2008 paper, Soros and colleagues (119) hypothesized that some obese children may
actually exhibit low cortisol levels (24-hour average), and that this is an adaptive response to the
insulin resistance associated with obesity. If this hypothesis were correct, then obese children
with elevated cortisol levels would be insulin resistant, as indicated by the two previous studies
(117; 118) and at higher risk for development of T2D. This is an interesting hypothesis that may
need to be researched further, as the mechanisms for the reduced or elevated cortisol responses
are unknown (119).
Psychosocial Function and Cortisol
In adults, blunted or low-variability in cortisol patterns throughout the day are associated
with adverse health effects, such as hypertension, increased abdominal deposition of adipose
tissue, and low HDL cholesterol (115). Altered cortisol responses to stress have also been shown
among children and adolescents, and are often associated with several psychosocial triggers (i.e.
bullying) and/or effects (i.e. behavior problems) (75)(120).
Ouellet-Morin et al. (75) examined cortisol responses in a sample of 12 year-old children
with a history of maltreatment. Children were twins selected from a longitudinal study cohort,
and were eligible for inclusion in the sub-study if only one twin had been maltreated in the past.
The Child Behavior Checklist (CBCL) was used to assess behavior problems and was completed
by parents, whereas a separate assessment was completed by teachers (Teacher’s Report Form).
In this study, children who had been maltreated demonstrated blunted cortisol responses to a

56

laboratory stressor (the Psychosocial Stress Test) when compared to non-maltreated children,
and these lower cortisol responses were also found to be associated with increased behavior
problems and social difficulties among maltreated children only. Thus, these results suggest that
early-life maltreatment is associated with altered cortisol responses to stress and also to
psychosocial difficulties later in childhood.
Further research by Gustafsson et al. (74) found an inverse u-shaped relationship between
cumulative stress exposure and diurnal cortisol patterns. In this study, 130 children and
adolescents (mean age 12.8) were again selected from a longitudinal twin study (only one twin
from each pair was included). In this study, the Coddington Life Events Scale was used to assess
psychosocial function; additional information was gathered regarding lifetime exposure to
chronic events (i.e. car accidents, witnessing a violent crime) and family socioeconomic status.
Then, thirty-two items from these questionnaires were combined (1 SES item, 24 negative life
events, 7 potentially tragic events) and counted to create a cumulative adversity “score,” such
that children with greater scores had experienced greater levels of adversity. Based on these
data, it was determined that adversity was significantly related to cortisol concentrations. When
cortisol was examined at individual time-points, cumulative adversity was related to postawakening cortisol (taken 30 minutes after waking) and the slope of the diurnal decline, but not
to immediately upon waking or bedtime levels. Interestingly, children who had experienced
moderate levels of adversity experienced higher 30-minute post-waking cortisol levels, a higher
cortisol awakening response (CAR), and a steeper diurnal decline than children who had low or
high adversity exposure, suggesting that some exposure to adversity causes heightened activity
of the HPA axis, but that there may be a threshold past which HPA axis activity is actually
reduced.

57

Conversely, a study by Adam et al. (73) of older adolescents found that a higher CAR
was related to increased risk of developing major depressive disorder (MDD). This was true
after adjusting for age, gender, health behaviors and health status, as well as various mental
health measures (i.e. current and past mood and anxiety disorders), suggesting that the increased
risk of MDD was likely to be related to the elevated CAR observed at baseline. However, other
cortisol measures, including waking and bedtime levels, were not associated with increased risk
of MDD.
In adults, a 2006 study of newly diagnosed breast cancer patients and healthy community
controls (121) examined the relationship between several measures of HPA axis activity (cortisol
AUC with respect to the ground, cortisol AUC with respect to the increase, early morning peak,
and diurnal rhythm) and measures of psychosocial functioning. Psychosocial distress was
measured using the General Health Questionnaire, which includes questions about general
psychological distress and somatic complaints, similarly to the PSC used in a youth population.
In this study, none of the psychosocial measures mentioned here predicted cortisol release in
cancer patients or controls. Correlations did indicate a positive correlation between cAUC with
respect to the increase (using morning cortisol as the baseline value) and distress among control
women, but not cancer patients.
Dietary Patterns and Hypercortisolemia
Insight on the role of diet in conditions of cortisol excess can be drawn from research on
Cushing’s syndrome and Cushing’s disease. Cushing’s syndrome is a clustering of several
disorders characterized by chronic glucocorticoid excess. Cushing’s disease is the most common
type of Cushing’s syndrome, and is the result of excessive ACTH release caused by a pituitary
tumor. Other types of Cushing’s syndrome can result from treatment with glucocorticoid
58

medications, tumors in tissues other than the pituitary, and adrenal tumors. Cushing’s syndrome
can also be food-dependent, resulting from over-expression of glucose-dependent insulinotropic
polypeptide (GIP), which is characterized by glucocorticoid excess after eating. Negativefeedback inhibition of ACTH release by circulating cortisol is suppressed, ACTH is released
from the pituitary without any regular pattern and the normal circadian rhythmicity in ACTH and
cortisol patterns is absent in patients with Cushing’s syndrome. Thus, the normal decline in
cortisol levels seen overnight does not occur, and Cushing’s syndrome is often diagnosed
following overnight cortisol sampling or an overnight dexamethasone suppression test (82).
Hypercortisolism is present in all types of Cushing’s syndrome, and chronic
overexposure to cortisol results in a number of metabolic and other clinical abnormalities.
Metabolic effects include obesity (particularly central fat distribution), hypertension, glucose
intolerance, hyperlipidemia, and diabetes; other effects include acne, hirsutism, osteopenia,
muscular weakness (especially in the lower extremities), and depression, among others (82).
Management of Cushing’s syndrome generally focuses on resolution of hypercortisolism
through surgical and pharmaceutical interventions, and little information is available on dietary
concerns. However, a Patient Information Publication from the NIH lists a number of guidelines
for Cushing’s patients. These patients need to be careful to reduce sodium intake since they are
prone to hypertension. Additionally, since hypercortisolism tends to lead to decreased bone
mineral density, it is especially important that Cushing’s patients consume adequate amounts of
both calcium and vitamin D; a separate dietary guideline has not been established, but patients
should be sure to consume the recommended intakes for both of these nutrients. Since high
cortisol levels also predispose these individuals to insulin resistance, they should consider
following a diet appropriate for individuals with diabetes, in order to maintain optimal blood
59

sugar and insulin levels (122). Since the DASH diet is effective in managing both hypertension
and insulin resistance, as discussed in a previous section, this may be a reasonable dietary pattern
for Cushing’s patients to follow. Also, the DASH diet emphasizes foods that are also good
sources of calcium and vitamin D (45).
Influence of Physical Activity on Stress and Cortisol
Disruptions in the relationships between food intake, PA, and stress can lead to negative
consequences. In ideal circumstances, increased physical activity has positive effects on stress
coping and appetite regulation. The resulting metabolic and physiological changes work to
reduce the risk of chronic disease. In cases of extreme stress or poor stress coping, however,
individuals may become less active and increase their intake of highly palatable “comfort foods”
that tend to be very calorie dense and offer few beneficial nutrients (4).
To date, only one study has specifically examined the role of PA in the relationship
between stress and obesity in youth (123). In a sample of 303 adolescents (mean age 16.6 years),
Yin and colleagues assessed measures of both personal- and community-level stress and their
relationship with BMI, skinfold thickness, and waist circumference. Personal stress, but not
community stress, was associated with adiposity after controlling for SES, and this relationship
was significantly modified by weekly days of PA. These results suggest a beneficial effect of
PA on the relationship between stress and obesity, although more research is necessary to further
clarify these relationships.
As previously mentioned, exercise of sufficient duration and/or intensity serves as a
strong enough stressor to augment circulating cortisol levels (124). With the onset of exercise,
there is an increase in circulating cortisol levels that is proportional to exercise intensity, which

60

peaks at a level dependent upon the duration of the exercise bout. After exercise training, cortisol
levels at rest may remain unchanged or decrease slightly. Additionally, there is some evidence
that the cortisol response to exercise is diminished after exercise training, such that exercise of
similar intensity results in a attenuated hormonal response after training (125; 126), possibly as a
result of increased clearance from the circulation. In athletes, nighttime cortisol levels are lower
following daytime exercise (127), suggesting that exercise may help to lower elevated cortisol
levels throughout the rest of the day.
Few papers have examined the differences in the cortisol response to exercise between
normal weight and obese children. In a study of 21 lean and obese children, salivary cortisol was
found to be elevated in response to exercise in obese, but not lean, children (128). However, in
this same study, post-exercise cortisol was 32% lower than pre-exercise levels in obese children;
the authors were not able to speculate as to the mechanism for this decrease. Exercise intensity
may be a limitation in this study, as participants exercised for 26 minutes at intensity
approximately equal to 60% of VO2max, which may not have been high enough to elicit a
response in all subjects (129). In light of increased glucocorticoid receptors and glucocorticoid
sensitivity in adipose tissue, this decrease may be representative of increased cortisol uptake by
the tissues.
A second study of obese boys found morning cortisol levels to be decreased following a
12-week aerobic exercise-training program. In this group, decreases in morning cortisol were
accompanied by decreases in weight, leptin, insulin, and LDL-cholesterol, as well as increases in
HDL-cholesterol (130). These positive changes suggest that exercise training may be
instrumental in improving the MetS profile of obese children. It is important to note, however,
that several methodological issues are unclear in this paper. These results are in contrast to those

61

of Ben Ounis and colleagues (131), who found resting cortisol levels to be significantly
increased following 2 months of aerobic exercise training. The variation in results between these
two studies may be explained, in part, by differences in the cortisol sampling procedure. In both
studies, cortisol was measured in serum following a venous blood draw obtained in the morning
(8 AM and between 7 and 8 AM, respectively) both at study enrollment and upon completion of
the exercise-training program (130; 131). Although an attempt was made to control for cortisol
sampling time by sampling at the same time of day, no information was given regarding the time
of sampling in relationship to waking time. Because cortisol levels change rapidly in the hour
following waking, even small variations in sampling with respect to waking time may result in
erroneous conclusions. Due to these methodological issues and general paucity of research,
additional studies are needed examining these relationships and possible mechanisms for
differences between normal weight and obese individuals.
Physical Activity, Stress, and Metabolic Syndrome
Although PA and stress are each independently associated with MetS, only one study has
examined their combined influence in a youth population. Holmes and colleagues assessed the
modifying effects of habitual PA on the relationship between several self-reported measures of
psychosocial stress and a continuous metabolic risk score (132). Among less active youth,
various measures of stress (i.e. school- and sports-related self-esteem, trait anxiety) were
positively associated with metabolic risk, while these associations did not exist in the high PA
group. Other measures of stress, such as depression measured by the Children’s Depression
Inventory, were not associated with metabolic risk. These results are encouraging, because they
suggest that PA can in fact modify the detrimental relationship between stress and MetS.
However, the sample included boys only, and the study did not include a physiological indicator

62

of stress. Thus, additional studies are needed to further examine these relationships and to
identify a biological mechanism for the modifying effect of PA.
Summary and Conclusions
With the alarming increase in pediatric obesity in the last several decades, conditions that
were once thought of as “adult” problems are being seen more frequently in the adolescent
population. It has been reported that up to 63% of adolescents possess at least one feature of
MetS (133), while nearly 50% of severely obese adolescents can be diagnosed with MetS (≥ 3
features). These youth are at increased risk for CVD, T2D, and other obesity-related diseases
later in life. Physical activity and diet are the most common therapeutic targets (25; 134),
however these two factors do not fully account for the recent increase in obesity rates. An
additional factor, garnering more attention recently, is psychosocial stress and dysregulation of
the HPA axis, resulting in altered daily cortisol patterns. Increased cortisol levels, either in the
morning or throughout the day, have been associated with the accumulation of visceral adipose
tissue, elevated blood pressure, insulin resistance, and MetS as a whole. It is hypothesized that
psychosocial stress contributes to chronically elevated cortisol, and that this increase in cortisol
is the biological mechanism through which psychosocial stress is associated with obesity and
MetS. Of course, it is also important to remember that obesity and MetS can exist independently
of each other (i.e., an individual can exhibit some or most of the features of MetS but not be
obese). Therefore, it is hypothesized that psychosocial stress contributes to chronically elevated
cortisol, and that this increase in cortisol is the biological mechanism through which
psychosocial stress is associated with obesity and MetS, beyond the energy balance issues
accounted for by PA and diet.

63

Physical activity has been shown to be an effective stress-management tool, and studies
in youth have shown that PA can moderate the relationship between psychosocial stress and
adiposity. It is less clear, however, whether PA can alter the relationship between stress and
MetS. Studies of athletes and obese youth have suggested that exercise training can lower
morning cortisol levels (127; 130); however, another study in youth found that exercise training
significantly increased resting cortisol levels (131). Methodological differences involving
cortisol sample timing, assessment of psychosocial stress, and measurement of PA may explain
some of these differences. Additionally, only one study in youth has examined the moderating
effects of PA on the relationship between stress and MetS (132). Thus, additional studies are
necessary to gain a better understanding of the role of PA in the management of stress and MetS
in an adolescent population. Because weight loss is a long-term process, identifying the role of
PA in the management of stress and MetS can be used to design interventions that maximize the
stress-management benefits of PA. The result may be short-term health improvements, even in
the absence of weight-loss.

64

REFERENCES

65

REFERENCES
1.

Power C, Lake JK, Cole TJ. Measurement and long-term health risks of child and
adolescent fatness. Int J Obes Relat Metab Disord. 1997 Jul;21(7):507-26.

2.

Wilks DC, Besson H, Lindroos a K, Ekelund U. Objectively measured physical activity
and obesity prevention in children, adolescents and adults: a systematic review of
prospective studies. Obesity Rev. 2011 May;12(5):e119-29.

3.

Steele RM, Brage S, Corder K, Wareham NJ, Ekelund U. Physical activity,
cardiorespiratory fitness, and the metabolic syndrome in youth. J Appl Physiol. 2008
Jul;105(1):342-51.

4.

De Vriendt T, Moreno L a, De Henauw S. Chronic stress and obesity in adolescents:
scientific evidence and methodological issues for epidemiological research. Nutr Metab
Cardiovasc Dis. 2009 Sep;19(7):511-9.

5.

Holmes ME, Ekkekakis P, Eisenmann JC. The physical activity, stress and metabolic
syndrome triangle: a guide to unfamiliar territory for the obesity researcher. Obes Rev.
2010 Jul;11(7):492-507.

6.

Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of obesity and trends in body
mass index among US children and adolescents, 1999-2010. JAMA. 2012
Feb;307(5):483-90.

7.

Koebnick C, Smith N, Coleman KJ, Getahun D, Reynolds K, Quinn VP, Porter AH, DerSarkissian JK, Jacobsen SJ. Prevalence of extreme obesity in a multiethnic cohort of
children and adolescents. J Pediatr. 2010 Jul;157(1):26-31.e2.

8.

Singh a S, Mulder C, Twisk JWR, van Mechelen W, Chinapaw MJM. Tracking of
childhood overweight into adulthood: a systematic review of the literature. Obes Rev.
2008 Sep;9(5):474-88

9.

Malina RM, Koziel S, Bielicki T. Variation in Subcutaneous Adipose Tissue Distribution
Associated With Age, Sex, and Maturation. Am J Hum Biol.1999;200(May 1998):189200.

10.

Baumgartner R, Roche AF. Tracking of fat pattern indices in childhood: The Melbourne
Growth Study. Hum Biol 1988;60(4):549.

11.

Reilly JJ. Descriptive epidemiology and health consequences of childhood obesity. Best
Pract Res Clin Endocrinol Metab. 2005 Sep;19(3):327-41.

12.

Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity. Int J
Obes Relat Metab Disord. 1999 Mar;23 Suppl 2:S2-11.

66

13.

Braet C, Mervielde I. Psychological Aspects of Childhood Obesity: A Controlled Study in
a Clinical and Nonclinical Sample. J Pediatr Psychol. 1997;22(1):59-71.

14.

Lauer RM, Connor WE, Leaverton PE, Reiter M a, Clarke WR. Coronary heart disease
risk factors in school children: the Muscatine study. J Pediatr. 1975 May;86(5):697-706.

15.

McGill HC, McMahan C a, Malcom GT, Oalmann MC, Strong JP. Relation of
glycohemoglobin and adiposity to atherosclerosis in youth. Pathobiological Determinants
of Atherosclerosis in Youth (PDAY) Research Group. Arterioscler Thromb Vasc Biol.
1995 Apr;15(4):431-40.Available from: http://www.ncbi.nlm.nih.gov/pubmed/7749853

16.

Chen W, Srinivasan SR, Li S, Xu J, Berenson GS. Clustering of long-term trends in
metabolic syndrome variables from childhood to adulthood in Blacks and Whites: the
Bogalusa Heart Study. Am J Epidemiol. 2007 Sep;166(5):527-33.

17.

Bao W, Srinivasan SR, Wattigney WA, Berenson GS. Persistance of multiple
cardiovascular risk clustering related to Syndrome X from childhood to young adulthood:
the Bogalusa Heart Study. Arch Intern Med. 1994;154:1842-1847.

18.

Ford ES. Risks for all-cause mortality, cardiovascular disease, and diabetes associated
with the metabolic syndrome: a summary of the evidence. Diabetes Care. 2005
Jul;28(7):1769-78.

19.

Huang PL. A comprehensive definition for metabolic syndrome. Dis Model Mech.
2009;2(5-6):231-7.

20.

Cook S, Weitzman M, Auinger P, Nguyen M, Dietz WH. Prevalence of a Metabolic
Syndrome Phenotype in Adolescents Findings From the Third National Health and
Nutrition Examination Survey, 1988-1994. Arch Pediatr Adolesc Med. 2003;157:19881994.

21.

Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW, Allen K, Lopes
M, Savoye M, Morrison J, Sherwin RS, Caprio S. Obesity and the metabolic syndrome in
children and adolescents. N Engl J Med. 2004 Jun;350(23):2362-74.

22.

Duncan GE, Li SM, Zhou X-H. Prevalence and trends of a metabolic syndrome phenotype
among u.s. Adolescents, 1999-2000. Diabetes Care. 2004 Oct;27(10):2438-43.

23.

Eisenmann JC. On the use of a continuous metabolic syndrome score in pediatric research.
Cardiovasc Diabetol. 2008 Jan;7:17.

24.

Cook S, Auinger P, Huang TT-K. Growth curves for cardio-metabolic risk factors in
children and adolescents. J Pediatr. 2009 Sep;155(3):S6.e15-26.

25.

Keith SW, Redden DT, Katzmarzyk PT, Boggiano MM, Hanlon EC, Benca RM, Ruden
D, Pietrobelli a, Barger JL, Fontaine KR, Wang C, Aronne LJ, Wright SM, Baskin M,

67

Dhurandhar NV, Lijoi MC, Grilo CM, DeLuca M, Westfall a O, Allison DB. Putative
contributors to the secular increase in obesity: exploring the roads less traveled. Int J Obes
(Lond). 2006 Nov;30(11):1585-94.
26.

Ekelund U, Anderssen SA, Froberg K, Sardinha LB, Andersen LB, Brage S. Independent
associations of physical activity and cardiorespiratory fitness with metabolic risk factors
in children: the European youth heart study. Diabetologia. 2007 Sep;50(9):1832-40.

27.

Pan Y, Pratt CA. Metabolic syndrome and its association with diet and physical activity in
US adolescents. J Am Diet Assoc. 2008 Feb;108(2):276-86.

28.

Martínez-Gómez D, Eisenmann JC, Moya JM, Gómez-Martínez S, Marcos A, Veiga OL.
The role of physical activity and fitness on the metabolic syndrome in adolescents: effect
of different scores. The AFINOS Study. J Physiol Biochem. 2009 Sep;65(3):277-89.

29.

Nguyen THHD, Tang HK, Kelly P, van der Ploeg HP, Dibley MJ. Association between
physical activity and metabolic syndrome: a cross sectional survey in adolescents in Ho
Chi Minh City, Vietnam. BMC Public Health. 2010 Jan;10:141.

30.

Moore JB, Davis CL, Baxter SD, Lewis RD, Yin Z. Physical activity, metabolic
syndrome, and overweight in rural youth. J Rural Health. 2008 Jan;24(2):136-42.

31.

Rizzo NS, Ruiz JR, Hurtig-Wennlöf A, Ortega FB, Sjöström M. Relationship of physical
activity, fitness, and fatness with clustered metabolic risk in children and adolescents: the
European youth heart study. J Pediatr. 2007 Apr;150(4):388-94.

32.

Ondrak KS, McMurray RG, Bangdiwala SI, Harrell JS. Influence of aerobic power and
percent body fat on cardiovascular disease risk in youth. J Adolesc Health. 2007
Aug;41(2):146-52.

33.

Katzmarzyk P, Malina R, Bouchard C. Physical Activity, Physical Fitness, and Coronary
Heart Disease Risk Factors in Youth: The Quebec Family Study. Prev Med.
1999;562:555-562.

34.

McMurray RG, Ainsworth BE, Harrell JS, Griggs TR, Williams OD. Is physical activity
or aerobic power more influential on reducing cardiovascular disease risk factors? Med
Sci Sports Exerc. 1998 Oct;30(10):1521-9.

35.

Lobelo F, Pate RR, Dowda M, Liese AD, Daniels SR. Cardiorespiratory fitness and
clustered cardiovascular disease risk in U.S. adolescents. J Adolesc Health. 2010
Oct;47(4):352-9

36.

Troiano RP, Briefel RR, Carroll MD, Bialostosky K. Energy and fat intakes of children
and adolescents in the united states: data from the national health and nutrition
examination surveys. Am J Clin Nutr. 2000 Nov;72(5 Suppl):1343S-1353S.

68

37.

Deshmukh-Taskar PR, Nicklas TA, O’Neil CE, Keast DR, Radcliffe JD, Cho S. The
relationship of breakfast skipping and type of breakfast consumption with nutrient intake
and weight status in children and adolescents: the National Health and Nutrition
Examination Survey 1999-2006. J Am Diet Assoc. 2010 Jun;110(6):869-78.

38.

Carlson JJ, Eisenmann JC, Norman GJ, Ortiz KA, Young PC. Dietary fiber and nutrient
density are inversely associated with the metabolic syndrome in US adolescents. J Am
Diet Assoc.2011 Nov;111(11):1688-95.

39.

Lytle LA, Himes JH, Feldman H, Zive M, Dwyer J, Hoelscher D, Webber L, Yang M.
Nutrient intake over time in a multi-ethnic sample of youth. Public Health Nutr.2002
Apr;5(2):319-28.

40.

Stang J, Story MT, Harnack L, Neumark-Sztainer D. Relationships between vitamin and
mineral supplement use, dietary intake, and dietary adequacy among adolescents. J Am
Diet Assoc.2000 Aug;100(8):905-10.

41.

Sugiyama T, Xie D, Graham-Maar RC, Inoue K, Kobayashi Y, Stettler N. Dietary and
lifestyle factors associated with blood pressure among U.S. adolescents. J Adolesc
Health.Feb;40(2):166-72.

42.

Volpe SL, Schall JI, Gallagher PR, Stallings VA, Bergqvist AGC. Nutrient intake of
children with intractable epilepsy compared with healthy children. J Am Diet Assoc.2007
Jun;107(6):1014-8.

43.

Moore LL, Singer MR, Qureshi MM, Bradlee ML. Dairy Intake and Anthropometric
Measures of Body Fat among Children and Adolescents in NHANES. J Am Coll Nutr.
2008 Dec;27(6):702–710.

44.

Bradlee ML, Singer MR, Qureshi MM, Moore LL. Food group intake and central obesity
among children and adolescents in the Third National Health and Nutrition Examination
Survey (NHANES III). Public Health Nutr. 2010 Jun;13(6):797-805.

45.

Appel LJ, Moore TJ, Obarzanek E, Vollmer WM, Svetkey LP, Sacks FM, Bray GA, Vogt
TM, Cutler JA, Windhauser MM, Lin PH, Karanja N. A clinical trial of the effects of
dietary patterns on blood pressure. DASH Collaborative Research Group. N Engl J Med.
1997 Apr;336(16):1117-24.

46.

Azadbakht L, Mirmiran P, Esmaillzadeh A, Azizi T, Azizi F. Beneficial effects of a
Dietary Approaches to Stop Hypertension eating plan on features of the metabolic
syndrome. Diabetes Care. 2005 Dec;28(12):2823-31.

47.

Barnard ND, Katcher HI, Jenkins DJA, Cohen J, Turner-McGrievy G. Vegetarian and
vegan diets in type 2 diabetes management. Nutr Rev. 2009 May;67(5):255-63.

69

48.

Ludwig DS, Pereira MA, Kroenke CH, Hilner JE, Van Horn L, Slattery ML, Jacobs DR.
Dietary fiber, weight gain, and cardiovascular disease risk factors in young adults. JAMA.
1999 Oct;282(16):1539-46.

49.

Pereira MA, Jacobs DR, Van Horn L, Slattery ML, Kartashov AI, Ludwig DS. Dairy
consumption, obesity, and the insulin resistance syndrome in young adults: the CARDIA
Study. JAMA. 2002 Apr;287(16):2081-9.

50.

Ventura E, Davis J, Byrd-Williams C, Alexander K, McClain A, Lane CJ, Spruijt-Metz D,
Weigensberg M, Goran M. Reduction in risk factors for type 2 diabetes mellitus in
response to a low-sugar, high-fiber dietary intervention in overweight Latino adolescents.
Arch Pediatr Adolesc Med. 2009 Apr;163(4):320-7.

51.

McPherson RS, Hoelscher DM, Alexander M, Scanlon KS, Serdula MK. Dietary
Assessment Methods among School-Aged Children: Validity and Reliability. Prev Med.
2000 Aug;31(2):S11-S33.

52.

Marquart LF, Cohen EA, Short SH. Nutrition Knowledge of Athletes and Their Coaches
and Surveys of Dietary Intake. In: Wolinsky I, editor. Nutrition in Exercise and Sport.
Boca Raton, FL: CRC Press LLC; 1998 p. 559-595.

53.

Whitney E, Rolfes S. Understanding Nutrition. 9th ed. Belmont, CA: Wadsworth; 2002.

54.

Nelson M, Black AE, Morris JA, Cole TJ. Between- and within-subject variation in
nutrient intake from infancy to old age: estimating the number of days required to rank
dietary intakes with desired precision. Am J Clin Nutr. 1989 Jul;50(1):155-67.

55.

Kennedy ET, Ohls J, Carlson S, Fleming K. The Healthy Eating Index: design and
applications. J Am Diet Assoc. 1995 Oct;95(10):1103-8.

56.

Guenther PM, Reedy J, Krebs-Smith SM. Development of the Healthy Eating Index-2005.
J Am Diet Assoc.2008 Nov;108(11):1896-901.

57.

McEwen BS, Seeman T. Protective and damaging effects of mediators of stress.
Elaborating and testing the concepts of allostasis and allostatic load. Ann N Y Acad Sci.
1999 Jan;896:30-47.

58.

Cannon W. Bodily changes in pain, hunger, fear, and rage: an account of recent researches
into the function of emotional excitement. New York: Appleton; 1920.

59.

Selye H. Stress and the general adaptation syndrome. Br Med J.1950;1(4667):1383-1392.

60.

Lazarus R. Stress, appraisal, and coping. New York: Springer Pub. Co.; 1984.

70

61.

Steptoe A, Feldman PJ. Neighborhood problems as sources of chronic stress: development
of a measure of neighborhood problems, and associations with socioeconomic status and
health. Ann Behav Med Jan;23(3):177-85.

62.

Seidman E, Allen L, Aber JL, Mitchell C, Feinman J, Yoshikawa H, Comtois KA, Golz J,
Miller RL, Ortiz-Torres B. Development and validation of adolescent-perceived
microsystem scales: social support, daily hassles, and involvement. Am J Community
Psychol 1995 Jun;23(3):355-88.

63.

Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc
Behav. 1983 Dec;24(4):385-96.

64.

Byrne DG, Davenport SC, Mazanov J. Profiles of adolescent stress: the development of
the adolescent stress questionnaire (ASQ). J Adolec. 2007 Jun;30(3):393-416.

65.

Jellinek MS, Murphy JM, Little M, Pagano ME, Comer DM, Kelleher KJ. Use of the
Pediatric Symptom Checklist to Screen for Psychosocial Problems in Pediatric Primary
Care A National Feasibility Study. Arch Pediatr Adolesc Med. 1999 Nov;153:254-260.

66.

Gardner W, Kelleher KJ, Pajer K a. Multidimensional adaptive testing for mental health
problems in primary care. Med Care. 2002 Sep;40(9):812-23.

67.

Stoppelbein L, Greening L, Jordan SS, Elkin TD, Moll G, Pullen J. Factor Analysis of the
Pediatric Symptom Checklist with a Chronically Ill Pediatric Population. J Dev Behav
Pediatr. 2005;26(5)

68.

Reed-Knight B, Hayutin LG, Lewis JD, Blount RL. Factor structure of the pediatric
symptom checklist with a pediatric gastroenterology sample. J Clin Psychol Med
Settings. 2011 Sep;18(3):299-306.

69.

Boothroyd RA, Armstrong M. An Examination of the Psychometric Properties of the
Pediatric Symptom Checklist With Children Enrolled in Medicaid. J Emot Behav Disord.
2009 Sep;18(2):113-126.

70.

Harris T, Brown G. Social origins of depression: A study of psychiatric disorder in
women. New York: Free Press; 1978.

71.

D’Amico D, Barbarito C. Health & Physical Assessment in Nursing. Prentice Hall; 2006.

72.

Porter JS, Bean MK, Gerke CK, Stern M. Psychosocial factors and perspectives on weight
gain and barriers to weight loss among adolescents enrolled in obesity treatment. J Clin
Psychol Med Settings. 2010 Jun;17(2):98-102.

73.

Adam EK, Doane LD, Zinbarg RE, Mineka S, Craske MG, Griffith JW. Prospective
prediction of major depressive disorder from cortisol awakening responses in adolescence.
Psychoneuroendocrinology 2010;35(6):921-931.

71

74.

Gustafsson PE, Anckarsa H, Lichtenstein P, Nelson N, Gustafsson PA. Does quantity
have a quality all its own ? Cumulative adversity and up- and down-regulation of circadian
salivary cortisol levels in healthy children. Psychoneuroendocrinology 2010;35:14101415.

75.

Ouellet-morin I, Odgers CL, Danese A, Bowes L, Shakoor S, Papadopoulos AS, Caspi A,
Moffitt TE, Arseneault L. Blunted cortisol responses to stress signal social and behavioral
problems among maltreated/bullied 12-year-old children. Biol Psychiatry.
2011;70(11):1016-1023.

76.

Taylor LM, Spence JC, Raine K, Plotnikoff RC, Vallance JK, Sharma AM. Physical
activity and health-related quality of life in individuals with prediabetes. Diabetes Res
Clin Pract. 2010;90(1):15-21.

77.

Nadeau K, D M, Kolotkin RL, Ph D, Boex R, Witten T, D R, Mcfann KK, Zeitler P,
Walders-abramson N. Health-Related Quality of Life in Adolescents With Comorbidities
Related to Obesity. J Adolesc Health. 2011;49(1):90-92.

78.

Pressman SD, Matthews KA, Cohen S, Martire LM, Scheier M, Baum A, Schulz R.
Association of Enjoyable Leisure Activities With Psychological and Physical Well-Being.
Psychosom Med.2009;71(7):725-732.

79.

McEwen BS. Protective and damaging effects of stress mediators: the good and bad sides
of the response to stress. Metabolism.2002 Jun;51(6 Suppl 1):2-4.

80.

Björntorp P. Do stress reactions cause abdominal obesity and comorbidities? Obes Rev.
2001 May;2(2):73-86.

81.

Chrousos GP, Kino T, Charmandari E. Evaluation of the hypothalamic-pituitary-adrenal
axis function in childhood and adolescence. Neuroimmunomodulation. 2009
Jan;16(5):272-83.

82.

Greenspan F, Gardner D. Basic and Clinical Endocrinology. 6th ed. New York: McGrawHill; 2004.

83.

Rask E. Tissue-Specific Dysregulation of Cortisol Metabolism in Human Obesity. J Clin
Endocrinol Metab. 2001 Mar;86(3):1418-1421.

84.

Wiegand S, Richardt A, Remer T, Wudy S a, Tomlinson JW, Hughes B, Grüters A,
Stewart PM, Strasburger CJ, Quinkler M. Reduced 11beta-hydroxysteroid dehydrogenase
type 1 activity in obese boys. Eur J Endocrinol. 2007 Sep;157(3):319-24.

85.

Gozansky WS, Lynn JS, Laudenslager ML, Kohrt WM. Salivary cortisol determined by
enzyme immunoassay is preferable to serum total cortisol for assessment of dynamic
hypothalamic--pituitary--adrenal axis activity. Clincal Endocrinol (Oxf). 2005
Sep;63(3):336-41.

72

86.

Ellison P. Puberty. In: Cameron N, editor. Human Growth and Development. San Diego:
Academic Press; 2002 p. 65-84.

87.

Roemmich JN, Rogol AD. Hormonal changes during puberty and their relationship to fat
distribution. Am J Hum Biol. 1999 Jan;11(2):209-224.

88.

Gunnar M, White B. Salivary cortisol measures in infant and child assessment. In: Singer
L, Zeskind P, editors. Biobehavioral Assessment of the Infant. New York: Guilford Press;
2001 p. 167-189.

89.

Larson MC, White BP, Cochran A, Donzella B, Gunnar M. Dampening of the cortisol
response to handling at 3 months in human infants and its relation to sleep, circadian
cortisol activity, and behavioral distress. Dev Psychobiol.1998 Dec;33(4):327-37.

90.

Goldberg S, Levitan R, Leung E, Masellis M, Basile VS, Nemeroff CB, Atkinson L.
Cortisol concentrations in 12- to 18-month-old infants: stability over time, location, and
stressor. Biol Psychiatry.2003 Oct;54(7):719-26.

91.

Hatzinger M, Brand S, Perren S, von Wyl A, von Klitzing K, Holsboer-Trachsler E.
Hypothalamic-pituitary-adrenocortical (HPA) activity in kindergarten children:
importance of gender and associations with behavioral/emotional difficulties. J Psychiatr
Res. 2007 Nov;41(10):861-70.

92.

Hanrahan K, McCarthy AM, Kleiber C, Lutgendorf S, Tsalikian E. Strategies for salivary
cortisol collection and analysis in research with children. Appl Nurs Res. 2006
May;19(2):95-101.

93.

Cutuli J, Wiik K, Herbers J, Gunnar M. Cortisol Function Among Early School-aged
Homeless Children. Psychoneuroendocrinology. 2010;35(6):612-625.

94.

Gustafsson PE, Gustafsson PA, Nelson N. Cortisol levels and psychosocial factors in
preadolescent children. Stress Health. 2006;9(September 2005):3-9.

95.

Gunnar MR, Wewerka S, Frenn K, Long JD, Griggs C. Developmental changes in
hypothalamus-pituitary-adrenal activity over the transition to adolescence: normative
changes and associations with puberty. Dev Psychopathol. 2009 Jan;21(1):69-85.

96.

Rosmalen JGM, Oldehinkel a J, Ormel J, de Winter a F, Buitelaar JK, Verhulst FC.
Determinants of salivary cortisol levels in 10-12 year old children; a population-based
study of individual differences. Psychoneuroendocrinology. 2005 Jun;30(5):483-95.

97.

Törnhage CJ. Reference values for morning salivary cortisol concentrations in healthy
school-aged children. J Pediatr Endocrinol Metab. 2002 Feb;15(2):197-204.

98.

Netherton C. Salivary cortisol and dehydroepiandrosterone in relation to puberty and
gender. Psychoneuroendocrinology. 2004 Feb;29(2):125-140.

73

99.

Knutsson U, Dahlgren J, Marcus C, Rosberg S, Brönnegård M, Stierna P, AlbertssonWikland K. Circadian cortisol rhythms in healthy boys and girls: relationship with age,
growth, body composition, and pubertal development. J Clin Endocrinol Metab. 1997
Feb;82(2):536-40.

100. Kelly SJ, Young R, Sweeting H, Fischer JE, West P. Levels and confounders of morning
cortisol collected from adolescents in a naturalistic (school) setting.
Psychoneuroendocrinology. 2008 Oct;33(9):1257-68.
101. Bouma EMC, Riese H, Ormel J, Verhulst FC, Oldehinkel AJ. Adolescents’ cortisol
responses to awakening and social stress; effects of gender, menstrual phase and oral
contraceptives. The TRAILS study. Psychoneuroendocrinology. 2009 Jul;34(6):884-93.
102. Tanner JM. Growth at adolescence. 2nd ed. Oxford: Blackwell Scientific Publications;
1962.
103. Matchock RL, Dorn LD, Susman EJ. Diurnal and seasonal cortisol, testosterone, and
DHEA rhythms in boys and girls during puberty. Chronobiol Int. 2007 Jan;24(5):969-90.
104. Oskis A, Loveday C, Hucklebridge F, Thorn L, Clow A. Diurnal patterns of salivary
cortisol across the adolescent period in healthy females. Psychoneuroendocrinology. 2009
Apr;34(3):307-16.
105. Romeo RD. Pubertal maturation and programming of hypothalamic-pituitary-adrenal
reactivity. Front Neuroendocrinol. 2010 Apr;31(2):232-40.
106. Carey MP, Deterd CH, de Koning J, Helmerhorst F, de Kloet ER. The influence of
ovarian steroids on hypothalamic-pituitary-adrenal regulation in the female rat. J
Endocrinol. 1995 Feb;144(2):311-21.
107. Evuarherhe O, Leggett J, Waite E, Kershaw Y, Lightman S. Reversal of the hypothalamopituitary-adrenal response to oestrogens around puberty. J Endocrinol. 2009
Aug;202(2):279-85.
108. Dowd JB, Simanek AM, Aiello AE. Socio-economic status, cortisol and allostatic load: a
review of the literature. Int J Epidemiol. 2009 Oct;38(5):1297-309.
109. Adler NE, Rehkopf DH. U.S. disparities in health: descriptions, causes, and mechanisms.
Ann Rev Public Health. 2008 Jan;29:235-52.
110. Shrewsbury V, Wardle J. Socioeconomic status and adiposity in childhood: a systematic
review of cross-sectional studies 1990-2005. Obesity (Silver Spring). 2008 Feb;16(2):27584
111. Lovasi GS, Hutson MA, Guerra M, Neckerman KM. Built environments and obesity in
disadvantaged populations. Epidemiol Rev. 2009 Jan;31(4):7-20.

74

112. Scott MM, Cohen DA, Evenson KR, Elder J, Catellier D, Ashwood JS, Overton A.
Weekend schoolyard accessibility, physical activity, and obesity: the Trial of Activity in
Adolescent Girls (TAAG) study. Prev Med. 2007 May;44(5):398-403.
113. Lupien SJ, King S, Meaney MJ, McEwen BS. Can poverty get under your skin? Basal
cortisol levels and cognitive function in children from low and high socioeconomic status.
Dev Psychopathol. 2001 Jan;13(3):653-76.
114. Chen E, Langer DA, Raphaelson YE, Matthews KA. Socioeconomic status and health in
adolescents: the role of stress interpretations. Child Dev. 2004;75(4):1039-52.
115. Rosmond R, Dallman MF, Björntorp P. Stress-related cortisol secretion in men:
relationships with abdominal obesity and endocrine, metabolic and hemodynamic
abnormalities. J Clin Endocrinol Metab. 1998 Jun;83(6):1853-9.
116. Kanaley JA, Weltman JY, Pieper KS, Weltman A, Hartman ML. Cortisol and growth
hormone responses to exercise at different times of day. J Clin Endocrinol Metab. 2001
Jun;86(6):2881-9.
117. Weigensberg MJ, Toledo-Corral CM, Goran MI. Association between the metabolic
syndrome and serum cortisol in overweight Latino youth. J Clin Endocrinol Metab. 2008
Apr;93(4):1372-8.
118. Barat P, Gayard-Cros M, Andrew R, Corcuff J-B, Jouret B, Barthe N, Perez P, Germain
C, Tauber M, Walker BR, Mormede P, Duclos M. Truncal distribution of fat mass,
metabolic profile and hypothalamic-pituitary adrenal axis activity in prepubertal obese
children. J Pediatr. 2007 May;150(5):535-9, 539.e1.
119. Soros A, Zadik Z, Chalew S. Adaptive and maladaptive cortisol responses to pediatric
obesity. Med Hypotheses. 2008 Sep;71(3):394-8.
120. Janicke DM, Marciel KK, Ingerski LM, Novoa W, Lowry KW, Sallinen BJ, Silverstein
JH, David M, Lisa M, Janet H. Impact of Psychosocial Factors on Quality of Life in
Overweight Youth. Obesity (Silver Spring). 2007;15(7):1799–1807.
121. Vedhara K, Tuinstra J, Stra JT, Miles JNV, Sanderman R, Ranchor AV. Psychosocial
factors associated with indices of cortisol production in women with breast cancer and
controls. Psychoneuroendocrinology. 2006 Apr;31(3):299-311.
122. National Insitutes of Health. Nutrition for Patients with Cushing Syndrome. Bethesda:
1997.
123. Yin Z, Davis CL, Moore JB, Treiber F a. Physical activity buffers the effects of chronic
stress on adiposity in youth. Ann Behav Med. 2005 Feb;29(1):29-36.

75

124. McMurray RG, Hackney A. Endocrine responses to exercise and training. In: Garrett W,
Kirkendall D, editors. Exercise and Sport Science. Philadelphia: Lippincot Williams &
Wilkins; 2000 p. 135-161.
125. Hackney AC, Sinning WE, Bruot BC. Reproductive hormonal profiles of endurancetrained and untrained males. Med Sci Sports Exerc. 1988 Feb;20(1):60-5.
126. De Souza MJ, Arce JC, Pescatello LS, Scherzer HS, Luciano AA. Gonadal hormones and
semen quality in male runners. A volume threshold effect of endurance training. Int J
Sports Med. 1994 Oct;15(7):383-91.
127. Hackney A, Viru A. Twenty-four-hour cortisol response to multiple daily exercise
sessions of moderate and high intensity. Clin Physiol. 1999 Mar;19(2):178-82.
128. Hershberger a M, McCammon MR, Garry JP, Mahar MT, Hickner RC. Responses of
lipolysis and salivary cortisol to food intake and physical activity in lean and obese
children. J Clin Endocrinol Metab. 2004 Sep;89(9):4701-7.
129. Hill EE, Zack E, Battaglini C, Viru M, Viru a, Hackney a C. Exercise and circulating
cortisol levels: the intensity threshold effect. J Endocrinol Invest. 2008 Jul;31(7):587-91.
130. Karacabey K. The effect of exercise on leptin, insulin, cortisol and lipid profiles in obese
children. J Int Med Res. 2009;37(5):1472-8.
131. Ben Ounis O, Elloumi M, Zouhal H, Makni E. Effect of an individualized physical
training program on resting cortisol and growth hormone levels and fat oxidation during
exercise in obese children. Ann Endocrinol (Paris).2011 Feb;72(1):34-41.
132. Holmes M, Eisenmann J, Ekkekakis P, Gentile D. Physical activity, stress, and metabolic
risk score in 8-to 18-year-old boys. J Phys Act Health. 2008;(5):294-307.
133. de Ferranti SD, Gauvreau K, Ludwig DS, Neufeld EJ, Newburger JW, Rifai N. Prevalence
of the metabolic syndrome in American adolescents: findings from the Third National
Health and Nutrition Examination Survey. Circulation. 2004 Oct;110(16):2494-7.
134. Barlow SE. Expert committee recommendations regarding the prevention, assessment, and
treatment of child and adolescent overweight and obesity: summary report. Pediatrics.
2007 Dec;120 Suppl:S164-92.
135. Ortiz-Andrellucchi A, Henríquez-Sánchez P, Sánchez-Villegas A, Peña-Quintana L,
Mendez M, Serra-Majem L. Dietary assessment methods for micronutrient intake in
infants, children and adolescents: a systematic review. Br J Nutr. 2009 Dec;102 Suppl
1:S87-117.

76

CHAPTER 3: METHODOLOGY

Specific Aims and Hypotheses
Specific aim 1: Determine the relationship between daily cortisol levels and the continuous
metabolic syndrome score (cMetS) in obese adolescents. It was hypothesized that cortisol area
under the curve (cAUC) would be positively related to cMetS.
Specific aim 2: Determine if habitual physical activity (MVPA and/or steps/day) attenuates the
relationship between daily cortisol and cMetS in obese adolescents. It was hypothesized that
there would be a significant interaction of PA and cAUC, such that PA would attenuate the
relationship between cAUC and cMetS.
Specific Aim #3: Determine the relationship between psychosocial functioning score and cAUC
in obese adolescents. It was hypothesized that psychosocial functioning would be positively
related to cAUC.
Specific Aim #4: Determine the relationship between psychosocial functioning score and cMetS
in obese adolescents. It was hypothesized that psychosocial functioning would be positively
related to cMetS.
Specific aim 5: Determine if habitual physical activity (MVPA and/or steps/day) attenuates the
relationship between psychosocial functioning score and cMetS in obese adolescents. It was
hypothesized that there would be a significant interaction of PA and psychosocial functioning,
such that PA would attenuate the relationship between psychosocial functioning and cMetS.

77

Methods
Study design and subjects.
The study design was cross-sectional. Subjects between the ages of 12-18 years were
recruited from The Healthy Weight Center (HWC; a pediatric obesity clinic) and Academic
General Pediatrics (AGP) clinic at Helen DeVos Children’s Hospital in Grand Rapids, MI.
Patient charts were screened by the investigator for body mass index (BMI), and those with a
th

BMI ≥95 age- and sex-specific percentile of CDC growth chart were identified for recruitment.
Eligible patients were approached by the investigator at the time of their clinic visit and invited
to participate at that time. A sample size of 50 adolescents was included due to budgetary
constraints (see power analysis at the end of this chapter). Each subject gave assent and parental
informed consent was obtained before study participation. The study protocol was approved by
Institutional Review Boards (IRB) at Michigan State University and Spectrum Health. Data
collection was completed between January and June of 2012.
Assessment of the main outcome – metabolic syndrome.
Height and body mass were measured according to standard procedures (1). Height was
measured in duplicate to the nearest 0.1 cm using a wall-mounted stadiometer (Harpenden, Great
Britain). Body mass was measured in duplicate to the nearest 0.1 kg using an electronic scale
(Scaltronix, White Plains, NY). Body mass index (BMI) was then calculated by dividing body
2

mass in kg by height in m and BMI percentile was determined according to the 2000 CDC
growth charts (2). Waist circumference (WC) was measured to the nearest 0.1 cm at the level of
the superior border of the iliac crest using a Gullick tape. Blood pressure was measured after 10
minutes of seated rest. Two were taken on the right arm with an automated sphygmomanometer

78

(Datascope, Accutorr Plus). Trained medical staff in each clinic obtained all physical
measurements, and values were reviewed by the researcher for consistency and repeated if
necessary. It was not possible for all measures to be obtained by the same person because doing
so would have interrupted clinic flow at the AGP clinic. For adults, good agreement has been
shown between clinic-based systolic blood pressure measures and standard measures taken by
trained research assistants (ICC=0.91), while agreement between clinic-based and standard
measures for diastolic blood pressure is somewhat lower (ICC=0.77)(3). Biochemical markers of
MetS were determined by blood work completed routinely as part of clinic visits, including
fasting measures of cholesterol (total, HDL, and LDL), triglycerides, glucose and insulin.
Metabolic syndrome score was derived by regressing subject values for each MetS variable onto
th

the age- and sex-specific 50 percentile defined by Cook and colleagues (4). Because Cook’s
data do not provide information regarding normative fasting blood glucose values, means and
SDs reported by Jolliffe (5) were used to calculate the glucose z-scores. Resultant z-scores for
triglycerides, HDL cholesterol (multiplied by -1 to account for the inverse relationship between
HDL and CVD risk), waist circumference, glucose, and systolic blood pressure were then
summed to produce the continuous MetS score (cMetS) as previously described (6).
Daily Cortisol Measures.
Salivary cortisol was collected during the course of one day. The daily cortisol profile has
been shown to remain stable for individuals over two days of measurement (7)(8), so values from
one day were expected to be a good indicator of day-to-day cortisol patterns in children. Salivary
cortisol has been shown to be highly correlated with serum measures (9; 10), while mixed results
have been found for day-to-day reliability, varying based on the time of day each sample was
taken and the time between each measurement occasion (interclass r=0.18-0.78) (11). Each

79

subject received saliva collection supplies and detailed instructions. Subjects were asked to
collect saliva samples by placing an oral swab under the tongue for one to two minutes, and then
placing the swab in a storage tube. As soon as possible after collection, the subject was to place
the sample in his/her home freezer until it was returned to the investigator by mail. Cortisol
samples are stable unfrozen for 5-7 days, and therefore freezing and then shipping by standard
mail is acceptable (12). Samples were collected at 6 time points: 1) immediately upon waking, 2)
30 minutes after waking, and 3) four additional samples taken at three-hour intervals
synchronized to waking time (e.g., 7:00 AM, 7:30 AM, 10:00 AM, 1:00 PM, 4:00 PM, and 7:00
PM). A graphical representation of sampling times is shown in Figure 3.1. Subjects were
instructed to avoid eating a major meal 60 minutes before sample collection, and to thoroughly
rinse the mouth 10 minutes prior to collection. Each tube was pre-labeled with a sample number,
and subjects were directed to record the time of sampling on a paper log.
To ensure compliance with the sample collection protocol, the investigator
communicated with subjects at specific times on the sampling day via text message and/or
telephone. If the family desired, the investigator called the subject night before sample
collection to inquire about his or her planned waking time and to remind the subject to take the
first sample immediately upon waking. The following morning, the investigator sent a text
message to the subject or called the specified phone number at the pre-arranged waking time.
This message served as a reminder to collect the sample, as well as an opportunity to verify that
the sample was collected at the appropriate time. If the sample was not collected at the correct
time, the investigator clarified that the subject knew when the next sample should be taken, and
the incorrect sample was not included in the analysis. Each collection tube was be labeled with a
code (i.e. blue circle), which the subject uncovered at the time of collection. The subject was

80

directed to reply to the investigator’s text message or call with the appropriate code and time of
sample collection. For each sample that was collected at the appropriate time, the investigator
added $1.00 to the subject’s final incentive amount. Previous studies that have employed similar
sampling protocols have included little data regarding compliance. Oskis et al. followed a
similar text message reminder protocol and reported that all subjects responded to the wake-time
reminder message, but did not provide any additional information regarding the number of
participants who completed all saliva collection (8). Kudielka et al. used a very similar sampling
protocol (samples at waking, 30-minutes post-waking, 11 AM, 3 PM, 8 PM, and 10 PM). In
their study, 74% complied with sampling instructions (monitored via an electronic device) while
26% failed to take at least one sample at the correct time, and most who missed one sample also
missed at least one more (13).
Once received, samples were stored at -80°C in a secure freezer at Spectrum Health.
Upon the completion of data collection, all samples were transferred on ice to a laboratory at
Michigan State University for analysis. Upon arrival at the laboratory, samples were allowed to
warm to room temperature, centrifuged for 15 minutes at 3,000 RPM, and clear saliva was
separated from the precipitants. Samples were analyzed in duplicate by enzyme-linked
immunoassay (ELISA) using a high sensitivity salivary cortisol kit (Salimetrics; State College,
PA). Subjects were required to collect a minimum of 3 samples (immediately upon waking, 30minutes after waking, and the final evening time point) to be included in the analysis. Cortisol
AUC was calculated according to the following formula:

 T1+ T 2 
 T 2 + T 3   T 3+ T 4   T 4 + T 5   T 5 + T 6 
AUC = 0.5
 + 2.5
 + 3
 + 3
 + 3

 2 
 2   2   2   2 

81

where T1=cortisol sample 1, T2=cortisol sample 2, T3=cortisol sample 3, T4=cortisol sample 4,
T5=cortisol sample 5, and T6=cortisol sample 6.
Figure 3.1. Example timeline for saliva sampling.
7:00 AM

7:30 AM

10:00 AM

1:00 PM

4:00 PM

7:00 PM

Wake up

Habitual physical activity.
Subjects were asked to wear the SenseWear Pro III armband (SWA, Body Media,
Pittsburgh, PA) for 7 consecutive days to assess habitual physical activity. The SWA has been
shown to be valid in healthy adults (14), children (15–17), and obese adults (18) and children
(19). The armband is worn over the right triceps and uses sensors to measure galvanic skin
response, skin temperature, heat flux, and near-body temperature; a 2-axis accelerometer is also
contained in the unit to measure movement (17). Additionally, subjects were asked to wear the
SWA to bed on two nights during the data collection period (one the night before saliva sample
collection). Following the 7-day period, activity data were downloaded and analyzed for
habitual physical activity (total energy expenditure; time spent in moderate-, and vigorous PA,
and steps/day) using version 7.0 of the manufacturer’s software.
In order to be included in the PA analysis, subjects were required to have at least 4 days
(3 weekdays and one weekend day) of data, each day including at least 10 hours of wear time
during waking hours. To insure compliance, the minimum standard of armband wear was
included in the requirements to be eligible for study incentives.

82

Psychosocial Functioning
Psychosocial functioning was evaluated using the Pediatric Symptoms Checklist (PSC)
(20). The PSC is a 35-item questionnaire designed for use by pediatric practitioners to detect
psychosocial problems among their patients. For each item, the child indicated whether the
statement describes the child never, sometimes, or often. Responses coded as “never” received a
score of 0 points, “sometimes” received a score of 1 point, and “often” received a score of 2
points, and these scores were summed to compute a total score. Recent factor analyses
completed with healthy children in pediatric primary care identified subscales indicative of
internalizing and externalizing difficulties (21). Although these factor analyses were completed
using the parent report form of the PSC, the developers of the questionnaire communicated that
the scales should be applicable to the youth report, as well (personal communication). For the
full PSC, as well as the internalizing and externalizing sub-scores, a higher score is indicative of
greater psychological dysfunction. For the youth report of the PSC, a total score ≥30 is said to
indicate impaired psychosocial functioning. For the purposes of this study, the total score on the
PSC was used as an indicator of psychosocial function.
Pubertal Stage
Pubertal stage was physician-assessed at the time of the clinic visit (first visit at the
HWC/well child visit at AGP). Physicians use Tanner’s pubic hair, breast, and genital criteria
(22). Of these, pubic hair is generally considered to be the most accurate criterion for use with
obese adolescents (23). Pubertal staging was used as a covariate in data analysis.
Dietary assessment.

83

Habitual dietary patterns were assessed using a 3-day diet record, which is included in
standard assessments at the HWC; subjects recruited from AGP were given detailed instructions
and asked to record all of their daily food intake on 3 days (2 days of their choice and the day of
cortisol sampling) during the week-long physical activity monitoring period. Subjects from AGP
were asked to return the diet records with their samples and SWA, while diet records for HWC
patients were retrieved from their medical record. Diet data were then analyzed and the healthy
eating index (HEI) was calculated according to Guenther et al. (24). In short, 10 components are
included in the HEI (the five major food groups, total dietary fat, saturated fat intake, cholesterol
intake, sodium intake, and dietary variety); each of these 10 components is scored on a range
from 0-10, so that subjects receive a diet quality score ranging from 0-100, with a higher score
indicating superior dietary quality. Scores for the major food groups are based on the age- and
sex-specific total daily serving recommendations for each individual. The two components
dealing with fat intake, components 6 and 7, are scored based on daily recommendations for each
nutrient, such that individuals consuming less than the recommended daily intake receive a score
of 10. Components 8 and 9, cholesterol and sodium, are based on milligrams consumed per day.
Finally, dietary variety is assessed by counting foods consumed at a minimum of one-half
serving in each food group and equivalents are added to determine how many different foods
each individual consumed. A perfect score of 10 is assigned to this category if an individual
consumes at least 16 different foods over the 3-day study period. The HEI was included to be
used as a covariate in cortisol analyses.
Data analysis.
Descriptive statistics were computed for physical characteristics and individual
components of MetS, habitual physical activity, psychosocial stress score (obtained via the PSC),
84

as well as cortisol AUC and at each individual time point. For continuous variables, differences
between boys and girls and between subjects recruited from the HWC and AGP clinics were
evaluated using an independent samples t-test. Differences in categorical variables between boys
and girls and between subjects recruited from HWC and AGP were evaluated using chi-square
analysis.
Specific aim 1: Determine the relationship between daily cortisol levels and the continuous
metabolic syndrome score (cMetS) in obese adolescents.
Linear regression, with and without controlling for pubertal stage, age and sex, was used to
examine the association between cortisol AUC and cMetS. Additionally, partial correlation,
again controlling for pubertal stage, age, and sex, was used to examine the relationship between
cortisol at specific time points (waking, 30-minutes post-waking, and 12-hours post-waking) and
individual components of MetS (lipids, blood pressure, waist circumference and BMI, glucose,
and insulin).
Specific aim 2: Determine if habitual physical activity (MVPA and/or steps/day)
attenuates the relationship between daily cortisol and cMetS in obese adolescents.
Possible effect modification of PA level was determined using multiple regression.
Predictors entered into the model included cortisol AUC and MVPA (min/day). The interaction
between cAUC and MVPA was examined first, with and without controlling for pubertal stage,
sex, and age. This technique was then repeated in order to examine the interaction between
cAUC and steps/day on cMetS. Differences in the resultant regression equations from aim 1 and
aim 2 were examined using an F test.

85

Specific Aim #3: Determine the relationship between psychosocial functioning score and
cAUC in obese adolescents.
Partial correlation, controlling for pubertal stage, age, and sex, was used to examine the
relationship between psychosocial function score (PSC total score, internalizing score, and
externalizing score) and cortisol AUC. The same technique was used to examine the relationship
between psychosocial function and cortisol at specific time points (i.e. waking, 30-minutes postwaking), for the total score and internalizing and externalizing sub-scores.
Specific Aim #4: Determine the relationship between psychosocial functioning score
and cMetS in obese adolescents.
Linear regression, with and without controlling for pubertal stage, sex, and age was
performed to examine the relationship between psychosocial function score and cMetS.
Additionally, partial correlation, again controlling for pubertal stage, age, and sex, was used to
examine the relationship between PSC scores (total, internalizing, and externalizing) and
individual components of MetS.
Specific aim 5: Determine if habitual physical activity (MVPA and/or steps/day)
attenuates the relationship between psychosocial functioning and cMetS in obese
adolescents.
Possible effect modification of PA level was determined using multiple regression.
Predictors entered into the model included total score from the PSC and MVPA (min/day). The
interaction between PSC score and MVPA was examined first, with and without controlling for
pubertal stage, sex, and age. This procedure was then repeated using PSC total score and

86

steps/day as predictors Differences in the resultant regression equations from aim 4 and aim 5
were examined using an F test.
An alpha level of 0.05 was used for statistical significance. A priori power analyses
identified a projected sample size of 68 adolescents in order to obtain a statistical power level of
0.80 with a Type I error rate of 5% (α=0.05) and a medium effect size. Due to budgetary
constraints, 50 adolescents were included in this study. The expected statistical power based on
this sample size was 0.76 for specific aims 1 and 2. All statistics were analyzed using SPSS
version 20.0.

87

REFERENCES

88

REFERENCES
1.

Malina RM, Bouchard C, Bar-Or O. Growth, Maturation, and Physical Activity. 2nd ed.
Champaign, IL: Human Kinetics; 2004.

2.

Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, Mei Z, Wei R,
Curtin LR, Roche AF, Johnson CL. 2000 CDC Growth Charts for the United States:
methods and development. Vital Health Stat. 2002 May;(246):1-190.

3.

Kim JW, Bosworth HB, Voils CI, Olsen M, Dudley T, Gribbin M, Adams M, Oddone EZ.
How well do clinic-based blood pressure measurements agree with the mercury standard?
J Gen Intern Med. 2005 Jul;20(7):647–9.

4.

Cook S, Auinger P, Huang TT-K. Growth curves for cardio-metabolic risk factors in
children and adolescents. J Pediatr. 2009 Sep;155(3):S6.e15–26.

5.

Jolliffe CJ, Janssen I. Development of age-specific adolescent metabolic syndrome criteria
that are linked to the Adult Treatment Panel III and International Diabetes Federation
criteria. J Am Coll Cardiol. 2007 Feb;49(8):891–8.

6.

Eisenmann JC. On the use of a continuous metabolic syndrome score in pediatric research.
Cardiovasc Diabetol. 2008 Jan;7:17.

7.

Edwards S, Clow a, Evans P, Hucklebridge F. Exploration of the awakening cortisol
response in relation to diurnal cortisol secretory activity. Life Sci. 2001
Mar;68(18):2093–103.

8.

Oskis A, Loveday C, Hucklebridge F, Thorn L, Clow A. Diurnal patterns of salivary
cortisol across the adolescent period in healthy females. Psychoneuroendocrinology. 2009
Apr;34(3):307–16.

9.

Gozansky WS, Lynn JS, Laudenslager ML, Kohrt WM. Salivary cortisol determined by
enzyme immunoassay is preferable to serum total cortisol for assessment of dynamic
hypothalamic--pituitary--adrenal axis activity. Clincal Endocrinol (Oxf). 2005
Sep;63(3):336–41.

10.

Cortisol Testing in Saliva & Salivary Cortisol Research - Salivary Assay - Salivary
Cortisol – Salimetrics. Available from: http://www.salimetrics.com/my-spitresearch/analytes/cortisol.php

11.

Golden SH, Wand GS, Malhotra S, Kamel I, Horton K. Reliability of hypothalamicpituitary-adrenal axis assessment methods for use in population-based studies. Eur J
Epidemiol. 2011 Jul;26(7):511–25.

89

12.

Hanrahan K, McCarthy AM, Kleiber C, Lutgendorf S, Tsalikian E. Strategies for salivary
cortisol collection and analysis in research with children. Appl Nurs Res. 2006
May;19(2):95–101.

13.

Kudielka BM, Broderick JE, Kirschbaum C. Compliance with saliva sampling protocols:
electronic monitoring reveals invalid cortisol daytime profiles in noncompliant subjects.
Psychosom Med. 2003 Mar;65(2):313–9.

14.

Malavolti M, Pietrobelli A, Dugoni M, Poli M, Romagnoli E, De Cristofaro P, Battistini
NC. A new device for measuring resting energy expenditure (REE) in healthy subjects.
Nutr Metab Cardiovasc Dis. 2007 Jun;17(5):338–43.

15.

Arvidsson D, Slinde F, Larsson S, Hulthén L. Energy cost in children assessed by
multisensor activity monitors. Med Sci Sports Exerc. 2009 Mar;41(3):603–11.

16.

Dorminy C a, Choi L, Akohoue S a, Chen KY, Buchowski MS. Validity of a multisensor
armband in estimating 24-h energy expenditure in children. Med Sci Sports Exerc. 2008
Apr;40(4):699–706.

17.

Calabró M a, Welk GJ, Eisenmann JC. Validation of the SenseWear Pro Armband
algorithms in children. Med Sci Sports Exerc. 2009 Sep;41(9):1714–20.

18.

Papazoglou D, Augello G, Tagliaferri M, Savia G, Marzullo P, Maltezos E, Liuzzi A.
Evaluation of a multisensor armband in estimating energy expenditure in obese
individuals. Obesity (Silver Spring). 2006 Dec;14(12):2217–23.

19.

Bäcklund C, Sundelin G, Larsson C. Validity of armband measuring energy expenditure
in overweight and obese children. Med Sci Sports Exerc. 2010 Jun;42(6):1154–61

20.

Jellinek MS, Murphy JM, Little M, Pagano ME, Comer DM, Kelleher KJ. Use of the
Pediatric Symptom Checklist to Screen for Psychosocial Problems in Pediatric Primary
Care A National Feasibility Study. Arch Pediatr Adolesc Med. 1999 Nov;153:254–260.

21.

Gardner W, Kelleher KJ, Pajer K a. Multidimensional adaptive testing for mental health
problems in primary care. Med Care. 2002 Sep;40(9):812–23.

22.

Tanner JM. Growth at adolescence. 2nd ed. Oxford: Blackwell Scientific Publications;
1962.

23.

Denzer C, Weibel a, Muche R, Karges B, Sorgo W, Wabitsch M. Pubertal development in
obese children and adolescents. Int J Obes (Lond). 2007 Oct;31(10):1509–19.

24.

Guenther PM, Reedy J, Krebs-Smith SM. Development of the Healthy Eating Index-2005.
J Am Diet Assoc. 2008 Nov;108(11):1896–901.

90

CHAPTER 4: RESULTS

Subject Characteristics
Patients at the Healthy Weight Center (HWC) and Academic General Pediatrics
(AGP) clinics were pre-screened for study eligibility according to the inclusion and
exclusion criteria outlined previously. A total of 110 were pre-determined to be eligible
based on BMI percentile and medical history free of known HPA axis disorder; 50 of
these 110 agreed to participate (15 boys, 35 girls; see Figure 4.1). Of these 50 subjects,
32 completed saliva collection, 32 completed physical activity monitoring, 41 had
complete glucose and lipid panels, and 38 returned a completed Pediatric Symptom
Checklist (PSC). Incomplete data were distributed as follows: 11 did not return any
saliva samples while 7 did not collect all required samples; 7 armbands were not returned
while 11 did not have sufficient wear time; 9 did not fulfill physician orders for lab work;
12 did not return the PSC. A total of 22 had complete data for anthropometry, PA,
cortisol, metabolic syndrome (MetS), and the PSC (see Table 4.1 for further explanation
of the distribution of measures). There were no significant differences for anthropometry,
cortisol, physical activity, or metabolic syndrome variables between those with and
without complete data.
Descriptive characteristics of the sample are shown in Table 4.2. In general, boys
and girls were similar in terms of anthropometric variables with boys being significantly
taller in terms of absolute values, but not when expressed as age- and sex-specific
percentiles. Subjects recruited from the HWC were heavier than those from the AGP
clinic (105.8 ± 28.1 kg vs. 88.4 ± 21.5 kg, respectively) and had a greater BMI (38.6 ±

91

2

2

7.2 kg/m vs. 34.0 kg/m , respectively). In the total sample, height approximated the
th

th

49 percentile while weight approximated the 98 percentile. Eighty-eight percent of
th

th

participants had a BMI ≥ the 97 percentile, and 46% of BMIs exceeded the 99

percentile. The distribution of ratings of secondary sex characteristics is shown in Table
4.3. The proportion of subjects in each stage of sexual maturation did not vary between
HWC and AGP clinics. It should be noted that there were not any girls classified in stage
1 for either pubic hair or breast development, which makes sense based on the mean age
of girls in this study (14.5 ± 2.0 y); similarly, it stands to reason that 8-9% (n=1 to n=2)
boys would be in stage 1 for either pubic hair or genital development.
Metabolic Syndrome
Mean values for the 42 subjects who completed the fasting lipid panel and
glucose assessments are shown in Table 4.4. Boys and girls did not differ on any
biochemical variables, waist circumference, or blood pressure. However, slight
differences existed for SBP (but not DBP or MAP) between subjects from the two clinics
(HWC: 123.2 ± 14.3 mmHg vs. AGP: 116.1 ± 8.1 mmHg; p= 0.06), as well as for WC
(HWC: 122.0 ± 16.5 cm vs. AGP: 99.3 ± 13.7 cm; p < 0.0001). Subjects were most likely
to meet MetS criteria for WC (88%), followed by HDL (64.3%) and BP (28%), while TG
(26.2%) and glucose (4.8%) were least likely to be adverse (see Figure 4.2). Adolescents
from the HWC were slightly more likely to have an elevated waist circumference (100%
vs. 81%, respectively) and a low HDL (76% vs. 56%) than AGP adolescents, though
these differences were not statistically significant (p>0.05). In total, 33.3% of the sample

92

Figure 4.1. Recruitment flowchart.

Total Eligible
n=110
(47 male, 63 female)

Healthy Weight Center
(n=31)
(10 male, 21 female)

Enrolled
n=19 (61.2%)
(6 male, 13 female)

Not enrolled
n=12 (38.7%)
(4 male, 8 female)

Academic General Pediatrics
(n=79)
(37 male, 42 female)

Enrolled
n=31 (39.2%)
(10 male, 21 female)

Not enrolled
n=48 (60.8%)
(27 male, 21 female)

Reasons for non-enrollment: HWC

Reasons for non-enrollment: AGP

No show: n=3 (25.0%)
Cancellation: n=2 (16.7%)
Unable to see: n=0 (0.0%)
Refused: n=3 (25.0%)

No show: n=23 (47.9%)
Cancellation: n=12 (25.0%)
Unable to see: n=4 (8.3%)
Refused: n=9 (18.8%)
Excluded: n=0 (0.0%)
93

Table 4.1. Distribution of subjects who completed data collection for study variables.
N (%)
N (%)
Anthropometry and labs
Anthropometry, labs, and saliva
Anthropometry, labs, saliva, PA
All

One measure
4 (8.0)
Anthropometry
50 (100.0)

Two measures
5 (10.0)
Labs
41 (82.0)
41 (82.0)

Three measures
6 (12.0)
Saliva
32 (64.0)

Four measures
14 (28.0)
Physical Activity
31 (62.0)

Five measures
22 (44.0)
PSC
38 (76.0)

28 (56.0)
22 (44.0)
22 (44.0)

94

(n=14 of 42) met the criteria for MetS. While the prevalence of MetS was slightly higher for
adolescents seen at the HWC (41% vs. 28%), this difference was not statistically significant
(p>0.05). Additionally, cMetS was positively correlated with BMI z-score (r=0.55, p<0.0001).
The mean cMetS score was 4.16 ± 4.30, and did not differ by clinic or sex.

Table 4.2. Descriptive characteristics of the sample. Values are mean (SD) unless otherwise
indicated.
Male
(n=15)
15.3 (1.7)

Female
(n=35)
14.5 (2.0)

Total
(n=50)
14.8 (1.9)

Min-Max

Height (cm)

169.6 (12.1)*

158.9 (7.0)

162.1 (10.0)

142.2-186.1

Height %ile

54.5 (27.1)

47.0 (28.1)

49.3 (27.8)

6.5-97.2

Weight (kg)

105.1 (33.3)

90.7 (20.2)

95.0 (25.4)

57.2-172.0

Weight %ile

97.8 (3.3)

97.4 (3.7)

97.5 (3.6)

82.7-99.9

BMI (kg/m )

2

35.7 (7.4)

35.7 (6.9)

35.7 (7.0)

25.8-53.8

BMI percentile

98.8 (1.0)

98.2 (2.0)

98.4 (1.8)

89.7-99.9

93.3%

85.7%

88%

--

Age (y)

% severe obesity
th
(BMI ≥ 97 )

12.3-18.9

* denotes statistically significant difference (p<0.05)

Table 4.3. Percent of subjects in each stage of pubertal development.

Male (%)

1
8.3

Pubic Hair
2
3
4
16.7 33.3 20.0

Female (%)

0.0

17.9

25.0

42.9

5
13.3
11.4

95

1
9.1

Breast/Genital
2
3
4
9.1
36.4 27.3

5
18.2

0.0

10.0

16.7

36.7

36.7

Table 4.4. Features of the Metabolic Syndrome within the sample. Values are mean (SD) unless
otherwise indicated.
Male

Female

Total

Min-Max

N

(n)
Waist circumference (cm) 110.9 (22.2) 106.6 (16.7) 107.9 (18.4) 76.5-154.1

50

Total cholesterol (mg/dl)

42

HDL-cholesterol (mg/dl)

155.2 (26.3) 157.3 (24.8) 156.7 (24.9) 94.0-212.0
42.5 (9.6)

46.3 (10.0)

45.0 (9.9)

30.0-71.0

42

Glucose (mg/dl)

112.8 (75.3)

86.2 (5.6)

93.8 (41.1)

76.0-122.0

42

Triglycerides (mg/dl)

148.8 (98.5) 112.6 (66.3) 122.9 (77.4) 37.0-394.0

42

SBP (mmHg)

121.9 (15.8)

117.5 (9.3)

50

DBP (mmHg)

64.5 (10.9)

67.2 (7.6)

66.4 (8.7)

47.0-86.0

50

MAP (mmHg)

83.5 (11.7)

83.4 (7.4)

83.4 (8.8)

64.0-109.2

50

118.8 (11.6) 98.0-155.7

Figure 4.2. Distribution of metabolic syndrome features.
100.0
Total

Percent meeting MetS criteria

90.0

Males

Females

80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
WC

HDL

BP

TG

GLU

Figure legend: WC = waist circumference; HDL = high density lipoprotein cholesterol; BP =
blood pressure (elevated SBP and/or elevated DBP); TG = triglyceride; GLU = fasting blood
glucose (the prevalence of adverse GLU in females was 0.0%). Prevalence of MetS = 33.3%
(males = 41.7%, females = 30.0%)

96

Habitual Physical Activity
Of the 50 subjects, 31 (62%) met the minimum wear-time criteria of 4 days for at least 10
h/day. Mean values for MVPA, steps per day, and TDEE are shown in Table 4.5. Subjects
recruited from AGP were more active than subjects from the HWC when measured by steps/day,
but no statistically significant differences were found between the sexes or clinics for MVPA or
TDEE. Subjects with and without MetS were similar in terms of TDEE and MVPA; however,
subjects without MetS took significantly more steps per day than those with MetS (9460 ± 2880
steps/day vs. 7053 ± 2637 steps/day, respectively).
Table 4.5. Physical activity and total daily energy expenditure within the sample. Values are
mean (SD) unless otherwise noted.
Male

Female

Total

(n=8)

(n=24)

(n=32)

MVPA
(min/day)

54.0 (22.4)

43.6 (26.2)

46.2 (25.3)

4.5-109.0

Steps/day

7774 (2479)

8673 (3223)

8441 (3035)

2858-16,010

3365.7 (858.8)

2766.8 (714.0)

2921.4 (785.4)

1876.1-4956.0

30.1 (4.8)

29.8 (6.8)

29.9 (6.3)

19.1-54.0

TDEE
(kcal/day)
TDEE
(kcal/kg/day)

Min-Max

Cortisol
A total of 39 subjects returned saliva samples, and 32 (64%) of the 50 subjects met the
criteria for inclusion in the cortisol analyses. Of these 32 subjects, 5 correctly obtained six
samples, 9 correctly obtained five, 10 correctly obtained four, and 8 correctly obtained three.
Inclusion criteria for the cortisol analysis required collection of at least the first, second, and
sixth time points so that cAUC could be calculated. Figure 3 displays the mean cortisol levels at
each sampling time. Mean cAUC was 1.337 ± 0.867 µg/dl (1.180 ± 0.753 µg/dl and 1.408 ±

97

0.922 µg/dl for boys and girls, respectively) and did not differ by sex or clinic. Subjects with
and without MetS were similar in terms of cAUC. Sixty-nine percent of the sample (n=22)
exhibited an increase in cortisol after waking, when defined as any increase in cortisol between
the waking and 30-min post sample. The intra-assay and inter-assay coefficients of variation
were 2.7% and 13.8%, respectively.
Figure 4.3. Mean cortisol level at each sampling time.
0.5

Mean

Males

Females

Cortisol (µg/dl)

0.4

0.3

0.2

0.1

0
0

2

4

6
8
Time since waking (h)

10

12

Pediatric Symptom Checklist
Completed PSCs were returned by 38 (76%) of the 50 subjects. Of these subjects, 23%
(n=8) were classified as having impaired psychosocial functioning (i.e., total score ≥ 30). There
was no difference in total scores between sexes or clinics. Additionally, internalizing and
externalizing scores did not differ by sex or clinic. Mean scores by sex are presented in Table 6.

98

The prevalence of impairment did not differ by clinic, but was slightly higher in girls than in
boys (24% vs. 20%, respectively; p = 0.50).
Table 4.6. Pediatric Symptom Checklist. Values are mean (SD) unless otherwise noted.
Male

Female

Total

Min-Max

(n=11)

(n=27)

(n=38)

Total score

22.5 (11.3)

23.9 (11.2)

23.5 (11.0)

1-45

Internalizing

18.2 (7.6)

20.4 (10.2)

19.8 (9.5)

1-43

Externalizing

6.6 (2.3)

7.1 (4.3)

7.0 (3.8)

0-17

Impaired (n(%))

2 (18.2)

7 (25.9)

9 (23.7)

--

Three-Day Diet Record
A total of 38 of the 50 subjects returned diet records. Records were frequently missing
serving sizes or sufficient detail to determine precisely what was eaten (i.e., “sandwich”,
“noodles”), and follow-up was not completed. Therefore, it was not possible to calculate the
healthy eating index (HEI) or analyze any dietary variables.
Specific Aim 1:
Specific aim 1 was to determine the relationship between the daily cortisol pattern,
expressed as cAUC, and cMetS in obese adolescents. Regression analyses, first without
controlling for potential covariates and second controlling for pubertal stage (pubic hair), sex,
and age, are shown in Table 7. No significant relationship was found between cAUC and cMetS
2

(R =0.001, p=0.91), and this relationship remained nonsignificant when pubertal stage, sex, and
2

age were added to the model (R =0.113, p=0.66). Figure 4.4 illustrates the relationship between
cAUC and cMetS in this sample.

99

Partial correlations between cAUC and individual components of MetS revealed a
significant positive relationship between cAUC and insulin levels (r=0.59, p=0.01) when
controlling for pubic hair stage, and sex. No significant relationships were found between cAUC
and triglycerides, HDL cholesterol, waist circumference, or blood pressure measures. When
examined by specific time points, cortisol sample 1 was positively related to DBP (r=0.49, p =
0.04) with a trend towards significance for MAP (r=0.43, p=0.07), while cortisol sample 2 was
positively correlated with insulin levels (r=0.58, p=0.01). Cortisol sample 6 was not associated
with any components of MetS.
Table 4.7. Independent and combined associations of cAUC and steps/day with cMetS.
2

2

F

P

-0.045

0.013

0.91

0.113

-0.074

0.606

0.66

1

0.029

-0.022

0.565

0.46

2

0.146

-0.067

0.686

0.61

1

0.045

-0.005

0.902

0.35

2

0.173

-0.034

0.838

0.52

R
cAUC
cAUC

Adjusted R

1

0.001

2

cAUC * steps/day
cAUC * steps/day
cAUC * MVPA
cAUC * MVPA
1
2

unadjusted model
adjusted for age, pubertal stage (pubic hair), and sex

100

Figure 4.4. Relationship between cAUC and cMetS.

Continuous Metabolic Syndrome Score
(cMetS)

9
8
7
6
5
4
3
2
1
0
-1
-2
-3
0

1

2

3

4

5

Cortisol Area Under the Curve (cAUC)

Specific Aim 2:
Specific aim 2 was to determine if PA attenuates the relationship between daily cortisol
and cMetS, first without adjusting for potential covariates and then including pubertal stage, sex,
and age as covariates. Neither the uncontrolled nor controlled models including steps/day were
2

2

significant (R =0.029, p=0.46 and R =0.146, p=0.61, respectively). These regression analyses
were then repeated including MVPA in the model rather than steps/day, and similar results were
2

2

found (uncontrolled R =0.045, p=0.35; controlled R =0.173, p=0.52).

Specific Aim 3:
Specific aim 3 was to determine the relationship between PSC score and cAUC. Partial
correlation, controlling for pubertal stage, age, and sex, indicated a significant inverse
101

relationship between PSC total score and cAUC (r=-0.45, p=0.04). Additional analysis of
specific time points (sample 1, sample 2, and sample 6) yielded a significant inverse relationship
between PSC total score and cortisol sample 2 (r=-0.54, p=0.01). No relationship was found
between cortisol sample 1 or cortisol sample 6 and PSC total score. Partial correlations, again
controlling for pubertal stage, age and sex, indicated a moderate inverse relationship between the
PSC internalizing score and cAUC (r=-0.48, p=0.03) and cortisol sample 2 (r=-0.57, p=0.01),
while there was a trend toward significance for cortisol sample 1 (r=-0.39, p=0.08). The PSC
externalizing score was inversely related to cortisol sample 2 (r=-0.50, p=0.02) with a trend
toward an inverse relationship with cAUC (r=-0.43, p=0.05).
Specific Aim 4:
Specific aim 4 was to determine the relationship between PSC score and cMetS. Simiarly
to aim 1, this was completed using regression analyses, first without controlling for potential
covariates and second controlling for pubertal stage (pubic hair), sex, and age, as shown in Table
2

8. No significant relationship was found between PSC score and cMetS (R =0.000, p=0.10), and
this relationship remained nonsignificant when pubertal stage, sex, and age were added to the
2

model (R =0.111, p=0.69).

102

Table 4.8. Independent association of PSC total score and combined PSC total score + steps/day
with cMetS.
2

2

F

P

-0.048

0.000

0.10

0.111

-0.086

0.564

0.69

1

0.049

0.003

1.077

0.31

PSC total * steps/day

2

0.164

-0.022

0.880

0.50

PSC total * MVPA1

0.053

0.008

1.176

0.29

2

0.193

0.013

1.074

0.40

R

Adjusted R

1

0.000

2

PSC total
PSC total

PSC total * steps/day

PSC total * MVPA
1
2

unadjusted model
adjusted for age, pubertal stage (pubic hair), and sex

Specific Aim 5:
Specific aim 5 was to determine if PA attenuates the relationship between PSC score and
cMetS. Results of the regression examining the interaction of PSC total score and steps/day on
cMetS are shown in Table 4.8. As shown, the interaction term resulted in a stronger relationship
than when the PSC total score was examined alone, however this remained non-significant
2

2

(uncontrolled R =0.049, p=0.31 and controlled R =0.164, p=0.49). These regression analyses
were then repeated including MVPA in the model rather than steps/day, yielding similar results
2

2

(uncontrolled R =0.053, p=0.29; controlled R =0.193, p=0.39).

103

CHAPTER 5: DISCUSSION

The purpose of this study was to examine the association of stress (measured by survey
and total daily cortisol) with metabolic syndrome in obese youth, and to evaluate whether
habitual physical activity attenuated the relationship. A total of fifty adolescents agreed to
participate in the study. Of these 50 adolescents, 32 completed saliva collection, 32 completed
physical activity monitoring, 41 had complete glucose and lipid panels, and 38 returned a
completed Pediatric Symptom Checklist (PSC)(see Table 4.1). Additional information regarding
cardio-metabolic health was collected during routine clinical visits at the Healthy Weight Center
and Academic General Pediatrics clinic at Helen DeVos Children’s Hospital in Grand Rapids,
MI. This chapter will begin by addressing primary and secondary findings and will then consider
potential explanations for these results, concluding with a summary of findings.
Metabolic Syndrome
In the total sample, 33% of participants met the criteria for metabolic syndrome as
outlined by Cook and colleagues (1). Previous studies of obese youth, including one from our
group (2), report the prevalence of MetS in obese youth ranging from 39-53% (2; 3) - estimates
vary largely as a result of the use of different cut-points and definitions of MetS described
previously (see Table 2.2). Adolescents in this study were most likely to meet Cook’s MetS
criteria for waist circumference, followed by HDL-cholesterol, BP, and triglyceride. Adverse
fasting blood glucose levels were less common, which is similar to our previous findings (2);
however, in our previous report, elevated blood pressure was the least common criterion met.
Clinic-measured systolic blood pressure has been shown to have good agreement (ICC = 0.91)

104

with measures taken by a trained researcher, while agreement between clinic- and researchermeasured diastolic blood pressure is slightly weaker (ICC = 0.77) (4).
Adolescents recruited from the HWC were more likely to meet the criteria for MetS, and
to have more adverse values for each component of MetS. This was expected since the HWC
sub-sample represents adolescents who are actively seeking treatment for obesity. Results are
consistent with available data showing that severely obese youth experience a higher prevalence
of MetS than moderately-obese youth (49.7% vs. 38.7%, respectively)(3). Although the
prevalence of MetS was slightly lower in our sample, the MetS was more common in the more
severely obese HWC subsample (BMI ≥ 99th percentile) (41%) vs. the AGP subsample (28%).
Although the prevalence of MetS appeared to be higher in boys (41.7%) vs. girls (30.0%), this
difference was not statistically significant. Although the prevalence of MetS was expected to be
higher in boys, our data demonstrate a small effect size (Cramer’s Φ = 0.11).
The average cMetS score in this sample was 4.16 ± 4.30. As expected, subjects with the
MetS had a significantly higher score (5.98 ± 1.60) than those without (2.38 ± 0.45; p=0.004).
Metabolic risk was moderately correlated with BMI z-score (r=0.55, p<0.0001), such that more
severely obese adolescents experienced higher cMetS scores. This is consistent with previous
studies showing that the prevalence of MetS is higher in obese youth than their normal weight
peers (3), and that heavier children tend to have higher cMetS scores (5). The mean and range of
scores reported in this study are similar to those reported previously in 7-9 year-old children of
all weight classifications (5). The range of cMetS scores found in this sample also supports the
notion that cardiovascular risk status varies even within a cohort of “high risk,” severely-obese
youth. In fact, 21% (n=9) of participants exhibited relatively healthy risk factor profiles, defined
as meeting zero or one cut-point for MetS. Thus, it is important to remember that while obese
105

youth are at greater risk of adverse health effects, not all will experience these effects during
adolescence.
Physical Activity
The SenseWear Armband (SWA) was used to assess habitual physical activity in terms of
steps/day, total daily energy expenditure (TDEE), and moderate-to-vigorous physical activity
(MVPA). Based on a previous validity study of v 7.0 of the SWA software, the lower cut-off for
moderate-intensity PA was raised to 4.0 METs to avoid an overestimation of time spent in
MVPA (6). Subjects averaged 8,434 steps per day with considerable inter-individual variation
(2,858 – 16,010 steps/day). Adolescents recruited from the AGP clinic were more active in
terms of steps/day than those recruited from the HWC, but no differences were found for MVPA
or TDEE. Physical activity levels did not differ between boys and girls.
Based on previously published data, subjects in this study were similar in activity levels
to other obese adolescents, and less active than the general adolescent population (7; 8). A
recent review outlining normative step/day values for children and adolescents presented median
values ranging from approximately 10,000 to 17,000 steps/day for boys and 10,000-13,000
steps/day for girls (7). Obese adolescents have been shown to take slightly fewer steps per day
than non-obese youth, averaging about 10,000 steps/day in a study by Olds (8) and
approximately 10,430 in a study of treatment-seeking obese children and adolescents (9). As
shown in a study by Tudor-Locke (7), boys generally average higher levels of steps/day than
girls; however, this was not found in the present study. Daily step counts between boys and girls
in the current study varied slightly, with girls taking more steps/day than boys, which is in
contrast to previous studies showing the boys generally take more steps than girls. This
discrepancy may be explained by our uneven and small sample size, as only 8 boys had complete
106

physical activity data compared to 24 girls; however, the observed effect size was small
(Cohen’s d = 0.11). The difference in steps per day seen between our study and others may also
be explained by the use of different devices (i.e., SWA vs. pedometer), and the validity of
steps/day as measured by the SWA has not been fully explored.
Total daily energy expenditure averaged 3365.7 ± 858.8 kcal/day in boys and 2766.8 ±
714.0 kcal/day in girls. Boys expended more calories than what has generally been shown for
obese boys in a previous study when measured by doubly-labeled water (2,621 kcal/day) and
earlier versions of the SWA software [vs. 5.1 = 2,577 kcal/day, and v. 6.1 = 2,112 kcal/day, each
for the total sample]; however, values for girls were similar to those previously reported by
doubly-labeled water (2,540 kcal/day)(10). Absolute values reported in the current study for the
total sample are also in excess of those reported in normal weight 9-16 and 10-14 year-old
children and adolescents (8; 11), as well as overweight and obese 9-16 year-olds (8). However,
when expressed as a relative value per kg of body weight, the values seen in this study (30.1 ±
4.8 kcal/kg/day and 29.8 ± 6.8 kcal/kg/day for boys and girls, respectively) are less than those
reported in normal weight youth. Although not reported directly in the paper, values published
by Dorminy and colleagues (11) based on whole-room calorimetry indicated relative TDEE of
43.2 kcal/kg/day for the total sample (it is not possible to calculate values for boys and girls
based on the data reported), which is 144% greater than the relative energy expenditure seen in
the current study (29.9 ± 6.3 kcal/kg/day for the total sample). Additionally, DLW values from
overweight and obese youth published by Backlund and colleagues (10) indicated relative TDEE
of 50.7 kcal/kg/day for the total sample, 48.2 kcal/kg/day for girls, and 57.0 kcal/kg/day for
boys. This discrepancy between the Backlund sample and the normal weight sample of Dorminy
is probably explained by differences in lean body mass; however, body composition data are not

107

available for either the current study or the Backlund study, so it is not possible to compare body
composition between the two.
Mean values of MVPA found in this study are similar to those reported elsewhere for
both normal weight and obese youth. Trost and colleagues (12) reported objectively-measured
(accelerometry) physical activity levels of approximately 70 min/day in obese youth while Ball
and colleagues (13) reported considerably lower amounts of approximately 43 min/day for obese
girls and 38 min/day for obese boys (ages 13-17 y). Studies reporting objectively measured
MVPA in normal weight youth or samples not divided according to weight status tend describe
higher values, including mean values of approximately 60 min/day (14; 15), 80 min/day (16), or
even 92 min/day (12). These values may vary due to the use of measurement instruments and
other methodological differences such as accelerometer cut-points. These values also vary
dramatically from self-reported values we have previously reported in a similar sample
(2)(approximately 17 min/day) and the approximately 726 min/week (104 min/day) reported in
European youth participating in the HELENA study (15). Again, the lack of statistically
significant sex differences in the current study may be explained by the relatively small sample
of boys. However, most physical activity differences were in the expected direction, with boys
participating in more MVPA and expending more kilocalories per day, although effect sizes for
MVPA and TDEE were small (Cohen’s d = 0.22 and 0.20, respectively).
In summary, participants in the current study took fewer steps per day and had lower
TDEE than normal weight youth and other obese youth reported in the literature. However,
levels of MVPA reported in this study are similar to those reported elsewhere in obese youth. In
this study, we raised the cut-point for moderate PA to 4 METs based on our previous validation
study (6); however, time spent in MVPA may still be slightly overestimated in this sample. The
108

difference in steps per day seen between our study and others may be explained by the use of
different devices (i.e., SWA vs. pedometer), and the validity of steps/day as measured by the
SWA has not been fully explored. Additionally, PA levels of boys and girls in the current study
did not differ as greatly as has been reported previously. This discrepancy between our results
and others may be explained by our relatively small sample of boys. Lastly, although subjects
recruited from AGP took significantly more steps/day than subjects from the HWC, no
significant differences were found between clinics for MVPA or TDEE. The reason for this
difference between clinics is not clear; however, HWC subjects displayed greater variability than
AGP subjects in both MVPA (range: 103.1 min vs. 86.0 min, respectively) and TDEE (range:
3,080 kcal/day and 2,125 kcal/day, respectively).
Daily Cortisol
In this study, salivary cortisol, or more specifically area under the curve of daily cortisol,
was used as an index of stress. Cortisol, the chief hormone of the hypothalamic-pituitary-adrenal
axis (HPA axis) has attracted attention in the field of obesity and metabolic syndrome research
because of the ability of the HPA axis to regulate cardiovascular, metabolic, and respiratory
responses to real or perceived threats (17). The HPA axis is activated in response to long-term
stressors and is often used as a biomarker of the stress response (18). When the HPA axis is
functioning appropriately, cortisol is released into the circulation in a predictable pattern.
Circulating cortisol levels are low just before waking, and increase quickly within the first 30-45
minutes after waking to reach their peak for the day. Following this peak, cortisol gradually
decreases throughout the remainder of the day. As shown in Figure 4.3, cortisol levels in this
study followed the expected patterns, with no significant differences between boys and girls at
any time point. Additionally, stressors such as a meal, physical activity, mental, or psychosocial

109

stressors stimulate further release of cortisol and can result in alteration of the normal diurnal
pattern at any time during the day (19). Because of the general predictability of this pattern,
cortisol is often assessed immediately upon waking and/or 30 minutes after waking in order to
assess the peak. The cortisol awakening response (CAR), or the difference between the waking
and 30-min post measures, is used frequently in the psychology literature as an assessment of
normal HPA axis activity (20). Cortisol area under the curve (cAUC) can also be used as an
assessment of total daily cortisol exposure (21).
The dynamic nature of cortisol in the period immediately following waking and the
responsivity of the HPA axis to stress pose several challenges to accurate assessment of cortisol
levels. In order to accurately compare cortisol measures obtained in the morning, it is necessary
to tightly control sampling time, since deviation of only a couple minutes can result in error
when cortisol levels are changing quickly. Unfortunately, the best way to tightly control
sampling time is via venipuncture or indwelling catheter following an overnight stay in a
controlled setting (i.e. hospital or general clinical research center). Doing this provides consistent
timing of sampling between subjects but induces substantial participant burden. The ability to
assess cortisol levels in saliva allows subjects to collect samples at home, in free-living
conditions, without the additional stress of venipuncture. However, asking subjects to collect
saliva samples at home reduces the ability to tightly control sampling time. Collection of
additional time points throughout the day and calculation of cAUC allows for an estimation of
total cortisol exposure while still allowing for examination of individual time points, providing
additional information about daily patterns.
In the current study, we attempted to control sampling time by polling subjects at the time
of recruitment about their usual waking time and sending text messages and/or making reminder
110

phone calls at each sampling time. Additionally, each correctly taken sample was incentivized
by adding one dollar to the subject’s gift card reward. Subjects were required to collect at least
sample 1 (immediately upon waking), sample 2 (30-min after waking), and sample 6 (12 hours
after waking) in order to receive a gift card and to be included in the cortisol analysis. Of the 50
subjects, a total of 39 returned saliva samples; of these, 32 met the minimum criteria for
inclusion in data analysis. The most frequently missed samples were at time points 3, 4 and 5,
which is not surprising since these three samples were not incentivized as strongly as samples 1,
2, and 6. The majority of subjects (69%) experienced a cortisol awakening response (CAR) when
defined as any increase over resting values. A total of 16 of the 32 subjects (50%) met the CAR
threshold of at least 2.5 nmol/L that has been reported previously (22). The mean change found
in the present study (4.66 ± 3.21 nmol/L) was smaller than that described in one study of healthy
female adolescents (9.16 ± 7.42 nmol/L) (22), but similar to another study of male and female
adolescents with and without major depressive disorder (4.77 ± 8.92 nmol/L) (23) –
unfortunately, neither study provides information regarding the participants’ weight status. A
number of subjects (n=10) in the present study experienced a decrease in cortisol levels from
sample 1 to sample 2. Because the percentage of non-responders in the present study is greater
than other studies (22), a lack of a response and perhaps especially a decrease might be
indicative of poor adherence to the study protocol. It is also possible that collecting the second
sample 30 minutes after waking missed the true cortisol peak in some female subjects, which
may occur as late as 45 minutes after waking (22). The variation in peak time may be related to
stage of sexual maturation, as well, with post-menarcheal girls peaking later in the morning (22).
However, cortisol levels at each time point, CAR, and cAUC did not vary by pubertal stage
(pubic hair) in the current sample.

111

Cortisol increases in the post-prandial period in both lean and obese youth (24). We
attempted to control for the increases in cortisol following meals by directing subjects not to eat
less than 1 hour before collecting each saliva sample. Time since last meal was recorded on the
sampling log filled out by each subject on the day of sampling and used to verify that subjects
followed directions regarding meal timing. As seen in Figure 4.2, meals did not have much of an
effect on cortisol levels in this sample.
Sexual dimorphism in cortisol levels (usually in the morning and often at school) has
been reported in previous studies (25–27), but was not evident in the present study. However,
previous studies that have compared boys and girls have varied in the method of cortisol
measurement (saliva vs. serum), and sampling time (morning, at school, etc.). This sexual
dimorphism was only evident in pubertal stage 2 in the study by Knuttson et al. (26) –
unfortunately, the pubertal development scale used in that study was not specified. The lack of
sex differences in the present study may be related to the relatively even distribution of pubertal
stages among the participants, as sex differences noted in previous studies have often matched
differences according to pubertal stage (27–30). It is difficult to explain the lack of sex
differences in the current study compared to others due to varying methodologies. However, the
current study included a much larger sample of female subjects than male subjects, which may
make it difficult to identify statistically significant differences between the sexes.
Psychosocial Functioning
In addition to the use of cortisol as a biomarker of stress, the 35-item youth report of the
Pediatric Symptom Checklist (PSC) was used to evaluate overall psychosocial function. The
PSC consists of 35 questions that assess symptoms and difficulties adolescents may experience
during daily life, including conflicts with peers and family members, difficulties in school,
112

somatic complaints, attention difficulties, and negative feelings, among others. The mean PSC
score in this sample was 23.5 ± 11.0, which falls below the clinical cut-off of 30 used to identify
impaired psychosocial functioning. Scores were slightly higher in HWC adolescents vs. AGP
adolescents, but this difference was not statistically significant. Similarly, HWC adolescents
were slightly more likely to be classified as having impaired psychosocial functioning than those
from AGP, though again this difference was not statistically significant. In the total sample, 24%
were classified as having impaired psychosocial functioning, which is higher than the prevalence
of impairment that has been noted in previous studies in the general pediatric primary care
population (31; 32). Another study of school-age children (9-14 y) assessed at primary care visits
that included a large number of low SES patients found a higher rate of impairment than the two
studies previously mentioned (20% vs. 8-14%)(33), but still lower than the proportion identified
in the current study. We previously found similar prevalence rates in a similar treatment-seeking
obese population (20.4% and 33.3%, respectively) (2; 34). Although SES was not directly
assessed in the current study, approximately 50% of all patients seen at the HWC and 85% of all
patients seen at AGP are eligible for Medicaid; therefore, it is reasonable to assume that the SES
levels of subjects in our study are similar. Our two previous studies showing similar prevalence
of psychosocial dysfunction (2; 34) were also drawn from patients at HWC and thus, SES should
be similar among the three studies.
Subscales for internalizing and externalizing difficulties, identified by factor analysis of
children in primary care (35), were also used to examine whether there are differences in cAUC
and cMetS based on different types of psychosocial dysfunction. The mean internalizing and
externalizing scores were 19.8 ± 9.5 and 7.0 ± 3.8, respectively; mean scores were not provided
in the Gardner paper, so it is not possible to compare these values to those reported previously.

113

Previous studies have indicated increased risk of internalizing problems (i.e. depression and
axiety) and externalizing problems (i.e. oppositional defiant disorder [ODD], attention-deficithyperactivity disorder [ADHD], or aggression) among overweight and obese children and
adolescents than their normal weight peers (36; 37). Although treatment-seeking obese children
(age 9-12 y) have been shown to score higher on internalizing and externalizing assessments than
non-treatment-seeking youth (37), this is not always supported in the literature (36; 38). The
present study did not show a difference between treatment-seeking (HWC) and non-treatment
seeking (AGP) adolescents, although we used a different instrument than these previous studies
(PSC vs. Child Behavior Checklist (CBCL)). The mechanism for potential differences between
treatment-seeking and non-treatment-seeking youth is unclear, but perception of psychosocial
difficulties may influence the decision to seek treatment (37).
Main Findings
Cortisol and Metabolic Syndrome
Researchers are interested in examining the relationships between cortisol and MetS
primarily because of previously documented relationships between cortisol levels and insulin
sensitivity in adults (39). Because cortisol is released in the body in response to stress – or
perceived threats – its main purpose is to ensure availability of glucose to properly fuel body
defenses (40). Increased glucose availability is achieved by decreasing responses to insulin at
target tissues, inhibiting β-cell release of insulin, and increasing hepatic glucose output (41). In
the present study, post-hoc data analysis shown a strong positive, correlation emerged between
cAUC and insulin when controlling for gender and pubic hair stage (r=0.59, p=0.01). However,
since insulin levels were only available for a portion of the subjects in this study (n=36), this was
not included in the main analysis and HOMA-IR was not calculated. Two previous studies have
114

also found that cortisol levels are associated with insulin resistance among obese adolescents,
and that this may also be influenced by abdominal fat (42; 43). An association between serum
cortisol and changes in β-cell function and acute insulin responses to glucose has also been
shown in a longitudinal study of overweight and obese Latino children and adolescents (44).
Taken together, these results suggest that elevated cortisol is related to increased release of
insulin in order to counteract cortisol-related insulin resistance. Over time, individuals with
chronically elevated cortisol may experience declines in pancreatic β-cell functioning, eventually
leading to the onset of type 2 diabetes. These results support the hypothesis posed by Soros and
colleagues in 2008 (45), which suggested that decreased HPA axis activity, manifested through
reduced cortisol release, may actually protect obese youth from the long-term effects of
chronically elevated cortisol on insulin action and glucose tolerance.
We did not find a relationship between cAUC and blood pressure in this sample. The lack
of a relationship between cortisol and blood pressure is in contrast to previous studies in adults
(46) and children (43). However, a previous study from our group (47) did not support a
relationship between morning cortisol and blood pressure. More recently, we have found an
inverse relationship between psychosocial function score and blood pressure in obese children
and adolescents (2). Tanaka and colleagues (49) also found an inverse relationship between
psychosocial and somatic symptoms and blood pressure in youth, suggesting that depressionrelated reductions in sympathetic activity may be responsible. Another study in adults (50) found
an inverse relationship between respiratory sinus arrhythmia (a marker of parasympathetic
nervous system activity) and pre-ejection period (a marker of sympathetic nervous system
activity) and systolic blood pressure; conversely, cortisol was not associated with blood pressure.

115

Mechanistically, specific stress interpretations or psychosocial dysfunction (such as depression)
may be associated with stronger effects on the autonomic nervous system than the HPA axis.
When cortisol was examined at each individual time point, no significant relationships
were found with MetS. Post-hoc analyses revealed a significant positive relationship between
sample 1 and DBP (r=0.49, p = 0.04) and a trend towards significance for MAP (r=0.43, p=0.07).
Sample 2 was positively correlated with insulin levels (r=0.58, p=0.01), but not any of the blood
pressure measures. Sample 6 was not significantly related to any MetS features. The
relationship between sample 2 (the early-morning peak) and insulin is consistent with those
results previously reported by Barat and Weigensberg, and the relationship between sample 1
(immediately upon waking) and blood pressure is consistent with those of Weigensberg (43).
Aims 1 and 2:
The results of this study indicate that daily cortisol levels are not independently
associated with metabolic syndrome in obese adolescents. The first aim of this study, to examine
the relationship between cAUC and cMetS, revealed no significant relationship between the two
variables. This was unexpected, as previous studies have shown cortisol levels to be related to
factors of the MetS among obese youth. Weigensberg and colleagues (43) found that morning
cortisol levels were positively correlated with the number of features of the MetS exhibited.
However, the relationship between morning cortisol and features of MetS was no longer
significant after adjusting for intra-abdominal adipose tissue (IAAT), suggesting that IAAT may
lie along the causal pathway between cortisol and MetS. Although there was a wide range in
absolute values in the present study, nearly all subjects met the criteria for elevated waist
circumference, which has been shown to be highly related to IAAT in this population (51).
However, neither IAAT nor body fat percentage were measured in the present study.
116

Physical activity, expressed as both steps/day and min/day of MVPA, was inversely
correlated with cMetS in this sample (r=-0.54, p=0.002 and r=-0.45, p=0.01, respectively).
However, there was not a significant relationship between TDEE and cMetS (r=-0.05, p=0.80).
The relationships shown with steps/day and MVPA are consistent with those reported previously
by the European Youth Heart Study (52), NHANES (53), AFINOS (54), and another study from
Vietnam (55). However, when physical activity was added to the regression examining the
relationship between cAUC and cMetS, the predictive ability of the model was improved, but
failed to significantly predict variance in cMetS. This may be explained in part by our relatively
small sample size, which was reduced by the failure of many subjects to collect complete data
for all measures.
Conversely to the idea that IAAT lies along the causal pathway between cortisol and
cMetS, results from the European Youth Heart Study (56; 57) suggest that PA influences cMetS
independently of adiposity. Therefore, there must be another mechanism, not directly related to
adiposity, which is responsible for this relationship. The present study was driven by the
hypothesis that dysregulation of the HPA axis, even at subclinical levels, might be the
mechanism responsible. To date, only one previous study has examined the role of PA in the
pathway between stress and obesity in youth. In this study of 303 adolescents, personal stress
was associated with adiposity after controlling for socioeconomic status, and this relationship
was modified by days of PA per week (58). These results support the notion that PA is involved
in the relationship between stress and obesity, however more research is necessary to clarify the
mechanisms driving this relationship. Similarly, only one study has examined the combined
influence of stress and PA on MetS in youth. Holmes and colleagues (16) found that, among less
active youth, certain measures of stress including school- and sports-related self esteem and trait

117

anxiety were positively associated with cMetS, while these associations did not exist in the high
PA group. These results suggest that PA modifies the relationship between stress and MetS.
However, this study did not include female subjects, nor did it include a physiological measure
of stress; thus, it is still not possible to identify the physiological basis for the relationships
found.
Aim 3:
The PSC was included in the present study in order to link the biological markers and
survey measures included in previous studies. Similar to the results for cAUC, no relationship
was found between PSC scores and cMetS, yet some interesting findings emerged. A moderate
inverse relationship was found between the PSC total score and cAUC (r=-0.45, p=0.04), and an
inverse relationship between PSC total score and cortisol sample 2 (r=-0.54, p=0.01). These
inverse relationships may further support the hypothesis (45) that cortisol release is suppressed in
obese adolescents who experience chronic stress in order to protect the body from the
detrimental metabolic effects of chronically elevated cortisol, especially in light of our findings
linking higher cAUC levels to elevated insulin levels. It is important to note, however, that this is
the first study to examine the relationships between the three PSC scores and cortisol levels,
making it difficult to compare our results to those of previous studies. Therefore, these results
should be interpreted with caution. Additional research is necessary to confirm the relationships
between the PSC scores and cortisol levels, and to evaluate whether the PSC properly captures
the types of stressors that influence adolescents’ HPA axis activity. Additionally, it is unclear
from our data if the PSC scores and cortisol levels are directly related, or whether a third
unknown variable may be involved in this relationship.

118

Additional analyses revealed a moderate inverse relationship between the PSC
internalizing score and cAUC (r=-0.48, p=0.03) and cortisol sample 2 (r=-0.57, p=0.01). This is
consistent with the relationships between self-esteem and depression (both internalizing
difficulties) and cMetS previously described by Holmes and colleagues (16). Conversely, PSC
externalizing score was found to be inversely and moderately correlated with cortisol sample 2
(r=-0.50, p=0.02). A previous study of adolescent boys undergoing psychological therapy
examined the relationship between cortisol samples collected immediately upon waking and in
the afternoon and externalizing behavior assessed using the CBCL (59). In this study, morning
cortisol levels were positively related to externalizing behavior and afternoon cortisol levels were
inversely related to externalizing behavior; however, these relationships were found after
treatment was complete, making comparison to the current sample inaccurate because
participants were either just beginning treatment at the HWC or not receiving specific care for
obesity as at AGP. The study by Schechter and colleagues also indicated that the type of
stressors participants experienced was important in the relationships between cortisol and
externalizing behavior, such that high daily hassles and high cortisol were related to
externalizing behavior, while high exposure to stress over the lifetime in combination with low
cortisol levels was associated with more externalizing behavior (59). Unfortunately, the PSC
does not allow for identification of specific types of stressors experienced by the participants.
Another study that examined acute responses to stress found that externalizing behavior
problems were associated with a blunted cortisol response to the Psychosocial Stress Test
among 12 year-old children who were bullied at younger ages or maltreated by an adult (60).
While this study assessed acute cortisol responses to stress rather than basal cortisol levels, the
results suggest that children with externalizing behavior problems may experience reduced

119

cortisol reactivity. Future studies are necessary to better understand how internalizing and
externalizing problems are related to cortisol levels, both chronically and in response to acute
stressors, and how this might relate to metabolic health during adolescence.
Aims 4 and 5:
Similar to aim 1, the results of this study indicated that our survey measure of stress, the
PSC, was not independently associated with cMetS in obese adolescents. This was unexpected as
it differs from the results of Holmes and colleagues (16), who found that specific measures of
stress (i.e. self-esteem, anxiety) were associated with a sample-specific cMetS score among 8-18
year old boys. Holmes found that this relationship between stress and MetS existed only among
the least active boys in the sample, while the most active boys did not exhibit this relationship.
This finding is also in contrast to the results of the present study, as our regression analyses for
aim 5 did not support a significant interaction effect between the PSC scores and PA for
predicting cMetS in our sample. Disagreement between these two studies may be explained by
several methodological differences. First, Holmes and colleagues used a number of
psychological assessments that specifically addressed depression, state- and trait-anxiety, selfesteem, appearance-related teasing and perceived stress; as previously mentioned, the PSC does
not allow for identification of specific psychosocial domains. Since the relationship between
stress and cMetS only existed for specific domains in the Holmes study, it is possible our method
was not sensitive enough to identify the same relationships. In addition, the Holmes study
included boys in all weight classifications, while the current study included only obese
adolescents and also included girls. This difference in the sample population, in addition to the
use of a sample-specific cMetS score in the Holmes study, may explain some of the
disagreement with our study.

120

Strengths and Limitations
This study is characterized by several strengths. Most notably, the assessment of cortisol
at several time points on one day improves upon previous studies that have employed only a
single waking sample or a single sample at another time of day. Although not all subjects
correctly obtained six saliva samples, 32 did collect at least three samples. Previous studies have
attempted to examine the relationships among a single cortisol sample obtained in the morning,
either immediately upon waking or 30-45 minutes after waking, and obesity and metabolic
syndrome. In contrast, daily cortisol levels may be better represented by a single sample taken
later in the day, such as an early afternoon sample (21). A mid-afternoon sample was included in
the present study, but unfortunately was not collected by enough subjects to be useful for
analysis (n=13 for cortisol sample 4 and n=12 for cortisol sample 5). Other studies that have
attempted to characterize the relationship between stress and MetS in either normal weight or
overweight and obese children and adolescents have relied on survey measures of stress and thus
have not been able to provide mechanistic evidence for any relationships found. The current
study improves on these previous studies by including both biological and survey measures. We
were also able to employ objective monitoring of physical activity as opposed to self-reported
activity. Furthermore, the use of the cMetS score based on nationally-representative data is a
more comprehensive measure of cardiovascular risk than previous examination of dichotomous
classifications of MetS, or exclusively measures of body size and composition. To our
knowledge, this is the first study to combine the cMetS score with a cumulative measure of
cortisol secretion in obese youth.
The most notable limitation to the present study is compliance with data collection,
especially for salivary cortisol, PA, and the PSC. Despite reminder phone calls and text messages

121

at the scheduled time for each saliva sample, only 32 (64%) subjects returned saliva samples that
met the minimum inclusion criteria, and only 5 (10%) correctly obtained all six samples.
Samples 1, 2, and 6 were more heavily incentivized than samples 3, 4, and 5 (i.e., samples 1, 2,
and 6 were required to be eligible for any compensation, whereas 3, 4, and 5 were only worth
one additional dollar), and this is reflected in the fact that samples 3, 4, and 5 were more likely to
be missed than the other three. Although most subjects spoke or exchanged text messages with
the primary investigator at all each sampling time and filled out a sample card documenting
sampling time, it is not possible to know for sure that samples were taken at the time reported.
This could be improved by having a trained research assistant personally observe each subject
collect each sample, or by conducting the study in a controlled environment such as a general
clinical research center. However, collecting samples in such a tightly controlled environment
might reduce the applicability of the results to a free-living setting and requiring a research
assistant’s presence for measures would be burdensome to both the researcher and participant.
Additionally, we assessed cortisol levels on a single day. Studies of day-to-day cortisol
stability in adolescents are limited. However, cortisol has been shown to be moderately- to
strongly-correlated on two sampling days in adolescent girls both before and after menarche
(22), and in adults (21; 61). Therefore, daily variation is not expected to have a large effect on
the results of the current study, but more research is necessary on day-to-day stability in
adolescents’ cortisol patterns.
The use of the SWA for assessment of physical activity allows for precise determination
of wear time. However, several adolescents who did not agree to participate in the study
expressed apprehension about wearing a monitor other kids would be able to see, or thought that
it might be uncomfortable. Efforts were made to address these concerns with subjects, but it is
122

possible that these issues affected the amount of time subjects wore the device. As with saliva
collection, requiring a minimum amount of wear time in order to be eligible for compensation
incentivized SWA compliance. This could potentially be improved by more heavily
incentivizing additional time, or by the use of a smaller, less noticeable and more comfortable
device.
The PSC was returned by only 38 (76%) subjects. In order to avoid disruption of clinic
flow, subjects recruited from AGP were given a copy of the PSC to take home and return with
their samples. They were reminded to return the survey in the box with their samples and
armbands; however, this was not successful in all cases. Compliance could be improved in
future studies by asking subjects to complete the PSC before leaving the clinic. The use of the
PSC as a pencil-and-paper assessment of psychosocial function can be seen as both a strength
and a weakness, in this case. On one hand, the PSC is commonly used in primary care in order
to identify children and adolescents who may experience psychosocial dysfunction and need
mental health care. In this instance, the use of the PSC in this study makes our results more
directly accessible to primary care physicians. Conversely, the choice of the PSC limits our
ability to compare our results to those of previous studies, and precluded the examination of
specific types of daily stressors and their relationship to cortisol and MetS.
Some previous studies have suggested that adolescent girls may reach a peak in salivary
cortisol about 45 minutes after waking, rather than the conventionally accepted 30 minutes (22).
Girls in the present study reached a slightly higher (though not statistically significant) peak than
boys, so timing of peak cortisol level is not expected to be a significant limitation in this study.
However, some subjects who had a relatively small CAR may have had a larger response if
sampling was completed at a later time. Also, as previously mentioned, it is not possible to
123

know with absolute certainty how long after waking each sample was collected. Therefore, some
samples may vary by a few minutes in either direction, adding uncertainty to the measurements.
Although attempts were made to control for biological maturation using secondary sex
characteristics, we did not control for menstrual cycle phase or hormonal contraceptive use
among post-menarcheal girls. A recent review suggests that both menstrual cycle phase and oral
contraceptive use may influence the CAR (through their influence on gonadal hormones), but
studies to date have been equivocal (20). Menstrual cycle phase influences plasma lipid levels in
premenopausal women, such that LDL, the TC:HDL ratio, and LDL:HDL ratio are all lower
during the luteal phase, when estrogen levels are higher than in the follicular phase (62).
Polycystic ovarian syndrome (PCOS), characterized by insulin resistance, hirsutism, acanthosis
nigricans, acne, and oligo- or amenorrhea, is reported in obese female adolescents. While PCOS
is associated with insulin resistance, it has not been shown to be associated with alterations in
lipid profiles when compared to girls without PCOS who were similar in terms of BMI, waist
circumference, and waist:hip ratio (63), or to samples of adolescent girls of all weight
classifications (64). Oral contraceptive use, however, has been associated with increased blood
pressure as well as increased TC, HDL, LDL, and triglycerides (65). Therefore, it is possible that
oral contraceptive use may have influenced our findings, while menstrual cycle phase probably
did not.
Several dietary factors, including vegetarian diets, fiber consumption, low fat dairy
consumption, are associated both with individual components of MetS and with MetS itself in
adolescent samples (66–69). Moreover, obese adolescents tend to eat diets that are relatively low
in many important nutrients such as mono- and poly-unsaturated fatty acids, calcium, and
magnesium, and high in saturated fat and sodium, which can contribute to adverse risk factor
124

profiles (70–73). Additionally, while less is known about the relationship between specific
nutrients and cortisol levels, patients with Cushing’s syndrome need to make efforts to reduce
sodium intake and meet recommendations for Vitamin D and calcium (74). Therefore, it is
important to consider dietary covariates when examining the relationships between cortisol and
MetS in obese adolescents. Unfortunately, data collected via 3-day diet records were not useable
in the current study because it was not possible to follow-up with subjects to clarify diary entries.
Although most subjects returned diet records, very few were detailed enough to glean any
information. Entries were frequently missing serving sizes or clarifying information. For
instance, lunch was frequently recorded as “sandwich,” without any information regarding the
type of sandwich, condiments, etc. Serving sizes were often listed as “bowl,” or “a little of
everything,” or missing entirely. Therefore, we were not able to control for dietary intakes in this
study. Consideration of energy balance may also be of importance in examining the relationships
between cortisol and MetS in obese adolescents who are trying to lose weight. Negative energy
balance results in increased cortisol levels meant to increase food intake. These increased
cortisol levels can, in turn increase the desire for highly-palatable foods (75). Future studies
should aim to collect quality dietary data, including total caloric intake as well as detailed
macronutrient information, in order to consider the interactions of dietary factors with cortisol
and MetS.
Lastly, and perhaps most importantly, this study was cross-sectional. The adolescent
sample is interesting because of changes in the overall hormonal milieu, growth, and insulin
sensitivity that are happening during puberty, but it is possible that earlier life influences affect
the relationship between cortisol and MetS in ways that could not be assessed with this study
design. On one hand, this study have been much more challenging to implement with younger

125

children and would require different saliva sampling techniques and more direct researcher
involvement with saliva collection. On the other hand, inclusion of a younger sample would
allow for the examination of relationships between cortisol and MetS without potential
confounding by sex hormones, and would make it possible to follow participants through
adolescence to assess change over time.
Future Directions
The results of this study offer several suggestions for further research. First, future
studies should attempt to provide better control over saliva sampling times. This could be done
by collecting data in a clinical setting using indwelling catheters or creating a more automated
system for monitoring sampling, perhaps using a smart phone or tablet computer. Collection in a
clinical setting is less ideal because an unfamiliar setting and frequent blood sampling, even with
a catheter, may increase acute stress and falsely elevate cortisol levels. Saliva sampling could
also be combined with 24-hour urine collection to allow for the analysis of both salivary cortisol
at specific time points, and urinary metabolites. Including more frequent sampling in the
morning may also allow for better determination of the morning peak, which may be helpful in
the face of potential variation in peak time between the sexes (22). Finally, reliable sample
collection throughout the afternoon would allow for analysis of the slope of the cortisol curve,
which may be related to a number of psychosocial factors (76).
As previously discussed, the poor compliance with physical activity monitoring in the
present study limits our ability to draw conclusions on the role of physical activity in the
relationship between stress and metabolic syndrome. Future studies should consider alternative
assessment methods, such as less conspicuous accelerometers worn at the waist, detailed
physical activity diaries, or a combination of assessments.
126

Future studies should also consider the use of additional psychological assessments, to
more closely examine specific psychosocial aspects that have been related to both cortisol and
metabolic syndrome. Assessment of depression, anxiety, experiences with bullying and trauma,
and familial stressors such as divorce or parental unemployment may provide a clearer picture of
the effect of specific dimensions of psychosocial functioning and stressors on health in obese
youth.
The results of the present study raise a number of research questions to be addressed in
the future. First, the positive relationship found between cAUC and insulin levels coupled with
the inverse relationships between PSC scores and cortisol levels seem to suggest the hypothesis
that cortisol levels may be suppressed in a subset of obese adolescents, in an effort to protect the
body from the negative effects of chronic cortisol excess (45). These relationships would be best
studied using a longitudinal study design beginning in early childhood and including more
comprehensive assessments of stress exposure and perception. Doing so would allow
researchers to identify the timing of altered cortisol secretion, metabolic effects, and
exposure/reaction to stressors. Additionally, the inclusion of markers of sympathetic and
parasympathetic nervous system activity, as discussed previously, would allow researchers to
consider the interactions of the different components of the autonomic nervous system for a more
comprehensive assessment of the relationships between stress and metabolic health. Finally,
recent research highlighting the importance of sedentary time in the development of obesity and
comorbidities including psychosocial health (77) suggests that time spent being sedentary may
be an important variable to include in future studies examining the role of physical activity in the
relationship between stress and metabolic syndrome.

127

REFERENCES

128

REFERENCES
1.

Cook S, Auinger P, Huang TT-K. Growth curves for cardio-metabolic risk factors in
children and adolescents. J Pediatr. 2009 Sep;155(3):S6.e15–26.

2.

Guseman EH, Eisenmann JC, Stratbucker W. Does physical activity modify the
relationship between psychosocial stress and metabolic syndrome in obese youth? In
Preparation. 2012.

3.

Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW, Allen K, Lopes
M, Savoye M, Morrison J, Sherwin RS, Caprio S. Obesity and the metabolic syndrome in
children and adolescents. N Engl J Med. 2004 Jun;350(23):2362–74.

4.

Kim JW, Bosworth HB, Voils CI, Olsen M, Dudley T, Gribbin M, Adams M, Oddone EZ.
How well do clinic-based blood pressure measurements agree with the mercury standard?
J Gen Intern Med. 2005 Jul;20(7):647–9.

5.

Eisenmann JC, Laurson KR, DuBose KD, Smith BK, Donnelly JE. Construct validity of a
continuous metabolic syndrome score in children. Diabetol Metab Syndr. 2010 Jan;2:8.

6.

Guseman EH, Eisenmann JC, Calabro MA, Morrison KM, Welk GJ. Validity of the
SenseWear Pro Armband in Obese Youth During Treadmill Walking. In Preparation.
2012.

7.

Tudor-Locke C, McClain JJ, Hart TL, Sisson SB, Washington TL. Expected values for
pedometer-determined physical activity in youth. Res Q Exerc Sport. 2009
Jun;80(2):164–74.

8.

Olds TS, Ferrar KE, Schranz NK, Maher C a. Obese adolescents are less active than their
normal-weight peers, but wherein lies the difference? J Adolesc Health. 2011
Feb;48(2):189–95.

9.

Eisenmann JC, Scott D, Hill EE, Hanba C, Atlas AB. Total daily energy expenditure and
physical activity in treatment-seeking overweight and obese youth. In Review. 2012.

10.

Bäcklund C, Sundelin G, Larsson C. Validity of armband measuring energy expenditure
in overweight and obese children. Med Sci Sports Exerc. 2010 Jun;42(6):1154–61.

11.

Dorminy CA, Choi L, Akohoue SA, Chen KY, Buchowski MS. Validity of a multisensor
armband in estimating 24-h energy expenditure in children. Med Sci Sports Exerc. 2008
Apr;40(4):699–706.

12.

Trost SG, Kerr LM, Ward DS, Pate RR. Physical activity and determinants of physical
activity in obese and non-obese children. Int J Obes. 2001;822–829.

129

13.

Ball GDC, Lenk JM, Barbarich BN, Plotnikoff RC, Fishburne GJ, Mackenzie K a.,
Willows ND. Overweight children and adolescents referred for weight management: are
they meeting lifestyle behaviour recommendations? Apply Physiol Nutr Metab. 2008
Oct;33(5):936–945.

14.

Garaulet M, Ortega F, Ruiz J, Rey-Lopez J, Beghin L, Manios Y, Cuenca-Garcia M, Plada
M, Diethelm K, Kafatos A, Molnar D, Al-Tahan J, Moreno L. 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. 2011;(35):1308–1317.

15.

Martinez-Gomez D, Gomez-Martinez S, Ruiz JR, Diaz LE, Ortega FB, Widhalm K,
Cuenca-Garcia M, Manios Y, De Vriendt T, Molnar D, Huybrechts I, Breidenassel C,
Gottrand F, Plada M, Moreno S, Ferrari M, Moreno L a, Sjöström M, Marcos A.
Objectively-measured and self-reported physical activity and fitness in relation to
inflammatory markers in European adolescents: the HELENA Study. Atherosclerosis.
2012 Mar;221(1):260–7.

16.

Holmes M, Eisenmann J, Ekkekakis P, Gentile D. Physical activity, stress, and metabolic
risk score in 8-to 18-year-old boys. J Phys Act Health. 2008;(5):294–307.

17.

McEwen BS. Protective and damaging effects of stress mediators: the good and bad sides
of the response to stress. Metabolism. 2002 Jun;51(6 Suppl 1):2–4.

18.

McEwen BS, Seeman T. Protective and damaging effects of mediators of stress.
Elaborating and testing the concepts of allostasis and allostatic load. Ann N Y Acad Sci.
1999 Jan;896:30–47.

19.

Chrousos GP, Kino T, Charmandari E. Evaluation of the hypothalamic-pituitary-adrenal
axis function in childhood and adolescence. Neuroimmunomodulation. 2009
Jan;16(5):272–83.

20.

Fries E, Dettenborn L, Kirschbaum C. The cortisol awakening response (CAR): facts and
future directions. Int J Psychophysiol. 2009 Apr;72(1):67–73.

21.

Edwards S, Clow a, Evans P, Hucklebridge F. Exploration of the awakening cortisol
response in relation to diurnal cortisol secretory activity. Life Sci. 2001 Mar;68(18):2093–
103.

22.

Oskis A, Loveday C, Hucklebridge F, Thorn L, Clow A. Diurnal patterns of salivary
cortisol across the adolescent period in healthy females. Psychoneuroendocrinology. 2009
Apr;34(3):307–16.

23.

Adam EK, Doane LD, Zinbarg RE, Mineka S, Craske MG, Griffith JW. Prospective
prediction of major depressive disorder from cortisol awakening responses in adolescence.
Psychoneuroendocrinology. 2010;35(6):921–931.

130

24.

Hershberger a M, McCammon MR, Garry JP, Mahar MT, Hickner RC. Responses of
lipolysis and salivary cortisol to food intake and physical activity in lean and obese
children. J Clin Endocrinol Metab.2004 Sep;89(9):4701–7.

25.

Rosmalen JGM, Oldehinkel a J, Ormel J, de Winter a F, Buitelaar JK, Verhulst FC.
Determinants of salivary cortisol levels in 10-12 year old children; a population-based
study of individual differences. Psychoneuroendocrinology. 2005 Jun;30(5):483–95.

26.

Knutsson U, Dahlgren J, Marcus C, Rosberg S, Brönnegård M, Stierna P, AlbertssonWikland K. Circadian cortisol rhythms in healthy boys and girls: relationship with age,
growth, body composition, and pubertal development. J Clin Endocrinol Metab. 1997
Feb;82(2):536–40.

27.

Törnhage CJ. Reference values for morning salivary cortisol concentrations in healthy
school-aged children. J Pediatr Endocrinol Metab. 2002 Feb;15(2):197–204.

28.

Netherton C. Salivary cortisol and dehydroepiandrosterone in relation to puberty and
gender. Psychoneuroendocrinology. 2004 Feb;29(2):125–140.

29.

Kelly SJ, Young R, Sweeting H, Fischer JE, West P. Levels and confounders of morning
cortisol collected from adolescents in a naturalistic (school) setting.
Psychoneuroendocrinology. 2008 Oct;33(9):1257–68.

30.

Bouma EMC, Riese H, Ormel J, Verhulst FC, Oldehinkel AJ. Adolescents’ cortisol
responses to awakening and social stress; effects of gender, menstrual phase and oral
contraceptives. The TRAILS study. Psychoneuroendocrinology. 2009 Jul;34(6):884–93.

31.

Jellinek MS, Murphy JM, Little M, Pagano ME, Comer DM, Kelleher KJ. Use of the
pediatric symptom checklist to screen for psychosocial problems in pediatric primary care:
a national feasibility study. Arch Pediatr Adolesc Med.1999 Nov;153:254–260.

32.

Hacker K a, Williams S, Myagmarjav E, Cabral H, Murphy M. Persistence and change in
pediatric symptom checklist scores over 10 to 18 months. Acad Pediatr. 2009;9(4):270–7.

33.

Pagano M, Cassidy L, Little M, Murphy JM, Jellinek MS. Identifying psychosocial
dysfunction in school-age children: the pediatric symptom checklist as a self-report
measure. Psychol Sch. 2000;37(2).

34.

Guseman EH, Eisenmann JC, Cadieux A, Stratbucker W. Changes in psychosocial
functioning after treatment for pediatric obesity. In Preparation. 2012

35.

Gardner W, Kelleher KJ, Pajer KA. Multidimensional adaptive testing for mental health
problems in primary care. Med Care. 2002 Sep;40(9):812–23.

36.

Gibson L, Byrne S, Blair E. Clustering of psychosocial symptoms in overweight children.
Aust N Z J Psychiatry. 2008;42:118–125.

131

37.

Braet C, Mervielde I. Psychological Aspects of Childhood Obesity: A Controlled Study in
a Clinical and Nonclinical Sample. J Pediatr Psychol. 1997;22(1):59–71.

38.

Van Vlierberghe L, Braet C, Goossens L, Mels S. Psychiatric disorders and symptom
severity in referred versus non-referred overweight children and adolescents. Eur Child
Adolesc Psychiatry. 2009 Mar;18(3):164–73.

39.

Björntorp P. Do stress reactions cause abdominal obesity and comorbidities? Obes Rev.
2001 May;2(2):73–86.

40.

Selye H. Stress and the general adaptation syndrome. Br Med J. 1950;1(4667):1383–1392.

41.

Kyrou I, Chrousos GP, Tsigos C. Stress, visceral obesity, and metabolic complications.
Ann N Y Acad Sci. 2006 Nov;1083:77–110.

42.

Barat P, Gayard-Cros M, Andrew R, Corcuff J-B, Jouret B, Barthe N, Perez P, Germain
C, Tauber M, Walker BR, Mormede P, Duclos M. Truncal distribution of fat mass,
metabolic profile and hypothalamic-pituitary adrenal axis activity in prepubertal obese
children. J Pediatr. 2007 May;150(5):535–9, 539.e1.

43.

Weigensberg MJ, Toledo-Corral CM, Goran MI. Association between the metabolic
syndrome and serum cortisol in overweight Latino youth. J Clin Endocrinol Metab. 2008
Apr;93(4):1372–8.

44.

Adam TC, Hasson RE, Ventura EE, Toledo-Corral C, Le K-A, Mahurkar S, Lane CJ,
Weigensberg MJ, Goran MI. Cortisol is negatively associated with insulin sensitivity in
overweight Latino youth. J Clin Endocrinol Metab. 2010 Oct;95(10):4729–35.

45.

Soros A, Zadik Z, Chalew S. Adaptive and maladaptive cortisol responses to pediatric
obesity. Med Hypotheses. 2008 Sep;71(3):394–8.

46.

Ward AMV, Fall CHD, Stein CE, Kumaran K, Veena SR, Wood PJ, Syddall HE, Phillips
DIW. Cortisol and the metabolic syndrome in South Asians. Clincal Endocrinol (Oxf).
2003 Apr;58(4):500–5.

47.

Hill EE, Eisenmann JC, Holmes ME, Heelan KA. Morning cortisol is not associated with
truncal fatness or resting blood pressure in children: cross-sectional and 1-2 year followup analyses. J Pediatr Endocrinol Metab. 2010 Oct;23(10):1031–7.

48.

Hill EE, Eisenmann JC, Gentile D, Holmes ME, Walsh D. The association between
morning cortisol and adiposity in children varies by weight status. J Pediatr Endocrinol
Metab. 2011 Jan;24(9-10):709–713.

49.

Tanaka H, Borres M, Thulesius O, Tamai H, Ericson MO, Lindblad L-E. Evidence of
decreased sympathetic function in children with psychosomatic symptoms. Clin Auton
Res. 2002 Dec;12(6):477–82.

132

50.

Licht CMM, Vreeburg S a, van Reedt Dortland AKB, Giltay EJ, Hoogendijk WJG,
DeRijk RH, Vogelzangs N, Zitman FG, de Geus EJC, Penninx BWJH. Increased
sympathetic and decreased parasympathetic activity rather than changes in hypothalamicpituitary-adrenal axis activity is associated with metabolic abnormalities. J Clin
Endocrinol Metab.2010 May;95(5):2458–66.

51.

Ball GDC, Huang TTK, Cruz ML, Shaibi GQ, Weigensberg MJ, Goran MI. Predicting
abdominal adipose tissue in overweight Latino youth. Int J Pediatr Obes. 2006
Jan;1(4):210–6.

52.

Ekelund U, Anderssen SA, Froberg K, Sardinha LB, Andersen LB, Brage S. Independent
associations of physical activity and cardiorespiratory fitness with metabolic risk factors
in children: the European youth heart study. Diabetologia. 2007 Sep;50(9):1832–40.

53.

Pan Y, Pratt CA. Metabolic syndrome and its association with diet and physical activity in
US adolescents. J Am Diet Assoc.2008 Feb;108(2):276–86.

54.

Martínez-Gómez D, Eisenmann JC, Moya JM, Gómez-Martínez S, Marcos A, Veiga OL.
The role of physical activity and fitness on the metabolic syndrome in adolescents: effect
of different scores. The AFINOS Study. J Physiol Biochem. 2009 Sep;65(3):277–89.

55.

Nguyen THHD, Tang HK, Kelly P, van der Ploeg HP, Dibley MJ. Association between
physical activity and metabolic syndrome: a cross sectional survey in adolescents in Ho
Chi Minh City, Vietnam. BMC Public Health. 2010 Jan;10:141.

56.

Brage S, Wedderkopp N, Ekelund U, Franks PW, Wareham NJ, Andersen LB, Froberg K.
Features of the metabolic syndrome are associated with objectively measured physical
activity and fitness in Danish children: the European Youth Heart Study (EYHS).
Diabetes Care. 2004 Sep;27(9):2141–8.

57.

Ekelund U, Brage S, Froberg K, Harro M, Anderssen S a, Sardinha LB, Riddoch C,
Andersen LB. TV viewing and physical activity are independently associated with
metabolic risk in children: the European Youth Heart Study. PLoS Med. 2006
Dec;3(12):e488.

58.

Yin Z, Davis CL, Moore JB, Treiber F a. Physical activity buffers the effects of chronic
stress on adiposity in youth. Ann Behav Med. 2005 Feb;29(1):29–36.

59.

Schechter JC, Brennan P a, Cunningham PB, Foster SL, Whitmore E. Stress, cortisol, and
externalizing behavior in adolescent males: an examination in the context of
multisystemic therapy. J Abnorm Child Psychol. 2012 Aug;40(6):913–22.

60.

Ouellet-morin I, Odgers CL, Danese A, Bowes L, Shakoor S, Papadopoulos AS, Caspi A,
Moffitt TE, Arseneault L. Blunted cortisol responses to stress signal social and behavioral
problems among maltreated/bullied 12-year-old children. Biol Psychiatry.
2011;70(11):1016-1023.

133

61.

Pruessner J, Wolf O, Hellhammer D. Free cortisol levels after awakening: a reliable
biological marker for the assessment of adrenocortical activity. Life Sci. 1997;61(26).

62.

Barnett JB, Woods MN, Lamon-Fava S, Schaefer EJ, McNamara JR, Spiegelman D,
Hertzmark E, Goldin B, Longcope C, Gorbach SL. Plasma lipid and lipoprotein levels
during the follicular and luteal phases of the menstrual cycle. J Clin Endocrinol Metab.
2004 Feb;89(2):776–82.

63.

Fulghesu A, Magnini R, Portoghese E, Angioni S, Minerba L, Melis GB. Obesity-related
lipid profile and altered insulin incretion in adolescents with polycystic ovary syndrome. J
Adolesc Health. 2010 May;46(5):474–81.

64.

Hart R, Doherty D a, Mori T, Huang R-C, Norman RJ, Franks S, Sloboda D, Beilin L,
Hickey M. Extent of metabolic risk in adolescent girls with features of polycystic ovary
syndrome. Fertil Steril. 2011 Jun;95(7):2347–53, 2353.e1.

65.

Du Y, Rosner BM, Knopf H, Schwarz S, Dören M, Scheidt-Nave C. Hormonal
contraceptive use among adolescent girls in Germany in relation to health behavior and
biological cardiovascular risk factors. J Adolesc Health. 2011 Apr;48(4):331–7.

66.

Appel LJ, Moore TJ, Obarzanek E, Vollmer WM, Svetkey LP, Sacks FM, Bray GA, Vogt
TM, Cutler JA, Windhauser MM, Lin PH, Karanja N. A clinical trial of the effects of
dietary patterns on blood pressure. DASH Collaborative Research Group. N Engl J Med.
1997 Apr;336(16):1117–24.

67.

Azadbakht L, Mirmiran P, Esmaillzadeh A, Azizi T, Azizi F. Beneficial effects of a
Dietary Approaches to Stop Hypertension eating plan on features of the metabolic
syndrome. Diabetes Care. 2005 Dec;28(12):2823–31.

68.

Barnard ND, Katcher HI, Jenkins DJA, Cohen J, Turner-McGrievy G. Vegetarian and
vegan diets in type 2 diabetes management. Nutr Rev. 2009 May;67(5):255–63.

69.

Ludwig DS, Pereira MA, Kroenke CH, Hilner JE, Van Horn L, Slattery ML, Jacobs DR.
Dietary fiber, weight gain, and cardiovascular disease risk factors in young adults.
JAMA. 1999 Oct;282(16):1539–46.

70.

Deshmukh-Taskar PR, Nicklas TA, O’Neil CE, Keast DR, Radcliffe JD, Cho S. The
relationship of breakfast skipping and type of breakfast consumption with nutrient intake
and weight status in children and adolescents: the National Health and Nutrition
Examination Survey 1999-2006. J Am Diet Assoc. 2010 Jun;110(6):869–78.

71.

Carlson JJ, Eisenmann JC, Norman GJ, Ortiz KA, Young PC. Dietary fiber and nutrient
density are inversely associated with the metabolic syndrome in US adolescents. J Am
Diet Assoc. 2011 Nov;111(11):1688–95.

134

72.

Moore LL, Singer MR, Qureshi MM, Bradlee ML. Dairy Intake and Anthropometric
Measures of Body Fat among Children and Adolescents in NHANES. J Am Coll Nutr.
2008 Dec;27(6):702–710.

73.

Bradlee ML, Singer MR, Qureshi MM, Moore LL. Food group intake and central obesity
among children and adolescents in the Third National Health and Nutrition Examination
Survey (NHANES III). Public Health Nutr. 2010 Jun;13(6):797–805.

74.

National Insitutes of Health. Nutrition for Patients with Cushing Syndrome. Bethesda:
1997.

75.

Warne JP. Shaping the stress response: interplay of palatable food choices,
glucocorticoids, insulin and abdominal obesity. Mol Cell Endocrinol. 2009 Mar;300(12):137–46.

76.

Vedhara K, Tuinstra J, Stra JT, Miles JNV, Sanderman R, Ranchor AV. Psychosocial
factors associated with indices of cortisol production in women with breast cancer and
controls. Psychoneuroendocrinology. 2006 Apr;31(3):299–311.

77.

Tremblay MS, LeBlanc AG, Kho ME, Saunders TJ, Larouche R, Colley RC, Goldfield G,
Gorber SC. Systematic review of sedentary behaviour and health indicators in school-aged
children and youth. Int J Behav Nutr Phys Act. 2011 Jan;8(1):98.

135

APPENDICES

APPENDIX A

INFORMED CONSENT – HEALTHY WEIGHT CENTER

137

This is the informed consent document that was approved by the Spectrum Health IRB and
Michigan State University IRB for participants recruited from the Healthy Weight Center.
Permission to Take Part in a Human Research Study &
HIPAA Authorization for Release of Health Information for
Research Purposes
Title of research study: Physical activity, stress, and metabolic syndrome in youth.
Investigator:

Emily E Hill, MA
Joey C Eisenmann, PhD

“You” refers you or your child.
“We” refers to Helen DeVos Children’s Hospital, Michigan State University, and the study
investigators.

We invite you to take part in a research study because you have been referred to the Healthy
Weight Center over concerns about your weight.

What you should know about a research study?
• Someone will explain this research study to you.
• You volunteer to be in a research study.
• Whether or not you take part is up to you.
• You can choose not to take part in the research study.
• You can agree to take part now and later change your mind.
• Whatever you decide it will not be held against you.
• Feel free to ask all the questions you want before you decide.
Who can I talk to?
If you have questions, concerns, or complaints, or think the research has hurt you talk to the
investigator or members of the research team at (616) 391-8675 or (919) 593-5851.

This research has been reviewed and approved by the Spectrum Health Institutional Review
Board. You may talk to them at (616) 486-2031 or irb@spectrum-health.org for any of the
following:

138

•
•
•
•
•

Your questions, concerns, or complaints are not being answered by the investigator or
research team.
You cannot reach the investigator or research team.
You want to talk to someone besides the investigator or research team.
You have questions about your rights as a research participant.
You want to get information or provide input about this research.

Why are we doing this research?
We would like to find out more about daily changes in a hormone called cortisol in teenagers.
Hormones help one part of the body send messages to other parts of the body. Stress (like
anxiety, illness, physical activity or temperature extremes) causes the body to produce a hormone
called cortisol. We will be measuring cortisol in saliva. The amount of cortisol the body
produces can be related to different health problems, such as metabolic syndrome.
Several factors make up the metabolic syndrome. Blood pressure is one of these factors. Others
are fats, sugar, and the hormone insulin in the blood. A mathematical formula can be used to
create a metabolic syndrome score based on these factors. A higher metabolic syndrome score is
related to a higher risk of health problems as people age.
We are interested in how the amount of cortisol in saliva changes at different times during the
day. Also, we hope to determine if these levels change based on a person’s metabolic syndrome
score, physical activity and/or stress level.
How long will I be in the research?
We expect that you will be in this research study for approximately seven days.
How many people will be studied?
We expect about 50 people will be in this research study.
What happens if I say yes, I want to be in this research?
You have already had your initial exam at the Healthy Weight Center. In addition to what you
have already been asked to do as part of your weight loss program (for example, wearing an
armband to measure your physical activity), we would like for you to collect some saliva
samples so that we can measure your cortisol level. We will ask you to collect a total of six
samples over the course of one full weekend day.
You will be given six plastic tubes with numbered labels on them. Each tube will contain a soft,
cotton-like cylinder. To collect a saliva sample, you will place one of these cylinders in your
mouth for one to two minutes. Once the sample has been collected, you will put the cylinder in
the plastic tube. We will ask you to write down the time each time you take a sample. We will
ask you to collect six samples over the course of one weekend day, beginning when you wake
up. The second sample will be taken 30 minutes after you wake up, and the remaining four
samples will be taken every three-hours based on the time you collected the first saliva sample.

139

For example, if your first saliva collection upon waking is at 7:00 AM, you will collect samples
as follows:
Figure A.1 Example timeline for collection of saliva samples.
7:00 AM

7:30 AM

10:00 AM

1:00 PM

4:00 PM

7:00 PM

First sample

You should not eat a meal within 1 hour before sample collection. You should also thoroughly
rinse your mouth with water 10 minutes before collection. You will also be given a saliva
collection log with detailed instructions. As soon as possible after collection, the samples should
be placed in your home freezer. Upon completion of the seven-day period, please return the
saliva samples and the armband to the researchers via mail. You will be provided with a prepaid container for this purpose.
We would also like to use information from your medical records (for example lab values,
weight, blood pressure) for this study. As part of your routine care through the Healthy Weight
Center you have already agreed to wear an armband to measure physical activity over a sevenday period. The information captured through the band will be used for this study as well. All
information will be kept confidential.
What happens if I say no, I do not want to be in this research?
You may decide not to take part in the research and it will not be held against you. A refusal to
participate will involve no penalty or loss of benefits to which you are otherwise entitled.
What happens if I say yes, but I change my mind later?
You can agree to take part in the research now and stop at any time it will not be held against
you.
Discontinuing participation will not result in penalty or loss of benefits to which you are
otherwise entitled.
If you stop being in the research, already collected data will be retained in the study database. If
you are participating in programs at the Healthy Weight Center, being in this study will not
affect your participation in those programs.
Could being in this study be bad for me?
There are no risks involved in this study.
Will I need to pay for any of the tests or procedures in the study?

140

There are no costs associated with participation in this study. Procedures involved in saliva
collection and analysis will not be billed to your insurance.
Will being in this study help me in any way?
There are no benefits to you from your taking part in this research. We cannot promise any
benefits to others from your taking part in this research. However, possible benefits to others
include gaining a better understanding of the relationship between physical activity and cortisol
level.
What happens to the information you collect?
Efforts will be made to limit your personal information, including research study and medical
records, to people who have a need to review this information. We cannot promise complete
confidentiality. Organizations that may inspect and copy your information for quality assurance
and data analysis include:
• The Investigator and his/her research staff
• Spectrum Health staff or its agents
• The Spectrum Health Institutional Review Board (IRB) and staff
• The Michigan State University Institutional Review Board (IRB) and staff
Some of these organizations may be given direct access to your medical records for verification
of the research procedures/data involved. By signing this document you are authorizing this
access.
Federal law provides additional protections of your personal information. These are described in
a later section.
Can I be removed from the research without my OK?
The person in charge of the research study can remove you from the research study without your
approval. A possible reason for removal includes failure to follow study procedures.
What else do I need to know?
Payment for participating will be provided in the form of incentives awarded for completion of
data collection. You will be awarded a small incentive (approximately $1 in value), such as a
code for one music download, for each requested saliva sample which was collected at the
appropriate time, as determined from the saliva data collection log. Upon completion of all
study data collection, you will be mailed one $25 gift card good for use at Walmart, Meijer, or a
similar store. In order to receive the $25 gift card, you must collect at least 3 saliva samples
(immediately upon waking, 30-minutes after waking, and final evening time point) and wear the
armband for at least 10 hours on at least 4 days.

HIPAA Authorization for Release of Health Information for Research Purposes

141

The information we are asking to use and share is called Protected Health Information (PHI). It
is protected by a federal law called the Privacy Rule of the Health Insurance Portability and
Accountability Act (HIPAA). In general, we cannot use or share your health information for
research without your permission.
What will be done with my information?
Your health information will be collected and entered in a database along with the information
from other people taking part in this study.
Why am I being asked to release it?
The purpose of our study is to determine how the amount of cortisol in saliva changes at
different times during the day. Also, we hope to determine if these levels change based on a
person’s metabolic syndrome score. We will also be measuring regular physical activity to see if
this changes cortisol levels.
What will be released?
To complete this research study, we will need to collect information about you. This
information may include:
•
•

Existing medical records and medical history.
New health information collected for purposes of this study.

Who will use it or share it?
Your information will be released by Dr. William Stratbucker. Dr. Stratbucker is the
medical director of the Healthy Weight Center at Helen DeVos Children’s Hospital. The
researcher, Emily E Hill, MA, will receive your information. Other people may see your
information. They include:
•
•
•
•
•

Spectrum Health staff, employees, or other agents
Agencies that accredit the hospital or the research program
The study doctor and his or her staff and study personnel.
The Spectrum Health Research and Human Rights Committee.
The Michigan State University Institutional Review Board.

Once your protected health information has been disclosed it is possible that someone may
share it again. Some people who receive your information may not be required by law to
keep it confidential. We cannot guarantee that your information will not be released or made
available to another party once it leaves Spectrum Health. We will share your information
only if necessary. We also use all reasonable efforts to request that other individuals who
receive your information protect your privacy.

142

How long will my health information be used?
This authorization has no expiration date.
Can I stop my protected health information from being collected?
You can cancel this authorization at any time. Your cancellation must be in writing. Once
you cancel your authorization, we will stop collecting your medical information except in
very limited cases if needed to comply with law, protect your safety, or make sure the
research was done properly. Any information that was collected before you stopped your
authorization will still be used as described above.
If you decide to stop the collection of your protected health information for this study, you
must send a written notice to:
William Stratbucker, MD
Healthy Weight Center MC232
330 Barclay NE Ste 303
Grand Rapids, MI 49503
Phone: (616) 391-7999
If you have more questions about the release of your health information, you may contact the
Spectrum Health HIPAA/Privacy and Information Security team at (616) 486-4113 or
patient.privacy@spectrum-health.org.
What happens if I do not want you to collect and release my information?
If you decide not to authorize release of your health information as part of this study, your
decision will in no way affect your medical care or cause you to lose any benefits to which you
are entitled. You cannot participate in this research study if you do not authorize the use or
release of your PHI.
When will it be destroyed?
Your information will be destroyed three years after study completion.
Your signature below documents your permission to take part in this research and to the use and
disclosure of your protected health information. You will receive a signed copy of this complete
form.

Signature of participant

Date

Printed name of participant

143

Signature of person obtaining consent

Date

Printed name of person obtaining consent

Your signature below documents your permission for the child named below to take part in this
research and to the use and disclosure of this child’s protected health information. You will
receive a signed copy of this complete form.

Printed name of child

Signature of parent or guardian

Date

q Parent
q Guardian (See note below)
Note on permission by guardians: An individual may provide permission for a child only if that
individual can provide a written document indicating that he or she is legally authorized to
consent to the child’s general medical care. Attach the documentation to the signed document.
Printed name of parent or guardian

Signature of person obtaining consent

Date

Assent

Printed name of person obtaining consent

q Verbal Assent Obtained
q Not obtained because the capability of the child is so limited that the child cannot
reasonably be consulted.

144

APPENDIX B:

ASSENT FOR CHILDREN – HEALTHY WEIGHT CENTER

145

This is the assent for children document that was approved by the Spectrum Health IRB and
Michigan State University IRB for participants recruited from the Healthy Weight Center.
RESEARCH ASSENT FOR
Physical activity, stress, and metabolic syndrome in youth
Research Assent for Children Ages 12-17

Principal Investigator:
Affiliation:

Emily E Hill, MA
Joey C Eisenmann, PhD
Michigan State University
Helen DeVos Children’s Hospital

What is a research study and why is this research study being done?
A research study is a way to learn more about something. For example, a research study might
involve doctors looking at something new, such as a new medicine or new treatment to see how
well it works. We would like to find out more about daily changes in a hormone called cortisol
in teenagers. Hormones help one part of the body send messages to other parts of the body.
Stress (like anxiety, illness, physical activity or temperature extremes) causes the body to
produce a hormone called cortisol. We will be measuring cortisol in saliva. The amount of
cortisol the body produces can be related to different health problems, such as metabolic
syndrome.
Several factors make up the metabolic syndrome. Blood pressure is one of these factors. Others
are fats, sugar, and the hormone insulin in the blood. A mathematical formula can be used to
create a metabolic syndrome score based on these factors. A higher metabolic syndrome score is
related to a higher risk of health problems as people age.
We are interested in how the amount of cortisol in saliva changes at different times during the
day. Also, we hope to determine if these levels change based on a person’s metabolic syndrome
score, physical activity and/or stress level.
Why am I being asked to be in this study?
You are being asked to be in this study because you have been referred to the Healthy Weight
Center over concerns about your weight.
How long will I be in the research?
We expect that you will be in this research study for approximately seven days.
How many people will be studied?
We expect about 50 people will be in this research study.

146

If I decide to join this research study, what will I have to do and what will happen to me?
You have already had your initial exam at the Healthy Weight Center. In addition to what you
have already been asked to do as part of your weight loss program (for example, wearing an
armband to measure your physical activity), we would like for you to collect some saliva
samples so that we can measure your cortisol level. We will ask you to collect a total of six
samples over the course of one full weekend day.
You will be given six plastic tubes with numbered labels on them. Each tube will contain a soft,
cotton-like cylinder. To collect a saliva sample, you will place one of these cylinders in your
mouth for one to two minutes. Once the sample has been collected, you will put the cylinder in
the plastic tube. We will ask you to write down the time each time you take a sample. We will
ask you to collect six samples over the course of one weekend day, beginning when you wake
up. The second sample will be taken 30 minutes after you wake up, and the remaining four
samples will be taken every three-hours based on the time you collected the first saliva sample.
For example, if you wake up at 7:00 AM, you will collect samples as follows:
Figure B.1 Example timeline for collection of saliva samples.

7:00 AM

7:30 AM

10:00 AM

1:00 PM

4:00 PM

7:00 PM

First sample

You should not eat a meal within one hour before sample collection. You should also
thoroughly rinse your mouth with water 10 minutes before collection. You will also be given a
saliva collection log with detailed instructions and asked to write down the time when you
collect each sample. As soon as possible after collection, the samples should be placed in your
home freezer. Upon completion of the seven-day period, please return the saliva samples and the
armband to the researchers via mail. You will be provided with a pre-paid container for this
purpose.
We would also like to use information from your medical records (for example lab values,
weight, blood pressure) for this study. As part of your routine care through the Healthy Weight
Center you have already agreed to wear an armband to measure physical activity over a sevenday period. The information captured through the band will be used for this study as well. All
information will be kept confidential.

147

Will anything good happen to me if I join this study?
There are no benefits to you from your taking part in this research. We cannot promise any
benefits to others from your taking part in this research. However, possible benefits to others
include gaining a better understanding of the relationship between physical activity and cortisol
level.
Will I be hurt if I am in the study?
There are no risks involved with this study. If you feel sick or are afraid that something is wrong
with you, tell your doctor, nurse or an adult at once.
Will I be paid if I join this study?
Payment for participating will be provided in the form of incentives awarded for completion of
data collection. You will be awarded a small incentive (approximately $1 in value), such as a
code for one music download, for each requested saliva sample which was collected at the
appropriate time, as determined from the saliva data collection log. Upon completion of all
study data collection, you will be mailed one $25 gift card good for use at Walmart, Meijer, or a
similar store. In order to receive the $25 gift card, you must collect at least 3 saliva samples
(immediately upon waking, 30-minutes after waking, and final evening time point) and wear the
armband for at least 10 hours on at least 4 days.
Will people know I am in this study?
Your family and your doctor, nurse and research staff will know that you are in this study. If
anyone else is given information about you, they will not know your name. A number will be
used instead of your name. When we are finished with this study we might write a report about
what was learned during this study. This report will not include your name or that you were in
the study.
What if I do not want to be in this study?
You do not have to join this study. It is your choice. If you say yes now, you can still change
your mind later. No one will be mad at you if you don’t want to be in the study or if you join the
study and change your mind later and stop.
Before you say yes or no to being in this study, we will answer any questions you have. If you
join the study, you can ask questions at any time. Just tell the doctor or nurse you have a
question.
I have read about this research study and my questions have been answered.
I have decided:
q Yes, I want to be in the study
q No, I do not want to be in the study

148

Printed Name of Child

Signature of Child (printed name only is acceptable)

Date of Assent

Printed Name of Person Obtaining Assent

Signature of Person Obtaining Assent

Date

149

APPENDIX C:

INFORMED CONSENT – ACADEMIC GENERAL PEDIATRICS

150

This is the informed consent document that was approved by the Spectrum Health IRB and
Michigan State University IRB for participants recruited from Academic General Pediatrics.
Permission to Take Part in a Human Research Study &
HIPAA Authorization for Release of Health Information for
Research Purposes
General Pediatrics
Title of research study: Physical activity, stress, and metabolic syndrome in youth.
Investigator:

Emily E Hill, MA
Joey C Eisenmann, PhD

“You” refers you or your child.
“We” refers to Helen DeVos Children’s Hospital, Michigan State University, and the study
investigators.
We invite you to take part in a research study because you have a BMI at or above the 95th
percentile for your age.
What you should know about a research study?
• Someone will explain this research study to you.
• You volunteer to be in a research study.
• Whether or not you take part is up to you.
• You can choose not to take part in the research study.
• You can agree to take part now and later change your mind.
• Whatever you decide it will not be held against you.
• Feel free to ask all the questions you want before you decide.
Who can I talk to?
If you have questions, concerns, or complaints, or think the research has hurt you talk to the
investigator or members of the research team at (616) 391-8675 or (919) 593-5851.
This research has been reviewed and approved by the Spectrum Health Institutional Review
Board. You may talk to them at (616) 486-2031 or irb@spectrum-health.org for any of the
following:
• Your questions, concerns, or complaints are not being answered by the investigator or
research team.
• You cannot reach the investigator or research team.
• You want to talk to someone besides the investigator or research team.
• You have questions about your rights as a research participant.

151

•

You want to get information or provide input about this research.

Why are we doing this research?
We would like to find out more about daily changes in a hormone called cortisol in teenagers.
Hormones help one part of the body send messages to other parts of the body. Stress (like
anxiety, illness, physical activity or temperature extremes) causes the body to produce a hormone
called cortisol. We will be measuring cortisol in saliva. The amount of cortisol the body
produces can be related to different health problems, such as metabolic syndrome.
Several factors make up the metabolic syndrome. Blood pressure is one of these factors. Others
are fats, sugar, and the hormone insulin in the blood. A mathematical formula can be used to
create a metabolic syndrome score based on these factors. A higher metabolic syndrome score is
related to a higher risk of health problems as people age.
We are interested in how the amount of cortisol in saliva changes at different times during the
day. Also, we hope to determine if these levels change based on a person’s metabolic syndrome
score, physical activity and/or stress level.
How long will I be in the research?
We expect that you will be in this research study for approximately seven days.
How many people will be studied?
We expect about 50 people will be in this research study.
What happens if I say yes, I want to be in this research?
Some of the procedures you have already completed as part of your well child visit will also be
used for this research. These include measurement of your height, weight, and blood pressure,
and blood work that has been ordered by your doctor. We would also like to measure your waist
circumference, which will be completed in a private exam room by one of the researchers. In
addition to what you have already been asked to do, we would like for you to collect some saliva
samples so that we can measure your cortisol level. We will ask you to collect a total of six
samples over the course of one full weekend day.
You will be given six plastic tubes with numbered labels on them. Each tube will contain a soft,
cotton-like cylinder. To collect a saliva sample, you will place one of these cylinders in your
mouth for one to two minutes. Once the sample has been collected, you will put the cylinder in
the plastic tube. We will ask you to write down the time each time you take a sample. We will
ask you to collect six samples over the course of one weekend day, beginning when you wake
up. The second sample will be taken 30 minutes after you wake up, and the remaining four
samples will be taken every three-hours based on the time you collected the first saliva sample.

For example, if your first saliva collection upon waking is at 7:00 AM, you will collect samples
as follows:
152

Figure C.1 Example timeline for collection of saliva samples.
7:00 AM

7:30 AM

10:00 AM

1:00 PM

4:00 PM

7:00 PM

First sample

You should not eat a meal within 1 hour before sample collection. You should also thoroughly
rinse your mouth with water 10 minutes before collection. You will also be given a saliva
collection log with detailed instructions. As soon as possible after collection, the samples should
be placed in your home freezer. Upon completion of the seven-day period, please return the
saliva samples and the armband to the researchers via mail. You will be provided with a prepaid container for this purpose.
To measure how much physical activity you do on a normal day, we will ask you to wear an
armband for one week. This armband uses sensors against the skin to measure changes in skin
temperature, sweating, and heart rate. It also contains a computer chip that senses movement.
The sensors are not sticky and will not prick your skin. You will be given detailed instructions
for wearing the armband. A researcher will discuss these instructions with you and be available
to answer any questions.
We will also ask you to complete two surveys. The first, the Pediatric Symptoms Checklist, tells
us about troubles you might experience as part of your daily life. The second is a 3-day diet
record. To complete this, we will ask you to write down everything you eat for 3 days.
We would also like to use information from your medical records (for example lab values,
weight, blood pressure) for this study. All information will be kept confidential.
What happens if I say no, I do not want to be in this research?
You may decide not to take part in the research and it will not be held against you. A refusal to
participate will involve no penalty or loss of benefits to which you are otherwise entitled.
What happens if I say yes, but I change my mind later?
You can agree to take part in the research now and stop at any time it will not be held against
you.
Discontinuing participation will not result in penalty or loss of benefits to which you are
otherwise entitled.
If you stop being in the research, already collected data will be retained in the study database. If
you are participating in any other programs, being in this study will not affect your participation
in those programs.

Could being in this study be bad for me?
153

There are no risks involved in this study.
Will I need to pay for any of the tests or procedures in the study?
There are no costs associated with participation in this study. Procedures involved in saliva
collection and analysis will not be billed to your insurance.
Will being in this study help me in any way?
There are no benefits to you from your taking part in this research. We cannot promise any
benefits to others from your taking part in this research. However, possible benefits to others
include gaining a better understanding of the relationship between physical activity and cortisol
level.
What happens to the information you collect?
Efforts will be made to limit your personal information, including research study and medical
records, to people who have a need to review this information. We cannot promise complete
confidentiality. Organizations that may inspect and copy your information for quality assurance
and data analysis include:
• The Investigator and his/her research staff
• Spectrum Health staff or its agents
• The Spectrum Health Institutional Review Board (IRB) and staff
• The Michigan State University Institutional Review Board (IRB) and staff
Some of these organizations may be given direct access to your medical records for verification
of the research procedures/data involved. By signing this document you are authorizing this
access.
Federal law provides additional protections of your personal information. These are described in
a later section.
Can I be removed from the research without my OK?
The person in charge of the research study can remove you from the research study without your
approval. A possible reason for removal includes failure to follow study procedures.
What else do I need to know?
Payment for participating will be provided in the form of incentives awarded for completion of
data collection. You will be awarded a small incentive (approximately $1 in value), such as a
code for one music download, for each requested saliva sample which was collected at the
appropriate time, as determined from the saliva data collection log. Upon completion of all
study data collection, you will be mailed one $25 gift card good for use at Walmart, Meijer, or a
similar store. In order to receive the $25 gift card, you must collect at least 3 saliva samples
(immediately upon waking, 30-minutes after waking, and final evening time point) and wear the
armband for at least 10 hours on at least 4 days.

154

HIPAA Authorization for Release of Health Information for Research Purposes
The information we are asking to use and share is called Protected Health Information (PHI). It
is protected by a federal law called the Privacy Rule of the Health Insurance Portability and
Accountability Act (HIPAA). In general, we cannot use or share your health information for
research without your permission.
What will be done with my information?
Your health information will be collected and entered in a database along with the information
from other people taking part in this study.
Why am I being asked to release it?
The purpose of our study is to determine how the amount of cortisol in saliva changes at
different times during the day. Also, we hope to determine if these levels change based on a
person’s metabolic syndrome score. We will also be measuring regular physical activity to see if
this changes cortisol levels.
What will be released?
To complete this research study, we will need to collect information about you. This
information may include:
•
•

Existing medical records and medical history.
New health information collected for purposes of this study.

Who will use it or share it?
Your information will be released by Dr. William Stratbucker. Dr. Stratbucker is the
medical director of the Healthy Weight Center and a pediatrician at Helen DeVos Children’s
Hospital. The researcher, Emily E Hill, MA, will receive your information. Other people may
see your information. They include:
•
•
•
•
•

Spectrum Health staff, employees, or other agents
Agencies that accredit the hospital or the research program
The study doctor and his or her staff and study personnel.
The Spectrum Health Research and Human Rights Committee.
The Michigan State University Institutional Review Board.

Once your protected health information has been disclosed it is possible that someone may
share it again. Some people who receive your information may not be required by law to
keep it confidential. We cannot guarantee that your information will not be released or made
available to another party once it leaves Spectrum Health. We will share your information
only if necessary. We also use all reasonable efforts to request that other individuals who
receive your information protect your privacy.
How long will my health information be used?
155

This authorization has no expiration date.
Can I stop my protected health information from being collected?
You can cancel this authorization at any time. Your cancellation must be in writing. Once
you cancel your authorization, we will stop collecting your medical information except in
very limited cases if needed to comply with law, protect your safety, or make sure the
research was done properly. Any information that was collected before you stopped your
authorization will still be used as described above.
If you decide to stop the collection of your protected health information for this study, you
must send a written notice to:
William Stratbucker, MD
Healthy Weight Center MC232
330 Barclay NE Ste 303
Grand Rapids, MI 49503
Phone: (616) 391-7999
If you have more questions about the release of your health information, you may contact the
Spectrum Health HIPAA/Privacy and Information Security team at (616) 486-4113 or
patient.privacy@spectrum-health.org.
What happens if I do not want you to collect and release my information?
If you decide not to authorize release of your health information as part of this study, your
decision will in no way affect your medical care or cause you to lose any benefits to which you
are entitled. You cannot participate in this research study if you do not authorize the use or
release of your PHI.
When will it be destroyed?
Your information will be destroyed three years after study completion.
Your signature below documents your permission to take part in this research and to the use and
disclosure of your protected health information. You will receive a signed copy of this complete
form.

Signature of participant

Date

Printed name of participant

Signature of person obtaining consent

156

Date

Printed name of person obtaining consent

Your signature below documents your permission for the child named below to take part in this
research and to the use and disclosure of this child’s protected health information. You will
receive a signed copy of this complete form.

Printed name of child

Signature of parent or guardian

Date

q Parent
q Guardian (See note below)
Note on permission by guardians: An individual may provide permission for a child only if that
individual can provide a written document indicating that he or she is legally authorized to
consent to the child’s general medical care. Attach the documentation to the signed document.
Printed name of parent or guardian

Signature of person obtaining consent

Date

Assent

Printed name of person obtaining consent

q Verbal Assent Obtained
q Not obtained because the capability of the child is so limited that the child cannot
reasonably be consulted.

157

APPENDIX D:

ASSENT FOR CHILDREN – ACADEMIC GENERAL PEDIATRICS

158

This is the assent for children document that was approved by the Spectrum Health IRB and
Michigan State University IRB for participants recruited from Academic General Pediatrics.
RESEARCH ASSENT FOR
Physical activity, stress, and metabolic syndrome in youth
Research Assent for Children Ages 12-17
General Pediatrics

Principal Investigator:
Affiliation:

Emily E Hill, MA
Joey C Eisenmann, PhD
Michigan State University
Helen DeVos Children’s Hospital

What is a research study and why is this research study being done?
A research study is a way to learn more about something. For example, a research study might
involve doctors looking at something new, such as a new medicine or new treatment to see how
well it works. We would like to find out more about daily changes in a hormone called cortisol
in teenagers. Hormones help one part of the body send messages to other parts of the body.
Stress (like anxiety, illness, physical activity or temperature extremes) causes the body to
produce a hormone called cortisol. We will be measuring cortisol in saliva. The amount of
cortisol the body produces can be related to different health problems, such as metabolic
syndrome.
Several factors make up the metabolic syndrome. Blood pressure is one of these factors. Others
are fats, sugar, and the hormone insulin in the blood. A mathematical formula can be used to
create a metabolic syndrome score based on these factors. A higher metabolic syndrome score is
related to a higher risk of health problems as people age.
We are interested in how the amount of cortisol in saliva changes at different times during the
day. Also, we hope to determine if these levels change based on a person’s metabolic syndrome
score, physical activity and/or stress level.
Why am I being asked to be in this study?
You are being asked to be in this study because you have a BMI at or above the 95th percentile
for your age.
How long will I be in the research?
We expect that you will be in this research study for approximately seven days.
How many people will be studied?

159

We expect about 50 people will be in this research study.
If I decide to join this research study, what will I have to do and what will happen to me?
Some of the procedures you have already completed as part of your well child visit will also be
used for this research. These include measurement of your height, weight, and blood pressure,
and blood work that has been ordered by your doctor. We would also like to measure your waist
circumference, which will be completed in a private exam room by one of the researchers. In
addition to what you have already been asked to do, we would like for you to collect some saliva
samples so that we can measure your cortisol level. We will ask you to collect a total of six
samples over the course of one full weekend day.
You will be given six plastic tubes with numbered labels on them. Each tube will contain a soft,
cotton-like cylinder. To collect a saliva sample, you will place one of these cylinders in your
mouth for one to two minutes. Once the sample has been collected, you will put the cylinder in
the plastic tube. We will ask you to write down the time each time you take a sample. We will
ask you to collect six samples over the course of one weekend day, beginning when you wake
up. The second sample will be taken 30 minutes after you wake up, and the remaining four
samples will be taken every three-hours based on the time you collected the first saliva sample.
For example, if you wake up at 7:00 AM, you will collect samples as follows:
Figure D.1 Example timeline for collection of saliva samples.
7:00 AM

7:30 AM

10:00 AM

1:00 PM

4:00 PM

7:00 PM

First sample

You should not eat a meal within one hour before sample collection. You should also
thoroughly rinse your mouth with water 10 minutes before collection. You will also be given a
saliva collection log with detailed instructions and asked to write down the time when you
collect each sample. As soon as possible after collection, the samples should be placed in your
home freezer. Upon completion of the seven-day period, please return the saliva samples and the
armband to the researchers via mail. You will be provided with a pre-paid container for this
purpose.
To measure how much physical activity you do on a normal day, we will ask you to wear an
armband for one week. This armband uses sensors against the skin to measure changes in skin
temperature, sweating, and heart rate. It also contains a computer chip that senses movement.
The sensors are not sticky and will not prick your skin. You will be given detailed instructions
for wearing the armband. A researcher will discuss these instructions with you and be available
to answer any questions.

160

We will also ask you to complete two surveys. The first, the Pediatric Symptoms Checklist, tells
us about troubles you might experience as part of your daily life. The second is a 3-day diet
record. To complete this, we will ask you to write down everything you eat for 3 days.

We would also like to use information from your medical records (for example lab values,
weight, blood pressure) for this study. All information will be kept confidential.
Will anything good happen to me if I join this study?
There are no benefits to you from your taking part in this research. We cannot promise any
benefits to others from your taking part in this research. However, possible benefits to others
include gaining a better understanding of the relationship between physical activity and cortisol
level.
Will I be hurt if I am in the study?
There are no risks involved with this study. If you feel sick or are afraid that something is wrong
with you, tell your doctor, nurse or an adult at once.
Will I be paid if I join this study?
Payment for participating will be provided in the form of incentives awarded for completion of
data collection. You will be awarded a small incentive (approximately $1 in value), such as a
code for one music download, for each requested saliva sample which was collected at the
appropriate time, as determined from the saliva data collection log. Upon completion of all
study data collection, you will be mailed one $25 gift card good for use at Walmart, Meijer, or a
similar store. In order to receive the $25 gift card, you must collect at least 3 saliva samples
(immediately upon waking, 30-minutes after waking, and final evening time point) and wear the
armband for at least 10 hours on at least 4 days.
Will people know I am in this study?
Your family and your doctor, nurse and research staff will know that you are in this study. If
anyone else is given information about you, they will not know your name. A number will be
used instead of your name. When we are finished with this study we might write a report about
what was learned during this study. This report will not include your name or that you were in
the study.
What if I do not want to be in this study?
You do not have to join this study. It is your choice. If you say yes now, you can still change
your mind later. No one will be mad at you if you don’t want to be in the study or if you join the
study and change your mind later and stop.
Before you say yes or no to being in this study, we will answer any questions you have. If you
join the study, you can ask questions at any time. Just tell the doctor or nurse you have a
question.
161

I have read about this research study and my questions have been answered.
I have decided:
q Yes, I want to be in the study
q No, I do not want to be in the study

Printed Name of Child

Signature of Child (printed name only is acceptable)

Date of Assent

Printed Name of Person Obtaining Assent

Signature of Person Obtaining Assent

Date

162

APPENDIX E:

DATA COLLECTION FORM

163

Data Collection Form
Subject ID:
Age:

Date:
Gender:

Inclusion Criteria:
Height
Weight

BMI
BMI %

English speaking?
Known disorder of HPA axis?
Use of glucocorticoid medication?
Eligible for study?
Consent obtained
☐ Yes ☐ No

Assent obtained

☐ Yes
☐ Yes
☐ Yes
☐ Yes
☐ Yes

☐ No
☐ No
☐ No
☐ No
☐ No

Data obtained from medical record:
Value

Date Obtained

Lipids:
Total cholesterol
HDL cholesterol
LDL cholesterol
Triglycerides
Glucose Tolerance:
Fasting glucose
Insulin
Blood Pressure
Systolic
Diastolic
Mean arterial pressure
Other:
Waist circumference
Pubertal stage
PSC score
☐ Yes

3-day diet record obtained

164

☐ No

APPENDIX F:

SALIVA COLLECTION INSTRUCTIONS AND SAMPLE COLLECTION LOG

165

Table F.1. Instructions for collection of saliva samples.
Saliva Collection Instructions
Before collecting saliva, please DO NOT:
When to take samples:
- Consume a major meal within 60 minutes
1. Right away when you wake up
- Consume dairy products within 30
2. 30-minutes after the first saliva
minutes
collection
- Consume acidic or high sugar foods
3. Three hours after the first saliva
- Brush teeth within 3 hours
collection
- Have dental work done within 24 hours
4. Three hours after the 3rd sample (6 hrs
- Drink alcohol within 24 hours of
after the first saliva collection)
collection
5. Three hours after the 4th sample (9 hrs
What to do:
after the first saliva collection)
- Rinse mouth with water
6. Three hours after the 5th sample (12
- Take sample at the directed time
hrs after the first saliva collection)
- Put a cotton swab under tongue or roll it
around across tongue. It may also be
Please use the form on the opposite side of
chewed slightly (Do not place it
this sheet to write down the time you take
each sample, and the colored shape that is on
between the cheek and gum)
- Make sure it’s very wet
each tube.
- Once saturated, place cotton inside
plastic insert inside the tube.
- Place the sample in the freezer.

Sampling Date:

_________

Wake-up Time:

_________

Figure F.1. Saliva sampling log.
Sample
Example
1

Goal
Time
7:00 AM

Actual
Time
7:06 AM

Tube
Code
Blue Square

What time did
you eat last?

Comments
Overslept

2
3
4
5
6
Do you have a cold or feel sick today? _____________________________________________

166

APPENDIX G:

THREE-DAY DIET RECORD

167

Subject ID

Day

Date

Figure G.1 Three-day diet record day 1.
FOOD DAY 1
TIME

WHAT YOU ATE AND/OR DRANK

HOW MUCH?
(Cup, Tablespoon, etc.)

Be as exact as possible. It is easier to record what you eat or drink right after doing so. If you
wait until the end of the day, you may forget something. Be sure to record water, diet soda, gum,
etc.

168

Subject ID

Day

Date

Figure G.2 Three-day diet record day 2.
FOOD DAY 2
TIME

WHAT YOU ATE AND/OR DRANK

HOW MUCH?
(Cup, Tablespoon, etc.)

Be as exact as possible. It is easier to record what you eat or drink right after doing so. If you
wait until the end of the day, you may forget something. Be sure to record water, diet soda, gum,
etc.

169

Subject ID

Day

Date

Figure G.3 Three-day diet record day 3.
FOOD DAY 3
TIME

WHAT YOU ATE AND/OR DRANK

HOW MUCH?
(Cup, Tablespoon, etc.)

Be as exact as possible. It is easier to record what you eat or drink right after doing so. If you
wait until the end of the day, you may forget something. Be sure to record water, diet soda, gum,
etc.

170

APPENDIX H:

SENSEWEAR ARMBAND INSTRUCTIONS

171

ARMBAND INSTRUCTIONS
Please wear the armband for one week (2 weekend days; 5 weekdays). In order to get a good
measurement of your physical activity the armband needs to be worn as much as possible; 10
hours per day minimum.
Please wear the armband to bed for 2 nights while sleeping.
Be sure to take the armband off for showering or other water activities such as swimming.
After a one week period, (____________ to ____________), return the armband in the mail with
your saliva samples and questionnaires.
If you have questions, please call or email Emily
Phone: 919-593-5851
Email: hillemi1@msu.edu

172

APPENDIX I:

PEDIATRIC SYMPTOM CHECKLIST – YOUTH REPORT

173

Pediatric Symptom Checklist - Youth Report (Y-PSC)
Please mark under the heading that best fits you:
Never
__

1. Complain of aches or pains............................
2. Spend more time alone.................................
3. Tire easily, little energy............................
4.
5.
6.
7.

Fidgety, unable to sit still..........................
Have trouble with teacher………………
Less interested in school.…………...
Act as if driven by motor.............................

Sometimes
___
___

___

___

Often
___
___
___
___
___
___

8. Daydream too much.....................................
9. Distract easily.......................................
10. Are afraid of new situations..........................
11. Feel sad, unhappy.....................................
12. Are irritable, angry..................................
13. Feel hopeless.........................................

___
___
___
___
___
___

14. Have trouble concentrating............................
15. Less interested in friends............................

___
___

16. Fight with other children.............................
17. Absent from school. …………………….
18. School grades dropping. ………………..
19. Down on yourself......................................
20. Visit doctor with doctor finding nothing wrong........
21. Have trouble sleeping.................................

___
___
___
___
___
___

22. Worry a lot...........................................
23. Want to be with parent more than before...............

___
___

24. Feel that you are bad.................................
25. Take unnecessary risks................................
26. Get hurt frequently...................................
27. Seem to be having less fun............................
28. Act younger than children your age....................

___
___
___
___
___

29. Do not listen to rules................................

___

174

Pediatric Symptom Checklist – Youth Report (Continued)

Please mark under the heading that best fits you:
Never
1. Complain of aches or pains............................
30. Do not show feelings..................................
31. Do not understand other people's feelings.............

Sometimes
___

Often
___
___

32. Tease others..........................................

___

33. Blame others for your troubles........................

___

34. Take things that do not belong to you.................
35. Refuse to share......................................

___
___

175