Intern
ati
o
n
a
l
Jo
urn
a
l
o
f
P
u
b
lic Hea
l
th Science (IJ
P
HS)
Vol
.
5,
N
o
.
1
,
M
a
rch
2
0
1
6
,
pp
.
60
~
69
I
S
SN
: 225
2-8
8
0
6
60
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJPHS
Association between Non-Ciga
rette/Smokeless Tobacco and
Hyp
ert
ension in the Nati
onal
Health Interview Survey:
A Pseudo-Panel Analysis
Olawunmi Ob
isesan
1,2
, Em
manuel
Thom
pson
2
, Adekunle
Obisesa
n
3
,
Olubus
ayo Akinola
4
,
Yv
on
ne Com
m
od
ore-
Men
s
ah
5
1
Emor
y
Univers
i
ty
, Rollins Scho
ol of
Publ
ic
Heal
th,
Atlanta, GA,
USA
2
Southeast Missouri State Univ
ersity
, C
a
pe Girar
d
eau, MO, USA
3
Southeast Hosp
ital
,
C
a
pe Girardeau, MO,
USA
4
Economic Com
m
unity
o
f
West
African
States (
E
COWAS), Abu
j
a, Niger
i
a
5
Emor
y
Univ
ersity
, Nell Hodgson Woodruff Sc
h
ool of Nursing
,
Atlanta, GA, USA
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Feb 02, 2016
Rev
i
sed
Feb
22
, 20
16
Accepted
Feb 28, 2016
This stud
y
sought to
examine
assu
mption between h
a
ving
ever
used non-
cigar
e
tt
e toba
cc
o or sm
okeless tobacco
, and a
diagnosis of hy
per
t
ensio
n
among a sample of 13, 086 United States
adults
participating in
the Nation
a
l
Health Interview Series from
2012-2014.
A pseudo-panel
analy
s
is of data
extracted
from the Integr
ated Health In
terview Series Surv
ey
was conducted
.
The gener
a
lized
linear mixed
model was used
to quantif
y
th
e effect of
a
histor
y
of non-
cigarette
tobacco
,
smoke
less tobacco,
and
socio-demographic
predictor var
i
ables on the r
e
sponse variab
le, a diagnosis of h
y
pertension. Th
e
transformed data, based
on the pseudo-pa
nel
techniqu
e, r
e
sulted in fif
t
y
-
seven (57) bir
t
h
cohorts and
followed in 2012
,
2013, and 2014
. The mean
age was 51.6
y
e
ars (±12.4).
The
findings of
this s
t
ud
y
revealed th
at th
e odds
of h
y
p
e
rtension
diagnosis for no
n-ci
gar
e
tte tob
a
cco users was 0.8846 times
lower (95% CI: 0.7907, 0
.
9896
) than non-users
after
adjusting
for possible
confounders such as age, langu
age, educ
ation, in
come and
y
e
ars of smoking.
Our stud
y
suggested th
at th
e association b
e
tween the use of n
on-cigar
e
tte
tobacco and the diagnosis of h
y
pert
ension amon
g the sample population is
consistent enou
gh to
assume a less pl
aus
i
ble
as
s
o
ciat
ion be
t
w
een th
e two
variab
les.
Keyword:
H
y
p
e
r
t
en
sion
No
n-
ci
garet
t
e
t
obacc
o
Sm
okeless tobacco
C
a
rdi
o
vasc
ul
ar
ri
sks
Pseudo-panel a
n
alysis
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Olawunm
i Obi
s
esan,
Em
o
r
y Un
iv
ersity,
Ro
llin
s Sch
o
o
l
of Pu
b
lic
Heal
th
,
1
518
Clifton
R
d
,
Atlan
t
a,
GA, USA.
Em
ail: oobises
a
n@gm
ail.co
m
1.
INTRODUCTION
Tobacco use i
s
a m
a
jor ris
k
for
hype
rtensi
on and
cardiovasc
u
lar
diseas
e (CVD)
a
n
d is the
m
o
st
im
portant caus
e
of prem
ature death in the
United Stat
es
[1]. Non-ci
garette tobacco a
nd sm
okeless tobacc
o
have
bee
n
sh
o
w
n
t
o
i
n
crease
t
h
e ri
s
k
of
C
V
D i
n
com
p
ari
s
on
t
o
n
o
n
-
sm
okers
[
1
]
.
Hy
pe
r
t
ensi
o
n
on
i
t
s
o
w
n
i
s
a
maj
o
r risk fact
or for CVD morbidity
and mortality [2]. Tobacco-relate
d illn
esses and hy
perte
n
sion in the US
co
sts m
o
re th
an
$
300
b
illio
n [3
] and
$4
6
b
illio
n
[4
] a year resp
ectiv
el
y. Man
y
stud
i
e
s h
a
v
e
ex
am
i
n
ed an
association
be
tween
hype
rte
n
sion a
n
d tobacco sm
oki
ng
(cigarette a
n
d non-c
i
g
are
tte) but the
exact
relatio
n
s
h
i
p
h
a
s rem
a
in
ed
un
clear [1
],[5
],[6
]. Exp
e
rts
h
a
v
e
p
r
oj
ected
t
h
at to
b
acco
u
s
e
wil
l
k
ill an
estim
a
t
ed
1
b
illio
n
p
e
o
p
l
e
in
th
e 2
1
s
t centu
ry [7
]. To
b
a
cco
u
s
e in
o
t
her form
s, b
e
sid
e
s cig
a
rette,
will also
co
n
t
rib
u
t
e to
in
creased
m
o
rbid
ity an
d
m
o
rtality rates g
l
ob
ally [7
].
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S I
S
SN
:
225
2-8
8
0
6
Asso
ci
a
tio
n b
e
t
w
een
Non
Ci
ga
rette/S
mo
keless To
ba
cco
a
n
d
Hyp
e
rten
sion in
th
e .... (
O
lawu
nmi Ob
isesan
)
61
The
continued use of non-ci
garette
tobacco is prom
oted
by
som
e
tobacco
com
p
anies under the clai
m
of a
ne
gl
i
g
i
b
l
e
or sm
al
l
e
r heal
t
h
ri
sk,
he
nc
e no
n
-
ci
garet
t
e
t
obacc
o (
N
C
T
) w
h
i
c
h
fal
l
un
de
r t
h
e t
h
re
e br
oa
d
cat
ego
r
i
e
s of
ot
he
r sm
oked pr
o
duct
s
, sm
okel
e
ss t
obacc
o
and ni
cot
i
n
e
pro
d
u
ct
s are l
e
ss regul
at
ed
t
h
an
ci
garet
t
e
s [6]
,
[
8
]
.
In
or
de
r t
o
m
i
nim
i
ze passi
ve ex
po
su
re t
o
sm
oke an
d t
o
red
u
ce ha
rm
att
r
i
but
e
d
t
o
i
n
v
e
t
e
rat
e
sm
okers, sm
okeless tobacco
(ST) was int
r
oduced as a su
bstitute for cigarette s
m
oking,
and a m
a
j
o
r a
d
junct
fo
r sm
oki
ng c
e
ssat
i
on [
9
]
.
Ex
pert
s ha
ve,
ho
weve
r, di
s
a
gree
d, wa
r
n
i
ng t
h
at
som
e
no
n
-
ci
garet
t
e
t
obac
c
o
p
r
od
u
c
ts are st
ill h
a
rm
fu
l since th
ese con
t
ain
certain
ingr
ed
ien
t
s, in
clud
i
n
g g
l
ycyrrh
izi
n
ic acid th
at lead
s t
o
pot
assi
um
wast
i
ng a
n
d s
odi
u
m
ret
e
nt
i
on w
h
i
c
h act
as
b
u
f
f
e
r
s
fo
r t
h
e faci
l
i
t
a
t
i
on
of
ni
c
o
t
i
n
e a
b
s
o
r
p
t
i
o
n.
These
have
bee
n
f
o
u
nd t
o
cau
se an
agg
r
a
v
at
i
on
o
f
hy
pe
rt
ensi
o
n
a
nd i
n
crease t
h
e
ri
sk
fo
r ch
r
oni
c di
seases i
n
cl
udi
ng
CV
D
[1
],[7
]-
[1
0
]
,[
11
]. W
h
il
e
so
m
e
r
e
sear
ch
er
s
h
a
v
e
f
ound
n
o
d
i
ff
er
en
ce
in
th
e pr
ev
al
en
ce o
f
h
y
p
e
r
t
en
sion
am
ong
users
of
no
n-ci
garet
t
e
t
obacc
o p
r
od
uct
s
[1
2]
-[
14]
,
ot
h
e
rs ha
ve fo
u
n
d
an associ
at
i
o
n
bet
w
een t
h
e
u
s
e of
no
n
-
ci
garet
t
e
t
obacc
o a
n
d
hy
pert
e
n
si
o
n
[1
0]
,[
15]
-
[
1
7
]
.
In
th
e Un
ited
States, th
e stati
s
tics fo
r th
ose who
u
s
e
NC
T i
s
unavai
l
a
bl
e. Ho
we
ver
,
an est
i
m
a
t
e
d 8.1
m
i
ll
i
on pe
o
p
l
e
(3
.2%
o
f
t
h
e
t
o
t
a
l
po
pul
at
i
o
n)
use S
T
, m
o
st
u
s
ers
bei
n
g m
a
les, ag
e 1
8
-
2
5
y
ears [
1
]
.
T
h
e
use
of
ST i
s
com
m
onest
am
ong
Nat
i
ve
Am
eri
cans (7%
)
a
n
d
Whi
t
e
s (
4
.
3
%) a
s
w
e
l
l
as am
ong
l
o
wer s
o
ci
o-ec
on
om
i
c
g
r
ou
p
s
, in
cl
u
d
i
n
g
b
l
u
e
-co
llar/u
n
s
k
illed
/
un
em
p
l
o
y
ed
worker
s an
d
no
n-colleg
e
g
r
adu
a
tes who
s
e
h
ealth
statu
s
es
m
a
y
al
ready
be i
n
fl
ue
nce
d
b
y
ot
her s
o
ci
al
det
e
rm
i
n
ant
s
o
f
heal
t
h
[
1
]
.
I
n
t
h
e 1
2
-
1
7 y
ear
age
gr
ou
p,
dat
a
fr
om
20
0
2
-
2
00
7 s
h
o
w
s a m
o
re t
h
an 1
0
-
f
o
l
d
hi
g
h
e
r p
r
eval
e
n
ce
of
use am
ong
boy
s
(6
.3%
)
t
h
an
gi
rl
s (
0
.
4
%) [
1
]
.
Desp
ite so
m
e
clai
m
s
th
at s
m
o
k
e
less tob
acco
is less h
a
rm
f
u
l th
an
cig
a
rett
e s
m
o
k
i
n
g
, th
e fact sti
ll re
m
a
i
n
s th
at
users a
r
e exposed to ha
rm
ful chem
icals
th
at
co
u
l
d
h
e
igh
t
en th
eir risk
fo
r sev
e
ral h
ealth
prob
lem
s
j
u
st as with
ci
garet
t
e
sm
oki
ng
[
9
]
.
Many epi
d
em
iological studies ha
ve e
x
am
ined th
e ass
o
ciation
between cigarettes a
n
d
hypertensi
on
but in
recent
years, only
m
i
nim
a
l
research has been done on the
health
effects of
non-
cigarette tobacco
[18
]
,[19
]. Giv
e
n
th
e m
o
rb
id
ity,
m
o
rtality
an
d
co
st asso
ciated
with
sm
o
k
i
n
g
, it is ex
p
e
dien
t to
estab
lish
the
rol
e
of
S
T
p
r
o
duct
s
i
n
t
h
e de
vel
o
pm
ent
of C
VD ri
sk fact
ors suc
h
as hy
pert
e
n
si
o
n
[
6
]
.
The pu
r
pose of
t
h
i
s
study
was
to e
x
am
ine the as
sociation bet
w
een e
v
er sm
ok
ed
no
n
-
ci
ga
ret
t
e
t
obacc
o i
n
one
’s l
i
f
et
i
m
e and
a
di
ag
no
si
s o
f
hy
pert
e
n
si
o
n
i
n
a
nat
i
onal
l
y
-
r
ep
r
e
sent
at
i
v
e sam
p
l
e
o
f
US
ad
ul
t
s
.
2.
R
E
SEARC
H M
ETHOD
2.
1.
Stud
y design and
p
o
pul
ation
Data was
extracted from
the Inte
g
r
ated
Health
In
terv
iew
Series
(IHIS) d
a
taset
wh
ich
co
m
p
rise
of
repeat
e
d
cr
oss
-
sect
i
onal
sam
p
l
e
s f
o
r
20
1
2
,
20
1
3
, an
d
20
14
. The
IH
IS
i
s
deri
ve
d f
r
o
m
t
h
e Nat
i
ona
l
Heal
t
h
In
terv
iew Su
rvey (NHIS)
pu
blic u
s
e
d
a
tasets. The NHIS
is
th
e prin
ci
p
a
l so
urce
o
f
informatio
n
on
th
e
h
ealth
of t
h
e
US
po
pul
at
i
o
n a
n
d
i
s
a cr
oss
-
sect
i
onal
h
ouse
h
ol
d i
n
t
e
r
v
i
e
w
s
u
rvey
.
The
sam
p
l
i
n
g
pl
a
n
f
o
l
l
o
ws
a
m
u
l
tistag
e
area p
r
ob
ab
ility
d
e
sig
n
th
at
p
e
rm
i
t
s th
e rep
r
esen
tative sa
m
p
lin
g
o
f
hou
seh
o
l
ds an
d
n
o
n
i
nstitu
tio
n
a
lized
in
d
i
v
i
du
als with
ov
ersam
p
l
i
n
g
o
f
Black
s and
Hisp
an
ics.
2.
2.
Vari
ables
defi
nition
and me
asures
Th
e d
a
ta fo
r t
h
is stu
d
y
was
ex
tr
acted from the Integrated
Health
In
terv
i
e
w Series (IHIS) d
a
taset
whi
c
h
com
p
ri
s
e
of re
peat
ed
c
r
oss
-
sect
i
o
ns f
o
r 2
0
1
2
, 2
0
1
3
, and 2
0
1
4
.
The
IH
IS
i
s
base
d on o
r
i
g
i
n
al
dat
a
fr
o
m
Natio
n
a
l
Health
In
terv
iew Surv
ey (NHIS)
pu
b
lic d
a
taset.
The vari
abl
e
s use
d
fr
om
t
h
e
IH
IS dat
a
set
i
n
cl
ude
d:
ever t
o
l
d
ha
d hy
pe
rt
ensi
o
n
o
n
2+ vi
si
t
s
(
H
YP
2TIM
E
)
, a
g
e (AG
E
)
,
l
a
ng
uage
of i
n
t
e
rvi
e
w (I
NTER
VL
AN
G)
,
ed
u
cation attain
m
e
n
t
(EDUCREC2
), p
e
rso
n
'
s to
tal
ea
r
n
i
n
gs (E
AR
N
I
NG
S),
e
v
er s
m
oked
n
o
n
-ci
g
aret
t
e
t
obacc
o
pr
od
u
c
t
s
(N
ONC
IG
EV)
,
an
d e
v
er
use
d
sm
okel
e
ss t
obac
c
o
(S
M
K
LSE
V
)
.
A
p
art
fr
om
HYP2T
I
M
E
t
h
at
was used
as t
h
e resp
onse
vari
abl
e
, t
h
e r
e
st
were pre
d
i
c
t
o
r va
ri
abl
e
s.
Hy
pe
rt
ensi
o
n
was de
fi
ne
d as t
w
o or
m
o
re occasions of ele
v
ated
blood press
u
re a
t
office
vis
its per national gui
d
elines [2
0] Educational attainm
e
nt
was d
e
fi
ne
d as
:
1 = ne
ver at
t
e
nde
d/
ki
nde
r
g
ar
t
e
n, 2 =
g
r
ade
1,
3 = g
r
a
d
e 2
,
4 = g
r
a
d
e 3
,
5
= gra
d
e
4,
6 =
gra
d
e
5
,
7 =
g
r
ad
e
6
,
8 =
g
r
ad
e
7
,
9
= gr
ad
e
8
,
10
= gr
ad
e
9
,
11
= gr
ad
e
1
0
,
1
2
=
g
r
ad
e 11
, 13
= 12
th gr
ad
e,
no
di
pl
om
a, 1
4
=
hi
g
h
sc
h
ool
g
r
aduat
e
,
15
=
G
E
D
or
e
qui
val
e
nt
,
1
6
=
som
e
col
l
e
ge,
n
o
de
g
r
ee,
1
7
=
A
A
d
e
gree:
Tech
ni
cal
/
voca
t
i
onal
/
o
cc
upat
i
onal
,
1
8
= A
A
deg
r
ee:
Aca
d
em
i
c
prog
ram
,
19 = B
ach
el
o
r
'
s
degree
(B
A
,
AB
,
B
S
, B
B
A
), 2
0
= M
a
st
er'
s
deg
r
ee (M
A, M
S
,
M
e
ng
, M
e
d, M
B
A),
21 = Pr
of
essi
onal
de
g
r
e
e
(M
D, D
D
S,
DVM
,
JD)
,
an
d 2
2
=
Doct
o
r
al
de
gr
ee. Lan
gua
ge
of i
n
t
e
r
v
i
e
w was categorized as Englis
h
,
S
p
ani
s
h, E
ngl
i
s
h an
d
Sp
an
ish
or
O
t
her
.
Earn
ing
s
wer
e
classif
i
ed
as:
1
=
$0
1
to
$49
99
,
2
= $
500
0
to
$
999
9, 3
= $
100
00
t
o
$
149
99,
4
=
$
150
00
t
o
$
199
99
, 5
= $2
000
0 to
$2
499
9,
6
=
$
250
00
to
$3
4999
,
7 = $3
500
0 to
$4
499
9,
8
=
$
4
5
000
t
o
$
549
99
, 9 =
$
5
5
000
to $64
999
,
10
=
$6
500
0 to
$
749
99
, and
1
1
=
$
750
00
an
d ov
er
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
I
J
PH
S Vo
l. 5
,
N
o
. 1
,
Mar
c
h
20
16
:
60
–
69
62
2.
3.
St
at
ist
i
ca
l analy
s
es
Fre
que
nci
e
s a
n
d pe
rce
n
t
a
ges
were
desc
ri
pt
i
v
e m
easur
es
us
ed to s
u
mm
arize th
e categori
cal variables
,
whi
l
e
m
ean an
d st
an
da
rd
de
v
i
at
i
on we
re
us
ed t
o
desc
ri
be
t
h
e co
nt
i
n
uo
u
s
vari
a
b
l
e
(
A
G
E
).
In t
h
e m
o
d
e
l
i
ng
process
,
age was transform
e
d to a
value on a logarithm
scale. The ge
ner
a
l
i
zed l
i
n
ear m
i
xed m
odel
(G
LM
M
)
with
a lo
g
it link
was
u
s
ed
to
q
u
a
n
tify th
e effect o
f
eac
h
of
th
e p
r
ed
ictors
o
n
t
h
e resp
on
se v
a
riab
le. A
detailed
d
e
scri
p
tio
n o
f
th
e
m
o
d
e
l
is
sh
own
u
n
d
e
r 2
.
4
.
Qu
ite
im
p
o
r
tan
tly,
th
e u
s
e
of
t
h
e IHIS dataset
h
a
s on
e
m
a
in
dra
w
back
beca
use sam
e
subj
ect
s were n
o
t
fol
l
o
we
d t
h
r
o
u
gh t
i
m
e, hence hi
st
ori
cal
i
n
f
o
rm
ati
on o
n
su
bject
s
were no
t av
ailab
l
e. Th
is
d
r
awb
a
ck
im
p
o
s
es li
m
ita
tio
n
to th
e
u
s
e
of
p
a
n
e
l d
a
ta m
o
d
e
ls. To ov
ercome th
is
lim
itation, Dea
t
on (1985) [21] sugge
sts foll
owi
ng c
o
hort
s due t
o
the acce
ssibility of household s
u
rveys
.
W
e
t
h
eref
o
r
e co
nst
r
uct
e
d a pseu
d
o
-
p
a
n
el
by
pul
l
i
ng i
ndi
vi
d
u
al
sub
j
ect
s sha
r
i
ng c
o
m
m
on y
e
ar of
bi
rt
h f
r
o
m
one
sur
v
ey
y
ear t
o
anot
her i
n
acc
or
da
nce wi
t
h
[
21]
. Eac
h
y
ear
of bi
rt
h f
o
rm
s a bi
rt
h co
ho
rt
.
The bi
rt
h co
h
o
rt
f
o
r
each i
ndi
vidua
l
subject ca
n be obtaine
d
by s
ubt
racting the
age
of the
subject
from
the year
he/she
pa
rticipated
in
th
e surv
ey.
2.
4.
Generaliz
ed
Linear Mixed Model
Th
is p
a
rt o
f
the stu
d
y
o
u
tlin
es th
e
m
a
in
featu
r
es of th
e GLMM. Th
e GLMM co
n
t
ain
s
b
o
t
h
fi
x
e
d
effect
s a
n
d ra
nd
om
effect
s.
It
i
s
a
n
ext
e
n
s
i
on
o
f
t
h
e
ge
neral
i
zed
l
i
n
ea
r m
odel
(
G
L
M
) by
a
d
di
n
g
ran
d
o
m
effect
s t
o
t
h
e l
i
near
pre
d
i
c
t
o
r
com
pone
nt
.
A G
L
M
i
s
a r
e
gres
si
o
n
m
odel
defi
n
e
d
wi
t
h
i
n
t
h
e di
st
ri
b
u
t
i
ona
l
st
ruct
u
r
e
of
t
h
e
ex
po
ne
nt
i
a
l
fa
m
i
ly
and i
s
gi
v
e
n as
f
o
l
l
o
w
s
:
f
y
e
x
p
θ
ψ
θ
ϕ
c
y,
ϕ
1
whe
r
e
μ
E
y
ψ
′
θ
,
Va
r
y
ϕψ
′′
θ
ϕ
V
μ
.
V
μ
capt
u
res
t
h
e
rel
a
t
i
o
nshi
p
bet
w
ee
n
E
y
and
Va
r
y
.
The GLM
ca
n therefore be
re
prese
n
ted
as:
g
μ
η
X
β
2
whe
r
e
, p
-
di
m
e
nsi
o
nal
v
ect
o
r
;
X,
mp
vect
or
o
f
pre
d
i
c
t
o
rs;
g
,
a l
i
nk
f
unct
i
o
n;
, a linear
predictor.
estim
a
tes are obtaine
d by
using an iterative
num
e
rical
app
r
oach
t
o
s
o
l
v
e
m
a
xim
u
m
li
kel
i
hoo
d e
quat
i
o
n
s
.
G
L
MMs ex
tend
G
L
Ms b
y
i
n
clu
d
i
ng r
a
n
dom
ef
f
ects
Zu
to th
e lin
ear
pred
i
c
to
r
s
how
n in
(
2
)
.
Th
e r
a
nd
om
effects allow
co
hort-sp
ecific p
r
ed
ictio
n, heterog
e
n
e
ity
between
co
horts, and
stru
ctural co
rrelation
with
in
coh
o
r
t
s
.
C
o
ndi
t
i
onal
o
n
a
q-
d
i
m
e
nsi
onal
vec
t
or
u
o
f
ra
nd
om
ef
fect
s
f
o
r
c
o
ho
rt
i,
GLMM
assu
m
e
s for th
e
j
th
r
e
spon
se
on
coh
o
r
t
i,
y
are:
y
|u
~f
|
y
|u
,
u
~f
u
,
f
|
y
|u
e
x
p
y
θ
ψ
θ
ϕ
c
y
,
ϕ
3
Th
e fo
llowing
relatio
n
s
ho
ld
:
1.
μ
E
y
|u
ψ
θ
2.
Va
r
y
|u
ϕ
ψ
θ
ϕ
V
μ
A f
u
nct
i
o
n
of
μ
wh
ich
is lin
ear
in
bo
th
and
u
i
s
gi
ve
n
by
:
gμ
X
βZ
u
4
Whe
r
e
X
and
Z
are p
an
d
q
d
i
m
e
nsi
onal
ve
ct
ors
of
k
n
o
w
n
pre
d
i
c
t
o
rs
f
o
r
fi
xe
d a
n
d r
a
nd
om
effect
s
respectively. T
h
e ra
ndom
effects signify
th
e i
m
p
act o
f
co
hort i o
n
its rep
e
ated
o
b
serv
ation th
at is n
o
t
cap
tu
red
by
t
h
e obse
r
ve
d pre
d
i
c
t
o
r.
Wi
t
h
t
h
ese ran
d
o
m
effect
s,
th
e id
ea is th
at, th
e sa
m
p
led
co
ho
rts are tho
ugh
t to
rep
r
ese
n
t
a p
o
pul
at
i
o
n
of al
l
po
ssi
bl
e c
o
h
o
r
t
s
an
d m
o
st
o
f
t
e
n ass
u
m
e
d t
o
f
o
l
l
o
w a
no
r
m
al
di
st
ri
but
i
o
n wi
t
h
zero m
ean and
varia
n
ce
σ
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S I
S
SN
:
225
2-8
8
0
6
Associ
ation bet
w
een Non Ci
garette/S
mokele
ss
T
o
bacc
o and Hype
rtension
in the
.... (
O
la
wunmi Obises
an)
63
2.
5.
Estima
tion
Th
e lik
elihoo
d
asso
ciated
with (3
) is g
i
v
e
n b
y
:
L
β,
D|y
f
|
f
udu
5
M
a
xi
m
u
m
l
i
k
el
i
hoo
d est
i
m
at
ion a
nd est
i
m
at
i
on o
f
and
D are ham
p
ered by the inte
gral in (5),
therefore approxim
a
te techniques ar
e
used for the estimation. Techni
ques
such as the Laplace,
pseudo
lik
elih
o
o
d
,
and
nu
m
e
rical i
n
tegratio
n
h
a
ve b
een
p
r
op
o
s
ed
in
th
e literatu
re [22
]
. In o
u
r case, we u
s
ed
num
eri
cal
i
n
t
e
grat
i
o
n
base
d
on
t
h
e a
d
a
p
t
i
v
e Ga
uss-
He
rm
i
t
e
qua
drat
ure
[2
3]
. T
o
est
i
m
at
e t
h
e rel
a
t
i
o
nshi
p
b
e
tween
HYP2
TIME and the six
(6
)
pred
i
c
to
rs, th
e pr
es
ent study ass
u
mes a norm
al
d
i
stribu
tion
fo
r th
e
random
effects and a logit link for
g. The logit link was appropri
ate
because
our re
sponse variable
“HY
P
2T
IM
E” can
t
a
ke one o
f
t
w
o p
o
ssi
bl
e val
u
es rep
r
ese
n
t
i
ng
“y
es” or “no”
t
o
eve
r
t
o
l
d
ha
d
hy
pert
e
n
si
o
n
o
n
2
+
v
i
sits. To
d
e
term
in
e statistical sig
n
i
fican
ce, we u
s
ed
o
d
d
s
ratio esti
m
a
ted
fro
m
th
e
GLMM
with
the
logit link a
nd the ass
o
ciated
1
α
%
confide
n
ce
interval
where
re
prese
n
ts t
h
e level
of
signi
ficance.
R
S
t
udi
o versi
o
n 0.
99
.4
4
6
(R
S
t
udi
o
,
Inc
.
) st
at
i
s
t
i
cal
package
was use
d
f
o
r
t
h
e
a
n
al
y
s
i
s
.
3.
R
E
SU
LTS AN
D ANA
LY
SIS
3.
1.
Descripti
v
e Statis
tics
Of 1
3
,
0
86
part
i
c
i
p
ant
s
i
n
v
o
l
v
ed i
n
t
h
e st
udy
, 4,
28
4 (
3
2
.
7
4
%
), 4
,
1
4
0
(3
1.
64
%), a
nd 4
,
6
62 (
3
5.
63
%)
were
f
r
om
20
1
2
,
2
0
1
3
,
a
n
d
2
0
1
4
su
rvey
y
ears
respect
i
v
el
y
.
The
t
r
a
n
sf
orm
e
d
dat
a
base
d
on
t
h
e
pse
u
do
-
p
ane
l
tech
n
i
qu
e
r
e
sulted
in
f
i
f
t
y-
sev
e
n
(5
7)
b
i
r
t
h
co
hor
ts and
each
coho
r
t
w
a
s
f
o
llow
e
d
in
2
0
1
2
, 201
3, an
d
2
014.
The y
o
u
nge
st
and
t
h
e
ol
dest
part
i
c
i
p
a
n
t
s
were
2
1
a
nd
79 years
old re
spectively.
Als
o
, t
h
e ave
r
a
g
e
age
of
p
a
rticip
an
ts
was 51
.6
years
with
a st
a
nda
rd
d
e
vi
at
i
on o
f
12
.
4
y
ears
.
Fi
gu
re
1 s
h
ow
s t
h
e
pr
o
p
o
r
t
i
o
n
of
pa
rt
i
c
i
p
an
t
s
fo
r eac
h
o
f
t
h
e
57
bi
rt
h c
o
h
o
rt
s
w
h
o
ha
d
h
y
p
ert
e
nsi
o
n
d
u
r
i
ng
th
e 2012
, 201
3, and
20
14
su
rv
eys. Each
lin
e r
e
pr
esen
ts a b
i
r
t
h
coh
o
r
t
and
th
e pro
por
tio
n
s
f
o
r
all th
e
bi
rt
h c
o
h
o
rt
s
ran
g
e f
r
om
0.40
(2
0
12 s
u
r
v
ey
) t
o
1.
00
(
2
0
1
4
su
rvey
).
The m
i
nim
u
m
and t
h
e m
a
xi
m
u
m
pr
o
p
o
r
t
i
o
n
s
d
u
r
i
n
g
t
h
e 2
0
1
2
sur
v
ey
were
a
s
soci
at
ed wi
t
h
t
h
e 19
9
1
(
0
.
4
0)
an
d
t
h
e 1
9
35 (1
.0
0)
bi
rt
h
c
o
ho
rt
s
respect
i
v
el
y
.
Whi
l
e
t
h
e m
i
nim
u
m
and m
a
xi
m
u
m
prop
o
r
t
i
ons
d
u
ri
ng t
h
e 2
0
1
3
su
r
v
e
y
were
respec
t
i
v
el
y
co
nn
ected
t
o
th
e 19
88
(0
.41)
and
19
45
(0.98
)
b
i
r
t
h
co
ho
r
t
s.
I
n
th
e
20
14
su
rv
ey, the
m
i
n
i
m
u
m a
n
d
th
e
max
i
m
u
m
p
r
op
or
tion
s
w
e
r
e
0
.
5
2
and 1.00
resp
ectiv
ely link
e
d to th
e
19
89
an
d th
e 193
9 b
i
r
t
h
co
hor
ts.
Br
o
a
d
l
y
spea
king, the
proportions am
ong all the
57 birth cohorts
we
re consistently
above
0.4 apa
r
t from
that of
1991
d
u
r
i
ng
t
h
e
2
012
su
rv
ey.
Fi
gu
re 1.
Pr
o
p
o
rt
i
o
n wi
t
h
hy
pert
e
n
si
o
n
i
n
e
ach bi
rt
h
co
h
o
r
t
Fi
gu
re 2
di
s
p
l
a
y
s
t
h
e
p
r
o
p
o
r
t
i
on of part
i
c
i
p
ant
s
w
ho ha
d
hy
pe
rt
ensi
on by
ed
ucat
i
o
nal
at
t
a
i
n
m
e
nt
.
Large
va
ri
at
i
o
ns
i
n
pr
op
o
r
t
i
ons
w
e
re o
b
se
rve
d
at
gr
ade
2 du
ri
n
g
t
h
e 2
0
1
3
a
n
d 20
1
4
sur
v
ey
y
ears
.
It
ca
n
furthe
r be
observe
d
that, t
h
e
varia
b
ilities te
nd t
o
dwindle
and
proportions
m
ove close
r
to each
othe
r as we
m
ove fr
om
l
o
wer l
e
vel
of
E
DUC
R
E
C
2
t
o
hi
g
h
er
l
e
vel
s
with
so
m
e
m
i
n
o
r
d
e
p
a
tu
re startin
g
fro
m
th
e th
ird
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
I
J
PH
S Vo
l. 5
,
N
o
. 1
,
Mar
c
h
20
16
:
60
–
69
64
h
i
gh
est lev
e
l in
20
12
.
Alm
o
s
t
all
th
e p
r
op
ortio
n
s
du
ri
n
g
the th
ree su
rv
ey years co
n
s
istently were ab
ov
e 0
.
60
in
ex
cep
tion
o
f
on
e
w
h
ich
w
a
s ob
serv
ed
at
gr
ad
e 2 (0
.3
3)
du
r
i
n
g
201
3 surv
ey.
Fi
gu
re 2.
Pr
o
p
o
rt
i
o
n wi
t
h
hy
pert
e
n
si
o
n
by
e
ducat
i
o
nal
at
t
a
i
n
m
e
nt
Fi
gu
re
3 s
h
o
w
s
t
h
e p
r
op
ort
i
o
n
of
pa
rt
i
c
i
p
ant
s
wh
o
ha
d
hy
pe
rt
ensi
o
n
by
l
a
n
gua
ge
of
i
n
t
e
r
v
i
e
w wi
t
h
i
n
th
e 3
-
year
sp
an
o
f
th
e su
rv
ey in
th
e Sp
an
i
s
h
eth
n
i
c popu
latio
n
s
. Th
ere w
a
s also
an o
b
s
er
v
e
d
in
crease in
hy
pe
rt
ensi
o
n
d
i
agn
o
si
s am
ong i
n
di
vi
d
u
al
s
wh
o s
p
eak
ot
h
e
r l
a
ng
ua
ges a
p
art
f
r
om
Eng
l
i
s
h and
Spa
n
i
s
h i
n
20
1
4
(Fi
g
u
r
e 3)
. It
coul
d fu
r
t
her be o
b
se
rv
ed t
h
at
cl
ear u
nde
rst
a
n
d
i
n
g o
f
di
ag
n
o
si
s an
d p
o
ssi
bl
e di
a
g
n
o
st
i
c
i
m
p
licatio
n
s
based
on
languag
e
h
a
d
po
sitiv
e eff
ect on
th
e pr
opo
r
tion
o
f
o
b
s
erv
e
d
hyp
er
ten
s
i
o
n
d
i
ag
no
sis
am
ong
st
t
h
e
E
ngl
i
s
h s
p
eaki
n
g
po
p
u
l
a
t
i
ons
wi
t
h
no
o
b
se
rv
ed c
h
an
ges
wi
t
h
i
n
t
h
e
3 y
ears
of
su
rv
ey
.
Fig
u
r
e
3
.
Propo
r
tion
r
e
sp
onded
yes t
o
H
Y
P2
TI
ME b
y
IN
TERV
LA
NG
Fi
gu
re 4 di
spl
a
y
s
t
h
e pr
op
o
r
t
i
o
n
o
f
part
i
c
i
p
a
n
t
s
wi
t
h
hy
pert
ensi
o
n
di
ag
n
o
s
i
s
by
ear
ni
n
g
s
.
The
l
o
west
pr
o
p
o
r
t
i
o
n
s
we
re o
b
se
rve
d
a
m
ong ear
ni
n
g
gr
o
ups
4,
3, a
n
d 6 i
n
2
0
1
2
,
2
0
1
3
, a
n
d 2
0
14
respect
i
v
el
y
.
F
u
rt
her,
t
h
e
hi
g
h
est
pr
op
o
r
t
i
ons
wer
e
rec
o
r
d
e
d
a
m
ong ear
ni
n
g
g
r
o
u
p
s
2,
5
,
an
d
1 i
n
2
0
1
2
,
20
1
3
,
an
d
2
0
1
4
respectively.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S I
S
SN
:
225
2-8
8
0
6
Associ
ation bet
w
een Non Ci
garette/S
mokele
ss
T
o
bacc
o and Hype
rtension
in the
.... (
O
la
wunmi Obises
an)
65
Fig
u
r
e
4
.
Propo
r
tion
r
e
sp
onded
yes t
o
H
Y
P2
TI
ME b
y
EARN
IN
GS
Fi
gu
res 5
a
nd Fi
gu
re 6
di
s
p
l
a
y
t
h
e
pr
o
p
o
r
t
i
on o
f
pa
rt
i
c
i
p
a
n
t
s
t
h
at
ha
d
a hy
pe
rt
ensi
o
n
d
i
agn
o
si
s
by
ever sm
oked
non-ciga
rette tobacco
pr
oducts
and eve
r
use
d
s
m
okeless to
bacco res
p
ectively.
The proportion
wi
t
h
hy
pe
rt
ens
i
on
di
a
g
n
o
si
s i
n
crease
d
m
a
rgi
n
al
l
y
fr
om
0.8
3
i
n
20
1
2
t
o
0
.
84
i
n
2
0
1
4
am
on
g t
h
o
s
e
wh
o
ha
d
neve
r
use
d
NC
T.
Ho
we
ver
t
h
e p
r
o
p
o
rt
i
o
n
d
ecl
i
n
ed
fr
om
abo
u
t
0.
8
1
t
o
0
.
79
bet
w
ee
n
20
12
an
d
2
0
1
4
f
o
r t
h
ose
wh
o had
e
v
er use
d
NC
T. Al
s
o
,
t
h
e pr
o
p
o
r
t
i
on wi
t
h
hy
pert
ensi
o
n
di
a
g
nos
i
s
rose by
1% (
0
.
82
t
o
0
.
8
3
) b
e
t
w
ee
n
20
1
2
an
d
20
1
4
am
ong t
hos
e had
ne
ver
use
d
ST.
Wh
i
l
e, t
h
e p
r
o
p
o
r
t
i
on ha
d e
v
er
use
d
ST
b
u
t
had
a
h
y
p
e
r
t
en
sion
diag
no
sis
d
r
op
ped
b
y
3% (0
.81-
0.78
) fr
o
m
2
0
1
2
to
20
14
.
Fig
u
r
e
5
.
Propo
r
tion
r
e
sp
onded
yes t
o
H
Y
P2
TI
ME b
y
NON
CIG
E
V
Fig
u
r
e
6
.
Propo
r
tion
r
e
sp
onded
yes t
o
H
Y
P2
TI
ME b
y
SM
K
L
SEV
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
I
J
PH
S Vo
l. 5
,
N
o
. 1
,
Mar
c
h
20
16
:
60
–
69
66
3.
2.
Odds R
a
ti
o A
n
al
ysi
s
Tabl
e 1 sh
ow
s
t
h
e est
i
m
a
t
e
d
od
ds rat
i
o
s a
n
d
t
h
ei
r corres
p
o
ndi
ng
95% c
o
n
f
i
d
e
n
ce i
n
t
e
rva
l
s from
t
h
e
GLMM m
o
d
e
l with th
e log
it lin
k
.
Here, we
fo
cu
s
on
si
g
n
i
ficant effects.
Our a
n
alysis
revealed that t
h
e odds
of
hy
pe
rt
ensi
o
n
di
a
g
n
o
si
s f
o
r
t
hose
wh
o r
e
s
p
o
n
d
ed
‘y
es’ t
o
t
h
e
use o
f
N
C
T was 0
.
8
8
4
6
t
i
m
e
s
l
o
wer
(9
5% C
I
.7
90
7, .
9
89
6
)
t
h
an t
hose
wh
o
resp
o
nde
d ‘
n
o’
, aft
e
r ad
j
u
st
i
n
g f
o
r t
h
e c
o
n
f
o
u
n
d
e
r
s o
f
age,
l
a
ng
uage
, ed
uc
at
i
on,
i
n
com
e
and y
e
ars
of
sm
oki
ng
.
Tab
l
e
1
.
Esti
mated
GLMM with
a Lo
g
it Li
nk
P
a
ra
m
e
t
e
r
Est
i
m
a
te
SE
Z
ra
t
i
o
P
value
O
dds ratio
95%
9CI
I
n
ter
cept -
5
.
2516
0.
5409
-
9
.
7090
0.
0000
0.
0052
(
0
.
0018,
0.
015
1)
L
og(
AGE) 1.
6849
0.
0834
20.
209
0
0.
0000
5.
3917
(
4
.
5789,
6.
348
9)
INTERVLANG 2
-0.3056
0.1459
-2.0940
0.0363
0.7367
(0
.5534, 0.980
7)
INTERVLANG 3
-0.0284
0.2033
-0.1400
0.8888
0.9720
(0
.6525, 1.447
8)
INTERVLANG 4
0.1609
0.
4895
0.3290
0.7424
1.1745
(0.4500,
3.065
5)
E
A
RNINGS 2
-
0
.
0273
0.
1389
-
0
.
1970
0.
8440
0.
9730
(
0
.
7411,
1.
277
5)
E
A
RNINGS 3
-
0
.
2353
0.
1275
-
1
.
8460
0.
0649
0.
7903
(
0
.
6155,
1.
014
7)
E
A
RNINGS 4
-
0
.
2308
0.
1306
-
1
.
7680
0.
0771
0.
7939
(
0
.
6146,
1.
025
4)
E
A
RNINGS 5
-
0
.
1011
0.
1298
-
0
.
7790
0.
4359
0.
9038
(
0
.
7008,
1.
165
6)
E
A
RNINGS 6
-
0
.
1775
0.
1165
-
1
.
5230
0.
1277
0.
8374
(
0
.
6664,
1.
052
2)
E
A
RNINGS 7
-
0
.
3167
0.
1191
-
2
.
6600
0.
0078
0.
7286
(
0
.
5769,
0.
920
1)
E
A
RNINGS 8
-
0
.
2218
0.
1255
-
1
.
7670
0.
0772
0.
8011
(
0
.
6264,
1.
024
5)
E
A
RNINGS 9
-
0
.
3605
0.
1324
-
2
.
7230
0.
0065
0.
6973
(
0
.
5379,
0.
903
9)
E
A
RNINGS 10
-
0
.
2360
0.
1486
-
1
.
5880
0.
1122
0.
7898
(
0
.
5903,
1.
056
8)
E
A
RNINGS 11
-
0
.
3572
0.
1201
-
2
.
9760
0.
0029
0.
6996
(
0
.
5529,
0.
885
2)
E
DUCRE
C2 2
0.
2848
0.
8946
0.
3180
0.
7503
1.
3295
(
0
.
2302,
7.
676
9)
E
DUCRE
C2 3
0.
2490
0.
7071
0.
3520
0.
7247
1.
2827
(
0
.
3208,
5.
128
6)
E
DUCRE
C2 4
0.
5391
0.
6469
0.
8330
0.
4047
1.
7144
(
0
.
4825,
6.
092
2)
E
DUCRE
C2 5
0.
4322
0.
6945
0.
6220
0.
5338
1.
5406
(
0
.
3949,
6.
009
9)
E
DUCRE
C2 6
-
0
.
0615
0.
5705
-
0
.
1080
0.
9142
0.
9403
(
0
.
3074,
2.
876
9)
E
DUCRE
C2 7
0.
2629
0.
4590
0.
5730
0.
5668
1.
3007
(
0
.
5290,
3.
198
4)
E
DUCRE
C2 8
-
0
.
0124
0.
5706
-
0
.
0220
0.
9827
0.
9877
(
0
.
3228,
3.
022
0)
E
DUCRE
C2 9
0.
3415
0.
4869
0.
7010
0.
4831
1.
4070
(
0
.
5418,
3.
653
6)
E
DUCRE
C2 10
0.
4315
0.
4606
0.
9370
0.
3489
1.
5396
(
0
.
6242,
3.
797
5)
E
DUCRE
C2 11
0.
3981
0.
4542
0.
8760
0.
3808
1.
4890
(
0
.
6113,
3.
626
9)
E
DUCRE
C2 12
0.
3281
0.
4459
0.
7360
0.
4619
1.
3883
(
0
.
5793,
3.
626
9)
E
DUCRE
C2 13
0.
6130
0.
4542
1.
3500
0.
1771
1.
8460
(
0
.
7579,
4.
496
3)
E
DUCRE
C2 14
0.
5352
0.
4216
1.
2700
0.
2043
1.
7077
(
0
.
7475,
3.
901
6)
E
DUCRE
C2 15
0.
5587
0.
4381
1.
2750
0.
2022
1.
7484
(
0
.
7409,
4.
126
0)
E
DUCRE
C2 16
0.
5286
0.
4223
1.
2520
0.
2107
1.
6965
(
0
.
7415,
3.
881
5)
E
DUCRE
C2 17
0.
5355
0.
4265
1.
2550
0.
2093
1.
7083
(
0
.
7404,
3.
941
2)
E
DUCRE
C2 18
0.
7646
0.
4351
1.
7570
0.
0789
2.
1481
(
0
.
9156,
5.
039
9)
E
DUCRE
C2 19
0.
5300
0.
4232
1.
2520
0.
2104
1.
6990
(
0
.
7412,
3.
894
1)
E
DUCRE
C2 20
0.
5985
0.
4289
1.
3950
0.
1629
1.
8194
(
0
.
7850,
4.
217
3)
E
DUCRE
C2 21
0.
4005
0.
4684
0.
8550
0.
3926
1.
4926
(
0
.
5959,
3.
738
2)
E
DUCRE
C2 22
0.
4012
0.
4607
0.
8710
0.
3838
1.
4937
(
0
.
6055,
3.
684
6)
NONCIG
EV
(Y
ES)
-0.1227
0.0573
-2.
1420
0.0322
0.8846
(0.7907, 0.989
6)
SM
KLSE
V (
Y
E
S
)
0.
0435
0.
0717
0.
6070
0.
5437
1.
0445
(
0
.
9076,
1.
202
1)
YE
AR 2
0.
0315
0.
0585
0.
5390
0.
5899
1.
0320
(
0
.
9203,
1.
157
4)
YE
AR 3
-
0
.
0131
0.
0566
-
0
.
2320
0.
8166
0.
9870
(
0
.
8833,
1.
102
8)
4.
DIS
C
USSI
ON
Th
is stud
y adop
ted
a
p
s
eudo
-p
an
el
data anal
ysis approac
h
to estim
a
t
e the effect of NCT
and
ST use
and
ot
he
r s
o
ci
o
-
dem
ogra
p
hi
c vari
a
b
l
e
s o
n
t
h
e di
ag
no
si
s of
hy
pe
rt
ensi
o
n
i
n
co
h
o
rt
s
defi
n
e
d by
y
ear
of
b
i
rt
h i
n
a th
r
ee-
year
per
i
od
, 201
2-
201
4. Th
e f
i
nd
ing
of
th
is stu
d
y
sh
ow
low
e
r
od
d
s
o
f
h
y
p
e
r
t
en
sion
d
i
ag
no
si
s w
ith
age am
ong
wh
o res
p
on
de
d ‘y
es’ t
o
t
h
e
use
of
NC
T com
p
a
r
ed t
o
t
h
ose w
ho
res
p
o
n
d
ed
‘
n
o
’
. F
r
i
e
dm
an et
al
.
(1
9
8
2
)
[
24]
re
p
o
rt
e
d
n
o
n
-sm
okers
ha
d a hi
gh
er m
ean di
ast
o
l
i
c
bl
oo
d
press
u
re (
1
m
m
Hg hi
ghe
r) t
h
an sm
oker
s
,
rega
rdl
e
ss
of g
e
nde
r. A
not
a
b
l
e
l
i
m
i
t
a
t
i
on of
t
h
i
s
st
udy
i
s
t
h
at
t
h
e charact
er
i
s
t
i
c
s of t
h
e ref
e
rence
gr
ou
p (t
hos
e
wh
o d
o
n
o
t
us
e ST and NC
T
,
bo
dy
m
a
ss i
ndex [B
M
I
]
st
atus) are
not
k
n
o
w
n
an
d m
a
y
have i
n
cl
ude
d
peo
p
l
e
wh
o use
ci
ga
re
t
t
e
s
or
w
h
o
we
re obese
. A
c
o
m
p
ari
s
on bet
w
een no
n
-
sm
oke
rs, sm
okers
a
n
d fo
rm
er
sm
okers
i
n
the Czech Re
public by
Pankova et
al (2015) [13]
showed no differe
n
ce in the
dia
g
nosis
of
hypertension
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S I
S
SN
:
225
2-8
8
0
6
Asso
ci
a
tio
n b
e
t
w
een
Non
Ci
ga
rette/S
mo
keless To
ba
cco
a
n
d
Hyp
e
rten
sion in
th
e .... (
O
lawu
nmi Ob
isesan
)
67
des
p
ite an adjustm
e
nt for age
and BM
I. Th
e
au
tho
r
s, ho
w
e
ver
,
fo
und
th
e diag
no
sis of
h
y
p
e
r
t
en
sion
w
a
s
m
o
r
e
l
i
k
el
y
i
n
non
-s
m
okers com
p
ared t
o
f
o
rm
er
sm
oker. A
not
h
e
r l
i
m
i
t
a
t
i
on i
s
we di
d n
o
t
ad
ju
st
fo
r al
l
pos
si
bl
e
con
f
ou
n
d
ers t
h
at
coul
d a
ffect
t
h
e rel
a
t
i
o
nshi
p
bet
w
ee
n
NC
T, ST
, a
n
d
t
h
e
di
ag
no
si
s o
f
hy
pert
e
n
si
o
n
, i
n
c
l
udi
n
g
bo
dy
m
a
ss i
n
d
e
x
(B
M
I
)
.
F
u
t
u
re
st
u
d
i
e
s s
h
o
u
l
d
i
nve
st
i
g
ate p
a
tien
t
s’
k
nowledg
e
o
f
th
e
d
e
leteriou
s effects of
no
n
-
ci
garet
t
e
t
obacc
o
o
n
ca
rd
i
ovasc
ul
ar
di
se
ase ri
s
k
.
5.
CO
NCL
USI
O
N
Gi
ve
n t
h
e s
odi
um
cont
ent
of
no
n
-
ci
garet
t
e
t
obacc
o
pr
od
uc
t
s
, rega
rdl
e
ss
o
f
l
o
we
r o
d
d
s o
f
de
vel
o
pi
n
g
h
y
p
e
rten
sion
, a larg
e nu
m
b
er o
f
ag
ed
ind
i
v
i
du
als wh
o
u
s
e th
ese produ
cts sh
ou
l
d
still b
e
co
nsid
ered
at risk
, i
n
som
e
form
, for
devel
o
pi
n
g
hy
pert
e
n
si
o
n
or
p
o
o
r
c
ont
r
o
l
of t
h
ei
r
bl
o
od
p
r
es
sure
. I
n
sc
reeni
ng
f
o
r t
obacc
o
use,
healthcare
provide
r
s
s
h
oul
d screen
for
the
use of bot
h
ci
garette
a
n
d non-ci
garette
tobacco products
in
their
p
a
tien
t
s as
wel
l
as in
itiate in
div
i
d
u
a
lized
h
e
alth
edu
cation
messag
e
s th
at t
a
rg
et t
o
b
acco
cessatio
n
i
n
all fo
rm
s.
REFERE
NC
ES
[1]
Pia
no MR.
,
Be
nowitz
NL
.,
Fi
tzGerald GA.
, Corbridge S.,
H
e
a
t
h
J
., H
a
hn
E.
, e
t
a
l
.,
“Impact of Smokeless Tobacco
Products on Car
d
iovascular Disease: Im
plication
s
for Policy
,
Prevention
,
an
d
Tr
eatment: A Polic
y
Statement Fro
m
the Am
erican Heart As
s
o
ciat
ion
”
,
Circulation
[Internet]
,
Ovid Technolog
ies (
W
olters Kluwer Health)
, vol/issue
:
122(15), pp
. 152
0–44, 2010
. Available from: http
://dx.doi.org/10
.
1
161/cir
.
0b013e3
181f432c3.
[2]
Am
erican Diabe
t
es
As
s
o
ciation,
“
C
ardiovas
c
ular
dis
eas
e and ris
k
m
a
nagem
e
nt”
,
Diabe
te
s Care
,
vol. 38, pp. S49-
S57, 2015. Available from: http:
//dx.doi.org
/doi: 1
0
.2337/dc15-S0
11.
[3]
Centers for
Disease Con
t
rol
an
d Prevent
i
on, “
E
conom
ic
f
a
c
t
s about
U.S.
tob
acco
produc
tion
and use
”
, 201
5.
Availab
l
e f
r
om:
http://www.cdc.
gov/tobacco/d
ata_statistics/f
act_
s
heets/economics/ econ_f
acts/.
[4]
Centers for Disease Control
and Prevention
,
“High
blood
pressure fact sheet”, 2015
. Available fro
m:
http://www.cdc.gov/dhdsp/data_
s
tatis
tics/fact_sh
eets/fs_bloodpr
essure.h
tm. Accessed November 1
5
2015.
[5]
American Heart Associati
on, “Tobacco and blood pressu
re”, 2015. Available from:
http://www.heart.org/HEARTOR
G
/Conditions/HighBloodP
ressure/Preven
tionTr
e
atme
ntofHighBloodPressure/Tob
acco-
a
nd-Blood-
Pressure_UCM_
301886_Articl
e
.j
sp#.
[6]
G
upta R., G
u
r
m
H
., Bartholo
m
ew
J
R
., “
S
m
o
ke
less Tobacco
and Card
iovas
c
ular Risk”,
Ar
chives of Intern
al
Medicine
[
I
n
te
rn
e
t
],
American
Medica
l Associa
tion (
A
MA)
, vol/issue: 164(17),
pp. 1845, 2004
. Availab
l
e from:
http://dx.doi.org/10.1001/
ar
chinte.164.17
.1845.
[7]
O’Connor RJ., “
N
on-cigare
tte
to
bacco produ
cts:
what have
we
le
arnt and wher
e a
r
e we head
ed
?
T
obacco Con
t
rol”
[Internet]
.
BMJ
,
vol/issue: 21(2)
,
pp. 181–90, 201
2. Avail
a
bl
e fro
m
:
http://dx
.
doi
.org/10.1136/
tob
acco
control-201
1-
050281.
[8]
University
of
North Carolin
a, “Non-cigarette toba
cco produ
cts and POS policie
s”, 2014
.
Availab
l
e from:
http:/
/count
ertob
acco
.org/non-
cig
a
ret
t
e-tob
a
c
c
o-p
r
oducts-and-pos-
polici
e
s.
[9]
Asplund K., “Smokeless tobacco
and card
i
ovas
c
ular d
i
sease”,
Progress in cardiovascular diseases
[Internet]
.
Elsev
i
er B
V
,
vol/
i
ssue: 45(5), pp.
383–94, 2003
. A
v
ail
a
ble
from
:
ht
tp://dx
.
doi
.org/1
0.1053/pcad.200
3.00102.
[10]
Benowitz NL.
,
“
S
odium Intake
from
Sm
okeles
s
Tobacco
”,
N
Engl J Med
[In
ternet]
.
New England Journal of
Medicine (
N
EJM/MMS)
,
vol/issue: 319(13), pp. 87
3–4, 1988. Availab
l
e from:
http://dx.doi.org
/10.1056/nejm198809293191318.
[11]
Valeriano J., Tu
cker P., Kattah
J., “An unus
ual cause Ofh
y
pok
alemic muscle
weakness”,
N
e
u
r
ology
[Intern
et]
.
Ovid Technolo
g
ies (
W
olters Kluwer Health)
,
vol/issue: 33
(9), pp. 1242–
1242, 1983.
Availab
l
e from:
http://dx.doi.org/10.1212/wnl.33
.9.1242.
[12]
Eliasson M., Asplund K., Nasic
S., Rodu
B., “Influence of smok
ing and snus
on the prevalen
ce
and incid
e
nce of
ty
p
e
2
diab
etes
amongst men: the north
ern Sw
eden MONICA
stud
y
”
,
Journal of In
ternal Medicin
e
[Internet]
,
Wiley
-
Blackwell, vol/issue: 2
56(
2), pp. 101
–10, 2004. Available from:
http://dx.do
i.o
r
g/10.1111/j.136
5-
2796.2004.0134
4.
[13]
Pankova A., Kr
alikova
E., Fraser
K., La
jk
a J., Sv
acin
a
S., M
a
toulek M., “No
difference
in h
y
per
t
ension prevalence
in smokers, former smokers
and non-smokers after
adjusting fo
r bod
y
mass
index and
age: a
cr
oss-sectional stu
d
y
from
the Czech
Republic
, 201
0”,
T
obacco In
duced Dis
e
as
es
[Internet]
,
Springer Sc
ie
nc
e
+ Busine
ss Me
dia
,
vol/issue: 13(1), 2015.
Availab
l
e from:
http://dx.d
o
i.org/10
.1186/s
12971-015-0049-4.
[14]
S
i
egel D.
, Beno
witz N.,
Erns
ter
VL., Grad
y D
G
., Hauck W
W
.
, “
S
m
okeles
s
tobacco
, c
a
rdiovas
c
ular r
i
s
k
factor
s
,
and nicotine
an
d cotinin
e levels
in
professional
base
ball play
ers”,
American Jou
r
nal of Public
Health
[Internet]
,
American Public Health
Association, v
o
l/issue: 82(3)
, pp. 417–2
1, 1992. Available from:
http://dx.doi.org/10.2105/ajph.82
.3.417.
[15]
Benowitz NL., “
D
aily
Use of Smoke
le
ss Tobac
c
o
: S
y
ste
m
ic
Effects”
,
Ann In
tern
Med
[Intern
et]
,
Am
erican Coll
e
g
e
of Ph
y
s
i
c
ians, v
o
l/issue: 1
11(2),
pp. 112
, 1989
. A
v
ail
a
ble from
:
ht
tp://dx
.
doi
.org/1
0.7326/0003-48
19-111-2-112.
[16]
Fernandez De
Bobadilla J., G
a
rrido P., Lóp
e
z De Sá
E.,
Sanz-De-Burgoa
V., “Cardiov
ascular r
i
sk factors in
sm
okers v
s
non-
sm
okers in a pri
m
ar
y
car
e settin
g: retratos stud
y
:
PP.20.305”,
Jo
urnal of Hypertension
[Internet]
,
Ovid Technologies (
W
olters Kluwer Health)
,
vol. 28, pp. 1, 2010. Availab
l
e from:
http://dx.doi.org/10.
1097/01.h
j
h.0
000379231.3948
2.f8.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
252
-88
06
I
J
PH
S Vo
l. 5
,
N
o
. 1
,
Mar
c
h
20
16
:
60
–
69
68
[17]
Hergens MP., Lambe M., Persh
a
gen G.
, Ye W., “Risk of h
y
p
e
rtension
am
ongst Swedish male snuff users:
a
prospectiv
e study
”,
Journal of In
ternal Medicine
[Internet]
,
Wiley
-
Blackwell,
vol/issue: 264(2), pp. 187–94, 2008.
Availab
l
e f
r
om:
http://dx.
doi.org/10.1111/j.1365-
2796.2008.0193
9.x
[18]
Bullock L. AHA., “Smokeless T
obacco Products Are Unsafe”,
Internal Med
i
cine News
[Intern
et]
,
Elsevier BV
,
vol/issue:
43(16)
, pp
. 30
, 2010
. A
v
ail
a
ble from
:
ht
tp://dx
.
doi
.org/1
0.1016/s1097-86
90(10)70838-6.
[19]
England LJ
.
,
Ki
m
SY., Tom
a
r S
L
., R
a
y
CS
.
,
Gupta P
C
., E
i
s
s
e
nberg T.
, et
al.
,
“
N
on-cigare
tte
to
bacco us
e am
on
g
women and adv
e
rse pregnancy
outcomes”,
Acta Obstet Gynecol Scand
[Inter
n
et]
,
W
ile
y
-
B
l
a
c
kwell
,
vol/issu
e:
89(4), pp
. 454–6
4, 2010
. Avai
lab
l
e from
:
http
://d
x.doi.org
/10.310
9/00016341003605719.
[20]
Weber MA., Sch
i
ffrin EL., White WB., Ma
nn S.,
Lindholm LH.,
Kenerson JG., et
al., “Clin
ical Pr
actice Guidelines
for the M
a
nage
m
e
nt of Hy
p
e
rte
n
s
i
on in the Com
m
unity
A S
t
at
em
ent b
y
th
e Am
erican S
o
cie
t
y
of Hy
p
e
rt
ens
i
on an
d
the In
terna
tion
a
l
Socie
t
y of H
y
p
e
rtension”
,
Journal of Hypertension
[Intern
et]
,
Ovid Techno
logies
(
W
olters Kluwer
Health)
,
vol/
i
ssue: 32(1)
, pp
. 3–1
5, 2014
. Avai
lab
l
e from
:
http
://d
x.doi.org
/10.109
7/hjh.000000000
0000065.
[21]
Deaton A., “
P
a
n
el dat
a
from
tim
e s
e
ries
of cros
s
-
s
ections
”,
Journal of Econ
ometrics
[Intern
et]
,
Elsevier BV,
vol/issue:
30(1-2
)
, pp
. 109–26
, 1
985. Availab
l
e
fr
om: http://dx.do
i.org/10.1016
/03
04-4076(85)90134-4.
[22]
Molenberghs, G., Verb
eke, G., “
M
odels for dis
c
r
e
te longitudinal
data”, Springer
,
New York, 2005
.
[23]
Liu Q.
, Pier
ce
DA., “
A
note o
n
Gauss—Hermite qu
adra
ture”
,
Biometr
i
ka
[In
ternet]
,
Oxford University
Press
(OUP), vol/issue: 81(3), pp. 624–
9, 1994
. Avai
lab
l
e from
:
http
://d
x.doi.org
/10.109
3/biom
et/81
.
3.6
24.
[24]
Friedman GD.,
Klatsk
y
AL., Siegelaub AB
., “Alcohol, tobacco, and
h
y
per
t
ens
i
on”,
Hy
pe
rte
n
sion
[Intern
et]
,
Ov
i
d
Technologies (Wolters Kluwer Health)
,
vol/issue: 4(5_Pt_
2), pp. I
II143
–III143, 1982. Availab
l
e fro
m:
http:/
/dx.doi
.org/
10.1161/01.h
y
p
.
4.5_pt_2.iii143
.
BIOGRAP
HI
ES OF
AUTH
ORS
Olawunm
i Obises
an is
an adju
nct facu
lt
y in t
h
e Departm
e
nt
of S
o
cial W
o
rk at S
outheas
t
Missouri Sta
t
e
Unive
r
sity
, Missouri,
USA.
Olaw
unmi holds several degr
ees in
cluding
a PhD in
Public Health (
E
pidemiolog
y
)
f
r
om Walden Un
ivers
i
t
y
.
S
h
e i
s
currentl
y
com
p
leting
anoth
e
r
doctora
l progra
m
in Health
Ed
ucat
ion from
A.T.
Still
Univer
sit
y
and
an MPH program
in
Prevention Sci
e
nce from
Em
ory
Universit
y
, R
o
llins School of
Public Health.
Olawunm
i has
worked with sev
e
ral hospi
ta
ls, an
d is curren
t
l
y
th
e Coordina
tor fo
r the R
e
giona
l A
r
thritis C
e
nter
(RAC) for 16 counties in Southeast Missour
i. Under th
e R
A
C program, Olawunmi has
im
plem
ented s
e
veral CDC-fund
ed evid
enc
e
-bas
ed
programs for
the State of
Missouri, in
cluding
the S
t
anford Un
ivers
i
t
y
’s
Chron
i
c Dis
eas
e
S
e
lf-
M
anagem
ent P
r
ogram
(CDS
MP). Olawunm
i’s
research
in
terests includ
e issues
that aff
e
ct
th
e health of
vuln
e
rabl
e populatio
ns, especially
i
mmi
gra
n
t
s
.
Emmanuel Tho
m
pson is an Assistant Professor of St
atisti
cs/Act
uaria
l
Scien
ce in
the Depar
t
m
e
nt
o
f
M
a
t
h
e
m
a
t
i
c
s
a
t
S
o
u
t
h
e
a
s
t
M
i
s
s
o
u
r
i
S
t
a
t
e
U
n
i
v
e
r
s
i
t
y
,
M
i
s
s
o
u
r
i
,
U
S
A
.
H
e
is
a
G
r
a
d
u
a
t
e
Statisti
cian m
e
m
b
er of the Am
erican Statist
i
c
a
l
Association and a
l
so an Affiliate
m
e
m
b
er of the
Institute
and Facult
y
of Actu
aries
in the Unit
ed K
i
ngdom
. Em
m
a
nuel holds a Ph.D
. in Statisti
cs
(Actuari
al S
c
i
e
nce option
)
an
d a M
S
c. in
Actuari
a
l S
c
i
e
n
ce both from
Univers
i
t
y
of
Calgar
y
in
Can
a
da.
He
also gr
a
duated
from
Univ
ers
i
t
y
of Cap
e
Coas
t
in Ghan
a
,
with
a
BS
c.
(First Class) in Statistics and h
a
s spent over
six
y
e
ars working in the insuran
ce industr
y
in
Ghana and briefl
y in Canad
a
. E
m
m
a
nuel’s
res
e
arch agend
a
covers
the areas
of ris
k
theor
y
& actuar
i
al
modeling, applied statistics
,
and applied econometrics. Emmanuel ha
s
published
and
presented d
i
ff
erent peer
rev
i
ew
ed papers
in
Applied Sta
tisti
cs and Ac
tuari
a
l
Scienc
e
at
the
na
tional
and
in
tern
ation
a
l
lev
e
ls
Olubusay
oAkin
o
la is a drug dema
nd reduction expert for the Eu
ropean Union Support Project to
Econom
ic Com
m
unit
y
of West
African States
(
E
COWAS) Regi
onal Act
i
on Plan on illi
cit drug
trafficking, related organ
i
zed
cr
ime and drug
abu
s
e
i
n
We
st
Africa
.
She i
s
a pha
rma
c
i
st
,
who
a
l
so
has a PhD in
Public Hea
lth
with speci
ali
z
a
t
i
on in Epid
em
io
log
y
. Her
lev
e
l
of work-based
experience in th
e region incr
eas
ed her passi
on for examining harm
minimization
among drug-
using population
s
through ef
fectiv
e public health
interv
entions.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PH
S I
S
SN
:
225
2-8
8
0
6
Asso
ci
a
tio
n b
e
t
w
een
Non
Ci
ga
rette/S
mo
keless To
ba
cco
a
n
d
Hyp
e
rten
sion in
th
e .... (
O
lawu
nmi Ob
isesan
)
69
Adekunle Obisesan is an Ad
ult Internal Medi
cin
e
/Critical
Care Hospitalis
t ph
y
s
ician at
Southeast Hospital, Cap
e
Girard
eau, Missouri.
Adekunle also h
o
lds a MPH degree from Johns
Hopkins Bloomberg School of
Public Health.A
dekunle’s work
experien
ces include workingas
the Integr
ated M
a
nagement of C
h
ildhood Illness
e
s and Immuniz
a
tions Program Officer with the
United Stat
es Agency
for Int
e
rn
ation
a
l Develo
p
m
ent’s Basic Support for Institu
t
i
onali
zing Ch
ild
Survival (BASICS) project. He has extensiv
e exper
i
ence d
e
velop
i
ng, implementing and
evalu
a
ting in
ter
v
entions and treatment protoco
l
s
for the prevention and management of various
childhood dise
a
s
es at the prim
ar
y
car
e lev
e
l
.
Ad
ekunle is als
o
anAdjunct Cli
n
ica
l
Precep
tor
(Medica
l
stude
nts’ educa
tion)
at A.T
.
Still
Universit
y
Kirk
sville Col
l
ege
of Osteopathi
c
Medicin
e
.
Yvonne Commo
dore-Mensah is an A
ssistant Research Professor in the Department of Nursing
at
Emor
y
Univ
ersity
, Nell Hod
g
son Woodruff
School of Nursing. Yvonne holds a Ph.D. in
Nursing from
Jo
hns Hopkins
Un
iversity
and BSN(
Honors) fro
m Fairleigh Dick
inson University
.
Yvonne’s research inter
e
st is primarily
ca
r
d
iovascular risk reduction in
ethnic minority
populations-p
articularly
thos
e of
African d
e
scen
t.
Evaluation Warning : The document was created with Spire.PDF for Python.