Bone mineral density is associated with carotid atherosclerosis in healthy postmenopausal women: a longitudinal analysis of randomized clinical trial data

PDF

Download

Open document

Flip pages

Copy asset link

Request this asset

Transcript (if available)

Content

Bone
Mineral
Density
is
Associated
with
Carotid
Atherosclerosis
in
Healthy

Postmenopausal
Women ：A
Longitudinal
Analysis
of
Randomized
Clinical

Trial
Data

By

Jiaonan
Li

Thesis

Submitted
in
Partial
Fulfillment

of
the
Requirements
for
the
Degree
of

Master
of
Science
in
Biostatistics

University
of
Southern
California

December
2015

I

ABSTRACT

Background
and
purpose:
Atherosclerosis
and
osteoporosis
are
major
causes
of

morbidity
and
mortality
in
postmenopausal
women.
Our
study
aimed
to
analyze
the

association
between
progression
of
carotid
artery
intima-‐media
thickness
(CIMT)

and
bone
mineral
density
(BMD),
providing
more
evidence
for
the
relation
between

osteoporosis
and
subclinical
atherosclerosis
in
postmenopausal
women.

Methods:
In
this
longitudinal
analysis,
349
postmenopausal
women
were
followed

for
3
years,
and
had
an
average
6
CIMT
measurements.
BMD
had
an
average
3

measurements
using
DEXA;
we
used
the
baseline
BMD
measurements
from
the
total

hip
and
lumbar
spine.
We
used
a
linear
mixed-‐effects
model
to
analyze
the

association
between
CIMT
progression
and
baseline
BMD
and
other
clinical

variables.

Results:
Longitudinally,
both
total
hip
and
lumbar
spine
BMD
were
significantly

inversely
associated
with
CIMT
progression
(P=0.01,
P=0.03).
The
mean
CIMT
rate

was
0.002
mm/year
(95%
CI:
-‐0.020,
-‐0.003)
lower
per
0.137g/cm
2

total
hip
BMD.

The
mean
CIMT
rate
was
0.001
mm/year
(95%
CI:
-‐0.016,
-‐0.001)
lower
per

0.122g/cm
2

mean
lumbar
spine
BMD.
After
adjusting
for
age,
the
longitudinal

association
between
total
hip
BMD
and
CIMT
did
not
change.

Conclusions:
Low
BMD
in
the
total
hip
and
lumbar
spine
may
represent
a
marker
of

both
osteoporosis
and
atherosclerosis
progression
for
healthy
postmenopausal

II

women.
Postmenopausal
women
at
higher
risk
for
osteoporosis
should
also
have

concern
about
the
risk
of
atherosclerosis.
These
results
have
clinical
and
public

health
implications
with
regard
to
co-‐occurrence
of
osteoporosis
and

atherosclerosis
for
postmenopausal
women
in
the
United
States.

III

DEDICATION

I
dedicate
my
thesis
to
my
big
family.
A
special
feeling
of
gratitude
to
my
loving

parents,
Weiguo
Li
and
Yan
Wen,
who
support
me
to
finish
my
graduate
school
and

encourage
me
to
pursue
the
things
I
love.

I
also
dedicate
my
thesis
to
my
friends
who
have
supported
me
throughout
the

process,
especially
Wendy
Cai
for
being
very
patient
to
answer
my
questions.

IV

ACKNOWLEDGEMENT

I
wish
to
thank
all
my
committee
members
who
were
generous
with
their
expertise

and
precious
time.

I
wish
to
thank
my
committee
chair
Wendy
Mack,
who
helps
me

revise
my
thesis
many
times
with
great
patient.
Without
her
help,
I
could
not
finish

my
thesis
timely.

I
wish
to
thank
my
committee
member
Dr.
Roksana
Karim,
who
helping
me
through

this
program.
Without
her
guidance
and
design,
the
results
of
this
research
could
not

have
been
found.

I
wish
to
thank
Dr.
Howard
N
Hodis,
who
provide
the
clinical
trial
dataset.

Finally,
I
wish
to
thank
all
the
staffs
in
preventive
medicine
department
for
their

excitement
to
provide
feedback
made
the
completion
of
thesis
an
enjoyable

experience.

V

TABLE
OF
CONTENTS

ABSTRACT
......................................................................................................................................................
I

DEDICATION
...............................................................................................................................................
III

ACKNOWLEDGEMENT
...........................................................................................................................
IV

TABLE
OF
CONTENT
................................................................................................................................
V

LIST
OF
TABLES
......................................................................................................................................
VII

INTRODUCTION
..........................................................................................................................................
1

METHODS
......................................................................................................................................................
4

Study
Population
........................................................................................................................
4

Assessment
of
Bone
Mineral
Density
................................................................................
4

Assessment
of
Atherosclerosis
Progression
..................................................................
5

Assessment
of
Clinical
Variables
.........................................................................................
6

Statistical
Analysis
....................................................................................................................
6

RESULTS
......................................................................................................................................................
10

Clinical
Variables,
CIMT
and
BMD
in
Postmenopausal
Women
.........................
10

Cross-‐sectional
Association
between
BMD
and
Clinical
Variables
...................
12

Univariate
Association
between
CIMT
and
Clinical
Variables
and
BMD
........
15

Multivariable
Association
between
Total
Hip
BMD
and
CIMT
............................
17

Effect
Modification
by
BMI
and
Smoking
.....................................................................
19

DISCUSSION
...............................................................................................................................................
20

Associations
of
BMD
with
Clinical
Variables
..............................................................
20

VI

Association
of
CIMT
with
Clinival
Variables
...............................................................
21

Association
of
CIMT
Rate
and
Total
Hip
BMD
............................................................
21

Biological
Link
and
Possible
Pathophysiological
Mechanisms
...........................
22

REFERENCES
.............................................................................................................................................
25

VII

LIST
OF
TABLES

Table
1.
Baseline
Characteristics
of
Study
Population
.............................................................
11

Table
2.
Baseline
Cross-‐sectional
Association

between
BMD
and
Clinical
Variables
............................................................................

14

Table
3.
Univeriate
Mixed
Effects
Associations
between

CIMT
and
Clinical
Variables
and
BMD
...........................................................................
16

Table
4.
Mixed
Effects
Model
of
CIMT
and

Total
Hip
BMD
with
Age
Adjustment
.............................................................................
18

1

INTRODUCTION

Atherosclerosis
is
the
primary
cause
of
heart
disease
and
stroke.
In
Westernized

societies,
it
is
the
underlying
cause
of
about
half
of
all
deaths
for
postmenopausal

women
[1].
Atherosclerosis
is
an
inevitable
degenerative
consequence
of
aging
[2].

Epidemiological
studies
have
revealed
several
additional
important
environmental

and
genetic
risk
factors
associated
with
atherosclerosis,
such
as
obesity,
smoking

and
metabolic
syndrome
[1].

Osteoporosis
is
a
metabolic
process
that
weakens
bone,
increasing
the
risk
for

fractures
[3].
In
the
United
States,
30%
of
postmenopausal
women
have

osteoporosis
[4].
Women
are
at
greater
risk
for
osteoporosis
than
men,
and
there
is

a
direct
relationship
between
the
reduction
of
endogenous
estrogen
levels
at

menopause
and
the
development
of
osteoporosis
[3].
Epidemiological
studies
have

revealed
important
risk
factors
for
osteoporosis,
including
older
age,
ethnicity
(in

particular
non-‐Hispanic
Whites
and
Asians
are
at
higher
risk),
lower
body
weight,

current
smoking
and
hypertension.

More
than
40
million
American
women
are
currently
postmenopausal
and
as

the
United
States
population
ages,
the
number
of
women
entering
menopause

steadily
increases
by
>1
million
American
women
per
year
[5].

Atherosclerosis
and

osteoporosis
are
major
causes
of
morbidity
and
mortality
in
postmenopausal

women.
Both
osteoporosis
and
atherosclerosis,
including
vascular
calcification
have

largely
been
attributed
to
the
aging
process
[6].
In
addition
to
menopause
and

2

advanced
age,
other
risk
factors
for
vascular
disease
such
as
dyslipidemia,
oxidative

stress,
inflammation,
hypertension
and
diabetes
have
also
been
related
to
lower

bone
mineral
density
[7].
Therefore,
the
interrelation
between
the
two
age-‐related

processes
of
atherosclerosis
and
osteoporosis
requires
inquiry.

Carotid
artery
intima-‐media
thickness
(CIMT),
measured
by
high-‐resolution
B-‐
mode
ultrasound,
is
a
commonly
accepted
measure
of
subclinical
atherosclerosis

and
is
a
biomarker
of
cardiovascular
disease
risk.
CIMT
is
a
suitable
measure
for

community-‐
and
population-‐based
studies
conducted
among
persons
without

clinically-‐defined
cardiovascular
disease.
Increased
CIMT
is
significantly
associated

with
atherosclerosis
[8].
Bone
mineral
density
(BMD),
which
is
measured
by
dual-‐
energy
X-‐ray
absorptiometry
(DEXA),
is
widely
used
for
diagnosing
osteoporosis
[9].

Several
cross-‐sectional
studies
have
evaluated
carotid
artery
atherosclerosis
and

demonstrated
a
positive
association
with
osteoporosis
among
elderly
women
[8.10].

One
study
reported
that
the
progression
of
vascular
calcification
(in
the
abdominal

aorta)
was
associated
with
increased
bone
loss
(metacarpal)
among
women
aged
45

to
57
years
[1].
However,
currently
there
is
no
longitudinal
research
with
sufficient

sample
size,
investigating
the
association
of
BMD
with
atherosclerosis
progression

in
postmenopausal
women.

In
the
present
community-‐based
longitudinal
study,
we
examined
the

association
between
progression
of
CIMT,
a
measure
of
sub-‐clinical
atherosclerosis,

and
total
hip
and
lumbar
spine
BMD
among
349
postmenopausal
women.
We
also

3

considered
the
effect
of
several
potential
confounding
factors
on
this
association.

4

METHODS

Study
Population

The
Women’s
Isoflavone
Soy
Health
(WISH)
trial
was
conducted
from
April
12,

2004
to
March
19,
2009.
WISH
was
a
randomized,
double-‐blinded,
placebo-‐
controlled
trial.
Participants
were
postmenopausal
women
without
vaginal

bleeding
>1
year
and
serum
estradiol
<20
pg/mL
[5].
Detailed
inclusion
and

exclusion
criteria
and
other
aspects
of
this
trial
have
been
previously
published
[5].

The
primary
WISH
trial
outcome
was
carotid
artery
ultrasonographic

measurement
of
CIMT
to
assess
progression
of
sub-‐clinical
atherosclerosis.
Clinic

visits
to
assess
CIMT
occurred
every
6
months
and
lasted
for
2.5
years.
BMD
was
a

secondary
trial
outcome.
Clinic
visits
to
assess
BMD
occurred
every
year
for
2.5

years.
Sample
size
based
on
CIMT
progression
required
350
participants
to
detect
a

difference
in
the
rate
of
CIMT
progression
of
12.4μm/year
at
a
0.05
significance
(2-‐
sided)
with
90%
power
[5].
The
350
women
were
recruited
from
the
general

population
of
the
Greater
Los
Angeles
area
predominantly
through
media

advertisement.
A
total
of
269
(77%)
postmenopausal
women
had
at
least
6
CIMT

measures;
25(7%)
participants
had
baseline
CIMT
only.

Assessment
of
Bone
Mineral
Density

Total
hip
and
lumbar
spine
BMD
were
measured
by
DXA
[5].
DXA
is
a
fast
and

low
dose
radiation
method
that
uses
two
different
X-‐ray
beams
to
estimate
bone

density;
higher
bone
density
allows
less
of
the
X-‐ray
beam
to
pass
through
the
bone.

5

The
amount
of
each
X-‐ray
beam
that
is
blocked
by
bone
and
soft
tissue
are
compared

with
each
other.
DXA
can
measure
as
little
as
2%
of
bone
loss
per
year
[3].
The
unit

of
BMD
is
expressed
as
g/cm
2
.

348
participants
had
baseline
total
hip
BMD
measurements
and
346

participants
had
baseline
lumbar
spine
BMD
measurements.
The
number
of
BMD

measurements
for
each
subject
ranged
from
1
to
3
and
the
median
number
of
BMD

measurements
was
3(80%).
The
scheduled
times
for
BMD
measurement
is
baseline,

12
months
and
30
months.
In
this
study,
we
only
used
the
baseline
BMD

measurement.

Assessment
of
Atherosclerosis
Progression

Carotid
artery
ultrasound
is
noninvasive
and
is
therefore
suitable
for
studies
in

subjects
who
are
free
from
symptomatic
cardiovascular
disease
[11].
In
WISH,

ultrasound
imaging
of
far
wall
CIMT
was
conducted
using
standardized
procedures

and
technology
specifically
developed
for
longitudinal
measurements
(24-‐27).
The

jugular
vein
and
carotid
artery
were
imaged
longitudinally
with
the
former
stacked

above
the
latter.
All
images
contained
internal
anatomical
landmarks
for

reproducing
probe
angulation.
The
baseline
image
for
each
individual
was
used
as

an
online
guide
for
follow-‐up
examinations
on
a
split-‐screen
system
designed
for

repeat
image
acquisition
for
longitudinal
studies
[5].
CIMT
was
the
average
of

approximately
70
to
100
individual
measurements
between
the
intima-‐lumen
and

media-‐adventitia
interfaces
along
a
1-‐cm
length
just
proximal
to
the
carotid
artery

6

bulb
at
the
same
point
of
the
cardiac
cycle
[12].
This
method
standardizes
the

location
and
the
distance
over
which
CIMT
is
measured,
ensuring
comparability

within
and
across
participants
(26,27).
The
coefficient
of
variation
of
the
350

repeated
baseline
CIMT
measurement
was
<1%
[5].
A
total
of
269
(77%)

postmenopausal
women
had
at
least
6
CIMT
measures;
25(7%)
participants
had

baseline
CIMT
only.

Assessment
of
Clinical
Variables

Height
and
weight
were
measured
at
the
clinic
screening.
Body
mass
index

(BMI)
was
calculated
as
weight
(kg)
/height
2

(m
2
)
formula.
Smoking
history
and

hormone
treatment
(HT)
history,
which
are
both
associated
with
BMD,
were

assessed
by
a
structured
questionnaire
at
baseline.

Statistical
Analysis

The
current
analysis
included
349
participants
who
had
at
least
one
CIMT

measurement,
and
total
hip
BMD
and
lumbar
spine
BMD
measurement
at
baseline.

We
only
used
baseline
BMD
measurement
data
in
the
analysis.
The
baseline
CIMT

measurement
contributed
to
estimation
and
testing
of
the
cross-‐sectional

association
of
baseline
CIMT
and
BMD.
The
follow-‐up
measurements
contributed
to

estimation
and
testing
of
the
association
between
BMD
and
CIMT
change
rate
(CIMT

progression).

Age
was
categorized
into
5
groups:
≤55
years
old,
55-‐59
years
old,
60-‐64
years

old,
65-‐69
years
old
and
70+
years
old.
Smoking
was
categorized
as
current,
former

7

and
never
smokers.
BMI
was
categorized
by
the
Center
for
Disease
Control
(CDC)

classification
scheme:
<18.5
kg/m
2
=underweight,
18.5-‐24.9
kg/m
2

=
ideal
weight,

25.0-‐29.9
kg/m
2

=
overweight,
≥30
kg/m
2

=obese.
Ethnicity
was
categorized
into

five
groups:
White
(non-‐Hispanic),
Black
(non-‐Hispanic),
Hispanic,
Asian
and
other.

Education
was
categorized
into
high
school
or
below,
college
and
bachelor’s
degree

or
above.

BMI,
age
and
race
are
commonly
considered
risk
factors
for
both
atherosclerosis

and
osteoporosis
of
postmenopausal
women
[3,4].
To
identify
possible
confounders

of
the
association
between
CIMT
and
BMD,
we
analyzed
the
univariate
associations

between
baseline
BMD
and
age,
race,
BMI,
past
use
of
hormone
therapy
and
smoking

status
by
use
of
linear
regression,
with
BMD
(total
hip
and
lumbar
spine)
as
the

dependent
variable.
We
centered
total
hip
BMD
on
the
baseline
mean
(0.937)
and

centered
lumbar
spine
BMD
on
the
baseline
mean
(0.898)
to
change
the
value
of
the

intercept.
After
setting
up
linear
regression
models,
we
examined
four
assumptions

for
linear
regression
by
checking
the
residual
of
the
model:
1)
The
linearity

assumption
is
satisfied
since
the
residuals
plotted
against
BMD
showed
a
flat
scatter

around
the
zero
line;
2)
The
independence
assumption
is
satisfied
since
subjects
in

Wish
were
not
correlated.
3)
Normality
assumption
is
satisfied
since
the
sample
size

is
large
and
both
mean
and
skewness
of
residuals
are
near
to
0
and
the
distribution

of
the
residuals
is
bell
shaped;
and
4)
Equal
variance
is
satisfied
since
the
residuals

have
common
variance
across
the
range
of
X-‐values.

8

The
associations
between
CIMT
rate
and
clinical
variables
(including
age,

ethnicity,
BMI,
education,
smoking
and
hormone
therapy)
were
analyzed
by

univariate
linear
mixed-‐effect
models,
with
random
intercept
and
time.
A
linear

mixed-‐effects
model
with
random
time
and
intercept
was
also
used
to
assess
the

correlation
between
CIMT
change
rates
and
baseline
BMD
(total
hip
and
lumber

spine
separately),
using
an
unstructured
correlation
matrix
to
model
the
repeatedly

measured
CIMT
dependent
variable.
CIMT
measured
over
the
trial
was
treated
as

the
dependent
variable.
In
the
mixed-‐effects
model,
the
primary
independent

variable
of
interest
was
baseline
BMD;
additional
variables
included
follow-‐up
time

(measured
as
years
since
randomization,
or
first
CIMT
measurement)
to
model
the

longitudinal
change
in
CIMT,
and
age
as
an
adjusting
variable.
Since
age
was

correlated
with
both
CIMT
and
BMD,
we
determined
whether
BMD
and
CIMT

progression
rate
were
associated
with
and
without
age
in
the
model.
Other
clinical

variables,
such
as
BMI,
smoking
status
and
hormone
therapy,
were
not
correlated

with
both
CIMT
and
BMD;
these
variables
were
therefore
not
considered

confounders.
The
regression
coefficient
for
BMD
evaluated
the
cross-‐sectional
effect

between
BMD
and
CIMT.
The
BMD
and
follow-‐up
time
interaction
coefficient

estimated
the
effect
of
BMD
on
the
annual
CIMT
progression
rate.
In
subsequent

models,
we
added
an
interaction
term
between
BMD
and
BMI
categories,
and

between
BMD
and
smoking
status
in
the
mixed-‐effect
model
to
test
whether
the

association
between
CIMT
and
BMD
differed
by
smoking
status
or
by
BMI

9

categories.
25
participants
had
baseline
CIMT
measurements
only.
These

participants
were
included
in
the
analyses,
contributing
information
on
cross-‐
sectional
but
not
longitudinal
associations
of
BMD
with
CIMT.

Statistical
analysis
used
SAS
9.3
software
(SAS,
Inc,
Cary,
NC);
statistical

significance
testing
was
conducted
at
a
2-‐tailed
0.05
significance
level.

10

RESULTS

Clinical
Variables,
CIMT
and
BMD
in
Postmenopausal
Women

Table
1
shows
the
characteristics
of
349
postmenopausal
women
randomized

into
the
WISH
trial
who
were
included
in
the
present
analysis.
The
mean
(SD)
age
of

participants
was
61(7)
years.
The
majority
(64%)
of
the
study
population
was
White

(non-‐Hispanic).
Of
the
349
participants,
60%
had
a
bachelor’s
degree
or
above

education
level,
59%
were
never
smokers
and
70%
had
hormone
therapy
before
the

trial.
The
mean
(SD)
BMI
was
26.6
(5.2)
kg/m
2
.
The
majority
(57%)
of
the
study

population
were
overweight
or
obese.
The
mean
baseline
CIMT
was
0.811
(0.100)

mm.
The
mean
baseline
total
hip
BMD
was
0.937
(0.137)
g/cm
2

and
the
mean

baseline
lumbar
spine
BMD
was
0.898
(0.122)
g/cm
2
.

11

Table1.
Baseline
Characteristics
of
Study
Population
(n=349)

Variables

Age,
years
61
(7)
1

Age
group

≤ 55
67
(19)
1

55-‐59
88
(25)
1

60-‐64
88
(25)
1

65-‐69
67
(19)
1

70+
39
(12)
1

Ethnicity

White
(non-‐Hispanic)
223
(64)
2

Black
(non-‐Hispanic)
20
(6)
2

Hispanic
55
(15)
2

Asian
38
(11)
2

Other
13
(4)
2

Education

High
school
or
below
19
(5)
2

College
122
(35)
2

Bachelor’s
degree
or
above
208
(60)
2

Smoking
history

Current

8
(2)
2

Former
135
(39)
2

Never
smoked
206
(59)
2

Body
mass
index,
kg/m
2

26.6
(5.2)
1

BMI
group

<
18.5
(underweight)
5
(2)
1

18.5-‐-‐24.9
(ideal
weight)
143
(41)
1

25.0-‐-‐29.9
(overweight)
116
(33)
1

30.0
and
above
(obese)
85
(24)
1

Past
use
of
hormone
therapy

Yes
245
(70)
2

No
104
(30)
2

CIMT,
mm
0.811(0.100)
1

Total
hip
BMD,
g/cm
2

0.937
(0.137)
1,3

Lumbar
Spine
BMD,
g/cm
2

0.898
(0.122)
1,3

1. Mean
(SD)
for
continuous
variables

2. Number
(%)
for
categorical
variables.

3. Total
hip
BMD
measurement
(n=348);
Lumber
spine
BMD
measurement
(n=346).

12

Cross-‐sectional
Association
Between
BMD
and
Clinical
Variables

Table
2
shows
the
cross-‐sectional
associations
between
baseline
BMD
(total
hip

and
lumbar
spine)
with
clinical
variables.
There
was
a
significantly
negative
linear

trend
between
mean
total
hip
BMD
and
age
(P<0.0001).
Mean
total
hip
BMD
in

women
≥
70
years
old
was
lower
than
women
<
age
55
(P<0.0001).
Mean
total
hip

BMD
significantly
differed
among
racial
groups
(P<0.0001).
Black
(non-‐Hispanic)

participants
had
higher
mean
total
hip
BMD
than
the
White
participants
(P<0.0001).

Hispanic
participants
had
a
marginally
significantly
higher
mean
total
hip
BMD
than

White
participants
(P=0.054).
Mean
total
hip
BMD
in
Asian
participants
was

significantly
lower
than
White
participants
(P<0.0001).
Mean
total
hip
BMD
was

significantly
positively
associated
with
BMI
(P<0.0001).
Mean
total
hip
BMD

significantly
differed
among
the
three
education
groups
(P<0.0001).
Mean
total
hip

BMD
in
the
bachelor’s
degree
or
above
education
group
was
significantly
lower
than

in
women
who
had
high
school
or
below
education
(P=0.02).
Mean
total
hip
BMD

significantly
differed
among
the
three
smoking
status
groups
(P<0.05).
Mean
total

hip
BMD
in
current
smokers
was
higher
than
in
never
smoker
(P=0.01).

In
contrast
to
total
hip
BMD,
age
was
not
significantly
associated
with
lumbar

spine
BMD
(P>0.05).
Mean
lumbar
spine
BMD
differed
among
racial
groups

(P<0.0001).
Compared
with
White
participants,
Black
(non-‐Hispanic)
participants

had
a
significantly
higher
mean
lumbar
spine
BMD
(P<0.0001);
Hispanic

participants
had
a
significantly
lower
mean
lumber
spine
BMD
(P<0.01)
and
Asian

13

participants
had
a
significantly
lower
mean
lumbar
spine
BMD
(P<0.0001)
than

White
participants.
Lumbar
spine
BMD
was
significantly
positively
associated
with

BMI
(P<0.0001).
Mean
lumbar
spine
BMD
significantly
differed
among
the
three

smoking
groups
(P<0.0001).
Mean
lumbar
spine
BMD
in
former
smokers
was

significantly
higher
than
never
smokers
(P<0.0001).
Postmenopausal
women
who

had
used
hormone
therapy
had
a
higher
mean
lumbar
spine
BMD
than
women
who

did
not
use
hormone
therapy
(P<0.01).

14

Table
2.
Baseline
Cross-‐sectional
Association
between
BMD
1

and
Clinical
Variables

Total
Hip
BMD
(n=348)
2

Lumbar
Spine
BMD
(n=346)
3

Variable
Coefficient
95%CI
P
Coefficient
95%CI
P

Age
group,
years

<0.0001

0.11

<55
Reference

Reference

55-‐59
-‐0.009
(-‐0.025,
0.007)

0.014
(-‐0.004,
0.032)

60-‐64
-‐0.001
(-‐0.016,
0.016)

0.009
(-‐0.010,
0.027)

65-‐69
-‐0.009
(-‐0.027,
0.008)

-‐0.001
(-‐0.020,
0.018)

70+
-‐0.057
(-‐0.077,
-‐0.036)

0.026
(0.002,
0.049)

Trend

<0.0001

0.48

Ethnicity

<0.0001

<0.0001

White
(non-‐Hispanic)

Reference

Reference

Black
(non-‐Hispanic)

0.101
(0.078,
0.124)

0.077
(0.052,
0.103)

Hispanic

0.014
(-‐0.001,
0.029)

-‐0.027
(-‐0.044,
-‐0.010)

Asian

-‐0.044
(-‐0.061,
-‐0.028)

-‐0.050
(-‐0.068,
-‐0.031)

Other

0.001
(-‐0.027,
0.028)

-‐0.030
(-‐0.061,
0.002)

BMI
1

0.011
(0.010,
0.012)
<0.0001
0.009
(0.008,
0.010)
<0.0001

BMI
group

<0.0001

<0.0001

Below
18.5
Reference

Reference

18.5-‐-‐24.9
0.090
(0.052,0.127)

0.087
(0.043,
0.132)

25.0-‐-‐29.9
0.137
(0.099,0.175)

0.111
(0.066,
0.156)

30.0
and
above
0.220
(0.182,0.259)

0.193
(0.148,
0.238)

Education

<0.01

0.72

≤
High
school

Reference

Reference

Some
college

-‐0.004
(-‐0.029,
0.020)

-‐0.013
(-‐0.041,
0.014)

≥
Bachelor’s
degree

-‐0.029
(-‐0.052,
-‐0.006)

-‐0.016
(-‐0.042,
0.010)

Smoking
history

0.01

<0.0001

Never
smoked

Reference

Reference

Former

0.053
(-‐0.003,
0.019)

0.025
(0.013,
0.037)

Current

0.008
(0.015,
0.090)

0.040
(-‐0.002,
0.082)

Past
hormone
therapy

-‐0.008
(-‐0.020,
0.003)
0.17
0.019
(0.006,
0.032)
<0.01

1. BMD,
bone
mineral
density;
BMI,
body
mass
index.

2. Total
hip
BMD
was
centered
on
the
mean
(0.937);
n=348.

3. Lumbar
spine
BMD
was
centered
on
the
mean
(0.898);
n=346.

15

Univariate
Association
between
CIMT
and
Clinical
Variables
and
BMD

Table
3
shows
the
CIMT
associations
with
clinical
variables
and
BMD
from
a

linear
mixed
effects
models.
BMD
of
the
total
hip
and
spine
were
divided
by
their

respective
SDs;
the
regression
coefficients
associated
with
BMD
are
therefore

expressed
per
SD.
Cross-‐sectionally,
the
mean
baseline
CIMT
significantly
increased

with
age
(P<0.0001).
The
mean
baseline
CIMT
marginally
significantly
differed

among
ethnic
groups
(P=0.06).
Asian
participants
had
a
significantly
lower
mean

baseline
CIMT
than
White
participants
(P=0.01).
Mean
lumbar
spine
BMD
was

positively
significantly
associated
with
CIMT
(P<0.01);
total
hip
BMD
was
not

associated
with
baseline
CIMT.

Longitudinally,
total
hip
BMD
was
significantly
inversely
associated
with
CIMT

progression
(P=0.01).
The
mean
CIMT
rate
was
0.002
mm/year
(95%
CI:
-‐0.020,
-‐
0.003)
lower
per
0.137g/cm
2

(1
SD)
total
hip
BMD.
Lumbar
spine
BMD
was
also

significantly
inversely
associated
with
CIMT
progression
(P=0.03).
The
mean
CIMT

rate
was
0.001
mm/year
(95%
CI:
-‐0.016,
-‐0.001)
lower
per
0.122g/cm
2

(1
SD)

mean
lumbar
spine
BMD.

16

Table
3.
Univariate
Mixed
Effects
Associations
between
CIMT
and
Clinical
Variables
and
BMD
1

Cross-‐sectional
effect
on
CIMT
baseline
Longitudinal
effect
on
CIMT
rate

Variable
Coefficient
95%CI
P
Coefficient
95%CI
P

Age
group,
years

<0.01

less
than
55
Reference

Reference

0.21

55-‐59
0.024
（-‐0.005,
0.053 ）

0.001
(-‐0.003,
0.003)

60-‐64
0.059
(0.030,
0.087)

0.001
(-‐0.002,
0.004)

65-‐69
0.076
(0.045,
0.107)

0.003
(-‐0.001,
0.006)

70+
0.148
(0.113,
0.184)

-‐0.002
(-‐0.006,
0.002)

Trend

<0.01
2

0.66
2

Ethnicity

0.06

0.70

White
(non-‐Hispanic)

Reference

Reference

Black
(non-‐Hispanic)

0.021
(-‐0.023,
0.066)

0.001
(-‐0.003,
0.006)

Hispanic

-‐0.006
(-‐0.035,
0.024)

-‐0.001
(-‐0.004,
0.002)

Asian

-‐0.044
(-‐0.078,
-‐0.009)

-‐0.001
(-‐0.005,
0.002)

Other

-‐0.036
(-‐0.092,
0.020)

-‐0.003
(-‐0.008,
0.003)

BMI

0.001
(-‐0.001,
0.003)
0.17
<0.001
(-‐0.001,
0.001)
0.77

BMI
categories

0.17

0.77

Below
18.5
Reference

18.5-‐-‐24.9
0.062
(-‐0.027,
0.151)

0.001
(-‐0.008,
0.009)

25.0-‐-‐29.9
0.081
(-‐0.009,
0.171)

-‐0.001
(-‐0.009,
0.008)

30.0
and
above
0.076
(-‐0.015,
0.166)

-‐0.001
(-‐0.009,
0.008)

Education

0.28

0.72

High
school
or
below

Reference

Reference

Some
college

0.010
(-‐0.039,
0.058)

-‐0.002
(-‐0.007,
0.003)

Bachelor’s
degree
or

above
-‐0.009
(-‐0.056,
0.038)

-‐0.002
(-‐0.007,
0.003)

Smoking
history

0.62

0.68

Never
smoked

Reference

Reference

Former

0.009
(-‐0.013,
0.031)

0.001
(-‐0.006,
0.008)

Current

-‐0.016
(-‐0.087,
0.055)

0.001
(-‐0.001,
0.003)

Past
use
of
hormone

therapy
0.014
(-‐0.009,
0.037)
0.24
0.001
(-‐0.001,
0.003)
0.36

Total
hip
BMD
3

0.070
(-‐0.017,0.157)
0.11
-‐0.012
(-‐0.020,
-‐0.003)
0.01

Lumbar
Spine
BMD
3

0.130
(0.054,
0.206)
<0.01
-‐0.008
(-‐0.016,
-‐0.001)
0.03

1. BMD,
bone
mineral
density;
BMI,
body
mass
index.

2. Trend
test
for
association
between
age
groups
and
CIMT.

3. The
range
of
total
hip
BMD
is
(0.720,
1.639),
mean
(SD)
is
0.937
(0.137).

The
range
of
lumbar
spine
BMD
is
(0.603,
1.399),
mean
(SD)
is
0.898
(0.122).

17

Multivariable
Associations
between
Total
Hip
BMD
and
CIMT

Table
4
shows
the
multivariable
mixed-‐effects
model
of
CIMT
and
total
hip
BMD,

adjusting
for
age.
Age
was
significantly
associated
with
baseline
CIMT
and
total
hip

BMD
(P<0.0001).
Cross-‐sectionally,
after
adjusting
for
age,
the
regression
parameter

of
adjusted
total
hip
BMD
changed
36%
from
the
unadjusted
total
hip
BMD

parameter
(age
treated
as
a
continuous
variable).
Age
therefore
confounded
the

cross-‐sectional
association
between
baseline
total
hip
BMD
and
CIMT.
Total
hip
BMD

was
significantly
positively
associated
with
CIMT,
adjusting
for
age
group
(P<0.01).

Mean
baseline
CIMT
increased
0.015
mm
(95%
CI:
0.031,
0.188)
per
0.137
g/cm
2

of

total
hip
BMD.

The
mean
total
hip
BMD
was
significantly
inversely
associated
with
CIMT

progression
(P<0.01).
The
mean
CIMT
rate
decreased
0.002
mm/year
(95%
CI:
-‐
0.021,
-‐0.003)
with
per
0.137g/cm
2

(1SD)
of
total
hip
BMD.

18

Table
4.
Mixed
Effects
Model
of
CIMT
and
Total
Hip
BMD
with
Age
Adjustment

Total
hip
BMD
model

Coefficient

95%CI
P

Intercept
0.749
(0.732,
0.765)
<0.01

Total
hip
BMD
1

0.109
(0.031,
0.188)
<0.01

Year
2,4

0.006
(0.004,
0.007)
<0.01

Total
hip
BMD
*
Year
1

-‐0.012
(-‐0.021,
-‐0.003)
<0.01

Age
3

0.034
(0.027,
0.042)
<0.01

Age
group
*Year
<0.001
(-‐0.001,
0.001)
0.97

1. Total
hip
BMD
coefficient
represent
cross-‐sectional
(baseline)
association
with
CIMT;
Total

hip
BMD*
Year
represents
longitudinal
association
with
CIMT
rate.

2. Unit
of
year
was
per
year.

3. Age
modeled
as
a
continuous
variable.

4. Year
variable
equal
year
since
randomization.

19

Effect
Modification
by
BMI
and
Smoking

Additional
interaction
term
of
BMD*BMI
and
BMD*smoking
evaluated
to
test

whether
the
cross-‐sectional
association
of
BMD
with
baseline
CIMT
differed
by
BMI

and
smoking.
Similarly,
interaction
terms
of
BMD*BMI*Year
and
BMD*Smoking*Year

tested
if
the
associations
of
BMD
with
CIMT
rate
differed
by
BMI
and
smoking.
None

of
these
interaction
terms
were
statistically
significant
(all
P>0.05).
Therefore,
the

association
between
BMD
(total
hip
BMD
and
lumbar
spine
BMD)
and
CIMT
did
not

significantly
differ
among
BMI
categories
(data
not
shown).

Similarly,
the
association
between
BMD
(total
hip
and
lumbar
spine
BMD)
and

CIMT
did
not
significantly
differ
among

smoking
categories
(data
not
shown).

20

DISCUSSION

In
the
randomized,
double-‐blinded,
placebo-‐controlled
WISH
trial,
our
analyses

showed
that
CIMT
progression
was
significantly
negatively
associated
with
BMD
of

the
total
hip
and
lumbar
spine
in
postmenopausal
women.
Epidemiology
and
clinical

trial
evidence
support
the
use
of
CIMT
as
a
validated
and
accepted
marker
for

generalized
atherosclerosis
burden
and
vascular
disease
risk
[14].
Thus,

postmenopausal
women
with
greater
BMD
loss
measured
at
the
hip
and
lumbar

spine
were
at
increased
risk
related
to
not
only
osteoporosis
but
also

atherosclerosis.

Estrogen-‐deficient
postmenopausal
women
are
at
greater
risk
for

atherosclerosis
and
osteoporosis.
These
are
not
merely
age-‐related
processes,
but

are
linked
by
many
common
factors.

Age,
ethnicity,
BMI,
education,
smoking
history

and
past
use
of
hormone
therapy
are
considered
as
common
risk
factors
for

osteoporosis
and
subclinical
atherosclerosis
in
postmenopausal
women
[5,13].

Osteoporosis
and
atherosclerosis
are
associated
with
significant
morbidity
and

mortality;
thus
analyzing
the
association
between
these
two
pathological
entities

has
important
public
health
impact.

Associations
of
BMD
with
Clinical
Variables

WISH
data
confirmed
many
of
the
known
BMD
associations.
BMD
decreased

with
age
and
significantly
differed
among
racial
groups,
with
Asian
participants

having
significantly
lower
BMD
than
White
participants.
Black
participants
had

21

significantly
higher
total
hip
and
lumbar
spine
BMD
than
White
participants

consistent
with
other
data
reporting
such
a
difference
[16].
While
smoking
is

considered
an
established
risk
factor
for
osteoporosis
[17],
BMD
was
significantly

higher
among
the
few
current
smokers
compared
to
non-‐smokers
in
the
WISH
trial.

This
finding
is
contrary
to
previous
studies
and
may
be
due
to
the
very
sample
size

of
current
smokers
(n=8).
A
positive
association
between
past
use
of
hormone

therapy
with
lumbar
spine
BMD
suggests
a
protective
effect
of
HT
on
lumbar
spine

osteoporosis,
which
is
consistent
with
previous
research
[13].

Association
of
CIMT
with
Clinical
Variables

After
analyzing
univariate
associations
between
CIMT
and
clinical
variables,
we

found
that
baseline
CIMT
was
significantly
associated
with
age,
which
is
consistent

with
other
study
[22].
Baseline
CIMT
significantly
differed
across
ethnic
groups;
the

average
baseline
CIMT
in
Asians
was
less
than
half
that
of
other
racial
groups.
While

we
did
not
find
a
difference
in
CIMT
progression
rates
across
racial
groups,
the

directions
of
race/ethnic
group
differences
in
CIMT
progression
rates
were
in
the

same
direction
as
seen
in
the
cross-‐sectional
analysis.

Association
of
CIMT
Rate
and
Total
Hip
BMD

Although
the
cross-‐sectional
association
between
CIMT
and
BMD
was
opposite

to
that
reported
in
previous
studies
[18],
the
longitudinally-‐analyzed
CIMT
rate
was

negatively
associated
with
BMD
(in
both
total
hip
and
lumbar
spine)
consistent
with

the
known
positive
association
between
osteoporosis
and
atherosclerosis.
The
mean

22

CIMT
rate
was
0.002
mm/year
(95%
CI:
-‐0.020,
-‐0.003)
lower
per
0.137g/cm
2

total

hip
BMD.
The
mean
CIMT
rate
was
0.001
mm/year
(95%
CI:
-‐0.016,
-‐0.001)
lower

per
0.122g/cm
2

mean
lumbar
spine
BMD.
This
result
is
consistent
with
and

importantly
extends
just
one
other
longitudinal
study
that
demonstrated
that

vascular
calcification
was
associated
with
increased
bone
loss
in
women
during

menopause
[1].

Our
model
ultimately
adjusted
for
age
as
age
confounded
the
cross-‐sectional

association
between
baseline
CIMT
and
total
hip
BMD.
However,
the
longitudinal

association
did
not
change
with
age
adjustment.
This
indicated
that
osteoporosis

and
atherosclerosis
are
not
associated
simply
due
to
a
common
age-‐related
process,

but
that
some
similar
pathophysiological
mechanisms
affect
both
processes.

Biological
Link
and
Possible
Pathophysiological
Mechanisms

Vascular
calcification
and
bone
mineralization
show
similarities
[1].
Calcifying

vascular
cells
appear
in
many
ways
similar
to
osteoblasts
[20]
and
specific
factors

and
proteins
crucial
to
bone
formation
are
also
associated
with
atherosclerosis

[1].

It
is
now
recognized
that
calcification
of
arterial
tissue
is
not
merely
a
passive

process
of
calcium
phosphate
precipitation
or
adsorption
in
end-‐stage

atherosclerosis,
but
instead
is
a
highly
organized
process
that
is
regulated
by

mechanisms
similar
to
those
involved
in
bone
mineralization
[23,24].

Epidemiological
data
suggest
that
estrogen
deficiency
is
a
risk
factor
for
both

cardiovascular
disease
and
osteoporosis
[19].
Arteries
and
bone
are
major
tissue

23

targets
for
estrogen.
Postmenopausal
women
who
are
estrogen
deficient
may

develop
aortic
artery
calcification
and
bone
loss
[1].

Aortic
calcification
influences
blood
flow
to
distal
regions
[21].
Impaired
blood

flow
to
the
lower
extremity
due
to
or
reflected
by
severe
atherosclerosis
in
the

lumbar
aorta
may,
at
least
to
some
extent,
contribute
to
the
complex
multi-‐factorial

pathogenesis
of
osteoporosis
in
the
hip
[21].

There
were
several
strengths
and
unique
features
of
our
study.
Most

importantly,
this
is
the
first
longitudinal
study
to
demonstrate
the
inverse

association
between
higher
BMD
and
lower
progression
of
atherosclerosis
in

healthy
postmenopausal
women.
The
WISH
participants
were
in
a
randomized

controlled
trial
of
postmenopausal
women
aged
45-‐92
years.
The
prevalence
of

prior
hormone
therapy
was
very
high
(70%).
Both
carotid
artery
ultrasound
and

bone
DXA
were
performed
under
strict
quality
control
standards
for
uniform
image

acquisition
and
minimization
of
measurement
error.
We
evaluated
and
found

consistent
associations
of
risk
factors
in
both
the
total
hip
and
lumbar
spine
BMD.

The
sample
size
that
contributed
to
estimation
and
testing
of
the
longitudinal
effect

was
large
(n=325)
and
77%
of
all
the
participants
contributed
more
than
6

measurements
of
CIMT;
this
provides
sample
size
and
power
to
detect
significant

associations
without
selection
bias.
We
used
mixed
effects
models
to
analyze
these

data;
this
method
allowed
us
to
flexibly
deal
with
the
longitudinal
data
involving

repeated
measures,
missing
measures
and
unequally
spaced
measures.
We
also

24

analyzed
the
association
between
CIMT
rate
and
BMD
stratified
by
BMI
and
smoking

categories;
we
failed
to
find
any
significant
effect
modification.

In
conclusion,
our
results
indicated
that
the
CIMT
progression
rate
is
negatively

associated
with
total
hip
and
lumbar
spine
BMD.
Low
BMD
in
the
total
hip
and

lumbar
spine
may
be
treated
as
a
marker
of
both
osteoporosis
and
atherosclerosis

progression
for
healthy
postmenopausal
women.
Postmenopausal
women
who
are

at
risk
for
osteoporosis
should
also
be
concerned
about
the
risk
for
atherosclerosis.

25

REFERENCES

[1]
A.Elisabeth
Hak,
Huibert
A.P.
Pols,
Albert
M.
van
Hemert,
Progression
of
Aortic

calcification
is
associated
with
metacarpal
bone
loss
during
menopause—A

population-‐based
longitudinal
study.Arterioscler
Thromb
Vasc
Biol.
2000;20:1926-‐
31.

[2]
Aldons
J.
Lusis.
Atherosclerosis.
Nature.Sep
14,
2000;407
(6801):233-‐241.

[3]
Retrieve
from
WebMed,
Osteoporosis
and
menopause.

http://www.webmd.com/menopause/osteoporosis-‐menopause

[4]Retrieve
from
international
osteoporosis
foundation,
Osteoporosis.

http://www.iofbonehealth.org/epidemiology

[5]
Howard
N.
Hodis,
Wendy
J.
Mack,
Isoflavone
Soy
protein
supplementation
and

atherosclerosis
progression
in
healthy
postmenopausal
women-‐
A
randomized

controlled
trial.
Stroke.2011;42:3168-‐3175.

[6]
Ihsane
Hamouchi,
Fadoua
Allali,
Hamza
Khazzani.
Low
bone
mineral
density
is

related
to
atherosclerosis
in
postmenopausal
Moroccan
women.
BMC
Public
Health.

2009;9:388.

[7]
A.
Luckish,
R.
Cernes,
M.
Boaz,
D.
Gavish,
Z.
Matas,
A.
Fux,
M.
Shargorodsky.
Effect

of
long-‐term
treatment
with
risedronate
on
arterial
compliance
in
osteoporotic

patients
with
cardiovascular
risk
factors.
Bone.
2008;43:279-‐283.

[8]
Chambless
LE,
Folsom
AR,
Clegg
LX,
Sharrett
AR,
Shahar
E,
Nieto
FJ,
Rosamond

WD,
Evans
G.
Carotid
Wall
Thickness
is
Predictive
of
Incident
Clinical
Stroke:
The

26

Atherosclerosis
Risk
in
Communities
(ARIC)
Study.

[9]
Small
RE.
Uses
and
Limitations
of
Bone
Mineral
Density
Measurements
in
the

Management
of
Osteoporosis.
MedGenMed.
2005;7:3.

[10]
Uyama
O,
Yoshimoto
Y,
Yamamoto
Y,
Kawai
A.
Bone
changes
and
carotid

atherosclerosis
in
postmenopausal
women.
Stroke.
1997;28:1730-‐1732.

[11]
Bots
ML,
Evans
GW,
Riley
W,
Meijer
R,
McBride
KH,
Paskett
ED,
Helmond
FA,

Grobbee
DE.
The
Osteoporosis
Prevention
and
Arterial
effects
of
tiboLone
(OPAL)

study:
design
and
baseline
characteristics.
Control
Clin
Trials.
2003;24,752-‐75.

[12]
Hodis
HN,
Mack
WJ,
Lobo
RA,
Shoupe
D,
Sevanian
A,
Mahrer
PR,
et
al.
Estrogen

in
the
prevention
of
atherosclerosis:
a
randomized,
double-‐blind,
placebo-‐controlled

trial.
Ann
Intern
Med.
2001;
135:939-‐953.

[13]
Karim
R,
Dell
RM,
Greene
DF,
Mack
WJ,
Gallagher
JC,
Hodis
HN.
Hip
fracture
in

postmenopausal
women
after
cessation
of
hormone
therapy:
results
from
a

prospective
study
in
a
large
health
management
organization.
Mnopause.

2011;18:1172-‐7.

[14]
De
Groot
E,
van
Leuven
SI,
Duivenvoorden
R,
Meuwese
MC,
Measurement
of

carotid
intima–media
thickness
to
assess
progression
and
regression
of

atherosclerosis.
Nat
Clin
Pract
Cardiovasc
Med.
2008;
5:280-‐8.

[15]
Sun
J1,
Zhang
C,
Xu
L,
Yang
M,
Yang
H.
The
transforming
growth
factor-‐β1
(TGF-‐
β1)
gene
polymorphisms
(TGF-‐β1
T869C
and
TGF-‐β1
T29C)
and
susceptibility
to

postmenopausal
osteoporosis:
a
meta-‐analysis.
Medicine
(Baltimore).
2015;94:e461.

27

[16]
Retrieved
from
NIH
osteoporosis
and
Africa
American
women,
osteoporosis

and
African
American
Women

http://www.niams.nih.gov/health_info/bone/osteoporosis/background/

[17]
Iqbal
J,
Sun
L,
Cao
J,
Yuen
T,
Lu
P,
Bab
I,
et
al.
Smoke
carcinogens
cause
bone
loss

through
the
aryl
hydrocarbon
receptor
and
induction
of
Cyp1
enzymes.
Proc
Natl

Acad
Sci
U
S
A.
2013;110:
11115-‐20.

[18]
Barengolts
EI,
Berman
M,
Kukreja
SC,
Kouznetsova
T,
Lin
C,
Chomka
EV.

Osteoporosis
and
coronary
atherosclerosis
in
asymptomatic
postmenopausal

women.
Calcif
Tissue
Int.
1998;62:209-‐13.

[19]
Tuomikoski
P1,
Mikkola
TS.
Postmenopausal
hormone
therapy
and
coronary

heart
disease
in
early
postmenopausal
women.
Ann
Med.
2014;46:1-‐7.

[20]
Watson
KE,
Boström
K,
Ravindranath
R,
Lam
T,
Norton
B,
Demer
LL.
TGF-‐beta
1

and
25-‐hydroxycholesterol
stimulate
osteoblast-‐like
vascular
cells
to
calcify.
J
Clin

Invest.
1994;93:2106-‐13.

[21]Tanko
LB,
Bagger
YZ,
Christiansen
C.
Low
bone
mineral
density
in
the
hip
as
a

marker
of
advanced
atherosclerosis
in
elderly
women.
Calcif
Tissue
Int.
2003;73:15–
20.

[22]
Nambi,
V.,
et
al.
Carotid
intima-‐media
thickness
and
cardiovascular
events.
The

Lancet.
379(9831):
2028-‐2030.

[23]
Vattikuti
R,
Towler
DA.
Osteogenic
regulation
of
vascular
calcification:
an
early

perspective.
Am
J
Physiol
Endocrinol
Metab.
2004;286:E686–E696.

28

[24]
Doherty
TM,
Detrano
RC.
Coronary
arterial
calcification
as
an
active
process:
a

new
perspective
on
an
old
problem.
Calcif
Tissue
Int.
1994;54:224–230.

Asset Metadata

Creator Li, Jiaonan (author)

Core Title Bone mineral density is associated with carotid atherosclerosis in healthy postmenopausal women: a longitudinal analysis of randomized clinical trial data

Contributor Electronically uploaded by the author (provenance)

School Keck School of Medicine

Degree Master of Science

Degree Program Biostatistics

Publication Date 08/27/2015

Defense Date 08/26/2015

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

Tag carotid atherosclerosis,lumbar spine BMD,OAI-PMH Harvest,postmenopausal,total hip BMD

Format application/pdf (imt)

Language English

Advisor Mack, Wendy (committee chair), Hodis, Howard (committee member), Karim, Roksana (committee member)

Creator Email jiaonanl@usc.edu,lijiaonan27@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-170432

Unique identifier UC11273401

Identifier etd-LiJiaonan-3850.pdf (filename),usctheses-c40-170432 (legacy record id)

Legacy Identifier etd-LiJiaonan-3850-0.pdf

Dmrecord 170432

Document Type Thesis

Format application/pdf (imt)

Rights Li, Jiaonan

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract Background and purpose: Atherosclerosis and osteoporosis are major causes of morbidity and mortality in postmenopausal women. Our study aimed to analyze the association between progression of carotid artery intima-media thickness (CIMT) and bone mineral density (BMD), providing more evidence for the relation between osteoporosis and subclinical atherosclerosis in postmenopausal women. ❧ Methods: In this longitudinal analysis, 349 postmenopausal women were followed for 3 years, and had an average 6 CIMT measurements. BMD had an average 3 measurements using DEXA