$549.5

.
. .... .5... L, $3.? .

«1-5... 3 ‘. k.
L'II‘ ‘ I

. .

x I: 5.!u.‘

gm
1.‘ ii.' "91‘.

§
7...... .5

K

.34..

1 .
zit tips-v:
.. $41. ‘33.}?
94:303.}...

.9 ill

 

, J... ‘nflh.
gggfi

 

 

 

W‘s ‘ LIBRARY
a Michigan State
" ”7 University

This is to certify that the
dissertation entitled

THE RELATIONSHIP OF BODY FAT DISTRIBUTION PATTERN
TO METABOLIC SYNDROME IN THE US AND TAIWAN

presented by
JIA—YAU DOONG

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Human Nutrition

 

 

411w

Major Professor’s Signature

W22—

Date

MSU is an Affirmative Action/Equal Opportunity Employer

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IPr0jIAcc&PrelelRC/DaIeDue.indd

 

THE RELATIONSHIP OF BODY FAT DISTRIBUTION PATTERN TO
METABOLIC SYNDROME IN THE US AND TAIWAN

By

JlA-YAU DOONG

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

Human Nutrition

2008

ABSTRACT

THE RELATIONSHIP OF BODY FAT DISTRIBUTION PATTERN TO
METABOLIC SYNDROME IN THE US AND TAIWAN

BY

JlA-YAU DOONG

The purposes of this study were: 1) to determine whether percent bodyfat
(%BF) added to the predictive power of waist circumference (WC) to assess
risks for modified metabolic syndrome (MMS, where WC was omitted as a
criteria) and/or each metabolic co-morbidity by gender; and 2) to examine
how the odds ratio (OR) for MMS and each co-morbidity differed by
distribution patterns of %BF in ovenlveight men and women with normal or
high WC in the United States and Taiwan. Two national survey datasets,
NHANES III from the US. and NAHSIT from] Taiwan, were used. A total of
960 males and 676 females non-Hispanic White Americans (BMI between 25
and 30 kg/mz) and 291 males and 312 females Taiwanese (BMI between 24
and 27 kg/mz) met these criteria. Percent BF was divided into 4 categories
based on the 50th percentile split for %BF and WC as normal vs high ( _>_102
cm for men, 88 cm for women in the US. and 90cm, 80 cm in Taiwan,
respectively). Odds Ratios (OR) equations were derived from logistic
regression models for MMS and metabolic co-morbidities [high triglycerides
(TO) 3150 mg/dL, low HDL (<40 mg/dl for male and <50 mg/dl for female),

high blood pressure (BP>85/135 mm Hg) and impaired fasting glucose

(IFG_>_100 mg/dL)]. The lower 50th percentile of %BF with normal WC in the
sample was used as the reference.

Results: In non-Hispanic whites, WC was the strongest positive predictor
for MMS and metabolic co-morbidities. WC was independent of gender,
except for low HDL and high blood pressure in men. The risk for MMS
increased with increasing WC, but not always with increasing BF, for people
with BMI 25<30. White males with high WC and %BF had the highest risk for
MMS (OR=2.1) and for high TG (OR= 1.8). White females with high WC but
normal %BF had the highest OR for MMS (OR= 2.2) and for impaired fasting
glucose (OR= 3.8).The relation of WC and %BF to MMS and co-morbidities
were weaker in the Taiwanese sample but similar to NHANES in that adding
%BF to WC did not increase the ability to predict MMS.

Conclusion: Percent BF provided no advantage over WC in assessing
obesity-related metabolic risks in the Non-Hispanic white US or in the
Taiwanese samples. The relationship of fat distribution patterns to the risk for
MMS differed somewhat by gender for both staples. Males with higher %BF
and high WC had the greatest risk for MMS, high T6. In women, %BF was
associated with decreased risk for high serum TG, low HDL, and high fasting

glucose and MMS syndrome when adjusted for WC.

Copyright by
JIA-YAU DOONG
2008

Acknowledgments

This dissertation could not have been written without the support and
friendship found at Michigan State and elsewhere. The love of my friends,
advisor and my committee members provided my inspiration and was my driving
force. I could not have come this far without the assistance of many individuals
and I want to express my deepest appreciation to them.

I would first and foremost like to express my deep gratitude to my
advisers, Dr. Sharon Hoerr for her guidance, understanding, patience, and most
importantly, his friendship during my graduate studies since 1990. Her care,
encouragement and support are more than word can express. I want to thank
each of my guidance committee members for their brilliant insight and for
providing numerous hours of advice and critiques: Dr. Lorraine Weatherspoon,
Dr. Beth Olson and Dr. James Pivarnik. Their insightful guidance is
indispensable to the accomplishment of this dissertation.

I am very grateful to all my colleagues at the Community Nutrition Lab.
The atmosphere has always been a perfect source of motivation. I would like to
thank some of my colleague and friends, Seung-Yeon Lee, Saori Obayashi, Sue
Gunnik,.Christy Mincy, Debra Keast, Julie Plasencia for their encouragement and
making these few years not only fun, but also very rewarding. Specially thanks to
Dr. Yi-Hsin Yang my statistics consultant in Taiwan, for her help to resolve the
statistical problems I encountered.

Finally, I would like to thank my dear mom, Mei-Chu Liu. She has always

believed in me and helped me reach my goals. Her support forged my desire to
achieve all that I could in life. I especially wish to express my love for Mei-Chen
who only knows the real price of this dissertation as we suffered and paid it

together. I thank for her endless love, patience, and understanding.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................ ix
LIST OF FIGURES .......................................................................................... xi
CHAPTER 1 INTRODUCTION ...................................................................... 1
Research purpose ........................................................................................ 9
Glossary ............................................................................................ 16
CHAPTER 2 REVIEW OF LITERATURE .................................................... 19
A. Definition of Metabolic Syndrome ................................................... 19
B. The etiological mechanism of metabolic syndrome ...................... 25
1. Clinical symptoms of metabolic syndrome .............................. 25
2. Insulin resistance and dyslipidemias ....................................... 28
3. Insulin resistance and hypertension ........................................ 32
C. Impact of total body fat, abdominal fat and visceral fat on
metabolic syndrome .............................................................. 34
1. Abdominal fat vs. peripheral fat in relation to the metabolic
syndrome .............................................................................. 35
2. The relation of adipose tissue derived proteins,
adipocytokines, to the metabolic syndrome ..................... 38
D. Ethnic differences in metabolic risk factors ................................... 43
1. The prevalence of the obesity related metabolic
risks in the US .................................................................. 43
2. The prevalence of the obesity related metabolic
risks in Asians ................................................................. 47
E. Life style factors (dietary factors, physical activity) related to
metabolic disorders. ....................................................................... 52
1. Weight reduction ....................................................................... 53
2. Relationships of dietary factors and metabolic syndrome ...55
a. Dietary Fat .................................................................... 55
b. Dietary carbohydrate .................................................. 57
0. Mediterranean diet pattern ......................................... 61
3. Relationship of exercise and metabolic syndrome ................ 62
F. Methods use to assess body fatness. ............................................. 67
1. Body mass index (BMI) ............................................................ 67
2. Waist Circumference ................................................................ 71
3. Percent body measured by bioelectrical impedance
analysis (BIA) .......................................................................... 75

vii

CHAPTER 3 METHOD .................................................................................. 80

Study design ..................................................................... 80
Subjects and Datasets ..................................................... 81
Statistical Analyses .......................................................... 87
CHAPTER 4 RESULTS ................................................................................. 92
CHAPTER 5 DISCUSSION ........................................................................... 113
CHAPTER 6 CONCLUSION ......................................................................... 125
A. Summary of Findings .................................................. 125
B. Strengths and Limitations ........................................... 127
C. Implications .................................................................. 129
D. Recommendations for Future Research .................... 130

APPENDICES
Appendix 1 Detailed laboratory procedures for metabolic risk and
anthropometric variables in NHANES III and NAHSIT ....... 132
Laboratory analyses data ........................................ 133
Anthropometric variables ........................................ 135
NAHSIT ...................................................................... 138
Covariates ................................................................. 138
BIBLIOGRAPHY ............................................................................................. 140

viii

Table 1

Table 2.1.

Table 2.2

Table 2.3

Table 2.4

Table 2.5

Table 3.1

Table 4.1

Table 4.2

Table 4.3

Table 4.4

Table 4.5

Table 4.6

Table 4.7

Table 4.8

Table 4.9

LIST OF TABLES

Criteria for diagnosis of Metabolic Syndrome by NCEP ATP

m‘, WHo2 and AAcrs3 .............................................................. 3
Criteria for diagnosis of Metabolic Syndrome by NCEP ATP
III1,WH02, ACE3 and IDF .......................................................... 24
Abnormalities associated with metabolic syndrome and
insulin resistance/compensatory hyperinsulinemia ............. 27
Proteins produced by adipose tissue, adipocytokines. ........ 42

Prevalence of metabolic syndrome in different ethnic groups

by different definitions ........................................................... 44
Treatment of Lifestyle Risk Factors for Long-Term
Prevention ................................................................................ 66

List of variables ......................................................................... 90

Descriptive characteristics and metabolic factors in the total
sample of overweight ............................................................ 93

Prevalence of metabolic risk factors in Non-Hispanic ............ 93

Pearson’s correlations of BMI, waist circumference and percent

body fat to the metabolic disorders ...................................... 96
Regression models using waist circumference and/or percent
body fat to predict the metabolic disorders ......................... 98
Prevalence of modified metabolic syndrome (MMS‘) and
metabolic related risks ........................................................... 99
Prevalence of modified metabolic syndrome and metabolic
related risks ............................................................................ 100
Odds ratio‘of metabolic related risks2 in low fat and normal
waist circumference (LFNWC) ............................................... 104
Descriptive characteristics and metabolic related risks in the
total sample of overweight (BMI 24-27) ................................ 105
Prevalence of metabolic risk factors in overweight ................ 106

ix

Table 4.10 Pearson’s correlations of BMI, waist circumference and percent
body fat to the metabolic disorders ...................................... 108

Table 4.11 Regression models using waist circumference and/or percent
body fat to predict metabolic syndrome ............................... 110

Table 4.12 Prevalence of modified metabolic syndrome and related
metabolic risk factors1 ............................................................ 112

Figure 1.1
Figure 1.2

Figure 1.3

Figure 2.1

Figure 2.2

Figure 2.3

Figure 4.1

LIST OF FIGURES

Conceptual Model for H14, H1.2, and H3.1 .............................. 11
Conceptual model for H24 ..................................................... 12

Four anthropometric categories based on two anthropometric
measures: .......................................................................... 14

Mechanisms Relating Insulin Resistance and Dyslipidemia

Modified from Lipid Online Slide Library ........................ 29
Role of CETP in HDL Metabolism. Modified from Lipid Online

Slide Library ....................................................................... 31
Patterns of metabolic syndrome in Taiwanese and in US

non-Hispanic whites ......................................................... 49
Prevalence of modified metabolic syndrome (MMS) ........... 102

xi

CHAPTER 1
INTRODUCTION

INTRODUCTION

Results from the 2003 to 2004 National Health and Nutrition Examination
Survey indicate that an estimated 66.3% of US. adults are either overweight or
obese (Ogden, Carroll et al., 2006). Obese individuals are at high risk of
hypercholesterolemia, hyperglycemia, hyperinsulinemia, type 2 diabetes,
cardiovascular disease (CVD), hypertension, stroke, and some cancers. In 2003
alone, the medical costs attributable to obesity and related diseases in the USA
were estimated at about $75 billion (Finkelstein, Fiebelkorn et al., 2004). The
increasing prevalences of obesity have been reported in Asian countries as well,
such as China, Taiwan, Japan, and Korea (Gu, Reynolds et al., 2005; Kim, Ahn et
al., 2005). As the prevalence of obesity increases worldwide along with
concomitant risks for chronic, metabolic diseases (WHO, 2000; Jacob et al.,
2000), health professionals search for improved ways to target those at greatest
risk for the diseases with high medical costs in these diverse populations.

The rapidly increasing prevalence of obesity and related diseases are
associated with a cluster of risk factors called metabolic syndrome (MS) or
Syndrome X. The concept of MS was introduced in the 1980’s as the
co-occurrence of cardiovascular risk factors, such as abdominal obesity,
atherogenic dyslipidemia (elevated triglyceride, small LDL particles, low HDL
cholesterol), raised blood pressure, insulin resistance (with or without glucose

intolerance), and prothrombotic and proinflammatory states (Alberti and Zimmet,

1998). Although MS is well known, the criteria for clinical diagnosis of MS varies
around the world with at least three recognized professional and medical
organizations reporting slightly different clinical criteria for diagnosis (Table 1).
These diagnostic differences confound a clear diagnosis, research on etiology,

and cross-cultural comparisons in risk trends.

800m ...m «0 cm>mmm .EocvaflflmgoEOoccm _mo_:__0 *0 cozfloowmxx cmotoE< 9:. .m
3mg. .umEPEN new Eon—.1 ”mam? .EngVEOfiNEmmLO £30... Oto>> .N
Goon ...m E cmEmma $9.20 uroom ..amozv tone. .: _mcma «cm—ENE... ~_:U< 925.69.“. cozmoauw 6.060.050 _mcozmz 9:. ._.

 

 

 

 

 

 

 

 

 

0&9ch
890m.— xm: :0 women 2.8905 Co 53:86 .8 22053 Emmoa «€3.50
E0562... _oo_:__o co mucoamu m_mocmm_o 220$ xm: 550 N mag 965335 9m 2293 xmc 85: on“ Co m Co n came—5&0
com 98:92 .
258:. 5358a . 0,35
05555» ago o=m>o>_od . 0mm. 0sz oc_:=m29:_E:n_m
..EmE 03A 35.65 383638 SoI-N . mo EEB: owM 429:
429: @963 $833 95mm”. . 2m. 5:998 5:593 tact: . co m mmooam 95mm“. 0
m: EE mmeX Samoa 083 599mm. . mIEEom \ 0.3M 9:305 32m . mIEE mmmxom FM 93me 5005 o
.65.: omv coEo>> 5.595 E .655 mmv ..EmE omv 552$
.555 ovv :22 .55 5 .555 mmv 535 ovv 8.2
6.38.05 .5 I . 6.93.98 .5: . 6.33.05 .5: o
555 02M 82.8.2.9: . 5&5 8m 8.58sz . 555 02M 828sz .
. E mmkv Eo wmk coEo;
. . 5:83 5 .3 0A E El E0 work :9:
cos c_ m on 0:9 95.58; mo:m.£E=o...o
N555. max .55 588585590 ._ .225. N59. 8A .55 . 5.5; mm .588 55503.9. . 2203 me
mcozficoo
28829.5 .o_Eoc__:m:_..ma>c
Love: 5259395 .695 c2538
9.59962 Lo“. 055.6 $826. 9:
323 9.3% $836 .329: or Fvv
m_o>m_ 383m 9sz .850: 55>
5323 Leon“ «6.5 8:222 382m 8538. o 3.33 .6on «M:
an BEE 92.6: 302.3 956m": 382m 95me 8535. . we BEE 85M 630:3 953E
52"..nt N 0%... .
6532.8
0:02 9: Co F 3 35:52 60:358. 5.3:. 0502 33569.95
mo<< oz; 5 a: -
me<< 25 N022, ..___ nEq $02 3 852.2% 2.8%.). so 58:85 .5 5.25 s 2.5

 

The three definitions shown in Table 1 are similar in that all use four
metabolic risk criteria for elevated triglycerides (TG), low high density lipoproteins
(HDL), high blood pressure or hypertension, and impaired fasting glucose.
However, the anthropometric criteria differ among all three. The National
Cholesterol Education Program’s Adult Treatment Panel Ill Report (NCEP, 2001)
focused on waist circumference (WC), the World Health organization (WHO)
focused on body mass index (BMI, where BMI=wt in kg/ht in m2) and/or waist to
hip ratio (WHR), and the American Association of Clinical Endocrinologists
(AACE) used BMI. Even the BMI cut-offs for risk differ among these three sets of
criteria. Such differences suggest a need for more research to elucidate the
relationship of anthropometric variables to metabolic complications in order to
generate useful anthropometric indicators for specific target populations.

In people of predominantly European descent, the risk for developing MS
increases strikingly at a BM|325 (Ford, Giles et al., 2002). In a recent study,
Park et al. reported that the prevalence of MS in nonnal-weight (BMI=18.5—24.9),
overweight (25.0—29.9), and obese ( 330) men were 4.8%, 22.8%, and
60.2%,respectively. Prevalence rates were similar for US. women. Studies
have shown, however, that individuals with greater degrees of central adiposity
develop this syndrome more frequently than those with peripheral body fat
distribution (Kissebah, Vydelingum et al., 1996). In all populations, current
studies showed that both increased total and abdominal adiposity are risks for
cardiovascular disease. The total amount of body fat and the amount of

abdominal fat are highly inter-correlated (Han TS, 1997), but the relative

contribution of each individual compartment to metabolic risks has not yet been
established, especially for different populations. It is evident that MS relates to the
degree of obesity, but it also appears to be critically dependent on body fat
distribution.

Anthropometric Indices (Al), such as BMI, waist circumference (WC),
skinfold measurements, and percent body fat (%BF) predicted by Bioelectrical
Impedance Analysis(BlA), are commonly used for screening obesity. BMI and
WC are the most widely used indicators of overweight and obesity and are useful
surrogate indicators of adiposity because they are simple, associated with body
adiposity, and relatively unaffected by body height (Heymsfield, Allison et al.,
1995). Many researchers have investigated whether BMI is better than WC or
vice versa for predicting risk for morbidities and mortality (Janssen, Katzmarzyk et
al., 2002; Katzmarzyk, Church et al., 2005). Since BMI and WC are highly
correlated and depend upon the age, gender, and racial structure of the study
sample, one or the other will perform better, and the magnitudes of their
association with related diseases often differ by a small amount. Janssen et al.
analyzed national health survey data from the US and reported that WC, not BMI,
best explained the obesity-related health risk (Janssen, Katzmarzyk et al., 2004).
BMI is an imperfect indicator for M8 for the following reasons. First, BMI cannot
distinguish between fat and fat-free mass, i.e., a person may be ovenNeight but
not overfat, or normal weight, yet overfat. Secondly, BMI cannot distinguish
truncal from peripheral body fatness (Gallagher, Visser et al., 1996). Another

limitation for BMI is that the universal cutoff points for BMI do not apply equally

well for all racial groups (Deurenberg et al., 1998). Obesity defined by WHO
criteria is BMI 330.0, but this cut-point results in a MS prevalence of only 2-3% of
the Japanese population, in contrast to 10-20% in Europe and the US (Shiwaku,
Nogi et al., 2005).

The advantages of WC include its high correlation with BMI (Lean, Han et al.,
1995) and with total body fat (Lean, Han et al., 1996), and its association with
cardiovascular disease risk factors independent of BMI (Janssen, Heymsfield et
al., 2002). Studies indicated that health risk increases in a graded fashion when
moving from normal weight to obese BMI categories and that within each BMI
category, individuals with high WC values are at a greater health risk than those
with normal WC values (NIH, 2000). Although WC is a criterion for the diagnosis
of M8 by ATPIII, the limitation of using WC to indicate adiposity is that it cannot
distinguish between the content of abdominal subcutaneous fat and
intra-abdominal visceral fat. It is the latter which is hypothesized to pose the
greatest risk for MS because it lies in the etiological pathway for the disease.
Also, WC only indicates primarily regional adiposity and not total body adiposity.
In addition, there is evidence that Asians should use smaller cut-points for waist
circumference than those for people of European descent in predicting increased
metabolic risk (Yagalla, Hoerr et al., 1996; Wildman, Gu et al., 2004).

Bioelectrical Impedance Analysis (BIA) holds promise as another
screening indicator for body composition, because it is inexpensive, easy to use,
free of observer bias and relatively precise when correctly used (Lukaski,

Bolonchuk et al., 1986; Roubenoff and Kehayias, 1991). Studies have shown

that percent body fat measured by BIA had statistically significant linear
relationships with body fat measured by dual-energy x-ray absorptiometry (DEXA)
(Bolanowski and Nilsson, 2001). However, BIA loses its accuracy in severely
obese persons and can be of limited usefulness for tracking changes in total body
fat in persons losing weight (NIH, 1998). Theoretically, BIA estimates total
body adiposity rather than body mass and should provide more precise
information about obesity as related to metabolic disease. Whether the use of
BIA for adiposity is better than BMI when combined with WC to predict
obesity-related metabolic risk is still unclear. As more than 60% of the US
population is either overweight or obese, it is difficult for health professionals to
target those at greatest risk for CVD in this group. Apparently, no single
anthropometric indicator can precisely target those with highest risk. A
combination of two or more screening indices should be able to help health
professionals target those in the greatest risk for MS and clarify whether the
regional or total adiposity contribute most to the etiology of MS.

As indicated earlier, racial and ethnicity variation in metabolic risks
complicate the determination of body adiposity and its distribution to metabolic
risks. In the US, Ford et al. reported age-adjusted prevalence of MS as 23.7%,
with the ethnic specific prevalence highest in Mexican Americans and lowest in
blacks. The prevalence of MS of Asian Americans in the US is unknown due to
insufficient sample size in both the National Health and Nutrition Examination
Survey (NHANES) III and NHANES 1999-2000. Researchers have recently

argued that it is precisely because people of Asian descent appear to have

greater central body fat deposition, that the BMI cutpoints for these populations
should be >23 rather than >25 for defining overweight (Ota, Takamura et al.,
2002; Kim, Ahn et al., 2005; Shiwaku, Nogi et al., 2005). One study also showed
(Deurenberg-Yap, Chew et al., 2002) that for the same percent body fat, the BMI
of Singaporean Chinese, Malays, and Asian Indians are about 3 kglm2 lower than
that of Caucasians. In Asian ethnic groups, specifically Chinese, Korean, Indian,
Paskistani, and Bangladeshi, the risk of hypertension, diabetes and
hypercholesterolemia starts to increase rapidly at levels of BMI and WC well
within the acceptable range of BMI or WC for Europeans (Deurenberg, Yap et al.,
1998; Deurenberg-Yap, Chew et al., 2002; Kim, Suh et al., 2004; Pan, Flegal et
al., 2004). Such complications appear to be associated also with both
abdominal fat distribution and total body fat adiposity. However, the relative
effect of total fat and of abdominal fat to MS and to specific metabolic disorders
within various populations still needs to be clarified. Comparative studies
between populations of European and Asian descent can help to delineate the
differences in amount and distribution of body fat between these groups. Such
comparisons are important because the WHO definitions for overweight and
obesity were established using data primarily from those of European descent
(WHO, 1995), yet people of Asian descent appear to exhibit metabolic risks at
lower anthropometric cutpoints and as a population, are increasing rapidly around

the globe.

Research purpose

There is an increasing threat of obesity as a global epidemic. Current
evidence demonstrates a high correlation between abdominal obesity and
metabolic syndrome, as well as elevated percent body fat and cardiovascular
risks at relatively low levels of BMI. Such trends and associations require that
appropriate measures be undertaken to prevent and manage obesity
complications among diverse population groups.
The purpose of this study was to identify the anthropometric indicators, waist

circumference and percent body fat, with the best predictive power to target those

at highest risk of metabolic syndrome related disorders1 (MSRD) in diverse

populations. The literature reviewed suggests that different patterns of body fat

distribution have different relative risks associated with metabolic syndrome2 and
MSRD. The study sample will be narrowed to people who are only overweight,
but not obese (BMI 25-299). There were several reasons to choose this
narrowed range of BMI. 1) For public health screening purposes, it is better to
target those in overweight rather than those already obese. 2) The wide variation
in body fat distribution that exists at a limited range of BMI can help determine the
effect of total fat and abdominal fat to metabolic risks more precisely than by using

the entire range of BMl’s. 3) At BMI>30, there are few people with either normal

 

1 Includes elevated triglycerides (_>_150 mg/dl); low HDL cholesterol ( 540 mg/dl in
men, _<_50 mg/dl in women); high blood pressure (3 130/85 mmHg); and impaired
fasting glucose (3110 mg/dl).

For the purposes of this study, MS will be defined as 2 or more of 4 risk
indicators: TG 3150 mg/dl, HDL <40 in men and <50 in women, blood pressure

3 130/85 mmHg, and fasting glucose3110 mg/dl as defined by ATP lll (NCEP,
2001), excluding the WC used here as a predictor.

waist circumference or body fat percentile categories. 4) Finally, studies have
shown that south Asians, Taiwanese, and Japanese have a much higher
prevalence of MS compared to the US population at this lower level of BMI and
the prevalence of BMI>30 is less than 4% in these countries (Deurenberg-Yap,
Chew et al., 2002; Lin, Yen et al., 2003; Kim, Suh et al., 2004).lt is expected that
the prevalence of each body fat distribution pattern will differ by race/ethnicity and
possibly by gender, resulting in different prevalences of metabolic risks within the
same BMI interval.

The specific aims and accompanying hypotheses for this study follow for
individuals within the defined BMI interval of overweight, 25-299.
Aim 1. To determine whether percent body fat assessed by Bioelectrical
Impedance Analysis adds to the predictive power of waist circumference in risk for
modified metabolic syndrome and/or each metabolic syndrome related disorder
(elevated triglycerides, low HDL cholesterol, high blood pressure, impaired fasting
blood glucose) in non-Hispanic white, overweight (BMI 25-299) men and women

(NHANES III). Figure 1.1 illustrates the conceptual model for this aim.

H1.1: Both waist circumference and percent body fat predict risk for modified

metabolic syndrome and/or each metabolic syndrome related disorder in

overweight non-Hispanic white men and women.

H1.2: Percent body fat adds to the predictive power of waist circumference to

detect risk for modified metabolic syndrome and/or each metabolic syndrome

related disorder in overweight non-Hispanic white men and women (NHANES III).

10

 

Non-Hispanic White men
and women BMI 25-2939

WC % Bodyfat

 

 

 

 

 

 

Confounding variables:
Age, smoking & drinking
status, Physicalactivitics, 37>
Dictaryfactors

 

 

 

 

 

 

‘\

r" .-t',...',.l,
L)Ԥ,"_':ll}_!ltlt.'llllail
"l C): t" D L

“x. l a

Modified Metabolic Syndrome1

 

 

 

 

 

lz‘r‘lg‘iAau‘j'tlé fwstmg leu

 

 

 

 

 

 

 

 

 

 

1 Based on NCEP ATPlll Criteria but not including high WC, i.e. having 2 or more of following

abnormalities: Triglycerides 3150 mg/dL; HDL: Men <40 mg/dL Women <50 mg/dL; Blood
pressure 3130/385 mmHg; Fasting glucose 3100 mg/dL.

Figure 1.1 Conceptual Model for H1.1, H1.2, and H34

Aim 2. To determine whether body fat distribution modifies the effect of percent
body fat and waist circumference on modified metabolic syndrome and each

metabolic syndrome related disorder differ in overweight non—Hispanic white men.

H24: For each percent fat quartile, those with high waist circumference will have

higher risk than those with normal waist circumference for modified

metabolic syndrome and each metabolic syndrome- related disorder in

11

overweight non-Hispanic white men and women Figure 1.2. illustrates the

conceptual model for H21, where the ovenlveight sample is categorized into

one of eight groups by percent body fat and waist circumference.

 

 

 

 

 

Total ad'p°5'ty Normal wc High wcl
(% Fat)
25th %tile 0‘1 01
50'“ %tile Q2 Q2
Q3 Q3
75‘” %tile
Q4 04

 

 

 

 

 

Confounding variables:
Age, smoking & drinking 3:’\
status, Physical activities,
Dictaryfactors

 

 

 

 

 

L‘\

 

 

%

 

 

 

 

 

 

 

 

lm p .21 i {(3 :53 fa st i n g G l u

i1 ./

 

 

 

Modified MetabolicSyndromeZ i

l

1WC: Waist Circumference, High: >102 cm in men, > 88 cm in women.

2 Based on NCEP ATPllI Criteria but not including high WC, i.e. having 2 or more of following
abnormalities: Triglycerides 3150 mg/dL; HDL: Men <40 mg/dL Women <50 mg/dL; Blood
pressure 3130/385 mmHg; Fasting glucose 3100 mg/dL.

Figure 1.2 Conceptual model for H21

H22 : Those with high body fat (>50th %tile), and high waist circumference

(HFHW) will have higher risk for modified metabolic syndrome and each

related disorder compared to those with low body fat (550th %tile) and high

waist circumference (LFHW), high body fat (>50th %tile) and normal waist

12

circumference (H FNW) and low body fat (550th %tile) and normal waist

circumference (LFNW) in ovenrveight non-Hispanic white men and women.

Figure 2.2. illustrates the conceptual model for H22, where the overweight

sample is categorized into one of four groups by WC (normal vs. high) and

median split for body fat.

13

WC1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Normal High
%Fat LFNW LFHW
h . Low %Fat Low %Fat
<50t %tlle NormalWC NormalWC
HFNW LHFHW
> 50'“ %tile Low %Fat Low %Fat
Confounding variables: Normal WC Normal WC
Age, smoking & drinking fl
status, Physical activity, [__,\
Dietary factors ”M":
A
Dyslipltiflil' ‘1 -, ”3,
EC), 2‘“: UL ‘ if! 'I.f>.{i l l ("(1 if"! blillg (.51 Ll

 

 

 

 

 

 

\ ll /

[Modified Metabolic Syndrome2

 

 

1 WC: Waist Circumference, High: >102 cm in men, > 88 cm in women.

2 Based on NCEP ATPllI Criteria but not including high WC, i.e. having 2 or more of following
abnormalities: Triglycerides 3150 mg/dL; HDL: Men <40 mg/dL Women <50 mg/dL; Blood
pressure 3130/385 mmHg; Fasting glucose 3100 mg/dL.

Figure 1.3 Four anthropometric categories based on two anthropometric
measures: waist circumference (WC) and percent body fat (% fat) in
ovenlveight people (BMI=25-29.9). The quartiles of adiposity are

specific to the non-Hispanic white (H22) or Taiwanese sample (H32)
by gender and for those who are overweight.

Aim 3. To determine the predictive power of waist circumference and percent
body fat to detect the risk for modified metabolic syndrome and/or each metabolic
syndrome related disorder in overweight Taiwanese men and women using data

from the Nutrition and Health Survey in Taiwan 1993-1996 (NAHSIT).

14

H3.1 : Both percent body fat and waist circumference predict risk for modified

metabolic syndrome and/or each metabolic syndrome related disorder
(elevated triglycerides, low HDL cholesterol, high blood pressure, and
impaired fasting blood glucose) in oveniveight Taiwanese men and women

(See previous Figure 1.1).

H3.2 Those with high body fat (>50th %tile) and high waist circumference

(HFHW) will have higher risk for modified metabolic syndrome and each
metabolic syndrome related disorder compared to those with low body fat
(55oth %tile) and high waist circumference (LFHW), high body fat (>50th
%tile) and normal waist circumference (H FNW) and those with both low body
fat (5 50th %tile) and normal waist circumference (LFNW) in overweight

Taiwanese men and women.

There are two expected outcomes of this study. First, findings should result
in evidence for using two to three anthropometric indicators together (instead of
just BMI or waist circumference) to screen for metabolic risks in diverse
populations. This is important because preventing obesity is easier and less
costly to manage than treating massive obesity or metabolic disorders
(Finkelstein, 2001). Secondly, the relative contribution of total fat vs. abdominal
fat to metabolic syndrome risk disorders within various populations should be
clarified by the study design and cross population comparisons. This outcome
should contribute to the understanding of how the distribution of adipose tissue

relates to the etiOlogy of metabolic syndrome in more than one race.

15

Glossary

Metabolic syndrome: defined as 3 or more of the following: waist circumference
3 102 cm in males, _>_ 88 cm in females, triacylglycerol concentration 3150 mgldl,
HDL cholesterol concentration 540 mgldl in men or _<_50 mgldl in women, blood
pressure 3 130/85 mmHg , and fasting glucose concentration _>_100 mgldL. In this
study, the diagnosis of the metabolic syndrome will not include a high waist
circumference.

Modified Metabolic syndrome (MMS): defined as two or more of the following:
triacylglycerol concentration 3150 mgldl, HDL cholesterol concentration 540
mgldl in men or 550 mgldl in women, blood pressure _>_ 130/85 mmHg, and fasting
glucose concentration 3100 mgldL. In this study, the diagnosis of the metabolic
syndrome will not include a high waist circumference.

Metabolic syndrome related disorders (MSRD): include elevated triglycerides
(3150 mgldl); low HDL cholesterol (<40 mgldl in men, 550 mgldl in women); high
blood pressure (_>_ 130/85 mmHg); and impaired fasting glucose (3100 mgldl).
High blood pressure: 3 130/85 mmHg. Individuals reporting a history of
hypertension and current blood pressure medication use were defined as having
high blood pressure, regardless of measured blood pressure values.

Impaired fasting glucose: fasting plasma glucose concentration _>_ 100 mgldl.
Individuals also met this criterion if they had a doctor diagnosis of diabetes or

were using hypoglycemic medication.

16

Total body adiposity: Expressed by percent body fat (%BF) from bio impedance
analysis (BIA).

Overweight: Having a BMI=25-29.9, in US non-Hispanic men and women;
Having a BMI=24-26.9 in Taiwanese men and women.

Normal waist circumference (WC): Having a WC <102 cm in US men and a WC
< 88 cm in US women, and having a WC < 90 cm in Taiwanese men and a WC <
80 cm in Taiwanese women.

Education level: In NHANES III education is divided into three categories: _<_8
years, 9 to 12 years, and >12 years of education. In NAHSIT, education was
divided into three categories: 59 years, 10 to 12 years, and more than 12 years of
education. The difference is due to a slight difference between the two
educational systems. In Taiwan, middle school goes from grades 7-9 and high
school goes from grades10 -12.

Smoking status: Subjects were considered cunent smokers if they smoked
cigarettes, cigars, or tobacco at the time of the interview; previous smokers if they
were not current smokers, but had smoked 3100 cigarettes in their entire life; and
nonsmokers if they had smoked less than these amounts.

Alcohol consumption: Subjects were non-drinkers if they drank no beer, wine,
or liquor in the past month, moderate drinkers if they drank 1-15 drinks in the past
month, or heavy drinkers if they drank >15 drinks in the past month.

Physical activity: This was based on the subject’s self- reported frequency of
common leisure physical activities within the past month. Activities in NHANES III

included walking a mile without stopping, jogging or running, swimming, regular

17

dancing, aerobic exercise or aerobic dancing, riding a bicycle, calisthenics,
garden or yard work, and weight lifting. The activities in NAHSIT included biking,
ball games, gymnastics/punching boxing, swimming, aerobic dancing/country
dancing, mountain climbing, jogging, walking, gardening, housework. Those
people reporting no activity within a weekly period, less than once per week, 2-5
activities/week, or >5 activities/week were divided into ‘physically inactive’, ‘light’,
‘moderate’, and ‘physically active’ categories, respectively.

Dietary factors: Both surveys assessed dietary intake from one 24 hour dietary
recall. The two dietary variables extracted from both for this study were percent of

energy from carbohydrate and percent of energy from fat.

18

CHAPTER TWO

REVIEW OF LITERATURE

The topics covered in this chapter relate to the rationale, significance and
methods for this study. The literature reviewed was limited to primarily that
published within the last 10 years. Three sections organize this review. First is
an in-depth review of metabolic syndrome including the definition, etiology, and
associated health complications. Next is a discussion regarding the amount and
distribution of body fat as relating to metabolic syndrome and metabolic related
disorders in different populations and ethnic groups. Finally, a review of selected
anthropometric indices elucidates the strengths and limitations of using different
measures and indices to assess the relation of body adiposity to metabolic
syndrome.

A. Definition of Metabolic Syndrome

Worldwide, the rapidly increasing prevalence of obesity and related diseases
has been associated with a cluster of risk factors called metabolic syndrome
(MS), earlier Syndrome X. People diagnosed with this syndrome have an
increased risk to develop Type 2 diabetes and/or cardiovascular diseases.
Metabolic syndrome is closely linked to a generalized metabolic disorder called
insulin resistance in which the normal actions of insulin are impaired (Reaven,
1988). Excess body fat, especially abdominal adiposity, and physical inactivity
promote the development of insulin resistance, as well as a genetic predisposition

for some individuals (Reaven, 1988).

19

Over the years, the number of metabolic disturbances associated with
metabolic syndrome has increased (Alberti, 1998). Risk factors for cardiovascular
disease (CVD) as well as chronic kidney disease are now included in some
definitions (Alberti and Zimmet, 1998; Reisin and Alpert, 2005). Currently, there
are three different sets of criteria for diagnosing MS, and this section will review
the similarities and differences among those that various organizations now
accept (Table 2.1).

The National Cholesterol Education Program (NCEP) Adult Treatment Panel
III (ATP Ill) diagnostic criteria for metabolic syndrome are currently recommended
and widely used, with minor modifications (NCEP, 2002). The ATP III definition
places treatment of metabolic syndrome as a secondary target of risk-reduction
therapy, after the primary target of LDL-cholesterol. NCEP ATP Ill proposed an
operational definition of MS to facilitate both diagnosis and preventative
interventions. Within this definition, a person must have at least three of the five
criteria for a diagnosis. The criteria include: 1) abdominal obesity, as determined
by a waist circumference of greater than 40 inches for men and greater than 35
inches for women; 2) triglyceride level of 150 mg/dL or greater; 3) HDL cholesterol
less than 40 mg/dL for men and less than 50 mg/dL for women; 4) blood pressure
of 130/85 mmHg or greater; 5) and a fasting glucose level of 110 mg/dL or
greater. In 2005, NCEP ATPIIl updated the definition, thus further lowering the
fasting glucose level to 100 mgldl (Grundy, Cleeman et al., 2005). Because
glucose tolerance testing is included within the ATP lll MS diagnosis criteria,

insulin concentration measurements and parameters for proinflammatory states

20

are not required. This set of criteria allows for ease of use in clinical practice and
requires only readily available clinical data.

A consultation group from the World Health Organization (WHO) established
the first international definition of metabolic syndrome in 1999 (WHO, 1999), and
recognized CVD as the primary outcome of the metabolic syndrome. The WHO
also viewed insulin resistance as a required component for diagnosis defined as
one of the four following conditions: 1)type 2 diabetes; 2)impaired fasting glucose
(IFG); 3)impaired glucose tolerance (IGT); or 4) (for those with normal fasting
glucose values) 3110 mgldL, a glucose uptake in the lowest quartile for the
background population under hyperinsulinemic, or euglycemic conditions. In
addition to insulin resistance, at least two additional of the following five risk

factors are necessary for a WHO diagnosis of metabolic syndrome: 1) BMI >30

kg/m2 and/or waist to hip ratio >0.9 inches in men and > 0.85 inches in women; 2)

TG >150 mg/dL; 3) HDL cholesterol <35 mgldL in men and <39 mgldL in women;
4) BP _>_ 140/90 mmHg; and 5) urinary albumin excretion rate 3 20 uglmin OR
albumin: creatinine ratio 3 30 mg/g. The WHO MS diagnostic criteria use a
higher blood pressure cut point than does the criteria used by the ATP Ill.
Recently, the International Diabetes Federation (IDF) proposed a second
international definition of MS that requires the presence of central obesity with
specific ethnic cut-points, in addition to two or more of the following factors: 1)
serum triglycerides 3150 mgldl and/or treatment for hypertriglyceridemia; 2) a low
serum HDL level of 550 mgldl for women and 540 mgldl for men and/or treatment

for this abnormality; 3) blood pressure elevated to _>_130/85 mmHg or previously

21

diagnosed hypertension; and 4) fasting glucose levels of _>_100 mgldl or previously
diagnosed type 2 diabetes mellitus (International Diabetes Federation, 2005). The
IDF appears to acknowledge in their MS definition that central obesity and insulin
resistance are important causative factors. The IDF definition builds upon the
NCEP criteria, but differs in two key aspects. First, the IDF definition has
lowered the threshold for waist circumference from 102 to 94 cm for Europid men
and from 88 cm to 80 cm for Europid women. Secondly, waist circumference is a
required component of MS under the IDF criteria, rather than an optional
component as it with the ATPIII definition by the NCEP.

These three definitions of MS have many similarities yet important
differences. Each recognizes the existence of a clinical entity with multiple risk
factors for CVD (Reisin and Alpert, 2005) including central obesity, high plasma
triglyceride (TG) levels, high-density lipoprotein cholesterol (HDL-C)
concentration and essential hypertension with some degree of glucose
intolerance. The definitions differ both in the diagnostic criteria and in their
usefulness to interpret etiology in cross-cultural comparisons for risk trends.

The NCEP ATP Ill, WHO and IDF definitions, include patients with diabetes
in the metabolic-syndrome population thus limiting the ability to reserve the
diagnosis of MS only for patients classified as pre-diabetic and at risk for
becoming diabetic (Ninomiya, L'ltalien et al., 2004). The WHO is the sole
organization with a recommendation to include microalbuminuria in its definition of
MS. Incorporating microalbuminuria within the criteria links metabolic syndrome

with the risk of developing chronic kidney disease and has been determined to be

22

valid (Chen, Wildman et al., 2004). All three definitions are in consensus

regarding central obesity as the key symptom for MS (Reisin and Alpert, 2005).
The anthropometric criteria, however, differ among all these definitions. The

NCEP ATP I" Report (NCEP, 2001) and IDF both focus on waist circumference

(WC), but IDF proposes stricter criteria including ethnic group specificity. The

WHO focuses on Body Mass Index (BMI = wt in kg/ht in m2) and/or waist-to-hip
ratio (VVHR). These differences in anthropometric criteria suggest the need for
additional research to elucidate the relationship of anthropometric variables to

metabolic complications, in order to generate useful anthropometric indicators for

specific target populations.

23

.moom 5000.000“. 00.0005 .000..0E0..._. 0000.000“. 00.0005 000.050.... .0
.000. ..0EE_N 0:0 E32 .000. ..tcescefiscmeo 500... 2.02. .0
.080 ..0 .0 $5020 .220 :80 .0002. :80. =. .20.. 205.00.. .32 0.50.00... 8.8300 0.0.0220 .2202 0... ..

 

 

0.0.00. ..0.. 080.000 2.30.0... .0 0.0§0.0 .0. ...0.0...30 .0000... 0.0 0.0.00. 000......
0.... .0 03. >00 03.0 3.0000 .0....00 0.0.00. .00 .050 N 03.0 0..0.300.0.n. 0.0.. 00.0.. 0... .0 m .0 0 00000005
00.20... 0 cc... 5.; 0008000
203030... .2000 30.9.. 90... 00M 20. 0...:..00.0.:.E30_0 «O 30.0... co .M
co. 00003.0 0.000... 05.00... . 0.....33 omM 0.0. ..0..0.0x0 0.0.30.0 E055 . 00003.0 05.00“. c
01...... 00.08%
.0... E... 0060.0 0.3000... 0005 . mIEEom . ov .M 0.3000... 0005 . 0.3000... 0005 c
30.0... on v ...0...0>> 30.9.. mmv ...0E0>> 30.9.. omv ...0Eo>>
.20... o... v .05. .20... 00v .02 .20... ovv ”cc...
..0.0.00_0..0 .0: . 60.00.05 .6... . 60.00.05 .51 c
i .20... 0.2m
30.0... om... A ..0>0. 0... c 30.0... om FA 00000020.... . 00000029.... 0
. . A... 00A. E0 00A :0Eo>>
00 ox .0ch 2. 2X. so No? :05.
0.? E05. 00:0.0....30..0 .002.

2.0. n..... n.0.0>> .2000 ~59. 8A .55 c 00 5.0000 .0...E00..< 0 0.0.00. .10.”.

 

0:00.008 20.00290

.0....0..._30....0n_>.. .0003 ..0..0m..00>...
00:: 02.0.3000 00393000 .0. 0....030
.0032 0... 2.0.00 9.0.03 00003.3 .303...

 

 

 

 

 

 

0030.0 0......0 .c0.0...0 .0. 0.0.8.30 o. .v. 0.00. 00003.0 35.00. .0....0: 55> . 050.00
0....0000 5.2. 00:0.0.E30..0 0000.06. 00003.0 02.00.... . .800. «0.: 00 403:.
.0.0>> .3 00.3000... 3.0000 .0....0o 00003.0 05.00. 00.00.... . 03M 00003.0 05.00...
00.200 N 0.... .
”00.2.2.0.
b.0000 .0....0o 0... .0 0:0 .3 00......00. .00..0.0.00. 5.30... 0:02 3.0.3.0205
00.0. NO..=.> LE a._.<

 

mun: 0:0 .NOI>> .2: n...< mmoz >9 mEo.0c>w £500.05. .0 0.09310... .0. 0:0...0 ...N 030..

24

B. The etiological mechanism of metabolic syndrome

A standardized definition of MS is difficult for a clinical entity that is
characterized by a complex pathophysiology, covered briefly in this section. By
understanding the metabolic evidence for the various criteria in different
definitions of the metabolic syndrome, researchers can design valid studies to
detect the risk factors related to metabolic syndrome and to avoid
misinterpretation of results from existing and future studies, especially in
comparisons across cultures.
1. Clinical symptoms of metabolic syndrome

Table 2.2 shows the spectrum of clinical abnormalities associated with the
metabolic syndrome. Abdominal obesity, impaired glucose tolerance,
dyslipidemia, and elevated blood pressure are commonly found in patients with
metabolic syndrome. Evidence is accumulating that insulin resistance is the
underlying factor that links atherosclerosis to the metabolic syndrome (NCEP,
2002; Reaven, 2005). Metabolic syndrome or insulin resistance syndrome is not a
disease of itself, but a physiological abnormality that increases the likelihood that
one or more of the abnormalities listed in Table 2.2 will be present (Reaven,
2005). Dyslipidemia is a hallmark of the metabolic syndrome and characterized
by elevated triglycerides (TG) and low levels of HDL cholesterol. Plasma LDL
cholesterol (LDL-C) levels are often normal in patients with the metabolic
syndrome. A common finding is that LDL particles are smaller and denser than
normal and associated with increased cardiovascular risk (Reilly and Rader,

2003). Additionally, disturbances such as microalbuminuria (WHO, 1997; Groop

25

and Orho—Melander, 2001) and hyperuricemia are associated with metabolic
syndrome. Furthermore, abnormalities in fibrinolysis and blood coagulation
concomitant with increased levels of plasminogen activator inhibitor 1 (PAl-1) and
fibrinogen have been attributed to the components of the metabolic syndrome.
Elevated markers of chronic inflammation have been described in relation to
metabolic syndrome (Han, Williams et al., 2002). Besides these abnormalities,
individuals with metabolic syndrome are susceptible to other conditions, notably
polycystic ovary syndrome, fatty liver, cholesterol gallstones, asthma, sleep
disturbances and some forms of cancer (lsomaa, 2003).

Researchers have suggested several etiological mechanisms to explain the
clustering of metabolic disturbances and cardiovascular risk factors described in
connection with the metabolic syndrome. The predominant underlying risk
factors for the syndrome appear to be abdominal obesity and insulin resistance
(Grundy, Cleeman et al., 2005). Other associated conditions can be physical
inactivity, aging, and hormonal imbalance (Apridonidze, Essah et al., 2005). In the
following section of this literature review, insulin resistance is reviewed as the key
etiological factor for the increased predisposition to dyslipidemia, hypertension,

and central adiposity.

26

TABLE 2.2 Abnormalities associated with metabolic syndrome and insulin
resistance/compensatory hyperinsulinemia (modified from Reaven, 2005 and

lsomaa, 2003) (lsomaa, 2003; Reaven, 2005)

 

 

 

Insulin resistance] compensatory
hyperinsulinemia

* T triglycerides
D sli idemia i HDL-C
y p J. LDL-particle diameter (small, dense LDL
particles)
T postprandial accumulation of triglyceride-rich
lipoproteins

 

*

Glucose intolerance impaired fasting glucose
impaired glucose tolerance

type 2 diabetes

Endothelial dysfunction Tmononuclear cell adhesion
T plasma concentration of cellular adhesion
molecules
Tplasma concentration of asymmetric
dimethylarginine
l endothelial-dependent vasodilation

Abnormal uric acid T plasma uric acid concentration
metabolism ivrenal uric acid clearance
Procoagulant factors T plasminogen activator inhibitor-1

T fibrinogen

Tlevels of von Willebrand factor
Markers of inflammation T C-reactive protein

white blood cell count
Others central obesity'

hypertension'

microalbumiuria
fatty liver disease
polycystic ovary syndrome

hemodynamic changes-T renal sodium retention

sleep disordered breathing

*Denotes core cluster of MS; HDL-C, high-density lipoprotein cholesterol; LDL, low-density
lipoprotein

27

2. insulin resistance and dyslipidemias

Insulin resistance— related dyslipidemia includes elevated serum triglycerides
and decreased high-density lipoprotein (HDL). Summarized in Figure 2.1 is
some of the considerable evidence that resistance to insulin-stimulated glucose
uptake leads to a compensatory increase in plasma insulin concentration
enhanced hepatic very low-density (VLDL) triglyceride (TG) secretion and
hypertriglyceridemia. For example, in individuals with normal insulin secretory
function, particularly non-diabetics, there is a relatively linear and positive
relationship between measures of insulin resistance and plasma insulin
concentration (Hollenbeck, Chen et al., 1984; Hollenbeck and Coulston, 1987;
Reaven, 1988). The more insulin resistant an individual is, the greater the
magnitude of hyperinsulinemia. In addition, resistance to insulin-stimulated
glucose uptake is positively associated with VLDL-TG secretion rate and plasma
TG concentration in both normal triglyceridemic and hypertriglyceridemic
individuals (Reaven, 1993). Increased triglyceride levels are implicated as an
underlying cause of the production of small, dense LDL cholesterol particles that
are more easily oxidized, hence more atherogenic, and less readily cleared from
the blood (Howard and Howard, 1994).

Insulin inhibits the release of F FA (non-esterified free fatty acids) from
adipose tissue and people with insulin resistance release FFA in the blood thus
contributing to dyslipidemia. FFAs are in plasma as the products of lipolysis of
triglycerides stored in adipose tissue and are the main substrates for energy

metabolism in the fasting state when insulin and serum glucose levels are

28

relatively low (Reaven and Laws, 1999). When insulin concentration increases,
usually during the fed state, it primarily suppresses plasma FFAs by inhibiting the
action of hormone-sensitive lipase (HSL). HSL causes breakdown of adipocyte
triglycerides into F FAs and glycerol. Plasma FFAs are also the major substrates
for liver triglyceride synthesis (Reaven and Laws, 1999). Insulin can decrease
plasma FFAs by promoting their re-esterification in adipose cells for storage as
triglycerides (Wolfe and Peters, 1987; Coppack, Evans et al., 1992).

@FFA'I
.%

*Liver

 

(hepatic

4‘ TG (CETP) lipase)

1‘Apo B

’I‘ VLDL L)
A A-1----
:EGETg: T6 [)0 )
Insufin
.

(lipoprotein or hepatic lipase)

Figure 2.1 Mechanisms Relating Insulin Resistance and Dyslipidemia
Modified from Lipid Online Slide Library (Rader, 2007)

FFA: Free fatty acids; VLDL: very Low-density lipoprotein; HDL: High-density
lipoprotein; LDL: Low-density lipoprotein; SDLDL: small dense low density
lipoprotein; TG: triglyceride: ApoB: Apoprotein B; CE: cholesterol ester;
CETP:cholesterol ester transfer protein

29

Many studies have demonstrated impaired FFA suppression in those with
insulin resistance (Byrne, Wareham et al., 1994). Byrne and colleagues showed
that FFA suppression is impaired in people with impaired glucose tolerance
compared to those with normal glucose tolerance. Yki-Jarvinen and Tasknen
demonstrated parallel impairment in insulin-stimulated glucose uptake and insulin
suppression of lipolysis in hypertriglyceridemic diabetic and non-diabetic subjects
(1988). Finally, with type 2 diabetes researchers have extensively documented
a defect in insulin suppression of FFA level (Bierman, Dole et al., 1957; Kashyap,
Magill et al., 1970; Reaven and Laws, 1999).

In insulin resistant subjects, day- long increases in plasma insulin (muscle
insulin resistance) and F FA concentrations, due to adipose tissue insulin
resistance, act on an insulin sensitive liver and stimulate hepatic TG synthesis
(Olefsky, Farquhar et al., 1974; Reaven, 2005). A consequence of hepatic TG
synthesis is an increase in very-low-density lipoprotein (VLDL) synthesis and
secretion, ultimately leading to hypertriglyceridemia (Ginsberg and Huang, 2000).
Insulin resistance also decreases lipoprotein lipase activity, the major mediator of
VLDL clearance that might also contribute slightly to elevated triglycerides. Low
levels of HDL-C are predominately found in certain populations, such as those
who smoke, are physically inactive, or who have type 2 diabetes, metabolic
syndrome, obesity, or hypertriglyceridemia (NCEP, 2001 ). Low HDL-cholesterol
concentrations in insulin- resistant or hyperinsulinemic persons are partly due to
the cholesterol ester transfer protein (CEPT) transferring of cholesterol from HDL

to VLDL (Swenson, 1991) (Fig 2.2). The higher the VLDL pool size, the greater

3O

the transfer rate from HDL to VLDL and the lower the resulting HDL-cholesterol

    

concentration.
Bile
Mature HDL
(' M Nascent HDL
- A-I
!
FC - .

 

“CE<————

  " \(x SR-BI

LD LR

VLDL/LDL

CET P = cholesteryl ester transfer protein; LDL = low-density
lipoprotein; LDLR = low-density lipoprotein receptor; VLDL = very-
low-density lipoprotein

Figure 2.2. Role of CETP in HDL Metabolism. Modified from Lipid Online
Slide Library (Rader, 2007)

Several studies have investigated the effects of insulin resistance on
lipoprotein abnormalities (Abbott, Lillioja et al., 1987; Laakso, Sarlund et al., 1990;
Laws and Reaven, 1992). In some dyslipidemic individuals, HDL particles may
be transformed to dysfunctional, lipid-poor, small HDLs with diminished residence
time in the plasma compartment. For instance, in moderate
hypertriglyceridemia, cholesterol ester transfer protein (CETP) activity
preferentially targets elevated levels of triglyceride—rich VLDL, leading to VLDL
enrichment with cholesterol but also equally enhanced TG content in HDL

(Guerin, Le Goff et al., 2001). Hepatic lipase hydrolyzes triglyceride-enriched HDL

31

from small dense HDL. Such small dense HDL of abnormal composition (high
triglyceridelcholesterol) has a shorter half-life in plasma than large HDL, and, as a
result, HDL levels decrease (Lewis and Rader, 2005). Small dense HDL
particles in hypertriglyceridemic states also exhibit impaired anti-oxidative activity
versus larger HDL particles, as demonstrated recently in a study that compared
patients with the metabolic syndrome from matched, normolipidemic healthy
controls (Hansel, Giral et al., 2004).

These complex mechanisms are the means by which prolonged insulin
resistance can lead to dislipidemia. Individual differences in genetic
susceptibility likely also play important roles. The next section reviews the
hypothesized mechanisms for insulin resistance and hypertension.

3. Insulin resistance and hypertension

The most complicated relationship between insulin resistance or
hyperinsulinemia and CVD relates to the role of essential hypertension. The
prevalence of insulin resistance among patients with hypertension is less in those
with hypertriglyceridemia or hyperglycemia (Bonora, Kiechl et al., 1998). No
more than half of patients with essential hypertension are also insulin resistant
(Zavaroni, Mazza et al., 1992), yet this subset of patients are at greatest risk for
CVD (Sheu, Jeng et al., 1992; Jeppesen, Hein et al., 2001). For example, patients
with essential hypertension with electrocardiograph evidence of ischemic
changes are more glucose intolerant and hyperinsulinemic compared with a
normotensive control group or a hypertensive patient with a normal

electrocardiogram (Sheu, Jeng et al., 1992). Measurement of insulin-mediated

32

glucose disposal has demonstrated that patients with essential hypertension and
ischemic electrocardiograph changes were insulin resistant, and dyslipidemic
changes associated with insulin resistance/hyperinsuIinemia were present in
these individuals compared with normotensive individuals or hypertensive
patients with normal electrocardiograms electrocardiogram (Sheu, Jeng et al.,
1992; Reaven, 2005).

Although insulin may have a vasodilatory effect, this effect seems to be active
in the microvasculature of the skeletal muscle, an important determinant of insulin
mediated glucose uptake (Vincent, Barrett et al., 2003). In vivo studies have
shown that insulin exerts a different effect on larger arteries and can cause
elevation in blood pressure in healthy humans (Arcaro, Cretti et al., 2002).

Arcaro et al. demonstrated that exogenous insulin infusion in healthy subjects
could cause endothelial dysfunction and diminish endothelium-dependent
vasodilation in the large conduit arteries by mechanisms involving the induction of
oxidative stress. In the presence of insulin resistance, as in patients with type 2
diabetes, the effect of insulin on the microvasculature in the skeletal muscle is
blunted (Baron, 1994).

It is reasonable to postulate that the sympathetic nervous system overrides
the normal vasodilatory effects of insulin under more extreme conditions such as
obesity, sucrose feeding and hypertension. For example, insulin has been
shown to contribute to sodium retention (Nosadini, Sambataro et al., 1993; Ling,
Matsunaga et al., 1995), to stimulate the sympathetic nervous systems

(Rahmouni, Morgan et al., 2004) and to induce oxidative stress. All these actions

33

could contribute to the pathogenesis of hypertension in people who are insulin
resistant (T seng, 2006). Recent studies on adipose tissue derived proteins
found that renin and angiotensinogen are produced by the adipocyte and might
contribute as well to the link between obesity and hypertension (Reaven, Scott et
al., 2005).

This section has described some likely mechanisms by which insulin
resistance and/or obesity could contribute to the manifestation of hypertension,
which is one criterion of MS in all three definitions. Elucidation of these
relationships is an active area of research as described next.

C. Impact of total body fat, abdominal fat and visceral fat on metabolic
syndrome

Worldwide, the prevalence of obesity is increasing dramatically. Twenty
years ago, as documented in NHANES III (1988 to 1991), 42% of men and 52% of
women in their fifties were ovewveight in the US. (Kuczmarski, Flegal et al.,
1994); and recent estimates from show NHANES 2003-2004 that 78% of men and
68% of women 40-59 years of age were overweight (Ogden, Carroll et al., 2006).
Using the criteria of the National Cholesterol Education Program Adult Treatment
Panel lll (ATPIII), the prevalence of the metabolic syndrome in US. adults 220
years of age was recently estimated to be 23.7% (Ford, Giles et al., 2002). 1
Prevalence increased with age, reaching 43.5% and 42.1% for those 60—69 and
270 years of age, respectively. According to NCEP ATP III, the “obesity epidemic”
is primarily responsible for the rising prevalence of metabolic syndrome. Obesity

is now known to be linked to insulin resistance, type 2 diabetes, and

34

cardiovascular disease and contributes to hypertension, high serum cholesterol,
low HDL cholesterol, and hyperglycemia. It is abdominal obesity in particular,
however, that especially correlates with metabolic risk factors and excess adipose
tissue releases several inflammatory factors that exacerbate these risk factors
(Grundy, Brewer et al., 2004).

The purpose of this section is to review the relationship of different regions of
body fat deposition and the inherent risks for metabolic syndrome. In addition,
the critical role that adipose tissue has in the manifestation of the insulin
resistance components will be discussed.

1. Abdominal fat vs. peripheral fat in relation to the metabolic syndrome

Many studies have shown that increased body fat adiposity elicits increased
insulin resistance. It is increasingly recognized that not only the degree of
overweight but also the distribution of body fat is important to the development of
insulin resistance and the metabolic syndrome. (Donahue and Abbott, 1987;
Ross, Leger et al., 1991; Bjorntorp, 1992). There is considerable evidence that
visceral or intra-abdominal fat, which increases with age, is linked to insulin
resistance. Adipose tissue, increased F FA turnover, and increased levels of
circulating FFA found in those who are obese are also highly associated with
insulin resistance (Bjorntarp, 1992). Over the last 10 years the use of computed
tomography, a tool used to study regional adipose tissue accumulation, has been
used to demonstrate significant positive associations between abdominal visceral
adipose tissue deposition and the risk factors of metabolic syndrome (Kvist,

Chowdhury et al., 1988).

35

Abdominal obesity, or “central adiposity”, is a preferential deposition of fat in
the abdominal region and often involves accumulation of visceral fat. Patients
with central adiposity have higher insulin levels and are more insulin resistant
than subjects at a similar weight with a peripheral type of obesity (Abate, Garg et
al., 1997; Samaras and Campbell, 2000). The importance of central obesity is
well recognized in the definitions of metabolic syndrome. Abdominal fat, however,
can be subdivided into two compartments: subcutaneous adipose tissue (SCAT)
and visceral adipose tissue (VAT) as identified by computed tomography or
magnetic resonance imaging scanning (Borkan, Oerzof et al., 1982; Kvist,
Chowdhury et al., 1988).

The excess accumulation of VAT appears to play a more significant
pathogenic role in metabolic syndrome than does SCAT. The abdominal SCAT is
located immediately beneath the skin and on top of the abdominal musculature.
Most lower body fat is SCAT, stored primarily in the femoral and gluteal regions
(Wajchenberg, 2000; Lafontan and Berlan, 2003). VAT deposits, located in the
body cavity beneath the abdominal muscles, are composed of the greater and
lesser omentum (peritoneum attached to the stomach and linked with other
abdominal organs) and the mesenteric fat. A lesser amount of VAT is located
retroperitoneally. In general, VAT accounts for up to 20 percent of total fat in men
and 5—8 percent in women. Studies have shown that lean individuals with normal
BMls can also have a significant accumulation of VAT with increased risk factors
for cardiovascular disease and diabetes (metabolically obese, but normal weight)

(Ruderrnan, Chisholm et al., 1998; Katsuki, Sumida et al., 2004). Meanwhile,

36

obese individuals with high BMI but relatively little VAT can present normal
metabolic profiles and a few risk factors for metabolic syndrome, cardiovascular
disease, and diabetes (Ruderman, Chisholm et al., 1998).

Numerous inherent differences between VAT and SCAT have been found in
the last decade. VAT is a major predictor for insulin resistance (Wagenknecht,
Langefeld et al., 2003) and metabolic syndrome (Carr, Utzschneider et al., 2004).
The "Portal Theory" hypothesized that insulin resistance and related features
could arise from VAT delivering FFAs at a high rate directly into the liver via the
portal vein (Bjorntorp, 1990; Amer, 1997). This, in turn, would increase hepatic
glucose production, reduce hepatic insulin clearance and ultimately lead to insulin
resistance, hyperinsulinemia, and hyperglycemia, as well as non-alcoholic fatty
liver disease. FFA flux could also lead to enhanced production of triglycerides
and apolipoprotein B-rich lipoproteins, both features of insulin resistance
syndrome reviewed earlier (Busetto, 2001). Compared to SCAT, VAT adipocytes
have a higher rate of lipolysis more readily stimulated by catecholamines and less
readily suppressed by insulin (Zierath, Livingston et al., 1998). VAT also
releases a larger amount of glycerol than the liver can convert to glucose, thus
contributing to hyperglycemia (Kishida, Kuriyama et al., 2000). Other investigators
have argued that the “Portal Theory” is inadequate as a sole explanation of VAT’s
role in metabolic syndrome (Freedland, 2004). The mechanism by which VAT
contributes to MS might be due in part to the proteins produced by the adipocytes

and will continue to be investigated.

37

2. Relationship of adipose tissue derived proteins, adipocytokines, to the

metabolic syndrome

The investigation of mechanisms for obesity’s intimate link to insulin

resistance, type 2 diabetes and cardiovascular disease has led to an increased
understanding of the complex function of adipocytes. Traditionally considered a
passive storage depot for triglycerides, adipocytes are now recognized as a
complex and active endocrine tissue secreting numerous immunomodulatory
factors that play major roles in the regulation of human metabolism and vascular
biology (Wisse, 2004; Hamdy, 2005; Hutley and Prins, 2005). In the last decade,
researchers have learned that adipocytes produce several proteins including
leptin, adiponectin, resistin ,tumor necrosis factor (TNF)-d, interleukins (lLs),
C-reactive protein (CRP), and plasminogen activator inhibitor (PAl)-1. Some of
the proteins are classical cytokines and others, including leptin, are structurally
related to cytokines (Wisse, 2004). This finding has led to the introduction of the
term ‘adipocytokines’ to describe the wide range of proteins adipose tissue
produces (Table 2.3). The term adipocytokine also highlights that several proteins
produced by adipocytes may act locally as autocrine and paracrine factors rather
than remote-acting endocrine factors (Frayn, Karpe et al., 2003). Through the
actions of such proteins, the adipocyte itself becomes central to the development
of insulin resistance and vascular disease.

For the adipocytokines and Lem, the endocrine function of adipose tissue is
well established. In the past decade, the endocrine role of leptin has expanded to

include regulation of reproduction (Cunningham, Clifton et al., 1999) and immune

38

function (Lord, 2002). Leptin exerts an effect on the central nervous system to
control satiety, increase energy expenditure, and reduce food intake. Obesity is
associated with leptin resistance and increased leptin secretion. Most overweight
and obese humans have elevated levels of leptin that do not suppress appetite.
This leptin resistance is postulated to be a fundamental pathology in obesity
(El-Haschimi, Pierroz et al., 2000; Hutley and Prins, 2005). Leptin helps regulate
insulin action within the liver and has anti-inflammatory actions (Klein, Horowitz et
al., 2000). Leptin secretion is considerably greater from subcutaneous fat stores
than from visceral fat depots (Montague, Prins et al., 1998; Van Harmelen,
Reynisdottir et al., 1998).

Adiponectin is an adipocytokine that may have a protective role in the
vascular wall (Ouchi, Kihara et al., 1999; Hotta, Funahashi et al., 2001).
Adiponectin can accumulate within injured vascular walls (Okamoto, Kihara et al.,
2006) and suppress TNFa-induced expression of adhesion molecules in vascular
endothelial cells (Ouchi, Kihara et al., 1999). Adiponectin has been strongly
implicated in cardiovascular health. Epidemiologic data demonstrates an inverse
association between serum adiponectin concentrations, vascular inflammatory
markers and manifestations of the metabolic syndrome, including altered CRP,
fibrinogen, hypertension, and endothelial function (Shimabukuro, Higa et al.,
2003; Schulze, Rimm et al., 2004). Furthermore, reduced levels of adiponectin
(hypoadiponectinemia) are associated with higher BMI, decreased insulin
sensitivity, less favorable plasma lipid profiles, and increased risk for development

of cardiovascular disease (Ryo, Nakamura et al., 2004; Trujillo and Scherer,

39

2005). The primary mechanisms by which adiponectin enhances insulin
sensitivity appears to be through increased fatty acid oxidation and inhibition of
hepatic glucose production (Lihn, Pedersen et al., 2005). Other actions of
adiponectin include suppression of macrophage to foam-cell transformation in
vitro (Ouchi, Kihara et al., 2001) and inhibition of endothelial signaling through
cyclic adenosine monophosphate (cAMP). In addition to its effects on the
endothelium and vascular wall, adiponectin secretion is reduced in obesity, which
further increases insulin resistance and inflammation (Reaven, Scott et al., 2005).

T‘umor necrosis factor (TNF-a) is a proinflammatory adipokine that stimulates
expression of other inflammatory mediators like leptin and lL-6 (Banks, Forbes et
al., 1995; Bullo, Garcia-Lorda et al., 2003). Expression and secretion of
anti-inflammatory adiponectin is reduced in response to TNF-a (Kappes and
Loffler, 2000). By releasing TNF-a, the adipocyte directly suppresses insulin
signaling, induces insulin resistance, and has direct inflammatory effects on the
macrophage and the endothelial cell. This promotes the development of
atherosclerosis.

Plasminogen-activator inhibitor 1 (PAI-1) is a prothrombotic factor which is
found more highly secreted by VAT than abdominal SCAT (Wagenknecht,
Langefeld et al., 2003; Fain, Madan et al., 2004; Pantanetti, Garrapa et al., 2004),
while leptin is more highly secreted by SCAT (Minocci, Savia et al., 2000).
Elevated levels of PAl-1 in inflammatory and obese states are a known risk factor
for thrombosis and a predictor for cardiovascular events and mortality (Fay,

2004).

40

Renin and angiotensinogen are involved in blood pressure control, causing
vasoconstriction. The fact that they are produced by the adipocyte may explain
the link between obesity and hypertension (Reaven, Scott et al., 2005).

The adipocytes and numerous other adipose tissue-derived molecules
described here have an impact on glucose homeostasis, vascular biology, tumor
development, lipoprotein metabolism, and inflammation (Rajala and Scherer,
2003). It is now clear that obesity, especially central obesity, dysregulated
expression and secretion of adipokines drives a proinflammatory state.
Adipokines from VAT can be delivered via the portal system directly to the liver
where they can affect hepatic, and ultimately systemic, inflammation. Chronic
inflammation contributes to the development of insulin resistance, glucose
intolerance, and atherogenesis, all characterizing the metabolic syndrome.
Understanding the molecular mechanisms linking obesity with components of the
metabolic syndrome represent key foci to determine the connection among

obesity, metabolic syndrome and cardiovascular disease.

41

Table 2.3. Proteins produced by adipose tissue, adipocytokines.

 

TNF-a

lL-6

lL-1j3

Lepfin

Adiponectin

Resistin
Acylation-stimulating protein
SAA3

d 1 acid glycoprotein
Pentraxin-3

lL-1 receptor antagonist
Macrophage migration inhibitor factor

 

Adapted from (Wisse, 2004)

42

D. Ethnic differences in metabolic risk factors

The prevalence of metabolic syndrome varies by race/ethnicity, by age, and
by the definition used for diagnosis as illustrated in Table 2.4. Environmental
factors, including health-related behaviors, lifestyles, and economic
disadvantages likely contribute to some of the race/ethnic disparities in the
prevalence of metabolic syndrome. However, because these factors cannot
explain all of the racial differences in disease patterns, genetic factors likely
operate at the metabolic level and in conjunction with and response to the
environment contributing as well to disparities in obesity-related co- morbidities
(Cossrow and Falkner, 2004). The aim of the next section of the literature review
is to evaluate current studies on the prevalence and variations in metabolic
syndrome in different ethnic groups and to explore possible contributions to the
disparities.
1. The prevalence of the obesity related metabolic risks in the US

Using the criteria of the NCEP-ATP III with the NHANES III database,
updated to the 2000 census, the overall prevalence of the metabolic syndrome in
the United States is high, approximately 47 million (23.8%) American adults (Ford,
Giles et al., 2002). The risk of metabolic syndrome progressively increases with
age, rising from approximately 7% for adults in their 30s to over 40% risk for those

who are

43

 

 

98 «.3 gm EH tom :22.
«.9 new New 0.8 at 2: 5.22
3m New 3: 3 92 am 3220
.02 98 «.8 o: 2: has 3.2 8% 2859.5 238.5% azemt
Avoow
rt 9 8A _Easfim .220 ...m 8 Box .93
Avoou
_._m Ho :20 .9528
m9 2: m9 S S: 3 0? 89.0. 533.
0.8 3m :2 82me
SN 4.? E< 58:2 88.0.
QNN QVN 9E>> _._m E mm=0 .20”:
«.mw 3 EN 0? 3% m:
v.8 58 1H 0.? 8m 53939525 308
5 $4 3m 3m :5 525:8 _._2m_§90u_.:___E
m 3: «.3 «.2 8-3 98 52.9.3
3m QB 83 m:
09.... 3: 88 25:2:
v.2 2: 83:. 53.3 608
wow 0.3 BA 88 9.8. 9.2.. :68 891.35%
:8 0.: 8-8
new NE 218 Efiécmoczo B8765 33.;
at g 30v g g g at g 3.;
295“. was. __< 295“. 222 __< 295“. was. __<
.850 0:3 __n_._.< o1W< T5 3:95 £53 3:35

 

oEEugm 0:03.322 no 52530

 

232536 523:“. 3 3:05 25.3 “c.2355 5 220.656 0:03.305 ho 3535.5 {N 03m...

44

 

 

0.0. 0.2 02 00:00:05.

0.: 0.3 ~00 000.8. .0000

0. : 0.2 01.0 000 0000000.. ...0 .0 .002 50.03.00.

0.2 0.8 0.00 .000 003 00.8. 080.65

.0000

0.0 :0003. .._m .0 {mm .003
3.0 3; .00. .00. £0 3... 3°. 3... as.
0.00.0“. 0.0.). __< 00.00“. 0.0.2 __< 0.050“. 0.0.2 ._<

.050 oz; :05 006 1"... 350% 0.050. 00.030

 

othugw 3.3522 ..o 52:50

.0003. .30 0.00..

45

60 years and older. Overall, the prevalence of the metabolic syndrome shows
similar patterns within each ethnic group, i.e., increasing with age and BMI. The
prevalence of the metabolic syndrome is highest in Hispanics and lower in
non-Hispanic whites and African Americans. Cossrow and colleagues found that
African Americans exhibit less dyslipidemia compared to Caucasians, but higher
rates of coronary heart disease and stroke (Cossrow and Falkner, 2004). The
lower prevalence of metabolic syndrome among African Americans might be
explained by the two lipid criteria in the ATP Ill definition of high triglycerides and
low HDL cholesterol. Lower prevalence of lipid abnormalization in African
Americans might offset their high rates of hypertension and glucose intolerance.
The NHANES 1999-2000 survey demonstrated that the prevalence of co-
morbidities for MS was high and increasing for various groups(Hajjar and
Kotchen, 2003). Currently, 28.7% of Americans have hypertension (Hajjar &
Kotchen, 2003). Non-Hispanic blacks had the highest prevalence of hypertension
(33.5%) among race-ethnic groups. Prevalence of hypertension increased with
age to 65.4% among those 360 years and older, and tended to be higher in
women (30.1%) than in men (27.1%). Mexican-Americans had the lowest
prevalence of hypertension during the same period of time (Hajjar and Kotchen,
2003). The prevalence of the metabolic syndrome using the ATPlll definition
was 86% among people with diabetes who were over age 50 in the NHANES
1999-2000 survey (Alexander, Landsman et al., 2003). A lower but still
higher-than-average prevalence of the metabolic syndrome was observed in

people with impaired glucose tolerance (31%) and impaired fasting glucose

46

(71%). Prevalence of the metabolic syndrome was 60% greater than that of type
2 diabetes in the same population. Similar differences in glucose tolerance
between African-Americans and non-Hispanic whites were found in healthy young
adults from the Coronary Artery Risk Development in young adults (Pereira et al.,
2002). This finding suggested that the predisposition to abnormalities in glucose
tolerance may be more common in African-Americans than in Caucasians
(Pereira, Jacobs et al., 2002). Thus, the increased prevalence of MS may be
related to an increase in BMI as well as to the aging and increasing ethnic-racial
diversity of the US. population (Mokdad, Ford et al., 2003; Kim and Beckles,
2004).
2. The prevalence of the obesity related metabolic risks in Asians

Several studies provide evidence to support that insulin resistance could be
the biological basis for the racial difference in expression and prevalence of
diabetes. Dickinson and colleagues examined possible variations in the
evolution of the insulin resistance syndrome across races by comparing Asians
and Caucasians (Dickinson, Colagiuri et al., 2002). Young, healthy, lean Asian
and Caucasian adults were fed white bread in an amount to provide 75 g of
available carbohydrate. Baseline plasma glucose and insulin as well as changes
in glucose and insulin concentration were measured for 2 hours post parandial.
Mean fasting glucose concentrations were similar between groups, but
postprandial glucose was markedly higher in Asian subjects compared with
Caucasians, with a two-fold higher mean incremental area under the curve.

Insulin responses were similar with a mean area under the curve greater than

47

two- fold higher for the Asian group. When insulin sensitivity was estimated by the
homeostasis model assessment, the Asians were less insulin sensitive than the
Caucasians. These results supported the likelihood of racial differences in
predisposition to the development of the insulin resistant syndrome. However,
cardiovascular fitness levels were not controlled nor matched in this study.

Pan et al. (2002) used data from the Nutrition and Health Survey in Taiwan
(NAHSIT) and the US National Health and Nutrition Examination Survey I"
(NHANES III) to estimate the prevalence rate of the metabolic syndrome defined
by NCEP-ATP lll guidelines. Findings demonstrated that the risk of increasing
BMI for hypertension, diabetes, and hypertriglyceridemia was greater for
Taiwanese than for US non-Hispanic whites in the US. As illustrated in Figure
2.3, the clustering of high blood pressure (BP), triglycerides (TG) and high-density
lipo-protein (HDL-C) with or without large waist circumference was most common
for Taiwanese men, while the US non-Hispanic white population showed a more
evenly distributed pattern. Most Taiwanese women had a clustering of high HDL,
BP and waist circumference, while most white women had a syndrome of high
HDL, TG, and waist circumference. Abdominal obesity appeared more prevalent
for women than for men across populations. The results of this study showed
that the patterns of MS varied by gender and among various ethnic/racial groups.
The implications of these different co-morbidity patterns should be considered in

cross cultural comparisons.

48

.ANOON 500.: E0: U030U<

00820 000.0 9500. cm... 5.0 00:20.6:26 00.0; 090. ”mo 05000.3. 083 :9: ”am 5.0.90.3: 3.05.0 20.5.26.
1.0... 00200205-... ”0.... 00.2.3 0.53.1-5: w: :. 0:0 00053.0... E 089930 o..onm.0E .o 0E0tmn. ad 050.“.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5:635

20 no

3.0 no 90 90 am am 3.0 mm

+ 3.0 am am .5... .5... + ..O...
.050 am .5... 0... 0... 0... .050 0...

m m m m m m v v m
:05.

 

 

 

 

$3.0...
._._mI<Z

 

 

 

 

 

 

 

 

 

 

 

at:

.5253...

 

 

 

 

 

 

 

 

 

 

3.0 90

+ 20
.050 mm
m n

no

.5...
m

5.5020.
00 no
no mm mm 3.0 am
mm ..o... ..01 + ..o...
O... 0... O... .050 0...
m m n v v m

:0Eo>>

49

Many Asians have a higher prevalence of metabolic disorders at a lower BMl
level compared to their Caucasians counterparts (Pan et al., 2002). It has been
debated for years whether the cut point for BMI to define obesity for Asian
countries should be set lower than that for populations with predominantly
Caucasian and African ancestry (Ota, Takamura et al., 2002; Grundy, Brewer et
al., 2004; Kim, Ahn et al., 2005; Shiwaku, Nogi et al., 2005). To this end,
investigators have examined the relationship between central body fat deposition

in Asians as a contributing risk factor to a higher prevalence of type 2 diabetes

and premature CVD.

Recently, it has been found that type 2 diabetes and premature CVD occurs
in individuals with lower BMIs in Asians compared to Caucasians (Ota,
Takamura et al., 2002; Grundy, Brewer et al., 2004; Kim, Ahn et al., 2005;
Shiwaku, Nogi et al., 2005). Bell et al. examined ethnic differences on the
strength of the association between BMI and hypertension (Colin Bell, Adair et al.,
2002). Cross-sectional data were used from adults aged 30-65 years in China (n
= 3,423), the Philippines (n = 1,929), and the United States (n = 7,957). A higher
BMI was associated with a higher prevalence of hypertension in all ethnic groups.
However, at BMI levels less than 25, prevalence rate increased more in Chinese
men and women, but not in Filipino women, compared to non-Hispanic whites.
Non-Hispanic blacks and Filipino women had a higher prevalence of hypertension
at every level of BMI compared with non-Hispanic whites and

Mexican-Americans. These ethnic differences, along with the strength of

50

association between BMI and hypertension and in underlying prevalence, warrant
further investigation of the use of ethnic-specific BMI cutoffs in clinical settings to
more accurately identify individuals at risk from obesity (Colin Bell, Adair et al.,

2002)

Another study examined Singaporean Chinese, Malays, and Indian
populations to determine the risk for selected co-morbidities at various BMI
categories and abdominal fat distribution using data from the Singapore National
Health Survey (Deurenberg-Yap, Chew et al., 2002). Results demonstrated that
at any given body fat percentage, the BMI of Singaporean Chinese, Malays, and

Asian Indians was approximately 3 kg/m2 lower than that of Caucasians.

In Asian ethnic groups, specifically Chinese, Korean, Indian, Pakistani, and
Bangladeshi, the risk of hypertension, diabetes, and hypercholesterolemia starts
to increase rapidly at levels of BMI and WC well within the acceptable range of
BMI or WC for Europeans (Deurenberg, Yap et al., 1998; Deurenberg-Yap, Chew
et al., 2002; Kim, Suh et al., 2004; Pan, Flegal et al., 2004). A possible
explanation of the difference in risk across races/ethnicities might be that Asians
have a higher level of body adiposity at a lower level of BMI or more visceral
adiposity distribution compared to Caucasians. Co-morbidities of metabolic
syndrome also appear to be associated with both abdominal fat distribution and
total body adiposity. The relative effect of total fat and of abdominal fat to MS
and to specific metabolic disorders within various populations, however, needs

clarification.

51

Although metabolic syndrome occurs in all populations, the disturbances of
lipid metabolism and distribution of body fat that accompanies insulin resistance
vary among populations. Genetics and environmental factors all contribute to
such disparities in co-morbidities. Varying definitions and age groups used in
different studies further complicate the differences. Regardless, cross cultural
studies can enhance understanding of etiology, so important to developing
effective prevention and treatments. Comparative studies between populations
of European and Asian descent can help to delineate differences in amount and
distribution of adipose tissue between groups. Such comparisons are important
because the WHO definitions for ovenNeight and obesity and NCEP-ATP lll
definitions for metabOlic syndrome were established using data from populations
primarily of European descent (WHO, 1995; Pan, 2002). However, people of
Asian descent appear to exhibit metabolic risks at lower anthropometric
cut-points, and finally, the marked increase in Asian populations around the globe

may call for the lowering of their BMI cut-point.

E. Lifestyle factors (diet and physical activity) related to metabolic

disorders.

The metabolic syndrome is closely linked to lifestyle factors of energy
intake in excess of needs and a sedentary lifestyle. One study showed that
middle- aged men engaging in over three hours per week of moderate to vigorous
leisure-time physical activity were half as likely as sedentary men to have MS

(Laaksonen, Lakka et al., 2002). This risk reduction was similar in two

52

prospective studies for the development of diabetes following lifestyle
interventions for obese subjects with impaired glucose tolerance (Tuomilehto,
Lindstrom et al., 2001; Diabetes Prevention Program Research Group, 2002).

Due in a large part to findings like these, the revised NCEP-ATP Ill report
targeted lifestyle interventions for prime consideration to reduce risk for metabolic
syndrome (Grundy, Cleeman et al., 2005) . Weight management, increased
physical activity, and dietary moderation were all recommended to reduce risk for
atherosclerotic cardiovascular disease (Grundy, Cleeman et al., 2005). The
following section will review the current evidence for the salutatory effect of
healthy diet and physical activity on the etiology of insulin resistance and the
metabolic syndrome.

1. Weight reduction
It is commonly observed that the probability of having metabolic

abnormalities, including metabolic syndrome, increases with the level of obesity.
The association between increased prevalence of obesity and metabolic
syndrome reflects the importance of adiposity as a contributor to insulin
resistance and metabolic syndrome. Recent clinical trials and observational
epidemiologic studies demonstrate the importance of lifestyle changes—such as
decreased intake of energy and dietary fat—to sustain weight loss, as well as
regular physical activity to improve insulin sensitivity, glucose tolerance and to
reduce the risk of cardiovascular disease (Ross, Janssen et al., 2004; Janssen,

Katzmarzyk et al., 2005).

53

Several researchers have emphasized that dietary treatment for metabolic
syndrome should primarily focus on weight reduction to improve both insulin
sensitivity and other aspects of the metabolic syndrome (Riccardi and Rivellese,
2000; Riccardi, Aggett et al., 2004; Reaven, 2005). Evidence from observational
and clinical trials of lifestyle interventions indicates that even moderate, sustained
weight loss can substantially improve insulin action and reduce therisk of type 2
diabetes. The Finnish Diabetes Prevention Study used an intensive lifestyle
modification and middle-aged, overweight subjects with impaired glucose
tolerance demonstrated reduced risk for diabetes of 58% over 3.2 years of
follow-up, accompanied only by a moderate weight reduction of 3.5 Kg
(T uomilehto, Lindstrom et al., 2001). The Finnish intervention included a diet low
in saturated and trans fats and high in fiber with regular moderate physical
activity. Hermansen reported similar results in a review of weight reduction and
blood pressure interventions (Hermansen, 2000). Most randomized control
studies demonstrated that even a modest weight loss of 3-9% was associated
with a significant reduction in systolic and diastolic blood pressure, roughly
3mmHg, in overweight and obese people.

Most recently in the Diabetes Prevention Program, 3,234 individuals with
impaired glucose tolerance were randomly assigned to receive intensive lifestyle
intervention, metformin, or a placebo (Ratner et al., 2005). Goals of the intensive
lifestyle modification program were to achieve and maintain a weight reduction of
at least 7% of initial body weight through consumption of a healthy, low-calorie,

low-fat diet and to engage in moderately intense physical activity for at least 150

54

min/week. After three years of annual assessments, the lifestyle interventions
improved CVD risk factor status compared to the placebo and metformin therapy
(Ratner, Goldberg et al., 2005). Findings from these studies demonstrate that
changes in diet and physical activity can reduce the risk for the co-morbidities of

the metabolic syndrome.

2. Relationships of dietary factors and metabolic syndrome

a. Dietary Fat

Consumption of energy-dense and/or high fat diets is strongly and positively
associated with overweight and obesity. This in turn impairs insulin sensitivity,
particularly when excess adiposity is located in the abdominal region (Feskens,
Virtanen et al., 1995). Insulin sensitivity is not only affected by the amount of
dietary fat, but also by the quality of dietary fat (Riccardi and Rivellese, 2000).
Epidemiological evidence and intervention studies reviewed here clearly show
that high intakes of saturated fats significantly worsened insulin-resistance, while
those of monounsaturated and polyunsaturated fatty acids improved it through
modifications in the composition of cell membranes reflecting dietary fat
composition (Riccardi, Giacco et al., 2004).

Astrup and colleagues conducted a meta-analysis of 16 ad libitum low-fat

dietary interventions to review the role of dietary fat on body fatness. The results

showed that a 10 percent reduction of total Kilocalories from fat would be

expected to decrease mean body weight of the population by 2.5 Kg (Astrup,
Grunwald et al., 2000; Frayn and Vessby, 2000). F urhtermore, such a reduction
in fat would reduce the obesity prevalence from 20% to 10%, which is an
important public health impact.

In a large, geographically based sample of Hispanic and non-Hispanic whites
in southern Colorado, high total and saturated fat intakes were associated with
hyperinsulinemia independent of age, gender, ethnicity, BMI, waist
circumference, total energy intake, or participation in vigorous physical activity
(Marshall, Bessesen et al., 1997). Findings from the Dutch and Finnish cohorts
of the Seven Countries Study, a 30—year follow-up survey, supported a role for
dietary fat in glucose intolerance and the development of type 2 diabetes
(Feskens, Virtanen et al., 1995). Although the regression coefficients were
small, habitual intake of total and saturated fat higher than the average was
positively associated with diabetes incidence 30 years later, adjusting for age,
cohort, BMI, and energy intake.

A recent multi-center study demonstrated that a shift from a diet rich in
saturated fatty acids to a diet rich in monounsaturated fat improved insulin
sensitivity in healthy people (Riccardi, Giacco et al., 2004). Rasmussen
conducted randomized multi-center clinical trial with 162 healthy, normotensive
subjects assigned to one of two isoenergetic diets: rich in monounsaturated fatty
acids (MUFA diet), or rich in saturated fatty acids (SFA diet) (Rasmussen, Vessby
et al., 2006) . After three months, those on the MUFA—rich diet had a significantly

lower diastolic blood pressure (DBP) compared to those on the SFA-rich diet.

56

Interestingly, the favorable effects of MUFA on DBP disappeared at a total fat
intake above the median of 37% total energy. Although the type of fat, rather than
the amount of fat, in the diet might be more important in terms of determining
health outcomes, it is noteworthy that a reduction on blood pressure from a
MUFA- rich diet is lost when the total fat intake is high.

Positive correlations between changes in dietary total and saturated fatty
acids and changes in total, LDL, and HDL cholesterol were in a meta-analysis of
37 dietary intervention studies observed (Yu-Poth, Zhao et al., 1999). In free-living
subjects, plasma total cholesterol decreased by 24 mgldl (10%) compared to
baseline, LDL cholesterol by 19 mgldl (12%), and triglycerides by 15 mgldl (8%) in
Step I diet interventions of 10% saturated fat and 300 mg cholesterol. In Step II
diet interventions of 7% saturated fat and 200 mg cholesterol, total cholesterol
decreased from baseline by 32 mgldl (13%), LDL cholesterol by 25 mgldl (16%)
and triglycerides by 17 mg/dl (8%). However, HDL cholesterol decreased by 7%
in response to Step II but not Step I dietary interventions (Yu-Poth, Zhao et al.,
1999).

b. Dietary carbohydrate

Three aspects of dietary carbohydrate might be relevant to the relation of
dietary carbohydrate to risk for metabolic syndrome. Findings from studies
investigating the replacement of dietary carbohydrate for fat, dietary fiber, and the
glycemic index have all been suggestive. Some of the research on these aspects
of carbohydrate and risks for co-morbidities of metabolic syndrome are reviewed

in this section.

57

Until recently, the prudent diet recommended for all Americans contained
approximately 15% of total energy from protein, 25%-30% from fat, and 55%-60%
from CHO (Gardner and Kraemer, 1995), an approach aimed at decreasing
saturated fat intake. The rationale for this recommendation was to reduce CVD
risk by keeping low-density lipoprotein cholesterol (LDL-C) concentration as low
as possible (Gardner and Kraemer, 1995). The problem with the prudent diet
approach was that replacing saturated fat with carbohydrate in isocaloric diets
might require more insulin secretion to maintain glucose homeostasis. As a result,
insulin resistant individuals will have a greater degree of compensatory
hyperinsulinemia (Reaven, 2005).

Although the beneficial effects of low-fat and high-CHO diets on LDL-C
concentrations are not lost in insulin resistant individuals, this dietary modification
in an insulin resistant person might increase postprandial plasma glucose, insulin
levels, and triglycerides and decrease HDL cholesterol (Garg, Bantle et al., 1994;
Garg, 1998). There is considerable evidence that increasing dietary intake of
CHO accentuates the postprandial accumulation of TG-rich lipoproteins (Chen,
Coulston et al., 1995; Abbasi, McLaughlin et al., 2000; Kim, Abbasi et al., 2001).
Thus, the predictable effect of low fat/high CHO diets in insulin resistant and
hyperinsulinemic individuals will be both an increase in fasting plasma TG
concentration and accentuation of the daylong accumulation of TG-rich remnant
lipoproteins. Two meta-analyses have shown that low-fat/high-CHO diets
result in lower HDL-C concentrations (Mensink and Katan, 1992; Gardner and

Kraemer, 1995). There is also an independent relationship between insulin

58

resistance and hyperinsulinemia and HDL-C concentration (Laws and Reaven,
1992).

Studies have demonstrated a reduced risk of type 2 diabetes with increased
intakes of dietary fiber, especially cereal fiber. In a cohort of 42,759 men followed
for six years, cereal fiber was inversely associated with type 2 diabetes risk.
Similar findings were observed in a large cohort of women, where the relative risk
of the extreme quintiles of cereal fiber intake was 0.72 (CI, 0.58-0.90, P
trend=.001) (Salmeron, Manson et al., 1997). Data from the Nurses’ Health
Study supported a role of whole grains in protection against the development of
type 2 diabetes (Liu, Manson et al., 2000; Liu, Manson et al., 2000). Comparing
the highest to the lowest quintiles of whole grain intake, and adjusting for age and
energy intake, the relative risk was 0.62 (95% CI: 0.37, 1.04; P: for trend = 0.07),
whereas increased consumption of refined grains was associated with elevated
risk of diabetes.

Manipulation of the glycemic index has been suggested to limit the metabolic
consequences of the conventional Iow-fat/high-CHO diets. The rationale for the
concept of the glycemic load is that if carbohydrate digestion and absorption are
slowed then carbohydrate-rich foods will induce rapid perturbation of the
metabolic steady fasting state (Riccardi and Rivellese, 2000). The glycemic index
(GI) quantifies the glycemic response produced by a standard amount of a
carbohydrate-containing food relative to the response produced by the same
amount of carbohydrate from white bread or glucose (Costacou and Mayer-Davis,

2003; Wolever, 2003). The product of the glycemic index and the carbohydrate

59

content of a food (glycemic load, GL) is an indicator of the quality as well as the
quantity of dietary carbohydrate. Although the glycemic index has been criticized
for its quantitative approximation, it nevertheless differentiates among the
carbohydrate-rich foods as ‘fast’ or ‘lente’ depending on the rate of digestion and
absorption, which clinical studies have shown as having divergent effects on most
of the cardiovascular risk factors clustered in the metabolic syndrome
(Gustafsson, Asp et al., 1993; Jenkins and Jenkins, 1995). The detrimental effects
of a high-carbohydrate diet on plasma glucose/insulin, triglyceride/HDL, or
fibrinolysis occur only when carbohydrate foods with a high GI are consumed,
while they are diminished when the diet is largely based on fiber-rich, low-GI
foods (Garg, Bantle et al., 1994; Frost, Leeds et al., 1999; Jenkins, Kendall et al.,
2000).

Based on the studies in this section, it appears that replacement of saturated
fat with MUFA and/or PUFA, rather than CHO, will lower LDL-C concentrations
and not lead to a significant increase in plasma insulin concentrations and the
manifestations of the metabolic syndrome. However, as mentioned before the
anti-athrogenic effect of a MUFA-rich diet seems to be lost at a high total fat intake
(Rasmussen, Vessby et al., 2006). Furthermore, diets high in fiber from whole
grains and diets with a low glycemic load have been found to have beneficial
effects on plasma glucose and insulin. Aspects of all these factors together likely
relates to the beneficial effects of the food pattern called the Mediterranean diet

reviewed next on risk for the metabolic syndrome.

60

c. Mediterranean diet pattern
Although some dietary factors have been linked to individual features of

the metabolic syndrome, endorsement of an entire dietary pattern, the
Mediterranean diet, has come from the scientific advisory committee of the
American Heart Association for its ability to reduce or halt the progression of
cardiovascular disease and co-morbidities of the metabolic syndrome (Robertson
and Smaha, 2001; Esposito, Marfella et al., 2004). Evidence for beneficial
health effects of the Mediterranean diet emerged from the classic studies of Keys
that showed diets enriched in monounsaturated fat (MUFA) related to reduced
incidence of coronary heart disease (Keys, 1997). The Lyon Diet Heart Study
showed that a Mediterranean-type diet reduced the rate of recurrence after an
initial myocardial infarction (de Lorgeril, Salen et al., 1999). Recently, Esposito
et al. conducted a randomized, single-blind trial to determine the dietary effect on
metabolic syndrome (Esposito, Marfella et al., 2004). For two years, randomly
assigned patients with metabolic syndrome followed either a Mediterranean-style
diet of daily consumption of whole grains, fruits, vegetables, nuts, and olive oil or
a prudent diet (carbohydrates, 50%-60% kcal; proteins, 15%-20% kcal; total fat,
<30% kcal). Those patients on the Mediterranean-style diet demonstrated
greater improvement of endothelial function and a significant reduction of markers
of systemic vascular inflammation compared to those on the prudent diet.

Food aspects of the Mediterranean diet other than MUFAs that might be
beneficial are low intake of alcohol and high intakes of antioxidants and

polyphenols. Alcohol intake in excess (more than 30 g/d) can increase both

61

plasma triglyceride and blood pressure levels (Kiechl, Willeit et al., 1996).
Increased oxidative stress might contribute to endothelial inflammation that could
play a role in the etiology of both diabetes and CVD. Thus, interest has focused
on the ability of antioxidant vitamins and polyphenols (from fruits, vegetables, and
whole grains) to protect against inflammation (Riccardi and Rivellese, 2000).
3. Relationship of exercise and metabolic syndrome

Regular physical activity has many benefits, including a reduced risk of
cardiovascular and all-cause mortality (Ekelund, Haskell et al., 1988). Benefits of
physical activity have been attributed to improved weight control, reduced LDL
cholesterol, and increased insulin sensitivity in a dose-response fashion(Kelley
and Goodpaster, 2001). Using data from NHANES III Park et al. demonstrated
that physical inactivity significantly predicted the prevalence of metabolic
syndrome (Park, Zhu et al., 2003) . In a separate NHANES III analysis, strong and
significant inverse associations were reported between physical activity and self-
reported stroke and angina/myocardial infarction (Ford and Giles, 2000).

Increased glucose tolerance was seen after oral and intravenous glucose
loading (Reaven and Laws, 1999) prior to both acute exercise and physical
training. For young, middle- aged, and older adults, several epidemiological
studies have reported significant associations between physical
activity/cardiorespiratory fitness with insulin resistance and other components of
the metabolic syndrome (Eriksson, Taimela et al., 1997; Siscovick, Fried et al.,
1997; Mayer-Davis, D'Agostino et al., 1998). Within the population-based

Kuoppio study, men who engaged in at least moderate-intensity (34.5 metabolic

62

equivalents [METs]) leisure time physical activity of <10 hr/wk were 60% more
likely to have metabolic syndrome (WHO definition) than those engaging in _>_3.0
hr/wk, even after adjustment for confounding variables (age, socio-economic
status, smoking, and alcohol consumption) (Laaksonen, Lakka et al., 2002).

Two intervention studies deserve attention. Among African American and
Caucasian sedentary participants recruited for the Health Risk Factors Exercise
Training and Genetics Family Study, prevalence of the metabolic syndrome
(ATPIII) decreased to 11.8% from 16.9% after 20 weeks of exercise training
(Katzmarzyk, Leon et al., 2003). Of 105 participants with the metabolic syndrome
at base line, 30.5% (32 participants) were no longer classified as having the
metabolic syndrome after the exercise training. No sex or race differences in the
efficacy of exercise in treatment were observed. In the Da Qing Study in China,
participants with impaired glucose intolerance were assigned to a dietary,
exercise, or diet-plus-exercise intervention group. The six-year incidence of
diabetes ranged from 41-46% across the intervention groups compared to 67.7%
in the control group (Pan, Li et al., 1997). Diabetes risk was reduced by 31%,
46%, and 42% in the diet, exercise, and diet-plus-exercise interventions,
respectively. Ross and Katzmarzyk found that men and women with high
cardiorespiratory fitness had smaller skinfold thicknesses and less abdominal fat
(by waist circumference) for a given BMI compared with men and women with low
fitness levels (Ross and Katzmarzyk, 2003).

The exercise induced enhancement of insulin action in muscle is

predominantly due to local contraction-dependent mechanisms (Dela, Mikines et

63

al., 1992). Insulin increases muscle blood flow and this effect tends to increase
with training (Dela, Mikines et al., 1992), whereas muscle blood flow may be
reduced by inactivity. It is also known that the risk of having insulin resistant
syndrome is higher in those with lower levels of cardiorespiratory fitness (Farrell,
Cheng et al., 2004) and that oven/veight and obese individuals have lower levels
of fitness than normal weight individuals (Farrell, Braun et al., 2002).
Furthermore, recent studies have demonstrated that physically fit men and
women, when compared with physically unfit men and women, have a lower
amount of total and abdominal fat for a given BMI (Ross and Katzmarzyk, 2003;
Janssen, Katzmarzyk et al., 2004). Thus, cardiorespiratory fitness may play an
important role in explaining the increased risk of morbidity and mortality
associated with metabolic syndrome across levels of body weight.

Lifestyle therapies are first-line interventions to reduce the metabolic
syndrome and individual metabolic risk factors (Grundy, Cleeman et al., 2005).
The major lifestyle interventions include weight loss in overweight or obese
subjects, increased physical activity, and modification of an atherogenic diet
(Table 2.5). These changes will produce a reduction in all of the metabolic risk
factors simultaneously. In the long run, the greatest benefit for those with the
metabolic syndrome will be derived from an effective lifestyle intervention.

The accumulation of evidence demonstrating that obesity and physical
inactivity, both modifiable, are the most important contributors in the development
of metabolic syndrome (Franz, Bantle et al., 2002). Therefore, it is important to

include dietary factors and physical activity within analytic models of risk

64

exposures and atherosclerotic cardiovascular disease outcomes, especially in
studies where dietary factors and physical activity measurements are available

and have been previously related with the study variables of interest.

Table 2.5. Treatment of Lifestyle Risk Factors for Long-Term Prevention of

atherosclerotic cardiovascular disease

or Prevention/Treatment of Type 2 Diabetes

 

Theraputic Target and Goals of
Therapy

Therapeutic Recommendations

 

Abdominal obesity

Goal: Reduce body weight by 7-10%
during first year of therapy. Continue
weight loss thereafter to extent possible
with goal to ultimately achieve desirable
weight (BMI<25kg/m2)

Consistently encourage weight
maintenance/reduction through
appropriate balance of physical activity,
caloric intake, and formal behavioral
programs when indicated to
maintain/achieve waist circumference
of <40 inches in men and <35 inches in
women. Aim initially at slow reduction of
~7-10% from baseline weight. Even
small amounts of weight loss are
associated with significant health
benefits.

 

Physical inactivity

Goal: Regular moderate-intensity
physical activity; at least 30 min of
continuous/intermittent (preferably 60
min) 5 d/wk, but preferably daily

In patients with established CVD,
assess risk with detailed physical
activity history and/or exercise test, to
guide preparation. Encourage 30-60
min moderate-intensity aerobic activity
(eg, brisk walking), preferably daily,
supplemented by increase in daily
lifestyle activities (eg, pedometer step
tracking, walking breaks at work,
gardening, household work). Higher
exercise times achieved by
accumulating exercise throughout day.
Encourage resistance training 2d/wk.
Advise medically supervised programs
for high risk patients (eg, recent acute
coronary syndrome or
revascularization, CHF).

 

Atherogenic diet
Goal: Reduce intakes of saturated fat,
trans fat, cholesterol

 

Recommendations: Saturated fat <7%
of total calories; reduce trans fat;
dietary cholesterol <200mg/d; total fat
25-35% of total calories. Most dietary
fat should be unsaturated, simple
sugars should be limited.

 

 

con estive heart failure.

j BMI indicates body mass index; CVD cardiovascular disease, and CHF,

Adapted from (Grundy, Cleeman et al., 2005).

66

 

F. Methods use to assess body fatness.

Obesity- related studies have used different anthropometric indices to assess
body fatness, including body mass index (BMI), waist circumference (WC),
waist-to-hip ratio (WHR), percent body fat measured by bioelectrical impedance
analysis (BIA), and subscapular thickness or triceps skin fold 0' SF)
measurement. However, specificity and adequacy of these indicators is still
controversial because they are indirect and do not allow a precise assessment of
body composition (Deurenberg, Yap et al., 1998; Prentice and Jebb, 2001). Direct
and accurate measurement of adiposity relies on complex technologies such as
dual energy X—ray absorptiometry (DEXA) or computed tomography that has little
epidemiological applicability due to high costs and methodological efforts needed.
This part of the literature review will focus on the definitions and uses BMI, WC,
and BIA as well as their limitations and strengths. Important considerations of
using these measures as indicators to interpreted obesity and related risk factors
will be discussed.

1. Body Mass Index (BMI)

Body mass index, or Quetelet Index, is a statistical measure of the weight of
a person scaled according to height. It was invented between 1830 and 1850 by
the Belgian polymath Adolphe Quetelet (Quetelet, 1973). BMI is defined as the
individual's _bgdv weight (Kg) divided by the square of their height (m2). BMI
became popular in the late 1980s, as obesity started to become an issue of

concern in Western society. Most medical professionals have been educated in its

67

use, and BMI has made significant progress into general public awareness and is
gaining acceptance by the mass media (Seidell, Kahn et al., 2001).

The main assumption of BMI is that body mass, adjusted for stature squared,
is closely associated with body fatness and consequent morbidity and
mortality(Bray, 2003). Because BMI does not measure the fat mass or
percentage of fat, the Expert Committee of WHO proposed classification of BMI in
terms of degrees of overweight rather than degrees of obesity, with the cut-off
points 25, 30, and 40 based principally on the association between BMI and
mortality (WHO, 1995). National Institutes of Health (NIH) and the World Health
Organization (WHO) then adopted similar body weight guidelines for overweight
and obesity (WHO, 1998; DHHS, 2000). According to the WHO, a BMI of 2 25.0
is classified as overweight and a BMI of 2 30.0 as obesity.

BMI has been used extensively to classify obesity highlighting associated
health risks. Based on NHANES III data, Park et al. (2003) reported a prevalence
of metabolic syndrome in 4.8%, 22.8%, and 60.2% of normal-weight (BMI:
18.5-24.9), overweight (25.0—29.9), and obese (2 30) men in the US population,
respectively. The same prevalence pattern were also found in US women (Park,
Zhu et al., 2003). The Asia Pacific Cohort Studies Collaboration (Ni Mhurchu,
Rodgers et al., 2004), using individual participant data from 33 prospective
studies in the Asia-Pacific region, showed that there were continuous positive
associations between baseline BMI and the risks of ischemic stroke, hemorrhagic

stroke, and ischemic heart disease (IHD), with each 2 kg/m2 lower BMI associated

68

with a 12% lower risk of ischemic stroke, 8% lower risk in hemorrhagic stroke, and
11% lower risk of IHD.

BMI is moderately correlated with body fatness but does not quantify total
body adiposity (Kuczmarski and Flegal, 2000). Because the BMI is dependent
only upon net weight and height, it makes simplistic assumptions about
distribution of muscle and bone mass, and thus may overestimate adiposity on
muscular individuals, e.g. athletes, while underestimating adiposity on those with
less lean body mass, eg. the elderly (Prentice and Jebb, 2001). This raises
concerns about over diagnosis of obesity and hence inappropriate attempts at
weight reduction. Moreover, for those with normal BMI (BMI=18.5-24.9) but with
higher total percent of body fat, physically inactive people may underestimate
their risks for obesity related disorders. Tanaka et al., using the sample of white
people aged from 17 to 60 y from the Quebec Family Study, showed that normal
BMI males with elevated adiposity had higher prevalence of risk factors than did
normal BMI males with less adiposity, and the prevalence in the former was rather
similar to that seen in overweight males (Tanaka, Togashi et al., 2002).

As discussed in the early part of this literature review, people with central
adiposity have higher insulin levels and are more insulin resistant than subjects at
a similar weight with a peripheral type of obesity. BMI does not take into account
the distribution of fat in the body. Studies have shown that lean individuals with
normal BMls can also have a significant accumulation of VAT with increased risk
factors for cardiovascular disease and diabetes (metabolically obese normal

weight) (Ruderman, Chisholm et al., 1998; Katsuki, Sumida et al., 2004). Aging is

69

accompanied with a progressive increase in body fat and decrease in lean body
mass. This occurs even in individuals who manage to maintain a constant BMI as
they become older. Thus, the relationship between BMI and body fat is
age-dependent (Prentice and Jebb, 2001). Because body fat is more likely to be
deposited in the abdominal cavity with increasing age, it is claimed that BMI
becomes a poor indicator of overall and abdominal fatness in older persons
(Seidell and Visscher, 2000).

In recent years there has been an increasing awareness that certain ethnic
groups display a very different relationship between BMI and body fat compared
to that described for Caucasians. The universal cutoff points for BMI do not apply
equally well for all racial groups. Many Asian races, such as Indian and Chinese,
tend to carry a proportionately higher fat mass for a given BMI than Caucasians
(Deurenberg-Yap, Chew et al., 2002; Deurenberg-Yap and Deurenberg, 2003;
Wildman, Gu et al., 2004). Data from the Singapore National Nutrition Survey
showed that at any given percent body fat, the BMI of Singaporeans (consisting
mainly Indian, Malay, and Chinese) was about 3kg/m2 lower than that of
Caucasians. In fact, the BMI cutoff points for ovewveight and obesity were based
on mainly those for Caucasians. Many Asian countries established their own
cut-offs; Singapore used BMI 23 and 27 as the cut-off for overweight and obesity
(Deurenberg-Yap, Chew et al., 2002), Taiwan used 24, 27(Yeh, Chang et al.,
2005) respectively. Japan uses BMI greater or equal to 25 as an obesity cut point

(Kanazawa, Yoshiike et al., 2002).

70

The BMI has been an invaluable tool to the backbone of the obesity
classification system and surveillance statistics. Even though BMI are available to
describe overweight for the medical management and treatment of overweight
and obesity, BMI alone should not be considered diagnostic. Additional risk
factors should be assessed, including the presence of abdominal obesity based
on a measure of waist circumference and the presence of concomitant risk factors
or comorbidities, such as hypercholesterolemia or diabetes (Kuczmarski and

Flegal, 2000).

2. Waist Circumference

Increased risk of cardiovascular disease has been found in individuals
presented with distribution of excess fat in the abdominal region. There is no
standard measure of abdominal obesity that is widely accepted at present. Waist
circumference (WC) has long been used as a surrogate measure of total
abdominal adiposity. Lean et al. (1995) first proposed the use of waist
circumference (WC) as part of clinical cardiovascular risk assessments and
interpretation of health risks associated with adiposity. Lean et al. (1995) have
found that WC 3 102 cm for men or 3 88 cm for women identified subjects with
body mass index 3 30. According to their results, Lean et al. suggested men with
waist circumference 3 94 cm and women with waist circumference 3 80 cm
should gain no further weight (waist action level 1), and men with waist
circumference 3 102 cm and women with waist circumference 3 88 cm should

reduce their weight (waist action level 2). The International Diabetes Foundation

71

(IDF) currently identifies WC limit as one of the necessary requirements for the
diagnosis of metabolic syndrome (Zimmet, Alberti et al., 2005).
Recommendations are based on ethnic origin; for men: Europids 3 94 cm, South
Asians and Chinese 3 90, Japanese 3 85; for women: Europids, South Asian and
Chinese 2 80 cm (Zimmet, Alberti et al., 2005).

The majority of current studies agree that waist circumference (WC) is
probably a better indicator of abdominal fatness and cardiovascular disease than
either BMI or waist-to-hip ratio (WHR)(Reeder, Senthilselvan et al., 1997;
Dobbelsteyn, Joffres et al., 2001). In the NHANES III study and in population
studies from Canada, Hong Kong, and Japan, WC was more closely related to
metabolic risk factors than was BMI (Ho, Lam et al., 2003; Zhu, Wang et al., 2003;
Janssen, Katzmarzyk et al., 2004). Besides identifying individuals with
cardiovascular disease risk factors, the WC measure has several advantages over
both BMI and WHR. Simplicity, low cost, acceptable accuracy, and ease of
interpretation have led to the use of WC measurement compared to BMI and WHR
in a clinical setting. It requires only the use of a tape measure, alleviating the
expense of the equipment and space needed to measure height and weight. A
single measurement (WC) as opposed to the ratio of two measures (WHR), is less
susceptible to measurement and calculation errors (Dobbelsteyn, Joffres et al.,
2001). Studies have shown a correlation between the change in WC and
abdominal fat mass for women and visceral fat mass for men and women (Ross,

Janssen et al., 2004) and a decrease in WC has been reported to correlate with

72

improvements in glucose metabolism. For these reasons, WC is an easy and
cost-effective field measure.

However, WC is not without limitations in public health applications (Seidell,
Kahn et al., 2001). The limitation .of WC is the inability to differentiate
subcutaneous adipose tissue from visceral adipose tissue. WC is a direct measure
of length rather than of area or volume. Defining fat distribution in the abdominal
region based on WC requires transformation of length into separate area or
volume (He, Engelson et al., 2004). Circumferences reflect internal as well as
subcutaneous adipose tissue and are influenced by variation in muscle and bone.
Waist circumference has been shown to be highly correlated with intra-abdominal
and total fat mass.

Although many investigators believe that waist circumference serves as a
surrogate for visceral fat, it still has a substantial contribution from metabolically
different subcutaneous fat (Misra, Wasir et al., 2005). Waist circumference seems
to quantify subcutaneous fat better than visceral fat (Bonora, Micciolo et al., 1995).
Furthermore, correlation of waist circumference to visceral fat depends on sex and
ethnicity because it is weaker in women and blacks (Weinsier, Hunter et al., 2001).
Weinsier et al. studied 46 white and black women during a weight loss program
and found that white women lost more intraabdominal adipose tissue (p < 0.001)
and less subcutaneous abdominal adipose tissue (p < 0.03) than did black women.
In this study, changes in waist circumference correlated with changes in
intraabdominal adipose tissue in white women but not in black women.

Investigators concluded that waist circumference was not a suitable surrogate

73

marker for tracking changes in the visceral fat compartment in black women
(Weinsier, Hunter et al., 2001).

The lack of a standard protocol for measurement for WC is unfortunate. WC
varies between subjects in relation to bone landmarks, muscle, adipose tissue, or
may be difficult to identify in obese subjects. In the Taiwan National Nutrition and
Health Survey (NAHSIT) and Canadian Heart Health Surveys (CHHS), WC was
measured at the visible narrowing of the waist after a normal exhalation, or in
extreme cases of obesity, at the level of the 12th rib (Wang, Thornton et al., 2000;
Yeh, Chang et al., 2005). In Japan’s national nutritional survey, the WC was
measured at the height of the navel (Nishimura, Nakagami et al., 2007), but in the
diagnostic criteria of the International Diabetes Foundation (IDF) it was measured
at the midpoint between the subcostal margin and the margin of the supracristal
plane (Zimmet, Alberti et al., 2005). In the United States, technicians in NHANES
III measured WC immediately above the iliac crest (DHHS, 1997). Several papers
reported considerable difference among different WC measure protocols (Wang,
Thornton et al., 2003; Bigaard, Spanggaard et al., 2005). In Wang’s study, there
was a statistically significant difference (-1.7 cm) between WC measured
immediately above the iliac crest (as in NHANES) and WC measured at the
narrowest waist (as in the CHHS) in men. In women, the difference between
measurement of WC at the iliac crest and narrowest waist was -4.6 cm. These
discrepancies in the position of WC measurement make comparisons with other

studies difficult. There is an urgent need to establish a single, universal bone

74

landmark to make the measurement of WC highly reliable and reproducible
(Seidell, Kahn et al., 2001).

It must be acknowledged that BMI and WC are highly correlated with each '
other, at least within the ranges typically observed in industrialized societies.
Because of this close correlation, it is difficult to separate the effects that each
may have on health (Seidell, Kahn et al., 2001) . The US. NIH has recommended
a two-tiered classification system using both BMI and WC (NIH, 2000). BMI and
WC might be useful as a first step in determining health risks that can be
confirmed by more complex and costly tests, such as blood analyses or

physiologic challenges.

3. Percent body measured by bioelectrical impedance analysis (BIA)
Obesity is characterized by an excess amount of body fat, to the extent that
heath may be impaired. Obesity as excess body fat was defined if percent body
fat was greater than 25% in male and 35% in female subjects (WHO, 1995).
However, the direct measurement of the body’s fat mass is only possible via
dissection and chemical analyses. Even indirect techniques like computed
tomography (CT) and dual-energy x-ray absorptiometry (DEXA) remains a
significant challenge for most body composition techniques. Several instruments
have been developed to measure total body fat and its distribution, such as
computer tomography (CT), magnetic resonance imaging (MRI) and dual-energy
X-ray absorptiometry (DEXA). In general, these techniques are more accurate

methods than anthropometric measurements for the assessment of body fat. But

75

these research techniques are not useful tools in large populations because of the
high initial capital investment, the need for a highly trained technical staff, and the
high annual maintenance and service costs (Ellis, 2001). Bioelectrical
impedance analysis (BIA) promises at least one measure of body composition
that can be applied to large populations. BIA is inexpensive, easy to use, free of
observer bias, and precise (Lukaski, Bolonchuk et al., 1986; Roubenoff and
Kehayias, 1991).

Bioelectrical impedance analysis (BIA) evaluates the electrical conductivity
characteristics of the human body (Foster and Lukaski, 1996). In particular,
single-frequency BIA evaluates resistance (the pure opposition of tissues to the
flow of electrons), reactance (related to the capacitance of cell membranes, tissue
interfaces, etc.), and phase angle (the shift between current and voltage). From a
nutritional standpoint, resistance (even if more associated with height in different
indexes) is considered to be related to total body water while phase angle
probably reflects the variability in body cell mass and the ratio intracellular
water/extracellular water (Foster and Lukaski, 1996). BIA measures of
resistance and impedance are proportional to body water volume, if body
electrolyte status is normal, and to the length of the conductor or stature (eg,
staturezl resistance). Based on the principles of BIA measurement, to estimate
the volume of total body water (TBW), three assumptions are used: the whole
body acts like a cylindrical conductor, the conductor’s length is proportional to the
subject’s height and the reactance component of the voltage signal can be

disregarded (Ellis, 2001). BIA uses regression analysis to derive prediction

76

models to estimate TBW and fat-free mass (F FM) (Guo, Roche et al., 1989;
Chumlea and Guo, 1994).

A large number of BIA equations in the literature predict F FM. These
equations vary in the parameters. included in the multiple regression equations
and their applicability to various subjects. Early BIA equations (before 1987) only
included heightzlresistance. Later equations include other parameters, such as
weight, age, gender, reactance, and anthropometric measurements of the trunk
and/or extremities to improve the prediction accuracy. Currently, the equations for
men and for women are as follows (Zhu, Wang et al., 2003):

Men:
FFM (kg) = -10.68 + 0.65 - 8le + 0.26 x weight + 0.02 x R

Women:

FFM (kg) = 0.53 + 0.69 x SZIR + 0.17 x weight + 0.02 x R
%BF = 100 - [(weight - FFM)/weight]

where 8le is stature squared divided by resistance (cm2/Q).

The use of BIA has increased because the equipment is portable and safe,
the procedure is simple and noninvasive, and the results are reproducible and
rapidly obtained (Kyle, Bosaeus et al., 2004). BMI and percent body fat
measured by BIA are highly correlated. Studies have shown that the correlation
coefficients between BMI and BIA are from 0.75 to 0.95 in different gender and
age groups (Nagaya, Yoshida et al., 1999). As previously reported, high BMI and

high percent body were strongly associated with high serum TC/HDLC ratio and

77

TG, and low serum HDLC rather than high serum TC and LDLC. Nagaya et al.
compared the BMI and percent body fat measured by BIA found that percent body
fat by BIA was best correlated with the serum indices, except for serum HDLC in
both sexes (Nagaya, Yoshida et al., 1999).

Using the BIA measured values from NHANES III sample compared to the
percent body fat determined by DXA, hydrostatic weighing, or data collected using
other reference body composition methods in other studies. The means reported
for the systematic differences among body composition methods and the
between-method limits of agreement are approximately 2.0 kg for FFM and 3 -5%
for percent body fat and should be considered when interpreting the data or
making inferences (Chumlea, Guo et al., 2002).

Although the daily precision of BIA is good, the accuracy for the assessment
of an individual’s body fatness remains unclear, due to body dehydration and
temperature. In conclusion, body fat can be determined by BIA provided that
hydration is normal and BIA equations used are applicable to the study
population, with regard to gender, age, and ethnic group (Kyle, Bosaeus et al.,
2004).

It is now accepted that the distribution of body fat is an important
determinant of metabolic abnormalities, possibly more so than the degree of
excess weight as measured by BMI, nor the percent body fat measured by BIA
(Walton, Lees et al., 1995). In particular, intraabdominal obesity or visceral fat is
strongly associated with metabolic disturbances and insulin resistance (Pouliot,

Despres et al., 1994), which could not be distinguished by simple WC

78

measurement. A combination of two or more screening indices should be able to
help health professionals target those in the greatest risk for MS and clarify

whether the regional or total adiposity contribute most to the etiology of MS.

79

CHAPTER THREE

METHODS

Study design

This study was a cross-sectional secondary data analysis of adults classified
as overweight, but not obese (BMI 25-29.9 in US and BMI 24-27 in Taiwan).
Data were obtained from two representative national health surveys, one from the
US (for non-Hispanic whites) and one from Taiwan (for those of Chinese
ancestry). The association and contribution of abdominal and total body
adiposity to the risks for modified metabolic syndrome and related disorder
(elevated triglycerides, low HDL cholesterol, high blood pressure, and impaired
fasting glucose) were assessed in both samples. Subjects were divided into
eight different anthropometric categories based on their waist circumference and
percent body fat (Figure 1.1). The odds ratio were calculated to determine the
odds of different body fat distribution patterns to modified metabolic syndrome
and each of the metabolic syndrome- related disorders and to compare the
race/ethnicity-specific differences across the non-Hispanic white and Taiwanese
populations. The main outcome measures were the presence of modified
metabolic syndrome (defined using NCEP ATP Ill guidelines, as two of four
disorders, not including high waist circumference), serum triglyceride (TG), low

high-density lipoprotein cholesterol (HDL), high blood pressure (BP), and elevated

80

fasting blood glucose, controlling for demographic factors, physical activity, and

energy from fat or from carbohydrates.

Subjects and Datasets

Two data sets were used in this study; NHANES III data were used to examine
Aim 1 and 2 for non-Hispanic whites. Both NHANES III and Nutrition and Health
Survey in Taiwan (NAHSIT) data were used to determine racial differences (for
non-Hispanic whites compared to Taiwanese) in body fat distribution in modified
metabolic syndrome and metabolic syndrome related disorders. Thi3rc_l Nationa_l_
filth and Nutritiongamination Survey (NHANES III) 1988—1994QHH38, 1996).
The National Center for Health Statistics (NCHS) conducted this survey to obtain
nationally representative information on the health and nutritional status of the US
population. NHANES III was conducted at 89 locations in 2 phases: 1988—1991
and 1991—1994. NHANES III used a stratified, multi-stage probability cluster
sampling, with each respondent’s data weighted by the probability of being
sampled. The results are representative of the US non-institutionalized
population, two months of age or over from the 50 states and the District of
Columbia (NCHS., 2007).

In NHANES III, standardized medical examinations were completed in
mobile centers that traveled to 89 pre-selected sites in the continental US. The
medical examinations included a blood chemistry panel and measurements of
blood pressure, plasma lipid and glucose concentrations, and anthropometric

measurements. Besides the medical examination, the NHANES III staff

81

conducted surveys in households, administering questionnaires to families, adults
and children. The household surveys included questions about demographics,
socioeconomic factors, dietary intakes, and health histories. The number of
sampled persons for this survey were 39,695 with 33,994 (86%) of these
interviewed and 30,818 (78%) both interviewed and examined (DHHS, 1996).
This NHANES III survey received human subject approval from the Centers for
Disease Control and Prevention. Details of the survey design and
questionnaires are published in the NHANES III Plan and Operation Reference
Manual (DHHS, 1997). The present study used data from the NHANES III instead
of NHANES 1999-2000 for two reasons. First, the protocol was nearly the same in
both NHANES III and NAHSIT, because the NAHSIT followed the NHANES III
design and procedures. Secondly, NHANESIII had the timeframe most
comparable to that of the NAHSIT.

N1utrition and Health Survey in Taiwan (NAHSIT), 1993-1996.

The NAHSIT was an island-wide survey in Taiwan of non-institutionalized
residents four years and older selected by a multi-stage complex sample design
(Pan, Yeh et al., 2003). A complex sampling scheme was used in th NAHSIT
survey, parallel to that of the NHANESIII. The 359 townships or city districts in
Taiwan were classified into seven strata according to their geographical locations
and degree of urbanization. A total sample of 9,961 participants was obtained
from door-to-door interviews, for a 74% response rate. Among those interviewed,
64.9% (6,464 individuals) participated and completed the physical examination.

The survey consisted of two parts: The health component included questionnaires

on lifestyle, disease history and measurement of blood pressure,
electrocardiogram, biochemical values, clinical signs, oral glucose tolerance test,
and anthropometric measures. Clinical data and anthropometric measurements
were obtained from a physical examination carried out in temporary clinics set- up
in the neighborhood of the survey sites. The dietary and nutrition component
included interviews for 24-hour dietary recalls, food frequency, nutrient
supplements, vegetarian diets, nutrition-related knowledge/attitude/practices, and
biochemical measurements for nutritional indicators. Details on the design and
operation of the survey have been published (Lin, Yen et al., 2003; Pan, Flegal et
al., 2004).

Myth samples. The original data were retrieved from the NHANES III CD ROM
(NCHS., 1998) and the NAHSIT data downloaded from
Mrdasinica.edu.tw/webpages/nahsit/index.htm. In this study, subjects from
both surveys were limited to those aged 20-74 years of age and having a
BMI=25-29.9 in NHANES III and BMI=24-27 in NAHSIT, respectively. To
compare the specific racial differences in metabolic syndrome related disorders,
only non-Hispanic whites were included from the NHANES III survey and only
those of Chinese descent were included from the NAHSIT. In NAHSIT, the
aboriginal island population, accounting for approximately 1.7% of the total
sample, was excluded from this study. Most of the aborigines live in the
mountainous areas and are thought to resemble the Malayo-Polynesians in their

genetic make-up (Lin, Yen et al., 2003) differing from the main population of

83

Taiwan who immigrated from mainland China in several waves over the last 200
years.

Participants’ data included anthropometric variables, blood pressure,
socioeconomic status, lifestyle-related information, blood studies and body
composition values from bioelectric impedance analysis (BIA). Further exclusion
criteria were any participants with missing data values, who consumed food or
beverages other than water within six hours before the venipuncture, and women
pregnant or lactating at baseline. Sample sizes from NHANES and NAHSIT
were approximately 1600 and 1000, respectively. Before analyzing the data,
approval from the university’s Institutional Review Board was obtained (Appendix
2).

Procedures

Three types of variables and categories of data were included in the analyses
and are shown in Table 3.1. A detailed description of how measurements for
these variables were collected and analyzed is found in Appendix 1.

%pendent variables were developed from the NHANES III and NAHSIT’s
medical examination data for serum triglycerides (TG), and high density
lipoprotein (HDL), systolic and diastolic blood pressures, and fasting plasma
glucose. Metabolic syndrome- related disorders were defined according to the
latest National Cholesterol Education Program guidelines (NCEP ATP lll)
(NCEP., 2001). That is, dyslipidemia was defined as high TG (serum TG 3150
mgldL) and low HDL cholesterol (<40 mgldL in men and <50 mgldl in women).

High blood pressure was defined as systolic blood pressure 3130 mm Hg,

84

diastolic blood pressure 385 mm Hg, or reports of a history of hypertension or
currently taking blood pressure medication, regardless of measured blood
pressure values. Impaired fasting glucose was defined as fasting plasma
glucose concentration 3100 mgldl (Grundy, Cleeman et al., 2005), previous
physician’s diagnosis of diabetes, or use of hypoglycemic medication. Modified
metabolic syndrome (MMS) was defined as having two or more of these
disorders.

Although the National Cholesterol Education Program guidelines include
high waist circumference as a component of the metabolic syndrome, in this
analysis the diagnosis of the metabolic syndrome does not include a high WC,
hence the name Modified Metabolic Syndrome. Waist circumference was used
instead as a predictor of the Metabolic Syndrome (Okosun, Liao et al., 2000;
Janssen, Katzmarzyk et al., 2004). Individuals who reported currently using
anti-hypertensive or anti-diabetic medication (insulin or hypoglycemic oral agents)
were defined as having hypertension or high fasting glucose regardless of
measured blood pressure and plasma glucose values.

Mndent variables in this study included WC and percent body fat (%BF)
as determined from bioelectrical impedance. WC and %BF were used to construct
the eight anthropometric categories according to normal and high values for WC
and quartiles of percent body fat (%BF). The cutpoints were 102 cm for men and
88 cm for women for the normal and high WC (NCEP, 2002). In this study, BMI
was calculated as the weight of the individual in kilograms divided by the height in

meters squared. Based on BMI, subjects were limited to those classified as

85

overweight with a BMI of 25—29.9, for both men and women (NIH, 1998). This
definition is also consistent with those of the World Health Organization (Lean,
Han et al., 1995).

Covariates from the demographic and lifestyle data included age, education,
gender, race, smoking and alcohol drinking status, and physical activity. Dietary
covariates were the percent of energy from total carbohydrates and from fat
extracted from the 24-hour dietary recall in each survey

The covariate definitions for the NHANES data are described in the following
paragraphs. The definitions for smoking and drinking are identical in NAHSIT to
NHANES.

Smoking was categorized as current, past, or never. Subjects were considered
current smokers if they smoked cigarettes, cigars, or tobacco at the time of the
interview; previous smokers if they were not current smokers, but had smoked
3100 cigarettes in their entire life; and nonsmokers if they had smoked less than
those amounts (Janssen et al., 2004). For this study, alcohol consumption was
graded as non-drinker, if subjects drank zero drinks of beer, wine or liquor in the
past month, moderate drinker if they drank 1—15 drinks, or heavy drinkers if they
drank >15 drinks.

Physical activity was based on the subject’s self- reported frequencies
involving common leisure physical activities in the past month in both data sets.
The activities in NHANES III included: walking a mile without stopping, jogging or
running, swimming, regular dancing, aerobic exercise, aerobic dancing, riding a

bicycle or exercise bicycle, calisthenics, garden or yard work, or weight lifting. The

86

activities in NAHSIT included: riding a bicycle, ball games, gymnastics,
punching/boxing, swimming, aerobic dancing/country dancing, mountain climbing,
jogging, walking, gardening, or housework. Respondents engaging in no activity
in a weekly period, less than one activity/week, 2-5 activities/week, and greater
than 5 activities/week were divided into ‘Physically inactive’, ‘light’, ‘moderately’,
and ‘physically active’ categories, respectively (Janssen, Katzmarzyk et al.,
2004). - Construction of rank order variables in NAHSIT from continuous data for
education and income are described in Appendix 1. Education level from
NHANES data was divided into three categories: 58 years, 9 to 12 years, and
more than 12 years of education. Dietary factors will use the energy intake from
carbohydrate and fat as covariates. Carbohydrate and fat were expressed as a
percentage of total energy, as one relevant measure of dietary composition that

might relate to risk of metabolic related disorders.

Statistical Analyses

Statistical analyses were conducted using the Statistical Package for the
Social Sciences, SPSS for Windows, version 14.0 (SPSS Inc., Chicago, IL,
USA,). All analyses were carried out with SUDAAN version 9 (Research
Triangle Institute, Research Triangle Park, NC), to account for the effect of the
complex sampling design in statistical testing and to derive means and standard
errors. Variance estimation for the statistics computed by SUDAAN used the
Taylor series linearization method or replication methods such as Jackknife.

Missing values were handled separately for each metabolic disorder in the

87

analysis. A p-value <0.05 was used to detect statistical significance.

Descriptive analyses included mean, standard deviation and standard error of the
mean for estimates of BMI, WC, %BF, TG, HDL, systolic and diastolic blood
pressure and fasting blood glucose. Values were calculated for each sex and
racial-ethnic group.

For Aim 1, an adjusted Wald test (Zhu, Wang et al., 2003) was used to
compare prevalence of modified MS, high TG, low HDL, hypertension and
impaired fasting glucose between high and normal waist circumference and
quartile categories of body fat. The data were age and sex standardized to the
1980 US population and design effect-adjusted standard errors were computed.
Multiple logistic regression analysis was used to determine the association of
waist circumference and percent body fat to the risk of modified metabolic
syndrome and each of the metabolic syndrome related disorders. This regression
was also to determine whether the percent body fat adds to the predictive power
of waist circumference to detect risk for modified metabolic syndrome and/or each
metabolic syndrome related disorder.

For Aim 2, anthropometric combinations were created based on high or
normal waist circumference and quartile of percent body fat to compare the odds
ratios for the modified metabolic syndrome and each of the metabolic syndrome
related disorders using multiple logistic regression analysis for men and women
adjusted by age, education, smoking, drinking habits, dietary factors, and physical
activity status. The reference cut-points for waist circumference was set below

102 cm for men and below 88 cm for women. The reference group was those

88

with normal waist circumference (WC) and having body fatness falling in the 50th

percentile (LF NW). The main comparisons of interest for odds ratios were three

groups: those with high body fat (>50th %tile) and high waist circumference
(HFHVV); those with low body fat (550th %tile) and high waist circumference
(LFHW); and those with high body fat (>50th %tile) and normal waist
circumference (HFNW) and low body fat (550th %tile)

For Aim3, H3.1 and H3.2, the prevalence of the modified metabolic syndrome and

each of the metabolic syndrome-related disorders were determined for the

Taiwanese people. The adjusted Wald Chi-square test was used to evaluate the

statistical significance of the prevalence rate. H33, logistic regression was used

to compare the prevalence rate ratios (odds ratios) in each of the four body fat
distribution categories for modified metabolic syndrome and each of the metabolic
syndrome related disorders across gender groups. The dependent variables
were presence of modified metabolic syndrome and the each of the metabolic

syndrome-related disorders. For independent variables, a variable of four

anthropometric categories based on high or normal waist circumference and 50th

quartile of percent body fat were used in this test, adjusted by age, education,

smoking, drinking habits, dietary factors, and physical activity status. 1

89

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

0.0.350? ”5.x. .o>> .0E.o:n0 H24.”...
0....x. 29.? "500.03 .0E.o: H>>zh=..
0.0.0903V “5503 .0E.o:00 H>>...h...
5.50.920 00.n0_.0>
00.00009 .00.. 6.5 0.0.02 ”2.2... 00000000 .0. .000 .0 000.: 0000000000.
0.0.x. 0.07 0>> E .I .r E E ..o. _0E.oz
0.000 smnv O>> U.0682 000E90505 .0
$0.0.
00002.50 .0 .0300... E0900.
:0Eo..> :_ E0 wwh .:0E :. E0 work 0>> H... .I .r E E __o. .0902
:0Eo..> :. Eo 00v .:0E :. Eo Norv O>> ..0E.oz 0:0E99.o5 .0 6.3. 00.n0_.0>
0:030:00 .0 00:90:03.6 .0.0>> E00:0a00:.
40.9: 007 :o..0..:0o:oo 008:.m 00:000.
.9. EE mwamem 0.2000... 0005 _r
_:0Eo.s 0Eo.0:>0 2.0090E 55>
:. 40.9: omv .0 :0E :_ ..0BE ovv :o..0..:0o:oo 6.900.000 .5... .0. 0000.050
...0\mE om 0M :o..0..:0o:oo 02.0029; 2.9.0.00: 59...; 5.2.2. 0Eo.0:>0
.m:..so..9 0... .o 0.9: .o N 001 030E9oco5 2.0090... 005022
.200. 0. V .00002 .0. 00.0 ..0. .0052
.03... 02k Em... 030E9oco5 00806 000.0. 9000“.
0... EE mwv .0852 m... E... 00A :91
05000.0 000.0. 2.9005 2.9005
0: 0:. 02100002 .0: 0.00 00.0 00.: _r 00.0 ..0. .00002 2.0.0.0
0.0000... 000... 2.90.5 0:0E9o:o5 05000.0 085
0.0E9 .9 .0. E omv 0.0E .9 .0. E ovv 26.. .0. cm... ..o. .0902
0.00:0. .9 _0.mE omM 0.0:. .9 ..0.mE ovM .0E.oz 0000:9205 .5...
..0. E 000A E .I .r E E ..o. .0902 0030...;
.39: omrv 00:09.6... E200 ..0E.oz 0:0E99.o5 002.0039; 0:00:0000
0:90.505 009C 00E02

 

 

003000., .0 .0... 0.0 0.000

90

 

>00 .00 00.000000 3.000 .99 E0.. .0. £00.00.
>00 .00 00E00000 3.000 .90. 05.. 90.02.0900 .00900

.0“.
90.030050

 

 

000000000 0.900. .0905
.o. 02.000.
02.00 2200.... 0 0. 3.2.00 00 0...... 09.000. ”02.000. .r .00...
200332280 000 00... 000. 09.000. 000... .N. 0.0.0005.
200.200.22.00 0N 09.000. ”90.0005. .0. 002.0...
200.200.22.00 0 00... .9005 09.000. 02.0.... 02.00900 3.2.00 .02030
.0000... 0.0. 0. 0. 0.0.. 0. .00000 00... 0... 000 000.0 0......
E0.. 0. .00000 0.00.E 00.5.0.0 0. .0E90..0 02.00000 0... 000.....00
0000.0...0 2.9.0 0 9 000 0. 0000.0...0 00... 02.00000 .0 0.00.. .m. 0.00.. N.

N. 00... 0.0E 000 0.00.. N.. 0. o. 0.00.. mm “00000900 00.... 0.0.
000.20 .....w...<z .00. 02.00000 .0 0.00.. N. 00... 0.0E 000 0.00...

000. 0.0E .N. 0.00.. NE...
T 02.00000 .0 0.00.. 0-0

 

 

 

 

 

 

N.. 0. m 0.00.. mm 00000900 00.... 0.0. 000.20 E mmz<Iz .00. 02.00900 02.0000m.
0.00E .000 0.... 00.50
.002. 0.0.. .0 .00.... ..000 00 200.0 00.5 000... “0.9.0.5002
0.00E
.000 0... 00.50 00.000 0. o. F 0002000 0000.. 0.0... .0 .00.; ..000 .r .02000002
:0... 000.050 2.00020 00 0.00.0 00.... 000... 0.9.0.5 0.0.0005. ..N. 0.0.000E
0.00E .000 00. 00.50 0000.. 0.0.. .0 .00.... ..m . >>000
..000 . 0000.050 2.00020 0 .A 0.00.0 0...... 0000. ”0.00.000 >>00I .0250900 00.205
.0E..0.._ 0. 0.000E0 000... 000.000. 0020E0 ..0.00.0E0000. .r .020E0-0oz
0... 02.00 ..0... 0. 000000. .0 000.0 8 .0 0.00.0 0N 09.9090 ..N. ”.000.
09M 000.0E0 000.00 0.0..0E0 .0950 .00 ..0.00.0E0 00050.0. ..m . 000.50
5.0.290. 00. .0 0E.. 0... .0 0020E0 ..0.00_0E0 .0050. 02.00900 000.0Em
.N. 0.0E0. ..r 0.0.2
02.00900 .0009...
0mm .0 0.00300 9 ON E0. H. 000000000 09.1 00.00050
00900.30 000.... 000.02

 

 

 

.0008...” 0.00.

91

CHAPTER 4
RESULTS

In this section, descriptive statistics of the study sample are first presented.
Next, each aim and hypothesis from the first chapter is repeated followed by the
data to refute or support it. Finally, data are shown that are relevant to the study
topic, but not directly pertinent to a hypothesis.
Demographics of the study subjects

Descriptive and metabolic characteristics are in Table 4.1 for the entire
sample of oven/veight (BMI 25-30) non-Hispanic White participants, ages from
20-74 years from the NHANES III. A total of 960 males and 674 females met the
selection criteria and were included. Within this overweight group, females were
about 4 years older than males. Males had a higher waist circumference (WC)
than females, but a lower percent body fat (%BF). For blood lipids, males had
higher triglycerides, systolic blood pressure and diastolic blood pressure; females
had higher HDL cholesterol. There was no difference in fasting glucose between

males and females.

92

Table 4.1. Descriptive characteristics and metabolic factors (mean:SE) in the

total sample of overweight (BMI 25-30) non-Hispanic White participants ages
20-74 yr from the National Health and Nutrition Examination Survey Ill, weighted

 

 

sample.
Men (n= 960) Women (n= 674) P value
Descriptive characteristics
A99 (Y) 44.2 :I: 0.5 48.5 i 0.8 “1001
3“" (kg/m2) 27.1 i 0.1 27.1 i 0.1 09‘”
Waist circumferences 975 i 0.6 91.7 i ()3 <0.001
(cm)
70 Bodyfat from BIA 24.9 i 0.3 37.7 i ()3 <0.001
Metabolic variables
Triglycerides (mgldl) 170.1: 0.8 1443 $4.6 <0.001
HDL cholesterol 434: 0.6 53.6 1' 0.6 <0.001
(mgldl)
Fasting Glucose 100.3: 1.4 98.2 i 1.5 0.333
(mgldl)
SystO'iC blood 124.0: 0.8 121.3 i 0.9 0005
pressure (mm Hg)
Diastolic blood 7743: 03 73.6 i 0.4 <0.001

pressure (mm Hg)

Table 4.2. Prevalence of metabolic risk factors in Non-Hispanic White men and
women ages 20-74 yr with BMI 25-30 from NHANES III, weighted sample.

Risk factors

Men

Women

 

% (95% CI)

% (95% Cl)

*

 

High Waist

circumference

Low HDL

thTG

High blood
pressure

High Fasting
Glucose
Metabolic Syndrome
Modified Metabolic

Syndrome

26.1(23.2,29.2)

45.2 (40.9, 49.5)
48.8 (41.1, 52.8)

43.7 (39.3, 48.3)

31.8 (27.7, 35.7)
32.1 (27.9, 38.8)
54.2 (49.1, 59.3)

87.1 (81.9, 71.9)

44.9 (40.8, 49.2)
32.3 (27.8, 37.3)

38.1 (33.9, 42.8)

22.7 (19.0, 28.9)
38.1 (31.0, 41.5)

39.5 (33.6, 45.7)

<0.001

0.935
0.000

0.070

0.006
0.165
<0.001

1 Based on NCEP ATPIII Criteria but not including high WC, i.e. having 2 or more of following
abnormalities: Triglycerides 3150 mgldL; HDL: Men <40 mg/dL Women <50 mg/dL; Blood
pressure 3130/385 mmHg; Fasting glucose 3100 mg/dL.

* Cochran-Mantel-Haenszel Chi-square Test

Table 4.2 shows the prevalence of metabolic risk factors for Non-Hispanic
white men and women ages 20—74 year with BMI=25<30. About two-thirds of
females had high WC. The prevalence of high WC was more than twice as high in
females compared to males. The twofold higher prevalence of elevated WC of
women in the present study might be expected to affect the use of WC to predict
the risk of MMS. In fact, when high WC was excluded as a criteria of M8, the
prevalence of MMS was higher in men than women (54.5% to 39.5%, P<0.01).
Males had a higher prevalence than did females of high serum triglycerides, high
fasting glucose and modified metabolic syndrome (MMS). There were no
significant differences between genders for the prevalence of low HDL or high
blood pressure. When high WC was excluded as a criterion, men had a higher

prevalence of MMS than women.

94

Aim 1. To determine whether percent body fat assessed by Bioelectrical
Impedance Analysis added to the predictive power of WC in risk for MMS and/or
each metabolic syndrome related disorder (MSRD) (elevated triglycerides, low
HDL cholesterol, high blood pressure, impaired fasting blood glucose) in

non-Hispanic white, oveniveight (BMI 25-29.9) men and women (NHANES III).

H1.1: Both WC and percent body fat predict risk for MMS and/or each MRSD

(elevated triglycerides, low HDL cholesterol, high blood pressure, and impaired

fasting glucose) in ovenNeight non-Hispanic white men and women.

H1.2: Percent body fat adds to the predictive power of WC to detect risk for MMS

and/or each MSRD in ovenNeight non-Hispanic white men and women (NHANES
III).

The correlations of BMI, WC and %BF to the modified metabolic disorders
are shown in Table 4.3. For this group of overweight, but not obese (BMI
25<30), non-Hispanic men and women, BMI was significantly correlated to both
WC (r=0.61, 0.44) and to %BF (r= 0.23, 0.47) respectively, with WC stronger for
men and %BF stronger for women. For both men and women, WC was weakly
to moderately associated to risk for most of the metabolic disorders. For men
%BF was weakly correlated with serum triglycerides and fasting blood glucose,
but only with HDL and blood pressure in women. WC was significantly and more
strongly associated with all MSRD compared to %BF in both men and women, but

no association exceeded r=0.38.

95

Table 4.3. Pearson’s correlations of BMI, waist circumference and percent

body fat to the metabolic disorders in Non-Hispanic White men and women
_ages 20—74 year with BMI 25-30 from NHANES III, wgighted sample.

 

 

 

Men (n=960) Women (n=674)
WC %BF WC %BF
Body Mass Index 061* 023* 044* 047*
(0.000) (0.000) (0.000) (0.000)
Waist 1.00 039* 1.00 036*
Circumference (0.000) (0.000)
Percent Body Fat 039* 1.00 036* 1.00
(%BF) (0.000) (0.000)
Serum 022* 010* 022* 0.06
triglycerides (0.000) (0.013) (0.000) (0.286)
(mgldL)
Serum HDL -0.11* 0.00 -0.03* -008*
(mg/dL) (0.004) (0.888) (0.563) (0.023)
Systolic blood 023* 0.07 038* 010*
pressure (mm Hg) (0.000) (0.085) (0.000) (0.011)
Diastolicblood 015* 0.05 017* 011*
pressure (mm Hg) (0.000) (0.096) (0.001) (0.010)
Fasting plasma 017* 009* 023* 0.06
glucose (mg/d L) (0.000) (0.010) (0.001) (0.120)

 

*Correiation is significant at the 0.05 level (2-tailed).

**

Correlation is significant at the 0.01 level (2-tailed).

96

Multiple regression models shown in Table 4.4 tested whether the %BF and
WC could predict risk for each MSRD (elevated triglycerides, low HDL cholesterol,
high blood pressure, and impaired fasting glucose) in the study population,
controlling for covariates. WC alone was a strong, positive predictor of metabolic
disorders independent of gender, except for systolic blood pressure and fasting
glucose in men and HDL in women. When %BF was added to the regression
equation, it did not add predictive power to any of the metabolic disorders, except
for serum triglycerides in men. In model 3, when both WC and %BF were
included in the regression, WC remained a predictor of all of the metabolic
disorders for both men and women, with the previous exceptions. For men, %BF
did not add to the predictive ability of WC for any of the metabolic disorders of
metabolic syndrome. For women, however, adding %BF to WC resulted in %BF
becoming a significant predictor of both serum HDL and fasting blood glucose.

For both men and women and for all the metabolic disorders, adding %BF to WC

did not change the R2 or increased it less than 1% when compared to WC alone
(Table 4.4, model 1 vs. model 3). After controlling for percent body fat (%BF),
the WC remained the most significant predictor for all of the metabolic risks in

which the odds ratio of MMS increased 7% for every 1 cm increase in WC (data

not shown).

97

.00. 0.02. 00.0 v 0

0000.058 ..0 .2 ac._.o..c8 .0. 0000.02.80". 00.0.5 .00. >08 E0200 .o 0.00:? 0:09.308. 05 5.? 0200050. 0.0.0.22 .

00.9.98 :0 .o. 9.50.28 .0. >08 0:00.00 .o 0.00:? 0000:0000... 05 5.3 00.80.00. 0.0.0.0.). .

.3390 9.29:0 .00. 0:0 0000.02.00.00 80.. 3.000 .0 €00.00 .:o_EE:mcoo

_o...8.0 ..0>0. 00.0.0.0 .093 00.9.98 .0. 05:20.8 00:00:08.6 8.0; .o 0.00:? 0.0000000... 05 5.3 02000.00. 0.00.0.2

.0290 9.0.8.0 ..0. 0:0 00.0.3598 Eo... 3.000 .0 £8 60.50588 .9620 .00. 00.0.90 .000 .o 030...? 5.2, 00.80.00. 0.0.0.0.). .

FN (0"!!!

 

8.2.05.0- 8.000.000 20 30.000. 0.0 0.00.0000 2.0 :0 000 880.0 000.0 0:09....

800.20 300.000 000 3.0.0.0 000 8.00.0030 00.0 00.0 000 0.88.0 .820 2.2.8.0
.2000. 8.0.0.30 30 0:00.00 00 8.0000000 30 00.0 000 0580.0 08.02920

8.305.083.8000- :0 $0.80 0.0 3.050. :0 2.0 .00 .0:
80.0000..- 8.9.0.000 ~00 5.0.00.0- 00 8.0.0.0... 000 00.0 ~00 000082055200
c0Eo>>

50.000 .8050 3.0 .9000 90 60.0.3.0 0.0 9.0 000 8820.803 0:08”.
80.00000- 8.000050 .00 800.000 00.0 8:00.000 .00 00.0 .00 0580.0 000.0 9.285
30:00 800.000 000 .:.0.00.0 00.0 £00000 00.0 00.0 .00 05.0.8.0 80.02.0090

 

 

600.30 8.00.0000. 3.0 £00.80 20 8.00.00. 30 20 E0 .0:
200.90 0.000.000 00.0 8.000003 .00 8.00.050. 00.0 00.0 000 800820.: 0.0.00
:05.
300.880 $0.880 $0.880 $0.880
5:000 8.. 00> mm .0. >80 .0 mm 02. mm mm 0
en .000... mm .0005 «F .000... FG .000...

 

0030.560 E00:0a0nc. mm .0. boon. o\e
.2050 005.038.... 0.0.0.... 9.0: E wmz<Iz Eo... om-m~ .55 5.2, .0an 0:0 :0... oEmaflIéoZ c. 0.00.8.0 00.0.0.
2.0.2.3 2.2.0.0... 05 00:00.0 0. .0. .62. 0:00.00 .220 00:0.00E8._o 00.0.5 9:0: 0.0uoE 05.00063. .vé 030...

 

98

Aim 2. To determine whether body fat distribution modifies the effect of percent
body fat and WC on MMS and each MSRD differently in overweight non-Hispanic

white men.

H24: For each percent fat quartile, those with high WC will have higher risk than

those with normal WC for MMS and each MSRD in overweight non-Hispanic

white men and women.

 

Table 4.5. Prevalence of modified metabolic syndrome (MMS1) and metabolic
related risks in different quartiles of percent body fat measured by Bio-electrical
Impedance within the normal and high WC categories in Non-Hispanic white mgr;
with BMI=25<30, ages 20-74 yr, from NHANES III, weighted sample.

 

 

 

Normal WC
Metabolic risks Quartile of % Bodyfatz
n Total 01 02 Q3 Q4
% (SE) % (SE) % (SE) % (SE) %(SE)
Have MMS 828 48.8128a 45314.1 52514.13 44.01588 57.8183:11
High triglycerides 835 42413.33 38514.3 43814.7a 47215.5 42.4174
Low HDL 626 41.9130a 44.7130 42815.0 38.01513 41.1142

High Blood Pressure 635 39.8128a 39013.7 42.1143 35.1149a 43014.5
High Fasting Glucose 635 28012.0a 25813.2 28013.0 25415.8a 40.1188

 

High WC
Have MMS 319 89813.8a 58.81110 89.8180a 78115.7a 70514.2a
High triglycerides 325 59414.4a 48219.4 84.51773 81.8188 59.4144
Low HDL 319 54513.33 49819.9 82.0178 59.01803 49.1142

High Blood Pressure 325 55313.03 43819.0 54.5188 84115.9a 54814.5
High Fasting Glucose 325 41713.73 25918.9 38.3177 50.1:l:5.6a 44.915]a

 

1 .
MMS: Modified metabolic syndrome, based on NCEP ATP Ill, 5 risk factors to diagnOSIs
metabolic syndrome (MS) but not including high waist circumference.

Measured by Bio-electrical Impedance

a - .
Values with the same superscript denote a significant difference between normal and high WC
groups by Wald test, P<0.05.

99

 

Table 4.6. Prevalence of modified metabolic syndrome and metabolic related
risks in different quartiles of percent bodyfat measured by Bio-electrical
Impedance within the normal and high WC categories in Non-Hispanic white

women with BMI=25<30, ages 20-74 yr, from NHANES III, weighted sample.

 

 

 

 

Normal WC
Metabolic risks Quartile of % Bodyfat2
n Total Q1 Q2 Q3 Q4

% (SE) % (SE) % (SE) % (SE) % (SE)
Have MMS1 19419.21418'189153ab 23.7188ab 20118.03b 9418.3ab
High triglycerides 194188133a 18815.3ab 22.1148ab 25.316.4ab 4913.33"
Low HDL 194 43.1143 49915.8 42017.1 40.41107 28319.7
High Blood a ab 8 ab
Pressure 194 19413.5 17.5148 24417.7 19.7181 12.4187
High Fasting a a a a a
Glucose 194 8913.0 5.3128 10214.1 11.4181 14417.8

High WC

Have MMS 470 49813.0a 51.5188a 48.3183a 57.015236 42.115.4ab
High triglycerides 480 3891278‘ 43.3187a 38415.3ab 44315.23” 33.0158ab
Low HDL 470 45.8123 45.9149 54.1150 45814.1 38.1148
High Blood 8 a a a
Pressure 480 47.4128 53.9184 39418.2 53813.5 43.2147
High Fasting a ab a a b
Glucose 480 29512.5 35.9144 28815.4 32215.0 24214.3
17

MMS: Modified metabolic syndrome, based on NCEP ATP Ill, 5 risk factors to diagnosis
metabolic syndrome (MS) but not including high waist circumference.

Measured by Bio-electrical Impedance; The quartile cut points of percent body fat were 22.9,
25.8 and 28.7 for men and 35.5, 38.1 and 40.3 for women, respectively.

a
Values with the same superscript denote a significant difference between normal and high WC
groups by Wald t-test, P<0.05.

Values with the same superscript denotes a significant difference between quartile groups in a
row, by Wald t-test, P<0.05

100

When BF distribution was divided into high and normal WC and quartiles of
%BF as shown in Tables 4.5 and 4.6, both men and women with high WC had
higher risk for MMS. The quartile cut-points of percent body fat were 22.9, 25.8
and 28.7 for men and 35.5, 38.1 and 40.3 for women, respectively. Men with high
WC also had higher risk for all metabolic risks compared to men with normal WC.
Women with high WC had higher risk for high triglycerides, high blood pressure
and high fasting glucose, but not for high HDL. In both men and women, the
trends of having MMS and most of the MSRD were highest in the third quartile,
but not in the 4th quartile in both normal and high WC groups. Figures 4.1
illustrate the pattern of risk response by %BF quartiles. For men, risk of MMS
increased in a stepwise fashion until the fourth quartile, when the risk dropped.
For men there were no significant differences in risks across quartile groups, but
for women there were. Most interesting was the apparent increase in risk for
high fasting glucose for women with normal WC, but a decline in risk for women

with high wc.

101

0.00.00 000.902, .... wmz<Iz E0: ...» 00.8 0000 .omvmmu_.2m 5.2, :0... 0:0 c0802, 02.000.11.02 ... 000000.00 0>> cm...
000 .0:..0: 0002000 0.03000 2:00.00 ..0 00.0030 0.0.0.20 0. 3.2.2. 00.9000 2.00000. 00500:. .0 00:0_0>0.n. 0.0 0.30.”.

003000. 0:00.00

 

 

 

 

 

0.00.00 Om Om O.
_ 1 o
1. OH
-.....ill ........ 4 --.-.-rlLa om
Om

 

o0
‘/.ii..lrorrirlll\.‘i 11....-. CW

 

 

 

 

 

00
i\ on

 

 

02. _0E.oz....... 0.5 coil.
0.00.0". ‘

ow

o\o 00:0_0>0.n_
205. I

m5:>_ 0201

102

H22: Those with high body fat (>50th %tile), and high waist circumference

(HFHVV) will have higher risk for MMS and each MSRD compared to those with

low body fat (_<_50th %tile) and high waist circumference (LFHWC), high body fat

(>50th %tile) and normal waist circumference (HFNWC) and low body fat (550th

%tile) and normal waist circumference (LFNWC) in overweight non-Hispanic
white men and women.

Table 4.7 shows what happened when the quartiles were collapsed into two
groups by a median split for %BF-~high and low %BF. The risks for MMS and

MSRD were higher in groups with high waist circumference when low body fat

(550th %tile) and normal waist circumference (LFNWC) used as a reference

group. For men this was only with high body fat (HFHWC). Men with HFHWC
were about twice as likely to have MMS and high TG compared to those with
LFNWC. Women with HFHWC were about twice as likely to have high blood
pressure compared to those with LFNWC. The risk of high fasting glucose was
about three times as high in both high WC groups (LFHWC, HFHWC) in women
but not for men. In groups with normal waist circumference (LFNWC, HFLWC)
the risks were not different regardless of their %BF. However, when
comparing low fat (LFNWC, LFHWC) to high fat (HFNWC,HFHWC) groups in
women, the risk for MMS and the most MSRD were both decreased in normal WC

and high WC group except for high blood pressure but statistically non-significant.

N

0.52“: E0... 20.0.50 5080.090 ...
5050.5 .0. 50 00 0:0 :05 .0. 5003 0.0.5 0.500 .00 0>> 0: b .:050.5 .0. $.00 0:0 :05 .0. 800.00 0.0.5.0. 500:
500.00 .50. .0 0.500 .30 0: .. .00:0.0.5:0..0 .0.0.5 :0.: 0.500080 :0... ”0.510... 00:0.0.5:0..0 .0.0.5 :0... 0.500080 .50.. ”03:"...
00:0.0.5:0..0 .0.0.5 .050: 0.50005O :0... ”032...... . 000.0 00:00.01 .00:0.0.5:0..0 .0.0.5 .050: 0.500080 .50.. .0320...V

0.0:.00. 5 005505.050 9000 0:. .0 0.05 .0 0.5. 5:0 90: .0.. .0>> :05 05020:. .0: .0: 0:050 20...... 0moz :0 000mm 0

20.0500 FM 000020 05.00”. .0155 mmM.om..M 05000.0 000.0 ...0.05 omv :050>> 20.05 00v :2... “.0... 7.0.05 03M 00000020.... m

.00005:0.00 5 0.9.0.5

00:00.58 ..0. 0:0 0.0.0.5058 50.. 50.0:0 80 0:0 0.0.5 .2020 050.050 .5500 00.0.50 .000 .0. 00.00.00 0050. 0000 F

 

 

 

 

 

1.0.0-5.: 00.0 500.000.: 0.0 80.1000. .0.. 00.. 880.0 0.500. 00....
1.0.0.0.... 00.. 600-000. 0: 80.7000. 0.0 00.. 0.008.008.0091
5.0-5.0. 00.. 8.0-00.0. .0.. 5.5.30.0. :0 00.. 020...
80.000. 00.0 5.0.0-000. 0.: .3000. 00.0 00.. .0: 28..
8.0-00.0. 00.. 50501.0: 5.0 60.7000. 000 00.. o__0080_.m:mm_..._w.m._.0.
Eon: :0Eo>>
800-000. 03. 0.0.7000. .00 60.0-5.0. 00.. 00.. 0080.0 0:000“. :0...
50.000. .0.. 6.0-00.0. .00 .00...-100. ~00 00.. 0.0808 08.0 80.:
1.00.000: 00.. 0.0.0-50. .0.. 80.7000. 00.. 00.. 05.0.:
80.0.00. 00.. 50.0.50. 00.. 5.2-00.0. R0 00.. .0128.
10.0.01... 3.0 6.1.0-000. 0.0.. 80.3.00. .0.. 00.. songwflmmwflmm
000...: :2...
0.51.: 0.5:“: 0.52.: 10.52“: 9.0.. 2.880....
0.0.5 00.... 8.05 08.02

 

0.0500 0050.05 ....

$20.12 50.. .... 3-00 0000 .omvmmuz20 5.5 :0505 0:0 :05 .255 0.:000.I-:0z :. AO>>In=rc 00:0.0.5:0..0 .0.0.5
:0.: 0:0 .0. >00: :0.: 0:0 ..0>>Z....I. 00:0.0.5:0..0 .0.0.5 .050: 0:0 .0. >00: :0.: .92."; 00:0.0.5:0..0 .0.0.5 :0.:
0:0 .0. >00: .50. .6320... 00:0.0.5:0..0 .0.0.5 .050: 0:0 .0. .50. 5 N00.0.. 00.0.0. 05000.05 .0.0..0. 0000 ...... 0.005

104

Aim 3. To examine the predictive power of WC and %BF to detect the risk for
MMS and/or each MSRD in overweight Taiwanese men and women using data
from the Nutrition and Health Survey in Taiwan 1993-1996 (NAHSIT).

A total of 291 males and 312 females were overweight, defined in Taiwan as
having a BMI 24<27, and comprised the sample for Aim 3 of this study. Within
this ovewveight group, females were about 4 years older than males. Males had a
higher WC than females, but a lower %BF (Table 4.8). For blood lipid profiles,
females had higher HDL cholesterol compared to males. There were no
differences by gender in fasting glucose or blood pressure. There was a trend

for men to have higher triglycerides, but the gender difference was not significant.

Table 4.8 Descriptive characteristics and metabolic related risks in the total
sample of overweight (BMI 24-27) Taiwanese participants ages 20-74 yr from
the Nutrition and Health Survey in Taiwan, weighted sample.

 

Descriptive characteristics

Men (n=291 ) Women (n= 312) P value

A9e (Y) 42.3.0.1 46211.3 0-02
BMI (kg/m2) 25.301 25.301 0-78
Waist circumferences (cm) 84.8i0.3 78fl3i03 <0.001
% Bodyfat from BIA 229:10 30.8i0.5 <0.001
Metabolic variables
Triglycerides (mgldl) 135.8i9.1 109.0169 0.053
HDL cholesterol (mgldl) 46.3i2.0 55.8i1.4 <0.001
Fasting Glucose (mgldl) 84.7:1.1 86.5:15 0.36
SYStOIiC DIOOd pressure 1257:12 1245:2'1 0.71
(mm Hg)
Diastolic blood pressure 82.7:0.8 78.8:1.4 0.07

(mm Hfl

105

Table 4.9 shows the prevalence of metabolic risk factors and MMS

between overweight Taiwanese men and women. The prevalence of high WC

was significantly higher in females compared to males (32.76% vs 14.05%), as

was low HDL. Males tended to have a higher prevalence of high triglycerides

compared to females. There were no significant differences in the prevalence of

MMS and high blood pressure between genders.

Table 4.9. Prevalence of metabolic risk factors in overweight (BMI 24 to 27)
Taiwanese men and women ages 20-74 yr from the Nutrition and Health Survey
in Taiwan, wejighted sample.

 

 

Men Women
Risk factors Percentage 95% CI Percentage 95% Cl
High Waist a b
circumference 14.05 (8.9. 21-6) 32.76 (28.9. 36.9)
Low HDL 28.31 (13.6, 49.7) 44.53 (35.4, 56.5)
High TG 29.31 (19.7, 41.2) 22.63 (15.7, 31.5)
High blood
pressure 53.24 (35.5, 70.2) 43.1 (36.1, 50.4)
”'9“ FaSt'"9 7.17 (3.9, 13.0) 10.27 (59,173)
Glucose
Metabolic Syndrome 16.1 (9.2, 26.5) 18.7 (13.5, 25.3)
M d'f d M t b I'
o ' 'e e a O 'C 35.5 (20.0, 55.0) 30.49 (25.8, 35.6)

Syndrome1

1
Based on NCEP ATPIII Criteria but not including high WC, i.e. having 2 or more of following

abnormalities: Triglycerides 3150 mgldL; HDL: Men <40 mg/dL Women <50 mgldL; Blood
pressure 3130/385 mmHg; Fasting glucose _>_100 mg/dL.

a
Different superscripts show significant differences between genders, P<0.01;

Cochran-Mantel-Haenszel Chi-square Test

106

H34 Both %BF and WC predict risk for MMS and/or each MSRD (elevated

triglycerides, low HDL cholesterol, high blood pressure, and impaired fasting
blood glucose) in overweight Taiwanese men and women.

The correlations of WC and %BF to the modified metabolic disorders in
Taiwanese oven/veight but not obese (BMI=24-27) men and women are shown in
Table 4.10. In both men and women, BMI was significantly correlated to WC at
(r=0.45, 0.29, respectively), but not to %BF. %BF was significantly correlated to
WC only in men, but not in the women. WC was strongly associated for women
with all of the metabolic disorders except for serum HDL. For men, WC only was
significantly correlated with systolic blood pressure and fasting blood glucose.
%BF was weakly correlated with all the metabolic risks in men. In women, only
systolic and diastolic blood pressure were positively correlated with %BF. WC
was significantly and more strongly associated with MSRD compared to %BF in

both men and women, but no association exceeded r=0.37.

107

Table 4.10. Pearson’s correlations of BMI, waist circumference and percent
body fat to the metabolic disorders in Taiwanese men and women ages 20-74
year with BMI 24-27 from NAHSIT, weighted sample.

 

 

 

 

Men Women

WC %BF WC %BF

Body Mass Index I) 045* 0.05 029* 0.07
(0.000) (0-519) (0.011) (0-144)

Waist Circumference 1.00 0.38 1.00 0.09
(0.015) (0.480)

Percent Body Fat 038* 1.00 0.09 1.00

Serum triglycerides 0.11 0.11 028* 0.17
(mgldL) (0.214) (0.326) (0016) (0.076)
Serum HDL (mgldL) -0.02 -0.12 -0.09 -0.06
(0.902) (0.428) (0.293) (0.552)

Systolic blood 0.37" 0.12 023* 021*
Pressure (mm H9) (0.001) (0085) (0.013) (0.036)
Diastolic blood 0.23 0.05 (111* 018*

Fasting plasma 020* 0.02 037* 0.05

 

*Correlation is significant at the 0.05 level (2-tailed).

108

Logistic regression models were developed to evaluate the relative
contributions of WC versus %BF to risk for MSRD adjusted by age, physical
activity, and alcohol intake. Percent energy intake from carbohydrates and fat
were dropped from the equation here, due to too many missing values. Smoking
status was also dropped from the regression equation in women, because only 2&
of Taiwanese women smoked (Table 4.11). Although the overall model for
each regression were significantly correlated to each of the MSRD, WC alone,
%BF alone and combined WC and %BF together were not a significant
predictor for most of the MSRD in both genders. Only %BF alone was a
significant predictor for systolic blood pressure in women.

When using R2 to determine the predictive power between WC and %BF,
WC had better predictive power than did %BF for most MSRD in both men and
women, but the differences were small. Percent body fat had slightly better

predictive power for blood pressure in women. When both WC and %BF were

combined in the equation model, the R2 increased only slightly (1% to 3%)

compared to WC alone (model 3 vs. model 1 in Table 4.11).

109

.02 20>) 00.0 v e 0

0009050 :0 .00 05:00:00 000 00:0000E3000 Em; .5 >000 E0800 00 0.00:? E00:0000£ 05 55, 0200060. 030.22 .0

00000900 :0 .00 05:00:00 000 >000 0:00.00 00 0.00:? 05080005 05 55, 0200050. 03:35. .0
.6390 000.000 .5 0:0 000000300000 E00 3.000 «.0 E0800

5000830000 .2020 ._0>0_ 020.000 09$ 00000900 .8 05:00:00 0000008380 503 00 0.00:? E00:0000£ 05 505 0200050. 03222 .N

.0330 950:5 5009:3008 6000.0 ._0>0_ 020.000 .000 00 0.00:? 5;) 0200050. 03222 ._.

 

 

 

20.00000- 600:3 8.0 80.00000- B0 @0000; 2.0 00.0 00m 0880 085 000000
A2050 60.00000 80 30.0050 30 $0000 00.0 00.0 08 05085 0830:9000
03.00000 80.050. 00.0 03.000000 00.0 60.00.20 00.0 00 05 0580500052600
$00080 $00000- 00.0 5.880 80 60.0300. 000 2.0 000 00:
6:00: 83:00 2.0 63:: :0 $000.0 2.0 000 now 000082055000
00.003
60.0050- 30.00000 :0 30000.00- :0 60.00000 20 00.0 00m 08020 0020 0008“.
$0020. 60.00000 00.0 30.00000- 30 80000.0 000 80 EN 050090 0830:9000
60002.0- 30.00%.0 00.0 8300000. 000 30:00 00.0 00.0 00 esmmmaeeegzoiw
30.0020- 3000000. 000 80.0020- 00.0 300000. 00.0 000 Km 00:
5.380 60:00.0 :0 30:30 00.0 60:0: :0 00.0 000 8008205,:200
:05.
$00080 000.080 $000080 300080
002000 as 03 mm 6:000 00 mm 0>> mm am e
00 .002: ...N .28.: NF .28.: F0 .002:

 

0.058 09000; .0885
050.050 0:00.300: 05 0:005 0301600 E0200 020:0 000000052000 E03 050: 0.0008 co_mm060m

.m0_00:0>00 E00:0000£ mm 000 >000 £0 0ch 00000005000 Em; mEm:

._._WI<Z E0: RN: VN _S—m 5:3 VNION 0mm c®E0>> DEN CmE mmwcm>>_m.._. E whmfluomfi Own—E9.

. —. EV 0.00...

110

H3.2 Those with high body fat (>50th %tile) and high waist circumference (HFHW)

will have higher risk for MMS and each MSRD compared to those with low body

fat (550th %tile) and high waist circumference (LFHW), high body fat (>50th %tile)

and normal waist circumference (H FNW) and those with both low body fat (5 50th

%tile) and normal waist circumference (LF NW) in overweight Taiwanese men and

women

Although the sample was somewhat small for examination by gender,

Table 4.12 shows the prevalence of MMS and related metabolic risk factors for
the sample divided into four groups of adults based on a median split by %BF.
The 50th percentile cutpoints of %BF were 23.3% for men and 31.4% for women,
respectively. In general, men and women with normal WC had lower prevalence
of MSRD compare to high WC groups, with the possible exception of low HDL’s
being somewhat higher in normal WC groups. Women with higher WC had
increased risk for high blood pressure and high fasting glucose compared to
normal WC group.

In normal WC category, women had higher prevalence of MMS and
metabolic related risks when %BF decreased. However, this phenomenon was
not seen in high WC group except for high fasting glucose. In the high WC group,
high percent bodyfat had higher risk for MMS and MSRD, the risk of having MMS
was almost double in the high %BF group compared to the low %BF group

(p=0.079).

111

.0058. 0_ 0000: 0250000000 0>000 0.: .0 0.0:. .0 020 >00 0>00 .0.. .0>> 090 050200. .00 50 2.00.0 ___n:< nE02 00 0000m

mod v a .30. 0 0. 0030.0 2.00:0 000.200 0000.000 “0000.090 0 0000000 .002 00.00 05 £3, 00:_0>
mod v a 200.0. 000 0200. 000.200 0000.000 00005090 0 0.000 02.00.0000 0800 05 5.2, 00:_0> u

000

m

202000000. 02003 .0. 000.5 000 000. .0. $0.0 6 .0. >000 0000.00 090 000 2,2 .0 0.060 50 0000
003E 00 FM 000020 @0000“. .mIEE mmMerM 05000.0 002m J0EE omv 00. L 09.. ovv 022 ADI 7.0.90 03M 00000039; 0

 

 

 

 

 

 

000.0 .00. e0 00. 0:00 at. I. 00.. 000.0 0.0.0 000020 000000 00.:
300.00 00.00 0.00 000.00 00.0 0.00.00 A.00 £0.00 05000.0 0003 00:
0.0.H 0.00 0.0 .2000 0.0 0.00.: 0.0.1.0000 00 00.:
0.: 00.00 0:00.00 0.: 02.00 0.0. 000.00 .5: 2,3
a. 0.02... 0.00 0 0.00.. 00.00 m 0.00 00.02 0. 0 $00.00 050.056 0.00002 000002
00 00 00 00 000": 00502,
0.0 H00.? 0.0 H000... 00... 00.0 0.0 H00 000020 00000. 00,0.
00.0 000.00 0.00.0 0020 2.0. : 000.00 00.0 000.00 050000 0000 09:
m 0.0 000.00 00. 000.00 0. : 000.00 00.0 00000 00 00.:
0.00 03.00 0.02 000.00 000 000.00 0.0 000.00 00: 2,00
0.0 fl. 3 000 000.3. 0.0. £0.00 00. 000.00 0020.000 0.00902 000.002
3 .00 00 000 000": :05.
00.2. 00.: 00.... 2,3 .030 09: 00.0. 2,3 200 0:00.202
0>> 0%- 0>> .0:..02

 

200.00 00390.5 00220.0 0_ >023 0:001 000 000032 05 0.0.. .> 3.8 00mm .0000 00000220..
€000 :20. 50.2006 0. 000000.00 02, 00... 000 0.0.00 05 e. 5.;

0000000000 00502005 >0 0.00000. 0003000

0000.00 090 000 2,2 0. P0.00000 20.. 2000000. 00020. 000 0Eo.00>0 260000;. 00500:. 00 80200.0. .36 0.00...

112

CHAPTER 5
DISCUSSION

Aim 1. To determine whether percent body fat assessed by Bioelectrical
Impedance Analysis adds to the predictive power of waist circumference (WC) in
risk for modified metabolic syndrome (MMS) and/or each metabolic syndrome
related disorder risk (elevated triglycerides, low HDL cholesterol, high blood
pressure, impaired fasting blood glucose; MSRD) in non-Hispanic white,
overweight (BMI 25-29.9) men and women (NHANES III).

This study was unique in limiting the examination of associations of
anthropometric indicators for metabolic co-morbidities to a restricted range of
BMI=25-30. The prevalence of modified metabolic syndrome, without high WC,
was 54.2% in males and 39.5% in female, much higher than it was for the ATPIII
definition for metabolic syndrome that includes WC. If using the modified ATP Ill
definition in our study, the prevalence of metabolic are 32.1% in males and 36.1%
in females which are still higher to the prevalence of metabolic syndrome in the
same population in Park’s study, since the prevalence reported used original
NCEP ATP ||| definition which the fasting glucose was set as 2 110 mg (Park, Zhu
etaL,2003)

The twofold higher prevalence of elevated WC of women in the present study
might be reasonably expected to affect the use of WC to predict the risk of
metabolic syndrome. In fact, when high WC was excluded as a criteria of
metabolic syndrome, the prevalence of modified metabolic syndrome was higher

in men than women (54.5% to 39.5%, P<0.01). The finding that MS was slightly

113

higher in women than men, usirg the ATPIII definition (32.1% in men and 36.1%
in women respectively), suggesis that WC makes a major contribution to MMS for
women.

The rates of heart disease by gender also support the findings from the
present study that show a decline in the gender gap in risk for CVD with
advancing age (Barrett-Connor, 1997; Li, Engstrom et al., 2006). Estrogen and
testosterone modulate glucose and lipid metabolism, such that a decline in
estrogen or a relative increase in testosterone induces insulin resistance and a
pro-atherogenic lipid profile (Regitz-Zagrosek, Lehmkuhl et al., 2006). Thus, it
was not surprising to find in the present study that the prevalence of MS or MMS
increased in women with advancing age. In the present study, the prevalence of
MMS in women was only slightly increased to 39.5% compared to the prevalence
by the ATPIII definition of 36.1%.

The main finding was that %BF provided no advantage over WC in screening
for MSRD. Current evidence clearly supports that central obesity is a major
contributor to metabolic syndrome (NCEP, 2001; Bergman, Kim et al., 2007). As
others have found, women averaged higher %BF values than did men, supporting
the necessity of using a sex-specific %BF range for predicting the risk of the MS
(Jakicic, Wing et al., 1998; Zhu, Wang et al., 2003). That women have high %BF
and men, high WC as found in the present study, has been well documented by
others (Lemieux, Prud'homme et al., 1993).

The weak to moderate correlations of BMI to %BF, (r= 0.23, 0.47 for men

versus women, respectively in the present study) is in contrast to other studies

114

that used wider spectrum of weights (r= 0.743 to 0.924) (Roubenoff, Dallal et al.,
1995; Nagaya, Yoshida et al., 1999). The weak correlations of both BMI and
%BF to metabolic risks in the present study were likely due to the narrowed BMI
range, clearly demonstrating the varied body composition within the BMI range of
25-30. Weaker correlation in the present study may also reflect greater
heterogeneity in factors other than adiposity that influenced lipid profiles.

In the present study, WC was more closely related to BMI than to %BF in
both genders and closely associated with metabolic risks, except for HDL. This
finding is in line with previous studies showing WC was the best predictor of
metabolic risk that compared with other anthropometric indicators (Janssen l,
2002; Wang, 2003; Janssen, Katzmarzyk et al., 2004; Phillips, Jing et al., 2008).

For women in the present study, when both WC and %BF were in the
prediction equation for MSRD, %BF related positively to HDL (higher %BF and
higher HDL) and negatively to fasting blood glucose (higher BF and lower
glucose). This observation is in line with some previous studies (Snijder, Dekker
et al., 2004; Sakai, Ito et al., 2005; Bosy-Westphal, Geisler et al., 2006), but
contrary to others (Nagaya, Yoshida et al., 1999; Deurenberg-Yap, Chew et al.,
2002). In the Quebec Family Study, using a comparable number of adults, %BF
added little additional information to BMI, as we also found with respect to
cardiovascular disease risk factors in females (Tanaka, Togashi et al., 2002).

In the Australian Diabetes study investigators found a high hip circumference
to be strongly associated with a reduced prevalence of undiagnosed diabetes and

dyslipidemia up to age 75 (Snijder, Zimmet et al., 2004). Several mechanisms

115

could explain the inverse association of %BF to the metabolic disturbances in
women after adjusting for WC. One possibility might be the effect of estrogen on
CVD risk factors (Tanaka, Togashi et al., 2002), i.e., the protective effects of
estrogen on plasma lipids might be stronger than the variability in adiposity within
the narrow range of weights used in the present study. Another possibility is
gender difference in the metabolism of lipids from the adipose depots on the
thighs and hips versus the abdomen. Abe and Fukunaga (1994) found thigh
adiposity in women associated with higher plasma HDL cholesterol levels and a
lower incidence of diabetes (Abe and F ukunaga, 1994). Subcutaneous fat tissue
from the hips and thighs has been suggested to play a protective role for MSRD,
because subcutaneous adipose tissue is less sensitive to lipolytic stimuli
compared to visceral fat (Busetto, 2001). Lipoprotein lipase (LPL) activity is lower
in femoral subcutaneous adipose fat compared to visceral fat (Arner, 1995).

This increased LPL activity in visceral region can lead to insulin resistance and
beta cell dysfunction (Frayn, 2002; McCarty, 2003). Larger reserves of body fat on
the buttocks and thighs of women versus marked accumulation of fat in the
intra-abdominal visceral fat depots of men using computed tomography has been
associated with more favorable levels of lipids and glucose in women (Williams,
2004; Tousignant, Faraj et al., 2008). Williams et al. (Williams, Hunter et al., 1997)
found that trunk adiposity was associated with unfavorable serum lipids and
lipoprotein levels and leg fatness was related to a favorable lipid profiles in women
aged 17—77 years.

In the present study, although WC had an independent ability to predict most

116

of the metabolic risks, it explained a low to moderate percentage of the variability
in the metabolic risks (total R2 = 0.04 to 0.41 in men and 0.06 to 0.41 in women
respectively). A relatively narrowed BMI range in this study population might
contribute to the low percentage of variability. Also, other factors such as
genetic predisposition to chronic diseases must be important as well.

It might seem an odd finding that in women %BF was inversely associated
with risk for MMS when adjusted for WC, but a similar finding has been reported
elsewhere. Hunter et al. found that the relationship of visceral fat to metabolic
complications was independent of concomitant variation in total body fat
(Williams, Hunter et al., 1997; Seidell, Perusse et al., 2001). One reasonable
explanation about the risk reduction when %BF was included in the equation in
the present study is that when WC increases, not only visceral fat increases, but
subcutaneous fat does as well. The ability of WC to predict abdominal
subcutaneous and visceral fat in addition to total fat (Janssen, Heymsfield et al.,
2002) should be considered as a potential mechanistic explanation for findings in
the present study. Cross-sectional studies have shown that, once the amount of
abdominal fat is taken into account, accumulation of fat and lean mass in the arms
and legs seems to be protective against cardiovascular risk factors (Ferreira,
Snijder et al., 2004; Snijder, Dekker et al., 2004). In most people, subcutaneous
fat accounts for the largest component of total fat (Abate, Garg et al., 1997;
Ross, Aru et al., 2002; Ross, Freeman et al., 2002), but for a given amount of total
body fat, men have about twice the visceral fat as do pre-menopausal women

(Lemieux, Prud'homme et al., 1993). The index of WC is comprised of both

117

subcutaneous fat and visceral fat in the abdominal region. It is the more
metabolically active visceral fat accumulation is that contributes most to MS
(Zierath, Livingston et al., 1998; Wong, Janssen et al., 2003; Hyun, Kim et al.,
2008) I

The findings from the present study clearly support that assessment of
cardiovascular risk in ovenNeight people solely from body weight or total body
fatness can be quite misleading. Although the best methods for estimating the
amount of visceral fat are imaging techniques, such as CT or magnetic resonance
(Borkan, Gerzof et al., 1982; Leite, Wajchenberg et al., 2002), these are
inappropriate for screening or routine clinical examinations due to high cost and/or
radiation exposure. Thus, WC is a good choice for predicting visceral obesity in

primary care and community-based studies.

Aim 2. To examine whether body fat distribution modifies the effect of percent
body fat and waist circumference on modified metabolic syndrome and each
metabolic syndrome related disorder differ in overweight non-Hispanic white men
and women.

In this study, the prevalence of women with low %BF but high WC was similar
to that from a previous report (St-Onge, Janssen et al., 2004), and consistent with
other findings that people with less visceral obesity have more favorable
metabolic profiles (Karelis, Faraj et al., 2005). Dvorak et al. (Dvorak, DeNino et

al., 1999) analyzed a cohort of 71 healthy normal weight (NW) women (21—35

118

years old) and divided them into metabolically obese3 and normal insulin
sensitivity groups. Investigators found a higher %BF (32 :I: 6 vs. 27 :I: 6%, P <
0.01) and higher amounts of subcutaneous fat (213 j: 61 vs. 160 1: 78 cm2, P =
0.03) and visceral adiposity (44 :I: 16 vs. 35 :I: 14 cm2, P < 0.05) in the
metabolically obese NW group versus the NW group, respectively.

In the present study, both genders with low %BF but high WC had a higher
prevalence of MMS and related metabolic risks than those with high %BF but
normal WC. This finding indicates that visceral obesity is a better predictor of the
progression to CVD than is total %BF. Visceral fat is more sensitive to lipolytic
activity, resulting in an increased release and accumulation of F FAs in multiple
organs contributing to insulin resistance (Despres, 2006). Concomitant
increases in adrenergic reactivity and blood pressure lead to vascular stiffness
(Egan, Lu et al., 1999). However, the high prevalence of MMS in both men and
women with normal WC indicates that WC cannot entirely be driving WC.

In this study, the metabolic risks increased with increasing %BF quartiles, but
not in a linear fashion, further supporting that %BF was not as good as WC in
predicting MMS and related metabolic disorders within this restricted BMI range.
The finding that high %BF did not predict the highest metabolic risks with
increasing %BF was consistent with the logistic regression that a protective effect
of %BF to the metabolic syndrome and related metabolic risks. Analysis of the

NHANES surveys found that, relative to the normal weight category, overweight

 

3 Metabolic ob ese NW women based on cut points for insulin sensitivity (normal
= glucose disposal >8 mg / min x kg fat-free mass (FFM); impaired = glucose
disposal <8 ml/min x kg FFM.

119

was associated with a statistically significant reduced number of deaths, after
adjusting for sex, age, smoking, race-ethnic group, and alcohol consumption

(F legal, Graubard et al., 2005). Other researchers have observed a protective
effect of peripheral obesity against cardiovascular disease. Tanko et al. reported
that peripheral fat mass showed an independent negative correlation with both
atherogenic metabolic risk factors and the most severe insulin resistance (2003).
Dyslipidemic syndrome was found in women with high percentage central fat and
low percentage peripheral fat (Tanko, Bagger et al., 2003).

Findings from present study support that WC may be a better predictor of
modified metabolic syndrome and related metabolic risks than %BF. In women,
higher levels of %BF were associated with reduced risk for high serum
triglycerides, low HDL, high fasting glucose and modified metabolic syndrome

when adjusted for WC.

Aim 3. To examine the predictive power of waist circumference and percent
body fat to detect the risk for modified metabolic syndrome and/or each metabolic
syndrome related disorder in overweight Taiwanese men and women using data
from the Nutrition and Health Survey in Taiwan 1993-1996 (NAHSIT).

The prevalence rates of MS in the Taiwanese sample were about half that of
the US white sample, but had rates that were similar by gender and increased
with age (Park, Zhu et al., 2003). The contributions of the MSRD to the
prevalence of MS and to MMS appear to differ by gender between these two

samples. Dividing the prevalence of MS by that for MMS in the Taiwanese

120

sample, about 45% (161/356) of men and 61% (18.7/30.5) in women had high
WC as one of the criteria. In contrast, in the non-Hispanic whites about 59%
(32.1/54.2) of men and 91% (361/395) of women had high WC as one of the
criteria. Thus, high WC contributed more strongly to the prevalence of MS in
non-Hispanic whites. This shows that not only the prevalence of metabolic
syndrome differed, but also the components of metabolic syndrome. About 55%
and 40% of Taiwanese men and women who have MS do not have high WC,
indicating factors other than abdominal adiposity play a major role in the etiology
of MS in Taiwan.

Using the NCEP ATP lll definition of MS, high blood pressure, high
triglycerides and low HDL with or without high WC was most common for
Taiwanese men, while the non-Hispanic white sample had a more evenly
distributed pattern of prevalence of co-morbidities (Pan, 2002). Most Taiwanese
women had a clustering of high WC, low HDL and high blood pressure, while
most white women had a syndrome of high WC, low HDL and high triglycerides
(Pan,2002)

Differences in the prevalence of MSRD for MS between the Taiwanese and
non Hispanic whites might have been due in part to differences in the protocol for
measuring WC between NHANES III and the NAHSIT. In NAHSIT, the WC was
measured midway between the iliac crest and the lowest rib, while in the
NHANES III survey, WC was measured at the iliac crest. Thus, in women, but not
in men, there can be a difference of up to 2 cm between a waist measurement

taken immediately above the iliac crest and a measure taken midway between the

121

iliac crest and the lowest rib (Wang, Thornton et al., 2003). This difference in
measurement location, cannot explain the difference entirely, becaUse although
the prevalence differed, the distribution patterns in WC were similar by gender. In
NAHSIT, the prevalence of high WC was over twice as in women compared to
men (32.8% vs. 14.1%), similar to that in non-Hispanic Whites (67% vs. 26%).

In the present study, the higher prevalence of hypertension in the Taiwanese
versus non-Hispanic white adults revealed considerable underlying variation
between these two groups. One mechanism by which body fat is thought to
influence hypertension is increased insulin resistance especially from central
adiposity. However, it has been known that after controlling for BMI, Asians have
a higher prevalence of hypertension than do whites (Chandalia, Abate et al.,
1999; Colin Bell, Adair et al., 2002; Pan, Flegal et al., 2004). One study
demonstrated that Asian Indian men are more insulin resistant than Caucasian
men independently of generalized or truncal adiposity (Chandalia, Abate et al.,
1999), suggesting that some Asian populations have a higher prevalence of
hypertension than Caucasians. It is likely that genetic and/or environmental
factors such as high sodium intake contribute to the higher prevalence of
hypertension in some Asians (Jones, 1995; Chandalia, Abate et al., 1999; Colin
Bell, Adair et al., 2002; Pan, Flegal et al., 2004).

Analysis of the NAHSIT dataset from Taiwan resulted in very weak
correlations between BMI and %BF for both men and women (r=0.05 and 0.07,
respectively), a finding that differs from a study on Japanese adults. Using

national survey data in Japan, Nagaya et al'. examined 12,287 Japanese men and

122

6657 women to find BMI strongly correlated with %BF across the entire
spectrum of BMI (r = 0.743 and 0.924, respectively) (Nagaya, Yoshida et al.,
1999). Again, this difference was due to the range in BMl’s, from 13.8 to 43.4
kg/m2 in Nagaya’s study to a range restricted to 24 to 27kg/m2 in the present
study. Such findings demonstrate a large inter-individual variation in %BF within a
narrowed range of BMI. It also explains the small R2 in the regression analysis of
WC and %BF to the MS and metabolic related risks.

Similar to findings in NHANES III data, the logistic regressions for MMS or
co-morbidities with the NAHSIT dataset provide compelling evidence that %BF
coupled with WC does not predict an increase in MMS and metabolic related
health risks better than does WC alone. One unexpected finding from the
NAHSIT regressions, though, was that adding %BF to WC in the equation
predicted high serum triglycerides and MMS in females (Table 4.12). Again, this
finding suggests that factors other than %BF or WC are also contributing to the
etiology of metabolic risks in Taiwanese ovenlveight, but not obese, adults. Both
the narrowed BMI range and smaller sample size in NAHSIT contribute to failure
of obtaining precise estimates in the present study. In addition, the reference
group for the prediction equations was normal WC and low %BF in overweight but
not obese subjects instead of the normal weight group.

In the present study, same as the American population, WC remain a better
predictor of modified metabolic syndrome and related metabolic risks than %BF in
Taiwanese overweight but not obese population. Men and women with normal

WC had lower prevalence of MSRD compare to high WC groups, with the

123

possible exception of low HDL’s being somewhat higher in normal WC groups.
Women with higher WC had increased risk for high blood pressure and high

fasting glucose compared to normal WC group.

124

CHAPTER 6

CONCLUSION

The concluding chapter provides a summary of major findings, and the
evaluation of the overall strengths and limitations of this research. The chapter
concludes with implications for future research suggested by findings from this

dissertation.

A. Summary of Findings

The aims of this research were as follows. 1) Determine whether %BF
assessed by Bioelectrical Impedance Analysis (BIA) adds to the predictive power
of WC in risk for MMS and/or each MSRD (elevated triglycerides, low HDL
cholesterol, high blood pressure, impaired fasting blood glucose) by gender. 2)
Examine whether BF distribution modifies the effect of %BF and WC on MMS and
each MSRD differently in non-Hispanic whites and Taiwanese overweight, but not
obese, men and women. Two national datasets NHANES III from the US. and
NAHSIT from Taiwan were used to address these aims.

The main finding was that %BF measured by BIA provided no advantage
over WC in screening for obesity-related metabolic risks. WC remained a strong
positive predictor for MMS and MSRD and independent of gender, except for low
HDL and high blood pressure in men. In this study, it was hypothesized that %BF
as an indicator of total BF would have predictive power for MSRD. However,

results demonstrated that neither %BF alone nor %BF combined WC improved

125

the predictive power over WC alone for most of the MSRD. These findings
confirmed that the MMS and co-morbidities are largely driven by central adiposity
rather than total body adiposity. Another study also demonstrated that WC, and
not BMI, explained obesity-related risks (Janssen, Katzmarzyk et al., 2004). In
the present study, %BF was not superior to BMI in predicting MSRD for people
classified as OW but not obese.

An important, and somewhat unexpected, finding of this study was that higher
levels of %BF were associated with reduced risk for high serum triglycerides, low
HDL, high fasting glucose and modified metabolic syndrome when adjusted for
WC in women. One reasonable explanation for the risk reduction when %BF
was included in the equation is that when WC increased, not only visceral fat
increased but subcutaneous fat did as well. The WC index is a combination of
both subcutaneous and intra-abdominal fat in the abdominal region. More and
more studies support that intra abdominal fat (visceral fat) accumulation is the
fundamental driver for MS (Kvist, Chowdhury et al., 1988; Abate, Garg et al.,
1997; Wong, Janssen et al., 2003; Janssen, Katzmarzyk et al., 2004). Therefore,
assessment of cardiovascular risk in oveniveight people solely from WC or total
BF can be misleading, especially for women.

Using the current WC cut-point, more than two-thirds of over-weight but not
obese white women had high WC, a rate twice that of white males in the same
category. Comparing the prevalence of each component of the MMS between
white men and women, all the metabolic risks were higher in men except for high

WC. However, the prevalence of MS defined by NCEP-ATP III was slightly higher

126

in women than men suggesting that high WC in women makes a major
contribution to MMS.

In examining BF distribution patterns and the prevalence of MS, MMS and
MSRD, ethnic differences became apparent between the US. white and
Taiwanese samples of over-weight adults. Both the WC and %BF were lower in
the Taiwanese versus US white samples for both genders. The %BF was about
2% and 7% lower in Taiwanese men and women, respectively, compared to US
non-Hispanic whites. The prevalence of high WC was also about half in the
Taiwanese compared to non-Hispanic whites. In general, the prevalence rate of
MS, MMS and each MSRD were all higher in non-Hispanic whites than in the
Taiwanese for both genders, except for high blood pressure. Non-Hispanic white
males had a higher prevalence of high serum triglycerides and high fasting

glucose while non-Hispanic white females had high prevalence of high WC.

B. Strengths and Limitations

Given that the subject pools of NHANES III and NAHSIT were large and
representative of the US. and Taiwanese populations, respectively, these are the
best data sets which to test the hypotheses. Findings in this study are
noteworthy, because for the first time differences among different body fat
distribution patterns could be examined by high and low WC and %BF in
over-weight but not obese people from two different populations. This is the first

report that demonstrates an independent and opposite relationship of WC and

127

%BF with serum HDL and fasting blood glucose in non-Hispanic white women in

both multiple regression analysis and in logistic regressions of this study.

This study has some limitations that warrant recognition. First, the
cross-sectional nature of this study precludes causal inferences about the
associations between WC, %BF, and MSRD. Ultimately, longitudinal data are
necessary to establish whether the predictive models accurately characterize
long-term risk for developing the MS. In addition, it would be desirable to control
more covariates in the regression models for the Taiwanese to parallel what was
done for the non-Hispanic whites. With hypertension so high in the Taiwanese
sample, it might have been desirable to include sodium data as a covariate as
well. Finally, there were some extreme cultural differences in the smoking
covariate, where in 98% of Taiwanese women reported never smoking then this
variable was dropped from the multiple regression analysis.

The %BF measured by BIA can be affected by factors that affect its validity
such as cold temperatures, type of food and recency of eating, beverages,
sweating, physical activity and other factors influencing conductivity and the body
hydration status (Kyle, Bosaeus et al., 2004; Kyle, Bosaeus et al., 2004). In this
study, Both BIA measurements in NHANES III and NAHSIT were carried out in
the morning after an overnight fast, a standard procedure necessary for accurate

measurements.

128

C. Implications

In this study, a narrowed range of BMI was used to test the hypotheses,
providing some important implications for clinical practice and public health where
it is desirable to target those over-weight, but not obese. The wide variation in
BF distribution found within a limited BMI range can help to evaluate the effect of
total BF and abdominal fat to metabolic risks more precisely than by using the
entire range of BMI’s. In both samples, at BMlz30, there were few subjects with
either normal WC or %BF. However, a narrowed range of BMI also raised some
difficulties for data analysis. A smaller sample size and a higher variation of %BF
within this narrowed range of BMI. This contributed to a wide range of standard
errors and confidence intervals. This high degree of variation means that was
more difficult to detect some relationships between BF and MSRD, especially in
the Taiwanese. Regardless, researchers and clinicians should be aware that for
people with BMI=25<30, WC is a better indicator of risk for MMS than is %BF or
even BMI. In order to screen high-risk individuals, health care professionals
should measure WC at all clinical visits and not just measure weight.

Based on the NCEP-ATP III definition, the MS criteria includes three or more
of following risk factors--high WC, high serum triglycerides, low HDL, high blood
pressure and high fasting blood glucose. In the present study, almost 90% of OW
but not obese non-Hispanic white women who had MS had high WC as one of the
three components. This might over estimate'the true MS in this group. i.e., for
those who with high WC might only have had high subcutaneous abdominal

adiposity instead of high intra-abdominal adiposity.

129

Findings from this study demonstrate that WC remained a strong positive
predictor for MMS and most of the co-morbidities. However, the finding that high
%BF did not predict the highest metabolic risks with increasing %BF percentiles
was consistent with the logistic regression demonstrating a protective effect of
%BF, especially in non-Hispanic white women. In that regard, future studies are
needed to demonstrate that a combination of risk factors, and particularly a
combination of insulin resistance syndrome features and indices of obesity, has

greater clinical utility than WC.

D. Recommendations for Future Research

In this study, results demonstrate an independent and opposite relationship
of WC and %BF with serum HDL and fasting blood glucose in non-Hispanic white
women. It is not clear if this phenomenon would also hold true for obese people.
That is to what extent is there a protective effect of total body adiposity to risk for
MS? More studies need to address this question to elucidate the true
relationship between BF distribution and the MS. Additionally, because neither
WC nor %BF can distinguish between subcutaneous BF and intra abdominal
adiposity, investigation of a new anthropometric screening tool might be useful for
a way to remove the thickness of the abdominal skinfold from the WC to assess
risk..

Racial and ethnic variation in metabolic risks complicates the determination
of body adiposity and its association with metabolic risks. In the present study, the

MS and co-morbidities were explained by WC and not by %BF. However, the

130

prevalence of high WC varied between non-Hispanic whites and Taiwanese
suggesting that further studies might be necessary to examine the possibility of a
differential disease/obesity relationship between Asians/Taiwanese and other
ethnic groups. Such studies should focus on the role of %BF, WC, and

subcutaneous versus intra abdominal fat in relation to risk for MS.

131

Appendix 1
Detailed laboratory procedures for metabolic risk and anthropometric

variables in NHANES III and NAHSIT

132

Both NHANES and NAHSIT data have similar sampling designs, data
collection procedures and analyses protocol. In this appendix, the NHANES data
are first described for each measurement and analysis procedure for each
variable that were used in this study. Then, any variables for which the analysis
differed for the NAHSIT, were described in the NAHSIT section.

Laboratory analyses data

In NHANES III, serum TG was measured enzymatically in serum or plasma
on a Hitachi 717 Analyzer (Hitachi, Tokyo, Japan) using commercial reagents
(Triglycerides/GPO reagent system pack, Boehringer Mannheim Diagnostics).
By using a series of coupled reactions in which triglycerides were hydrolyzed to
produce glycerol, glycerol was then oxidized by using glycerol oxidase and H202,
One of the reaction products, was measured quantitatively in a
peroxidase-catalyzed reaction that produces a color similar to that for cholesterol

and absorbancy is measured at 500 nm. The reaction sequence was as follows.

(1) Triglycerides lipase \ glycerol + fatty acids

 

(2) Glycerol + ATP glycerokinase glycerol-3-phosphate + ADP

 

High-density lipoprotein-cholesterol (HDL-C) was measured after precipitation
of other lipoproteins with a polyanion/divalent cation mixture based on the
following steps:

a. Add 100 pL of heparin sulfate-MnCI mixture to 1 mL of serum for each

133

sample. The reaction yields apo-B-containing lipoprotein, precipitate, and
soluble HDL-cholesterol

b. Remove precipitate by centrifuging at 1500 x g for 30 min. Remove the

clear supernatant and placing it into a 20-mL glass vial.

c. Place 500 pL of the supernatant and 50 pL sodium bicarbonate into an

Eppendorf tube. Vortex intermittently. Let the tubes stand at 20-25 C for
10 minutes; then centrifuge at 10,000 x g for two minutes.

d. Measure HDL-cholesterol in clear supernatant.

Two quality control pools were used to monitor the analysis of and triglyceride
and HDL. The precision of lipid and lipoprotein analyses was determined from
duplicate analysis of each pool analyzed in each run. The triglyceride control
pools (Q) and HDL-cholesterol control pools (MO and A0) were prepared by
Centers for Disease Control (CDC). These pools were also analyzed in
duplicate in every analytical run for a maximum of 50 runs. They were used to
assess accuracy with respect to CDC reference methods and analytical precision.

Blood pressure was reported as the average of six readings, three taken
during the household interview and three taken during the physical exam. At the
household interview, an interviewer measured blood pressure three times with the
participant resting quietly, sitting in a chair at home. A physician took three
additional blood pressure measurements in the mobile examination center. In
both venues, blood pressure was measured with a standard sphygmomanometer
(W. A. Baum, Copiague, NY). One of four cuff sizes (pediatric, regular adult,

large, or thigh) was chosen based on the circumference of the participant’s arm,

134

as indicated by the manufacturer's guidelines.

Plasma glucose was measured by using a hexokinase enzymatic method.
Venous whole blood was collected into a vacuum tube containing the glycolytic
inhibitors potassium oxalate and sodium fluoride and was centrifuged immediately
at 15009 for 10 minutes. The plasma was frozen at -70°C, shipped on dry ice to
the University of Missouri Diabetes Diagnostic Laboratory, and stored at -70°C
until analysis. Plasma glucose was measured using a modified hexokinase
enzymatic method (Cobas Mira assay; Roche, Basel, Switzerland), both within
assay and between-assay quality control procedures were used; the coefficient of

variation of the method was 1.6—3.7% during the six years of the survey.

Anthropometric variables

In NHANES III, weight and height were measured using standardized
techniques (DHHS 1997). Body weight was measured with an electronic load
cell scale to the nearest 0.01 kg (Toledo 8136 Scale). Participants wore only
personal undergarments and disposable paper shirts, pants and foam slippers
were provided, but no adjustment was made in the analyses from the minimal
clothing weight (0.18 kg).

Stature was measured using a fixed stadiometer (Holtain Height
Stadiometer). Participants stood on the floor board of the stadiometer with his or
her back to the vertical backboard of the stadiometer. The heels, buttocks,
scapulae and head were against the upright surface of the stadiometer with the

head positioned in the Frankfort horizontal plane. Hair ornaments, buns, braids,

135

etc. were removed to obtain an accurate measurement. Afterwards the
examiner read the measurement and the assistant recorded it to the nearest 0.1
cm.

Waist circumference was measured in a standing position. The examiner
marked a horizontal line at the high point of the iliac crest and then crossed the
line to indicate the mid-axillary line of the body. The examiner then stood on the
participant’s right side and placed the measuring tape around the trunk in a
horizontal plane at this level and marked on the right side of the trunk. The
measurement was made at minimal respiration to the nearest 0.1 cm.

NHANES III used the following steps to ensure the quality of
measurements: 1) one participant was selected by the computer to have body
measurements replicated by another technician every other day; 2) tolerance
levels/ranges were set for weight, height and waist circumference measurement
to function as a quality control measure by minimizing possible measuring and
recording errors; 3) anthropometric technicians were periodically observed by
the body measurement consultant to ensure standardization. The consultant
reviewed any deviations from the protocol with the technicians. There was no
validation testing described in the NHANES III operation and reference manual
(DHHS 1996).

A body composition analyzer for BIA (model 19908; Valhalla Scientific,
San Diego) was used for the measurement of whole body electrical resistance
and impedance. During the BIA procedure, subjects lay supine on a

nonconductive examination table without a pillow under their head. The

136

technician instructed the subject to remain motionless and relaxed with their arms
and legs slightly apart and not touching any other part of the body. Participants
had a single, tetrapolar BIA measurement of resistance (Res) and reactance at 50
kHz taken between the right wrist and ankle while in a supine position. The
percent body fat estimates were derived from prediction equations for fat-free
mass (FF M) that were validated and cross-validated for men and women
separately and for blacks and whites between the ages of 12 and 94 y. The
converted NHANES III, RJL resistance values were used to calculate the
impedance index of stature in cm squared divided by resistance (Slees)in the
TBW and FFM prediction equations (Zhu, Wang et al. 2003). The equations for
men and for women were as follows:

Males TBW =1.203 +0.176 weight +0449 Slees

Females TBW=3.747 + 0.113 weight + 0.45 Slees

Males FFM =10.678 + 0:262 weight + 0.652 Slees + 0.015 Res

Females FFM =9.529 + 0.168 weight + 0.696 SZIRes + 0.016 Res

The National Institute of Standards and Technology independently certified
measurement accuracy of the Valhalla impedance machines used in NHANES III
after the survey. Detailed information on the BIA procedure is presented
elsewhere (Kuczmarski 1996; NCHS. 1998). The percent body fat was
calculated as follows:

Total body Fat = weight — Fat Free Mass and %BF = BF/weight.

137

NAHSIT

Laboratory, clinical and anthropometric measurements were identical for
NHASIT with three exceptions: 1) fasting whole-blood glucose was measured
with heparinized blood by the glucose oxidase method using a glucose analyzer
(Model 23A, YSI, USA) immediately after blood drawing. The blood sample was
not frozen as in NHANES and a different laboratory was used in the analysis.
For NAHSIT, a Hitachi 747 analyzer (Hitachi 717 analyzer in NHANES III) was
used to measure serum triglyceride, cholesterol, and HDL-C values in the central
laboratory; 2) blood pressure was measured by standard sphygmomanometers at
home after the participants rested at least five minutes. A cuff of appropriate size
was used for each participant. Two, and not six, blood pressure measurements
were made 30 seconds apart with the arm at the level of the heart. If the two
measurements were more than 10 mmHg apart, a third measurement was made.
The two closest blood pressure values were averaged to obtain the mean blood
pressure; 3) The BIA machine used in NAHSIT was the SELCO SIF891 and the
formula used to calculate percent body fat were based on healthy Japanese male

and females.

Covariates.

The covariates in NAHSIT were continuous rather than categorical such as
years of education and income. These were converted to ranked categories as
similar as possible to the NHANES data. In NAHSIT, physical activity was

assessed using both the frequency and average time spent on activities in the

138

preceding month, but in NHANESIII only frequency was asked. Occupation
related activities were also included in the NAHSIT but not in NHANES III. For
comparison purpose, only frequency of leisure time activities of NAHSIT data was
used in this study and was converted to ranked categories use the same method

as for NHANES III.

139

BIBLIOGRAPHY

140

BIBLIOGRAPHY

Abate, N., A. Garg, et al. (1997). "Prediction of total subcutaneous abdominal,
intraperitoneal, and retroperitoneal adipose tissue masses in men by a
single axial magnetic resonance imaging slice." Am J Clin Nutr 65(2):
403-8.

Abbasi, F., T. McLaughlin, et al. (2000). "High carbohydrate diets, triglyceride-rich
lipoproteins, and coronary heart disease risk." Am J Cardiol 85(1): 45-8.

Abbott, W. G., S. Lillioja, et al. (1987). "Relationships between plasma lipoprotein
concentrations and insulin action in an obese hyperinsulinemic population."
Diabetes 36(8): 897-904.

Abe, T. and T. F ukunaga (1994). "Relationships between subcutaneous fat and
muscle distributions and serum HDL-cholesterol." J Atheroscler Thromb
1(1): 15-22.

Alberti, K. G. (1998). "Impaired glucose tolerance: what are the clinical
implications?" Diabetes Res Clin Pract 40 Suppl: 83-8.

Alberti, K. G. and P. Z. Zimmet (1998). "Definition, diagnosis and classification of
diabetes mellitus and its complications. Part 1: diagnosis and classification
of diabetes mellitus provisional report of a WHO consultation." Diabet Med
15(7): 539-53.

Alexander, C. M., P. B. Landsman, et al. (2003). "NCEP-defined metabolic
syndrome, diabetes, and prevalence of coronary heart disease among
NHANES III participants age 50 years and older." Diabetes 52(5): 1210-4.

Apridonidze, T., P. A. Essah, et al. (2005). "Prevalence and characteristics of the
metabolic syndrome in women with polycystic ovary syndrome." J Clin
Endocrinol Metab 90(4): 1929-35.

Arcaro, G., A. Cretti, et al. (2002). "Insulin causes endothelial dysfunction in
humans: sites and mechanisms." Circulation 105(5): 576-82.

Arner, P. (1995). "Differences in lipolysis between human subcutaneous and
omental adipose tissues." Ann Med 27(4): 435-8.

Arner, P. (1997). "Regional adipocity in man." J Endocrinol 155(2): 191-2.

Astrup, A., G. K. Grunwald, et al. (2000). "The role of low-fat diets in body weight
control: a meta-analysis of ad libitum dietary intervention studies." Int J
Qbes Relat Metab Disord 24(12): 1545-52.

Banks, R. E., M. A. Forbes, et al. (1995). "The acute phase protein response in
patients receiving subcutaneous IL-6." Clin Exp Immunol 102(1): 217-23.

141

Baron, A. D. (1994). "Hemodynamic actions of insulin." Am J Physiol 267(2 Pt 1):
E187-202.

Barrett-Connor, E. (1997). "Sex differences in coronary heart disease. Why are
women so superior? The 1995 Ancel Keys Lecture." Circulation 95(1):
252-64. '

Bergman, R. N., S. P. Kim, et al. (2007). "Abdominal obesity: role in the
pathophysiology of metabolic disease and cardiovascular risk." Am J Med
120(2 Suppl 1): S3-8; discussion 829-32.

Bierman, E. L., V. P. Dole, et al. (1957). "An abnormality of nonesterified fatty acid
metabolism in diabetes mellitus." Diabetes 6(6): 475-9.

Bigaard, J., I. Spanggaard, et al. (2005). "Self-reported and technician-measured
waist circumferences differ in middle-aged men and women." J Nutr
135(9): 2263-70.

Bjorntorp, P. (1990). ""Portal" adipose tissue as a generator of risk factors for
cardiovascular disease and diabetes." Arteriosclerosis 10(4): 493-6.

Bjorntorp, P. (1992). "Abdominal fat distribution and disease: an overview of
epidemiological data." Ann Med 24(1): 15-8.

Bolanowski, M. and B. E. Nilsson (2001). "Assessment of human body
composition using dual-energy x-ray absorptiometry and bioelectrical
impedance analysis." Med Sci Monit 7(5): 1029-33.

Bonora, E., S. Kiechl, et al. (1998). "Prevalence of insulin resistance in metabolic
disorders: the Bruneck Study." Diabetes 47(10): 1643-9.

Bonora, E., R. Micciolo, et al. (1995). "Is it possible to derive a reliable estimate of
human visceral and subcutaneous abdominal adipose tissue from simple
anthropometric measurements?" Metabolism 44(12): 1617-25.

Borkan, G. A., S. G. Gerzof, et al. (1982). "Assessment of abdominal fat content
by computed tomography." Am J Clin Nutr 36(1): 172-7.

Bosy—Westphal, A., C. Geisler, et al. (2006). "Value of body fat mass vs
anthropometric obesity indices in the assessment of metabolic risk factors."
Int J Obes (Long) 30(3): 475-83.

Bray, G. A. (2003). "Evaluation of obesity. Who are the obese?" Postgrad Med
114(6): 19-27, 38.

Bullo, M., P. Garcia-Lorda, et al. (2003). "Systemic inflammation, adipose tissue
tumor necrosis factor, and leptin expression." Obes Res 11(4): 525-31.

142

Busetto, L. (2001). "Visceral obesity and the metabolic syndrome: effects of
weight loss." Nutr Metab Cardiovasc Dis 11(3): 195-204.

Byrne, C. D., N. J. Wareham, et al. (1994). "Hypertriglyceridaemia in subjects with
normal and abnormal glucose tolerance: relative contributions of insulin
secretion, insulin resistance and suppression of plasma non-esterified fatty
acids." Diabetologia 37(9): 889-96.

Carr, D. B., K. M. Utzschneider, et al. (2004). "Intra-abdominal fat is a major
determinant of the National Cholesterol Education Program Adult
Treatment Panel lII criteria for the metabolic syndrome." Diabetes 53(8):
2087-94.

Chandalia, M., N. Abate, et al. (1999). "Relationship between generalized and
upper body obesity to insulin resistance in Asian Indian men." J Clin
_E_r_t_docrinol Metab 84(7): 2329-35.

Chen, J., R. P. Wildman, et al. (2004). "Association between inflammation and
insulin resistance in US. nondiabetic adults: results from the Third National
Health and Nutrition Examination Survey." Diabetes Care 27(12): 2960-5.

Chen, Y. D., A. M. Coulston, et al. (1995). 'Why do low-fat high-carbohydrate
diets accentuate postprandial lipemia in patients with NIDDM?" Diabetes
Care 18(1): 10-6.

Chuang, S. Y., C. H. Chen, et al. (2004). "Prevalence of metabolic syndrome in a
large health check-up population in Taiwan." J Chin Med Assoc 67(12):
61 1-20.

Chumlea, W. C. and S. S. Guo (1994). "Bioelectrical impedance and body
composition: present status and future directions." Nutr Rev 52(4): 123-31.

Chumlea, W. C., S. S. Guo, et al. (2002). "Body composition estimates from
NHANES III bioelectrical impedance data." Int J Obes Relat Metab Disord
26(12): 1596-609.

Colin Bell, A., L. S. Adair, et al. (2002). "Ethnic differences in the association
between body mass index and hypertension." Am J Epidemiol 155(4):
346-53.

Coppack, S. W., R. D. Evans, et al. (1992). "Adipose tissue metabolism in obesity:
lipase action in vivo before and after a mixed meal." Metabolism 41(3):
264-72.

Cossrow, N. and B. Falkner (2004). "Race/ethnic issues in obesity and
. obesity-related comorbidities," J Clin Endgcrinol Metab 89(6): 2590-4.

143

Costacou, T. and E. J. Mayer-Davis (2003). "Nutrition and prevention of type 2
diabetes." Anng Rev Nutr 23: 147-70.

Cunningham, M. J., D. K. Clifton, et al. (1999). "Leptin's actions on the
reproductive axis: perspectives and mechanisms." Biol Reprod 60(2):
216-22.

de Lorgeril, M., P. Salen, et al. (1999). "Mediterranean diet, traditional risk factors,
and the rate of cardiovascular complications after myocardial infarction:
final report of the Lyon Diet Heart Study." Circulation 99(6): 779-85.

Dela, F., K. J. Mikines, et al. (1992). "Effect of training on insulin-mediated
glucose uptake in human muscle." Am J Physiol 263(6 Pt 1): E1134-43.

Despres, J. P. (2006). "Is visceral obesity the cause of the metabolic syndrome?"
Ann Med 38(1): 52-63.

Deurenberg-Yap, M., S. K. Chew, et al. (2002). "Elevated body fat percentage
and cardiovascular risks at low body mass index levels among
Singaporean Chinese, Malays and Indians." Obes Rev 3(3): 209-15.

Deurenberg-Yap, M. and P. Deurenberg (2003). "Is a re-evaluation of WHO body
mass index cut-off values needed? The case of Asians in Singapore." Nutr
_Rfl 61(5 Pt 2): 880-7. '

Deurenberg, P., M. Yap, et al. (1998). "Body mass index and percent body fat: a
meta analysis among different ethnic groups." Int J Obes Relat Metab
Disord 22(12): 1164-71.

DHHS, N. C. f. H. S. (1996). Third National Health and Nutrition Examination
Survey, (NHANES III, 1988-1994) analytic and reporting guidelines (CD-ROM).
;. Hyattsville, MD:, Centers for Disease Control and Prevention.

DHHS, N. C. f. H. S. (2000). The practical guide-identification, evaluation, and
treatment of overweight and obesity in adults. Bethesda, MD, National
Institutes of Health,.

DHHS, N. C. f. H. S. U. (1997). The Third National Health and Nutrition
Examination Survey, Ill 1988-94. Reference Manuals and Reports
[CD-ROM]. Hyattsville, MD, Center for Disease Control and Prevention.

Diabetes Prevention Program Research Group (2002). "The Diabetes Prevention
Program (DPP): description of lifestyle intervention." _Diabetes Care 25(12):
2165-71.

Dickinson, 8., S. Colagiuri, et al. (2002). "Postprandial hyperglycemia and insulin

sensitivity differ among lean young adults of different ethnicities." J Nutr
132(9): 2574-9.

144

Dobbelsteyn, C. J., M. R. Joffres, et al. (2001). "A comparative evaluation of waist
circumference, waist-to-hip ratio and body mass index as indicators of
cardiovascular risk factors. The Canadian Heart Health Surveys." ELL
Qbes Relat Metab_Disorg_ 25(5): 652-61.

Donahue, R. P. and R. D. Abbott (1987). "Central obesity and coronary heart
disease in men." Lancet 2(8569): 1215.

Dvorak, R. V., W. F. DeNino, et al. (1999). "Phenotypic characteristics associated
with insulin resistance in metabolically obese but normal-weight young
women." Diabetes 48(11): 2210-4.

Egan, B. M., G. Lu, et al. (1999). "Vascular effects of non-esterified fatty acids:
implications for the cardiovascular risk factor cluster." Prostaglangins
ngkot Essent Fatty Acid_s 60(5-6): 411-20.

Einhorn, D., G. M. Reaven, et al. (2003). "American College of Endocrinology
position statement on the insulin resistance syndrome." Endocr Pract 9(3):
237-52.

Ekelund, L. G., W. L. Haskell, et al. (1988). "Physical fitness as a predictor of
cardiovascular mortality in asymptomatic North American men. The Lipid
Research Clinics Mortality Follow-up Study." N Engl J Med 319(21):
1379-84.

El-Haschimi, K., D. D. Pierroz, et al. (2000). "Two defects contribute to
hypothalamic leptin resistance in mice with diet-induced obesity." J Clin
Invest 105(12): 1827-32.

Ellis, K. J. (2001). "Selected body composition methods can be used in field
studies." J Nutr 131(5): 15898-958.

Eriksson, J., S. Taimela, et al. (1997). "Exercise and the metabolic syndrome."
Diabetologia 40(2): 125-35.

Esposito, K., R. Marfella, et al. (2004). "Effect of a mediterranean-style diet on
endothelial dysfunction and markers of vascular inflammation in the
metabolic syndrome: a randomized trial." Jama 292(12): 1440-6.

Fain, J. N., A. K. Madan, et al. (2004). "Comparison of the release of adipokines
by adipose tissue, adipose tissue matrix, and adipocytes from visceral and
subcutaneous abdominal adipose tissues of obese humans."
Endocrinology 145(5): 2273-82.

Farrell, S. W., L. Braun, et al. (2002). "The relation of body mass index,
cardiorespiratory fitness, and all-cause mortality in women." Obes Res
10(6): 417-23.

145

Farrell, S. W., Y. J. Cheng, et al. (2004). "Prevalence of the metabolic syndrome
across cardiorespiratory fitness levels in women." Obes Res 12(5): 824-30.

Fay, W. P. (2004). "Plasminogen activator inhibitor 1, fibrin, and the vascular
response to injury." Trends Cardiovasc Med 14(5): 196-202.

Ferreira, l., M. B. Snijder, et al. (2004). "Central fat mass versus peripheral fat and
lean mass: opposite (adverse versus favorable) associations with arterial
stiffness? The Amsterdam Growth and Health Longitudinal Study." J Clin
Endocrinol Metab 89(6): 2632-9.

Feskens, E. J., S. M. Virtanen, et al. (1995). "Dietary factors determining diabetes
and impaired glucose tolerance. A 20-year follow-up of the Finnish and
Dutch cohorts of the Seven Countries Study." Diabetes Care 18(8):
1104-12.

Finkelstein, E. A., l. C. Fiebelkorn, et al. (2004). "State-level estimates of annual
medical expenditures attributable to obesity." Obes Res 12(1): 18-24.

Finkelstein, M. M. (2001). "Obesity, cigarette smoking and the cost of physicians'
services in Ontario." Can J PQlic Health 92(6): 437-40.

Flegal, K. M., B. I. Graubard, et al. (2005). "Excess deaths associated with
undenrveight, overweight, and obesity." JAMA 293(15): 1861-7.

Ford, E. S. and W. H. Giles (2000). "Serum C-reactive protein and fibrinogen
concentrations and self-reported angina pectoris and myocardial infarction:
findings from National Health and Nutrition Examination Survey Ill." J Clin
Epidemiol 53(1): 95-102.

Ford, E. S., W. H. Giles, et al. (2002). "Prevalence of the metabolic syndrome
among US adults: findings from the third National Health and Nutrition
Examination Survey." Jama 287(3): 356-9.

Foster, K. R. and H. C. Lukaski (1996). 'Whole-body impedance--what does it
measure?" Am J Clin Nlfl 64(3 Suppl): 3888-3968.

Franz, M. J., J. P. Bantle, et al. (2002). "Evidence-based nutrition principles and
recommendations for the treatment and prevention of diabetes and related
complications." Diabetes Care 25(1): 148-98.

Frayn, K. N. (2002). "Adipose tissue as a buffer for daily lipid flux." Diabetologia
45(9): 1201-10.

Frayn, K. N., F. Karpe, et al. (2003). "Integrative physiology of human adipose
tissue." Int J Obes Relat Metab Disord 27(8): 875-88.

146

Frayn, K. N. and B. Vessby (2000). "Foreword " The British Jogrnal of Ngtrition
83(Supplement s1): S1-S4

Freedland, E. S. (2004). "Role of a critical visceral adipose tissue threshold
(CVA'I‘I') in metabolic syndrome: implications for controlling dietary
carbohydrates: a review." Mtr Metab (Lond) 1(1): 12.

Frost, G., A. A. Leeds, et al. (1999). "Glycaemic index as a determinant of serum
HDL-cholesterol concentration." Lancet 353(9158): 1045-8.

Gallagher, 0., M. Visser, et al. (1996). "How useful is body mass index for
comparison of body fatness across age, sex, and ethnic groups?" Am J
Epidemiol 143(3): 228-39.

Gardner, C. D. and H. C. Kraemer (1995). "Monounsaturated versus
polyunsaturated dietary fat and serum lipids. A meta-analysis." Arterioscler
Thromb Vasc Biol 15(11): 1917-27.

Garg, A. (1998). "High-monounsaturated-fat diets for patients with diabetes
mellitus: a meta-analysis." Am J Clin Nutr 67(3 Suppl): 5778-5828.

Garg, A., J. P. Bantle, et al. (1994). "Effects of varying carbohydrate content of
diet in patients with non-insulin-dependent diabetes mellitus." Jama
271(18): 1421—8.

Ginsberg, H. N. and L. S. Huang (2000). "The insulin resistance syndrome: impact
on lipoprotein metabolism and atherothrombosis." J Cardiovasc Risk 7(5):
325-31.

Groop, L. and M. Orho—Melander (2001). "The dysmetabolic syndrome." J Intern
Med 250(2): 105-20.

Grundy, S. M., H. B. Brewer, Jr., et al. (2004). "Definition of metabolic syndrome:
report of the National Heart, Lung, and Blood Institute/American Heart
Association conference on scientific issues related to definition."
Arterioscler Thromb VascfloJ 24(2): e13-8.

Grundy, S. M., J. I. Cleeman, et al. (2005). "Diagnosis and management of the
metabolic syndrome. An American Heart Association/National Heart, Lung,
and Blood Institute Scientific Statement. Executive summary." CarrLol Rev
13(6): 322-7.

Grundy, S. M., J. I. Cleeman, et al. (2005). "Diagnosis and management of the
metabolic syndrome: an American Heart Association/National Heart, Lung,
and Blood Institute Scientific Statement." Circulation 112(17): 2735-52.

Gu, D., K. Reynolds, et al. (2005). "Prevalence of the metabolic syndrome and
overweight among adults in China." Lancet 365(9468): 1398-405.

147

Guerin, M., W. Le Goff, et al. (2001). "Atherogenic role of elevated CE transfer
from HDL to VLDL(1) and dense LDL in type 2 diabetes : impact of the
degree of triglyceridemia." Arterioscler Thromb Vasc m 21(2): 282-8.

Guo, S. M., A. F. Roche, et al. (1989). "Fat-free mass in children and young adults
predicted from bioelectric impedance and anthropometric variables." Am J
Clin Nutr 50(3): 435-43.

Gustafsson, K., N. G. Asp, et al. (1993). "Effects of different vegetables in mixed
meals on glucose homeostasis and satiety." Eur J Clin Nutr 47(3):
192-200.

Hajjar, I. and T. A. Kotchen (2003). "Trends in prevalence, awareness, treatment,
and control of hypertension in the United States, 1988-2000." Jama 290(2):
199-206.

Hamdy, O. (2005). "Lifestyle modification and endothelial function in obese
subjects." Expert Rev Cardiovasc Ther 3(2): 231—41.

Han TS, S. J., Currall JE, Morrison CE, Deurenberg P, Lean ME. (1997). "The
influences of height and age on waist circumferences as an index of
adiposity in adults." Int J Obes Relat Metab Disord 12: 83-9.

Han, T. S., K. Williams, et al. (2002). "Analysis of obesity and hyperinsulinemia in
the development of metabolic syndrome: San Antonio Heart Study." Obes
Q 10(9): 923-31.

Hansel, B., P. Giral, et al. (2004). "Metabolic syndrome is associated with
elevated oxidative stress and dysfunctional dense high-density lipoprotein
particles displaying impaired antioxidative activity." J Clin Endocrinol Metab
89(10): 4963-71.

He, Q., E. S. Engelson, et al. (2004). "Validation of an elliptical anthropometric
model to estimate visceral compartment area." Obes Res 12(2): 250-7.

Hermansen, K. (2000). "Diet, blood pressure and hypertension." Br J Nutr 83
Suppl 1: 8113-9.

Heymsfield, S., D. Allison, et al. (1995). Assessment of human body composition.
Handbook of Assessment Methods for Eating Behaviors and
Weight-RelatedrPgblems. A. DB. Thousand Oaks, CA, SAGE
Publications, Inc: 515-560.

Ho, S. Y., T. H. Lam, et al. (2003). 'Waist to stature ratio is more strongly

associated with cardiovascular risk factors than other simple
anthropometric indices." Ann Epidemiol 13(10): 683-91 .

148

Hollenbeck, C. B., N. Chen, et al. (1984). "Relationship between the plasma
insulin response to oral glucose and insulin-stimulated glucose utilization in
normal subjects." Diabetes 33(5): 460-3.

Hollenbeck, C. B. and A. M. Coulston (1987). "Effect of variation in diet on
lipoprotein metabolism in patients with diabetes mellitus." Diat_)e_tes Metat;
_ng 3(3): 669-89.

Hotta, K., T. Funahashi, et al. (2001). "Circulating concentrations of the adipocyte
protein adiponectin are decreased in parallel with reduced insulin
sensitivity during the progression to type 2 diabetes in rhesus monkeys."
Diabetes 50(5): 1126-33.

Howard, B. V. and W. J. Howard (1994). "Dyslipidemia in non-insulin-dependent
diabetes mellitus." Endocr Rev 15(3): 263-74.

Hutley, L. and J. B. Prins (2005). "Fat as an endocrine organ: relationship to the
metabolic syndrome." Am J Med Sci 330(6): 280-9.

Hyun, Y. J., O. Y. Kim, et al. (2008). "Evaluation of metabolic syndrome risk in
Korean premenopausal women: not waist circumference but visceral fat."
Circ J 72(8): 1308-15.

International Diabetes Federation (2005). The IDF consensus worldwide definition
of the metabolic syndrome.

lsomaa, B. (2003). "A major health hazard: the metabolic syndrome." Life Sci
73(19): 2395-411.

Jakicic, J. M., R. R. Wing, et al. (1998). "Bioelectrical impedance analysis to
assess body composition in obese adult women: the effect of ethnicity." [M
J Obes Relat Metal;_Disorq 22(3): 243-9.

Janssen, l., S. B. Heymsfield, et al. (2002). "Body mass index and waist
circumference independently contribute to the prediction of nonabdominal,
abdominal subcutaneous, and visceral fat." Am J Clin Ntfl 75(4): 683-8.

Janssen l, H. 8., Allison DB, Kotler DP, Ross R (2002). "Body mass index and
waist circumference independently contribute to the prediction of
nonabdominal subcutaneous, and isceral fat." Am J Clin Nutr 75: 683-8.

Janssen, I., P. T. Katzmarzyk, et al. (2005). "Comparison of overweight and
obesity prevalence in school-aged youth from 34 countries and their
relationships with physical activity and dietary patterns." Obes Rev 6(2):
123-32.

149

Janssen, I., P. T. Katzmarzyk, et al. (2002). "Body mass index, waist
circumference, and health risk: evidence in support of current National
Institutes of Health guidelines." Arch Intern Med 162(18): 2074-9.

Janssen, I., P. T. Katzmarzyk, et al. (2004). "Waist circumference and not body
mass index explains obesity-related health risk." Am J Clin Nutr 79(3):
379-84.

Janssen, l., P. T. Katzmarzyk, et al. (2004). "Fitness alters the associations of BMI
and waist circumference with total and abdominal fat." Obes Res 12(3):
525-37.

Jenkins, D. J. and A. L. Jenkins (1995). "Nutrition principles and diabetes. A role
for "Iente carbohydrate"?" Diabetes Care 18(11): 1491-8.

Jenkins, D. J., C. W. Kendall, et al. (2000). "Viscous and nonviscous fibres,
nonabsorbable and low glycaemic index carbohydrates, blood lipids and
coronary heart disease." Curr Opin Lipidol 11(1): 49-56.

Jeppesen, J., H. O. Hein, et al. (2001). "Low triglycerides-high high-density
lipoprotein cholesterol and risk of ischemic heart disease." Arch Intern Meg
161(3): 361-6.

Jia, W. P., K. S. Xiang, et al. (2004). "[A comparison of the application of two
working definitions of metabolic syndrome in Chinese populationj."
@mghua Yi Xue Za Zhi 84(7): 534-8.

Jones, D. W. (1995). "Hypertension in East Asia." Am J Hypertens 8(12 Pt 2):
111s-114s.

Kanazawa, M., N. Yoshiike, et al. (2002). "Criteria and classification of obesity in
Japan and Asia-Oceania." Asia Pac J Clin Nutr 11 Suppl 8: $732-$737.

Kappes, A. and G. Loffler (2000). "Influences of ionomycin, dibutyryI-cycloAMP
and tumour necrosis factor-alpha on intracellular amount and secretion of
apM1 in differentiating primary human preadipocytes." Horm Metag Res
32(11-12): 548-54.

Karelis, A. D., M. Faraj, et al. (2005). "The metabolically healthy but obese
individual presents a favorable inflammation profile." J Clin Endocrinol
Metab 90(7): 4145-50.

Kashyap, M. L., F. Magill, et al. (1970). "Insulin and non-esterified fatty acid
metabolism in asymptomatic diabetics and atherosclerotic subjects." Can
MecLAssoc J 102(11): 1165-9.

Katsuki, A., Y. Sumida, et al. (2004). "Increased oxidative stress is associated
with serum levels of triglyceride, insulin resistance, and hyperinsulinemia in

150

Japanese metabolically obese, normal-weight men." Diabetes Care 27(2):
631-2.

Katzmarzyk, P. T., T. S. Church, et al. (2005). "Metabolic syndrome, obesity, and
mortality: impact of cardiorespiratory fitness." Diabetes Care 28(2): 391-7.

Katzmarzyk, P. T., A. S. Leon, et al. (2003). "Targeting the metabolic syndrome
with exercise: evidence from the HERITAGE Family Study." Med Sci
Sports Exerc 35(10): 1703-9.

Kelley, D. E. and B. H. Goodpaster (2001). "Effects of exercise on glucose
homeostasis in Type 2 diabetes mellitus." Med Sci Sports Exerc 33(6
Suppl): 8495-501; discussion 8528-9.

Keys, A. (1997). "Coronary heart disease in seven countries. Nutrition 13(3):
250-252.

Kiechl, S., J. Willeit, et al. (1996). "Insulin sensitivity and regular alcohol
consumption: large, prospective, cross sectional population study (Bruneck
study)" _B_rpj 313(7064): 1040-4.

Kim, C. and G. L. Beckles (2004). "Cardiovascular disease risk reduction in the
Behavioral Risk Factor Surveillance System." Am J Prev Meg 27(1): 1-7.

Kim, D. M., C. W. Ahn, et al. (2005). "Prevalence of obesity in Korea." Obes Rev
6(2): 117-21.

Kim, H. S., F. Abbasi, et al. (2001). "Effect of insulin resistance on postprandial
elevations of remnant lipoprotein concentrations in postmenopausal
women." Am J Clin Nutr 74(5): 592-5.

Kim, Y., Y. K. Suh, et al. (2004). "BMI and metabolic disorders in South Korean
adults: 1998 Korea National Health and Nutrition Survey." Obes Res 12(3):
445-53.

Kishida, K., H. Kuriyama, et al. (2000). "Aquaporin adipose, a putative glycerol
channel in adipocytes." J Biol Chem 275(27): 20896-902.

Kissebah, A., N. Vydelingum, et al. (1996). "Predicting body composition by
densitometry from simple anthropometric measurements." Am J Clin Nu_tr
63: 4-14.

Klein, 8., J. F. Horowitz, et al. (2000). "Leptin production during early starvation in
lean and obese women." Am J thsiol Endpcrinol Metap 278(2): E280-4.

Kuczmarski, R. J. and K. M. Flegal (2000). "Criteria for definition of oveniveight in

transition: background and recommendations for the United States." Am J
Clin Nutr 72(5): 1074-81.

151

Kuczmarski, R. J., K. M. Flegal, et al. (1994). "Increasing prevalence of
ovenrveight among US adults. The National Health and Nutrition
Examination Surveys, 1960 to 1991." Jama 272(3): 205-11.

Kvist, H., B. Chowdhury, et al. (1988). "Total and visceral adipose-tissue volumes
derived from measurements with computed tomography in adult men and
women: predictive equations." Am J Clin Nutr 48(6): 1351-61.

Kyle, U. G., I. Bosaeus, et al. (2004). "Bioelectrical impedance analysisupart I:
review of principles and methods." Clin Nutr 23(5): 1226-43.

Kyle, U. G., I. Bosaeus, et al. (2004). "Bioelectrical impedance analysis-part II:
utilization in clinical practice." Clin Nutr 23(6): 1430-53.

Laakso, M., H. Sarlund, et al. (1990). "Insulin resistance is associated with lipid
and lipoprotein abnormalities in subjects with varying degrees of glucose
tolerance." Arteriosclerosis 10(2): 223-31.

Laaksonen, D. E., H. M. Lakka, et al. (2002). "Low levels of leisure-time physical
activity and cardiorespiratory fitness predict development of the metabolic
syndrome." Diabetes Care 25(9): 1612-8.

Lafontan, M. and M. Berlan (2003). "Do regional differences in adipocyte biology
provide new pathophysiological insights?" Trends Pharrnacol Sci 24(6):
276-83.

Laws, A. and G. M. Reaven (1992). "Evidence for an independent relationship
between insulin resistance and fasting plasma HDL-cholesterol,
triglyceride and insulin concentrations." J Intern Meg 231(1): 25-30.

Lean, M. E., T. S. Han, et al. (1996). "Predicting body composition by
densitometry from simple anthropometric measurements." Am J Clin Nutr
63(1): 4-14.

Lean, M. E., T. S. Han, et al. (1995). 'Waist circumference as a measure for
indicating need for weight management." _B_lT_lj 311(6998): 158-61.

Lee, W. Y., J. S. Park, et al. (2004). "Prevalence of the metabolic syndrome
among 40,698 Korean metropolitan subjects." Diabetes Res Clin Pra_c_t
65(2): 143-9.

Leite, C. C., B. L. Wajchenberg, et al. (2002). "lntra-abdominal thickness by
ultrasonography to predict risk factors for cardiovascular disease and its
correlation with anthropometric measurements." Metatfism 51(8):
1034-40.

152

Lemieux, S., D. Prud'homme, et al. (1993). "Sex differences in the relation of
visceral adipose tissue accumulation to total body fatness." Am J Clin Nu_tr
58(4): 463-7.

Lewis, G. F. and D. J. Rader (2005). "New insights into the regulation of HDL
metabolism and reverse cholesterol transport." Circ Res 96(12): 1221-32.

Li, C., G. Engstrom, et al. (2006). "Sex-specific cardiovascular morbidity and
mortality in a cohort treated for hypertension." J vaertens 24(8): 1523-9.

Li, Z. Y., G. B. Xu, et al. (2006). "Prevalence rate of metabolic syndrome and
dyslipidemia in a large professional population in Beijing." finerosclerosis
184(1): 188-92.

Lihn, A. S., S. B. Pedersen, et al. (2005). "Adiponectin: action, regulation and
association to insulin sensitivity." Obes Rev 6(1): 13-21.

Lin, Y. C., L. L. Yen, et al. (2003). "Prevalence of overweight and obesity and its
associated factors: findings from National Nutrition and Health Survey in
Taiwan, 1993-1996." Prev Med 37(3): 233-41.

Ling, B. N., H. Matsunaga, et al. (1995). "Role of growth factors in mesangial cell
ion channel regulation." Kidney Int 48(4): 1158-66.

Liu, S., J. E. Manson, et al. (2000). "Fruit and vegetable intake and risk of
cardiovascular disease: the Women's Health Study." Am J Clin Nlfl 72(4):
922-8.

Liu, S., J. E. Manson, et al. (2000). "A prospective study of whole-grain intake and
risk of type 2 diabetes mellitus in US women." Am J IMIIC Hea_lt_h 90(9):
1409-15.

Lord, G. (2002). "Role of leptin in immunology." Nutr Rev 60(10 Pt 2): 335-8;
discussion 868-84, 85-7.

Lukaski, H. C., W. W. Bolonchuk, et al. (1986). "Validation of tetrapolar
bioelectrical impedance method to assess human body composition." _J_
Appl Physiol 60(4): 1327-32.

Marshall, J. A., D. H. Bessesen, et al. (1997). "High saturated fat and low starch
and fibre are associated with hyperinsulinaemia in a non-diabetic
population: the San Luis Valley Diabetes Study." Diabetologia 40(4):
430-8.

Mayer-Davis, E. J., R. D'Agostino, Jr., et al. (1998). "Intensity and amount of

physical activity in relation to insulin sensitivity: the Insulin Resistance
Atherosclerosis Study." Jama 279(9): 669-74.

153

McCarty, M. F. (2003). "A paradox resolved: the postprandial model of insulin
resistance explains why gynoid adiposity appears to be protective." Med
Hypotheses 61(2): 173-6.

Mensink, R. P. and M. B. Katan (1992). "Effect of dietary fatty acids on serum
lipids and lipoproteins. A meta-analysis of 27 trials." Arterioscler Thromfib
12(8): 911-9.

Minocci, A., G. Savia, et al. (2000). "Leptin plasma concentrations are dependent
on body fat distribution in obese patients." Int J Obes Relat Metabgisord
24(9): 1139-44.

Misra, A., J. S. Wasir, et al. (2005). 'Waist circumference criteria for the diagnosis
of abdominal obesity are not applicable uniformly to all populations and
ethnic groups." Nutrition 21(9): 969-76.

Mokdad, A. H., E. 8. Ford, et al. (2003). "Prevalence of obesity, diabetes, and
obesity-related health risk factors, 2001." Jama 289(1): 76-9.

Montague, C. T., J. B. Prins, et al. (1998). "Depot-related gene expression in
human subcutaneous and omental adipocytes." Diabetes 47(9): 1384-91.

Nagaya, T., H. Yoshida, et al. (1999). "Body mass index (weight/height2) or
percentage body fat by bioelectrical impedance analysis: which variable
better reflects serum lipid profile?" Int J Obes Relat Metab_Disord 23(7):
771-4.

NCEP (2001). "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 llI)." Jama 285(19): 2486-97.

NCEP (2002). "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(25): 3143-421.

NCEP. (2001). "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 lll)." Jama 285(19): 2486-97.

NCHS. (1998). Third National Health and Nutrition Examination Survey (NHANES
III, 1988- 1994) laboratory data file (CDROM). Hyattsville, MD, Centers for
Disease

Control and Prevention;.

NCHS. (2007). National Health and Nutrition Examination Survey.

154

Ni Mhurchu, C., A. Rodgers, et al. (2004). "Body mass index and cardiovascular
disease in the Asia-Pacific Region: an overview of 33 cohorts involving 310
000 participants." nt J Epidemiol 33(4): 751-8.

NIH (1998). Clinical (Melines on the Identification, Evalpation, and Treatment of
Ovenrveight and Obesitv in Adults. Bethesda, Maryland, Department of
Health and Human Services; National Institutes of Health; National Heart,
Lung, and Blood Institute,.

NIH (2000). The Practical Gme: Identification, Evaluation and Treatment of
Oveniveight and Obesity in Adults. Bethesda, MD, NIH.

Ninomiya, J. K., G. L'ltalien, et al. (2004). "Association of the metabolic syndrome
with history of myocardial infarction and stroke in the Third National Health
and Nutrition Examination Survey." Circulation 109(1): 42-6.

Nishimura, R., T. Nakagami, et al. (2007). "Prevalence of metabolic syndrome
and optimal waist circumference cut-off values in Japan." Diabetes Res
mumd.

Nosadini, R., M. Sambataro, et al. (1993). "Role of hyperglycemia and insulin
resistance in determining sodium retention in non-insulin-dependent
diabetes." Kidney Int 44(1): 139-46.

Ogden, C. L., M. D. Carroll, et al. (2006). "Prevalence of overweight and obesity in
the United States, 1999-2004." Jama 295(13): 1549-55.

Okamoto, Y., S. Kihara, et al. (2006). "Adiponectin: a key adipocytokine in
metabolic syndrome." Clin Sci (Lond) 110(3): 267-78.

Okosun, I. S., Y. Liao, et al. (2000). "Abdominal adiposity and clustering of
multiple metabolic syndrome in White, Black and Hispanic americans." A_n_r1
Epidemiol 10(5): 263-70.

Olefsky, J. M., J. W. Farquhar, et al. (1974). "Reappraisal of the role of insulin in
hypertriglyceridemia." Am J Med 57(4): 551-60.

Ota, T., T. Takamura, et al. (2002). "Preobesity in World Health Organization
classification involves the metabolic syndrome in Japanese." Diabetes
Care 25(7): 1252-3.

Ouchi, N., S. Kihara, et al. (1999). "Novel modulator for endothelial adhesion
molecules: adipocyte-derived plasma protein adiponectin." Circulation
100(25): 2473-6.

Ouchi, N., S. Kihara, et al. (2001). "Adipocyte-derived plasma protein,
adiponectin, suppresses lipid accumulation and class A scavenger

155

receptor expression in human monocyte-derived macrophages."
Circulation 103(8): 1057-63.

Pan, W. H. (2002). "Metabolic Syndrome - An Important but Complex Disease
Entity for Asians." Acta Cardiol Sin(18): 24-26.

Pan, W. H., K. M. Flegal, et al. (2004). "Body mass index and obesity-related
metabolic disorders in Taiwanese and US whites and blacks: implications
for definitions of overweight and obesity for Asians." Am J Clin Nutr 79(1):
31-9.

Pan, W. H., W. T. Yeh, et al. (2003). "Prevalence and awareness of diabetes and
mean fasting glucose by age, sex, and region: results from the Nutrition
and Health Survey in Taiwan, 1993-1996." Diabet Med 20(3): 182-5.

Pan, X. R., G. W. Li, et al. (1997). "Effects of diet and exercise in preventing
NIDDM in people with impaired glucose tolerance. The Da Qing IGT and
Diabetes Study." Diabetes Care 20(4): 537-44.

Pantanetti, P., G. G. Garrapa, et al. (2004). "Adipose tissue as an endocrine
organ? A review of recent data related to cardiovascular complications of
endocrine dysfunctions." Cling!) vaertens 26(4): 387-98.

Park, Y. W., S. Zhu, et al. (2003). "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 Meg
163(4): 427-36.

Patel, A., K. C. Huang, et al. (2006). "Is a single definition of the metabolic
syndrome appropriate?--A comparative study of the USA and Asia."
Atherosclerosis 184(1): 225-32.

Pereira, M. A., D. R. Jacobs, Jr., et al. (2002). "Dairy consumption, obesity, and
the insulin resistance syndrome in young adults: the CARDIA Study." Jama
287(16): 2081-9.

Phillips, G. B., T. Jing, et al. (2008). "Does insulin resistance, visceral adiposity, or
a sex hormone alteration underlie the metabolic syndrome? Studies in
women." Metagflism 57(6): 838-44.

Pouliot, M. C., J. P. Despres, et al. (1994). 'Waist circumference and abdominal
sagittal diameter: best simple anthropometric indexes of abdominal
visceral adipose tissue accumulation and related cardiovascular risk in
men and women." Am J Cardiol 73(7): 460-8.

Prentice, A. M. and S. A. Jebb (2001). "Beyond body mass index." Obes Rev 2(3):
141-7.

156

Quetelet, L. A. J. (1973). A treatise on man and the development of his faculties
Edinburgh, United Kingdom: William and Robert Chambers, 1842.
Comparative statistics in the 19th century. . Farnborough, United Kingdom,
Gregg International Publishers.

Rader, D. J. (2007). Physiology and Pathophysiology of HDL Metabolism. Lipids
Online Slide Liprarv www.lipidsonline.org.

Rahmouni, K., D. A. Morgan, et al. (2004). "Hypothalamic PI3K and MAPK
differentially mediate regional sympathetic activation to insulin." J Clin
Invest 114(5): 652-8.

Rajala, M. W. and P. E. Scherer (2003). "Minireview: The adipocyte--at the
crossroads of energy homeostasis, inflammation, and atherosclerosis."
Endocrinology 144(9): 3765-73.

Rasmussen, B. M., B. Vessby, et al. (2006). "Effects of dietary saturated,
monounsaturated, and n-3 fatty acids on blood pressure in healthy
subjects." Am J Clin Ngt; 83(2): 221-6.

Ratner, R., R. Goldberg, et al. (2005). "Impact of intensive lifestyle and metformin
therapy on cardiovascular disease risk factors in the diabetes prevention
program." Diabetes Care 28(4): 888-94.

Reaven, G. and A. Laws (1999). Insulin Resistance. Insulin Resistance. Totowa,
New Jersey, Human Press inc.: 269, 97.

Reaven, G. M. (1988). "Banting Lecture 1988: Role of insulin resistance in human
disease." Diabetes 37: 1595-1607.

Reaven, G. M. (1993). "Role of insulin resistance in human disease (syndrome X):
an expanded definition." Annu Rev Med 44: 121-31.

Reaven, G. M. (2005). "The insulin resistance syndrome: definition and dietary
approaches to treatment." Annur Rev Nutr 25: 391-406.

Reaven, G. M. (2005). "Insulin resistance, the insulin resistance syndrome, and
cardiovascular disease." Panminerva ng 47(4): 201-10.

Reaven, G. M. (2005). "The metabolic syndrome: requiescat in pace." Clin Chem
51(6): 931-8.

Reaven, G. M., E. M. Scott, et al. (2005). "Hemostatic abnormalities associated
with obesity and the metabolic syndrome." J Thromb Haemost 3(5):
1074-85.

157

Reeder, B. A., A. Senthilselvan, et al. (1997). "The association of cardiovascular
disease risk factors with abdominal obesity in Canada. Canadian Heart
Health Surveys Research Group." Cmai 157 Suppl 1: 839-45.

Regitz-Zagrosek, V., E. Lehmkuhl, et al. (2006). "Gender differences in the
metabolic syndrome and their role for cardiovascular disease." Clin Res
Cardiol 95(3): 136-47.

Reilly, M. P. and D. J. Rader (2003). "The metabolic syndrome: more than the
sum of its parts?" Circulation 108(13): 1546-51.

Reisin, E. and M. A. Alpert (2005). "Definition of the metabolic syndrome: current
proposals and controversies." Am J Med Sci 330(6): 269-72.

Riccardi, G., P. Aggett, et al. (2004). "PASSCLAlM--body weight regulation,
insulin sensitivity and diabetes risk." Eur J Nutr 43 Suppl 2: II7-ll46.

Riccardi, G., R. Giacco, et al. (2004). "Dietary fat, insulin sensitivity and the
metabolic syndrome." Clin Nutr 23(4): 447-56.

Riccardi, G. and A. A. Rivellese (2000). "Dietary treatment of the metabolic
syndrome--the optimal diet." Br J Nutr 83 Suppl 1: 8143-8.

Robertson, R. M. and L. Smaha (2001). "Can a Mediterranean-style diet reduce
heart disease?" Circulation 103(13): 1821-2.

Ross, R., J. Aru, et al. (2002). "Abdominal adiposity and insulin resistance in
obese men." Am J thsiol Endocrinol Metab 282(3): E657-63.

Ross, R., J. Freeman, et al. (2002). "Abdominal obesity, muscle composition, and
insulin resistance in premenopausal women." J Clin Endocrinol Metap
87(11): 5044-51.

Ross, R., l. Janssen, et al. (2004). "Exercise-induced reduction in obesity and
insulin resistance in women: a randomized controlled trial." Obes Res
12(5): 789-98.

Ross, R. and P. T. Katzmarzyk (2003). "Cardiorespiratory fitness is associated
with diminished total and abdominal obesity independent of body mass
index." Int J Obes Relat Metab Disord 27(2): 204-10.

Ross, R., L. Leger, et al. (1991). "Adipose tissue distribution changes during rapid
weight loss in obese adults." Int J Obes 15(11): 733-9.

Roubenoff, R., G. E. Dallal, et al. (1995). "Predicting body fatness: the body mass
index vs estimation by bioelectrical impedance." Am J Public Health 85(5):
726-8.

158

Roubenoff, R. and J. J. Kehayias (1991). "The meaning and measurement of lean
body mass." Nutr Rev 49(6): 163-75.

Ruderman, N., D. Chisholm, et al. (1998). "The metabolically obese,
normal-weight individual revisited." Diabetes 47(5): 699-713.

Ryo, M., T. Nakamura, et al. (2004). "Adiponectin as a biomarker of the metabolic
syndrome." Circ J 68(11): 975-81.

Sakai, Y., H. Ito, et al. (2005). "Favourable association of leg fat with
cardiovascular risk factors." J Intern Med 257(2): 194-200.

Salmeron, J., J. E. Manson, et al. (1997). "Dietary fiber, glycemic load, and risk of
non-insulin-dependent diabetes mellitus in women." Jama 277(6): 472-7.

Samaras, K. and L. V. Campbell (2000). "Increasing incidence of type 2 diabetes
in the third millennium: is abdominal fat the central issue?" Diabetes Gag
23(4): 441-2.

Schulze, M. B., E. B. Rimm, et al. (2004). "Relationship between adiponectin and
glycemic control, blood lipids, and inflammatory markers in men with type 2
diabetes." Diabetes Care 27(7): 1680-7.

Seidell, J. C., H. S. Kahn, et al. (2001). "Report from a Centers for Disease
Control and Prevention Workshop on use of adult anthropometry for public
health and primary health care." Am J Clin Nu_t; 73(1): 123-6.

Seidell, J. C., L. Perusse, et al. (2001). 'Waist and hip circumferences have
independent and opposite effects on cardiovascular disease risk factors:
the Quebec Family Study." Am J Clin Nutr 74(3): 315-21.

Seidell, J. C. and T. L. Visscher (2000). "Body weight and weight change and their
health implications for the elderly." Eur J Clin Nutr 54 Suppl 3: 833-9.

Sheu, W. H., C. Y. Jeng, et al. (1992). "Insulin resistance and abnormal
electrocardiograms in patients with high blood pressure." Am J vaertens
5(7): 444-8.

Shimabukuro, M., N. Higa, et al. (2003). "Hypoadiponectinemia is closely linked to
endothelial dysfunction in man." J Clin Endocrinol Metab 88(7): 3236-40.

Shiwaku, K., A. Nogi, et al. (2005). "Prevalence of the metabolic syndrome using
the modified ATP III definitions for workers in Japan, Korea and Mongolia."
J Occup Health 47(2): 126-35.

Siscovick, D. S., L. Fried, et al. (1997). "Exercise intensity and subclinical
cardiovascular disease in the elderly. The Cardiovascular Health Study."
Am J Epidemiol 145(11): 977-86.

159

Snijder, M. B., J. M. Dekker, et al. (2004). "Trunk fat and leg fat have independent
and opposite associations with fasting and postload glucose levels: the
Hoorn study." Diabetes Care 27(2): 372-7.

Snijder, M. B., P. Z. Zimmet, et al. (2004). "Independent and opposite
associations of waist and hip circumferences with diabetes, hypertension
and dyslipidemia: the AusDiab Study." Int J Obes Relat Metaflisord
28(3): 402-9.

St-Onge, M. P., I. Janssen, et al. (2004). "Metabolic syndrome in normal-weight
Americans: new definition of the metabolically obese, normal-weight
individual." Diabetes Cfl 27(9): 2222-8.

Swenson, T. L. (1991). "The role of the cholesteryl ester transfer protein in
lipoprotein metabolism." Diabetes Metab Rev 7(3): 139-53.

Tan, C. E., S. Ma, et al. (2004). "Can we apply the National Cholesterol Education
Program Adult Treatment Panel definition of the metabolic syndrome to
Asians?" Diabetes Care 27(5): 1182-6.

Tanaka, S., K. Togashi, et al. (2002). "Is adiposity at normal body weight relevant
for cardiovascular disease risk?" Int J @es Relat Metab Disord 26(2):
176-83.

Tanko, L. B., Y. Z. Bagger, et al. (2003). "Peripheral adiposity exhibits an
independent dominant antiatherogenic effect in elderly women." Circulation
107(12): 1626-31.

Tillin, T., N. Forouhi, et al. (2005). "Metabolic syndrome and coronary heart
disease in South Asians, African-Caribbeans and white Europeans: a UK
population-based cross-sectional study." Diabetologia 48(4): 649-56.

Tousignant, B., M. Faraj, et al. (2008). "Body fat distribution modulates insulin
sensitivity in post-menopausal overweight and obese women: a MONET
study." Int J Obes (Lond) 32(11): 1626-32.

Trujillo, M. E. and P. E. Scherer (2005). "Adiponectin--journey from an adipocyte
secretory protein to biomarker of the metabolic syndrome." J Intern Med
257(2): 167-75.

Tseng, C. H. (2006). "Exogenous insulin use and hypertension in adult patients
with type 2 diabetes mellitus." Arch Intern Med 166(11): 1184-9.

Tuomilehto, J., J. Lindstrom, et al. (2001). "Prevention of type 2 diabetes mellitus
by changes in lifestyle among subjects with impaired glucose tolerance." N
Engl J Med 344(18): 1343-50.

160

Van Hannelen, V., S. Reynisdottir, et al. (1998). "Leptin secretion from
subcutaneous and visceral adipose tissue in women." Diabetes 47(6):
913-7.

Vincent, M. A., E. J. Barrett, et al. (2003). "Inhibiting NOS blocks microvascular
recruitment and blunts muscle glucose uptake in response to insulin." Am J
Physiol Endocrinol Metab 285(1): E123-9.

Wagenknecht, L. E., C. D. Langefeld, et al. (2003). "Insulin sensitivity, insulin
secretion, and abdominal fat: the Insulin Resistance Atherosclerosis Study
(IRAS) Family Study." Diabetes 52(10): 2490-6.

Wajchenberg, B. L. (2000). "Subcutaneous and visceral adipose tissue: their
relation to the metabolic syndrome." Endocr Rev 21(6): 697-738.

Walton, C., B. Lees, et al. (1995). "Body fat distribution, rather than overall
adiposity, influences serum lipids and lipoproteins in healthy men
independently of age." Am J Med 99(5): 459-64.

Wang, J. (2003). 'Waist circumference: a simple, inexpensive, and reliable tool
that should be included as part of physical examinations in the doctor's
office." Am J Clin Ntfl 78(5): 902-3.

Wang, J., J. C. Thornton, et al. (2003). "Comparisons of waist circumferences
measured at 4 sites." Am J Clin Nutr 77(2): 379-84.

Wang, J., J. C. Thornton, et al. (2000). "Anthropometry in body composition. An
overview." Ann N Y Acacfici 904: 317-26.

Weinsier, R. L., G. R. Hunter, et al. (2001). "Body fat distribution in white and
black women: different patterns of intraabdominal and subcutaneous
abdominal adipose tissue utilization with weight loss." Am J Clin Nu_tr 74(5):
631-6.

WHO (1995). Physical Status: The Use and Interpretation of Anthropbmetrv.
Geneva, WHO.

WHO (1997). Obesity, preventing and managing the global epidemic. Geneva,
WHO.

WHO (1998). Obesity, preventing and managing the global epidemic. Geneva,
WHO.

WHO (1999). Definition, diagnosis and classification of diabetes mellitus and its
complications: report of a WHO Consultation. Part 1: diagnosis and
classification of diabetes mellitus, World Health Organization.

161

 

Wildman, R. P., D. Gu, et al. (2004). "Appropriate body mass index and waist
circumference cutoffs for categorization of overweight and central adiposity
among Chinese adults." Am J Clin Nutr 80(5): 1129-36.

Williams, C. M. (2004). "Lipid metabolism in women." Proc thr Soc 63(1):
153-60.

Williams, M. J., G. R. Hunter, et al. (1997). "Regional fat distribution in women and
risk of cardiovascular disease." Am J Clin Nutr 65(3): 855-60.

Wisse, B. E. (2004). "The inflammatory syndrome: the role of adipose tissue
cytokines in metabolic disorders linked to obesity." J Am Soc NeLhrol
15(1 1): 2792-800.

Wolever, T. M. (2003). "Carbohydrate and the regulation of blood glucose and
metabolism." Nutr Rev 61(5 Pt 2): 840-8.

Wolfe, R. R. and E. J. Peters (1987). "Lipolytic response to glucose infusion in
human subjects." Am J thsiol 252(2 Pt 1): E218-23.

Wong, 8., I. Janssen, et al. (2003). "Abdominal adipose tissue distribution and
metabolic risk." Sports Med 33(10): 709-26.

Yagalla, M. V., S. L. Hoerr, et al. (1996). "Relationship of diet, abdominal obesity,
and physical activity to plasma lipoprotein levels in Asian Indian physicians
residing in the United States." J Am Diet Assoc 96(3): 257-61.

Yeh, W. T., H. Y. Chang, et al. (2005). "Do centrally obese Chinese with normal
BMI have increased risk of metabolic disorders?" Int J Obes Relat Metab_
Disord.

Yu-Poth, S., G. Zhao, et al. (1999). "Effects of the National Cholesterol Education
Program's Step I and Step II dietary intervention programs on
cardiovascular disease risk factors: a meta-analysis." Am J Clin th_r 69(4):
632-46.

Zavaroni, I., S. Mazza, et al. (1992). "Prevalence of hyperinsulinaemia in patients
with high blood pressure." J Intern Med 231(3): 235-40.

Zhu, 8., Z. Wang, et al. (2003). "Percentage body fat ranges associated with
metabolic syndrome risk: results based on the third National Health and
Nutrition Examination Survey (1988-1994)." Am J Clin ng 78(2): 228-35.

Zierath, J. R., J. N. Livingston, et al. (1998). "Regional difference in insulin
inhibition of non-esterified fatty acid release from human adipocytes:
relation to insulin receptor phosphorylation and intracellular signalling
through the insulin receptor substrate-1 pathway." Diabetologia 41(11):
1343-54.

162

Zimmet, P. 2., K. G. Alberti, et al. (2005). "Mainstreaming the metabolic
syndrome: a definitive definition. This new definition should assist both
researchers and clinicians." Med J Aust 183(4): 175-6.

163

is

. :42

554.1. .

...
ima-sfimwfi ..
a...

. I).
....

.-.
.umwmtt

O It
fir.

191

. r: 3
.. sue...
.meut...

 

.tt .f
.. 1.er
.4953
$.19“.
1&4?“