Diabetes Research and Clinical Practice 69 (2005) 249–255
www.elsevier.com/locate/diabres
Increased erythropoiesis and subclinical inflammation
as part of the metabolic syndrome
T. Mardi a,d,e, S. Toker a,b,d,e, S. Melamed c,d,e, A. Shirom b,e, D. Zeltser a,d,e,
I. Shapira a,d,e, S. Berliner a,d,e,*, O. Rogowski a,d,e
a
Department of Medicine ‘‘D’’ and Institute for Special Medical Examinations (MALRAM),
Tel Aviv Sourasky Medical Center, Tel Aviv, affiliated to Sackler Faculty of Medicine,
Tel Aviv University, Tel Aviv, 6 Weizman Street, Tel Aviv 64239, Israel
b
Faculty of Management, Tel Aviv University, Tel Aviv, Israel
c
National Institute of Occupational and Environmental Health, Raanana, Israel
d
Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
e
Tel Aviv Medical Center Inflammation Survey (TAMCIS) Group, Tel Aviv, Israel
Received 7 June 2004; received in revised form 25 November 2004; accepted 10 January 2005
Available online 3 March 2005
Abstract
Recent studies have suggested the insulin resistance might be accompanied by enhanced erythropoiesis. We have examined
this association in individuals with the metabolic syndrome (MS) who in addition to insulin resistance harbor a chronic low
grade inflammation. This study is relevant because chronic inflammation might have a suppressive effect on erythropoiesis. 280
and 554 non-smoking women and men with respective age of 46.4 9.3 (mean S.D.) and 44.0 11.0 years are included. A
significant correlation was noted between the numbers of the components of the MS and the inflammatory biomarkers including
the white blood cell count, high sensitivity C-reactive protein, fibrinogen concentrations and the erythrocyte sedimentation rate.
In addition, a significant correlation (r = 0.157, p = 0.008) was noted between the number of components of the MS and the
number of red blood cells in the peripheral blood in women. The same was true for men (r = 0.192, p < 0.0005). We conclude
that enhanced erythropoiesis could be a new, hitherto unrecognized component of the MS. The enhanced erythropoiesis could
give an erroneous impression of general ‘‘good’’ health in these individuals.
# 2005 Elsevier Ireland Ltd. All rights reserved.
Keywords: Erythropoiesis; Metabolic syndrome; Inflammation
Abbreviations: MS, metabolic syndrome; TAMCIS, Tel Aviv Medical Center Inflammation Survey; WBCC, white blood cell count; ESR,
erythrocyte sedimentation rate; hs-CRP, high sensitivity C-reactive protein; HOMA, homeostasis model assessment; BMI, body mass index;
RBC, red blood cells; PMN, polymorphonuclears; BP, blood pressure
* Corresponding author. Tel.: +972 3 6973133; fax: +972 3 6974961.
E-mail address: [email protected] (S. Berliner).
0168-8227/$ – see front matter # 2005 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.diabres.2005.01.005
250
T. Mardi et al. / Diabetes Research and Clinical Practice 69 (2005) 249–255
1. Introduction
The metabolic syndrome (MS) conveys a significant risk for future atherothrombotic cardiovascular
events. It has been suggested recently that increased
erythropoiesis may be a new aspect of the insulin
resistance syndrome [1,2]. This is rather unexpected
due to the low grade inflammation that accompanies
the MS [3] that could contribute to the presence of an
anemia of chronic disease.
Here, we investigate the existence of enhanced
erythropoiesis in individuals with the MS in the
presence of low grade inflammation. The results point
to the presence of enhanced erythropoiesis in
individuals with the MS despite harboring a low
grade state of inflammation. These findings have
relevance since the enhanced erythropoiesis could
give an erroneous impression of general ‘‘good’’
health in these individuals.
in the present survey gave their written informed
consent according to the instructions of the Institutional Ethics Committee. Recruitment was based on
local announcements and advertisements in the
monthly pay bill of the medical personnel as well
as personal appeal to the patients in the various
outpatient clinics to participate in the inflammation
survey. Individuals with an underlying inflammatory
disease (arthritis, inflammatory bowel disease, etc.) as
well as individuals with any infections or other
inflammatory condition, including infarction, surgery
or angiography during the 6 months before recruitment are excluded. We also excluded individuals with
anemia (hemoglobin lower than 13.5 g/dL for men and
11.7 g/dL for women), those treated with steroid or
non-steroidal anti-inflammatory medication, except
for aspirin (at doses lower then 325 mg/day), smokers
as well as females receiving hormone replacement
therapy, and any patients taking lipid lowering
medications including HMG-CoA Reductase Inhibitors (Statins) and fibrates.
2. Methods
2.2. Definitions of the MS
2.1. The participants
We have used the data base of the Tel Aviv Medical
Center Inflammation Survey (TAMCIS) [4–6]. The
rationale of this survey was to determine the presence
of microinflammation in apparently healthy individuals and those with atherothrombotic risk factors in
Israel. Large cross sectional studies are being
performed in many other countries including Europe
and North America but this is the first large scale
survey that is currently taking place in Israel. It will
enable us to find out if the findings related to low grade
inflammation that were reported in other countries are
also relevant for ours. TAMCIS is a cross sectional
study to which we invited apparently healthy employees of the Tel Aviv Medical Center and Tel Aviv
Municipality (Israel) and individuals with atherothrombotic risk factors who are followed up in the
outpatient clinics of the Medical Center. Members of
the medical staff, retired employees of the medical
center and the Tel Aviv Municipality, individuals who
are investigated in our outpatient health screening
program as well as those who are in routine follow up
in diabetes, hypertension and metabolic disorders
clinics are also included. All the individuals included
The National Cholesterol Education Program
(NCEP) Adult Treatment Panel–III guidelines define
the MS as the presence of 3 of the following risk
determinants: (1) increased waist circumference
(>102 cm for men, >88 cm for women), (2) elevated
triglycerides of 1.70 mmol/L, (3) low HDL cholesterol
(1.03 mmol/L for men, 1.29 mmol/L for women), (4)
hypertension (systolic blood pressure 130 mmHg or
diastolic pressure 85 mmHg) or antihypertensive
medication use, and (5) impaired fasting glucose
6.1 mmol/L [7].
2.3. Laboratory tests
White blood cell count (WBCC) and differential
were performed by using the Coulter STKS (Beckman
Coulter, Nyon, Swiss) electronic cell analyzer. The
erythrocyte sedimentation rate (ESR) was determined
by the method of Westergren [8], quantitative
fibrinogen by the method of Clauss [9] and a Sysmex
6000 (Sysmex Corporation, Hyaga, Japan) autoanalyzer and the high sensitivity C-reactive protein (hsCRP) was performed by nephelometry using a
Boering BN II Nephelometer (DADE Boering,
T. Mardi et al. / Diabetes Research and Clinical Practice 69 (2005) 249–255
Marburg, Germany) [10]. The coefficient of variation
(CV) for hs-CRP concentrations below 1 mg/L is 0.11
while it is 0.05 for concentrations >1 mg/L. The CV
for fibrinogen is 0.06.
2.4. Statistical analyses
Statistical analysis was performed separately for
men and women. All data was summarized and
displayed as mean S.D. for the continuous variables
(age, the MS components and the inflammation
markers), and as number of patients plus the
percentage in each group for categorical variables
(cardiovascular risk factors and medication). The
crosstabs and descriptive procedures were used to
produce frequencies of categorical variables (cardiovascular risk factors and medication) and
means S.D. of continuous variables (age, the MS
components and the inflammation markers). Participants were divided into four groups according to the
number of components of the MS: 0, 1, 2 and 3 or
more components. For all continuous variables, a oneway ANOVA analysis was performed to compare the
251
various parameters between the different groups.
Hochberg’s multiple comparison technique was used
for pairwise comparison between patient’s categories.
Pearson partial correlations for confounding variables
were used to evaluate the association between the
number of components of the MS and the different
inflammation markers, as well as the different
component of the blood count, and between the
erythrocyte parameters to the individual components
of the MS. All correlations were adjusted for age and
cardiovascular risk factors (hypertension and hyperlipidemia). In order to assess which components of the
MS contribute to the change in erythrocyte parameters
(hemoglobin, hematocrit and red blood cells), we
performed linear regression using those three erythrocyte parameters as the dependent variable and all
the different components of the MS as the covariates
using the stepwise method. The hs-CRP has non
normal distribution, thus we used a logarithmic
transformation, and the results of hs-CRP are reflected
as back transformed geometric means and S.D. and for
the correlation the log (hs-CRP) was used. The level of
significance used for all of the above analyses was two
Table 1
Age, body mass index, hypertension and dyslipidemia as well as hemoglobin, hematocrit and red blood cells and the different inflammatory
variables (N, mean and S.D.) for women, in relation to the number of the components of the metabolic syndrome
Number of components
0 (N = 123)
1 (N = 82)
2 (N = 44)
3 (N = 31)
P-value
Hochberg
Women (N = 280)
Age (years)
43.8 9.9
46.5 8.6
51.1 8.1
50.0 6.5
<0.0005
0–2
0–3
1–2
<0.0005
0.004
0.037
BMI (kg/m2)
23.9 3.2
25.7 3.9
29.3 4.2
30.8 4.6
<0.0005
0–1
0,1–2,3
0.005
<0.0005
Hypertension (%)
2.4
11.0
15.9
32.3
<0.0005
Dyslipidemia (%)
4.9
11.0
14.0
19.4
N.S.
Hemoglobin (mmol/L)
8 0.6
7.9 0.7
8.1 0.6
8.3 05
N.S.
N.S.
N.S.
Hematocrit (%)
37.7 2.7
37.5 3.2
38.2 2.7
38.8 2.3
N.S.
N.S.
N.S.
RBC (106 cells/mL)
4.4 0.3
4.4 0.3
4.5 0.4
4.5 0.3
0.016
N.S.
N.S.
6.64 1.47
6.68 1.54
6.54 1.36
7.45 1.90
0.048
N.S.
N.S.
Polymorphonuclears (10 cells/mL)
3.91 1.21
3.89 1.26
3.89 1.14
4.57 1.55
N.S.
N.S.
N.S.
ESR (mm/h)
16.3 6.8
20.1 11.3
21.4 9.6
20.1 11.2
0.003
0–1
0–2
0.027
0.013
Fibrinogen (mmol/L)
8.4 1.41
8.3 1.4
8.8 1.5
9.1 1.8
0.029
N.S.
N.S.
hs-CRP (mg/L)
1.0 2.9
1.4 2.9
2.3 3.2
3.9 3.1
<0.0005
0–2,3
1–3
<0.0005
<0.0005
3
WBCC (10 cells/mL)
3
252
T. Mardi et al. / Diabetes Research and Clinical Practice 69 (2005) 249–255
tailed, p < 0.05. The SPSS statistical package was
used to perform all statistical evaluation (SSPS Inc.,
Chicago, IL, USA).
3. Results
We have examined 554 men and 280 women of
mean age 44.8 10.6 years. There were 123 women
with no components of the MS, 82 with 1, 44 with 2 and
31 with 3 components; the respective numbers for
men being 184, 190, 107 and 73. Several subjects were
taking medications but the percentage of individuals
taking low dose aspirin did not exceed 8%, 4% for alpha
blockers, 16% for beta blockers, 7% for calcium
channel blockers, 6% for ACE inhibitors, 3% angiotensin II receptor blockers and 3% for diuretics.
The results of the hematological variables and that
of the inflammatory Biomarkers are reported in
Table 1 for women and Table 2 for men. It is noted
that the multiplicity of the components of the MS is
associated with low grade inflammation as expressed
by the inflammatory biomarkers. These inflammatory
biomarkers correlate with the number of components
of the MS in both women and men (Table 3). A
significant correlation between the number of these
components and the number of red blood cells in
women (r = 0.157, p = 0.008) and men (r = 0.192,
p < 0.0005) is noted.
We have further analyzed the correlations between
the individual components of the MS and the
hematological parameters (Table 4). A significant
correlation with the number of erythrocytes was noted
for waist in both genders and triglycerides in men.
Finally, a linear regression model was performed
by using the individual components of the MS as the
independent variables, and the hematological variables as the depend ones. The results are reported in
Table 2
Age, body mass index, hypertension and dyslipidemia as well as hemoglobin, hematocrit and red blood cells and the different inflammatory
variables (N, mean and S.D.) for men, in relation to the number of the components of the metabolic syndrome
Number of components
0 (N = 184)
1 (N = 190)
2 (N = 107)
3 (N = 73)
P-value
Hochberg
Men (N = 554)
Age (years)
40.4 10.6
44.1 11.0
47.2 10.0
48.3 10.7
<0.0005
0–1
0–2,3
1–3
0.005
<0.0005
0.025
BMI (kg/m2)
24.9 2.7
25.9 2.7
28.0 3.2
30.4 3.5
<0.0005
0–1
0,1–2,3
2–3
0.006
<0.0005
<0.0005
Hypertension (%)
0
6.8
17.8
30.1
<0.0005
Dyslipidemia (%)
5.5
8.4
24.3
23.3
<0.0005
Hemoglobin (mmol/L)
9.2 0.6
9.2 0.6
9.3 0.6
9.3 0.6
N.S.
N.S.
N.S.
Hematocrit (%)
43.2 2.7
43.3 2.7
43.7 2.6
43.9 2.7
N.S.
N.S.
N.S.
5.0 0.4
5.0 0.4
5.1 0.3
5.1 0.3
0.003
0–3
1–3
0.003
0.035
6.56 1.75
6.59 1.42
7.09 1.48
7.11 1.49
0.004
0–2
0.033
Polymorphonuclears (10 cells/mL)
3.73 1.10
3.82 1.11
4.14 1.15
4.07 1.12
0.011
0–2
0.020
6
RBC (10 cells/mL)
WBC (103 cells/mL)
3
ESR (mm/h)
6.4 4.9
9.3 7.4
10.2 7.4
10.9 8.3
<0.0005
0–1,2,3
<0.0005
Fibrinogen (mmol/L)
7.2 1.4
7.7 1.5
7.8 1.4
8.2 1.4
<0.0005
0–1
0–2
0–3
0.004
0.002
<0.0005
hs-CRP (mg/L)
0.9 2.4
1.3 3.0
1.5 2.6
2.2 2.2
<0.0005
0–1
0–2,3
1–3
0.007
<0.0005
<0.0005
T. Mardi et al. / Diabetes Research and Clinical Practice 69 (2005) 249–255
253
Table 3
Results of the age adjusted Pearson partial correlation coefficients between the number of components of the metabolic syndrome and the
inflammatory variables for women (upper part) and for men (lower part)
ESR
Women (N = 280)
Number of components
Men (N = 554)
Number of components
Fibrinogen
Log (hs-CRP)
WBCC
PMN
Correlation coefficient
Significance
0.143
0.017
0.073
N.S.
0.341
<0.0005
0.154
0.010
0.156
0.009
Correlation coefficient
Significance
0.184
<0.0005
0.147
0.001
0.250
<0.0005
0.150
<0.0005
0.133
0.002
Table 5 and show that the number of erythrocytes is
mainly influenced by waist in both genders and the
concentration of triglycerides in men.
4. Discussion
It has been realized recently that the MS is
associated with an erythrocytotic response [1] and that
this might be a hitherto unrecognized feature of the
syndrome [2]. In fact, Choi et al. recently reported a
study in 1314 non-diabetic subjects over the age of 60,
selected from a cross-sectional study in Seoul [2]. The
authors measured fasting and post-load 2h plasma
glucose, insulin concentrations, lipid profile, and
anthropometric as well as hematological parameters.
The authors found a positive correlation between
insulin resistance also and hemoglobin concentrations
in non-smoking men (r = 0.2, p = 0.018). In nonsmoking women, insulin resistance correlated with
hemoglobin (r = 0.1, p = 0.001). Hemoglobin concentrations were also associated with other components of the insulin resistance syndrome such as body
man index, blood pressure, lipid profile and fasting
plasma insulin levels used as a surrogate for insulin
resistance. Furthermore, the group in the upper
quartile of insulin resistance showed higher hemoglobin concentrations than the lower quartile, independent of smoking status and serum iron concentrations.
The authors suggested that increased erythropoiesis
and subclinical inflammatory, expressed as the number
of white blood cells in peripheral blood, could be part
of the MS [2].
In another study, Barbieri et al. [1] found a
correlation between insulin resistance and red cell
count (r = 0.14, p < 0.0001), plasma hemoglobin
Table 4
Results of the age adjusted Pearson partial correlation coefficients between the hemoglobin, hematocrit and red blood cells to the individual
components of the metabolic syndrome for women (upper part) and for men (lower part)
Waist
HDL
TG
Glucose
Systolic BP
Diastolic BP
Correlation coefficient
Significance
0.035
N.S.
0.035
N.S.
0.110
N.S.
0.066
N.S.
0.114
N.S.
0.092
N.S.
Hematocrit
Correlation coefficient
Significance
0.045
N.S.
0.018
N.S.
0.094
N.S.
0.062
N.S.
0.125
N.S.
0.098
N.S.
RBC
Correlation coefficient
Significance
0.177
0.003
0.101
N.S.
0.192
0.001
0.141
0.018
0.116
0.052
0.052
N.S.
Correlation coefficient
Significance
0.052
N.S.
0.046
N.S.
0.163
<0.0005
0.027
N.S.
0.062
N.S.
0.104
0.016
Hematocrit
Correlation coefficient
Significance
0.070
0.028
0.019
N.S.
0.155
<0.0005
0.011
N.S.
0.099
0.021
0.121
0.005
RBC
Correlation coefficient
Significance
0.152
<0.0005
0.088
0.040
0.221
<0.0005
0.048
N.S.
0.146
0.001
0.153
<0.0005
Women (N = 280)
Hemoglobin
Men (N = 554)
Hemoglobin
254
T. Mardi et al. / Diabetes Research and Clinical Practice 69 (2005) 249–255
Table 5
Linear regression model coefficients using the individual components of the metabolic syndrome as the independent variables, and the
hemoglobin, hematocrit and the red blood cells as the dependent variables in women (upper part) and in men (lower part)
Variables entered
the model
Unstandardized
coefficients
Standardized
coefficients
Significance
Partial
correlation
Women
Hemoglobin
Systolic BP
0.009
0.148
0.013
0.148
Hematocrit
Systolic BP
0.030
0.176
0.003
0.176
RBC
Triglycerides
Waist
0.001
0.004
0.151
0.129
0.017
0.042
0.143
0.122
Triglycerides
Glucose
0.002
0.005
0.176
0.093
<0.0005
0.036
0.169
0.090
Hematocrit
Triglycerides
0.005
0.149
<0.0005
0.149
RBC
Triglycerides
0.001
0.200
<0.0005
0.200
Men
Hemoglobin
(r = 0.16, p < 0.001), hematocrit (r = 0.15,
p < 0.001) and plasma iron (r = 0.1, p < 0.05). Red
blood cell count was also associated with other
biological markers of insulin resistance syndrome in
this Chianti study that included 608 subjects aged 22–
96 years [1].
The notion that insulin and insulin growth factors I
and II support in vitro [11–15] and vivo [16–18]
erythropoiesis is not new. However, the MS is
accompanied by low grade inflammation, a factor
that could support potential down regulation of
erythropoiesis in these individuals [19]. Thus, it
remained to be seen if the factors that prevail are the
erythropoietic or the erythrosuppressive ones. This is
especially important due to the fact that anemia by
itself has been shown to be a risk factor for
cardiovascular disease [20]. In fact, the presence of
anemia was independently associated with an
increased risk of cardiovascular disease in the
atherosclerosis risk in communities (ARIC) study, a
community cohort of subjects between the ages of 45
and 64 years [20].
In the present study, we evaluated several
established inflammatory biomarkers including the
WBCC, ESR, fibrinogen and hs-CRP concentrations
in addition to the hematological ones. The significant
finding is therefore that erythropoiesis was evident,
especially in men, despite there being clear indications
of chronic inflammation. The erythropoiesis correlated with waist circumference, a classical marker of
central obesity and insulin resistance.
The limitation of the present study is that it did not
included measurements of insulin or the homeostasis
model assessment (HOMA) index. In addition, we do
not know what effect, if any, this enhanced
erythropoiesis might have on the course of the
atherothrombotic disease. Another limitation is the
lack of IL-6 measurements. This cytokine is probably
increased in individuals with multiplicity of components of the MS and could contribute to both the
production of inflammation sensitive proteins in the
liver [21] and at the same time be involved in
erythropoiesis [22]. Finally, it should be stressed that
our study is not a truly population based cross
sectional one. In fact, we cannot exclude the
possibility of a potential selection bias introduced
by the recruitment of subjects from TAMCIS.
We have included several individuals who were on
medication that might have had an effect on
inflammation (e.g. ACE inhibitors) or insulin resistance (thiazide diuretics). However, due to the small
number of subjects who were taking these medications, it is unlikely that they had a major effect on the
results of the present study.
We conclude that the multiplicity of components of
the MS is associated with erythropoiesis despite of the
presence of concomitant low grade inflammation. The
lack of anemia in these individuals might give an
erroneous impression of general ‘‘good’’ health. It
remains to be seen whether this enhanced erythropoiesis might have significance in terms of viscosity,
especially in the presence of hyperfibrinogenemia.
T. Mardi et al. / Diabetes Research and Clinical Practice 69 (2005) 249–255
References
[1] M. Barbieri, E. Ragno, E. Benvenuti, et al. New aspects of the
insulin resistance syndrome: impact on haematological parameters, Diabetologia 44 (2001) 1232–1237.
[2] K.M. Choi, J. Lee, Y.H. Kim, et al. Relation between insulin
resistance and hematological parameters in elderly KoreansSouthwest Seoul (SWS) Study, Diab. Res. Clin. Pract. 60
(2003) 205–212.
[3] P.M. Ridker, J.E. Buring, N.R. Cook, N. Rifai, C-reactive
protein, the metabolic syndrome, and risk of incident cardiovascular events: an 8-year follow-up of 14719 initially healthy
American women, Circulation 107 (2003) 391–397.
[4] S. Cohen, T. Tolshinsky, O. Rogowski, et al. Real time, control
adjusted evaluation of intensity of the inflammatory response,
J. Inform. Techn. Healthcare 1 (2003) 195–207.
[5] O. Rogowski, I. Shapira, A. Zimran, et al. Automated system
to detect low-grade underlying inflammatory profile: Gaucher
disease as a model, Blood Cells Mol. Dis. 34 (2005) 26–29.
[6] O. Rogowski, S. Toker, I. Shapira, et al. Values of high
sensitivity C-reactive protein in each month of the year in
apparently healthy individuals, Am. J. Cardiol. 95 (2005) 152–
155.
[7] Executive Summary of The Third Report of The National
Cholesterol Education Program (NCEP) Expert Panel on
Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III), JAMA 285 (2001)
2486–2497.
[8] International committee for standardization in hematology.
Recommendation of measurement of erythrocyte sedimentation rate of human blood, Immunochemistry 2 (1965) 235–254.
[9] A. Clauss, Gerinnungsphysiologische Schnellmethode zur
Bestimmung des Fibrinogens, Acta Haematol. Basel 17
(1957) 237–246.
[10] N. Rifai, R.P. Tracy, P.M. Ridker, Clinical efficacy of an
automated high-sensitivity C-reactive protein assay, Clin.
Chem. 45 (1999) 2136–2141.
[11] S. Miyagawa, M. Kobayashi, N. Konishi, T. Sato, K. Ueda,
Insulin and insulin-like growth factor I support the proliferation of erythroid progenitor cells in bone marrow through the
sharing of receptors, Br. J. Haematol. 109 (2000) 555–562.
255
[12] J. Ratajczak, Q. Zhang, E. Pertusini, B.S. Wojczyk, M.A.
Wasik, M.Z. Ratajczak, The role of insulin (INS) and insulin-like growth factor-I (IGF-I) in regulating human erythropoiesis. Studies in vitro under serum-free conditions–
comparison to other cytokines and growth factors, Leukemia
12 (1998) 371–381.
[13] K. Sawada, S.B. Krantz, E.N. Dessypris, S.T. Koury, S.T.
Sawyer, Human colony-forming units-erythroid do not require
accessory cells, but do require direct interaction with insulinlike growth factor I and/or insulin for erythroid development, J.
Clin. Invest. 83 (1989) 1701–1709.
[14] N. Dainiak, S. Kreczko, Interactions of insulin, insulinlike
growth factor II, and platelet-derived growth factor in erythropoietic culture, J. Clin. Invest. 76 (1985) 1237–1242.
[15] N. Bersch, J.E. Groopman, D.W. Golde, Natural and biosynthetic insulin stimulates the growth of human erythroid progenitors in vitro, J. Clin. Endocrinol. Metab. 55 (1982) 1209–
1211.
[16] B.S. Stonestreet, M. Goldstein, W. Oh, J.A. Widness, Effects of
prolonged hyperinsulinemia on erythropoiesis in fetal sheep,
Am. J. Physiol. 257 (1989) R1199–R1204.
[17] M. Petakov, L. Biljanovic-Paunovic, G. Jovcic, N. Stojanovic,
V. Todorovic, W. Jelkmann, The influence of acute sterile
inflammation on erythropoiesis in rats, Exp. Hematol. 26
(1998) 222–227.
[18] J.A. Widness, J.B. Susa, J.F. Garcia, et al. Increased erythropoiesis and elevated erythropoietin in infants born to diabetic
mothers and in hyperinsulinemic rhesus fetuses, J. Clin. Invest.
67 (1981) 637–642.
[19] G. Weiss, Pathogenesis and treatment of anaemia of chronic
disease, Blood Rev. 16 (2002) 87–96.
[20] M.J. Sarnak, H. Tighiouart, G. Manjunath, et al. Anemia as a
risk factor for cardiovascular disease in the Atherosclerotic
Risk Communities (ARIC) study, J. Am. Coll. Cardiol. 40
(2002) 27–33.
[21] J.M. Fernandez-Real, W. Ricart, Insulin resistance and chronic
cardiovascular inflammatory syndrome, Endocr. Rev. 24
(2003) 278–301.
[22] T. Klausen, N.V. Olsen, T.D. Poulsen, J.P. Richalet, B.K.
Pedersen, Hypoxemia increases serum interleukin-6 in
humans, Eur. J. Appl. Physiol. Occup. Physiol. 76 (1997)
480–482.