127
maori
tage
issue
mortal
supplied
fitzpatrick
taken
london
lynch
pearce
potentially
britain21
medicine
conclusions
determi
shown
practical
invalids
excluded
trust
previous
reduce
350
access
figures
153
land
fh
101
rothman
address
munity
me
thank
equalities
type
percentiles
commercialization
estimate
smith
2002
context
death
situation
bartley
disparity
nature
there
eur
everyone
intercept
rate
cen
cutoff
importance
provides
identify
conditions
source
measuring
processes
weakened
sequent
striking
occur
bg
education
varying
peter
directly
groupings
marked
dunnell
inter
few
reducing
depended
york
salmond
even
1975
1991
should
inequali
301
involve
456
calculated
gradients
selected
predominantly
widen
ity
272
1980s
interest
blane
poor
1983
foster
kind48
excess
resulting
men
sci
drives
decedent
pb
coefficient
137
index
100th
sult
latter
708
known
323
differences
191
visits
756
position
rent
lowest
person
personal
correspondence
reporting
aetiology
numerator
straightforward
table
denomi
propor
companying
philadelphia
fisher
232
result
142
mation
codes
intractable
references
findings
real
set
house
country
divisions
449
need
segi
could
dc
rates
per
process
whereas
families
vation
certificate
research
infor
survey
analyses11
working
425
difficult
both
statistics
explanations43
1950
neil
power
occupations
period
although
male
important
rev
explanations
comparison
all
periods
british
1584
904
648
inequality
jw
would
no
systems
european
gdp
ratios
total
educa
js
1459
1984
support
181
existence
154
psychosocial
itself
mapping
1957
third
mcmichael
profound
mortality
disease
black
agree
digit
ries
solely
budget
planning
467
474
1976
cluding
during
dr
academic
229
ministry
self
elsewhere
methods
222
recently
preventable
rather
revised
unfair
age
281
158
decreasing
classes
higher
draft
environment
terpretation
711
132
capital
results
bd
educational
key
537
valid
prime
pared
relative
inverse
appear
magnitude
generating
area
auckland
forms
assets
demonstrating
284
necessary
165
729
classifiable
funding
extent
147
millennium
measures
pooled
slope
they
johnston
investigated
marmot
davis
validation
whatever
individual
457
aj
spec
represented
according
previously
greenland
editors
disadvan
predicted
505
showing
pationally
625
vol
nzseia
promotion
763
smoking
jobs
factors
expected
submitted
655
1998
summarised
classification
never
mail
studies
105
effects
suggestions
559
median
life
larly
system
productive
socioeco
irving
economically
throw
kt
widening
veloped
model
129
comments
environ
national
appears
commentary
unhealthy
policy
indi
stand
receives
munication
145
risk
tohoku
124
implications
1974
care
understanding
centile
kj
percentile
1972
shows
723
vidual
measured
presented
cates1
study
595
488
affect
fundamental
835
accompanying
populations
standards
family
113
case
account
deaths
cryer
309
smedley
disad
about
sites
1987
sociol
jp
lifestyle
beyond
together
justment
1968
increase
score
568
soc
catego
916
008
population17
329
own
inequalityb
2001
published
0600
security
improved
possible
aff
803
investigators
copeland
sary
varied
past
cost
unemployment
returns
ethnic
britain
agricultural
examine
random
proportions
1982
certifi
excluding
under
contrast
498
justification
our
distal
differential
earlier
its
nants
med
neo
263
group
eu
wel
accepted
relatively
attempt
methodological
hierarchy
hadden
permit
records
lington
theory
accuracy
olds
funded
long
757
particular
classified
artefactual
than
force
just
measure
effective
department
850
category
seems
tion
strength
899
0th
fax
2000
borgan
nzdep91
methodology
emphatic
caution
comparable
depart
stud
ple
might
record
order
kiro
280
tend
data
vestigated
labour
131
corresponding
council
discussion
sendai
mackenbach
back
555
gradient
ae
employment
1960
second
inequalities
1997
allocated
skill
repeat
423
interven
furthermore
reported
174
child
work
strongly
indicator
behavioural
coding
construct
pational
new
conducted
centre
entirely
service
whether
downstream
pro
three
states
reference
classifications
inaccurate
was
moting
communication
two
examples
each
neric
923
wish
148
start
assess
attention
raven
differ
japan
responding
204
depri
generated
massey
proportional
1990
countries
generally
howard
228
perpetuated
describe
changed
strengthened
macintyre
realised
close
belonging
available
238
consistent
drawn
however
regards
compared
determinant
consequently
one
trend
alderson
find
sought
any
499
more
material
year
patterns
141
groups
kaplan
persistence
among
nzsei
report
450
class
detailed
most
analyse
correlated
worst
had
dataset
information
diminished
chronic
adelstein
proportion
reasons
measurement
private
light
interpretations
451
egories
validity
undiminished
again
of
relevant
still
objective
ga
lim
will
much
part
tively
average
iii
337
paradigm
982
experiment
fact
ings
theless
weighting
theories
future
quite
cause
classify
very
only
status
requiring
problem
kawachi
aust
structural
fox
unclassifiable
ransom
grouping
analyses
reasonable
producing
lished
strong
chapman
epidemiol
into
107
438
pos
students
journal
ratio
concern
campus
particu
retired
then
013
stratifi
probably
drift
associated
direct
205
modify
farmers
increasing
goldblatt
come
less
midpoint
1993
due
argued
regression
volunteering
percentage
analysis
cur
household
analyzing
151
1977
increased
slopea
six
illsley
paper
mcdowall
138
patterning
notes
almost
decline
thus
restricted
similar
1995
overview
srr
nations
differing
areas
described
same
indicate
ardised
science
rich
merely
mr
make
expectancy
90s
epidemiology
ited
mg
agenda
occu
younger
individuals
affects
doubled
proximate
023
applied
some
here
lancet
167
cat
graham
biases
non
658
274
differentials
against
ensure
877
aged
indicators
250
occupation
614
incomes
validating
society
separate
official
sible
focus
rg
basis
article
wellington
simi
link
between
convenient
likely
employed
ency
clin
easton
female
proposes
illustrated
347
levels
strengths
com
materialistic
320
patrick
communi
updates
organisations
presumably
publication
women
385
971
entials
am
older
public
479
fully
diseases
why
product
ties
children
598
racial
problems
377
socioeconomic
phelan
lower
1072
must
specific
pc
characteristics
illness
accounts
conceptual
using
package
657
males
since
components
338
1996
ing
occurred
categorisation
involved
united
623
other
socio
significance
328
122
term
except
confined
university
sets
literature
afflictions
startling
view
lost
313
tre
limited
manuscript
397
within
obtained
459
disparities
771
concluded
ferent
186
washington
1986
generate
choice
insufficient
materialist
inadequate
well
zealand
tobias
recent
because
skills
potential
455
engl
nator
datasets
variance
sources
level
980
followed
wb
occurring
industrialised
ble
linked
youngest
173
gra
061
parisons
scales
divided
indicating
wider
156
neces
analyses7
complicated
world
decrease
greatest
864
slightly
equal
certificates
ne
959
nzsco
give
scores
unique
437
them
used
figure
1989
being
considerable
december
press
examined
when
572
various
tality
attempts
derived
rela
argue
challenge
publique
1999
costa
may
zea
different
broader
wilkinson
ii
terms
exclusion
requested
2nd
previously1
census
grouped
1992
ah
debate
andrew
244
further
income
regard
change
lev
379
november
int
number
developing
nevertheless
summary
error
societies
causes
sizes
now
al41
population
237
standardised
step
sa
100
380
neglect
ported
bias
revision
considerations
their
scale
therefore
excludes
elley
vantage
interpretation
possi
denote
point
hand
carried
behav
williams
also
distribution
000
related
cancer
but
too
small
over
denominator
found
largely
association
197
albeit
instead
mcleod
analy
routledge
1980
assume
statistical
uni
former
repeated
nomic
behavioral
fund
sante
tony
unlikely
evidence
low
126
jc
across
through
while
europe
stronger
paul
caused
clear
australian
take
sporle
sutton
based
1200
educ
developed
purposes
cigarette
peo
nationwide
linear
martikainen
school
1994
rivation
estab
tions
fell
1815
once
589
blakely
estimation
persisting
215
greater
such
pappas
150
245
versity
mobility
persistent
apply
occupational
current
decreased
discussed
exception
decades
intervention
see
borman
assessed
analysed
perceptions
economic
time
ment
ind
version
429
health
409
son
human
489
framework
comparing
mckinlay
separately
interpreted
general
104
mechanisms
five
lippincott
apparent
christchurch
international
mor
represent
housing
has
upstream
academy
out
unemployed
reason
103
marshall
falls
howden
consist
regressed
production
1219
categories
inactive
membership
strategies
social
woodward
helped
balance
modern
grant
consequence
words
dif
dep
trends
syme
use
markedly
unlike
despite
497
paid
169
942
misclassification
queen
190
instance
kunst
davey
weighted
504
box
planations
forces
indices
1966
existing
8714
792
rep
els
march
dient
hewlett
yielded
cludes
anything
oldest
observed
noted
sl
growth
interventions
highest
established
jk
643
checkoway
suggestion
not
meta
years
019
po
practitioner
crampton
avoidable
remain
srrs
range
carpenter
cial
selection
133
123
304
groenhof
particularly
effect
perhaps
resented
declined
cation
times
standard
workers
perspective
166
community
valkonen
403
kelsey
ongley
persons
abstract
major
issues
primary
place
acknowledgements
high
availability
above
first
combined
ses
coded
overall
sub
1985