University of Southern California Dissertations and Theses

The vicious cycle of inactivity, obesity, and metabolic health consequences in at-risk pediatric populations

(USC Thesis Other)

The vicious cycle of inactivity, obesity, and metabolic health consequences in at-risk pediatric populations

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE VICIOUS CYCLE OF INACTIVITY, OBESITY, AND METABOLIC HEALTH
CONSEQUENCES IN AT-RISK PEDIATRIC POPULATIONS

by

Ya-Wen Hsu

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PREVENTIVE MEDICINE (HEALTH BEHAVIOR))

May 2011

Copyright 2011 Ya-Wen Hsu

ii

ACKNOWLEDGEMENTS
The completion of this dissertation and my program of study would not have been
possible were it not for the generous support and assistance of many people.
I sincerely thank my dissertation committee members for their guidance and
committed effort. I would like to express my deepest gratitude to my advisor, Dr. Donna
Spruijt-Metz, for her assistance and guidance in getting my graduate career started on the
right foot and providing me with the foundation for becoming a synthetic researcher. Her
creative scientific thinking has been a constant inspiration to me. I am grateful for her
support in both my academic and personal life. I am heartily thankful to my other mentor,
Dr. Chih-Ping Chou, for his encouragement and statistical wisdom. Not only was he
readily available for me, but he always patiently directed me through analysis. He has
enlightened me through his logical thinking. Our casual Chinese chatting warms up my
abroad academic journey. I appreciate Dr. Stanley Azen for his support and for imparting
his statistical knowledge in a fun and creative spirit. His class, PM510, inspired my
interests in statistics. I give a heartfelt thanks to Dr. Jennifer Unger for fine tuning the
statistics and direction of my dissertation. I am very grateful for Dr. Lawrence Palinkas,
who was so generous with his unparalleled sociocultural perspective of obesity and his
noteworthy ability to ask insightful questions. I have been fortunate to learn from Dr.
Jaimie Davis, who provided the vision and valuable feedback that increased the depth of
this dissertation.

iii

In addition to my committee, there are many teachers and colleagues who assisted me
throughout my study. A special thanks goes to Mary Ann Murphy, who helped me and
advanced my writing skills. To Dr. Selena Nguyen-Rodriguez, an extremely
knowledgeable and patient teacher. I am also thankful for the company and support of
other students in the lab: Britni Belcher, Arianna McClain, and Alicia Thornton, as well
as other colleagues: Melissa Gunning, Hee-Sung Shin, Ernest Shen, Dr. Emily Ventura,
Dr. Adar Emken, and Dr. Courtney Byrd-Williams. I am also grateful for the support of
Marny Barovich, who shepherded me with care and enthusiasm.
I also would like to thank my friends, Cathy, Pei-Chung, Hui-Ting, Maru, Sherry, Allen,
Roger, Naco, Penny, Evaline, Ting, Anita, Jason, Robin, Kevin, and Christine, who
supported me in many ways and accompanied me through the ups and downs over the
years of my study. They are like my family in the United States. My abroad study life
would have been much less memorable without them.
Finally, my very special appreciation goes to my parents, Jui-Ming Hsu and Dr.
Jung-Hsiu Chen, my brother, Dr. Che-Wei Hsu, and my late grandfather, Chin-Ho Chen,
for their heartwarming support and endless love. Their unwavering faith and confidence
in my abilities, and in me, is what shaped me into the person I am today. I would also like
to thank my cousins, Dr. Meng-Wei Wan, Meng-Lin Wan, and my aunt Jung-Li Chen.
They were always encouraging me with their best wishes. I am thankful for this
opportunity and hope that I can use what I have learned to serve others first and foremost.

iv

TABLE OF CONTENTS
ACKNOWLEDGEMENTS ............................................................................................ ii
LIST OF TABLES .......................................................................................................... vi
LIST OF FIGURES ...................................................................................................... viii
ABBREVIATIONS ......................................................................................................... ix
ABSTRACT .................................................................................................................... x
Chapter 1 INTRODUCTION.................................................................................... 1
1.1 Significance .............................................................................................................. 1
1.2 Background .............................................................................................................. 4
Pediatric Obesity and At-risk Pediatric Populations .................................................. 4
Pediatric Obesity in the United States .................................................................... 5
Pediatric Obesity in China ...................................................................................... 6
Obesity Risk in Chinese Adolescents and Public Health Implications ...................... 7
Relationships between Obesity and Metabolic Syndrome: Role of Insulin
Resistance ................................................................................................................. 10
The Metabolic Syndrome: Definition and Prevalence in Youth .............................. 11
Metabolic Risk among Latino and African American Youth in the U.S. ................ 13
Physical Activity versus Sedentary Behavior ........................................................... 14
Objective and Subjective Measures of Activity Levels............................................ 15
Activity Levels and Overweight (Study 1) ............................................................... 19
Activity Levels and the Metabolic Syndrome (Study 2) .......................................... 22
Longitudinal Relationships between Activity Levels and the Metabolic
Syndrome (Study 3) .................................................................................................. 24
1.3 Specific Aims and Hypotheses ............................................................................... 28
1.4 Overall Summary ................................................................................................... 31
1.5 Study Samples ........................................................................................................ 32
Chapter 2 CORRELATES OF OVERWEIGHT STATUS IN CHINESE
YOUTH: AN EAST-WEST PARADOX ..................................................................... 34

v

Chapter 3 INFLUENCE OF PHYSICAL ACTIVITY AND SEDENTARY
BEHAVIOR ON METABOLIC SYNDROME IN MINORITY YOUTH ............... 73
Chapter 4 LONGITUDINAL DIFFERENCES IN PHYSICAL ACTIVITY
AND SEDENTARY BEHAVIOR BY METABOLIC SYNDROME AMONG
YOUNG LATINA AND AFRICAN AMERICAN FEMALES ................................. 95
Chapter 5 SUMMARY AND CONCLUSIONS .................................................. 120
5.1 Summary of Findings ........................................................................................... 120
5.2 Strengths and Limitations .................................................................................... 122
5.3 Implications .......................................................................................................... 124
5.4 Overall Conclusions ............................................................................................. 130
BIBLIOGRAPHY ........................................................................................................ 132
APPENDICES .............................................................................................................. 160
Appendix A: Adult Definition of the Metabolic Syndrome Proposed by ATP III .... 160
Appendix B: Pediatric Definition of the Metabolic Syndrome Based on ATP III .... 161

vi

LIST OF TABLES

Table 2-1: Characteristic of Covariates of Interest by Overweight Status ....................... 55
Table 2-2: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest (N=9023) ............... 57
Table 2-3: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for ChengDu ........ 59
Table 2-4: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for HangZhou ........ 61
Table 2-5: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for ShenYang ....... 63
Table 2-6: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Wuhan ............ 65
Table 2-7: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Harbin ............ 67
Table 2-8: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Kunming ........ 69
Table 2-9: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Qingdo ........... 71
Table 3-1: Demographic Characteristics by Metabolic Syndrome ................................... 90
Table 3-2: Adjusted Odds Ratios (OR) Examining the Associations of Activity
Levels with Metabolic Syndrome ..................................................................................... 91
Table 3-3: Pearson Correlations/Partial Correlations between Activity Levels
and Features of Metabolic Syndrome ............................................................................... 93
Table 3-4: Summary of Significant Findings on Physical Activity, Sedentary
Behavior, and Metabolic Syndrome by Measurements .................................................... 94
Table 4-1: Comparisons of Baseline Demographic Characteristics between
Participants With and Without Complete Baseline Data ................................................ 113

vii

Table 4-2: Comparisons of Baseline Demographic Characteristics by Attrition Status in
Current Analytical Sample .............................................................................................. 114
Table 4-3: Baseline Demographic Characteristics .......................................................... 115
Table 4-4: Results of the Growth Curve Models to Assess the Influence of Metabolic
Syndrome on Activity Levels Measured by Accelerometer over the 1-Year Period ..... 116
Table 4-5: Results of the Growth Curve Models to Assess the Influence of Metabolic
Syndrome on Activity Levels Measured by 3-Day Physical Activity Recall (3DPAR)
over the 1-Year Period .................................................................................................... 117

viii

LIST OF FIGURES
Figure 1-1: Conceptual Model of the Vicious Circle of Inactivity, Obesity, and Metabolic
Health Consequences .......................................................................................................... 3
Figure 3-1: Activity Levels by Metabolic Syndrome ....................................................... 89
Figure 4-1: Changes in Activity Levels (as measured by Accelerometry) According to
Visit Number and Metabolic Syndrome ......................................................................... 118
Figure 4-2: Changes in Activity Levels (as measured by 3DPAR) According to Visit
Number and Metabolic Syndrome .................................................................................. 119

ix

ABBREVIATIONS
BMI=Body Mass Index
CDC=Center for Disease Control and Prevention
FSIVGTT=Frequently Sampled Intravenous Glucose Tolerance Test
GCRC= General Clinical Research Center
HDL= High Density Lipoprotein
IOTF= International Obesity Task Force
MET =Metabolic Equivalent
MetS= Metabolic Syndrome
MVPA=Moderate-to-Vigorous Physical Activity
SES=Socioeconomic Status
WHO=World Health Organization
3DPAR=3-Day Physical Activity Recall

x

ABSTRACT
Purpose:
This dissertation examined the associations between physical activity, sedentary
behavior, overweight, and the metabolic syndrome (MetS) in at-risk pediatric populations
in the United States and in China. Study 1 identified the independent influences of physical
activity, sedentary behavior, and other weight-related correlates on overweight status in
Chinese youth. Study 2 explored the influences of physical activity and sedentary behavior
on MetS in US minority youth. Study 3 compared the longitudinal trends of physical
activity and sedentary behavior between youth with and without MetS in a sample of US
minority female youth.
Methods:
Participants were youth (ages 8-18 years) in the United States and in China. Study 1
used baseline data from of a longitudinal smoking prevention and health promotion study
conducted in the 7 large cities in China for Chinese youth ages 11-18 years. Study 2 used
baseline data from three related pediatric obesity studies that share a set of common
methods and measures (US Latino and African American youth ages 8-18 years). Data
for Study 3 are from a longitudinal, observation study for Latina and African American
female youth ages 8-11 years at baseline.

xi

Results:
In Study 1, Chinese youth were more likely to be overweight if they spent more time
being sedentary, slept <7 hours/night, were male, were younger, participated more in
vigorous physical activity, had higher levels of parental education, better self-perceived
health status, a higher frequency of vegetable intake, and a lower frequency of sweet/fast
food intake. In Study 2, lower levels of moderate-to-vigorous physical activity (by
accelerometry) and higher levels of sedentary behavior (by 3-day physical activity recall)
are associated with increased the metabolic risk independent of each other and body
composition. In Study 3, a significant decline in MVPA and an increase in sedentary
behavior were observed over one year. Sedentary behavior as measured by accelerometry
increased 23.42 minutes/per quarterly visit, adding up to 93.68 minutes/per year more in
youth with MetS than in those without.
Conclusion:
Overweight-related correlates seem to play different roles in the Chinese culture
than in Western cultures. Findings from this dissertation support a vicious cycle of
increasing inactivity, obesity, and metabolic complications. These findings, coupled with
longitudinal evidence of the effects of activity levels on obesity and MetS, suggests that
physical activity and sedentary behavior could function as antecedents as well as
consequences of overweight or MetS in youth.

1
Chapter 1 INTRODUCTION
1.1 Significance
During the last two decades, there have been dramatic shifts in human environment,
behaviors, and lifestyles. These transitions have resulted in a global epidemic of pediatric
overweight(168) and have made the metabolic syndrome (MetS) a global public health
problem.(241) MetS, a suggested early indicator for type 2 diabetes and cardiovascular
disease, has gained attention recently as one of the major health consequences of the
emerging obesity epidemic.(177, 193) Behavioral correlates such as physical activity and
sedentary behavior have been indicated as precursors of both obesity(47) and MetS.(40)
However, findings regarding activity levels and obesity have been mixed, and the
relationships between physical activity levels and MetS are poorly understood in youth.
The overall goal of this dissertation is to examine the associations between physical
activity, sedentary behavior, overweight, and MetS in at-risk pediatric populations.
Specifically, this dissertation examines aspects of pediatric obesity and MetS in the
United States and China.
In the United States, minority youth, especially Latino and African American youth,
have been identified as having higher rates of obesity(154) and higher insulin resistance
than Caucasian youth, independent of adiposity.(77) The emergence of MetS has
paralleled this rise in obesity.(177, 193) In the U.S., Latino youth have been identified as
having highest prevalence of MetS. These rates vary widely from 2.0 - 9.4%, depending

2

on the definition used.(35) Nevertheless, little is known regarding relationships between
physical activity, sedentary behavior, and MetS in minority youth.(55, 57, 159, 193)
Furthermore, the majority of existing studies have assessed activity levels via either
subjective or objective measurements, but few have used both, which may provide more
comprehensive estimations in activity levels. Finally, to our knowledge, no other studies
have assessed the longitudinal associations between MetS and levels of physical activity
and sedentary behavior in Latino and African American youth.
In China, adolescent overweight and obesity have

increased substantially in the past
decade.(94) However, because much of the existing research on obesity-related correlates
(e.g., physical activity, sedentary behavior, diet, and sleep) has focused on the Western
populations, there has been limited research examining these associations in Chinese
adolescents. Because the Chinese culture and economic climates represent a distinctly
different context from Western countries, it is possible that the correlates of adiposity that
have been identified in Western youth may have different (or no) relationships to
adiposity in Chinese youth.
The conceptual model of the proposed three studies is illustrated in Figure 1-1. In
Study 1, associations between physical activity, sedentary behavior, and overweight will
be examined in Chinese youth. Study 2 aims to investigate the influences of physical
activity and sedentary behavior on MetS in the US minority youth. Study 3 will compare
the longitudinal trends of physical activity and sedentary behavior between youth with
and without MetS in the US minority populations. Together, these three studies illustrate

3

a proposed vicious circle of increasing inactivity, obesity, and worsening metabolic
health - physical activity and sedentary behavior can thus be conceptualized of as both
antecedents and consequences of overweight or MetS. This idea of a vicious cycle offers
a more comprehensive explanation for the associations between obesity, metabolic health,
and activity levels during childhood development.
Figure 1-1: Conceptual Model of the Vicious Circle of Inactivity, Obesity, and
Metabolic Health Consequences

4

1.2 Background
Pediatric Obesity and At-risk Pediatric Populations
Pediatric obesity has reached epidemic proportions around the world.(168) Of
particular concern is that overweight and obesity are associated with a variety of adverse
physiological and psychological health outcomes, including MetS,(83) cardiovascular
disease,(69, 146) type 2 diabetes,(124) some cancers,(22) pulmonary problems (e.g.,
obstructive sleep apnea),(180) orthopedic complications,(197) impaired quality of life,(15)
and stigmatization and discrimination.(145) This is especially alarming because
overweight adolescents have as much as a 70% chance of becoming overweight or obese
adults,(Whitaker, Wright et al. 1997; Dietz 1998; Freedman, Khan et al. 2005) and
overweight youth are likely to develop serious and long-term adiposity-related health
consequences.(42, 149, 175) Evidence has also shown that adolescent obesity is an early
risk indicator for adult morbidity and mortality independent of adult obesity status.(14, 21)
Therefore, from a public health perspective, in order to prevent overweight and related
chronic diseases in adults, it is important to identify modifiable correlates of overweight
and obesity in youth as well as to target these correlates for change in youth.
The proposed studies will examine the influences of behavioral correlates (e.g.
physical activity and sedentary behavior) on overweight and the MetS in three pediatric
populations at high risk for obesity: Chinese youth (Study 1), US Latino youth (Study 2
and 3), and US African American youth (Study 2 and 3).

5

Pediatric Obesity in the United States
In the United Sates, the total cost of obesity-related care for children and adults in
the United States is estimated to be $151 billion in 2010.(213) Among children and
adolescents aged 6-17 years, estimated hospital expenditures associated with obesity
increased 3-fold between 1979 and 2001, reaching $127 million per year. (25) The
prevalence of obesity (at or above the 95
th
percentile of BMI for age) among children
ages 6 to 11 tripled from 6.5% in 1980 to 19.6% in 2008. There was also a comparable
increase from 5.0% to 18.1% among adolescents ages 12 and 19.(106, 149, 155) Based
on the most recent estimates from the National Health and Nutrition Examination Survey
(NHANES) from 2007 to 2008, 31.7% of children and adolescents aged 2 through 19
years were overweight [at or above the 85
th
percentile of Body Mass Index (BMI) for
age], and 16.9% were obese.(154) While the obesity epidemic affects all ethnic groups in
the U.S., the rates are particularly high in ethnic minority populations such as Latino and
African Americans. In 2007-2008, the rates of obesity among children and adolescents
ages 6-19 were 43.0% for Mexican Americans and 38.7% for African Americans, as
compared to 32.5% of non-Hispanic whites.(154) Not only is the prevalence
disproportionately higher in these two ethnic groups, so is the growth rate,(213)
a
placing

a
Based on NHANES data collected between 1970 and 2004, Wang et al
28
found that average
annual increase in prevalence of overweight is highest among Mexican American boys and
African American girls.

6

them at disproportionately high risk of obesity-related chronic diseases.(37, 76) For
example, MetS is most prevalent in Mexican American youth, (35) and Latino and
African American youth are more insulin resistant than Caucasian youth, independent of
adiposity (77). Therefore, to develop successful interventions to prevent obesity and
related chronic diseases for these two high-risk youth population, it is crucial to identify
the factors contributing to their health problems. The proposed Study 2 and 3 will
examine the relationships between physical activity, sedentary behavior, and MetS
among Latino and African American youth in the United States.
Pediatric Obesity in China
The epidemic in pediatric obesity has not only been observed in developed countries,
but also in developing countries, especially in developing countries undergoing rapid
economic transitions.(136) China, one of the most rapidly developing countries in the
world in the past two decades, has had a corresponding rapid increase in the rates of
overweight and obesity, particularly in youth.(20, 28, 214) In fact, from 1992 to 2002, the
prevalence of overweight and obesity in children (younger than 6 years old) and in youths
(aged 7-17 years) increased by 31.7% and 17.9%, respectively.(123, 228) These increases
have been found in both urban and rural areas, with a particularly sharp growth in urban
areas.(20) For example, in cities like Beijing and Shanghai, the overweight rate for
youths aged 7 to 18 years increased more than 5-fold between 1985 and 2000, from
1.5~1.6% to 8.8~14.5%.(20) Although the prevalence rate in China is relatively low
compared to developed countries like U.S, a combination of the alarming growth in

7

obesity rates and the lack of knowledge about this population puts Chinese youth at
increased risk. Hence, to fill the gaps in the existing literature, the proposed Study 1 will
investigate the associations between weight-related correlates and overweight status in
Chinese youth.
Obesity Risk in Chinese Adolescents and Public Health Implications
Despite the fact that Asian populations have a lower BMI than other ethnic
minorities,(45, 79, 224) they are at increased risk of obesity-related diseases(74) and have
a 60% higher rate of Type Ⅱ diabetes than Non-Hispanic Whites with the same
BMI.(133) In addition, Asian populations have higher visceral fat, a risk factor for
cardiovascular disease and Type Ⅱ diabetes, with adjustments for age and total body
fat.(160) Previous research has indicated that Chinese populations have higher amounts
of body fat at lower BMIs and waist circumferences than do Western populations.(222)
This, coupled with the upward trend in overweight and obesity has led to the increasing
development of chronic diseases.(238) Therefore, assessing factors contributing to
obesity in this population is urgently needed.
In addition to these adverse impacts on health, obesity in China has led to a growing
economic burden.(213) In 2003, estimates of the direct medical cost attributable to
overweight and obesity amounted to 21.11 billion Yuan (RMB) (equivalent to $2.74 U.S.
billion) accounting for 25.5% of the total direct medical costs of the four major chronic
diseases (hypertension, diabetes, coronary heart disease and stroke), or 3.7% of national

8

total medical costs.(237) Hence, it is imperative to design interventions for controlling
the obesity epidemic in order to prevent chronic diseases.
There is considerable evidence that the prevalence of overweight as well as its
determinants vary by culture and ethnicity.(3, 70, 74, 79) It is possible that the
adiposity-related correlates observed in Western youth may not be generalizable to those
in China. Aside from cultural contexts, it has been suggested that differences in
socioeconomic status and environmental factors may also contribute to the varying
obesity prevalence across populations.(212) For example, unlike Western society,(144) a
positive association between socioeconomic status (SES) and overweight prevalence has
been found in China, both in nationally representative samples (e.g. The China Health
and Nutrition Survey is a diverse sample that covered eight provinces with substantially
variations in geography and economic development(212)) and regional samples (e.g.
Wuhan(99), Xi’an(114)). This positive relationship between SES and overweight
prevalence remains significant even after adjustment of residential urbanization (eg. rural,
urban).(114, 212) Since much of the existing research has been conducted in Western
populations, more research is needed to understand roles that adiposity-related correlates
play in weight status in order to develop effective interventions to prevent obesity for
Chinese youth.
One challenge of studying overweight and obesity in China is that a variety of BMI
references [e.g., Center for Disease Control and Prevention (CDC),(108) World Health
Organization (WHO),(226) the International Obesity Task Force (IOTF)(34)] have been

9

used to define pediatric overweight and obesity. The lack of a commonly accepted
standard makes it difficult to compare the findings between studies. Another challenge is
that the two most commonly-used BMI references (CDC and WHO) are based on
Western youth, thus, are not appropriate for Chinese youth. This is important as Chinese
and other Asian populations tends to have higher amounts of body fat at lower BMIs than
do Western populations, the cut-offs for overweight are likely to be lower among
Chinese.(220, 239)
In 2000, IOTF published a series of sex- and age-specific BMI cut-offs that were
developed using international data sets from Brazil, Russia, Hong Kong, Singapore, the
Netherlands, the United Kingdom, and the United States.(34) The IOTF references are
international BMI cut-offs used to define “overweight” and “obesity” for children and
adolescent aged 2-18 years. The age and gender specific BMI curves pass through a BMI
of 25 kg/m
2
for “overweight” and 30 kg/m
2
for “obesity” at age 18. These BMI cut-offs,
25 kg/m
2
and 30 kg/m
2
, are linked to the widely used CDC adult cut-offs for overweight
and obesity, respectively.
The proposed Study 1 will define “overweight” based on the IOTF cut-offs, which is
more appropriate in Chinese youth for the following reasons: 1) As it is an international
reference based on the data from seven countries (two of them are ethnically Chinese –
Hong Kong and Singapore), it thus generally takes into account the lower BMI cut points
in Chinese populations., 2) The IOTF reference has been recommended for international
use because the cut-offs are based on international data.(34) In addition, it has been used

10

in several studies on Chinese population;(99, 114, 230) using such reference makes our
study findings comparable to other Chinese studies and international studies.
Relationships between Obesity and Metabolic Syndrome: Role of Insulin Resistance
In line with the rising prevalence of overweight and obesity, MetS, an important risk
factor for cardiovascular disease and type 2 diabetes, is also increasing in pediatric
populations worldwide.(193) The most accepted underlying pathophysiology mechanism
linking obesity to MetS and increased chronic disease risk is thought to be insulin
resistance.(183) Obesity is closely associated with insulin resistance. Studies have
indicated that adipose tissue released amounts of non-esterified fatty acids, glycerol,
hormones, pro-inflammatory cytokines and other factors that related to the development
of insulin resistance.(100) Indeed, obesity has been found to be the most common and the
most important risk factor for insulin resistance in children and adolescents.(23, 113)
Insulin resistance refers to an insensitivity of the peripheral tissues (e.g., muscle,
liver, adipose tissue) to the effects of insulin.(113) It is a condition in which the body
cells have become less responsive to the actions of insulin in transporting glucose from
the bloodstream into tissues for energy. As a result, the pancreatic beta cells may
over-compensate by releasing more insulin, leading to hyperinsulinemia as the body
attempts to maintain glucose homeostasis in the blood.(44) Previous research suggests
that when insulin resistance is accompanied by beta cell dysfunction, the dysregulation of
glucose homeostasis results.(100) Such impaired glucose tolerance and other
pathophysiology related to insulin resistance are closely associated with the increased

11

risk of MetS and the development of type 2 diabetes and cardiovascular complications.(9,
40, 184)
The Metabolic Syndrome: Definition and Prevalence in Youth
MetS, also known as syndrome X and insulin resistance syndrome, was first
described by Reaven in 1988.(171) It was not until a decade later, however, that the
World Health Organization proposed the first definition for adults. In 2001, the Adult
Treatment Panel (ATP) III of the National Cholesterol Program established their clinical
criteria for defining MetS; since then this has become widely used (see Appendix A).(82)
As defined by the ATP III, MetS is the presence of three or more of the following five
risk factors: elevated triglycerides, low HDL-cholesterol, abdominal adiposity,
hyperglycemia, and elevated blood pressure.(199) Previous research has conclusively
shown that MetS is predictive of subsequent cardiovascular disease and Type 2 diabetes
in adults.(93, 110, 125) In adolescents, however, the clinical relevance of MetS is more
controversial. Some studies have found that MetS in childhood is related to the
development of Type 2 diabetes in adulthood .(143) On the other hand, there is evidence
suggesting that instability exists in the categorical diagnosis of MetS during adolescence,
so its clinical utility may be reduced for youth.(75) Additional research on MetS in youth
is therefore warranted.
In contrast to the definition of MetS in adults, several criteria for MetS in children
and adolescents have been proposed(36, 41, 218) (See Appendix B), but there is no
standardized pediatric definition.(68) This lack of consensus may be due to the

12

puberty-related physiological changes that affect metabolic profiles.(75) The varying
definitions have led to a wide range of estimates of prevalence and made it difficult to
compare across studies. In NHANES from 1999 to 2002, the prevalence rates of MetS
varied from 2.0-9.4 % in adolescents aged 12-19 years old and from 12.4-44.2% among
those who were obese, depending on the definitions used.(35) Recently, Shaibi and
Goran(184) compared the rates of the MetS in youth using three common pediatric
criteria based upon the ATP III definition and showed moderate agreement between them
(kappa=0.5-0.7). They concluded that a uniform pediatric definition of MetS is needed to
assist clinicians to identify youth who are at risk for prevention interventions.
In the proposed Study 2 and 3, a combination of pediatric definitions proposed by
Cruz et al(41) and Cook et al,(36, 184)
a
who applied pediatric cutoffs to the ATP III
definition(64) will be used. This definition, which has been applied in our previous
studies, (41, 184) classifies adolescents as having MetS if they have at least three of the
following features: abdominal obesity [waist circumference ≥ 90th percentile (percentiles
from NHANES Ш 1988-1994 data) for age, sex, and Hispanic and African American
ethnicity].(65) hypertriglyceridemia [triglycerides ≥ 90th percentile (percentiles from
NHANES Ш 1988-1994 data) for age, sex, and Hispanic and African American

a
Cook et al
101
originally used fasting glucose ≥110 mg/dl to define Impaired Fasting Glucose. The criteria for
Impaired Fasting Glucose has been updated by the American Diabetes Association
108
as a fasting glucose
≥100 mg/d. The revised cut-point has been used in Cook et al’s article in 2008
9
and will be used in the
proposed Study 2 and 3.

13

ethnicity],(87) low HDL-cholesterol [HDL-cholesterol ≤ 10th percentile (percentiles
from NHANES Ш 1988-1994 data) for age, sex, and Hispanic and African American
ethnicity],(87) elevated blood pressure [systolic or diastolic blood pressure ≥ 90th
percentile (percentiles from NHANES 1999-2000 data) adjusted for age, sex, and height],
and hyperglycemia (impaired fasting glucose ≥ 100 mg/dl).(5, 35, 36)
a

Metabolic Risk among Latino and African American Youth in the U.S.
Among U.S. adolescents, the prevalence of MetS is highest among Mexican
Americans (2.6-11.1%), followed by Whites (2.2-10.7%), and African Americans
(1.6-5.2%).(35) Specifically for overweight Latino youth, the prevalence rates of MetS
could reach as high as 30% according to previous studies in the Los Angeles area.(39) It
is important to note that although African Americans have relatively low rates of MetS
compared with other ethnic groups, they are at greater risk for specific components of
MetS, such as hypertension(51, 198) and abdominal adiposity.(51) In addition, African
American and Latino youth have highest rates of obesity(154) and that their levels of
insulin resistance are higher than Caucasian youth, independent of adiposity.(77) Hence,
more studies are warranted to examine metabolic risk factors in Latino and African
Americans youths, not only because these two populations are understudied, but also
because they are disproportionately affected by obesity and insulin resistance, two crucial
risk factors linked with MetS.

14

Physical Activity versus Sedentary Behavior
Increasing physical activity and decreasing sedentary behavior are key elements in
the prevention of weight gain and the treatment of obesity.(47, 216) Physical activity is
defined as any bodily movement produced by a contraction of skeletal muscle that
substantially increases energy expenditure.(90) Sedentary behavior is conceptualized as a
range of activities that involve low energy expenditure, rather than merely the absence of
physical activity.(13) Although engaging in physical activity and being sedentary may
seem like two sides of the same coin, research has suggested that these two behaviors
should be treated as different dimensions, rather than as a continuum (e.g.
time-displacement hypothesis, which proposes that sedentary behaviors displaces time
spent in physical activity).(148, 208)
This perspective is supported by the findings from two pediatric reviews,(127, 178)
in which the relationship between physical activity and sedentary behavior was shown to
be near zero, suggesting two independent constructs. Sallis et al(178) summarized
correlates of physical activity and found that physical activity and sedentary behaviors
(TV viewing and video games) were not strongly related in children and unrelated in
adolescents. Another recent review conducted by Marshall et al.(127) showed a small
negative correlation between physical activity and sedentary behavior (r=-0.096) in youth.
Although this small correlation may provide some evidence for a displacement
hypothesis, the authors suggested that the effect is too small to be of practical
significance and that sedentary behavior may not necessarily be the main deterrent to or

15

replacement for physical activity.(13) For instance, when one chooses to watch TV,
TV-viewing time is equally as likely to displace time spent in any other behavior (e.g.
sleep), as it is to displace physical activity. Additional evidence against the displacement
hypothesis shows how physical activity and sedentary behavior might coexist.(152, 157)
One study(128) using cluster analysis to describe activity patterns in a large youth cohort
from the UK and the US found that, in both girls and boys, certain clusters with high
sedentary behavior also showed high levels of physical activity. Therefore, the inverse
relationship between physical activity and sedentary behavior cannot always be assumed.
In summary, sedentary behavior should be studied as a distinct construct. In fact,
Owen et al. contend that sedentary behavior can sometimes compete with and sometimes
co-occur with physical activity.(157) This further lends support for studying physical
activity and sedentary behavior separately and simultaneously in relation to obesity
(Study 1) and MetS (Study 2), which, will be further discussed in page 26.
Objective and Subjective Measures of Activity Levels
Accurate assessments of physical activity and sedentary behavior are central to
evaluating their influences on health outcomes. However, obtaining valid measures of a
behavior that is as complex as physical activity is challenging. Multiple measurement
modalities have been utilized to capture activity levels and could be categorized into two
primary domains: subjective (e.g., questionnaire, activity log) and objective (e.g., direct
observation, accelerometers, pedometers, heart rate monitors).

16

Compared to objective measures, subjective measures such as self-reported
assessment of activity levels are less expensive, easier to implement, and more commonly
used, especially for large-scale studies.(190) It has been suggested that physical activity
occurs over four dimensions: frequency, intensity, time, and type.(147) Depending on the
survey tool that is used, information of all four dimensions can be collected. As
self-reported activity levels are based on individual perception, it is prone to the bias
created by social desirability, recall error, and misinterpretation.(96) Overcoming the
biases inherent to subjective measures, accelerometer-based activity monitors provide
objective estimates of the frequency, duration, and intensity. Among the objective
modalities, electric monitors such as accelerometers have become popular, possibly due
to the recent reductions in price. However, they are usually not able to obtain contextual
information on the types of activities. Other shortcomings of accelerometry are possible
lacunas in accuracy (e.g inability to accurately detect activity intensity during movements
with static hip position such as biking) and also that its usability (e.g. inability to wear
them while playing water sports) is dependent on activity type.(187)
A growing number of studies have been conducted to compare the agreement
between subjective assessments and objective measurements. Concordances (as measured
by correlations) between questionnaires and accelerometry have been demonstrated to be
poor to fair, ranging from -0.26 to 0.04.(84, 132, 186, 202, 225) Possible explanations
have been proposed for this considerable disagreement between methods. First,

17

inconsistencies may reflect the differences in how activity levels are assessed between
measures. Responses from questionnaires are subjective and thus are influenced by
respondent’s fitness levels; while estimates of physical activity using accelerometry are
absolute values.(84) In addition, some state-of-art subjective questionnaires, such as
3DPAR, could provide estimations for each intensity level using the compendium of
physical activities.(1) However, applying the indices from compendium may be
imprecise, depending on the degree of between-subject variability in a given population
to the sample of the compendium study. Second, estimates by questionnaires are usually
quantified by respondents based on their recall of the dominant intensity level for the
dominant activity during a certain period of time (e.g., if they were running for 15 min
from 10:00 to 10:15 am, they may record that they had been running for an entire
half-hour block from 10.00 to 10:30 am). This, in many cases, may lead to an
overestimation of time spent in physical activity, since participants may also include time
spent warming up, resting, or cooling down, especially for time-based questionnaires.
Accelerometers, on the other hand, only record the actual time spent in specific intensity
level while participants are physically active.(84) Finally, some of the existing studies
that compared the agreements between methods are those that evaluated the validity of
questionnaires using accelerometers for comparison. Such evaluations, however, are
problematic since there is no one measure that can be considered the ‘gold standard’.(31)
For example, accelerometry tends to underestimate certain activates such as biking and
swimming. Qualitative data from our Latino and African American youth suggest that

18

swimming, in particular, is a favorite activity. Indeed, it has been suggested that high
correlations are less likely to be observed by comparing a questionnaire with an
accelerometer since two modalities assess different components of the activity
behaviors.(31)
Thus, each measure has its shortcomings, and a combination of both subjective and
objective approaches might be preferred since this would enable complementary insights
that would enrich understandings of physical activity behaviors.(190) Thus, the proposed
Study 2 and 3 aim to expand existing knowledge on relationships between activity levels
and metabolic health by assessing activity levels both objectively (using accelerometry)
and subjectively (using 3DPAR). It is important to note that although some studies have
compared estimates of activity levels between measures, to our knowledge, none have
investigated the associations between activity levels and metabolic health outcomes using
both subjective and objective techniques. Such approach has two unique advantages. First,
it provides a more comprehensive understanding of the relationships between physical
activity, sedentary behavior, and MetS. As previously mentioned, the agreement between
objective and subjective measures is poor; it is likely that they capture different
dimensions of activity. Since each type of activity measure has its own limitations in
accuracy, relying on one modality may not provide the complete picture of how activity
levels influence metabolic health.(190) Second, it allows explorations of whether or not
the relationships between activity levels and metabolic health differ by measure. If there

19

are differences in these relationships between activity modality, in order to provide
conclusive or consistent findings on activity levels and metabolic outcomes, future
research should compare its findings with previous work using the same type of activity
measure.
Activity Levels and Overweight (Study 1)
Overweight is a condition in which excess body fat results from a positive energy
balance (i.e., when energy intake exceeds energy expenditure).(78) Two of the main
strategies that have been recommended for reducing overweight are increasing physical
activity and decreasing sedentary behavior.(47) According to the 2008 Physical Activity
Guidelines for Americans,(209) children and adolescents should participate in 60 minutes
or more of moderate -to-vigorous physical activity daily. Although there are not yet
official guidelines or limits for sedentary behavior, the American Academy of Pediatrics
recommends that children and adolescents should not spend more than 2 hours of media
time per day.(4)
Physical activity has been linked to positive health outcomes. (178) Several
reviews(88, 104, 147, 242) have summarized the role of physical activity in overweight
and obese youth and their findings have been mixed. Nevertheless, the majority of the
studies have found an inverse association between physical activity and overweight or
obesity, especially for vigorous intensity.(242) Inconsistencies regarding this association
across studies might be due to variations in methods used to measure physical activity as

20

well as adiposity.(88) The most commonly utilized measure is self-reported
questionnaires, either completed by the youth themselves or as proxy measures
completed by their parents.
Sedentary behavior is an increasingly important public health threat in adolescents,
considering the fact that youth spend the majority of the day being sedentary(201) and the
increasing availability of attractive low-intensity media-related activities prompted by
new, affordable and accessible technologies. However, while the literature on the role of
physical activity in the development of obesity has been widely studied, sedentary
behavior has received relatively little attention.(147) Recently, more studies are
beginning to focus on sedentariness as a public health issue. This growing trend may be
promoted by the increasing support of viewing sedentary behavior as a separate construct
from physical activity. Among the available literature regarding sedentary behavior and
obesity, findings on the relationship between sedentary behavior and adiposity have also
been inconsistent. Patrick et al.(166) showed that while only six of twenty-one studies in
adolescent populations reported a significant relationship between weight status and
physical activity, sedentary behavior, especially time spent on television viewing, was
consistently related to overweight. In one meta-analysis review among children and
adolescents,(127) there was a significant negative relationship between sedentary
behavior and adiposity. Nevertheless, the authors indicated that the magnitude of this
relationship was too small to be of substantial clinical relevance. In these two reviews,
most studies only assessed TV viewing as sedentary behavior. Relying on a single

21

indicator of sedentary behavior may not be sufficient to understand the influence of
sedentary behavior on obesity.(13) They suggested that expanding the research definition
of ‘sedentary behavior’ to include other types of activities that clearly fall into this
category may increase the clinical relevance of sedentary behavior on adiposity.(13)
One major gap in the literatures on activity levels and overweight/obesity is that
most studies do not investigate both physical activity and sedentary behavior
simultaneously. Such consideration is important because not only these two behaviors are
distinct constructs that could coexist (see p.19), but also because the protective effects of
physical activity on heath could be attenuated by prolonged sedentary behavior.(57) For
example, one study showed that among highly-active individuals, those watching 4 hours
of TV per day were twice as likely to be overweight as those watching less than 1 hour of
TV per day.(179) Another gap is that growth and development is often not considered
when assessing effects of physical activity and sedentary behavior on weight status
during childhood and adolescence.(88) Puberty is a critical period of development with
dynamic changes in body composition(102) and activity levels.(79, 105, 162) It is
possible that maturation could affect the relationships between physical activity levels
and MetS. Lack of attention to the influence of pubertal stage may further confound
interpretations in youth studies. Therefore, the proposed Study 1 aims to fill these gaps by
including both physical activity and sedentary behavior in the same model, which allows
the investigation of their independent effects on overweight. In addition, pubertal

22

development is measured and is adjusted for the relationships between activity levels and
overweight.
Activity Levels and the Metabolic Syndrome (Study 2)
It has been widely documented that higher levels of physical activity and lower
levels of sedentary behavior is associated with a reduced risk of MetS in adults.(53, 67)
The precise mechanism by which physical activity acts on MetS have yet to be
established. One potential explanation proposed is that physical activity may mediate its
favorable effects on insulin sensitivity and metabolic features by decreasing subclinical
inflammation involving cytokines expressed by adipose tissues, such as adiponectin,
leptin, and soluble TNF-α 1.(46, 101, 167, 173) It is also likely that physical activity
reduces adiposity and thereby decreases the chance of developing insulin resistance and
associated metabolic risks.(134)
In the existing pediatric literature on activity levels on MetS, four major weaknesses
have been identified. First, while higher levels of physical activity and lower levels of
sedentary behavior have been shown to reduce the risk of MetS in adults, (53, 67) these
relationships have been rarely investigated in adolescents. For example, in their review
on physical activity and MetS, the Physical Activity Guidelines Advisory Committee
found only eight studies on pediatric populations. (208) Availability of data on minority
youth is even more limited. Another review(193) of the relevant pediatric literature only

23

identified one study(18) which assessed the associations between activity levels and
MetS specifically in minority youth
a
. That study,(18) which was conducted on a sample
of 4- to19-year-old Hispanic children and adolescents, demonstrated that
moderate-to-vigorous physical activity as measured by accelerometry was inversely
related to the number of risk factors of MetS. To inform development of interventions for
minority ethnic groups who are at greater metabolic risk, more studies on minority youth
are needed to assist in establishing the link between physical activity levels and MetS.
Another gap in the existing literature is that sedentary behavior has received less
attention than physical activity. As sedentary behavior is known to be associated with
progression toward MetS in adults,(53, 67) future research needs to understand this
relationship in youth. Another problem common to the literature on activity levels and
obesity(147) is that studies examining relationships between activity levels and MetS are
often confined to self-reported assessments of physical activity and sedentary behavior.
There is a lack of studies using objective measures of physical activity. Indeed, only three
pediatric studies identified used objectively measured physical activity (all measured by
acclerometry).(8, 57, 172) However, these studies were conducted on European youths
and little is known regarding activity levels and MetS in US minority youth.

a
Another study was conducted in a predominantly minority adolescent sample. Walker at el.
227
found no
association between physical activity and MetS in a racially/ethnically diverse adolescent sample from
NHANES [ Non-Hispanic Black (34%), Mexican American (38%), and Non-Hispanic White (28%)].

24

Finally, most of the studies that evaluate the influences of physical activity and
sedentary behavior on MetS do not control for body composition, a potential confounder
of the relationship between activity levels and metabolic health.(167) The few adult
studies(111, 161, 240) that did take body composition into consideration found that body
composition generally attenuated the protective effects of increasing physical activity and
reducing sedentary behavior on MetS.(32) In children and adolescents, however, little is
known regarding these relationships. More studies are thus warranted to investigate
whether the associations between activity levels and MetS partially are independent of
body composition. If these relationships were established, such knowledge would have
important public health implications., i.e. while minimal increases in physical activity
and reductions in sedentary behavior may not necessarily lead to lower adiposity, these
changes may still have favorable effects on metabolic risk profiles.(193) Therefore, the
proposed Study 2 aims to fill the aforementioned gaps by exploring the independent
influences of physical activity and sedentary behavior on MetS in Latino and African
American youth, using both subjective and objective measure of activity levels. These
associations will be adjusted for body compositions.
Longitudinal Relationships between Activity Levels and the Metabolic Syndrome (Study 3)
Much of the prior literature regarding relationships between physical activity,
sedentary behavior, and MetS have relied on cross-sectional data(53, 67), disallowing
causal inference or clear conclusions on directionality of relationships, and leaving open
the possibility that relationships could be bidirectional or reciprocal. Higher levels of

25

physical activity and lower levels of sedentary behavior may lead to poor metabolic
health outcomes, such as MetS. On the other hand, poor metabolic health may contribute
to declines in physical activity and increases in sedentariness. A recent study in a
nationally representative sample in youth aged 6 to 19 years showed that overweight
youth are significantly less active than normal weight youth, regardless of age.(10) This
suggests that, while inactivity certainly leads to poor metabolic outcomes such as
overweight or MetS, the presence of overweight or MetS might in turn lead to lower
activity levels and more sedentariness, resulting in a vicious circle of increasing inactivity
and worsening metabolic health - physical activity and sedentary behavior can thus be
conceptualized of as both antecedents and consequences of overweight or MetS. In order
to unveil the causal relationships between physical activity, sedentary behavior, and MetS,
more longitudinal investigations that allow for the study of temporal relationships are
needed.
In adults, there is longitudinal evidence indicating that low levels of physical
activity predict the development of MetS.(19, 56, 58, 89, 112) Leisure time physical
activity recorded in middle age man was a significant protector against MetS 28 years
later in life.(89) In another study of 393 adult men and women, increasing levels of
physical activity over a period of 5.6 years was found to reduce metabolic risk
independent of aerobic fitness and body fatness.(58) In children and adolescents,
however, there is a lack of relevant longitudinal research. Two pediatric studies were
found. Raitakari et al(170) showed that persistent physical activity lasting 6 years was

26

related to low metabolic risk while persistent inactivity to high risk. Yang et al(232)
found that sustained sport participation in youth protects against MetS in adulthood as
measured 21 years later.
While most studies have focused on investigating the preventive effects of physical
activity on metabolic risk, the possibility that overweight status and other metabolic
health issues such as MetS might in turn negatively impact activity levels, sending youth
into a spiral of health risk and health risk behavior, has received less attention. This idea
of a vicious cycle offers a more comprehensive explanation for the associations between
metabolic health and activity levels during childhood development. Children who are
overweight might have lower levels of physical activity as a consequence of their weight
status. (59) One pediatric study(135) examined whether inactivity was the cause of
fatness or fatness was the cause of inactivity over the 3-year period, and concluded that
physical inactivity was the result of fatness rather than its cause. Previous research has
demonstrated that overweight children find sedentary activities more reinforcing than
active pursuits,(62) have poor self-efficacy of overcoming barriers of being active(205),
and have high rates of peer victimization (e.g. bullying behaviors and peer isolation).(59,
95) Especially during childhood, when physical activity is often organized group
activities, these negative social and emotional consequences may discourage overweight
children from being active and increase their time spent in sedentary behaviors.(59, 147)
Following this logic, youth with MetS (whom are usually overweight) may have lower
self-efficacy for physical activity, greater social/emotional barriers, or, their poorer

27

physical fitness may limit their engagement in physical activity. Furthermore, as declines
in physical activity during puberty have been widely documented,(79, 105, 162) it is
possible that the presence of MetS may accelerate the reduction in physical activity and
the increase in sedentary behavior as youth traverse puberty.
To date, no studies have compared whether youth with MetS have sharper decreases
in physical activity and greater increases in sedentary behavior than those without MetS
over time. Investigations of these associations has particular public health significance,
because if the decline in physical activity is greater in youth with MetS, such a finding
would not only reinforce the focus on prevention, but also underscore the importance of
early intervention to mitigate low physical activity levels for those with MetS. In addition,
if high sedentary and low physical activity levels are found in those with MetS, it would
support the usefulness of regular MetS screenings among youth to reduce their future
disease risk. Hence, the goal of the proposed study (Study 3) is to examine the
longitudinal changes in levels of physical activity and sedentary behavior between youth
with and without MetS. It is important to note that the only possibility of truly
determining cause and effect here would be to use data from a nationally representative
prospective birth cohort. However, the strength of Study 3 is that we can look at temporal
relationships, and thus provide a meaningful addition to the current literature by testing
the hypothesis that the presence of MetS contributes to longitudinal (over the course of
12 months) decline in physical activity and increase sedentary behavior.

28

1.3 Specific Aims and Hypotheses
Paper 1: Cross-sectional analysis of physical activity, sedentary behavior, and odds of
being overweight in Chinese adolescents
The overall objective of the first study is to investigate the cross-sectional
associations between physical activity and sedentary behavior on overweight status in
Chinese adolescents. The paper closes with a discussion of whether these correlates of
adiposity that have been identified in Western youth may have different (or no)
relationships to adiposity in Chinese youth.
• Aim 1: to compare physical activity, sedentary behavior, and other weight -related
correlates in a large sample of overweight and non-overweight Chinese
adolescents.

o Exploration: the differences in levels of physical activity, sedentary
behavior, and other weight -related correlates in overweight and
non-overweight Chinese adolescents will be explored.

• Aim 2: to examine the independent influences of physical activity, sedentary
behavior, and other weight-related correlates on overweight status in Chinese
adolescents.

o Exploration: the independent influences of physical activity, sedentary
behavior, and other weight -related correlates on overweight status in
Chinese adolescents will be explored.

• Aim 3: to investigate whether the roles that weight-related correlates play in
pediatric overweight in Chinese youth are similar to the roles they play in
Western youth.

o Exploration: Findings from this study will be compared to findings on
weight-related correlates in Western youth, and the possibility that
correlates play different roles in the Chinese culture than in Western
cultures will be explored. As this aim is exploratory, there are no specific
hypotheses.

29

Paper 2: Cross-sectional analysis of physical activity, sedentary behavior, and the
metabolic syndrome in Latino and African American youth
The overall objective of the second study is to examine the associations between
physical activity, sedentary behavior, and the metabolic syndrome in Latino and African
American youth. Activity levels were measured by accelerometry as well as the 3-Day
Physical Activity Recall (3DPAR), and the relationships of activity levels to the
metabolic syndrome were compared across measures.
• Aim 1: to compare physical activity and sedentary behavior of Latino and African
American youth with and without the metabolic syndrome.

o Hypothesis: Participants with the metabolic syndrome will have lower
levels of physical activity and higher levels of sedentary behavior than
those without the metabolic syndrome.

• Aim 2: to examine the influences of physical activity and sedentary behavior on
the odds of developing the metabolic syndrome in Latino and African American
youth.
o Hypotheses:
Physical activity will be negatively associated with the odds of
developing the metabolic syndrome.
Sedentary behavior will be positively associated with the odds of
developing the metabolic syndrome.

• Aim 3: to examine the associations between physical activity, sedentary behavior,
and individual features of the metabolic syndrome.
o Hypotheses:
Physical activity will be negatively correlated with triglycerides
levels, waist circumference, fasting glucose levels, and blood
pressure while positively correlated with high-density
lipoprotein-cholesterol levels.

30

Sedentary behavior will be positively correlated with triglycerides
levels, waist circumference, fasting glucose levels, and blood
pressure while negatively correlated with high-density
lipoprotein-cholesterol levels.

• Aim 4: To explore differences in associations with metabolic outcomes between
physical activity /sedentary behavior measured by self-report as opposed to
physical activity /sedentary behavior measured objectively.

o Exploration: Findings based on accelerometry will be compared to those
based on 3DPAR. As this aim is exploratory, there are no specific
hypotheses.

Paper 3: Longitudinal Differences in Physical Activity and Sedentary Behavior by
Metabolic Syndrome among Young Latina and African American Females
The overall objective of the third study is to examine if the metabolic syndrome in
young Latina and African American females is associated with longitudinal changes of
physical activity and sedentary behavior over the course of 12 months. Specifically, the
patterns of physical activity and sedentary behavior across 12 months between those with
and without the metabolic syndrome are compared.
• Aim 1: to determine if participants with and without the metabolic syndrome have
different initial levels of physical activity and sedentary behavior.
o Hypothesis:
Participants with the metabolic syndrome will have lower initial
levels of physical activity as compared to those without.
Participants with the metabolic syndrome will have higher initial
levels of sedentary behavior as compared to those without.

31

• Aim 2: to determine if participants with and without the metabolic syndrome have
different longitudinal changes in physical activity and sedentary behavior over 12
months.
o Hypothesis:
Participants with the metabolic syndrome will show a greater
decrease in physical activity over time as compared to those
without.
Participants with the metabolic syndrome will show a greater
increase in sedentary behavior over time as compared to those
without.
1.4 Overall Summary
In summary, the proposed Study 1 aims to expand current knowledge on the
relationship between physical activity and overweight in Chinese youth by taking the
possible influence of sedentary behavior on that relationship into account. Study 1 is also
one of the first studies to be able examine the relationship between sedentary behavior
and overweight after adjusting for the influence of physical activity. Additionally,
whether or not the correlates of obesity that have been identified in Western youth have
different (or no) relationships to obesity in Chinese youth will also be explored. The
purpose of Study 2 is to investigate the influences of physical activity and sedentary
behavior on MetS in minority youth. Study 2 aims to fill the gaps in existing research on
activity levels and MetS by examining both physical activity and sedentary behavior
simultaneously, taking into account body composition, using both subjective and
objective activity measures, and adding to the limited data in Latino and African
American youth. Finally, the proposed Study 3 will compare the longitudinal trends of
physical activity and sedentary behavior between youth with and without MetS, adjusting

32

for body composition. Study 3 is strengthened by the use of both objectively and
subjectively measured activity levels. Moreover, it will be the first study that examines
group differences in activity levels over time.
1.5 Study Samples
The dissertation draws upon data from four separate research studies conducted
through the University of Southern California (USC) Department of Preventive Medicine.
Paper 1 uses baseline data from the China Seven City Study (CSCS, PI: Dr. Anderson
Johnson). Paper 2 uses baseline data from three related pediatric obesity studies that share
a set of common methods and measures: the Insulin Resistance and Declining Physical
Activity Levels in African American and Latina girls (TRANSITIONS, PI: Dr. Donna
Spruijt-Metz) study, the Strength and Nutrition Outcomes for Latino Adolescents (SANO,
PI: Dr. Michael Goran) study, and the Strength Training and Nutrition Development for
African American Youth (STAND, PI: Dr. Michael Goran) study. Paper 3 uses data over
the first year from the TRANSITIONS study. Detailed study methodologies can be found
in the individual papers included in subsequent sections.
The CSCS study is a longitudinal smoking prevention and health promotion study
initiated by a consortium of researchers in the U.S. and China in 2001. The CSCS
collected data of four age groups: middle school students, high school students, college
students, and parents of the middle and high school students. The overall goal of the

33

study is to monitor tobacco use and related health practices and outcomes among youth
and adults in China.
The TRANSITIONS study is an ongoing 4-year longitudinal study of Latina and
African American girls, ages 8-11 years at study entry in 2007-2009. Participants are
evaluated annually for clinical measures and quarterly each year for psychosocial and
behavioral measures. The overall goal of the study is to determine physiological and
psychological determinants of the decline in physical activity in Latina and African
American girls during puberty.
The SANO study is a 16-week randomized trial developed to examine the effects of
a modified carbohydrate nutrition program combined with strength training on insulin
sensitivity, adiposity, and associated metabolic risks in overweight Latino adolescents
aged 14-18 years. There were three intervention groups including: 1) control, 2) modified
carbohydrate nutrition, and 3) modified carbohydrate nutrition with strength training.
Data collection began in February 2006 and ended in June 2007. The STAND study is the
sister study of the SANO study conducted on African American adolescents aged 13-18
years.

34

Chapter 2 CORRELATES OF OVERWEIGHT STATUS IN CHINESE
YOUTH: AN EAST-WEST PARADOX
2.1 INTRODUCTION
Global trends in pediatric overweight and obesity have reached epidemic proportions
in both developed countries and developing countries undergoing fast economic
transitions.(168) China is one of the most rapidly developing countries in the world. In
China, the prevalence of pediatric overweight has escalated substantially in the past
decades, and this trend has been found in both urban and rural areas.(20, 215) From 1992
to 2002, the prevalence of overweight in the Chinese population increased by 31.7% for
children younger than 6 years old and 17.9% for youths aged 7-17 years.(123, 228) The
overweight rate for youths age 7 to 18 years increased more than 5-fold between 1985
and 2000, from 1.5~1.6% to 8.8~14.5% in large urban cities like Beijing and Shanghai,
where the prevalence has reached the same epidemic levels found in some developed
countries.(26) Although the growth in the prevalence of overweight in the rural cities is
slower, research suggested that the epidemic of childhood overweight might be in its
early stages.(20)
The prevalence of overweight as well as the determinants of overweight differ
significantly by culture and ethnicity.(3, 70, 74) Thus, influences of certain correlates on
overweight observed for Western populations may not be generalizable to those in China.
In contrast to Western society,(144) a positive association between socioeconomic status

35

(SES) and overweight prevalence has been found in China.(99) Previous research
suggests that overweight epidemic in China may partially be attributed to a widespread of
social belief that overweight represents health and prosperity in Chinese culture.(227)
A healthy diet and physical activity have long been emphasized as two important
strategies for the prevention of obesity.(103) Studies conducted in Western societies also
show consistently that diet, physical activity, and sedentary behaviors are lifestyle factors
that significantly influence weight status.(166) Associations between these lifestyle
factors and weight status was examined in a sample of U.S. Latino and African American
youth (from the proposed Study 2) and a higher level of moderate-to-vigorous physical
activity was found to be related to body fat mass (β=-0.12, P=0.02). In China, the
adoption of an open-market policy and rapid economical growth have led to major shifts
in individual dietary patterns.(176) Not only scarcity of food has been reduced on a
national level, but also average food consumption has increased.(168) In the wake of
these changes, adolescents have begun to consume more energy dense and nutrient poor
foods.(50) Coinciding with China’s increasing industrialization are shifts in physical
activity behaviors.(29) Increased household automobile ownership and the affordability
of TV and computers have been linked to reductions in physical activity(30) and
increases in sedentary behavior(122) among Chinese youth.
Sleep is another lifestyle factor that has been linked with overweight.(27) Growing
evidence of an association between a insufficient sleep and overweight is coming out of

36

research conducted in Western cultures.(121) In China, academic pressure and stress
induced by preparation for the National College Entrance Exam has been indicated as a
major reason for sleep deprivation in adolescents.(119, 235) In one cross-country
comparison of sleep patterns,(118) Chinese youth went to bed later, woke up earlier, and
on average had one hour shorter sleep duration in comparison to American youth. Greater
sleep debt may contribute to the sharp increase in adolescent overweight in China.
Much of the existing research on weight-related behavioral correlates has been
conducted in Western populations. Because the Chinese culture and economic climate
represent a distinctly different context from Western society, it is possible that the
behavioral correlates of adiposity that have been identified in Western youth may be
differentially related to adiposity in Chinese youth. Furthermore, while some studies have
assessed the influence of weight-related correlates separately in Chinese youth, to our
knowledge, none to date have examined the influences of diet, physical activity,
sedentary behavior and sleep simultaneously. Therefore, the objective of the study was to
explore the independent influences of sleep duration, physical activity, sedentary
behavior, and dietary behavior on overweight status in Chinese adolescents.
2.2 METHODS
Sample
The cross-sectional data presented were collected from the baseline survey of a
longitudinal smoking prevention and health promotion study conducted in the 7 large

37

cities in China (China Seven City Study, CSCS).(98) To obtain a geographically
representative and socioeconomically diverse sample, CSCS included cities in 4 regions
of China: Northeast (Harbin, Shenyang), Central (Wuhan), Southwest (Chengdu,
Kunming), and Coastal (Hangzhou, Qingdao). The baseline data presented were collected
from students in middle and high schools as well as from their parents or guardians in
2002. A stratified sampling strategy was adopted to randomly select middle school and
high school samples in each city.(231) In each city, high, middle, and the low-income
residential districts were identified. Within each identified district, middle and high
schools were grouped according to 3 levels of academic performance. One middle school
and one high school were randomly selected from each of the 9 (3 levels in each of 3
districts) clusters to participate in the study. From these schools, one class each from
grades 7, 8 (middle school), 10, and 11 (high school) were randomly selected for the
study. All students from each selected classroom were invited to participate in the study.
Parental consent forms were distributed to students in their schools. Students with
parental consent completed the questionnaire during class time. Trained data collectors
administered questionnaires and answered any questions that students had about the
surveys. Teachers were not in the room during questionnaire administration. Parents
completed the questionnaire at home. Height/weight data were collected by trained
personnel from the Health Bureaus from each city. Study procedures were approved by
the Institutional Review Boards of the University of Southern California and each of the
7 participating cities’ Health Bureaus.

38

Across the 7 cities, a total of 14,434 students and their parents provided consent and
filled out questionnaires. Participants who have completed responses to all measures of
interests were included in the current study. The final sample therefore consisted of 9023
(62.51%) students and their parents who responded to the questionnaires.
Measures
Sedentary behavior
A total of 4 items adapted from the Youth Risk Behavior Surveillance Survey
(YRBSS) were used to measure time spent on 2 sedentary activities: watching TV/video
and using a computer outside of school.(16) For each sedentary activity, participants were
asked to report how much time they spent on school days and on a usual weekend or
holiday, respectively. Reponses to the questions ranging from “none/hardly any”, “< 30
minutes”, “30 - 59 minutes”, “1 – 1.99 hours”, “2 – 2.99 hours”, “3 – 3.99 hours”, and “≥
4 hours”. The responses were converted into minutes. Minutes spent on an average day
for each sedentary activity were then calculated via the following equation: minutes spent
on an average day = [(minutes spent on school days)*5 + (minutes spent on an average
weekend)*2] / 7. Finally, total minutes spent on sedentary behaviors were obtained by
summing the minutes spent on an average day for watching TV/video and for using a
computer outside of school.

39

Vigorous physical activity (VPA)
The frequency of VPA was assessed by one item adapted from YRBSS,(16) which
asked participants “how many times a week do you breathe hard and sweat for over 20
minutes while riding a bicycle, walking fast, jogging, dancing or doing other exercise or
hard physical labor?” The response options ranged from “none” to “8 or more times” and
were further dichotomized into “less than 3 times a week” and ”greater or equal to 3
times a week”.
Sleep duration
Sleep duration on an average school night and on an average weekend or holiday
night were assessed through 2 questions, respectively. Hours of sleep obtained on an
average night were then calculated via the following equation: sleep time on an average
night = [(sleep time on school days)*5 + (sleep time on an average weekend night)*2] / 7.
Self-perception of health status Participants were asked to describe their health status
based on one item in terms of the following response options: excellent, good, fair, and
poor.
Dietary behavior
Dietary intake during the last month was measured by food frequency items adapted
from the YRBSS.(16) Five food items were included in the current analyses to assess the
consumption of different foods: vegetable intake (eg, fresh or cooked vegetables), fruit
intake (eg, fresh fruits including fruit juice), sweets intake (eg, desserts, ice cream, candy,
or soda), snack intake (eg, potato chips, corn chips and tortilla chips), and fast food intake

40

(eat something from a fast food restaurant, eg, McDonald's, KFC, Pizza Hut etc.).
Answer categories for frequency of consumption during the last month were: “less than
once a week”, “once a week”, “2-3 times per week”, “4-6 times a week”, “once a day”,
and “2 or more times per day”. Following procedures that have been used
previously,(117) these 6 categories were converted into the following values to indicate
the frequency of dietary intake per week: “0 times”, “1 time”, “2.5 times”, “5 times”, “7
times” and “14 times”.
BMI and weight status
Height and weight were measured by a standard calibrated scale and stadiometer.
BMI was calculated as weight in kg / (height in meters).
2
Cut-offs for overweight were
based on the criteria proposed by the International Obesity Task Force (IOTF) for
children and adolescents aged 2-18 years.(34) The IOTF published a series of sex- and
age-specific BMI cut-offs that were developed using international data sets from Brazil,
Russia, Hong Kong, Singapore, the Netherlands, the United Kingdom, and the United
States. The age and gender specific BMI curves pass through a BMI of 25 kg/m
2
(the
widely used adult cut-off) for “overweight” at age 18. In our study, overweight is defined
as a BMI equal to or greater than to the IOTF age- and gender- specific cut-offs.
Age and school level
Age in year and school grade at the time of measurement were collected by
questionnaires. In China, middle school level refers to students in 7
th
to 9
th
grade while

41

high school level refers to those in 10
th
to 12
th
grade. Accordingly, participants were
stratified into middle school (7
th
and 8
th
grade) and high school (10
th
and 11
th
grade).”
Parental income
Monthly parental income was reported by parents and was collapsed into 3
categories: “≤ 500 Yuan ($60 USD)”, “501 - 2000 Yuan ($61 - 240 USD)”, and “> 2000
Yuan (> $240 USD)”. During the period that the data was collected for this study,
poverty was defined as having an income less than 869 Yuan per capita (poverty
line).(175)
Parental education level
The highest education level attained by either father or mother was provided by students’
parents. The responses were divided into 3 categories: “below high school”, “high
school”, and “college”.
All measurements were translated into Chinese and back-translated into English.
Statistical Analysis
Participants with complete data for all measurements included in the present study
were compared to participants with incomplete data on any of the measures included in
the present study on demographic variables and BMI using independent sample t tests or
chi-square tests. Frequencies and means were used to present the descriptive statistics of
the study sample. The primary outcome of interest was overweight status (yes/no).
Independent sample t tests or chi-square tests were conducted to compare covariates such

42

as participation in VPA, time spent in sedentary behavior, sleep duration, vegetable
intake, fruit intake, sweets intake, snack intake, fast food intake, self-perception of health
status, parental education, parental income, school level, and gender by overweight status.
To correct for multiple comparisons, Bonferroni correction was applied.
Considering the hierarchical structure of the data (students nested within schools),
multi-level models were used to control for the random effects of schools, in which the
random intercepts were allowed to vary between schools. Multivariate multi-level logistic
modeling procedures (Proc Glimmix) were utilized to examine the associations between
overweight status and the following covariates: participation in VPA, time spent in
sedentary behaviors, sleep duration, vegetable intake, fruit intake, sweets intake, snack
intake, fast food intake, self-perception of health status, parental education, parental
income, gender, and school level. The association between overweight status and each
independent variable was evaluated by the odds ratio (OR) for overweight status, after
adjustments for city and all other covariates in the model. These same sets of multi-level
logistic models were also conducted for each city.
Logarithmic transformations were applied where necessary to achieve better
normality due to the skewness of some continuous variables (time spent in sedentary
behaviors). All continuous variables used in the multi-level modeling analyses were first
centered to the school mean of each school sub-sample to facilitate the interpretation of
the intercept and were then standardized to a mean of 0 and a standard deviation of 1 to

43

obtain standardized parameter estimates and strengthen the comparability. Analyses were
conducted with SAS 9.2 (SAS Institute Inc., Cary, NC, USA). The significance of the
findings was evaluated at the P<0.05 level.
2.3 RESULTS
There were no statistical differences on BMI, age, and parental education between
participants with and without complete data. However, those with complete data were
more likely to be female (P<0.001) and higher monthly parental income (P=0.011) (data
not shown).
Sample Characteristics
The study sample consisted of 9023 students (52.71% female; 15±1.71 years) from
140 schools across 7 cities. The average BMI was 20.79 kg/m
2
with 17.87% (N=1612)
classified as overweight and 82.13% (N=7411) classified as non-overweight. On average,
students participated in 2.89 VPA sessions of at least 20 minutes per week, spent 119.14
minutes in sedentary behaviors per day, and obtained 7.97 hours of sleep per night. Mean
frequency of vegetable consumption per week during the past 30 days was 7.54 times;
fruit 6.50 times; sweets 4.06 times; snack 3.11 times; and fast food 0.43 times.

44

Correlates by Overweight Status
As shown in Table 2-1, overweight youths obtained fewer hours of sleep per night
(P=0.002) than non-overweight youths. Additionally, those who were overweight were
more likely to be male (P<0.001), were younger (P<0.001), engaged in VPA more
frequently (P<0.001), reported better perceived health status (P<0.001), consumed
vegetables (P<0.001) and fruits (P=0.024) more frequently, consumed sweets (P<0.001),
snack (P<0.001), and fast food (P<0.001) less frequently, had higher parental levels of
education (P<0.001), and had higher parental levels of income (P=0.017) when compared
to those who were not overweight. After Bonferroni adjustment of the significance level
of P=0.005 for multiple comparisons, most of the significant associations remained. Only
findings for parental income and fruit intake became non-significant.
Multi-level Analyses
Table 2-2 presents the effects of the covariates on the odds of being overweight. In
comparison to youths who slept fewer than 7 hours per night, those who slept 7 - 7.99
hours per day, 8 - 8.99 hours per day, and 9 hours or more per night had a 23%, 27%, and
35% lower odds of being overweight, respectively (7-7.99 hours: 95% CI=0.63-0.92;
8-8.99 hours: 95% CI=0.63-0.84, ≥ 9 hours: 95% CI=0.52-0.82). Youths who spent more
time in sedentary activities were more likely to be overweight (OR=1.11, 95%
CI=1.04-1.17); high school students had a lower likelihood of being overweight than
middle school students (OR=0.74, 95% CI=0.64-0.84). Some results were inconsistent

45

with the obesity research from the West. Boys were 1.61 times more likely to be
overweight than females (95% CI=1.43-1.81). Youths whose parents had a college
(OR=1.47, 95% CI=1.24-1.75) or a high school education (OR=1.29, 95% CI=1.10-1.50)
were more likely to be overweight than youths whose parents had not finished high
school. Overweight was more common among youths who reported better perceived
health status (OR=1.11, 95% CI=1.05-1.18). Finally, higher vegetable intake and lower
sweets and fast food intake were related with greater odds of being overweight
(vegetables: OR=1.10, 95% CI=1.04-1.17; sweets: OR=0.77, 95% CI=0.72-0.82; fast
food: OR=0.93, 95% CI=0.87-0.99), while those who participated in VPA more
frequently (OR=1.14, 95% CI=1.01-1.28) had a greater chance of being overweight.
Table 2-3~2-9 show the effects of the covariates on the odds of being overweight by
city
a
. Similar to the findings based on whole study sample, Chinese youth were more
likely to be overweight if they were boys (found in ChengDu City, HangZou City,
ShenYang City, Wuhan City, Harbin City, and Qingdo City), were younger (found in
HangZou City, ShenYang City, and Qingdo City), had higher SES (found in ChengDu
City, ShenYang City, Wuhan City, Harbin City, and Qingdo City), participated more in
VPA (found in Wuhan City), spent more time being sedentary (found in ShenYang City,

a
For ShenYang City, estimates from logistic regression model were presented as the multi-level logistic
modeling did not converge. This may be explained by the low intraclass correlation for ShenYang (0.01) as
compared to other cities (0.02-0.1). Additionally, the same logistic regression models were tested for all other
6 cities and these results were consistent to those from mixed modeling procedures.

46

Wuhan City, and Harbin City), had shorter sleep duration (found in ShenYang City,
Wuhan City, and Qingdo City), had a greater intake of vegetable (found in ShenYang
City and Wuhan City), had a lower intake of sweets (found in ChengDu City, HangZou
City, ShenYang City, and Qingdo City), had a lower intake of fast food (found in
HangZou City), and had a better self-perceived health status (found in ShenYang City
and Harbin City).
2.4 DISCUSSION
This is one of the first studies to examine the associations between overweight status,
sleep duration, physical activity, sedentary behavior, diet, and general demographics in a
geographically representative and socioeconomically diverse sample of Chinese
adolescents. Our findings provide important insights into roles that weight-related
behavioral correlates identified from the Western literature play in pediatric overweight
in Chinese adolescents. While certain factors concurred with obesity research conducted
in the West, the majority of our findings run contradictory to findings in Western
cultures.
Consistent with studies of Western populations,(6, 38, 66) we found screen-related
sedentary activities (eg, TV and computer) and short sleep duration were related to
greater odds of being overweight. After adjusting for VPA, SES, and all other covariates
in the model, the odds of being overweight were 1.1 times higher for each 83 minute

47

increment (1 SD) of sedentary behavior. These technology-related sedentary activities
were originally available only to the higher SES populations in China. However, as rapid
economic development has made these technologies more readily available to a wider
population, the adverse influence of small screen recreation on pediatric adiposity
appears to be crossing socioeconomic as well as cultural boundaries. Several hypotheses
have been proposed to explain the sleep-overweight relationship.(27, 235) One theory
proposes that individuals with shorter sleep duration have lower energy expenditure,
given that sleep deprivation generally leads to changes in the structure of sleep stages and
results in fatigue, daytime sleepiness, and low physical activity levels.(181) Another
theory proposes that sleep deprivation leads to an elevation of ghrelin and a depression of
leptin, and that these changes are sufficient to stimulate appetite and hunger.(116, 195)
While our data on sedentary behavior and sleep concurred with studies in Western
cultures, several East-West paradoxes emerged. First, we found that higher parental
income was related to overweight, but when parental education and other covariates were
adjusted, parental income was not significant while parental education remained
significant. A mix of higher parental income and higher parental education seems to be
related to pediatric obesity in China, while they are related to lower BMI in Western
countries.(79) It has been suggested earlier that higher SES youth were more likely to be
overweight in China.(212) Similar to the concept of fundamental causes of disease,(115)
previous research indicated that the relations of SES and overweight varied across

48

countries with different socioeconomic development levels: the positive relations
between SES and overweight were commonly seen in countries at an early stage of
socioeconomic development, but the overweight burden could then shift to lower-SES
population with ongoing socioeconomic change.(189) One possible explanation for the
positive association between parental education and overweight in China is that
modernization and westernization accompanied by the rapid economic changes has
resulted in a transformation of lifestyle.(212)Unhealthy shifts in lifestyles and
weight-related behaviors that are currently only available to the wealthy in China include
increased automobile ownership (bicycling used to be the popular mode of
transportation),(29) increased time in screen-related sedentary behavior,(122) increased
availability of energy dense foods (meat and processed foods),(85) and the adoption of
Western fast food (usually more expensive than traditional Chinese foods).(29) In our
study, parental education and parental income were moderately and positively correlated.
Therefore, it is reasonable to expect that the youth of more highly educated parents would
be wealthier and have greater access to modern and more expensive lifestyles. The above
factors may explain why overweight is more common in youths whose parents have
achieved higher education levels.
According to the National Health and Nutrition Examination Survey (NHANES) in the
US population,(153) the prevalence of overweight was similar between boys (35.3%) and
girls (34.1%) aged 6-19 years. In contrast to this and to most studies in Western

49

populations, we found that boys were more likely to be overweight than girls. This has
been found by other studies on Chinese youth.(20, 26) One possible explanation for the
significant gender difference in overweight prevalence is that body shape ideals differ
between boys and girls in China.(229) One study found that girls preferred a thinner body
figure while boys preferred a heavier and a more muscular shape.(176) In addition, girls
expressed greater weight concerns than did boys.(229) Thus, Chinese girls are more
likely to be aware of their weight and might be more prone to diet to prevent weight gain.
Another potential explanation is the social-cultural preference for sons over daughters in
Chinese culture. (212) Traditionally, girls were expected to help with the housework,
while boys may be overprotected from labor and somewhat ‘overindulged’ by
parents.(99)
Inconsistent with findings in Western society,(153) we found that younger youths were
more likely to be overweight than older youths in China after adjusting for SES. Among
Chinese youth, participation in physical activity outside of school is relatively low;
walking or biking to school is the most common type of moderate physical activity.(207)
Younger youths were less likely to ride a bicycle or walk to school by themselves and
more likely to be transported by their parents.(114) Perhaps, a lower incidence of active
commuting to school, an important opportunity for physical activity, offers a partial
explanation for the fact that overweight was more common in the younger youth.

50

In contrast to most findings in studies carried out in Western cultures, we found that,
frequent consumption of vegetables was actually related to greater odds of overweight in
Chinese youth after controlling for SES. Diets rich in fruits and vegetables have generally
associated with lower caloric intake in studies conducted in the West.(61, 221) However,
our findings might be explained by differences in cooking methods. In China, the 2 most
common methods of cooking vegetables are deep-frying and stir-frying, both of which
involve generous use of oil.(126) Hence, frequent vegetable consumption could result in
higher intake of energy from vegetable oil(185) and be linked to weight gain.
Interestingly, we also found that overweight Chinese youth reported a lower frequency of
sweets intake and fast food intake. The traditional Chinese diet consists of a variety of
high-glycemic carbohydrates with rice as the staple grain.(48) Previous research indicates
that a high-glycemic diet favors weight gain and obesity.(120) Overweight youth might
be consuming greater amounts of foods (eg, rice and noodles) associated with increased
caloric intake and increased adiposity. However, intake of these high glycemic and
possibly obesogenic foods were not measured in the current study. In addition, only
frequency of food intake was measured in our study. Information related to portion size
was not collected. Portion size has been shown to be related to obesity(234) and might be
a crucial determinant of overweight in Chinese youth. Another potential explanation is
that overweight youth might tend to under-report their intake of sweets and fast food, 2
types of high-energy dense foods, due to social desirability.(54, 97)

51

A final finding that was inconsistent with research conducted in Western populations
was that a higher frequency of participation in VPA was associated with greater odds of
being overweight. Three potential explanations are proposed here. First, as the level of
VPA was measured subjectively, overweight youth might be more likely to perceive
certain activities as having vigorous intensity while the same activities might be
described as moderate or light among non-overweight youth. In addition, no information
was provided on moderate or light intensity levels of physical activity, to which youth
devoted most of their time. It is possible that overweight youth engage less in moderate
or light intensity levels of physical activity than non-overweight youth. Second, social
desirability might also explain why overweight Chinese youths reported a higher
frequency of participation in VPA. However, this over-reporting bias could not be
determined in our study as no objective measure of VPA was used. Future studies that
utilize both subjective and objective measures of physical activity and diet intake are
needed to investigate the effect of social desirability. Third, it is possible that some
overweight Chinese youth were trying to lose weight by engaging in physical activity or
by reducing intake of high-energy dense foods. This in turn, may explain our inconsistent
findings.
Stratifying by city, additional analyses were conducted to further investigate whether
the associations between weight-related correlates and odds of overweight were
consistent across the 7 cities. In general, the results by city were similar to reported for

52

the overall sample. More specifically, these results indicated that the associations
between the correlates and being overweight were similar across 7 cities during 2002,
though these cities may differ in their regional cultures, geographical locations, and
socioeconomic development. With the ongoing economic transitions in China, the speed
of socioeconomic development may continue to vary by city, possibly impacting the
moderating effects of city characteristic (e.g., urban vs. rural cities) on the associations
between weight-related correlates and overweight risk in the future.
Strengths and Limitations
The large and diverse Chinese youth population used in the present analysis affords us
the statistical power necessary to meaningfully evaluate stratified results. Multivariate
multi-level logistic modeling procedures allowed us to examine the true relationships
between the odds of overweight and the variables of interest with adjustments for
possible confounders. There are several limitations that should be considered in
interpreting the results. First, the cross-sectional nature of our study limits inferences
regarding causality. Second, some bias might have been introduced by the use of
self-report rather than objective measures of activity level. In addition, this study only
assessed time spent doing VPA and 2 kinds of sedentary activities. Future research
should also measure moderate and light intensity of physical activity. Also, more
sedentary activities such as reading, talking on the phone, playing video games should be
included. Third, only participants with complete data were included in the current study,

53

hence, the findings may not generalizable to those without complete data. Fourth, in order
to keep the survey short, assessments of some constructs relied on a small number of
items.
2.5 CONCLUSIONS
This study contributes new information on the modifiable determinants of overweight
that could be targeted for interventions in Chinese youth. Our results indicate that, in
keeping with findings in Western populations,(6, 38, 121) both short sleep duration per
night and greater time spent in sedentary activities were significant risk factors for
overweight. Therefore, interventions to decrease sedentary behaviors and improve sleep
habits may be promising approaches for preventing pediatric overweight. In contrast with
what has been found among Western youth, we found that overweight Chinese youth
were male, were younger, participated more in VPA, had higher SES, a greater intake of
vegetables, and a lower intake of sweets and fast food. This ‘East-West paradox’ might
be driven by rapid shifts in the economic climate in China that makes attractive yet
unhealthy lifestyle and behavioral choices available primarily to the wealthy and
educated population. As the Chinese economy continues to grow, it is crucial to track
these paradoxical relationships, which may or may not ‘flip’ to match relationships we
now see in Western countries. The problem of pediatric overweight in China is complex,
and it is due, in part, to cultural phenomena specific to China. Beyond the prevention
strategies learned from Western countries, solutions to the pediatric obesity epidemic in

54

China should consider the impact of the interactions between economy, culture and
lifestyle on overweight in Chinese youth.

55

Table 2-1: Characteristic of Covariates of Interest by Overweight Status
Variable

Overweight
(N=1612)
Non-overweight
(N=7411)
P-value

Mean (±SD) /
N (%)
Mean (±SD) / N
(%)

Gender <0.001 *** †
Female 669 (41.50%) 4087 (55.15%)
Male 943 (58.50%) 3324 (44.85%)
School Level <0.001 *** †
Middle School 765 (47.46%) 3030 (40.89%)
High School 847 (52.54%) 4381 (59.11%)
Parental Education <0.001 *** †
Below High School 348 (21.59%) 2024 (27.31%)
High School 669 (41.50%) 3142 (42.40%)
College 595 (36.91%) 2245 (30.29%)
Monthly Parental Income 0.017 *
0 - 500 Yuan 179 (11.10%) 945 (12.75%)
501 - 2000 Yuan 987 (61.23%) 4641 (62.62%)
> 2000 Yuan 446 (27.67%) 1825 (24.63%)
Participation in VPA
<0.001 *** †
< 3 times per week 826 (51.24%) 4212 (56.83%)
≥ 3 times per week 786 (48.76%) 3199 (43.17%)
Time Spent on Sedentary
Behavior (minutes per day)
122.10
(±86.96) 118.50 (±82.15)
0.130
Sleep Duration (hours per night)
7.89 (±1.12) 7.99 (±1.09)
0.002 ** †
Frequency of Vegetable Intake
a
7.88 (±4.53) 7.46 (±4.48) <0.001 *** †
Frequency of Fruit Intake
a
6.72 (±4.10) 6.46 (±4.15) 0.024 *

56

Table 2-1 (continued)
Variable

Overweight
(N=1612)
Non-overweight
(N=7411)
P-value

Mean (±SD) /
N (%)
Mean (±SD) / N
(%)

Frequency of Sweets Intake
a
3.36 (±3.08) 4.21 (±3.55) <0.001 *** †
Frequency of Snack Intake
a
2.48 (±2.90) 3.24 (±3.42) <0.001 *** †
Frequency of Fast Food Intake
a
0.35 (±1.02) 0.44 (±1.12) <0.001 *** †
Self-perception of Health Status 3.01 (±0.81) 2.88 (±0.82) <0.001 *** †
Abbreviations: VPA= vigorous physical activity
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
* P<0.05, ** P<0.01, *** P<0.001 for 2-tailed tests
† P< 0.004 adjusted significance for 2-tailed tests with application of Bonferroni
correction for multiple comparisons
a. The frequency of dietary intake refers to times per week during last 30 days

57

Table 2-2: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest (N=9023)
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 1.61 (95% CI:1.43-1.81)
School Level
Middle school 1.00
High school 0.74 (95% CI: 0.64-0.84)
Parental Education Level
Below High School 1.00
High School 1.29 (95% CI: 1.10-1.50)
College 1.47 (95% CI: 1.24-1.75)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 1.06 (95% CI: 0.88-1.27)
> 2000 Yuan 1.23 (95% CI: 0.98-1.54)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 1.14 (95% CI: 1.01-1.28)
Time Spent on Sedentary Behavior (+1 SD)
b,c
1.11 (95% CI :1.04-1.17)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 0.77 (95% CI: 0.63-0.92)
8 - 8.99 hours 0.73 (95% CI :0.63-0.84)
≥ 9 hours 0.65 (95% CI :0.52-0.82)
Frequency of Vegetable Intake (+1 SD)
c, d
1.10 (95% CI : 1.04-1.17)
Frequency of Fruit Intake (+1 SD)
c, d
1.03 (95% CI : 0.97-1.10)

58

Table 2-2 (continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Sweet Intake (+1 SD)
c, d
0.77 (95% CI: 0.72-0.82)
Frequency of Snack Intake (+1 SD)
c, d
0.93 (95% CI : 0.87-1.00)
Frequency of Fast Food Intake (+1 SD)
c, d
0.93 (95% CI: 0.87-0.99)
Self-perception of Health Status (+1 SD)
c
1.11 (95% CI: 1.05-1.18)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for city and all other covariates included
in the model
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

59

Table 2-3: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for ChengDu
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 1.97 (95% CI:1.39-2.80)
School Level
Middle school 1.00
High school 0.93 (95% CI: 0.62-1.42)
Parental Education Level
Below High School 1.00
High School 1.21 (95% CI: 0.74-1.99)
College 1.92 (95% CI: 1.13-3.27)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 1.57 (95% CI: 0.79-3.12)
> 2000 Yuan 2.19 (95% CI: 1.03-4.67)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 1.03 (95% CI: 0.73-1.46)
Time Spent on Sedentary Behavior (+1 SD)
b, c
0.95 (95% CI :0.79-1.13)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 0.76 (95% CI: 0.38-1.52)
8 - 8.99 hours 0.89 (95% CI :0.57-1.39)
≥ 9 hours 1.02 (95% CI :0.61-1.69)
Frequency of Vegetable Intake (+1 SD)
c, d
1.05 (95% CI : 0.89-1.24)

60

Table 2-3 (continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Fruit Intake (+1 SD)
c, d
1.15 (95% CI : 0.97-1.36)
Frequency of Sweet Intake (+1 SD)
c, d
0.63 (95% CI: 0.52-0.78)
Frequency of Snack Intake (+1 SD)
c, d
0.96 (95% CI : 0.79-1.17)
Frequency of Fast Food Intake (+1 SD)
c, d
0.88 (95% CI: 0.71-1.10)
Self-perception of Health Status (+1 SD)
c
0.98 (95% CI: 0.83-1.17)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for all other covariates included in the
model
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

61

Table 2-4: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for HangZhou
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 2.34 (95% CI:1.54-3.57)
School Level
Middle school 1.00
High school 0.52 (95% CI: 0.32-0.87)
Parental Education Level
Below High School 1.00
High School 1.32 (95% CI: 0.74-2.35)
College 1.70 (95% CI: 0.95-3.04)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 0.44 (95% CI: 0.12-1.54)
> 2000 Yuan 0.43 (95% CI: 0.12-1.49)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 1.03 (95% CI: 0.68-1.56)
Time Spent on Sedentary Behavior (+1 SD)
b, c
1.06 (95% CI :0.86-1.31)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 0.64 (95% CI: 0.32-1.27)
8 - 8.99 hours 0.62 (95% CI :0.37-1.05)
≥ 9 hours 0.71 (95% CI :0.33-1.50)
Frequency of Vegetable Intake (+1 SD)
c, d
1.08 (95% CI : 0.88-1.31)

62

Table 2-4 (continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Fruit Intake (+1 SD)
c, d
1.11 (95% CI : 0.90-1.36)
Frequency of Sweet Intake (+1 SD)
c, d
0.73 (95% CI: 0.55-0.97)
Frequency of Snack Intake (+1 SD)
c, d
0.96 (95% CI : 0.76-1.22)
Frequency of Fast Food Intake (+1 SD)
c, d
0.73 (95% CI: 0.55-0.96)
Self-perception of Health Status (+1 SD)
c
1.11 (95% CI: 0.90-1.36)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for all other covariates included in the
model
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

63

Table 2-5: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for ShenYang
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 1.59 (95% CI:1.41-1.78)
School Level
Middle school 1.00
High school 0.75 (95% CI: 0.67-0.84)
Parental Education Level
Below High School 1.00
High School 1.27 (95% CI: 1.09-1.46)
College 1.46 (95% CI: 1.24-1.72)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 1.05 (95% CI: 0.87-1.25)
> 2000 Yuan 1.10 (95% CI: 0.89-1.36)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 1.09 (95% CI: 0.97-1.22)
Time Spent on Sedentary Behavior (+1 SD)
b, c
1.11 (95% CI :1.05-1.17)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 0.71 (95% CI: 0.59-1.85)
8 - 8.99 hours 0.76 (95% CI :0.66-0.87)
≥ 9 hours 0.66 (95% CI :0.53-0.82)
Frequency of Vegetable Intake (+1 SD)
c, d
1.10 (95% CI : 1.04-1.16)

64

Table 2-5 (Continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Fruit Intake (+1 SD)
c, d
1.03 (95% CI : 0.97-1.10)
Frequency of Sweet Intake (+1 SD)
c, d
0.78 (95% CI: 0.72-0.83)
Frequency of Snack Intake (+1 SD)
c, d
0.94 (95% CI : 0.88-1.01)
Frequency of Fast Food Intake (+1 SD)
c, d
0.94 (95% CI: 0.88-0.10)
Self-perception of Health Status (+1 SD)
c
1.11 (95% CI: 1.05-1.18)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for all other covariates included in the
model. As the multi-level logistic modeling did not converge, estimates from logistic
regression model were presented as the approximations for multi-level logistic modeling.
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

65

Table 2-6: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Wuhan
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 1.60 (95% CI:1.14-2.25)
School Level
Middle school 1.00
High school 0.95 (95% CI: 0.61-1.46)
Parental Education Level
Below High School 1.00
High School 1.57 (95% CI: 0.86-2.88)
College 2.26 (95% CI: 1.17-4.37)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 1.08 (95% CI: 0.65-1.79)
> 2000 Yuan 0.97 (95% CI: 0.51-1.85)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 0.98 (95% CI: 0.70-1.39)
Time Spent on Sedentary Behavior (+1 SD)
b, c
1.20 (95% CI :1.01-1.44)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 0.52 (95% CI: 0.31-0.87)
8 - 8.99 hours 0.63 (95% CI :0.42-0.93)
≥ 9 hours 0.48 (95% CI :0.24-0.97)
Frequency of Vegetable Intake (+1 SD)
c, d
1.23 (95% CI : 1.04-1.45)
Frequency of Fruit Intake (+1 SD)
c, d
1.07 (95% CI : 0.89-1.29)

66

Table 2-6 (Continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Sweet Intake (+1 SD)
c, d
0.83 (95% CI: 0.67-1.02)
Frequency of Snack Intake (+1 SD)
c, d
0.89 (95% CI : 0.73-1.09)
Frequency of Fast Food Intake (+1 SD)
c, d
0.98 (95% CI: 0.82-1.17)
Self-perception of Health Status (+1 SD)
c
1.13 (95% CI: 0.95-1.34)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for all other covariates included in the
model
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

67

Table 2-7: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Harbin
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 1.56 (95% CI:1.12-2.17)
School Level
Middle school 1.00
High school 0.83 (95% CI: 0.59-1.17)
Parental Education Level
Below High School 1.00
High School 1.65 (95% CI: 1.08-2.52)
College 1.87 (95% CI: 1.13-3.10)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 0.76 (95% CI: 0.42-1.39)
> 2000 Yuan 0.85 (95% CI: 0.56-1.27)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 1.42 (95% CI: 1.02-1.97)
Time Spent on Sedentary Behavior (+1 SD)
b, c
1.23 (95% CI :1.03-1.45)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 1.24 (95% CI: 0.69-2.20)
8 - 8.99 hours 0.90 (95% CI :0.59-1.35)
≥ 9 hours 0.71 (95% CI :0.39-1.31)
Frequency of Vegetable Intake (+1 SD)
c, d
0.88 (95% CI : 0.75-1.04)
Frequency of Fruit Intake (+1 SD)
c, d
0.97 (95% CI : 0.82-1.15)

68

Table 2-7 (Continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Sweet Intake (+1 SD)
c, d
0.98 (95% CI: 0.81-1.18)
Frequency of Snack Intake (+1 SD)
c, d
0.87 (95% CI : 0.71-1.07)
Frequency of Fast Food Intake (+1 SD)
c, d
0.94 (95% CI: 0.79-1.12)
Self-perception of Health Status (+1 SD)
c
1.25 (95% CI: 1.08-1.46)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for all other covariates included in the
model
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

69

Table 2-8: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Kunming
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 1.23 (95% CI:0.85-1.18)
School Level
Middle school 1.00
High school 0.89 (95% CI: 0.60-1.33)
Parental Education Level
Below High School 1.00
High School 1.23 (95% CI: 0.79-1.91)
College 1.56 (95% CI: 0.95-2.55)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 1.20 (95% CI: 0.68-2.14)
> 2000 Yuan 1.24 (95% CI: 0.62-2.48)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 1.13 (95% CI: 0.79-1.61)
Time Spent on Sedentary Behavior (+1 SD)
b, c
1.14 (95% CI :0.95-1.38)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 0.80 (95% CI: 0.49-1.31)
8 - 8.99 hours 0.82 (95% CI :0.54-1.27)
≥ 9 hours 0.79 (95% CI :0.39-1.57)
Frequency of Vegetable Intake (+1 SD)
c, d
1.09 (95% CI : 0.92-1.30)

70

Table 2-8 (Continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Fruit Intake (+1 SD)
c, d
0.92 (95% CI : 0.76-1.11)
Frequency of Sweet Intake (+1 SD)
c, d
0.93 (95% CI: 0.76-1.14)
Frequency of Snack Intake (+1 SD)
c, d
0.92 (95% CI : 0.76-1.12)
Frequency of Fast Food Intake (+1 SD)
c, d
0.94 (95% CI: 0.78-1.13)
Self-perception of Health Status (+1 SD)
c
1.16 (95% CI: 0.97-1.39)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for all other covariates included in the
model
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

71

Table 2-9: Adjusted Odds Ratio (OR) and 95% CI for Multi-level Logistic
Regression Model of Overweight Status and Covariates of Interest for Qingdo
Variable
a
Adjusted OR (95% CI)
Gender
Female 1.00
Male 1.97 (95% CI:1.51-2.57)
School Level
Middle school 1.00
High school 0.61 (95% CI: 0.47-0.81)
Parental Education Level
Below High School 1.00
High School 1.57 (95% CI: 0.89-2.80)
College 1.92 (95% CI: 1.03-3.57)
Family Income
≤ 500 Yuan 1.00
501 - 2000 Yuan 1.19 (95% CI: 0.84-1.68)
> 2000 Yuan 1.02 (95% CI: 0.69-1.51)
Participation in Vigorous Physical Activity
< 3 times per week 1.00
≥ 3 times per week 1.00 (95% CI: 0.77-1.31)
Time Spent on Sedentary Behavior (+1 SD)
b, c
1.10 (95% CI :0.94-1.29)
Sleep Duration Per Night
< 7 hours 1.00
7 - 7.99 hours 0.56 (95% CI: 0.34-0.92)
8 - 8.99 hours 0.75 (95% CI :0.54-1.03)
≥ 9 hours 0.72 (95% CI :0.43-1.21)
Frequency of Vegetable Intake (+1 SD)
c, d
1.11 (95% CI : 0.97-1.26)
Frequency of Fruit Intake (+1 SD)
c, d
1.09 (95% CI : 0.95-1.26)

72

Table 2-9 (Continued)
Variable
a
Adjusted OR (95% CI)
Frequency of Sweet Intake (+1 SD)
c, d
0.75 (95% CI: 0.64-0.88)
Frequency of Snack Intake (+1 SD)
c, d
0.88 (95% CI : 0.74-1.05)
Frequency of Fast Food Intake (+1 SD)
c, d
0.99 (95% CI: 0.85-1.15)
Self-perception of Health Status (+1 SD)
c
1.02 (95% CI: 0.89-1.17)
Overweight was defined based on the IOTF age- and gender- specific cut-offs in children
and adolescents (34)
a. Each parameter estimate has been adjusted for all other covariates included in the
model
b. Log-transformed values were used
c. Variables were standardized to a mean of 0 and a standard deviation of 1 to obtain
standardized parameter estimates
d. The frequency of dietary intake refers to times per week during last 30 days

73

Chapter 3 INFLUENCE OF PHYSICAL ACTIVITY AND SEDENTARY
BEHAVIOR ON METABOLIC SYNDROME IN MINORITY YOUTH
3.1 INTRODUCTION
The metabolic syndrome (MetS) refers to a cluster of cardiovascular and Type 2
diabetes risk factors which include: elevated triglycerides, low high-density lipoprotein
(HDL)-cholesterol, abdominal adiposity, hyperglycemia, and elevated blood
pressure.(199) Based on current estimates from the National Health and Nutrition
Examination Survey (NHANES) from 1999 to 2002, the prevalence of MetS in youth
aged 12-19 years old varies widely from 2.0-9.4% reaching 44.2% among obese
children.(35) Our group has shown that over 30% of overweight Latino children in the
Los Angeles area have MetS.(39) We also have shown that children with persistent MetS
were at progressively greater risk for Type 2 diabetes.(211) This is especially important
in light of the fact that other research has found that MetS in childhood predicted adult
MetS and Type 2 diabetes mellitus as much as 25 to 30 years later.(143) Thus, to reduce
the increasing prevalence of such chronic diseases in adults, it is important to identify
lifestyle behaviors that can be modified during childhood.
Two of the key modifiable lifestyle factors which influence metabolic health are
physical activity(159) and sedentary behavior.(67) Lower levels of physical activity and
higher levels of sedentary behavior are known to be associated with progression toward
MetS in adults.(53, 67) However, little is known regarding these relationships in youth,

74

particularly in minority populations.(55, 57, 159, 193) In a review article on physical
activity and MetS in youth,(193) only one (18) of the six studies conducted focused on a
minority population (Latino). More research is therefore needed to understand how
physical activity and sedentary behavior affect MetS in minority youth.
Another gap in the literature on activity levels and MetS is that while there are some
studies on physical activity, less is known about sedentary behavior. Although engaging
in physical activity and being sedentary may seem like two sides of the same coin,
research has suggested that these two behaviors should be treated as different dimensions,
rather than a continuum.(208) The majority of existing studies measure physical activity
either by subjective or objective measures. Subjective measures are susceptible to
memory bias and social desirability,(57) while objective measures suffer from issues of
compliance and the inability to register certain activities.(187, 190) Using both types of
measures may complement the limitations inherent in each measurement modality and
hence provide more comprehensive estimations in activity levels.
Therefore, the purpose of this study was to investigate the associations between
physical activity, sedentary behavior, and MetS in Latino and African American youth.
Activity levels were measured objectively by accelerometry and subjectively by the
3-Day Physical Activity Recall (3DPAR). We hypothesized that youth with MetS would
have lower levels of MVPA and higher levels of sedentary behavior than those without.
We also hypothesized that higher levels of sedentary behavior would be associated with

75

MetS while higher levels of physical activity would be protective against MetS.
Differences in relationships between MetS and activity levels by measurement modality
were examined.
3.2 METHODS
Participants
Data for this sample was taken from baseline measures in three related pediatric
obesity studies that share a set of common methods and measures: TRANSITIONS
(Insulin Resistance and Declining Physical Activity Levels in African American and
Latina girls), SANO (Strength and Nutrition Outcomes for Latino Adolescents), and
STAND (Strength Training and Nutrition Development in African American Youth). Of
187 participants who completed a baseline visit, 7 were missing for activity data by
3DPAR and 75 were missing for activity data by accelerometry. The final sample
therefore consisted of 105 participants (Latino 74.3%, female 75.2%, 13±3 years) who
have completed responses to both 3DPAR and accelerometry.
Participants were recruited through medical clinics, churches, community centers,
local schools, and advertisements in Los Angeles County. Detailed methods for these
studies have been published previously.(43, 154, 210) Participants were excluded if they
were using a medication or had been diagnosed with a condition known to influence body
composition, fat distribution, or insulin/glucose metabolism or if they had been diagnosed
with diabetes by fasting plasma glucose (≥ 126 mg/dl) at the time of entry to the study.

76

The studies were approved by the University of Southern California (USC) Institutional
Review Board. Written informed consent and assent were obtained from all parents and
children before any testing procedures.
Protocol
Youth arrived in the afternoon at USC General Clinical Research Center (GCRC).
Weight and height were measured three times to the nearest 0.1 kg and 0.1 cm,
respectively, using a beam medical scale and wall-mounted stadiometer by trained
medical professionals. Body mass index (BMI) percentiles for age and gender were
determined based upon established Centers for Disease Control and Prevention normative
curves.(222) Sitting blood pressure was measured in triplicate using the right arm after
the participant had rested for 5 minutes. Waist circumference was measured in triplicate
at the umbilicus and recorded to the nearest 0.1 cm. Body composition (total fat mass and
total lean tissue mass) was measured by air plethysmography (BodPod
TM
) and pubertal
Tanner stage was assessed by a licensed pediatric health care provider.(196) Fasting
blood samples were measured for triglycerides, HDL-cholesterol, and fasting glucose.

77

Definition of the Metabolic Syndrome (MetS)
Because no standard definition for MetS exists in pediatrics, we used a combination
of pediatric definitions proposed by Cruz et al(41) and Cook et al,(36, 184)
a
who applied
pediatric cutoffs to the Adult Treatment Panel III definition(150). This definition has
been applied in our previous studies.(41, 184) To be classified as having MetS,
participants had to have at least three of the following features: abdominal obesity (waist
circumference ≥ 90th percentile for age, sex, and ethnicity from Third National Health
and Nutrition Examination Survey data),(65) hypertriglyceridemia (triglycerides ≥ 90th
percentile for age, sex, and ethnicity),(87) low HDL-cholesterol (HDL-cholesterol ≤ 10th
percentile for age, sex, and ethnicity),(87) elevated blood pressure (systolic or diastolic
blood pressure ≥ 90th percentile adjusted for age, sex, and height), and hyperglycemia
(impaired fasting glucose ≥ 100 mg/dl).(5, 35, 36)
a

Physical Activity and Sedentary Behavior
Accelerometry data
Objective assessments of activity levels were obtained using the ActiGraph
(Actigraph, LLC; FT. Walton Beach, FL) model GT1M accelerometer. The ActiGraph is

a
Cook et al
101
originally used fasting glucose ≥110 mg/dl to define Impaired Fasting Glucose. The criteria for
Impaired Fasting Glucose has been updated by the American Diabetes Association as a fasting glucose ≥100
mg/d.
108
The revised cut-point has been used in Cook et al’s article in 2008
9
and will be used in the proposed
Study 2 and 3.

78

a uniaxial accelerometer that measures acceleration in the vertical plane and it has been
shown as a valid and reliable measure of physical activity for children and
adolescents.(169) The participants were instructed to wear the device on the right hip
during waking hours, with the exception of time spent in bathing or swimming activities.
Accelerometer data were downloaded and processed using a SAS program developed for
use with NHANES physical activity monitor data (available at:
http://riskfactor.cancer.gov/tools/nhanes_pam). The raw data is processed to calculate the
minutes of non-wear (defined by an interval of sixty or more consecutive minutes of zero
activity intensity counts, with exceptions for up to 3 minutes of 0-100 counts) and the
minutes of wear time.
Although participants were instructed to wear the accelerometers for seven
consecutive days, the wear time varied. In keeping with prior research,(130, 204) a valid
day of wear was defined as having at least ten hours of wear-time, and only participants
with four or more valid days of data were included in the analyses. The total number of
minutes spent in moderate-to-vigorous physical activity (MVPA) and sedentary activity,
was determined by summing minutes in a day where the count met the criterion for that
intensity. Then, these total minutes were averaged across the number of valid days to
obtain the mean minutes per day. The cut point for MVPA [≥4 metabolic equivalent
(METs)] was age-adjusted using the criteria from the Freedson group.(206) The
sedentary cut point of 100 counts, which was previously defined by Matthews et al.,(129)
has been validated in adolescents.(203)

79

Self-reported physical activity data
3DPAR was used to assess self-reported activity levels.(164, 219) 3DPAR has been
validated in adolescents.(164) Students identified different activities (from a list of 71
activities provided) to describe their activity in half-hour intervals during a day from 7:00
am – 12:00 am for three days, and rated how much effort (intensity level) they put into
each activity (light, moderate, hard, or very hard). Activity types were converted into
half-hour blocks of either light, moderate, or vigorous physical activity using a
combination of the intensity ratings provided by the participants and the compendium of
physical activities.(1) MVPA (≥4 METs) was created to be consistent to the
accelerometry variable. Half-hour blocks spent watching television/movies, playing video
games/surfing the internet, and talking on phone were coded separately as sedentary
behaviors. Daily total time spent in MVPA and sedentary behavior was obtained by
summing over the half-hour blocks over one day. Mean minutes per day was then
obtained by averaging total minutes across three days.
Statistical analysis
Participants with complete data included in the present study were compared to
participants without complete data on demographic variables and MetS using
independent samples t tests or chi-square tests.
For the preliminary analyses, participants were dichotomized into those with and
without MetS. Independent sample t-tests and chi-square tests were conducted to
compare demographic variables between these two groups. Differences in MVPA and

80

sedentary behavior by MetS were examined using analysis of covariance (ANCOVA).
Associations between MVPA, sedentary behavior, and the odds of MetS were evaluated
by multivariate logistic regressions for both accelerometry and 3DPAR. The effects of
MVPA and sedentary behavior on Mets were first examined separately and then were
included in the same model showing the independent associations between of MVPA and
sedentary behavior with MetS after adjusting for each other. Pearson and partial
correlations were used to assess the relationships of MVPA and sedentary behavior with
each individual feature of MetS. Following variables were included as covariates to
control for the potential confounding in ANCOVA models, multivariate logistic
regressions, and partial correlation: age, gender, ethnicity, pubertal Tanner stage, fat mass,
and lean tissue mass.
Logarithmic transformations were applied where necessary to achieve better
normality where variables were skewed (i.e. fat mass, lean tissue mass, MVPA, sedentary
behaviors, triglycerides, HDL-cholesterol, fasting glucose, and systolic/diastolic blood
pressure). All continuous independent variables were centered on the sample’s mean to
prevent collinearity. Analyses were performed with SAS v9.1 (SAS Institute, Cary, NC).
The significance of the findings was evaluated at the P <0.05 level.

81

3.3 RESULTS
There were no statistical differences in MetS and gender between participants with
and without complete data. However, those with complete data were more likely to be
younger (P =0.001) and Latino (P=0.015) (data not shown).
In this study, 16% (n=17) participants met the criteria for MetS. Demographic
characteristics by participants with and without MetS are shown in Table 3-1. Those with
MetS were more likely to be male (P =.02) and have a higher Tanner stage (P =.02).
Additionally, youth with MetS had higher BMI percentiles (P <.001), greater fat mass (P
<.001), greater lean tissue mass (P =.02), higher percent body fat (P =.002), and lower
percent lean tissue mass (P =.002). There were no statistically significant differences in
MetS prevalence by ethnic group and age. Activity levels by MetS are shown in Figure
3-1. Based on the accelerometry, youth with MetS spent 16% less time in MVPA (P
=.008) than those without. Based on the 3DPAR, those with MetS spent 39% more time
in sedentary behavior than those without (P =.008). No significant differences in
sedentary behavior as measured by accelerometry or in MVPA as measured by 3DPAR
between the two groups were found.
Table 3-2 presents the results of logistic regression models demonstrating the
associations between activity levels and odds of having MetS. When both MVPA and
sedentary behaviors were included in the same model for each measure, greater time
engaging in MVPA (by accelerometry: OR=0.49, 95%CI=0.25-0.98) was related to lower

82

odds of MetS while greater participation in sedentary behavior was related to higher odds
of MetS (by 3DPAR: OR=4.44, 95%CI=1.33-14.79). In addition, to determine if the
relationship between activity levels and odds of MetS were sensitive to measurement
modality, the same relationships were assessed adjusting for the other type of measure.
The relationship between activity levels and odds of MetS remained significant after
including MVPA by accelerometry and sedentary behavior by 3DPAR in the same model;
less time spent in MVPA as measured by accelerometry (OR=0.34, 95%CI=0.14-0.84)
and greater time spent in sedentary behaviors by 3DPAR (OR=6.42, 95%CI=1.65-25.02)
were related to greater odds of MetS.
We next examined Pearson and partial correlations between activity levels and each
individual feature of MetS (Table 3-3). We found that accelerometer and 3DPAR data
were significantly correlated to different individual risk factors. Compared to the
unadjusted correlations, the number of the significant findings reduced after adjusted for
body composition and other covariates. Based on the partial correlations, there were
inverse correlations between MVPA by accelrometry and fasting glucose (r=-0.21, P
=.03) as well as systolic blood pressure (r=-0.25, P =.01). In addition, sedentary behavior
by 3DPAR was negatively correlated with HDL-cholesterol (r=-0.21, P =.04) and
positively correlated systolic blood pressure (r=0.26, P =.009).
Table 3-4 provides an overall summary of our significant findings from the
subjective and objective measures.

83

3.4 DISCUSSION
To our knowledge, this is the first study to examine the relationships between MetS
and both physical activity and sedentary behavior in minority youth. Our results
confirmed the hypothesis that youth with MetS would have lower levels of MVPA and
higher levels of sedentary behavior than those without. Additionally, we found that lower
levels of physical activity and higher levels of sedentary behavior are associated with
greater metabolic risk and individual MetS features, after adjusting for body composition.
In general, our results support findings from adult studies(51, 53) in that physical
activity was shown to be inversely associated with metabolic risk. Similar results have
been found in previous studies in youth,(8, 101, 131, 137, 159, 172) the majority of
which relied on subjective self-report activity levels.(101, 131, 137, 159) However, that
studies using objective measures of physical activity (specifically accelerometry)(8, 57,
172) have been less conclusive. In the current study, although we examined physical
activity both subjectively and objectively, significant relationships between MVPA and
odds of MetS were found with objective measurement by accelerometry only.
Accordingly, we compared our findings with previous work using accelerometry. Among
the three studies in which physical activity was measured by accelerometer, Barge et al(8)
and Ekelund et al(57) found that increased physical activity levels were related to
reduced risk for MetS among youth in Denmark, Estonia, and Portugal. However, another
study in Swedish children and adolescents(172) found that physical activity was only
significantly related to MetS in adolescent girls, but not in adolescent boys or in young

84

children, regardless of gender. All three studies found that the association between
physical activity and metabolic risk became non-significant after adjusting for
cardiorespiratory fitness, which has been shown to have stronger associations with the
prevention of MetS than physical activity.(8, 172) It should be noted that two(8, 172) of
these three studies did not control for body composition, an important confounder, when
examining the influence of physical activity on metabolic risk. Although physical fitness
was not assessed in the current study, our findings add to the limited number of studies
applying objective measures of physical activity to assess its influence on MetS in
minority youth. We showed that higher levels of MVPA are related to lower metabolic
risk and that this association persists even after body composition, sedentary behavior,
and other relevant covariates are included in the models.
In contrast to the growing number of studies on physical activity and the MetS, very
few have examined the role of sedentary behavior on metabolic health. In adults, time
spent watching television and using a computer has been shown to positively linked with
MetS.(53, 67) Very little research has been conducted in youth, and findings were
inconclusive. One study(57) found a trend for a positive association between
self-reported TV viewing and risk for MetS (p=0.053), after adjustment for physical
activity. Given that youth spend the majority of their time in sedentary behaviors and
inactivity,(92) more research is needed to understand how sedentary behavior is related to
MetS. This study is the first to show that sedentary behavior (as measured subjectively by
3DPAR) is positively associated with metabolic risk, independent of body composition.

85

This association remains significant even after adjusting for self-reported and
objectively-measured physical activity. Our results showed that increased MVPA and
decreased sedentary behavior may lower all five metabolic risk factors, including low
HDL-cholesterol and larger waist circumference, the two most commonly seen criteria
for MetS in minority youth.(39) These findings suggest that interventions to prevent
MetS, need to target sedentary behavior in addition to physical activity.
When comparing the results by measurement method in this study (objectively and
subjectively measured activity levels), the significant findings regarding MVPA and
MetS were based on the accelerometer data while the majority of significant findings
regarding sedentary behavior were based on the data by 3DPAR. Thus, it appears that
accelerometer and 3DPAR may capture different aspects of activity levels. The
accelerometer, because it assesses physical activity objectively, avoids the recall bias
inherent in self-report measures and hence appears to detect differences in activity levels
better among youth with and without MetS. However, the accuracy of the accelerometry
is dependent on activity type.(187) Accelerometry is not able to capture activity intensity
well during movements with static hip position (i.e. biking, strength training) or for water
sports (i.e. swimming).(187) Additionally, because accelerometry does not accurately
measure movement of the upper body performed during low levels of activities, (109) it
may not accurately differentiate sedentary from light intensity activities. A unique
advantage of self-report measures, and in particular the 3DPAR, is their ability to provide
rich contextual data on the types sedentary activities, allowing researchers to focus on the

86

influence of activities that are considered recreational sedentary behavior such as small
screen recreation. Self-report of activity levels allows research on sedentary activities
directly related to advances in technology that promote prolonged inactive life styles and
have been found to have adverse effects on metabolic health. Despite its advantages,
there are limitations to self-reported measures of behavior. One relevant limitation is that
overweight youth may underreport their time in sedentary behavior due to social
desirability.(57) However, in this study this type of bias would only make the association
between sedentary behavior and MetS stronger than observed, and indicating that our
results add support for utilizing a carefully validated self-report measure (3DPAR) to
assess sedentary behavior.
Finally, consistent with the most recent NHANES data among adolescents,(35, 51)
we found that MetS is more common in males than in females. Based on the same
data,(35, 51) MetS is more prevalent in Latino youth compared to African American
youth. In our study, there was a similar trend, but this ethnic difference did not reach
significance and this is probably due to small number of African Americans in our study
with only 2 having MetS. Additionally, it is important to note that we conducted the same
set of analyses in Latino youth only, and the findings on the relationships between
physical activity, sedentary behavior, and Mets remained unchanged. We therefore
decided to report results using both ethnic groups to increase statistical power.

87

Strengths and Limitations
A particular strength of this study is that it uses both subjective and objective
measures of activity which allows comparison of the relationships to MetS across
measures. Another key strength is that we examine both physical activity and sedentary
behavior. Recent research has suggested that the protective effect of physical activity on
heath could be attenuated by sedentary behavior.(57) By including both physical activity
and sedentary behavior in the same model, we are able to investigate their independent
effects on metabolic risk.
In addition to the aforementioned limitations for accelerometer and 3DPAR, other
limitations of this study should also be considered. First, the cross-sectional study design
limits causal inferences. Second, the uneven sample sizes of the MetS, gender, and ethnic
groups could preclude additional findings. Third, this study was not designed to
specifically compare the true differences between accelerometer and 3DPAR and
although both measures were usually collected at the same time, there are some instances
where the data collection period for the two methods did not overlap completely. Future
studies with careful designs and population-based samples are warranted to determine the
differences in activity levels and their associations to metabolic-related biological
markers between accelerometer and self-reported physical activity.
3.5 CONCLUSIONS
In conclusion, our results contribute to the limited data for youth regarding the
beneficial effects of higher levels of physical activity on metabolic health. Furthermore,

88

this study extends previous findings by demonstrating that higher levels of sedentary
behaviors are related to greater metabolic risk. The relationships between physical
activity and metabolic risk are independent of sedentary behavior and body composition,
and, vice versa. Future interventions that aim to improve metabolic health in minority
youth should target both the promotion of physical activity and the reduction of sedentary
behavior. Additionally, it is recommended that both objective and subjective measures of
activity levels, when used in conjunction and carefully administered, may provide more
complete information on the impact of activity-related behaviors on metabolic health.

89

Figure 3-1: Activity Levels by Metabolic Syndrome

Abbreviations: MVPA, moderate-to-vigorous physical activity; MetS, metabolic
syndrome
Analysis were adjusted for age, gender, ethnicity, Tanner stage, fat mass, and lean tissue
mass
* P <.05, ** P <.01

**
**

90

Table 3-1: Demographic Characteristics by Metabolic Syndrome

MetS (N=17)
N (± SD/%)
Non-MetS (N=88)
N (± SD/%)
P-value
a

Gender
Female 9 (11.39%) 70 (88.61%)
.02 *
Male 8 (30.77%) 18 (69.23%)
Ethnicity
Latino 15 (19.23%) 63 (80.77%)
.15
African American 2 (7.41%) 25 (92.59%)
Age (Years) 14.00 (±2.37) 12.86 (±3.11) .16
Pubertal Tanner stage
.02 *
1 0 (0.00%) 19 (21.59%)
2 3 (17.65%) 17 (19.32%)
3 2 (11.76%) 2 (2.27%)
4 6 (35.29%) 12 (13.64%)
5 6 (35.29%) 38 (43.18%)
BMI percentile 98.47 (±1.63) 89.02 (±18.05) <.001 ‡
Total fat mass (Kg) 41.80 (±19.08) 25.21 (±15.83) <.001 ‡
Percent body fat 41.21 (±8.71) 32.21 (±10.97) .002 **
Total lean tissue mass (Kg) 56.03(±14.02) 45.96 (±16.25) .02 *
Percent lean body fat 58.79 (±8.71) 67.79 (±10.97) .002 **
Abbreviations: MetS, metabolic syndrome.
* P <.05, ** P <.01, ‡ P <.001
a. x
2
tests were used for categorical variables and independent t tests for continuous
variables.

91

Table 3-2: Adjusted Odds Ratios (OR) Examining the Associations of Activity
Levels with Metabolic Syndrome
Adjusted OR for Metabolic
Syndrome

(3+ or more)
a

Accelerometer
b

Model A
MVPA 0.49 (95% CI: 0.25-0.95)
Model B
Sedentary Behavior 2.56 (95% CI: 0.06-119.63)
Model C
MVPA 0.49 (95% CI: 0.25-0.98)
Sedentary Behavior 1.01 (95% CI: 0.02-61.70)
3DPAR
b

Model D
MVPA 0.86 (95% CI: 0.58-1.26)
Model E
Sedentary Behavior 4.44 (95% CI: 1.41-14.02)
Model F
MVPA 0.99 (95% CI: 0.64-1.56)

92

Table 3-2 (Continued)
Adjusted OR for Metabolic
Syndrome

(3+ or more)
a

Sedentary Behavior 4.44 (95% CI: 1.33-14.79)
MVPA by Accelerometry and Sedentary Behavior
by 3DPAR
b

Model G
MVPA by Accelerometry 0.34 (95% CI: 0.14-0.84)
Sedentary Behavior by 3DPAR 6.42 (95% CI: 1.65-25.02)
Abbreviations: MVPA, moderate-to-vigorous physical activity.
a. Parameters were adjusted for age, gender, ethnicity, Tanner stage, fat mass, and lean
tissue mass.
b. Variables were not normally distributed so statistical tests were run with
log-transformed data.

93

Table 3-3: Pearson Correlations/Partial Correlations between Activity Levels and
Features of Metabolic Syndrome
* P <.05, ** P <.01, ‡ P <.001
a. Variables were not normally distributed so statistical tests were run with
log-transformed data.
b. Parameters were adjusted for age, gender, ethnicity, Tanner stage, fat mass, and lean
tissue mass.
Pearson Correlations
Accelerometry
a
3DPAR
a

MetS Feature MVPA
Sedentary
Behavior
MVPA
Sedentary
Behavior
Triglycerides
a
-0.21*
HDL-cholesterol
a
-0.22 *
Waist circumference -0.42 ‡ 0.32 **
Fasting glucose
a

Systolic Blood Pressure
a
-0.33 ** 0.21 * 0.31**
Diastolic Blood Pressure
a
-0.27**

Partial Correlations
b

Accelerometry
a
3DPAR
a

MetS Feature MVPA
Sedentary
Behavior
MVPA
Sedentary
Behavior
Triglycerides
a

HDL-cholesterol
a
-0.21*
Waist circumference
Fasting glucose
a
-0.21*
Systolic blood pressure
a
-0.25* 0.26 **
Diastolic blood pressure
a

94

Table 3-4: Summary of Significant Findings on Physical Activity, Sedentary
Behavior, and Metabolic Syndrome by Measurements
Accelerometer 3DPAR
Time in MVPA↑ Triglycerides↓
Waist circumference↓
Fasting glucose↓
Systolic blood pressure↓
Diastolic blood pressure↓
Odds of MetS↓

Time in Sedentary Behavior↑ Waist circumference↑
Systolic blood pressure↑

HDL-cholesterol↓
Systolic blood pressure↑
Odds of MetS↑
Abbreviations: MVPA, moderate-to-vigorous physical activity; MetS, metabolic
syndrome

95

Chapter 4 LONGITUDINAL DIFFERENCES IN PHYSICAL ACTIVITY AND
SEDENTARY BEHAVIOR BY METABOLIC SYNDROME AMONG YOUNG
LATINA AND AFRICAN AMERICAN FEMALES
4.1 INTRODUCTION
The metabolic syndrome (MetS) refers to a cluster of cardiovascular and Type 2
diabetes risk factors.(199) Longitudinal studies in pediatric populations have
demonstrated that engagement in physical activity and reductions in sedentary behavior
lower the risk of developing MetS. Raitakari et al(170) showed that in youth persistent
physical activity (equivalent to at least 2 hours of intense aerobic activity per week)
lasting 6 years was related to lower metabolic risk, while persistent inactivity (equivalent
to 1 hour of light aerobic activity per week) was related to higher metabolic risk. Yang et
al(232) found that sustained sport participation in youth reduces risk for developing MetS
in adulthood after a 21-year follow-up. To date, while most studies have focused on
investigating the preventive effects of physical activity on metabolic risk, no research has
examined the longitudinal impact of MetS on activity levels.
Our previous cross-sectional data(91) indicate that youth with MetS have lower
levels of physical activity and higher levels of sedentary behavior than those without
MetS; however, without examining temporal associations, it is possible that associations
are reciprocal. Overweight status, a crucial risk factor for developing MetS,(193) has
been found to be associated with adverse physical(146) and psychological(15) and health
outcomes. Is it thus possible that poor metabolic health and overweight status might

96

contribute to an inactive lifestyle. With the rising rates of MetS in pediatric
populations,(193) this exploratory investigation is important because the presence of
MetS along with unfavorable activity patterns places youth at ‘double jeopardy’ for risk
for type 2 diabetes and cardiovascular chronic diseases.
Evidence illustrating that overweight may lead to physical inactivity supports the
possibility that activity levels are impacted by metabolic health. One pediatric study(135)
examined whether inactivity was the cause of fatness or fatness was the cause of
inactivity over a 3-year period, and concluded that adiposity at baseline led to lower
levels of physical inactivity over time, suggesting that low levels of physical activity
were the result of fatness rather than the cause. Though the underlying mechanism for
how adiposity influences activity levels has not yet been discovered, there is evidence of
links between overweight status and unfavorable correlates of physical activity. Previous
research has indicated that overweight children had greater body-related barriers to
physical activity and received less social support for physical activity than normal weight
children.(236) Other work reported that overweight youth had poor physical competence
(140) and lower motor abilities.(107) Accordingly, because youth with MetS are usually
overweight, it is likely that their greater barriers to physical activity, or their poorer
physical ability could limit their engagement in physical activity and increase their time
being sedentary.

97

Puberty is a critical period of development marked by dynamic biological changes.
Increasing body fat(102) and insulin resistance(138) during puberty may increase the risk
of developing MetS. Pubertal insulin resistance tends to be more severe in females.(138)
Compared to other ethnic groups, Latina and African American female youth have the
most pronounced pubertal decline in physical activity, (79, 105, 162) less moderate to
vigorous physical activity, (80) and higher levels of sedentary behavior.(80) As the
presence of MetS might negatively impact activity levels, it is also possible that the
presence of MetS may accelerate the pubertal decline in physical activity and increase in
sedentary behavior.
Therefore, the purpose of this study was to explore the differences in longitudinal
patterns of activity levels during puberty by MetS status. We hypothesized that 1) youths
with MetS will have lower initial values of physical activity and higher baseline levels of
sedentary behavior than those without MetS, and 2) youths with MetS will have greater
declines in physical activity and greater increases in sedentary behavior over the course
of 12 months than those without.
4.2 METHODS
Participants
Data for this sample was taken from the TRANSITIONS study (Insulin Resistance
and Declining Physical Activity Levels in African American and Latina girls), a
longitudinal, observational study of psychosocial and physiological determinants of the
pubertal decline in physical activity in minority girls. Of 79 participants who completed

98

their baseline inpatient visit, 14 had missing activity data by accelerometry, 8 were
missing data for MetS, 2 were missing body composition data. The final analytical sample
therefore consisted of 55 female participants (Latino 76%, 9.4±0.9 years at baseline) who
had complete baseline data for all measures of interests.
Participants were recruited through medical clinics, churches, community centers,
local schools, and advertisements in Los Angeles County. Detailed study methods have
been published previously.(49, 60, 182, 191) Participants were excluded if they were
using a medication or had been diagnosed with a condition known to influence body
composition, fat distribution, insulin/glucose metabolism, or if they were diagnosed
with diabetes by fasting plasma glucose (≥ 126 mg/dl) at the time of entry to the study.
Written parental informed consent and child assent were obtained before any data were
collected. The study was approved by the University of Southern California’s (USC)
Institutional Review Board.
Protocol
Participants in the current study had up to five activity assessments points over the
course of 12 months. Metabolic and demographic measures were collected during the
baseline inpatient visit. Both subjective and objective measures of activity levels were
obtained at the inpatient visit and the four subsequent quarterly assessments.
During the baseline inpatient visit, youth arrived in the afternoon at USC General
Clinical Research Center (GCRC). Participants completed questionnaires, including the

99

3DPAR on self-reported activity levels. Assessments of anthropometry, body
composition, and pubertal Tanner stage were also obtained. After the examination,
participants were served a standardized dinner and a snack before 8 pm, which marked
the beginning of an overnight fast. Only water was permitted during this period. At
approximately 8 am the following day, the Frequently Sampled Intravenous Glucose
Tolerance Test (FSIVGTT) was performed. Fasting blood samples from FSIVGTT
[before glucose (25% dextrose, 0.3 g/kg body weight) was administered intravenously at
time 0] were evaluated for triglycerides, HDL-cholesterol, and glucose. Participants were
instructed on how to wear an accelerometer for the week following the inpatient visit.
Quarterly home visits were conducted by research staff approximately every three
months subsequent to the inpatient visit. Participants were given an accelerometer to
wear and the 3DPAR questionnaire to complete.
Measures
Body Composition and Pubertal Tanner Stage
Weight and height were measured three times to the nearest 0.1 kg and 0.1 cm,
respectively, using a beam medical scale and wall-mounted stadiometer by trained
medical professionals. Body mass index (BMI) percentiles for age and gender were
determined based upon established Centers for Disease Control and Prevention normative
curves.(222) Body composition (total fat mass and total lean tissue mass) was measured

100

by air plethysmography (BodPod
TM
) and pubertal Tanner stage was assessed by a
licensed pediatric health care provider.(196)
Metabolic Measures
Sitting blood pressure was measured in triplicate using the right arm after the participant
had rested for 5 minutes.(150) Waist circumference was measured in triplicate at the
umbilicus and recorded to the nearest 0.1 cm. Blood samples taken during the FSIVGTT
were centrifuged immediately for 10 min at 2500 rpm and 8–10 °C to obtain plasma, and
aliquots were frozen at −70 °C until assayed. Glucose was assayed using a Yellow
Springs Instruments 2700 analyzer (Yellow Springs Instrument, Yellow Springs, OH)
that uses a membrane bound glucose oxidase technique. Fasting lipids including
triglycerides and HDL-cholestero were assessed using Vitros Chemistry DT Slides
(Johnson and Johnson Clinical Diagnostics, Inc., Rochester, New York).
Definition of the Metabolic Syndrome (MetS)
Because no standard definition for MetS exists for pediatric populations, we used a
combination of pediatric definitions proposed by Cruz et al(41) and Cook et al,(36, 184)
a

who applied pediatric cutoffs to the Adult Treatment Panel III definition.(64) This
definition has been applied in our previous studies.(41, 184) To be classified as having

a
Cook et al originally used fasting glucose ≥110 mg/dl to define Impaired Fasting Glucose. The criteria for
Impaired Fasting Glucose has been updated by the American Diabetes Association as a fasting glucose ≥100
mg/d.(16) The revised cut-point has been used.

101

MetS, participants had to have at least three of the following features: abdominal obesity
(waist circumference ≥ 90th percentile for age, sex, and ethnicity from Third National
Health and Nutrition Examination Survey data),(65) hypertriglyceridemia (triglycerides ≥
90th percentile for age, sex, and ethnicity),(87) low HDL-cholesterol (HDL-cholesterol ≤
10th percentile for age, sex, and ethnicity),(87) elevated blood pressure (systolic or
diastolic blood pressure ≥ 90th percentile adjusted for age, sex, and height),(151) or
hyperglycemia (impaired fasting glucose ≥ 100 mg/dl).(5, 36)
a
Physical Activity and Sedentary Behavior
Accelerometry data
Objective assessments of activity levels were obtained using the ActiGraph
(Actigraph, LLC; FT. Walton Beach, FL) model GT1M accelerometer. The ActiGraph is
a uniaxial accelerometer that measures acceleration in the vertical plane and it has been
shown to be a valid and reliable measure of physical activity for children and adolescents
(169). The participants were instructed to wear the device on the right hip during waking
hours, with the exception of time spent bathing or swimming. Accelerometer data were
processed using an updated version of the SAS program developed for use with the
NHANES physical activity monitoring data (available at:
http://riskfactor.cancer.gov/tools/nhanes_pam). The raw data was processed to calculate
the minutes of non-wear (defined by an interval of sixty or more consecutive minutes of

102

zero activity intensity counts, with exceptions for up to 3 minutes of 0-100 counts) and
the minutes of wear time.
A valid day of wear was defined as having at least ten hours of wear-time, and only
participants with four or more valid days of data were included in the analyses (204). The
total number of minutes spent in moderate-to-vigorous physical activity (MVPA) and
sedentary activity, was determined by summing minutes in a day where the count met the
criterion for that intensity. Then, these total minutes were averaged across the number of
valid days to obtain the mean minutes per day. The cut point for MVPA [≥4 metabolic
equivalent (METs)] was age-adjusted using the criteria that the Freedson group proposed
for youth between the ages of 6 and 17 years (206). The sedentary cut point of 100 counts,
which was previously defined by Matthews et al. (129), has been validated in
adolescents.
Self-reported physical activity data
A 3DPAR was used to assess self-reported activity levels.(164, 219) 3DPAR has
been validated in adolescents.(164) Students identified different activities (from a list of
71 activities provided) to describe their activity in half-hour intervals during a day from
7:00 am – 12:00 am for three days, and rated how much effort (intensity level) they put
into each activity (light, moderate, hard, or very hard). Activity types were converted into
half-hour blocks of either light, moderate, or vigorous physical activity using a
combination of the intensity ratings provided by the participants and the compendium of
physical activities.(1) The criteria for MVPA (≥4 METs) was used to be consistent to the

103

accelerometry variable. Half-hour blocks spent watching television/movies, playing video
games/surfing the internet, and talking on phone were coded separately as sedentary
behaviors. Daily total time spent in MVPA and sedentary behavior was obtained by
summing over the half-hour blocks over one day. Mean minutes per day was then
obtained by averaging total minutes across three days.
Statistical Analysis
Attrition is a common problem in longitudinal analysis. To assess the impact of
attrition on our data, two sets of attrition analysis were performed. First, of participants
who completed their first annual visit, those with complete baseline data for all
measurements included in the present study (final analytical sample) were compared to
those with incomplete baseline data on demographic variables using independent samples
t-tests/Wilcoxon rank sum tests for continuous variables or chi-square tests/ Fisher’s
exact tests for categorical variables. In the final analytical sample, these same sets of
analyses were applied to compare participants who remained in the present study (from
visit 1-5) and participants who dropped out prior to the 5
th
visit. Participants were
dichotomized into those with and without MetS for descriptive analyses. Independent
samples t-tests/ Wilcoxon rank sum tests and chi-square tests/ Fisher’s exact tests were
conducted to compare demographic variables between these two groups.
Growth curve modeling was used to evaluate whether longitudinal trends (initial
level and slope) reflected by the repeated measurements of physical activity and
sedentary behavior differed by MetS status (yes/no) over the course of 12 months. An

104

important element of growth curve methodology is its ability to account for individual
differences in trajectories over time.(52) Specifically, it allows us to investigate
differences in physical activity and sedentary behavior over time by MetS status (yes/no),
with adjustments for the individual factors that may influence these changes. Four growth
curve models were developed to investigate trends of physical activity and sedentary
behavior repeatedly measured by both accelerometry and 3DPAR. For all models,
quarterly visit (visit1-5) was used as the measurement of time. The following variables
were included as common covariates in each model: age, ethnicity, pubertal Tanner stage,
fat mass, and lean tissue mass.
All continuous independent variables were centered on the sample’s mean to prevent
collinearity, for ease of interpretation and to ensure meaningful explanations of potential
interaction effects. Logarithmic transformations were applied to achieve normality for
MVPA by accelerometry, MVPA by 3DPAR, and sedentary behaviors by 3DPAR, which
were skewed. Analyses were performed using SAS v.9.2 (SAS Institute, Cary, NC). The
significance of the findings were evaluated at the p < 0.05 level.
4.3 RESULTS
No statistical differences in baseline pubertal Tanner stage, baseline age, baseline
BMI percentile, and baseline activity levels by both accelerometry and 3DPAR were
found between participants with complete baseline data for all measurements and those
with incomplete baseline data (Table 4-1). More African Americans participants had
incomplete baseline data than Latina (p=0.049). In the final analytical sample (N=55),

105

there were no statistical differences in ethnicity, baseline pubertal Tanner stage, baseline
age, baseline BMI percentile, and baseline activity levels by both accelerometry and
3DPAR between participants who completed 5 visits and those who dropped out (Table
4-2).
Baseline characteristics
Descriptive baseline characteristics by participants with and without MetS are
shown in Table 4-3. In this study, 13% (n=7) of participants met the criteria for MetS. At
baseline, those with MetS had higher BMI percentiles (p=0.001), greater fat mass
(p=0.003), higher percent body fat (p=0.002), and lower percent lean tissue mass
(p=0.002). No significant difference in baseline activity levels were found between those
with and without MetS.
Longitudinal trends of activity levels by MetS status
Longitudinal trends of activity levels as assessed by accelerometry and 3DPAR
are illustrated in Figure 4-1 and 4-2, respectively. Adjusted results for the effects of
baseline MetS status on initial levels and longitudinal changes of MVPA and sedentary
behavior by accelerometry are presented in Table 4-4. For the initial levels of activity
levels, Latina youth engaged 20.07 minutes fewer in MVPA (p=0.006) than African
Americans; youth in Tanner stage 2 spent 44.14 more minutes in sedentary behavior than
those in Tanner stage 1 (p=0.022). In addition, age (p=0.034) and total fat mass (p=0.006)
were inversely related to lower levels of MVPA at the baseline. No significant

106

differences in both initial MVPA and initial sedentary behavior by MetS status were
found. There were statistically significant linear associations between longitudinal
changes in activity levels and visit. In the study sample, MVPA declined on an average of
2.17 minutes (-4.67%) per quarterly assessment (p=0.004), adding up to 8.68 minutes
(-18.67%) per year decline and sedentary behavior increased by 5.21 minutes (1.24%) per
quarterly assessment (p=0.002), adding up to 20.84 (4.97%) minutes per year increase,
after adjusting for age, ethnicity, pubertal Tanner stage, fat mass, and lean tissue mass.
When comparing the MetS-related rates of change in activity levels, sedentary behavior
increased 23.42 minutes/per quarterly visit, adding up to 93.68 minutes/per year more in
youth with MetS than in those without (p=0.014).
Table 4-5 shows the effects of baseline MetS status on initial levels and
subsequent changes in MVPA and sedentary behavior by 3DPAR. MVPA declined by
8.64 minutes in MVPA (-7.34%) per visit adding up to 34.56 minutes (-29.36%) over a
year in our sample (p=0.014). At the baseline, Latina youth spent an average of 54.89
minutes more in sedentary behavior than African Americans (p=0.008). Baseline MetS
status had no effects on either the initial levels or the subsequent rate of change in MVPA
and sedentary behavior over time.
4.4 DISCUSSION
This study is one of the few longitudinal studies to examine the developmental
trajectory of activity levels in minority adolescent girls. Consistent with other
studies,(165, 200) we observed both a significant decline in MVPA and an increase in

107

sedentary behavior over one year. This study is the first to evaluate the influence of MetS
on initial activity levels and longitudinal changes in physical activity and sedentary
behavior during puberty. Our primary finding is that baseline MetS status predicts a faster
rate of increase in sedentary behavior (as measured by accelerometry), independent of the
initial pubertal stage, initial body composition, and other covariates.
To our knowledge, the underlying mechanism for how MetS influences activity
levels has not yet been identified as there has been a lack of relevant research.
Considering that youth with MetS are often overweight, evidence from studies on
adiposity and activity levels might shed some light on how MetS affects physical activity
and sedentary behavior in youth. It has been found that overweight children might have
lower levels of physical activity as a consequence of their weight status.(59) One
potential explanation could be that widely-documented fact that overweight youth report
greater numbers of negative psychosocial experiences involving more weight-related
distress(86, 233) and peer victimization (e.g. bullying behaviors and peer isolation).(59,
95) These negative experiences may result in a lack of confidence in their ability to
overcome barriers to physical activity,(205) which may make sedentary activities more
reinforcing than active pursuits.(62) While all these unfavorable factors have been linked
with a greater preference for sedentary and decreased inclination towards physical
activities among youth,(86) appearance-related psychosocial distress was more prevalent
among overweight adolescent girls than overweight boys.(236) Particularly during
childhood, when physical activity often involves organized group activities, it is likely

108

that these negative social and emotional consequences may discourage overweight girls
from being active and also may increase time spent in sedentary behaviors.(59, 147)
Physical factors may also contribute to the relationships between MetS and activity levels.
Studies have reported that overweight children have poorer physical abilities,(139) poorer
motor skills performance,(42) and lower physical fitness levels(2) compared to their
normal weight peers.
Following the same logic, youth with MetS (whom are usually overweight) may
have lower self-efficacy for physical activity, greater activity-related social/emotional
distress, or poorer physical competence and muscular strength, which may increase their
participation in sedentary activities and/or limit their engagement in physical activity.
Evidence supporting this claim were found in two cross-sectional studies. It has been
showed that youth with MetS had higher levels of stress-related serum cortisol,(217)
suggesting a potential link between greater chronic life stress and the presence of MetS.
Additionally, waist circumference and blood pressure, two individual MetS components,
have been found to be inversely associated with physical fitness in children.(17) Another
potential explanation of how MetS influences activity levels is related to leptin, an
adipocyte derived hormone that plays a central role in the regulation of energy.(174)
Higher leptin levels have been shown to be positively related to hours of TV viewing(71)
as well as negatively related to physical activity.(7, 11) In line with other pediatric(141)
and adult(72) studies, we found a higher level of leptin in youth with MetS than those
without (MetS: 27.42±7.32 ng/ml; Non-MetS: 12.76±8.84 ng/ml, p<0.001) in our

109

participants. Thus, the relatively unhealthy activity patterns seen in participants with
MetS may be related to higher levels of leptin.
This adverse impact of MetS on sedentary behavior has two important implications.
First, it highlights the utility of considering screening for MetS regularly in pediatric
clinics. Second, it underscores the need to conduct interventions specifically among
youth with MetS. Given that youth with MetS had a steeper increase in sedentary
behavior, detection of MetS at an early age could help to identify individuals at increased
metabolic risk when clinical manifestations of chronic disease risk are not as obvious, as
well as help to target interventions to initiate lifestyle changes to reduce future health
risks. Future efforts should not only aim to prevent the development of MetS, but also
aim to promote physical activity and decrease sedentary behavior among those with
MetS.
It is important to note that, while there was a longitudinal impact of MetS on the rate
of change in sedentary behavior, the longitudinal impact of MetS on MVPA did not reach
significance. Participants in the current study had very low levels of daily activity (47
minutes of MVPA as measured by accelerometry), as compared to a national youth
sample at similar age (88 minutes of MVPA as measured by accelerometry),(10) thus
there was limited minutes of physical activity to “lose” over a year. In addition, previous
research has indicated that school-related physical activity accounts for almost 50% of
children’s MVPA.(73) It is thus possible that in our school-aged sample, the MetS-related
differences in total engagement of MVPA may be attenuated by children’s similar

110

opportunities for school-related physical activity such as their physical activity for travel
(e.g., walking to and from school) and required participation in physical education classes.
On the other hand, we did fund significant associations between MetS and sedentary
behavior. This may be due to the fact that there is generally more opportunity to
participate in attractive sedentary behaviors than in physical activity.(62) Perhaps, a more
pronounced decline in MVPA occurs among those with MetS as they enter early
adulthood and are no longer mandated to participate in school-based physical activities.
Thus, perhaps MetS-related increases in sedentary behavior manifest in early puberty,
while effects on MVPA could emerge in late adolescence or early adulthood.
When comparing the results by measurement method in this study (objectively and
subjectively measured activity levels), the significant findings regarding MetS and
activity levels were observed based on the accelerometry data only. It poses a challenge
to explain why there were no significant findings on MetS and activity levels based on
the 3DPAR data. One possible explanation lies in the bias introduced by the self-reported
nature of the 3DPAR. As the estimates obtained by 3DPAR are subjective, they are
influenced by respondent’s perception of fitness levels, social desirability, and memory
introducing larger between-subject variations in activity levels. These larger variations
could further have been exaggerated by the small size of our MetS sample, resulting in
unstable activity estimates. This explanation is supported by the activity patterns
demonstrated in Figure 4-2, where a more consistent trend of activity levels was
observed in the non-MetS group (N=48) but not in the MetS group (N=7). Collectively, it

111

is likely that the larger variations in activity levels introduced by the self-reported nature
of the 3DPAR might dilute the differences in activity levels between MetS group. Future
studies with a larger cohort of both MetS and non-MetS individuals are needed to further
investigate this explanation.
Strengths and Limitations
The strengths of this study include the use of both objective and subjective physical
activity measurements, the longitudinal design, and the use of growth curve modeling
that allowed us to model changes in activity patterns at the individual level. Some
limitations of this study warrant consideration. First, only the baseline MetS was
measured. As we know that MetS status is subject to change across puberty,(211) the
potential dynamic changes in MetS status throughout the year was thus not taken into
account when assessing the association between MetS and activity levels. Second, the
relatively small overall sample size and the uneven sample sizes of MetS, pubertal
Tanner stage, and ethnic groups may preclude more detailed exploration of findings and
impede the generalizability of our results. Third, measurement bias of activity levels
could exist considering the self-reported nature of 3DPAR and the fact that the accuracy
of the accelerometry is dependent on activity type (e.g., cannot adequately capture
swimming, biking, and other activities). Fourth, although no significant differences on
demographic characteristics were found between completers and dropouts, bias due to the
high attrition rate may have occurred in the current study.

112

4.5 CONCLUSIONS
Despite the above limitations, this is the first longitudinal study to demonstrate that
MetS leads to adverse activity patterns in minority girls. Our findings suggest a vicious
cycle of inactivity and metabolic complications. Overweight status and other metabolic
health issues such as MetS might lead to low activity levels and increased sedentary
behaviors, triggering a spiral of health risk and health risk behavior. This offers a more
comprehensive explanation for the associations between metabolic health and activity
levels, accounting for pubertal development and the accompanying behavioral and
metabolic changes that occur at this time. Given the precipitous annual increase in
sedentary behavior shown in peripubertal minority girls with MetS, it could be important
to develop targeted interventions to reduce sedentary behavior and increase physical
activity in these children. The combination of the psychosocial, physical, and biological
mechanisms involved in obesity and MetS may contribute to an increase of sedentary
behavior in youth. Future research that incorporates longitudinal designs in large and
diverse populations with multiple follow-ups for metabolic health, activity levels, and
psychosocial correlates are needed to uncover mechanisms by which MetS might
influence sedentary behavior and physical activity over time.

113

Table 4-1: Comparisons of Baseline Demographic Characteristics between
Participants With and Without Complete Baseline Data
With complete
baseline data (N=55)
N (± SD/%)
Without complete
baseline data (N=24)
N (± SD/%)
P-value
a

Ethnicity
Latino 42 (76.36%) 13 (23.64%) 0.049 *
African American 13 (54.17%) 11 (45.83%)
Age (Years) 9.42 (± 0.92) 9.00 (± 1.02) 0.076
Pubertal Tanner
stage

1 27 (69.23%) 12 (30.77%) 0.941
2 28 (70.00%) 12 (30.00%)
BMI percentile 79.55 (± 23.84) 80.57 (± 23.63) 0.788
Accelerometer
(Minutes)

MVPA 46.48 (±26.89) 87.79 (±151.18) 0.993
Sedentary behavior 419.13 (±69.61) 391.39 (±76.92) 0.258
3DPAR (Minutes)
MVPA 117.73 (±73.94) 93.70 (±84.39) 0.087
Sedentary behavior 137.00 (±108.07) 163.48 (±150.95) 0.767
x
2
tests/Fisher’s exacts tests were used for categorical variables and independent t tests/
Wilcoxon rank sum tests for continuous variables.
* P <.05

114

Table 4-2: Comparisons of Baseline Demographic Characteristics by Attrition
Status in Current Analytical Sample
Completers (N=23)
N (± SD/%)
Dropouts (N=32)
N (± SD/%)
P-value
a

Ethnicity
Latino 16 (38.10%) 26 (61.90%) 0.314
African American 7 (53.85%) 6 (46.15%)
Age (Years) 9.48 (± 0.79) 9.38 (± 1.01) 0.684
Pubertal Tanner stage
1 8 (29.63%) 19 (70.37%) 0.072
2 15 (53.57%) 13 (46.42%)
BMI percentile 82.17 (± 24.37) 77.67 (± 23.66) 0.239
Accelerometer
(Minutes)

MVPA 42.58 (±28.99) 49.28 (±25.38) 0.103
Sedentary behavior 424.50 (±60.86) 415.27 (±76.00) 0.632
3DPAR (Minutes)
MVPA 117.39 (±84.53) 117.97 (±66.72) 0.614
Sedentary behavior 143.91 (±104.32) 132.03 (±112.08) 0.657
x
2
tests/Fisher’s exacts tests were used for categorical variables and independent t tests/
Wilcoxon rank sum tests for continuous variables.

115

Table 4-3: Baseline Demographic Characteristics

MetS (N=7)
N (± SD/%)
Non-MetS (N=48)
N (± SD/%)
P-value
a

Ethnicity
Latino 5 (11.90%) 37 (88.10%)
0. 664
African American 2 (15.38%) 11 (84.62%)
Age (Years) 9.29 (±1.11) 9.44 (±0.90) 0.686
Pubertal Tanner stage
0.101 1 1 (3.73%) 26 (96.30%)
2 6 (21.43%) 22 (78.58%)
BMI percentile 98.08 (±1.83) 76.85 (±24.37) 0.001 **
Total fat mass (Kg) 22.46 (±8.56) 10.93 (±7.07) 0.003 **
Percent body fat 38.26 (±5.74) 24.22 (±9.98) 0.002 **
Total lean tissue mass
(Kg)
34.69 (±6.29) 30.48 (±6.76) 0.127
Percent lean tissue mass 61.74 (±5.74) 75.78 (±9.98) 0.002 **
Accelerometer (Minutes)
MVPA 33.00 (±11.60) 48.44 (±27.98) 0.120
Sedentary behavior 408.40 (±57.02) 420.70 (±71.64) 0.667
3DPAR (Minutes)
MVPA 124.30 (±62.14) 116.80 (±76.03) 0.781
Sedentary behavior 171.40 (±143.50) 132.00 (±102.90) 0.587
a. x
2
tests/Fisher’s exacts tests were used for categorical variables and independent t tests/
Wilcoxon rank sum tests for continuous variables.
* P <.05, ** P <.01, ‡ P <.001

116

Table 4-4: Results of the Growth Curve Models to Assess the Influence of Metabolic
Syndrome on Activity Levels Measured by Accelerometer over the 1-Year Period
MVPA
a
Sedentary behavior
β (SE) P-value
b
β (SE) P-value
b

Intercept 4.08 (0.14) <0.001 ‡ 377.12
(20.53)
<0.001 ‡
Visit -0.05 (0.03) 0.004 * 5.21 (3.98) 0.002 **
Initial lean mass (Kg) 0.01 (0.01) 0.638 0.71 (1.91) 0.712
Initial fat mass (Kg) -0.03 (0.01) 0.006 ** 0.78 (1.49) 0.604
Ethnicity (Latino vs. AA) -0.38 (0.13) 0.006 ** 25.31 (18.21) 0.170
Initial age (years) -0.14 (0.07) 0.034 * 4.93 (8.98) 0.584
Tanner (2 vs. 1) -0.01 (0.14) 0.952 44.14 (18.81) 0.022 *
Baseline MetS (1 vs. 0) 0.13 (0.22) 0.546 -56.72 (33.98) 0.102
Visit * Baseline MetS (1 vs. 0) -0.09 (0.07) 0.168 23.42 (9.06) 0.014 *
Abbreviations: MetS, metabolic syndrome; MVPA, moderate-to-vigorous physical
activity; AA, African American.
a. Log-transformed data was used.
b. Parameters were adjusted for all variables included in the model.
* P <.05, ** P <.01, ‡ P <.001

117

Table 4-5: Results of the Growth Curve Models to Assess the Influence of Metabolic
Syndrome on Activity Levels Measured by 3-Day Physical Activity Recall (3DPAR)
over the 1-Year Period
MVPA
a
Sedentary behavior
a

β (SE) P-value
b
β (SE) P-value
b

Intercept 4.08 (0.41) <0.001 ‡ 3.60 (0.39) <0.001 ‡
Visit -0.18 (0.10) 0.014 * 0.01 (0.09) 0.464
Initial lean mass (Kg) 0.04 (0.04) 0.252 -0.02 (0.04) 0.544
Initial fat mass (Kg) -0.03 (0.03) 0.394 -0.03 (0.03) 0.332
Ethnicity (Latino vs. AA) 0.45 (0.36) 0.224 0.93 (0.33) 0.008 **
Initial age (years) -0.04 (0.18) 0.824 0.06 (0.17) 0.700
Tanner (2 vs. 1) 0.18 (0.38) 0.642 -0.15 (0.35) 0.664
Baseline MetS (1 vs. 0) 0.85 (0.65) 0.202 0.99 (0.67) 0.146
Visit * Baseline MetS (1 vs. 0) -0.22 (0.22) 0.320 0.13 (0.20) 0.512
Abbreviations: MetS, metabolic syndrome; MVPA, moderate-to-vigorous physical
activity; AA, African American.
a. Log-transformed data was used.
b. Parameters were adjusted for all variables included in the model.
* P <.05, ** P <.01, ‡ P <.001

118

Figure 4-1: Changes in Activity Levels (as measured by Accelerometry) According
to Visit Number and Metabolic Syndrome

119

Figure 4-2: Changes in Activity Levels (as measured by 3DPAR) According to Visit
Number and Metabolic Syndrome

120

Chapter 5 SUMMARY AND CONCLUSIONS
5.1 Summary of Findings
The overall goal of this dissertation was to investigate the associations between
physical activity, sedentary behavior, overweight, and MetS in at-risk pediatric
populations in the United States and in China. The first objective was to examine the
independent influences of physical activity, sedentary behavior, and other weight-related
correlates on overweight status in Chinese youth. The second objective was to investigate
the influences of physical activity and sedentary behavior on MetS in US minority youth.
The third objective was to compare the longitudinal trends of physical activity and
sedentary behavior between youth with and without MetS in a sample of US minority
female youth.
The results from Study 1 showed that relationships between overweight and some
factors (e.g., sleep, sedentary behavior) were the same as found in the majority of studies
conducted in youth in Western societies. However, the majority of relationships were
contradictory to findings in Western youth (e.g., overweight Chinese youth were younger,
participated more in VPA, had higher parental SES, had better self-perceived health
status a greater intake of vegetables, and a lower intake of sweets and fast food). These
observed ‘East-West paradoxes’ may be due to rapid shifts in the economic climate and
the cultural phenomena specific to China.

121

In Study 2, cross-sectional influences of physical activity and sedentary behavior on
MetS were examined in Latino and African American youth. Our findings demonstrated
that youth with MetS had lower levels of physical activity and higher levels of sedentary
behavior than those without. MVPA was inversely related to odds of developing MetS
and individual MetS features, while sedentary behavior was positively related to the odds
of developing MetS and individual MetS features, independent of body compositions. In
Study 2, both subjective (3DPAR) and objective (accelerometry) measures of activity
levels were utilized. Significant relationships were found between MVPA and MetS as
measured by accelerometry, while significant relationships between sedentary behavior
and MetS were found only with sedentary behavior as measured by self-report (3DPAR).
It seems that the accelerometer and 3DPAR might capture different components of
activity levels.
Study 3 explored the longitudinal impacts of baseline MetS on changes of physical
activity and sedentary behavior over one year between youth with and without MetS in a
sample of minority female peripubertal females. We found that MVPA declined on an
average of 2.17 minutes per quarterly measurement, while sedentary behavior increased
by 5.21 minutes per quarterly measurement over a year. On average, sedentary behavior
in youth with MetS increased by 23.42 minutes/per quarterly visit faster as compared to
those without. The unfavorable activity patterns predicted by MetS may be driven by a

122

combination of the negative social/emotional experiences and biological mechanisms
involved in obesity and poor metabolic health.
These three studies examined the cycle of increasing inactivity, obesity, and
worsening metabolic health in several pediatric populations. As hypothesized, physical
activity and sedentary behavior were related to overweight status (Study 1) and metabolic
health (Study 2), and MetS had an adverse impact on sedentary behavior over time
(Study 3). These findings, coupled with longitudinal evidence of the effects of activity
levels on obesity(147, 242) and MetS(19, 56, 58, 89, 112), suggest that physical activity
and sedentary behavior could function as antecedents as well as consequences of
overweight or MetS in youth. Understanding this vicious cycle can enhance our
understanding of the associations between obesity, metabolic health, and activity levels in
youth during puberty.
5.2 Strengths and Limitations
The results of this dissertation should be considered in light of several limitations.
First, the unique demographic (e.g., ethnic, socioeconomic and geographic location)
composition of the study samples restricts the generalizability of the findings. For each
study, the findings may not be applicable to other populations with different
characteristics or living in different societies. Second, the cross-sectional design of Study
1 and 2 precludes the determination of causal inferences. It is important to note that
though our findings on impacts of activity levels on overweight and MetS were

123

cross-sectional, there are other longitudinal evidences to confirm these causal
relationships(19, 56, 58, 89, 112, 147, 242). Third, the small and uneven sample size of
MetS and ethnic groups impede additional sub-group analyses in Study 2 and 3. Fourth,
bias pertaining to measurements may jeopardize the validity of study findings. These
potential threats include recall errors and social-desirability in reporting introduced by
subjective self-report of physical activity (e.g., assessments of VPA, sedentary behavior,
diet behaviors, and sleep in Study 1; activity levels obtained by 3DPAR in Study 2 and 3)
as well as concerns regarding accuracy of the objectively-measured activity data(190)
(e.g., estimates by accelerometry in Study 2 and 3).
Despite above limitations, this dissertation has several strengths. By assessing both
physical activity and sedentary behavior simultaneously, it allows us to identify the
independent influences of physical activity and sedentary behavior on adiposity and MetS.
Another strength is that these associations regarding activity levels and MetS are adjusted
for body composition using rigorous measures of body fat. In addition, the inclusion of
three distinct studies offers a unique opportunity to explore research questions using
different measurement modalities among diverse pediatric populations. The use of both
subjective and objective measures of activity levels not only enables the comparisons
between these two types of measurements, but also provides a more comprehensive
understanding of the associations between activity levels and obesity-related health risk.
The focus on the at-risk, yet understudied pediatric populations in China and the United

124

States adds to the limited literatures on activity levels, overweight, and MetS among
these populations. Finally, Study 3 is the first to propose and examine how MetS affects
activity levels over time. Such longitudinal evidence advances our understanding of the
associations regarding MetS and activity levels by offering an alternative reverse
causality.
5.3 Implications
Cultural and socioeconomic considerations for obesity prevention
Findings from Study 1 suggest that weight-related correlates might play different
roles in Chinese culture than they do in Western cultures. The observed “East-West
paradox” lends support to the importance of cultural influences in shaping individual
health behaviors that may affect obesity. Another crucial finding in Chinese youth is the
positive SES-obesity association. This relationship was consistently observed in
developing societies across age and gender groups in a review by Sobal and
Stunkard.(188) However, in another review (136) of adults from developing countries,
only half of the reviewed studies illustrated a positive SES-obesity relationship, while
some inverse SES-obesity relationships in adults were observed. Social conditions have
been posited as fundamental causes of disease because social factors (e.g., SES) embody
access to important resources (e.g., access to health care, money, power, and social
connections) and are thus influential to determine disease risks.(115) Following from this,
it is likely that the positive relationships between SES and overweight were commonly

125

seen in countries in an early stage of socioeconomic development. Nevertheless, the
obesity burden could shift to lower SES groups, as the country’s gross national product
(GNP) rises. Taken together, these findings suggest that a nation’s development stage
may moderate the relationships between SES and obesity. This idea creates an important
perspective that weight-related correlates (e.g., SES, ethnicity, nation’s economic
development stage) could interact to influence obesity. Relevant examples have been
described elsewhere.(24) For example, the interaction between SES and ethnicity on
obesity had been documented in a recent report. In this study, the inverse SES-obesity
relationship was found to vary across ethnic groups.(156) In summary, results from Study
1 suggest that caution should be used when translating findings across cultures and
socioeconomic development stages. More specifically, the SES-obesity association is
dynamic and more complicated in developing societies undergoing rapid economic
growth. Obesity prevention and treatment should be culturally appropriate and consider
the socio-economical aspects of the target population to enhance its effectiveness. Aside
from the direct influences of correlates on obesity, future research should also consider
the inter-relationships between these correlates and how they interact to affect obesity
risk.
Usefulness of regular MetS screening to reduce future disease risk in youth
The adverse impact that MetS was found to have on sedentary behavior (Study 3)
has three clinical implications. It suggests that screening for MetS on a regular basis in

126

pediatric clinics or at schools might be help to identify youth at high risk for increasingly
poor physical activity patterns. The finding that peripubertal females with MetS had a
much steeper increase in time spent in sedentary behavior over one year also emphasizes
the need for early intervention to protect against the development of unfavorable activity
patterns and prevent their health risks. In addition, physicians and clinicians should
receive training and resources to refer youth with MetS to appropriate activity programs
to increase their activity levels. There is little research examining how MetS status might
influence activity levels. Potential mechanisms involving psychosocial (e.g., body image
disturbances,(158, 194) embarrassment,(33, 142) bullying,(59, 95) or teasing(86)) and
biological (e.g., higher levels of leptin, poor motor skills) factors were proposed in Study
3. Future studies are required to identify the modifiable factors that account for the
increased inactivity among youth with MetS. These factors could then be addressed in
interventions to reduce sedentary behavior and promote physical activity among at-risk
youth with MetS.
Physical activity and sedentary behavior should both be addressed in future interventions
In Study 2, we found that the inverse relationship between MVPA and MetS was
independent of sedentary behavior, and vice versa, the positive relationship between
sedentary behavior and odds of MetS was independent of MVPA. These results suggest
that future interventions aiming to improve metabolic health and reduce obesity in youth
should target both the promotion of MVPA and the reduction of sedentary behavior. This
strategy is essential to obtain greater beneficial health effects. The time-displacement

127

hypothesis, states that sedentary behavior displaces time spent in physical activity,
however this may not always hold.(13) In other words, it cannot be assumed that
decreases in sedentary behavior will be replaced or balanced by increases in physical
activity. Several interventions have reduced sedentary behavior without increasing
physical.(81, 192) This is most likely due to the fact that sedentary pursuits such as TV
viewing are being replaced by light activities that do not reach the level of moderate
physical activity. Moreover, even one chooses to engage in physical activity rather than
sedentary behaviors, the health benefits may be limited if physical activity levels fail to
reach the duration or intensity level needed to reap health benefits.(209) To date,
sedentary behavior has received less attention than physical activity in existing obesity
prevention programs, although research has shown that sedentary behavior has a negative
impact on health independent of physical activity levels.(12, 243) Furthermore, it has
been shown that energy-dense, nutritionally poor snack food consumption is related to
TV viewing and other sedentary activities.(63). Future interventions should therefore
include a focus on sedentary behavior in addition to physical activity.
Use of a combination of both subjective and objective measures
Accurate assessments of physical activity and sedentary behavior are central to
being able to evaluate relationships with health outcomes. In Study 2, the significant
associations between activity levels and MetS differed by measurement modality. It
appears that relying on just one measure may not be sufficient to accurately estimate

128

activity levels. Accelerometers and 3DPAR both have different weaknesses.
Accelerometers have the disadvantage of being unable to accurately detect certain
activities (e.g., biking) and cannot worn in water.(84) Self-report measures such as
3DPAR are prone to recall bias and social desirability issues.(187) Given that the perfect
activity measure is currently lacking, future research should use both subjective and
objective measures in conjunction to complement the limitations inherent in each
measurement modality.
Aside from the differences in associations between activity levels and MetS,
accelerometers and 3DPAR also differ in their estimates of activity levels (Figure 3-1). A
better understanding of how activity is measured by each modality may help explain
these differences. For the time dimension of physical activity, the higher estimates
obtained by 3DPAR might be due to the fact that it assesses the dominant activity for
each half-hour block while accelerometers record actual time spent in specific intensity
levels of physical activity. In addition, relatively larger variations in physical activity
were observed for the estimates obtained by 3DPAR. This could be explained by the fact
that 3DPAR is subjective; therefore, more variations could be introduced because
subjective estimates are influenced by respondent’s perception of fitness levels and
memory. Estimates from accelerometers, on the other hand, are determined by continuous
wearable monitoring so they are not as prone to human error as 3DPAR.

129

For the time dimension of sedentary behavior, accelerometers capture all activities
under certain cut-offs and the types of these activities were unknown. These cut-offs were
determined by calibration of accelerometers against suitable criterion measures of energy
expenditure. To accurately identify sedentary behavior, it is recommended that these
cut-offs need to be age-specific for children.(163) However, there is a lack of calibration
studies among children and the commonly-used cut-off (e.g., 100 counts per minute) was
defined based on a small group of 13-14 year-old female youths (N=74). Another concern
on accelerometers’ sedentary behavior estimations is that sedentary time may not be
correctly distinguished from the non-wear period(223) or light physical activity.(163)
Previous research has showed that accelerometry estimation may substantially
misclassify sedentary time as non-wear, particularly for individuals who are older and
have higher BMI.(223) Future studies are needed to refine and explore the proper
population-specific algorithms. Self-report by the 3DPAR has the unique advantage of
providing rich contextual data on the types sedentary activities and thus allows
researchers to focus on specific activities and contexts that have been found to have
adverse effects on metabolic health. It is important to note that in Study 3, the activity
patterns were less consistent among those with MetS as compared to those without MetS,
based on 3DPAR data (Figure 4-2). As previously discussed, this lack of consistency
may be due to the fact that between-subject variations inherent in self-reported data were
exaggerated in our small sample size of the MetS group (N=7). This observation
illustrates the common problem of utilizing self-report instrument with smaller sample.

130

Therefore, we suggest a larger sample size is needed to obtain valid estimations by
3DPAR.
Taken together, a combination of subjective and objective measures is
recommended to better understand activity levels. Considering that these two modalities
assess different aspects of activity levels, caution should be undertaken when attempting
to compare results across measurement modalities.
5.4 Overall Conclusions
This dissertation expands the current literature on obesity and MetS in at-risk
pediatric populations in China and in the United States. Findings from this dissertation
suggest a vicious cycle of increasing inactivity, obesity, and metabolic complications.
Such knowledge points to new directions for preventing obesity and improving metabolic
health in minority youth, and identifies the pivotal role of sedentary behavior. Future
effort should focus on the development of interventions that target reductions in
sedentariness , rather than targeting physical activity alone. Additionally, in light of the
rising rates of overweight and MetS in pediatric populations,(193) more attention should
be paid to youth with MetS. Given the precipitous annual increase in sedentary behavior
shown in peripubertal minority girls with MetS, it could be important to develop targeted
interventions to reduce sedentary behavior and increase physical activity in these
children. Lastly, since research on how MetS affects activity levels is still in its infancy,

131

little is known about the underlying mechanisms. Further research is necessary to better
understand why people with MetS have a marked increase in sedentary behavior.

132

BIBLIOGRAPHY
1. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, O'Brien
WL, Bassett DR, Jr., Schmitz KH, Emplaincourt PO, Jacobs DR, Jr., and Leon AS.
Compendium of physical activities: an update of activity codes and MET intensities.
Medicine & Science in Sports & Exercise 32: S498-504, 2000.
2. Aires L, Silva P, Santos R, Santos P, Ribeiro JC, and Mota J. Association of
physical fitness and body mass index in youth. Minerva Pediatr 60: 397-405, 2008.
3. Allen ML, Elliott MN, Morales LS, Diamant AL, Hambarsoomian K, and Schuster
MA. Adolescent participation in preventive health behaviors, physical activity, and
nutrition: differences across immigrant generations for Asians and Latinos compared with
Whites. Am J Public Health 97: 337-343, 2007.
4. American Academy of Pediatrics. Children, adolescents, and television. Pediatrics
107: 423-426, 2001.
5. American Diabetes Association. Diagnosis and Classification of Diabetes Mellitus.
Diabetes Care S37–S42, 2005.
6. Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, and Pratt M. Relationship of
physical activity and television watching with body weight and level of fatness among
children: results from the Third National Health and Nutrition Examination Survey.
JAMA 279: 938-942, 1998.
7. Banks WA. The many lives of leptin. Peptides 25: 331-338, 2004.
8. Barge S, Wedderkropp N, Ekelund U, Franks PW, Wareham NJ, Andersen LB, and
Froberg K. Features of the metabolic syndrome are associated with objectively measured
physical activity and fitness in Danish children. Diabetes Care 27: 2141-2148, 2004.

133

9. Barr EL, Zimmet PZ, Welborn TA, Jolley D, Magliano DJ, Dunstan DW, Cameron
AJ, Dwyer T, Taylor HR, Tonkin AM, Wong TY, McNeil J, and Shaw JE. Risk of
cardiovascular and all-cause mortality in individuals with diabetes mellitus, impaired
fasting glucose, and impaired glucose tolerance: the Australian Diabetes, Obesity, and
Lifestyle Study (AusDiab). Circulation 116: 151-157, 2007.
10. Belcher BR, Berrigan D, Dodd KW, Emken BA, Chou CP, and Spuijt-Metz D.
Physical Activity in US Youth: Impact of Race/Ethnicity, Age, Gender, & Weight Status.
Med Sci Sports Exerc [Epub ahead of print].
11. Belcher BR, Chou C-P, Rodriguez SN, Hsu Y-H, Byrd-Williams C, Weigensberg
MJ, and Spuijt-Metz D. The Longitudinal Effects of Leptin on Physical Activity in
Minority Female Children. Obesity Under review.
12. Bertrais S, Beyeme-Ondoua JP, Czernichow S, Galan P, Hercberg S, and Oppert JM.
Sedentary behaviors, physical activity, and metabolic syndrome in middle-aged French
subjects. Obes Res 13: 936-944, 2005.
13. Biddle SJ, Gorely T, Marshall SJ, Murdey I, and Cameron N. Physical activity and
sedentary behaviours in youth: issues and controversies. J R Soc Promot Health 124:
29-33, 2004.
14. Biro FM, and Wien M. Childhood obesity and adult morbidities. Am J Clin Nutr 91:
1499S-1505S.
15. Bjorntorp P. Obesity. Lancet 350: 423-426, 1997.
16. Brener ND, Collins JL, Kann L, Warren CW, and Williams BI. Reliability of the
Youth Risk Behavior Survey Questionnaire. Am J Epidemiol 141: 575-580, 1995.
17. Brunet M, Chaput JP, and Tremblay A. The association between low physical
fitness and high body mass index or waist circumference is increasing with age in
children: the 'Quebec en Forme' Project. Int J Obes (Lond) 31: 637-643, 2007.

134

18. Butte NF, Puyau MR, Adolph AL, Vohra FA, and Zakeri I. Physical activity in
nonoverweight and overweight Hispanic children and adolescents. Med Sci Sports Exerc
39: 1257-1266, 2007.
19. Byberg L, Zethelius B, McKeigue PM, and Lithell HO. Changes in physical activity
are associated with changes in metabolic cardiovascular risk factors. Diabetologia 44:
2134-2139, 2001.
20. C.-Y. Ji and Cooperative Study on Childhood Obesity: Working Group on Obesity
in China (WGOC). The prevalence of childhood overweight/obesity and the epidemic
changes in 1985–2000 for Chinese school-age children and adolescents Obesity Reviews
9: 78-81, 2008.
21. Cali AM, and Caprio S. Obesity in children and adolescents. J Clin Endocrinol
Metab 93: S31-36, 2008.
22. Calle EE, and Kaaks R. Overweight, obesity and cancer: epidemiological evidence
and proposed mechanisms. Nat Rev Cancer 4: 579-591, 2004.
23. Caprio S. Insulin resistance in childhood obesity. J Pediatr Endocrinol Metab 15
Suppl 1: 487-492, 2002.
24. Caprio S, Daniels SR, Drewnowski A, Kaufman FR, Palinkas LA, Rosenbloom AL,
and Schwimmer JB. Influence of race, ethnicity, and culture on childhood obesity:
implications for prevention and treatment: a consensus statement of Shaping America's
Health and the Obesity Society. Diabetes Care 31: 2211-2221, 2008.
25. Centers for Disease Control and Prevention. Obesity - At A Glance 2009
http://www.cdc.gov/chronicdisease/resources/publications/AAG/obesity.htm. [08/16,
2010].
26. Chen C. Overview of obesity in Mainland China. Obes Rev 9: 14-21, 2008.

135

27. Chen X, Beydoun MA, and Wang Y. Is Sleep Duration Associated With Childhood
Obesity? A Systematic Review and Meta-analysis. Obesity (Silver Spring) 16: 265-274,
2008.
28. Cheng TO. Diabetes and obesity epidemics in China: a national crisis. Int J Cardiol
123: 1-2, 2007.
29. Cheng TO. Fast food, automobiles, television and obesity epidemic in Chinese
children. Int J Cardiol 98: 173-174, 2005.
30. Cheng TO. Physical Activity in Young Children. Lancet 363: 1164, 2004.
31. Chinapaw MJ, Slootmaker SM, Schuit AJ, van Zuidam M, and van Mechelen W.
Reliability and validity of the Activity Questionnaire for Adults and Adolescents
(AQuAA). BMC Med Res Methodol 9: 58, 2009.
32. Churilla JR, and Zoeller RF. Physical Activity: Physical Activity and the Metabolic
Syndrome: A Review of the Evidence. American Journal of Lifestyle Medicine 2:
118-125, 2008.
33. Cohen ML, Tanofsky-Kraff M, Young-Hyman D, and Yanovski JA. Weight and its
relationship to adolescent perceptions of their providers (WRAP): a qualitative and
quantitative assessment of teen weight-related preferences and concerns. J Adolesc
Health 37: 163, 2005.
34. Cole TJ, Bellizzi MC, Flegal KM, and Dietz WH. Establishing a standard definition
for child overweight and obesity worldwide: international survey. BMJ 320: 1240-1243,
2000.
35. Cook S, Auinger P, Li C, and Ford ES. Metabolic syndrome rates in United States
adolescents, from the National Health and Nutrition Examination Survey, 1999-2002. J
Pediatr 152: 165-170, 2008.

136

36. Cook S, Weitzman M, Auinger P, Nguyen M, and 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 157: 821-827,
2003.
37. Crawford PB, Story M, Wang MC, Ritchie LD, and Sabry ZI. Ethnic issues in the
epidemiology of childhood obesity. Pediatric Clinics of North America 48: 855-878,
2001.
38. Crespo CJ, Smit E, Troiano RP, Bartlett SJ, Macera CA, and Andersen RE.
Television watching, energy intake, and obesity in US children: results from the third
National Health and Nutrition Examination Survey, 1988-1994. Arch Pediatr Adolesc
Med 155: 360-365, 2001.
39. Cruz ML, and Goran MI. The metabolic syndrome in children and adolescents. Curr
Diab Rep 4: 53-62, 2004.
40. Cruz ML, Shaibi GQ, Weigensberg MJ, Spruijt-Metz D, Ball GD, and Goran MI.
Pediatric obesity and insulin resistance: chronic disease risk and implications for
treatment and prevention beyond body weight modification. Annu Rev Nutr 25: 435-468,
2005.
41. Cruz ML, Weigensberg MJ, Huang TT, Ball G, Shaibi GQ, and Goran MI. The
metabolic syndrome in overweight Hispanic youth and the role of insulin sensitivity. J
Clin Endocrinol Metab 89: 108-113, 2004.
42. D'Hondt E, Deforche B, De Bourdeaudhuij I, and Lenoir M. Childhood obesity
affects fine motor skill performance under different postural constraints. Neurosci Lett
440: 72-75, 2008.
43. Davis JN, Kelly LA, Lane CJ, Ventura EE, Byrd-Williams CE, Alexandar KA, Azen
SP, Chou CP, Spruijt-Metz D, Weigensberg MJ, Berhane K, and Goran MI. Randomized
control trial to improve adiposity and insulin resistance in overweight Latino adolescents.
Obesity (Silver Spring) 17: 1542-1548, 2009.

137

44. DeFronzo RA, Bonadonna RC, and Ferrannini E. Pathogenesis of NIDDM. A
balanced overview. Diabetes Care 15: 318-368, 1992.
45. Deurenberg P, Deurenberg-Yap M, and Guricci S. Asians are different from
Caucasians and from each other in their body mass index/body fat percent relationship.
Obesity Review 3: 141-146, 2002.
46. Devaraj S, Rosenson RS, and Jialal I. Metabolic syndrome: an appraisal of the
pro-inflammatory and procoagulant status. Endocrinol Metab Clin North Am 33: 431-453,
table of contents, 2004.
47. Dietz WH. The role of lifestyle in health: the epidemiology and consequences of
inactivity. Proc Nutr Soc 55: 829-840, 1996.
48. Ding EL, and Malik VS. Convergence of obesity and high glycemic diet on
compounding diabetes and cardiovascular risks in modernizing China: An emerging
public health dilemma. Global Health 4: 4, 2008.
49. Donoho CJ, Weigensberg MJ, Emken BA, Hsu JW, and Spruijt-Metz D. Stress and
Abdominal Fat: Preliminary Evidence of Moderation by the Cortisol Awakening
Response in Hispanic Peripubertal Girls. Obesity (Silver Spring) Epub ahead of print.
50. Du S, Lu B, Zhai F, and Popkin B. A new stage of the nutrition transition in China.
Public Health Nutr 5: 169-174, 2002.
51. Duncan GE, Li SM, and Zhou X-H. Prevalence and trends of a metabolic
syndrome phenotype among US adolescents, 1999–2000. Diabetes Care 27: 2438-2443,
2004.
52. Duncan TE, Duncan SC, Strycker AL, Li F, and Alpert A. An introduction to latent
variable growth curve modeling: Concepts, issues, and applications. Mahwah, NJ:
Lawrence Erlbaum Associates, 1999.

138

53. Dunstan DW, Salmon J, Owen N, Armstrong T, Zimmet PZ, Welborn TA,
Camerona J, Dwyer T, Jolley D, and Shaw JE. Associations of TV viewing and physical
activity with the metabolic syndrome in Australian adults. Diabetologia 48: 2251-2261,
2005.
54. Ebbeling CB, Sinclair KB, Pereira MA, Garcia-Lago E, Feldman HA, and Ludwig
DS. Compensation for energy intake from fast food among overweight and lean
adolescents. JAMA 291: 2828-2833, 2004.
55. Ekelund U, Anderssen SA, Froberg K, Sardinha LB, Anderssen LB, and Brage S.
Independent associations of physical activity and cardiorespiratory fitness with metabolic
risk factors in children : the European youth heart study. Diabetologia 50: 1832-1840,
2007.
56. Ekelund U, Brage S, Franks PW, Hennings S, Emms S, and Wareham NJ. Physical
activity energy expenditure predicts progression toward the metabolic syndrome
independently of aerobic fitness in middle-aged healthy Caucasians: the Medical
Research Council Ely Study. Diabetes Care 28: 1195-1200, 2005.
57. Ekelund U, Brage S, Froberg K, Harro M, Anderssen SA, Sardinha LB, Riddoch C,
and Andersen LB. TV viewing and physical activity are independently associated with
metabolic risk in children: the European Youth Heart Study. PLoS Med 3: e488, 2006.
58. Ekelund U, Franks PW, Sharp S, Brage S, and Wareham NJ. Increase in physical
activity energy expenditure is associated with reduced metabolic risk independent of
change in fatness and fitness. Diabetes Care 30: 2101-2106, 2007.
59. Elgar FJ, Roberts C, Moore L, and Tudor-Smith C. Sedentary behaviour, physical
activity and weight problems in adolescents in Wales. Public Health 119: 518-524, 2005.
60. Emken BA, Richey J, Belcher B, Hsu Y-W, and Spruijt-Metz D. Objectively
measured physical activity is negatively associated with plasma adiponectin levels in
minority female youth. Int J Pediatr Endocrinol In press.

139

61. Epstein LH, Gordy CC, Raynor HA, Beddome M, Kilanowski CK, and Paluch R.
Increasing fruit and vegetable intake and decreasing fat and sugar intake in families at
risk for childhood obesity. Obes Res 9: 171-178, 2001.
62. Epstein LH, and Roemmich JN. Reducing sedentary behavior: role in modifying
physical activity. Exerc Sport Sci Rev 29: 103-108, 2001.
63. Epstein LH, Roemmich JN, Paluch RA, and Raynor HA. Influence of changes in
sedentary behavior on energy and macronutrient intake in youth. Am J Clin Nutr 81:
361-366, 2005.
64. Expert Panel on Detection Evaluation and Treatment of High Blood Cholesterol in
Adults. 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 285: 2486-2497, 2001.
65. Fernandez JR, Redden DT, Pietrobelli A, and Allison DB. Waist circumference
percentiles in nationally representative samples of African-American,
European-American, and Mexican-American children and adolescents. J Pediatr 145:
439-444, 2004.
66. Fogelholm M, Kronholm E, Kukkonen-Harjula K, Partonen T, Partinen M, and
Harma M. Sleep-related disturbances and physical inactivity are independently associated
with obesity in adults. Int J Obes (Lond) 31: 1713-1721, 2007.
67. Ford ES, Kohl HWI, Mokdad AH, and Ajani UA. Sedentary Behavior, Physical
Activity, and the Metabolic Syndrome among U.S. Adults. Obes Res 13: 608-614, 2005.
68. Ford ES, and Li C. Defining the metabolic syndrome in children and adolescents:
will the real definition please stand up? J Pediatr 152: 160-164, 2008.
69. Freedman DS, Dietz WH, Srinivasan SR, and Berenson GS. The relation of
overweight to cardiovascular risk factors among children and adolescents: the Bogalusa
Heart Study. Pediatrics 103: 1175-1182, 1999.

140

70. Freedman DS, Wang J, Thornton JC, Mei Z, Pierson RN, Jr., Dietz WH, and Horlick
M. Racial/ethnic differences in body fatness among children and adolescents. Obesity
(Silver Spring) 16: 1105-1111, 2008.
71. Fung TT, Hu FB, Yu J, Chu NF, Spiegelman D, Tofler GH, Willett WC, and Rimm
EB. Leisure-time physical activity, television watching, and plasma biomarkers of
obesity and cardiovascular disease risk. Am J Epidemiol 152: 1171-1178, 2000.
72. Gannage-Yared MH, Khalife S, Semaan M, Fares F, Jambart S, and Halaby G.
Serum adiponectin and leptin levels in relation to the metabolic syndrome, androgenic
profile and somatotropic axis in healthy non-diabetic elderly men. Eur J Endocrinol 155:
167-176, 2006.
73. Gidlow CJ, Cochrane T, Davey R, and Smith H. In-school and out-of-school
physical activity in primary and secondary school children. J Sports Sci 26: 1411-1419,
2008.
74. Gomez SL, Kelsey JL, Glaser SL, Lee MM, and Sidney S. Immigration and
acculturation in relation to health and health-related risk factors among specific Asian
subgroups in a health maintenance organization. Am J Public Health 94: 1977-1984,
2004.
75. Goodman E, Daniels SR, Meigs JB, and Dolan LM. Instability in the diagnosis of
metabolic syndrome in adolescents. Circulation 115: 2316-2322, 2007.
76. Goran MI. Metabolic precursors and effects of obesity in children: a decade of
progress, 1990-1999. American Journal of Clinical Nutrition 73: 158-171, 2001.
77. Goran MI, Bergman RN, Cruz ML, and Watanabe R. Insulin resistance and
associated compensatory responses in african-american and Hispanic children. Diabetes
Care 25: 2184-2190, 2002.
78. Goran MI, and Treuth MS. Energy expenditure, physical activity, and obesity in
children. Pediatr Clin North Am 48: 931-953., 2001.

141

79. Gordon-Larsen P, Adair LS, and Popkin BM. Ethnic differences in physical activity
and inactivity patterns and overweight status. Obesity Research 10: 141-149, 2002.
80. Gordon-Larsen P, McMurray RG, and Popkin BM. Adolescent physical activity and
inactivity vary by ethnicity: The National Longitudinal Study of Adolescent Health.
Journal of Pediatrics 135: 301-306, 1999.
81. Gortmaker SL, Peterson K, Wiecha J, Sobol AM, Dixit S, Fox MK, and Laird N.
Reducing obesity via a school-based interdisciplinary intervention among youth: Planet
Health. Arch Pediatr Adolesc Med 153: 409-418, 1999.
82. Grundy SM. A constellation of complications: the metabolic syndrome. Clin
Cornerstone 7: 36-45, 2005.
83. Grundy SM. Obesity, metabolic syndrome, and cardiovascular disease. J Clin
Endocrinol Metab 89: 2595-2600, 2004.
84. Ham SA, Reis JP, Strath SJ, Dubose KD, and Ainsworth BE. Discrepancies between
methods of identifying objectively determined physical activity. Med Sci Sports Exerc 39:
52-58, 2007.
85. Hawkes C. Agro-food industry growth and obesity in China: what role for regulating
food advertising and promotion and nutrition labelling? Obes Rev 9: 151-161, 2008.
86. Hayden-Wade HA, Stein RI, Ghaderi A, Saelens BE, Zabinski MF, and Wilfley DE.
Prevalence, characteristics, and correlates of teasing experiences among overweight
children vs. non-overweight peers. Obes Res 13: 1381-1392, 2005.
87. Hickman TB BR, Carroll MD, Rifkind BM, Cleeman JI, Maurer, and KR JC.
Distributions and trends of serum lipid levels among United States children and
adolescents ages 4-19 years: Data from the Third National Health and Nutrition
Examination Survey. Prev Med 27: 879-890, 1998.

142

88. Hills AP, King NA, and Armstrong TP. The contribution of physical activity and
sedentary behaviours to the growth and development of children and adolescents:
implications for overweight and obesity. Sports Med 37: 533-545, 2007.
89. Holme I, Tonstad S, Sogaard AJ, Larsen PG, and Haheim LL. Leisure time physical
activity in middle age predicts the metabolic syndrome in old age: results of a 28-year
follow-up of men in the Oslo study. BMC Public Health 7: 154, 2007.
90. Howley ET. Type of activity: resistance, aerobic and leisure versus occupational
physical activity. Med Sci Sports Exerc 33: S364-369; discussion S419-320, 2001.
91. Hsu Y-W, Belcher BR, Ventura EE, Byrd-Williams CE, Weigensberg MJ, Davis J,
McClain A, Goran MI, and Spruijt-Metz D. Physical Activity, Sedentary Behavior, and
the Metabolic Syndrome in Minority Youth. Med Sci Sports Exerc Revise and Resubmit.
92. Hsu Y-W, Chou C-P, Nguyen-Rodriguez ST, McClain AD, Belcher BR, and
Spruijt-Metz D. Influences of Social Support, Perceived Barriers, and Negative Meanings
of Physical Activity on Physical Activity in Middle School Students. Journal of Physical
Activity & Health 8: 210-219, 2011.
93. Hunt KJ, Williams K, Hazuda HP, Stern MP, and Haffner SM. The metabolic
syndrome and the impact of diabetes on coronary heart disease mortality in women and
men: the San Antonio Heart Study. Ann Epidemiol 17: 870-877, 2007.
94. Iwata F, Hara M, Okada T, Harada K, and Li S. Body fat ratios in urban Chinese
children. Pediatr Int 45: 190-192, 2003.
95. Janssen I, Craig WM, Boyce WF, and Pickett W. Associations between overweight
and obesity with bullying behaviors in school-aged children. Pediatrics 113: 1187-1194,
2004.
96. Janz KF. Physical activity in epidemiology: moving from questionnaire to objective
measurement. Br J Sports Med 40: 191-192, 2006.

143

97. Jebb SA. Energy intake and body weight. In: Eating Disorders and Obesity: A
Comprehensive Handbook, edited by Fairburn CG, and Brownell KD. New York:
Guilford, 2002, p. 37-42.
98. Johnson CA, Palmer PH, Chou CP, Pang Z, Zhou D, Dong L, Xiang H, Yang P, Xu
H, Wang J, Fu X, Guo Q, Sun P, Ma H, Gallaher PE, Xie B, Lee L, Fang T, and Unger
JB. Tobacco use among youth and adults in Mainland China: the China Seven Cities
Study. Public Health 120: 1156-1169, 2006.
99. Johnson CA, Xie B, Liu C, Reynolds KD, Chou CP, Koprowski C, Gallaher P,
Spruitj-Metz D, Guo Q, Sun P, Gong J, and Palmer P. Socio-demographic and cultural
comparison of overweight and obesity risk and prevalence in adolescents in Southern
California and Wuhan, China. J Adolesc Health 39: 925 e921-928, 2006.
100. Kahn SE, Hull RL, and Utzschneider KM. Mechanisms linking obesity to insulin
resistance and type 2 diabetes. Nature 444: 840-846, 2006.
101. Kelishadi R, Razaghi EM, Gouya MM, Ardalan G, Gheiratmand R, Delavari A,
Motaghian M, Ziaee V, Siadat ZD, Majdzadeh R, Heshmat R, Barekati H, Arabi MSM,
Heidarzadeh A, and Shariatinejad K. Association of Physical Activity and the Metabolic
Syndrome in Children and Adolescents: CASPIAN Study. Horm Res 67: 46-52, 2006.
102. Kelly LA, Lane CJ, Ball GD, Weigensberg MJ, Vargas LG, Byrd-Williams CE,
Ventura EE, and Goran MI. Birth weight and body composition in overweight Latino
youth: a longitudinal analysis. Obesity (Silver Spring) 16: 2524-2528, 2008.
103. Key JK, Schatzkin A, and Willet WCea. Diet, nutrition and the prevention of cancer.
Public Health Nutr 7: 187-200, 2004.
104. Kim Y, and Lee S. Physical activity and abdominal obesity in youth. Appl Physiol
Nutr Metab 34: 571-581, 2009.
105. Kimm SY, Glynn NW, Kriska AM, Barton BA, Kronsberg SS, Daniels SR,
Crawford PB, Sabry ZI, and Liu K. Decline in physical activity in black girls and white
girls during adolescence. N Engl J Med 347: 709-715, 2002.

144

106. Kirk S, Zeller M, Claytor R, Santangelo M, Khoury PR, and Daniels SR. The
relationship of health outcomes to improvement in BMI in children and adolescents.
Obes Res 13: 876-882, 2005.
107. Korsten-Reck U, Kaspar T, Korsten K, Kromeyer-Hauschild K, Bos K, Berg A, and
Dickhuth HH. Motor abilities and aerobic fitness of obese children. Int J Sports Med 28:
762-767, 2007.
108. Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, Mei Z, Wei
R, Curtin LR, Roche AF, and Johnson CL. 200 CDC Growth Charts for the United States:
methods and development. Vital Health Stat 11 246: 1-190, 2002.
109. Kumahara H, Schutz Y, Ayabe M, Yoshioka M, Yoshitake Y, Shindo M, Ishii K,
and Tanaka H. The use of uniaxial accelerometry for the assessment of
physical-activity-related energy expenditure: a validation study against whole-body
indirect calorimetry. Br J Nutr 91: 235-243, 2004.
110. Laaksonen DE, Lakka HM, Niskanen LK, Kaplan GA, Salonen JT, and Lakka TA.
Metabolic syndrome and development of diabetes mellitus: application and validation of
recently suggested definitions of the metabolic syndrome in a prospective cohort study.
Am J Epidemiol 156: 1070-1077, 2002.
111. Laaksonen DE, Lakka HM, Salonen JT, Niskanen LK, Rauramaa R, and Lakka TA.
Low levels of leisure-time physical activity and cardiorespiratory fitness predict
development of the metabolic syndrome. Diabetes Care 25: 1612-1618, 2002.
112. Laaksonen DE, Lakka HM, Salonen JT, Niskanen LK, Rauramaa R, and Lakka TA.
Low levels of leisure-time physical activity and cardiorespiratory fitness predict
development of the metabolic syndrome. Diabetes Care 25: 1612-1618, 2002.
113. Lee JM, Okumura MJ, Davis MM, Herman WH, and Gurney JG. Prevalence and
determinants of insulin resistance among U.S. adolescents: a population-based study.
Diabetes Care 29: 2427-2432, 2006.

145

114. Li M, Dibley M, Sibbritt D, and Yan H. Factors associated with adolescents'
physical inactivity in Xi'an City, China. Med Sci Sports Exerc 38: 2075-2085, 2006.
115. Link BG, and Phelan J. Social conditions as fundamental causes of disease. J Health
Soc Behav Spec No: 80-94, 1995.
116. Littman AJ, Vitiello MV, Foster-Schubert K, Ulrich CM, Tworoger SS, Potter JD,
Weigle DS, and McTiernan A. Sleep, ghrelin, leptin and changes in body weight during a
1-year moderate-intensity physical activity intervention. Int J Obes (Lond) 31: 466-475,
2007.
117. Liu C, Xie B, Chou CP, Koprowski C, Zhou D, Palmer P, Sun P, Guo Q, Duan L,
Sun X, and Anderson Johnson C. Perceived stress, depression and food consumption
frequency in the college students of China Seven Cities. Physiol Behav 92: 748-754,
2007.
118. Liu X, Liu L, Owens JA, and Kaplan DL. Sleep patterns and sleep problems among
schoolchildren in the United States and China. Pediatrics 115: 241-249, 2005.
119. Liu X, Zhao Z, Jia C, and Buysse DJ. Sleep patterns and problems among chinese
adolescents. Pediatrics 121: 1165-1173, 2008.
120. Ludwig DS. The glycemic index: physiological mechanisms relating to obesity,
diabetes, and cardiovascular disease. JAMA 287: 2414-2423, 2002.
121. Lumeng J, Somashekar D, Appugliese D, Kaciroti N, Corwyn R, and Bradley R.
Shorter Sleep Duration Is Associated With Increased Risk for Being Overweight at Ages
9 to 12 Years. Pediatrics 12: 1020-1029, 2007.
122. Ma GS, Li YP, Hu XQ, Ma WJ, and Wu J. Effect of television viewing on pediatric
obesity. Biomed Environ Sci 15: 291-297, 2002.

146

123. Ma GS, Li YP, Wu YF, Zhai FY, Cui ZH, Hu XQ, Luan DC, Hu YH, and Yang XG.
The prevalence of body overweight and obesity and its changes among Chinese people
during 1992 to 2002. Zhonghua Yu Fang Yi Xue Za Zhi 39: 311-315, 2005.
124. Maffeis C, Moghetti P, Grezzani A, Clementi M, Gaudino R, and Tato L. Insulin
resistance and the persistence of obesity from childhood into adulthood. Journal of
Clinical Endocrinology & Metabolism 87: 71-76, 2002.
125. Malik S, Wong ND, Franklin SS, Kamath TV, L'Italien GJ, Pio JR, and Williams
GR. Impact of the metabolic syndrome on mortality from coronary heart disease,
cardiovascular disease, and all causes in United States adults. Circulation 110:
1245-1250, 2004.
126. Malin AS, Qi D, Shu XO, Gao YT, Friedmann JM, Jin F, and Zheng W. Intake of
fruits, vegetables and selected micronutrients in relation to the risk of breast cancer. Int J
Cancer 105: 413-418, 2003.
127. Marshall SJ, Biddle SJ, Gorely T, Cameron N, and Murdey I. Relationships between
media use, body fatness and physical activity in children and youth: a meta-analysis. Int J
Obes Relat Metab Disord 28: 1238-1246, 2004.
128. Marshall SJ, Biddle SJ, Sallis JF, McKenzie TL, and Conway TL. Clustering of
sedentary behaviors and physical acitvity among youth: a cross-national study. Pediatr
Exercise Science 14: 401-147, 2002.
129. Matthews CE, Chen KY, Freedson PS, Buchowski MS, Beech BM, Pate RR, and
Troiano RP. Amount of Time Spent in Sedentary Behaviors in the United States,
2003-2004. Am J Epidemiol 167: 875-881, 2008.
130. Matthews CE, Chen KY, Freedson PS, Buchowski MS, Beech BM, Pate RR, and
Troiano RP. Amount of time spent in sedentary behaviors in the United States,
2003-2004. Am J Epidemiol 167: 875-881, 2008.

147

131. McMurray RG, Bangdiwala SI, Harrell JS, and Amorim LD. Adolescents with
metabolic syndrome have a history of low aerobic fitness and physical activity levels.
Dynamic Medicine 7: 1-6, 2008.
132. McMurray RG, Ring KB, Treuth MS, Welk GJ, Pate RR, Schmitz KH, Pickrel JL,
Gonzalez V, Almedia MJ, Young DR, and Sallis JF. Comparison of two approaches to
structured physical activity surveys for adolescents. Med Sci Sports Exerc 36: 2135-2143,
2004.
133. McNeely MJ, and Boyko EJ. Type 2 Diabetes Prevalence in Asian Americans:
Results of a national health survey Diabetes Care 27: 66-69, 2004.
134. McTiernan A. Mechanisms linking physical activity with cancer. Nat Rev Cancer 8:
205-211, 2008.
135. Metcalf BS, Hosking J, Jeffery AN, Voss LD, Henley W, and Wilkin TJ. Fatness
leads to inactivity, but inactivity does not lead to fatness: a longitudinal study in children
(EarlyBird 45). Arch Dis Child.
136. Monteiro CA, Moura EC, Conde WL, and Popkin BM. Socioeconomic status and
obesity in adult populations of developing countries: a review. Bull World Health Organ
82: 940-946, 2004.
137. Moore JB, Davis CL, Baxter SD, Lewis RD, and Yin Z. Physical Activity,
Metabolic Syndrome, and Overweight in Rural Youth. The Journal of Rural Health 24:
136-142, 2008.
138. Moran A, Jacobs DR, Jr., Steinberger J, Hong CP, Prineas R, Luepker R, and
Sinaiko AR. Insulin resistance during puberty: results from clamp studies in 357 children.
Diabetes 48: 2039-2044, 1999.
139. Morano M, Colella D, and Capranica L. Body image, perceived and actual physical
abilities in normal-weight and overweight boys involved in individual and team sports. J
Sports Sci 1-8, Epub ahead of print.

148

140. Morano M, Colella D, Robazza C, Bortoli L, and Capranica L. Physical
self-perception and motor performance in normal-weight, overweight and obese children.
Scand J Med Sci Sports Epub ahead of print.
141. Moreno LA, Pineda I, Rodriguez G, Fleta J, Giner A, Juste MG, Sarria A, and
Bueno M. Leptin and metabolic syndrome in obese and non-obese children. Horm Metab
Res 34: 394-399, 2002.
142. Morris V. Community-based programming to treat childhood obesity. Curr Diab
Rep 6: 395-400, 2006.
143. Morrison JA, Friedman LA, and Gray-McGuire C. Metabolic syndrome in
childhood predicts adult cardiovascular disease 25 years later: the Princeton Lipid
Research Clinics Follow-up Study. Pediatrics 120: 340-345, 2007.
144. Murasko JE. Socioeconomic status, height, and obesity in children. Econ Hum Biol
7: 376-386, 2009.
145. Must A, and Strauss R. Risks and consequences of childhood and adolescents
obesity. Int J Obes Relat Metab Disord 23: S2-11, 1999.
146. Must A, and Strauss RS. Risks and consequences of childhood and adolescent
obesity. Int J Obes Relat Metab Disord 23 Suppl 2: S2-11, 1999.
147. Must A, and Tybor DJ. Physical activity and sedentary behavior: a review of
longitudinal studies of weight and adiposity in youth. Int J Obes (Lond) 29 Suppl 2:
S84-96, 2005.
148. Mutz DC, Roberts DF, and van Vuuren DP. Reconsidering the displacement
hypothesis. Commun Res 20: 51-75, 1993.
149. National Center for Health Statistics. Health, United States, 2004 with Chartbook on
Trends in the Health of Americans. Hyattsville, MD: 2004.

149

150. National Cholesterol Education Program (NCEP) Expert Panel on Detection E, And
Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III),. 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) final report. JAMA 285: 2486-2497, 2001.
151. National High Blood Pressure Education Program Working Group on High Blood
Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation, and
treatment of high blood pressure in children and adolescents. Pediatrics 114(suppl 4th
rep): 555-576., 2004.
152. Nelson MC, Gordon-Larsen P, Adair LS, and Popkin BM. Adolescent physical
activity and sedentary behavior: patterning and long-term maintenance. Am J Prev Med
28: 259-266, 2005.
153. Ogden CL, Carroll MD, Curtin LR, Lamb MM, and Flegal KM. Prevalence of high
body mass index in US children and adolescents, 2007-2008. JAMA 303: 242-249.
154. Ogden CL, Carroll MD, Curtin LR, Lamb MM, and Flegal KM. Prevalence of high
body mass index in US children and adolescents, 2007-2008. JAMA 303: 242-249, 2010.
155. Ogden CL, Flegal KM, Carroll MD, and Johnson CL. Prevalence and trends in
overweight among US children and adolescents, 1999-2000. JAMA 288: 1728-1732,
2002.
156. Ogden CL, Lamb MM, Carroll MD, and Flegal KM. Obesity and socioeconomic
status in children and adolescents: United States, 2005-2008. NCHS Data Brief 1-8, 2010.
157. Owen N, Leslie E, Salmon J, and Fotheringham MJ. Environmental determinants of
physical activity and sedentary behavior. Exerc Sport Sci Rev 28: 153-158, 2000.
158. Pallan MJ, Hiam LC, Duda JL, and Adab P. Body image, body dissatisfaction and
weight status in South Asian children: a cross-sectional study. BMC Public Health 11:
21.

150

159. Pan Y, and Pratt CA. Metabolic syndrome and its association with diet and physical
activity in US adolescents. J Am Diet Assoc 108: 276-286; discussion 286, 2008.
160. Park YW, Allison DB, Heymsfield SB, and Gallagher D. Larger amounts of visceral
adipose tissue in Asian Americans. Obes Res 9: 381-387, 2001.
161. Park YW, Zhu S, Palaniappan L, Heshka S, Carnethon MR, and Heymsfield SB.
The metabolic syndrome: prevalence and associated risk factor findings in the US
population from the Third National Health and Nutrition Examination Survey, 1988-1994.
Arch Intern Med 163: 427-436, 2003.
162. Pate RR, Long BJ, and Heath G. Descriptive epidemiology of physical activity in
adolescents. Pediatric Exercise Science 6: 434-447, 1994.
163. Pate RR, O'Neill JR, and Lobelo F. The evolving definition of "sedentary". Exerc
Sport Sci Rev 36: 173-178, 2008.
164. Pate RR, Ross R, Dowda M, Trost SG, and Saland JM. Validation of a 3-Day
Physical Activity Recall Instrument in Female Youth Pediatric Exercise Science 15:
257-265, 2003.
165. Pate RR, Stevens J, Webber LS, Dowda M, Murray DM, Young DR, and Going S.
Age-related change in physical activity in adolescent girls. J Adolesc Health 44: 275-282,
2009.
166. Patrick K, Norman G, Calfas K, Sallis J, Zabinski M, Rupp J, and Cella J. Diet,
physical activity, and sedentary behaviors as risk factors for overweight in adolescence.
Arch Pediatr Adolesc Med 158: 385-390, 2004.
167. Platat C, Wagner A, Klumpp T, Schweitzer B, and Simon C. Relationships of
Physical Activity with Metabolic Syndrome Features and Low-grade Inflammation in
Adolescents. Diabetologia 49: 2078-2085, 2006.

151

168. Popkin BM, and Gordon-Larsen P. The nutrition transition: worldwide obesity
dynamics and their determinants. Int J Obes Relat Metab Disord 28 Suppl 3: S2-9, 2004.
169. Puyau MR, Adolph AL, Vohra FA, Zakeri I, and Butte NF. Prediction of activity
energy expenditure using accelerometers in children. Med Sci Sports Exerc 36:
1625-1631, 2004.
170. Raitakari OT, Porkka KV, Taimela S, Telama R, Rasanen L, and Viikari JS. Effects
of persistent physical activity and inactivity on coronary risk factors in children and
young adults. The Cardiovascular Risk in Young Finns Study. Am J Epidemiol 140:
195-205, 1994.
171. Reaven GM. Role on insulin resistance in human disease. Diabetes 37: 1595-1607,
1988.
172. Rizzo NS, Ruiz JR, Hurtig-Wennlof A, Ortega FB, and Sjostrom M. Relationship of
physical activity, fitness and fatness with clustered metabolic risk in children and
adolescents: The European Youth Heart Study. J Pediatr 150: 388-394, 2007.
173. Robinson LE, and Graham TE. Metabolic syndrome, a cardiovascular disease risk
factor: role of adipocytokines and impact of diet and physical activity. Can J Appl
Physiol 29: 808-829, 2004.
174. Romon M, Lafay L, Bresson JL, Oppert JM, Borys JM, Kettaneh A, and Charles
MA. Relationships between physical activity and plasma leptin levels in healthy children:
the Fleurbaix-Laventie Ville Sante II Study. Int J Obes Relat Metab Disord 28:
1227-1232, 2004.
175. Rural Survey Organization of National Bureau of Statistics of China. Poverty
Statistics in China 2004 (On-line) National Bureau of Statistics of China.
http://www.nscb.gov.ph/poverty/conference/papers/4_Poverty%20Statistics%20in%20Ch
ina.pdf. [June 22, 2010].

152

176. Sakamaki R, Toyama K, Amamoto R, Liu CJ, and Shinfuku N. Nutritional
knowledge, food habits and health attitude of Chinese university students--a cross
sectional study. Nutr J 4: 1-11, 2005.
177. Saland JM. Update on the metabolic syndrome in children Curr Opin Pediatrics 19:
183-191, 2007.
178. Sallis JF, Prochaska JJ, and Taylor WC. A review of correlates of physical activity
of children and adolescents. Med Sci Sports Exerc 32: 963-975, 2000.
179. Salmon J, Bauman A, Crawford D, Timperio A, and Owen N. The association
between television viewing and overweight among Australian adults participating in
varying levels of leisure-time physical activity. Int J Obes Relat Metab Disord 24:
600-606, 2000.
180. Schwartz AR, Patil SP, Laffan AM, Polotsky V, Schneider H, and Smith PL.
Obesity and obstructive sleep apnea: pathogenic mechanisms and therapeutic approaches.
Proc Am Thorac Soc 5: 185-192, 2008.
181. Sekine M, Yamagami T, Handa K, Saito T, Nanri S, Kawaminami K, Tokui N,
Yoshida K, and Kagamimori S. A dose-response relationship between short sleeping
hours and childhood obesity: results of the Toyama Birth Cohort Study. Child Care
Health Dev 28: 163-170, 2002.
182. Shahabi A, Bernstein L, Azen SP, McClain AD, Belcher B, Hsu Y-H,
Nguyen-Rodriguez ST, and Spruijt-Metz D. Recruitment and retention of
African-American and Latina pre-adolescent females into a longitudinal bio-behavioral
study. Ethnicity and Disease In Press.
183. Shaibi GQ, Cruz ML, Ball GD, Weigensberg MJ, Kobaissi HA, Salem GJ, and
Goran MI. Cardiovascular fitness and the metabolic syndrome in overweight latino
youths. Med Sci Sports Exerc 37: 922-928, 2005.
184. Shaibi GQ, and Goran MI. Examining metabolic syndrome definitions in overweight
Hispanic youth: a focus on insulin resistance. J Pediatr 152: 171-176, 2008.

153

185. Shi Z, Hu X, Yuan B, Hu G, Pan X, Dai Y, Byles JE, and Holmboe-Ottesen G.
Vegetable-rich food pattern is related to obesity in China. Int J Obes (Lond) 32: 975-984,
2008.
186. Sirard JR, and Pate RR. Physical activity assessment in children and adolescents.
Sports Med 31: 439-454, 2001.
187. Slootmaker SM, Schuit AJ, Chinapaw MJ, Seidell JC, and van Mechelen W.
Disagreement in physical activity assessed by accelerometer and self-report in subgroups
of age, gender, education and weight status. Int J Behav Nutr Phys Act 6: 17, 2009.
188. Sobal J, and Stunkard AJ. Socioeconomic status and obesity: a review of the
literature. Psychol Bull 105: 260-275, 1989.
189. Song YM. Commentary: varying relation of socioeconomic status with obesity
between countries at different stages of development. Int J Epidemiol 35: 112-113, 2006.
190. Spruijt-Metz D, Berrigan D, Kelly L, McConnell R, Dueker D, Lindsey G, Atienza
AA, Nguyen-Rodriguez S, Irwin ML, Wolch J, Jerrett M, Tatalovich Z, and Redline S.
Measures of physical activity and exercise. In: Handbook of Assessment Methods for
Eating Behaviors and Weight-Related Problems: Measures, Theory, and Research, edited
by Allison DB, and Baskin ML. Thousand Oaks, CA: Sage Publications, 2009, p. pp.
187-254.
191. Spruijt-Metz D, Emken BA, Richey J, Hsu Y-W, Nguyen-Rodriguez ST, Belcher
BR, McClain AD, Goran MI, and Weigensberg MJ. CRP is related to higher leptin levels
in minority prepubertal females regardless of adiposity levels. Obesity In Press.
192. Spruijt-Metz D, Nguyen-Michel ST, Goran MI, Chou CP, and Huang TT. Reducing
sedentary behavior in minority girls via a theory-based, tailored classroom media
intervention. Int J Pediatr Obes 3: 240-248, 2008.
193. Steele RM, Brage S, Corder K, Wareham NJ, and Ekelund U. Physical activity,
cardiorespiratory fitness, and the metabolic syndrome in youth. J Appl Physiol 105:
342-351, 2008.

154

194. Stigler MH, Arora M, Dhavan P, Shrivastav R, Reddy KS, and Perry CL.
Weight-related concerns and weight-control behaviors among overweight adolescents in
Delhi, India: A cross-sectional study. Int J Behav Nutr Phys Act 8: 9.
195. Taheri S, Lin L, Austin D, Young T, and Mignot E. Short sleep duration is
associated with reduced leptin, elevated ghrelin, and increased body mass index. PLoS
Med 1:e62.: 2004.
196. Tanner JM. Growth and maturation during adolescence. Nutr Rev 39: 43-55, 1981.
197. Taylor ED, Theim KR, Mirch MC, Ghorbani S, Tanofsky-Kraff M, Adler-Wailes
DC, Brady S, Reynolds JC, Calis KA, and Yanovski JA. Orthopedic complications of
overweight in children and adolescents. Pediatrics 117: 2167-2174, 2006.
198. Taylor H, Liu J, Wilson G, Golden SH, Crook E, Brunson CD, Steffes M, Johnson
WD, and Sung JH. Distinct component profiles and high risk among African Americans
with metabolic syndrome: the Jackson Heart Study. Diabetes Care 31: 1248-1253, 2008.
199. 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) Final Report. Circulation 106: 3143-3421, 2002.
200. Treuth MS, Baggett CD, Pratt CA, Going SB, Elder JP, Charneco EY, and Webber
LS. A longitudinal study of sedentary behavior and overweight in adolescent girls.
Obesity (Silver Spring) 17: 1003-1008, 2009.
201. Treuth MS, Hou N, Young DR, and Maynard LM. Accelerometry-Measured
Activity or Sedentary Time and Overweight in Rural Boys and Girls. Obes Res 13:
1606-1614, 2005.
202. Treuth MS, Hou N, Young DR, and Maynard LM. Validity and reliability of the
Fels physical activity questionnaire for children. Med Sci Sports Exerc 37: 488-495,
2005.

155

203. Treuth MS, Schmitz K, Catellier DJ, McMurray RG, Murray DM, Almeida MJ,
Going S, Norman JE, and Pate R. Defining accelerometer thresholds for activity
intensities in adolescent girls. Med Sci Sports Exerc 36: 1259-1266, 2004.
204. Troiano RP, ., Berrigan D, Dodd KW, Masse LC, Tilert T, and McDowell M.
Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc
40: 181-188, 2008.
205. Trost SG, Kerr LM, Ward DS, and Pate RR. Physical activity and determinants of
physical activity in obese and non-obese children. Int J Obes Relat Metab Disord 25:
822-829, 2001.
206. Trost SG, Pate RR, Sallis JF, Freedson PS, Taylor WC, Dowda M, and Sirard J. Age
and gender differences in objectively measured physical activity in youth. Med Sci Sports
Exerc 34: 350-355, 2002.
207. Tudor-Locke C, Ainsworth B, Adair L, Du S, and Popkin B. Physical activity and
inactivity in Chinese school-aged youth: the China Health and Nutrition Survey. Int J
Obes (Lond) 27: 1093-1099, 2003.
208. U.S. Department of Health and Human Services. Physical Activity Guidelines
Advisory Committee Report. 2008.
209. US Department of Health and Huaman Services. 2008 Physical Activity Guidelines
for Americans. 2008.
210. Ventura E, Davis J, Byrd-Williams C, Alexander K, McClain A, Lane CJ,
Spruijt-Metz D, Weigensberg M, and 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 163: 320-327, 2009.
211. Ventura EE, Lane CJ, Weigensberg MJ, Toledo-Corral CM, Davis JN, and Goran
MI. Persistence of the metabolic syndrome over 3 annual visits in overweight Hispanic
children: association with progressive risk for type 2 diabetes. J Pediatr 155: 535-541,
2009.

156

212. Wang Y. Cross-national comparison of childhood obesity: the epidemic and the
relationship between obesity and socioeconomic status. Int J Epidemiol 30: 1129-1136,
2001.
213. Wang Y, Beydoun MA, Liang L, Caballero B, and Kumanyika SK. Will all
Americans become overweight or obese? estimating the progression and cost of the US
obesity epidemic. Obesity (Silver Spring) 16: 2323-2330, 2008.
214. Wang Y, Monteiro C, and Popkin BM. Trends of obesity and underweight in older
children and adolescents in the United States, Brazil, China, and Russia. Am J Clin Nutr
75: 971-977, 2002.
215. Wang Y, and Wang JQ. A comparison of international references for the assessment
of child and adolescent overweight and obesity in different populations. Eur J Clin Nutr
56: 973-982, 2002.
216. Wareham N. Physical activity and obesity prevention. obesity reviews 8 109-114,
2007.
217. Weigensberg MJ, Toledo-Corral CM, and Goran MI. Association between the
metabolic syndrome and serum cortisol in overweight Latino youth. J Clin Endocrinol
Metab 93: 1372-1378, 2008.
218. Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW, Allen K,
Lopes M, Savoye M, Sherwin RS, and Caprio S. Obesity and the metabolic syndrome in
children and adolescents. N Engl J Med 350: 2362-2374, 2004.
219. Weston AT, Petosa R, and Pate RR. Validation of an instrument for measurement of
physical activity in youth. Medicine & Science in Sports & Exercise 29: 138-143, 1997.
220. WHO Expert Consultation. Appropriate body-mass index for Asian populations and
its implications for policy and intervention strategies. Lancet 363: 157-163, 2004.

157

221. WHO/FAO. Diet, nutrition and the prevention of chronic diseases: Report of the
joint WHO/FAO expert consultation Geneva: World Health Organization, 2003.
222. Wildman RP, Gu D, Reynolds K, Duan X, and He J. Appropriate body mass index
and waist circumference cutoffs for categorization of overweight and central adiposity
among Chinese adults. Am J Clin Nutr 80: 1129-1136, 2004.
223. Winkler EA, Gardiner PA, Healy GN, Clark BK, Sugiyama T, Matthews CE, and
Owen NG. Distinguishing True Sedentary From Accelerometer Non-wearing Time. Med
Sci Sports Exerc 41: 171-172, 2009.
224. Wolf AM, Gortmaker SL, Cheung L, Gray HM, Herzog DB, and Colditz GA.
Activity, inactivity, and obesity: racial, ethnic, and age differences among schoolgirls.
American Journal of Public Health 83: 1625-1627, 1993.
225. Wong SL, Leatherdale ST, and Manske SR. Reliability and validity of a
school-based physical activity questionnaire. Med Sci Sports Exerc 38: 1593-1600, 2006.
226. World Health Organization. The WHO Child Growth Standards
http://www.who.int/childgrowth/standards/bmi_for_age/en/index.html. [07/15, 2010].
227. Wu Y. Overweight and obesity in China. BMJ 333: 362-333, 2006.
228. Wu YF, Ma GS, Hu YH, Li YP, Li X, Cui ZH, Chen CM, and Kong LZ. The current
prevalence status of body overweight and obesity in China: data from the China National
Nutrition and Health Survey. Zhonghua Yu Fang Yi Xue Za Zhi 39: 316-320, 2005.
229. Xie B, Chou CP, Spruijt-Metz D, Liu C, Xia J, Gong J, Li Y, and Johnson CA.
Effects of perceived peer isolation and social support availability on the relationship
between body mass index and depressive symptoms. Int J Obes (Lond) 29: 1137-1143,
2005.

158

230. Xie B, Chou CP, Spruijt-Metz D, Reynolds K, Clark F, Palmer PH, Gallaher P, Sun
P, Guo Q, and Johnson CA. Weight perception and weight-related sociocultural and
behavioral factors in Chinese adolescents. Prev Med 42: 229-234, 2006.
231. Xie B, Chou CP, Spruijt-Metz D, Reynolds K, Clark F, Palmer PH, Gallaher P, Sun
P, Guo Q, and Johnson CA. Weight perception, academic performance, and
psychological factors in Chinese adolescents. Am J Health Behav 30: 115-124, 2006.
232. Yang X, Telama R, Hirvensalo M, Mattsson N, Viikari JS, and Raitakari OT. The
longitudinal effects of physical activity history on metabolic syndrome. Med Sci Sports
Exerc 40: 1424-1431, 2008.
233. Young-Hyman D, Tanofsky-Kraff M, Yanovski SZ, Keil M, Cohen ML, Peyrot M,
and Yanovski JA. Psychological status and weight-related distress in overweight or
at-risk-for-overweight children. Obesity (Silver Spring) 14: 2249-2258, 2006.
234. Young LR, and Nestle M. The contribution of expanding portion sizes to the US
obesity epidemic. Am J Public Health 92: 246-249, 2002.
235. Yu Y, Lu BS, Wang B, Wang H, Yang J, Li Z, Wang L, Liu X, Tang G, Xing H, Xu
X, Zee PC, and Wang X. Short sleep duration and adiposity in Chinese adolescents. Sleep
30: 1688-1697, 2007.
236. Zabinski MF, Saelens BE, Stein RI, Hayden-Wade HA, and Wilfley DE.
Overweight children's barriers to and support for physical activity. Obes Res 11: 238-246,
2003.
237. Zhao W, Zhai Y, Hu J, Wang J, Yang Z, Kong L, and Chen C. Economic burden of
obesity-related chronic diseases in Mainland China. Obes Rev 9 Suppl 1: 62-67, 2008.
238. Zhou B, Wu Y, Yang J, Li Y, Zhang H, and Zhao L. Overweight is an independent
risk factor for cardiovascular disease in Chinese populations. Obes Rev 3: 147-156, 2002.

159

239. Zhou BF. Predictive values of body mass index and waist circumference for risk
factors of certain related diseases in Chinese adults--study on optimal cut-off points of
body mass index and waist circumference in Chinese adults. Biomed Environ Sci 15:
83-96, 2002.
240. Zhu S, St-Onge MP, Heshka S, and Heymsfield SB. Lifestyle behaviors associated
with lower risk of having the metabolic syndrome. Metabolism 53: 1503-1511, 2004.
241. Zimmet P, Magliano D, Matsuzawa Y, Alberti G, and Shaw J. The metabolic
syndrome: a global public health problem and a new definition. J Atheroscler Thromb 12:
295-300, 2005.
242.Zoelle RF. Physical Activity, Sedentary Behavior, and Overweight/Obesity in Youth:
Evidence From Cross-sectional, Longitudinal, and Interventional Studies. Am J Lifestyle
Med 3: 110-114, 2009.
243.Zoelle RF. Physical Activity, Sedentary Behavior, and Overweight/Obesity in Youth:
Evidence From Cross-sectional, Longitudinal, and Interventional Studies. American
Journal of Lifestyle Medicine 3: 110-114, 2009.

160

APPENDICES
Appendix A: Adult Definition of the Metabolic Syndrome Proposed by ATP III
(199)

Feature At least 3 of the following 5 features:
Triglycerides ≥150 mg/dl
HDL cholesterol Males: ≤40 mg/dl
Females: ≤50 mg/dl
Abdominal obesity Males: waist circumference >102 cm
Females: >88 cm
Blood glucose Fasting glucose ≥110 mg/dl
(The ATP definition has since been updated to use a cutoff of
fasting glucose ≥100 mg/dl to reflect the current American
Diabetes Association definition of impaired fasting glucose)
Blood pressure Blood pressure ≥130/85 mm Hg(5)

161

Appendix B: Pediatric Definition of the Metabolic Syndrome Based on ATP III

At least 3 of the following 5 features:
Feature Cook et al (36)/
Ford et al 2005
(67)
Cruz et al (41) Weiss et al 2004
(218)
Triglycerides ≥110 mg/dL ≥90
th
% for age
and sex
≥95
th
% for age, sex,
and ethnicity
HDL cholesterol ≤40 mg/dL ≤10
th
% for age
and sex
≤5
th
% for age, sex,
and ethnicity
Abdominal obesity ≥90
th
% for age
and sex
≥90
th
% for age,
sex, and ethnicity
BMI-z ≥2
Blood glucose Impaired fasting
glucose
(≥110 mg/dL)
Impaired glucose
tolerance
(≥140<200
mg/dL)
Impaired glucose
tolerance
(≥140<200 mg/dL)
Blood pressure ≥90
th
% for age,
sex, and height
≥90
th
% for age,
sex, and height
≥90
th
% for age, sex,
and height

Abstract (if available)

Abstract PURPOSE: This dissertation examined the associations between physical activity, sedentary behavior, overweight, and the metabolic syndrome (MetS) in at-risk pediatric populations in the United States and in China. Study 1 identified the independent influences of physical activity, sedentary behavior, and other weight-related correlates on overweight status in Chinese youth. Study 2 explored the influences of physical activity and sedentary behavior on MetS in US minority youth. Study 3 compared the longitudinal trends of physical activity and sedentary behavior between youth with and without MetS in a sample of US minority female youth.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Objectively measured physical activity and related factors in minority youth

PDF

Motivation and the meanings of health behavior as factors associated with eating behavior in Latino youth

PDF

Influences of specific environmental domains on childhood obesity and related behaviors

PDF

The metabolic syndrome in overweight Latino youth: influence of dietary intake and associated risk for Type 2 diabetes

PDF

Effects of sugar and fiber consumption in minority adolescents and self-tracking as a potential dietary intervention tool

PDF

Psychosocial and behavioral ractors associated with emotional eating in adolescents

PDF

Competing risks: the role of the perceived consequences of refusing to share injection equipment among injection drug users

PDF

The acute relationship between affective states and physiological stress response, and the moderating role of moderate-to-vigorous physical activity

PDF

Effects of parent stress on weight-related parenting practices and child obesity risk

PDF

Effects of progestin-only long-acting reversible contraception on metabolic markers in obese women

PDF

Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California

PDF

Objectively measured physical activity and body fat distribution in overweight Hispanic and African American adolescents

PDF

Applications of mindfulness toward promoting psychosocial health in diverse communities

PDF

A network analysis of online and offline social influence processes in relation to adolescent smoking and alcohol use

PDF

The role of depression symptoms on social information processing and tobacco use among adolescents

PDF

Marginalization in acculturation is related to objectively measured physical activity in Latina adolecents

PDF

System dynamics of in-home family eating behavior: insights from intensive longitudinal data using Ecological Momentary Assessment and wearable sensors

PDF

Mixed methods investigation of user engagement with a smoking cessation app

PDF

Examining tobacco regulation opinions and policy acceptance among key opinion leaders and tobacco retailers in low socioeconomic status African American, Hispanic, and non-Hispanic White communities

PDF

Free-living and in-lab effects of sedentary time on cardiac autonomic nervous system function in youth with overweight and obesity

Asset Metadata

Creator Hsu, Ya-Wen (author)

Core Title The vicious cycle of inactivity, obesity, and metabolic health consequences in at-risk pediatric populations

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior)

Degree Conferral Date 2011-05

Publication Date 04/26/2012

Defense Date 03/09/2011

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Chinese youth,metabolic syndrome,minority youth,OAI-PMH Harvest,obesity,physical activity,sedentary behavior

Place Name China (countries), USA (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chou, Chih-Ping (committee chair), Spruijt-Metz, Donna (committee chair), Azen, Stanley Paul (committee member), Palinkas, Lawrence A. (committee member), Unger, Jennifer B. (committee member)

Creator Email yawenhsu@usc.edu,ywxxbirdy@yahoo.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3779

Unique identifier UC1159931

Identifier etd-Hsu-4455 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-468058 (legacy record id),usctheses-m3779 (legacy record id)

Legacy Identifier etd-Hsu-4455.pdf

Dmrecord 468058

Document Type Dissertation

Rights Hsu, Ya-Wen

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu