Content uploaded by Tanja Pessi
Author content
All content in this area was uploaded by Tanja Pessi on May 12, 2015
Content may be subject to copyright.
ORIGINAL ARTICLE
Use of combined oral contraceptives alters metabolic determinants and genetic regulation
of C-reactive protein. The Cardiovascular Risk in Young Finns Study
Atte Haarala
1
, Carita Eklund
1
, Tanja Pessi
1
, Terho Lehtima¨ki
2
, Risto Huupponen
3,4
, Antti Jula
5
,
Jorma Viikari
6
, Olli Raitakari
7
and Mikko Hurme
1
1
Department of Microbiology and Immunology, University of Tampere, Tampere, Finland;
2
Department of Clinical Chemistry,
Tampere University Hospital and University of Tampere, Finland;
3
Department of Pharmacology, Drug Development and
Therapeutics, Turku, Finland;
4
Health Care District of Southwest Finland, Clinical Pharmacology, TYKSLAB;
5
National Public
Health Institute, Turku, Finland;
6
Department of Medicine, University of Turku, Turku, Finland;
7
Department of Clinical
Physiology, University of Turku, Turku, Finland
Background. Use of combined oral contraceptives (COCs) is known to increase concentrations of C-reactive
protein (CRP), an important predictor of cardiovascular disease. The inflammatory nature of the disease is well
acknowledged. The aim of this study was to find out whether the metabolic, lifestyle and genetic determinants of
CRP differ between women who use COCs and those who do not use any hormonal contraceptives (non-users).
Material and methods. A total of 1,257 women (24–39 years) participated in the ongoing Cardiovascular Risk in
Young Finns Study, a population based cross-sectional follow-up study. Use of hormonal contraceptives was
determined by questionnaire. Plasma CRP and other cardiovascular risk factors were measured; five CRP gene
polymorphisms were genotyped (2717AwG, 2286CwTwA, +1059GwC, +1444CwT and +1846GwA) and
CRP haplotypes were constructed. Results. Multivariate regression analysis revealed that BMI and leptin were
the main determinants of CRP in non-users, whereas in COC users the main determinants were BMI, leptin and
triglycerides. The median CRP and triglyceride values were significantly higher in COC users than in non-users.
The correlations between triglyceride and CRP were tested separately in different COC users in accordance with
progestagen content and dosage, the analysis revealing significant association only in women using a high dosage
of progestagen or cyproterone. The haplotypes of CRP gene had no significant association with CRP
concentration in COC users, while independent effects on CRP were found in non-users. Conclusion. Our study
suggests that use of COCs alters the metabolic determinants and genetic regulation of CRP.
Keywords: Body mass index; haplotypes; leptin; progestins; triglycerides
Introduction
C-reactive protein (CRP) is an acute phase protein
that has been widely used as a marker of acute
inflammation. It is now known that even a minor
elevation of CRP, i.e. low-grade inflammation, is
associated with an increased risk of cardiovascular
disease (CVD) morbidity and mortality [1,2]. Many
demographic, social, metabolic and lifestyle factors
are known to have an effect on CRP concentration,
e.g. age, sex, ethnicity, socio-economic status, birth-
weight, dietary pattern, physical activity, alcohol
consumption, diabetes mellitus, insulin concentra-
tions, glucose concentrations, blood pressure, body
mass index (BMI), HDL-cholesterol, triglycerides
and oestrogen/progestogen use [3,4].
Genetics plays a part in determination of CRP
concentration. Twin studies suggest that CRP plasma
concentrations are over 40 % heritable, and there is
an increasing amount of evidence showing an
association between CRP genetics and CRP concen-
tration [5,6]. For example, in an American popula-
tion, haplotypes constructed from seven selected
CRP gene SNPs have been associated with different
CRP concentrations [7]. In Finns, data from the
ongoing Cardiovascular Risk in Young Finns Study,
the population used in the present study, have shown
that at least five CRP SNPs are associated with
different CRP concentrations [8]. Haplotypes based
on those five SNPs have been associated with life-
long differences in average CRP concentrations, and
seem to explain about 5 % of circulating CRP
concentrations [9].
Combined oral contraceptives (COCs) have been
widely used as a safe and efficient method for
preventing unwanted pregnancies. Over the years,
the amounts of oestrogen and progestagen in COCs
Correspondence: Atte Haarala, Department of Microbiology and Immunology, University of Tampere Medical School, FIN-33014 University of Tampere,
Finland. Tel: +358 3 3551 7723. Fax: +358 3 3551 6173. Email: atte.haarala@uta.fi
(Received 30 April 2008; accepted 15 August 2008)
The Scandinavian Journal of Clinical & Laboratory Investigation,
Vol. 69, No. 2, April 2009, 168–174
ISSN 0036-5513 print/ISSN 1502-7686 online #2009 Informa UK Ltd (Informa Healthcare, Taylor & Francis AS).
DOI: 10.1080/00365510802449642
have been decreased and chemical formations have
been modified in order to reduce the risk of
thrombosis. The progestagens can be divided into
second-generation (norgestrel, levonorgestrel, norges-
trione) and third-generation (desogestrel, gestodene)
compounds. Some progestagens are not classifiable
into second or third generation (e.g. cyproterone,
norgestimate). Despite these developments, with use
of COCs there is still an increased the risk of
developing myocardial infarction, especially in
women with other cardiovascular risk factors [10].
Similarly, the risk of thromboembolic disease is still
increased [11].
We have previously shown in this population that
women using oral contraceptives have higher CRP
concentrations than non-users [12]. It is usually
considered that increased CRP is mostly due the
oestrogen component in COCs, although there are
some studies suggesting that CRP concentration
might also depend on progestagen content in COCs
[13–15]. In the present analysis, our aim was to find
out whether the metabolic, lifestyle and genetic
determinants of CRP differ between women who
use COCs and those who do not use any hormonal
contraceptives.
Methods
Subjects
The subjects in the study were participants in the
ongoing Cardiovascular Risk in Young Finns Study,
a five-centre follow-up study involving five university
hospital cities in Finland. The study began in 1980,
when 3,596 participants aged 3, 6, 9, 12, 15 and 18
were randomly selected for the study [16]. The most
recent follow-up was conducted in 2001, when 1,257
women and 1,026 men were 24–39 years of age.
Cardiovascular risk factors, including serum lipids,
BMI, blood pressure values, CRP, alcohol consump-
tion, diabetes and smoking habits were recorded [17].
The study was approved by local ethics committees.
Clinical and chemical analyses
BMI was calculated from measured height and
weight. A random zero sphygmomanometer
(Hawksley & Sons Ltd, Lancinn UK) was used to
measure blood pressure and a mean of three
measurements was used in the analysis. From fasting
plasma sample CRP, insulin, leptin, total cholesterol,
HDL-cholesterol and triglyceride concentrations
were drawn. LDL-cholesterol concentration was
calculated using the Friedewald formula (detailed
description in [17,18]). Smoking habits, alcohol
consumption, hormonal contraceptive use, physical
activity [19], history of recent infection, diabetes
and chronic rheumatic disease were elicited by
questionnaire.
The study included women who did not use any
hormonal contraceptives (non-users) (n5811) and
those who used COCs (COC users) (n5305), proges-
tin only pills (n512), intrauterine devices (n5119) or
subcutaneous capsules (n53). This data was unavail-
able in 7 subjects. The COCs contained the following
substances: ethinylestradiol (n5291), oestradiol vale-
rate (n514), gestodene (n5136), desogestrel (n578),
cyproterone (n550), levonorgestrel (n547), norethis-
terone (n54), lynestrenol (n52). All intrauterine
devices released levonorgestrel.
Fasting plasma CRP concentrations were ana-
lysed using a high-sensitive latex turbidometric
immunoassay (Wako Chemicals GmbH, Neuss,
Germany); detection limit 0.06 mg/L. DNA was
extracted from whole blood using a commercially
available kit (Qiagen Inc., Hilden, Germany) in 2001.
CRP gene polymorphisms 2717AwG (rs2794521),
2286CwTwA (rs3091244), +1059GwC (rs1800947),
+1444CwT (rs1130864) and +1846GwA (rs1205)
were genotyped using the ABI Prism 7900HT
Sequence Detection System for both PCR and allelic
discrimination (Applied Biosystems, Foster City,
Calif., USA). For SNP +1059, a commercial kit from
Applied Biosystems was used (Assay On Demand,
C_177490_10 CRP). The SNPs 2717, +1444 and
+1846 were genotyped using Assays By Design from
Applied Biosystems under standard conditions. The
triallelic tagSNP 2286 was genotyped as previously
described [7], except for genotype calling, which was
done manually from the PCR run component tab.
Statistical analyses
The haplotypes were constructed using the PHASE v.
2.0.2 program [20] from the five CRP SNPs. This
program calculates from the genotype data the most
likely haplotype pairs for each individual using a
Bayesian statistical method. The haplotypes are
shown in the order 2717, 2286, +1059, +1444 and
+1846. The three most common haplotypes (fre-
quency w0.10) in this cohort were: A-T-G-T-G
(frequency 0.350), A-C-G-C-A (0.301), G-C-G-C-G
(0.207) (previously published in [8]).
The data were analysed with SPSS for Windows
statistical software (versions 14.0 and 15.0; SPSS Inc.,
Chicago, IL., USA). We excluded subjects whose
CRP concentrations were above 10 mg/L (n551),
triglycerides above 4 mmol/L (n54), who had a
history of recent infections (n583), diabetes (n511)
or chronic rheumatic disease (n525), who were
Combined oral contraceptives and CRP 169
pregnant (n562) or lactating (n554). Haplotyping
was unsuccessful in four subjects. No separate
analyses were performed for subjects using pills
containing progestin only, subcutaneous capsules,
oestradiol valerate, norethisterone or lynestrenol,
because the number of subjects in these groups was
low (nv10). As even vaginal administration of
hormones can affect protein synthesis of hepatocytes
[21], the intrauterine device users were excluded from
the non-user group. The total number of subjects
after the exclusions was 841.
Since the distributions of CRP, insulin, leptin
and triglyceride values were skewed, the non-
parametric Mann-Whitney test was used in statis-
tical analysis. For linear regression analysis and
for analysis of covariance (ANCOVA), the values
were log-transformed prior to analysis. Correlation
between skewed variables was estimated with
Spearman’s test. For normally distributed variables,
the t-test for independent samples was used to detect
differences in mean values among different groups.
The effects of metabolic and lifestyle factors on
CRP concentration were analysed using a linear
regression model, and therefore scale variables or
dummy variables were used. The regression model
for logarithmic CRP was constructed from the
following variables: BMI, waist circumference, age,
HDL-cholesterol, LDL-cholesterol, (log)triglycer-
ides, diastolic blood pressure, systolic blood pres-
sure, (log)insulin, glucose, (log)leptin, physical
activity index, alcohol consumption and smoking.
The variables were tested in a univariate model and
values that were associated with the dependent
variable (pv0.15) were selected for the multivariate
model. From co-linear variables (e.g. BMI and waist
circumference) only the one that showed a stronger
association with (log)CRP was selected for the
model. All non-significant variables (pw0.05) were
excluded one by one from the multivariate model,
starting from the least significant.
The effects of the haplotypes on CRP concentra-
tion were compared between carriers and non-
carriers using the Mann-Whitney test. The difference
between groups after adjustment by the variables
showing significance in the regression model
(pv0.05) was calculated with the ANCOVA method.
Results
Characteristics of the study subjects are given in
Table I. Women using COCs had higher median CRP
concentrations (pv0.001), triglyceride concentrations
(pv0.001), insulin concentrations (p50.032), mean
BMI (p50.002) and HDL-cholesterol (pv0.001) than
non-users. The COC users also had significantly lower
waist circumference (pv0.001), LDL-cholesterol
concentrations (pv0.001), glucose concentrations
(p50.009) and were significantly younger (pv0.001).
The intrauterine device users did not differ signifi-
cantly from the non-users according to the parameters
in data characteristics (data not shown).
To determine whether metabolic and lifestyle
determinants of CRP differ between non-users and
COC users, we built separate linear regression models
for these groups (Table II). In the multivariate
analysis of non-users, variables that remained sig-
nificant in the model were BMI and (log)leptin. In the
Table I. Characteristics of study subjects.
Variable
Non-users (n5591) COC users (n5250)
pfor difference*Mean SD Mean SD
Body mass index (kg/m
2
) 24.21 ¡4.49 26.21 ¡3.32 0.002
Waist circumference (cm) 79.37 ¡11.65 75.67 ¡8.62 v0.001
Age (years) 32.17 ¡4.99 29.36 ¡4.73 v0.001
HDL-cholesterol (mmol/L) 1.34 ¡0.28 1.54 ¡0.30 v0.001
LDL-cholesterol (mmol/L) 3.19 ¡0.75 2.93 ¡0.71 v0.001
Systolic blood pressure (mmHg) 116.13 ¡12.71 118.03 ¡12.49 0.053
Diastolic blood pressure (mmHg) 71.85 ¡8.74 72.52 ¡8.79 0.318
Glucose (mmol/L) 4.93 ¡0.43 4.84 ¡0.42 0.009
Physical activity index 16.50 ¡14.75 17.90 ¡14.05 0.258
Smoking (daily) (%) 20.07 % 20.99 %
Median Quartiles Median Quartiles pfor difference**
CRP (mg/L) 0.54 0.26–1.33 1.66 0.81–3.30 v0.001
Triglycerides (mmol/L) 0.90 0.70–1.20 1.20 1.00–1.60 v0.001
Insulin (mU/L) 6.00 5.00–9.00 7.00 5.00–9.00 0.032
Leptin (mU/L) 12.59 7.59–19.95 12.88 8.66–19.50 0.538
*T-test for difference between non-users and COC users. **Mann-Whitney test for difference between non-users and COC users.
170 A. Haarala et al.
multivariate analysis of COC users, variables that
remained significant were BMI, (log)leptin and
(log)triglycerides. To find out whether the effect
of triglycerides was dependent on a certain proges-
tagen formulation in COCs, we calculated correla-
tion between triglycerides and CRP inside different
progestagen-containing subgroups (Table III). The
correlation was significant only in subjects using
cyproterone (r50.508, p50.001). To assess the
progestagen dosage effect, we calculated the corre-
lation in women using continuous COCs with low
dosages of progestagen (v3150 mg/month) and in
women using continuous COCs with high dosages
of progestagen (>3150 mg/month). Cyproterone
users were not included in this analysis. The
correlation was significant only in women using
continuous high dosages of progestagen (r50.298,
p50.012). Differences in median CRP values
between different progestagen users did not reach
statistical significance.
The effects of the haplotypes on CRP concentra-
tions were analysed separately in women according
to contraceptive use before and after adjustment of
metabolic and lifestyle factors. The results are
presented in Table IV (ANCOVA). In non-users,
all haplotypes had significant effects on CRP
Table II. Univariates and adjusted multiple linear regression model of (log)CRP in women without hormonal contraceptives
and with COCs.
Variable
Non-users (n5591) COC users (n5250)
Univariate Multivariate Univariate Multivariate
BSEpBSE pBSE pBSEp
Body mass index
(kg/m
2
)
0.053 ¡0.004 v0.001 0.028 0.005 v0.001 0.048 ¡0.008 v0.001 0.020 0.009 0.029
Waist circumference
(cm)
0.002 ¡0.001 v0.001 0.002 ¡0.001 v0.001
Age (years) 0.002 ¡0.004 0.542 0.000 ¡0.006 0.956
HDL-cholesterol
(mmol/L)
20.294 ¡0.070 v0.001 20.040 ¡0.097 0.680
LDL-cholesterol
(mmol/L)
0.101 ¡0.026 v0.001 0.014 ¡0.042 0.742
(log) Triglycerides
(mmol/L)
0.797 ¡0.107 v0.001 0.666 ¡0.169 v0.001 0.358 0.156 0.023
Systolic blood
pressure (mmHg)
0.007 ¡0.002 v0.001 0.006 ¡0.002 0.006
Diastolic blood
pressure (mmHg)
0.010 ¡0.002 v0.001 0.009 ¡0.003 0.003
(log) Insulin (mU/L) 0.697 ¡0.081 v0.001 0.440 ¡0.128 0.001
Glucose (mmol/L) 0.163 ¡0.045 v0.001 0.063 ¡0.065 0.331
(log) Leptin (mU/L) 0.869 ¡0.058 v0.001 0.557 0.080 v0.001 0.760 ¡0.101 v0.001 0.524 0.124 v0.001
Physical activity
index
20.004 ¡0.001 0.003 20.005 ¡0.002 0.050
Smoking (daily) 20.007 ¡0.046 0.879 20.123 ¡0.069 0.076
Alcohol
(drinks per week)
20.001 ¡0.003 0.863 20.007 ¡0.005 0.214
R
2
50.309 R
2
50.225
Table III. Median CRP and triglycerides levels compared to the type of progestagen included in the COCs.
Progestagen compound n
CRP Triglycerides Correlation*
Median Quartiles Median Quartiles rp
Gestodene 116 1.56 0.89–3.33 1.20 1.00–1.40 0.128 0.171
Desogestrel 60 1.90 0.79–3.26 1.30 1.00–1.70 0.210 0.106
Levonorgestrel 37 1.29 0.54–2.34 1.10 0.90–1.50 0.290 0.082
Cyproterone 41 2.00 0.52–4.06 1.60 0.90–2.00 0.508 0.001
Low-dosage continuous 96 1.56 0.89–3.31 1.20 1.00–1.50 0.042 0.686
High-dosage continuous
(without cyproterone)
71 2.02 0.82–3.43 1.30 1.00–1.80 0.298 0.012
*Spearman’s test for correlation between CRP and triglycerides.
Combined oral contraceptives and CRP 171
concentrations, the p-values ranging from v0.001 to
0.041. In COC users, the haplotypes had no signi-
ficant effects on CRP concentration. Further analy-
sis in relation to progestagen/oestrogen content or
dosage did not change the result.
Discussion
CRP has various roles in the development of
atherosclerosis. It can bind to modified LDL-
cholesterol particles, especially to the non-esterified
cholesterol in LDL [22], after which CRP-opsonized
LDL can be taken up by macrophages via CRP
receptor CD32 [23]. This can lead to increased foam-
cell formation in atherosclerotic plaques. CRP can
also induce the expression of adhesion molecules [24]
and inhibit nitric oxide expression in the human
endothelial cells [25], both of which facilitate the
atherosclerosis processes further. In addition, CRP
promotes apoptosis of endothelial progenitor cells
that are responsible for vascular regenerative poten-
tial [26].
Although the risk of both myocardial infarction
(MI) and venous thromboembolism (VTE) is gen-
erally low in young women, COC use is known to
increase the risk of both. This is especially true in the
case of women who have other cardiovascular risk
factors – environmental and genetic. However, the
risk of MI might be lower in women who use third-
generation oral contraceptives [10], while the risk of
VTE diseases might be higher in women who use
third-generation COCs [11]. Because CRP is one
possible pathological agent behind CVD, and COC
usage affects CRP levels, we assessed the effects of
COC usage and the different progestagen effects on
CRP in young Finnish women.
The data shown in this report demonstrate that
CRP and triglyceride levels are higher in COC users
than in non-users. Also, the determinants of plasma
CRP concentration are different in COC users than in
non-users. The most striking difference was seen in
triglycerides, i.e. in COC users there was a positive
association between triglyceride and CRP concen-
trations, while in non-users there was no association
at all. In COC users, the association was found only
in users of COCs containing cyproterone or other
high progestagen dosage COCs. To the best of our
knowledge, these findings are now described for the
first time.
It has previously been shown that COC use
increases both plasma triglyceride concentrations
[21,27] and CRP concentrations [28]; this was also
observed in the present study. It is known that COCs
increases CRP concentrations without increasing IL-
6 concentrations [29], suggesting that COCs stimu-
lates hepatocytes directly to synthesize CRP, and not
via IL-6-mediated inflammation. The mechanism of
how COCs stimulates hepatocytes protein synthesis is
still not known. However, the role of oestrogen might
be more important than that of progestagen, because
studies in postmenopausal women have shown that
oestrogen alone can induce CRP concentrations
[14,30], and in premenopausal women it has been
observed that pills containing only progestagen do
not elevate CRP concentrations [28]. The role of
progestagens on CRP concentration is still unclear.
Studies comparing different progestagen effects on
CRP suggest that the effect might depend on specific
progestaten content [13–15]. Our findings support the
idea that progestagens do have a role in CRP
regulation. In particular, the amount of progestagen
seems to be important. Although there were differ-
ences in correlation between CRP and triglycerides
depending on the progestagens that had been used,
there were no significant differences in CRP concen-
trations. The cyproterones are usually prescribed to
women with androgen excess, e.g. polycystic ovary
syndrome. This underlying disorder affects the CRP
[31] and triglyceride [32] levels, so we cannot rule out
that correlation between triglycerides and CRP
Table IV. Effect of CRP haplotype carriage on CRP levels before and after adjustment of metabolic and lifestyle factors.
Carriage of
haplotype
Non-users COC users
nMedian Quartiles p* Adjusted p** nMedian Quartiles p* Adjusted p**
ATGTG+308 0.59 0.28–1.36 124 1.65 0.72–3.44
ATGTG2210 0.43 0.23–0.96 0.008 0.004 94 1.71 0.82–3.27 0.776 0.688
ACGCA+261 0.48 0.24–1.05 103 1.67 0.80–3.33
ACGCA2257 0.56 0.28–1.35 0.131 0.041 115 1.67 0.89–3.25 0.572 0.391
GCGCG+195 0.41 0.20–0.86 87 1.87 0.89–3.18
GCGCG2323 0.66 0.28–1.49 v0.001 v0.001 131 1.54 0.79–3.33 0.584 0.419
Haplotypes are composed of SNPs 2717AwG, 2286CwTwA, +1059GwC, +1444CwT, 1846GwA. **Mann-Whitney test for
difference between carrier and non-carrier. *ANCOVA test with (log)CRP values: Non-users after adjustment of BMI, (log)leptin. COC
users after adjustment of BMI, (log)triglycerides, (log)leptin.
172 A. Haarala et al.
among cyproterone users is due to the underlying
disorder, because its frequency is unknown in this
cohort. To avoid this bias, all the analyses were also
performed without cyproterone users, but it did not
change the results. The other significant determinants
of plasma CRP in our data were BMI and leptin.
They had similar effects on CRP in COC users and
non-users, so it seems that the IL-6 mediated
stimulation of CRP by BMI [33] and leptin [34] is
not affected by COC use.
In addition to the changes in the role of metabolic
factors in the regulation of CRP concentrations, we
found that the contribution of genetic control on
CRP concentrations was entirely different between
COC users and non-users. Haplotypes of the CRP
gene were associated with CRP concentrations in
non-users, but not in COC users. At present, the
molecular background of this is not known.
However, hepatic induction of CRP gene via IL-6
and without IL-6 is different; the CRP gene IL-6
responsive elements are probably involved in the
former but not in the latter. In other words, the CRP
production induced or enhanced by COCs is pre-
sumably regulated by different transcription factors
from the IL-6 mediated CRP production. Of the 5
SNPs used for construction of the haplotypes, two
are promoter region polymorphisms (2717 and
2286), one is exonic (+1059) and two are in the
39UTR region (+1444 and +1846), and there is no
obvious steroid receptor binding sequences in these
positions. Therefore, understanding the molecular
mechanism of this observed difference requires a
more thorough analysis of the transcriptional control
of this gene. The haplotypes analysed can explain
only about 5 % of the circulating CRP [9], so it is
possible that the genetic effect is overwhelmed by the
strong COC effect.
The most limiting factor in this study was its
cross-sectional design. There was a relatively large
variation in the use of different COCs with different
contents and amounts of oestrogen and progestagens,
which made it difficult to create comparable and
representative subgroups in relation to COC usage.
In order to keep comparison clear and representative
in both groups, the main comparison was done
between COC users and non-users. There were also
other disadvantages. The use of contraceptives was
elicited only by questionnaire; COC users and non-
users might not be comparable in every respect. For
example, there might be some infertile women who
do not need COCs and also women wishing to
become pregnant and therefore not using COCs.
There could also be other reasons possibly altering
the use of COCs that we could not account for,
such as temporary intermission due to broken
relationships, inconstant use of COCs (time and
brand, etc.). Unfortunately we were not able to
analyse all of these confounders, but still we believe
that these cases are low in number and do not affect
our results. Despite these limitations, we believe that
this study included a relatively large and representa-
tive number of young women (n5841) and that our
conclusions are valid.
In conclusion, our findings support a direct role
for COCs in CRP determination by affecting its
metabolic and genetic regulation. These findings
highlight the importance of thorough adjustment of
CRP concentration with confounding variables in
the analysis of genetic polymorphism, i.e. COC
usage diminishes the effect of CRP genetics on CRP
concentration. Future research of COC usage-
related changes in CRP determination and genetic
regulation are needed, especially random controlled
trials, and the clinical relevance of the findings has to
be seen.
Acknowledgements
We thank Sinikka Repo-Koskinen, Eija Spa˚re and
Nina Peltonen for their skilful technical assistance
and Heini Huhtala for her help with statistical
problems. The study was financially supported by
the Tampere University Central Hospital Medical
Fund, the Emil Aaltonen Foundation (to T.L.), the
Tampere Tuberculosis Foundation, the Academy of
Finland (grant nos. 117941, 77841, 210283), the Juha
Vainio Foundation, the Finnish Foundation of
Cardiosvascular Research, the Finnish Cultural
Foundation and Special Federal Grants for the
Turku University Central Hospital.
Declaration of interest: The authors report no
conflicts of interest. The authors alone are respon-
sible for the content and writing of the paper.
References
[1] Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive
protein and other markers of inflammation in the prediction
of cardiovascular disease in women. N Engl J Med
2000;342:836–43.
[2] Danesh J, Wheeler JG, Hirschfield GM, Eda S, Eiriksdottir G,
Rumley A, et al. C-reactive protein and other circulating
markers of inflammation in the prediction of coronary heart
disease. N Engl J Med 2004;350:1387–97.
[3] Kushner I, Rzewnicki D, Samols D. What does minor
elevation of C-reactive protein signify? Am J Med
2006;119:166 e117–28.
[4] Wilson AM, Ryan MC, Boyle AJ. The novel role of C-reactive
protein in cardiovascular disease: risk marker or pathogen. Int
J Cardiol 2006;106:291–7.
Combined oral contraceptives and CRP 173
[5] MacGregor AJ, Gallimore JR, Spector TD, Pepys MB.
Genetic effects on baseline values of C-reactive protein and
serum amyloid a protein: a comparison of monozygotic and
dizygotic twins. Clin Chem 2004;50:130–4.
[6] Hage FG, Szalai AJ. C-reactive protein gene polymorphisms,
C-reactive protein blood levels, and cardiovascular disease
risk. J Am Coll Cardiol 2007;50:1115–22.
[7] Carlson CS, Aldred SF, Lee PK, Tracy RP, Schwartz SM,
Rieder M, et al. Polymorphisms within the C-reactive protein
(CRP) promoter region are associated with plasma CRP
levels. Am J Hum Genet 2005;77:64–77.
[8] Eklund C, Kivimaki M, Islam MS, Juonala M, Kahonen M,
Marniemi J, et al. C-reactive protein genetics is associated with
carotid artery compliance in men in The Cardiovascular Risk
in Young Finns Study. Atherosclerosis 2008;196:841–8.
[9] Kivimaki M, Lawlor DA, Smith GD, Eklund C, Hurme M,
Lehtimaki T, et al. Variants in the CRP gene as a measure of
lifelong differences in average C-reactive protein levels: the
Cardiovascular Risk in Young Finns Study, 1980–2001. Am J
Epidemiol 2007;166:760–4.
[10] Tanis BC, van den Bosch MA, Kemmeren JM, Cats VM,
Helmerhorst FM, Algra A, et al. Oral contraceptives and the
risk of myocardial infarction. N Engl J Med 2001;345:
1787–93.
[11] Battaglioli T, Martinelli I. Hormone therapy and thromboem-
bolic disease. Curr Opin Hematol 2007;14:488–93.
[12] Raitakari M, Mansikkaniemi K, Marniemi J, Viikari JS,
Raitakari OT. Distribution and determinants of serum high-
sensitive C-reactive protein in a population of young adults:
The Cardiovascular Risk in Young Finns Study. J Intern Med
2005;258:428–34.
[13] Skouby SO, Gram J, Andersen LF, Sidelmann J, Petersen KR,
Jespersen J. Hormone replacement therapy: estrogen and
progestin effects on plasma C-reactive protein concentrations.
Am J Obstet Gynecol 2002;186:969–77.
[14] Kluft C, Leuven JA, Helmerhorst FM, Krans HM. Pro-
inflammatory effects of oestrogens during use of oral contra-
ceptives and hormone replacement treatment. Vascul
Pharmacol 2002;39:149–54.
[15] Buchbinder S, Kratzsch J, Fiedler GM, Yar V, Brugel M,
Leichtle A, et al. Body weight and oral contraceptives are the
most important modulators of serum CRP levels. Scand J Clin
Lab Invest 2008;68:140–4.
[16] Akerblom HK, Viikari J, Uhari M, Rasanen L, Byckling T,
Louhivuori K, et al. Atherosclerosis precursors in Finnish
children and adolescents. I. General description of the cross-
sectional study of 1980, and an account of the children’s and
families’ state of health. Acta Paediatr Scand Suppl 1985;
318:49–63.
[17] Juonala M, Viikari JS, Hutri-Kahonen N, Pietikainen M,
Jokinen E, Taittonen L, et al. The 21-year follow-up of the
Cardiovascular Risk in Young Finns Study: risk factor levels,
secular trends and east–west difference. J Intern Med
2004;255:457–68.
[18] Viikari LA, Huupponen RK, Viikari JS, Marniemi J, Eklund
C, Hurme M, et al. Relationship between leptin and C-reactive
protein in young Finnish adults. J Clin Endocrinol Metab
2007;92:4753–8.
[19] Telama R, Yang X, Viikari J, Valimaki I, Wanne O, Raitakari
O. Physical activity from childhood to adulthood: a 21-year
tracking study. Am J Prev Med 2005;28:267–73.
[20] Stephens M, Smith NJ, Donnelly P. A new statistical method
for haplotype reconstruction from population data. Am J
Hum Genet 2001;68:978–89.
[21] Sitruk-Ware RL, Menard J, Rad M, Burggraaf J, de Kam
ML, Tokay BA, et al. Comparison of the impact of vaginal
and oral administration of combined hormonal contraceptives
on hepatic proteins sensitive to estrogen. Contraception
2007;75:430–7.
[22] Taskinen S, Kovanen PT, Jarva H, Meri S, Pentikainen MO.
Binding of C-reactive protein to modified low-density-
lipoprotein particles: identification of cholesterol as a novel
ligand for C-reactive protein. Biochem J 2002;367:403–12.
[23] Zwaka TP, Hombach V, Torzewski J. C-reactive protein-
mediated low density lipoprotein uptake by macrophages:
implications for atherosclerosis. Circulation 2001;103:1194–7.
[24] Pasceri V, Willerson JT, Yeh ET. Direct proinflammatory
effect of C-reactive protein on human endothelial cells.
Circulation 2000;102:2165–8.
[25] Verma S, Wang CH, Li SH, Dumont AS, Fedak PW,
Badiwala MV, et al. A self-fulfilling prophecy: C-reactive
protein attenuates nitric oxide production and inhibits
angiogenesis. Circulation 2002;106:913–19.
[26] Verma S, Kuliszewski MA, Li SH, Szmitko PE, Zucco L,
Wang CH, et al. C-reactive protein attenuates endothelial
progenitor cell survival, differentiation, and function: further
evidence of a mechanistic link between C-reactive protein and
cardiovascular disease. Circulation 2004;109:2058–67.
[27] Godsland IF. Biology: risk factor modification by OCs and
HRT lipids and lipoproteins. Maturitas 2004;47:299–303.
[28] Williams MJ, Williams SM, Milne BJ, Hancox RJ, Poulton R.
Association between C-reactive protein, metabolic cardio-
vascular risk factors, obesity and oral contraceptive use in
young adults. Int J Obes Relat Metab Disord 2004;28:
998–1003.
[29] van Rooijen M, Hansson LO, Frostegard J, Silveira A,
Hamsten A, Bremme K. Treatment with combined oral
contraceptives induces a rise in serum C-reactive protein in
the absence of a general inflammatory response. J Thromb
Haemost 2006;4:77–82.
[30] Pradhan AD, Manson JE, Rossouw JE, Siscovick DS,
Mouton CP, Rifai N, et al. Inflammatory biomarkers,
hormone replacement therapy, and incident coronary heart
disease: prospective analysis from the Women’s Health
Initiative observational study. J Am Med Assoc
2002;288:980–7.
[31] Kelly CC, Lyall H, Petrie JR, Gould GW, Connell JM, Sattar
N. Low grade chronic inflammation in women with polycystic
ovarian syndrome. J Clin Endocrinol Metab 2001;86:2453–5.
[32] Valkenburg O, Steegers-Theunissen RP, Smedts HP, Dallinga-
Thie GM, Fauser BC, Westerveld EH, et al. A more
atherogenic serum lipoprotein profile is present in women
with polycystic ovary syndrome: a case-control study. J Clin
Endocrinol Metab 2008;93:470–6.
[33] Banks RE, Forbes MA, Storr M, Higginson J, Thompson D,
Raynes J, et al. The acute phase protein response in patients
receiving subcutaneous IL-6. Clin Exp Immunol 1995;102:
217–23.
[34] Santos-Alvarez J, Goberna R, Sanchez-Margalet V. Human
leptin stimulates proliferation and activation of human
circulating monocytes. Cell Immunol 1999;194:6–11.
174 A. Haarala et al.
