Erythrocyte sedimentation rate: a possible marker of atherosclerosis and a strong predictor of coronary heart disease mortality

opossumoozeMechanics

Feb 21, 2014 (3 years and 5 months ago)

111 views

European Heart Journal (2000) 21,1614–1620
doi:10.1053/euhj.2000.2148,available online at http://www.idealibrary.com on
Erythrocyte sedimentation rate:a possible marker of
atherosclerosis and a strong predictor of coronary
heart disease mortality
G.Erikssen
1
,K.Liestøl
2
,J.V.Bjørnholt
3
,H.Stormorken
4
,E.Thaulow
3
and
J.Erikssen
1
1
Medical Department,Central Hospital of Akershus,Nordbyhagen;
2
Institute for Informatics,University of Oslo;
3
Medical Department,Section of Cardiology,National Hospital,University of Oslo;
4
Institute for Medical
Research,National Hospital,University of Oslo,Norway
Aims Since atherosclerosis is a chronic inflammation and
the erythrocyte sedimentation rate is an appropriate test for
monitoring chronic inflammatory responses,we wanted
to investigate whether the erythrocyte sedimentation
rate might carry prognostic information on the risk of
sustaining coronary heart disease events.
Method The erythrocyte sedimentation rate was deter-
mined in 2014 apparently healthy men aged 40–60 years
during an extensive cardiovascular survey in 1972–75,and
the test was repeated in an identical follow-up examination
7 years later.Cause-specific mortality and rates of non-fatal
myocardial infarction were followed for 23 years.
Results The erythrocyte sedimentation rate was strongly
correlated with age,haemoglobin level,smoking status,
total cholesterol level and systolic blood pressure.After
adjusting for all these associations in multivariate Cox
regression analyses,the erythrocyte sedimentation rate
emerged as a strong short- and long-term predictor of
coronary heart disease mortality,particularly in men
who had developed angina pectoris and/or had a positive
exercise ECG test at the second survey.Increases in
non-coronary heart disease deaths and in non-fatal myo-
cardial infarctions were only seen in the upper erythrocyte
sedimentation rate range.
Conclusions The erythrocyte sedimentation rate is a
strong predictor of coronary heart disease mortality,and
appears to be a marker of aggressive forms of coronary
heart disease.The erythrocyte sedimentation rate probably
gives substantial information in addition to that given by
fibrinogen on the risk of coronary heart disease death.
(Eur Heart J 2000;21:1614–1620,doi:10.1053/euhj.2000.
2148)
￿2000 The European Society of Cardiology
Key Words:Erythrocyte sedimentation rate,atheroscler-
osis,coronary heart disease mortality,non-fatal myocardial
infarction,prediction.
See page 1567 for the Editorial comment on this article
Introduction
Regardless of initiating causes atherosclerosis is
basically chronic inflammation in the arterial wall
[1]
.
This has prompted a search for inflammatory markers
as indicators of coronary heart disease risk,and,for
example,fibrinogen and C-reactive protein have gained
some popularity in this regard
[2,3]
.In 1988 the Inter-
national Committee for Standardization in Haematol-
ogy suggested that the erythrocyte sedimentation rate
might be a particularly robust and appropriate test for
monitoring chronic inflammatory processes since it is
sensitive to both fibrinogen and immunoglobulins
[4]
.
Accordingly,a fast and inexpensive erythrocyte
sedimentation rate test might be a valuable tool for
assessing the strength of the inflammatory response
associated with atherosclerosis,and might conceiv-
ably carry important short- and long-term prog-
nostic information.Earlier we reported an association
between increased erythrocyte sedimentation rate and
progression of angiographically documented coronary
heart disease
[5]
.The aim of the present study was to
assess whether a high erythrocyte sedimentation rate
might be associated with an increased risk of sustain-
ing fatal coronary heart disease events,as judged from
Revision submitted 1 February 2000,and accepted 2 February
2000.
Correspondence:Gunnar Erikssen,MD,Medical Department,
Central Hospital of Akershus,N-1474 Nordbyhagen,Norway.
0195-668X/00/211614+07 $35.00/0 ￿2000 The European Society of Cardiology
data collected during 23 years of follow-up in 2014
apparently healthy men initially aged 40–60 years.
Material and methods
The baseline survey (Survey 1) conducted in 1972–75 is
described in detail elsewhere
[6]
.The participants were
recruited from five companies in Oslo,Norway.All
apparently healthy men aged 40–60 years in these
companies were invited to participate in the study
(n=2341),and 2014 of them accepted the invitation
(86%).‘Apparently healthy’ implied absence of recog-
nized heart diseases,hypertension,diabetes mellitus,
cancer and a number of other serious diseases.None
were using drugs regularly.A prerequisite for inclusion
in the study was absence of febrile illnesses or surgery
during the previous 2 weeks,and all were expected
to conduct a symptom-limited exercise ECG test —
although a positive test result was not among the
exclusion criteria.Among survivors at 31 December
1982,1756 (91%) participated in a virtually identical
examination in 1980–82 (Survey 2).At this stage 1428/
1756 men remained healthy according to the baseline
inclusion criteria
[7]
,and this cohort was used to check
the results based on the Survey 1 data and for assessing
the possible significance of alterations in the erythrocyte
sedimentation rate.
Examination procedures and definitions
All were asked to abstain fromsmoking and eating for 8
and 12 h,respectively.The examinations took place
between 0730 and 1030h at the National University
Hospital of Oslo,Norway,and included a comprehen-
sive case history,complete physical examination,height
and weight measurements,chest X-ray,spirographic
tests,a symptom-limited bicycle exercise ECGtest and a
number of fasting blood tests including measurement of
the erythrocyte sedimentation rate using Westergren’s
technique.The participants were classified as current
smokers or non-smokers
[8]
,and as physically active or
not,as reported previously
[9]
.The exercise ECG test
technique and interpretation are described elsewhere
[6]
.
Body mass index was calculated as body weight (kg)
divided by squared height (metres).High-density lipo-
protein cholesterol level,available as a screening pro-
cedure in Norway since 1974,was measured in only 269
men (13%) during the final part of Survey 1,but in 1678
of the 1756 (96%) during Survey 2.High-density lipo-
protein cholesterol,total cholesterol and triglyceride
levels were measured by standardized methods
[10]
.The
fasting blood glucose level was only determined during
Survey 1
[11]
.
Mortality and morbidity data
Permission to study the death records in Statistics
Norway and relevant patient records for all who had
been hospitalized in any Norwegian hospital during
follow-up was given by the Norwegian Board of Health
and the Norwegian Data Inspectorate.Mortality data
are complete as at 31 December 1996,and morbidity
data are updated until 31 December 1994.
Coronary heart disease deaths include sudden
deaths and deaths caused by myocardial infarction.
Cardiovascular deaths also include deaths from stroke
and other arteriosclerotic diseases.The New York
Health Insurance questionnaire was used for establish-
ing the diagnosis of angina pectoris
[12]
,and myocardial
infarctions according to standardized criteria described
earlier
[5]
.The years of first and recurrent myocardial
infarctions were also recorded.
Statistical methods
Associations between the erythrocyte sedimentation rate
and time to death from coronary heart disease and all
causes were studied applying proportional hazards
(Cox) models.Based on preliminary analyses we decided
to use the erythrocyte sedimentation rate mainly as a
continuous variable.
Plots of log S(t) versus log (t) show acceptable
compliance with the proportional hazards assumption.
Results are presented as relative risks (RR).For con-
tinuous variables,the RRs associated with increases of
2 SDs are presented.
Standard mortality ratios were computed using as
reference the 1990 mortality rates reported by Statistics
Norway.Standard Pearson correlation coefficients were
computed between the erythrocyte sedimentation rate
and selected variables.All statistical calculations were
done using the StatView(version 5.0) computer package.
Results
Erythrocyte sedimentation rates were distributed
asymmetrically (Fig.1) with a majority of low values
and a long upper range ‘tail’.Participants were divided
into five subgroups according to erythrocyte sedimen-
tation rate-intervals 0–4,5–9,10–14,15–29 and
￿30 mm.h
￿1
.Significant correlations with the eryth-
rocyte sedimentation rate were found for age,haemo-
globin level,smoking status,total cholesterol level and
systolic blood pressure (Table 1).Weaker but significant
correlations with the erythrocyte sedimentation rate
were found for triglycerides,fasting blood glucose and
physical activity levels.Subjects with low erythrocyte
sedimentation rates had more favourable values for a
number of coronary heart disease risk factors than men
with erythrocyte sedimentation rates in the higher
ranges.
Table 2 shows a steep increase in 23 years’ coronary
heart disease,cardiovascular disease and total mor-
tality with increasing erythrocyte sedimentation rate.
This increase is mainly caused by the coronary heart
disease–erythrocyte sedimentation rate association.For
Erythrocyte sedimentation rate 1615
Eur Heart J,Vol.21,issue 19,October 2000
non-cardiovascular disease and cancer mortality a sub-
stantial increase in mortality is only observed in the
highest range.
Proportional hazards analyses show that the Survey 1
erythrocyte sedimentation rate was a strong predictor of
coronary heart disease mortality after 23 years (Table 3,
columns 1–3),less so for cardiovascular disease
mortality,and non-significant for cancer and
non-cardiovascular disease mortality (details not
shown).The associations in Table 3,columns 1,4 and 5
0
300
ESR
Count
250
200
150
100
50
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
31
32
33
34
35
36
37
38
39
40
42
43
44
45
48
53
95
30
Figure 1 Distribution of values of the erythrocyte sedimentation rate (ESR) measured among 2014
apparently healthy men during Survey 1 in 1972–75.
Table 1 Baseline values (SD) of a number of parameters within subgroups of erythrocyte sedimentation rate (ESR),
and associations between erythrocyte sedimentation rate and these parameters.Data from 2014 men examined during
Survey 1 in 1972–75
Parameter
ESR
0–4
(n=805)
ESR
5–9
(n=745)
ESR
10–14
(n=256)
ESR
15–29
(n=172)
ESR
￿30
(n=36)
All
(n=2014)

ESR (mm.h
￿1
) 2∙9 (0∙9) 6∙5 (1∙3) 11∙6 (1∙4) 19∙1 (3∙7) 38∙9 (11∙2) 7∙4 (6∙7) 1∙00
Age (years) 48∙9 (5∙4) 50∙0 (5∙4) 50∙8 (5∙5) 51∙2 (5∙6) 53∙2 (5∙5) 49∙8 (5∙5) 0∙17***
Haemoglobin (g.l
￿1
) 15∙6 (1∙1) 15∙2 (0∙9) 14∙9 (1∙3) 14∙8 (1∙0) 14∙5 (1∙0) 15∙3 (1∙1) ￿0∙23***
Smokers (%) 38∙0 44∙2 50∙8 57∙0 50∙0 43∙8 0∙10***
Total cholesterol (mmol.l
￿1
) 6∙43 (1∙11) 6∙73 (1∙17) 6∙92 (1∙16) 6∙94 (1∙44) 6∙76 (1∙36) 6∙65 (1∙19) 0∙12***
Triglycerides (mmol.l
￿1
) 1∙27 (0∙67) 1∙32 (0∙76) 1∙40 (0∙64) 1∙43 (0∙79) 1∙27 (0∙56) 1∙32 (0∙71) 0∙06*
Fasting blood glucose (g.l
￿1
)† 79∙8 (9∙4) 80∙4 (10∙0) 80∙5 (9∙2) 81∙0 (14∙4) 81∙8 (8∙8) 80∙3 (10∙1) 0∙05*
Systolic blood pressure (mmHg) 129∙0 (17∙9) 129∙7 (17∙2) 130∙5 (16∙1) 134∙7 (20∙2) 138∙6 (23∙4) 130∙1 (17∙9) 0∙11***
Positive exercise ECG (%)‡ 8∙0 10∙9 13∙3 11∙6 19∙4 10∙2 0∙06
Physically active (%) 16∙8 14∙6 12∙1 9∙9 5∙6 14∙6 ￿0∙06*
Body mass index (kg.m
￿2
) 24∙5 (2∙6) 24∙5 (2∙8) 24∙8 (2∙7) 24∙7 (3∙3) 24∙0 (2∙4) 24∙6 (2∙8) 0∙01
*P<0∙05,***P<0∙0001.
†16 missing values of fasting blood glucose during Survey 1.
‡See Erikssen
[6]
.
§Pearson correlation coefficient.
1616 G.Erikssen et al.
Eur Heart J,Vol.21,issue 19,October 2000
based on Survey 2 data on the 1428 still healthy men at
Survey 2 and 16 years of follow-up are almost identical
with those for columns 1–3.The predictive power of
Survey 1 erythrocyte sedimentation rate remained virtu-
ally unchanged during the whole observation period,
and was strong (Cox RR=2∙00,P<0∙0001,using the
same covariates as in Table 3,columns 1–3) even after
only 7 years (at the time of Survey 2).Changes in
erythrocyte sedimentation rate fromSurvey l to Survey 2
did not add significant information (details not shown).
At Survey 2,121 of the 1756 examined (7%) had
developed angina pectoris with or without a positive
exercise ECG test,and 282 had a positive exercise ECG
test only (‘silent ischaemia’).Table 4 shows a marked
increase in coronary heart disease mortality with
erythrocyte sedimentation rates ￿15 both among men
with angina pectoris and those with silent ischaemia,
although coronary heart disease mortality among
angina patients was highest in all erythrocyte sedimen-
tation rate subgroups.In a Cox analysis including all
the common covariates except exercise ECG findings,
the erythrocyte sedimentation rate was the strongest
predictor of coronary heart disease mortality in the
merged group of 403 men (RR=1∙92,P<0∙0001,details
not shown).
Among men alive at 31 December 1994,the associ-
ation between erythrocyte sedimentation rate and inci-
dence of non-fatal myocardial infarctions was weak,in
contrast to fatal first-time heart attacks (Cox RR=1 90,
P<0∙0001,Table 5).
Discussion
Although markedly correlated with a number of known
coronary heart disease risk-factors,the erythrocyte
Table 2 Total mortality and mortality from various causes after 23 years,associated with different levels of
erythrocyte sedimentation rate (ESR) determined at Survey 1 in 1972–75
ESR
(mm.h
￿1
)
n SMR* Total % CVD % CHD % Cancer % Non-CVD† %
0–4 805 0∙72 210 26∙1 104 12∙9 78 9∙7 66 8∙2 106 13∙2
5–9 745 0∙66 197 26∙4 103 13∙8 88 11∙8 56 7∙5 94 12∙6
10–14 256 0∙73 77 30∙1 35 13∙7 29 11∙3 25 9∙8 42 16∙4
15–29 172 1∙09 73 42∙4 43 25∙0 39 22∙7 14 8∙1 30 17∙4
￿30 36 1∙54 22 61∙1 12 33∙3 9 25∙0 6 16∙7 10 27∙8
All 2014 0∙73 579 28∙7 297 14∙7 243 12∙1 167 8∙3 282 14∙0
SMR=standard mortality ratio (reference:Norwegian male population,1990).
CVD=cardiovascular disease;CHD=coronary heart disease.
†Non-CVD mortality=mortality from cancer+mortality from other non-CVD causes.
Table 3 Relative risks (RR) of coronary heart disease (CHD) mortality among
2014 apparently healthy men in Survey 1 during 23 years’ follow-up,and among 1428
apparently healthy men in Survey 2 during 16 years’ follow-up (Cox analyses)
Risk factor
CHD mortality in
Survey 1 —1996
(n=243)
CHD mortality in
Survey 2 —1996
(n=112)
RR 95% CI￿ RR 95% CI￿
ESR (mm.h
￿1
) 1∙59 (1∙27–1∙97)*** 1∙72 (1∙32–2∙25)***
Age (years) 2∙18 (1∙69–2∙81)*** 2∙02 (1∙36–3∙00)**
Haemoglobin (g.l
￿1
)‡ 1∙31 (0∙96–1∙80) 1∙16 (0∙95–1∙43)
Smoking status (yes/no) 1∙95 (1∙50–2∙53)*** 2∙10 (1∙41–3∙12)***
Total cholesterol (mmol.l
￿1
) 1∙33 (1∙03–1∙72)*
HDL cholesterol (mmol.l
￿1
)§ 0∙52 (0∙33–0∙83)**
Triglycerides (mmol.l
￿1
) 1∙18 (0∙95–1∙47) 1∙07 (0∙77–1∙49)
Fasting blood glucose (g.l
￿1
) 1∙23 (1∙02–1∙52)*
Systolic blood pressure (mmHg) 1∙38 (1∙07–1∙77)** 1∙70 (1∙19–2∙41)**
X-test ECG (positive vs negative) 1∙47 (1∙04–2∙07)* 1∙96 (1∙27–3∙01)**
Physical activity (active/not active) 0∙86 (0∙59–1∙26) 0∙74 (0∙40–1∙34)
Body mass index (kg.m
￿2
) 1∙08 (0∙81–1∙42) 1∙17 (0∙79–1∙74)
*P<0∙05,**P<0∙01,***P<0∙0001.
†Increase of 10 years.
‡Five (0∙4%) missing haemoglobin values during Survey 2.
§65 (4∙6%) missing values of high-density lipoprotein cholesterol during Survey 2.
￿Confidence interval.
Erythrocyte sedimentation rate 1617
Eur Heart J,Vol.21,issue 19,October 2000
sedimentation rate emerged as a strong long- and short-
term predictor of coronary heart disease mortality,and
carried the same relative prognostic information both
among healthy men,men with a positive exercise ECG
test and men with angina pectoris at Survey 2.More-
over,with increasing erythrocyte sedimentation rate
there was a much steeper gradient in percentages of men
dying from coronary heart disease without prior myo-
cardial infarction than in the percentages of coronary
heart disease survivors having had one or more myo-
cardial infarctions.Although our findings unquestion-
aby suggest that the erythrocyte sedimentation rate is a
marker of increased coronary heart disease risk,the
explanations are less clear.
The erythrocyte sedimentation rate is influenced by
the size and number of erythrocytes.The predictive
power of the erythrocyte sedimentation rate in the
present study is virtually identical to its power in a
previous study based on 488/2014 men in which haema-
tocrit replaced haemoglobin in multivariate analyses
[13]
.
Thus,differences in these parameters can hardly explain
our findings.
Accelerated erythrocyte aggregation is caused by
large,asymmetrical plasma proteins inhibiting the
negative electrical forces which normally keep the
erythrocytes apart.The proteins known to be mainly
responsible for this effect are fibrinogen,immunoglobu-
lins,lipoproteins and alpha-2 macroglobulin
[14]
.How
various cytokines,chemokines and other substances
liberated fromatheromatous tissue
[1]
influence the eryth-
rocyte sedimentation rate is largely unknown,but pro-
teins which facilitate thrombocyte aggregation and
cellular adhesion and migration might conceivably
also alter erythrocyte membranes and increase their
propensity to aggregate.This should be studied further.
Fibrinogen Since fibrinogen is an important determinant
of the erythrocyte sedimentation rate,most of our
erythrocyte sedimentation rate findings might be
claimed only to reflect differences in fibrinogen levels.
Age,smoking,obesity,elevated blood pressure,elevated
cholesterol and triglyceride levels,diabetes,season
(winter) and infection are known to increase fibrinogen
levels in men,while exercise,alcohol consumption and
hepatitis B surface antigen carriage have the opposite
effect
[15]
.However,our analyses (Table 3) show that the
erythrocyte sedimentation rate retains considerable pre-
dictive power even after adjustment for most of these
factors.Notably,a prerequisite for participation in the
study was an absence of febrile illnesses or surgery
during the 2 weeks before the erythrocyte sedimentation
rate was measured.Less than 25% of the variance in
Table 4 Relationship between erythrocyte sedimentation rate (ESR) and coronary
heart disease mortality among 403 men having developed angina pectoris and/or a
positive exercise ECG test at Survey 2*
Angina pectoris or
positive exercise ECG test
Angina
pectoris
Positive exercise ECG test,
not angina pectoris
n % Dead n % Dead n Dead
ESR 0–4 26/159 16∙4 10/47 21∙2 16/112 14∙3
ESR 5–9 19/116 16∙4 7/34 20∙6 12/82 14∙6
ESR 10–14 9/56 16∙1 4/17 23∙5 5/39 16∙1
ESR 15–29 17/60 28∙3 7/18 38∙9 10/42 23∙8
ESR ￿30 6/12 50∙0 3/5 60∙0 3/7 42∙9
Mean 77/403 19∙1 31/121 25∙6 46/282 16∙3
*16 years of follow-up.
Table 5 Relations between ESR and proportions of CHD survivors having had one
or more myocardial infarction versus proportions of men dying from coronary heart
disease without prior MI (21 years follow-up)
Number
in group
CHD survivors having had
one or more MI
Men dying from CHD
without prior MI
n % n %
ESR 0–4 805 89 11∙1 38 4∙7
ESR 5–9 745 90 12∙1 31 4∙2
ESR 10–14 256 32 12∙5 7 2∙7
ESR 15–29 172 20 11∙6 23 13∙4
ESR ￿30 36 8 22∙2 7 19∙4
Total 2014 240 11∙9 106 5∙3
ESR=Erythrocyte sedimentation rate (Survey 1);CHD=coronary heart disease;MI=myocardial
infarction.
1618 G.Erikssen et al.
Eur Heart J,Vol.21,issue 19,October 2000
erythrocyte sedimentation rate appears to be explained
by differences in fibrinogen levels in healthy individ-
uals
[16]
.Accordingly,the erythrocyte sedimentation rate
probably gives substantial additional information to
fibrinogen on the risk of coronary heart disease death,
although the magnitude of this effect cannot be esti-
mated in our study because of a lack of fibrinogen data.
Immuloglobulins The inflammatory responses in athero-
mas at various stages of development are heterogenous
and there are substantial differences in the numbers and
types of inflammatory cells involved
[17]
.Chronic inflam-
mation in the vessel wall may also be associated with
viral and bacterial infections
[18]
,and with autoimmune
diseases
[19]
—thus indicating that both humoral and
cell-mediated immune mechanisms may contribute to
atherogenesis
[20,21]
.High cellularity within the protective
fibrous cap markedly increases the risk of plaque
rupture
[22]
,and,in accord with the Duguid hypothesis,
plaque rupture may be one important pathway for the
progression of atherosclerosis and development of
acute atherothrombotic complications
[23]
.Conceivably,
humoral responses associated with highly cellular
atheromas may be stronger than in less cellular lesions.
The ensuing rise in plasma levels of fibrinogen,and in
some cases possibly also immunoglobulins,would
contribute to the increased erythrocyte sedimentation
rate.
Lipoproteins Although total cholesterol,high-density
lipoprotein cholesterol and triglycerides were included
as covariates in the multivariate analyses (Table 3),the
erythrocyte sedimentation rate emerged as a strong
predictor of coronary heart disease mortality.Exclusion
of these parameters had virtually no effect on the
predictive power of the erythrocyte sedimentation rate
(details not shown).
Alpha-2 macroglobulin The regulation of alpha-2
macroglobulin is poorly understood.Its levels remain
stable in most disease states,rise with increased loss of
other plasma proteins,but do not generally increase
during inflammation
[14]
.
Association with atherosclerosis
Autopsy studies have shown that at the time of coronary
heart disease death,atherosclerosis is a generalized
disease involving large amounts of tissue and usually
more than 50% of the total arterial surface
[24]
.An
elevated erythrocyte sedimentation rate may therefore
be a marker of the extent and/or intensity of a general
atherosclerotic process and thus a marker for advanced
atherosclerosis heralding increased risk of arterial
thrombosis.Recently,a similar strong association
between elevated C-reactive protein levels and thrombo-
occlusive complications of coronary heart disease was
suggested
[25]
.
The observation that the erythrocyte sedimentation
rate is correlated with established coronary heart dis-
ease risk factors (Table 1),also suggests that a high
erythrocyte sedimentation rate is a marker of advanced
atherosclerosis.
Association with cardiovascular disease
To our knowledge erythrocyte sedimentation rate
and cardiovascular disease associations have only been
reported from three prospective studies besides ours,
although these were less extensive and had shorter
follow-up
[26–30]
.Two
[26–29]
had approximately similar
erythrocyte sedimentation rate ranges at baseline but
shorter follow-up,and the other
[30]
reported associations
only among men with an erythrocyte sedimentation rate
>21 mm.h
￿1
.
Reference values for erythrocyte sedimentation rate
levels
Reference values for erythrocyte sedimentation rates
among healthy men are not established and probably
vary slightly in different parts of the world.Values of
<10 mm.h
￿1
have been recommended for men below
50 years and <14 for men above 60 years
[31]
.In all
but one
[30]
of the studies referred to above the mean
erythrocyte sedimentation rate at baseline was low (in
the range 6–9 mm.h
￿1
),and in all studies the partici-
pants were ‘apparently healthy’ at baseline.The present
study,having the longest follow-up,confirms and
extends the findings in these earlier reports.Moreover,
our data appear to fulfil most other requirements for
accepting epidemiological findings such as being consist-
ent,strong,graded,coherent,appropriately sequenced
and reasonably unbiased
[32]
.
Although we found a strong association between
elevated erythrocyte sedimentation rate and increased
coronary heart disease mortality,most coronary heart
disease decedents had normal or only slightly elevated
erythrocyte sedimentation rates.This only reflects what
is also observed for other coronary heart disease risk
factors.In one study for example
[33]
,normal C-reactive
protein levels were seen on admission to hospital in
three-quarters of patients with acute myocardial infarc-
tion.Cox regression analyses using the erythrocyte sedi-
mentation rate as a discrete variable in the present study
indicate that the dose-response relationship between the
erythrocyte sedimentation rate and coronary heart dis-
ease risk is roughly linear.Nevertheless,a high-risk
group can be defined:about 10% of the men had an
erythrocyte sedimentation rate ￿15 mm.h
￿1
(Table 2).
Among these,20% of the coronary heart disease deaths
occurred and,in the present material,the risk associated
with an erythrocyte sedimentation rate ￿15 mm.h
￿1
was comparable to having total cholesterol
>8 mmol.l
￿1
.
Conclusion
Besides being an inexpensive and easy test to do,
the erythrocyte sedimentation rate appears —in absence
of confounding conditions —to be a strong short- and
Erythrocyte sedimentation rate 1619
Eur Heart J,Vol.21,issue 19,October 2000
long-term predictor of coronary heart disease mortality
in apparently healthy,middle-aged men.Since the eryth-
rocyte sedimentation rate also carries strong prognostic
information in men with known or suspected coronary
heart disease and,since an increasing erythrocyte sedi-
mentation rate was associated with a particularly steep
gradient in the percentages of men dying from coronary
heart disease without prior myocardial infarction,it is
hypothesized that a high erythrocyte sedimentation
rate may be an indicator of aggressive,malignant forms
of coronary heart disease,conceivably by being a
marker of activated humoral immune mechanisms in
widespread atheromatous tissues.
References
[1] Ross R.Atherosclerosis —an Inflammatory disease.NEngl J
Med 1999;340:115–26.
[2] Meade TW,Brozovic M,Chakrabarti RR et al.Haemostatic
function and ishaemic heart disease:Principal results fromthe
Northwick Park Heart Study.Lancet 1986;2:533–7.
[3] Ridker PM,Cushman M,Stampffer MJ,Tracy RP,
Henneskens CH.Inflammation,aspirin,and the risk of
cardiovascular disease in apparently healthy men.N Engl J
Med 1997;336:973–9.
[4] International Committee for Standardization in Hematology
(Expert Panel on Blood Rheology).Guidelines on selection of
laboratory tests for monitoring the acute phase response.
J Clin Pathol 1988;41:1203–12.
[5] Erikssen J,Mundal R.The non-infarct patient with coronary
artery disease:can a high risk group be identified?Ann NY
Acad Sci 1982;382:438–9.
[6] Erikssen J.Aspects of latent coronary heart disease:a preva-
lence and methodological validation study of apparently
healthy,working,middle-aged men.PhD thesis,University
Hospital (Rikshospitalet) Medical Department B,Oslo,
Norway,1978.
[7] Erikssen G,Liestøl K,Bjørnholt J,Thaulow E,Sandvik L,
Erikssen J.Changes in physical fitness and changes in
mortality.Lancet 1998;352:759–62.
[8] Sandvik L,Erikssen G,Thaulow E.Long term effects of
smoking on physical fitness and lung function:a longitudinal
study of 1393 middle-aged men for seven years.Br Med J
1995;311:715–8.
[9] Mundal R,Erikssen J,Rodahl K.Assessment of physical
activity by questionnaire and personal interview with particu-
lar reference to fitness and coronary mortality.Eur J Appl
Physiol 1987;56:145–52.
[10] Erikssen J,Skrede S.Serum lipids and latent coronary
insufficiency.Scand J Clin Lab Invest 1977;37:243–50.
[11] Bjørnholt JV,Erikssen G,Aaser E et al.Fasting blood
glucose:an underestimated risk factor for cardiovascular
death.Diabetes Care 1999;22:45–9.
[12] Frank CV,Weinblatt E,Shapiro S.Angina pectoris in men.
Prognostic significance of selected medical factors.Circulation
1973;47:509.
[13] Erikssen G,Thaulow E,Sandvik L,Stormorken H,Erikssen
J.Haematocrit:a predictor of cardiovascular mortality?
J Intern Med 1993;234:494–9.
[14] Fernlund P,Hanson A,Laurell C-B,Lundh B.Klinisk kemi i
praktisk medisin (5th edn).Studentlitteratur,Lund,Sweden,
1986.
[15] Montalescot G,Collet JP,Choussat R,Thomas D.Fibrino-
gen as a risk factor for coronary heart disease.Eur Heart J
1998;19 (Suppl H):H11–H17.
[16] Bain BJ.Some influences on the ESR and the fibrinogen level
in healthy subjects.Clin Lab Haemat 1983;5:45–5.
[17] van der Wal AC,Becker AE,van der Loos CM,Das PK.Site
of intimal rupture or erosion of thrombosed coronary athero-
sclerotic plaques is characterized by an inflammatory process
irrespective of the dominant plaque morphology.Circulation
1994;89:36–44.
[18] Metha JL,Saldeen TG,Rand K.Interactive role of infection,
inflammation and traditional risk-factors in atheroscleosis and
coronary artery disease.J AmColl Cardiol 1998;31:1217–25.
[19] Schafer Elinder L,Frostegard J.Ateroskleros inte bara en
vellevnadssjukdom.Infeksjoner kan vara en riskfaktor for
hjart-karlsjukdom.La¨kartidningen 1998;95:2203–6.
[20] Schmitz G,Herr AS,Rothe G.T-lymphocytes and monocytes
in atherogenesis.Herz 1998;23:168–77.
[21] Ridker PM,Henneskens CH,Roitman-Johnson B,Stampfer
MJ,Allen J.Plasma concentration of soluble intercellular
adhesion molecule 1 and risk of future myocardial infarction
in apparently healthy men.Lancet 1998;351:88–92.
[22] Falk E,Shah PF,Fuster V.Coronary plaque disruption.
Circulation 1995;92:657–71.
[23] Libby P.Molecular bases of the acute coronary syndromes.
Circulation 1995;91:2844–50.
[24] Berenson GS,Srinivasan SR,Bao W,Newman WP,Tracy
RE,Wattigney WA.Association between multiple cardiovas-
cular risk factors and atherosclerosis in children and young
adults.N Engl J Med 1998;338:1650–6.
[25] Koenig W,Sund M,Fro¨hlich M et al.C-reactive protein,a
sensitive marker of inflammation,predicts future risk of
coronary heart disease in initially healthy middle-aged men.
results from the MONICA (Monitoring Trends and Determi-
nants in Cardiovascular Disease) Augsburg Cohort Study,
1984 to 1992.Circulation 1999;99:237–42.
[26] Natvig H,Borchgrevink CF,Dedichen J,Owren PA,Schio¨tz
EH,Westlund K.A controlled trial of the effect of linolenic
acid on incidence of coronary heart disease.Scand J Clin Lab
Invest 1968;22:1–20.
[27] Carlson LA,Bo¨ttiger LE.Ischæmic heart-disease in relation
to fasting values of plasma triglycerides and cholesterol.
Stockholm Prospective Study.Lancet 1972;1:865–8.
[28] Carlson LA,Bo¨ttiger LE,Arhfeldt PE.Risk-factors for myo-
cardial infarction in the Stockholm Prospective Study.A
14-year follow-up focussing on the role of plasma triglycerides
and cholesterol.Acta Med Scand 1979;206:351–60.
[29] Bo¨ttiger LE,Carlson LA.Risk factors for ischaemic
vascular death for men in the Stockholm Prospective Study.
Atherosclerosis 1980;36:389–408.
[30] Gillum RF,Mussolino ME,Makuc DM.Erythrocyte
sedimentation rate and coronary heart disease:the NHANES
I epidemiologic follow-up study.J Clin Epidemiol 1995;48:
353–61.
[31] Lewis SM.Erythrocyte sedimentation rate and plasma
viscosity.Association of Clinical Pathologists Broadsheet
1980;94:1–7.
[32] Statistical evidence and inference.In:Bradford Hill A.
Principles of medical statistics,9th edn.New York:Oxford
University Press,1971:309–23.
[33] Maseri A.Inflammation in ischemic heart disease.Br Med J
1996;312:1049–50.
1620 G.Erikssen et al.
Eur Heart J,Vol.21,issue 19,October 2000