European Journal of Cancer 124 (2020) 102e109

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.ejcancer.com

Original Research

Oral contraceptive and intrauterine device use and the

risk of cervical intraepithelial neoplasia grade III or
worse: a population-based study

Diede L. Loopik a,*, Joanna IntHout b, Willem J.G. Melchers c,

Leon F.A.G. Massuger d, Ruud L.M. Bekkers e,1, Albert G. Siebers f,g

a
Department of Obstetrics and Gynaecology, Radboud Institute for Molecular Life Sciences, Radboud university medical
center, PO Box 9101, 6500HB, Nijmegen, the Netherlands
b
Department of Biostatistics, Radboud Institute for Health Sciences, Radboud university medical center, PO Box 9101,
6585KM, Nijmegen, the Netherlands
c
Department of Medical Microbiology, Radboud university medical center, PO Box 9101, 6500HB, Nijmegen, the
Netherlands
d
Department of Obstetrics and Gynaecology, Radboud university medical center, PO Box 9101, 6500HB, Nijmegen, the
Netherlands
e
Department of Obstetrics and Gynaecology, Catharina Hospital, PO Box 1350, 5602ZA, Eindhoven, the Netherlands
f
Department of Pathology, Radboud university medical center, PO Box 9101, 6500HB, Nijmegen, the Netherlands
g
PALGA, Randhoeve 225a, 3995 GA, Houten, the Netherlands

Received 2 September 2019; accepted 10 October 2019

Available online 21 November 2019

KEYWORDS Abstract Objective: Hormonal contraceptive use has been associated with the development
Cervical intraepithelial of cervical cancer, although inconsistent results are reported on the association with intrauter-
neoplasia; ine device (IUD) use. The aim of this study was to evaluate the association between the type of
Contraceptives; contraceptive use and the development of cervical intraepithelial neoplasia grade III or worse

Abbreviations: ASC-H, atypical squamous cells e cannot rule out high-grade squamous intraepithelial lesion; ASC-US, atypical squamous cells of
undetermined significance; CIN, cervical intraepithelial neoplasia; hrHPV, high-risk human papillomavirus; HSIL, high-grade squamous intra-
epithelial lesion; IUD, intrauterine device; LSIL, low-grade squamous intraepithelial lesion; NILM, negative for intraepithelial lesion or malig-
nancy; NOS, not otherwise specified; OC, oral contraceptive; PALGA, the nationwide network and registry of histo- and cytopathology in the
Netherlands.
* Corresponding author: Department of Obstetrics and Gynaecology, Radboud Institute for Molecular Life Sciences, Radboud university medical
center, PO Box 9101, 6500HB, Nijmegen, the Netherlands. Fax: þ31 243668597.
E-mail addresses: diede.loopik@radboudumc.nl (D.L. Loopik), Joanna.intHout@radboudumc.nl (J. IntHout), willem.melchers@radboudumc.
nl (W.J.G. Melchers), leon.massuger@radboudumc.nl (L.F.A.G. Massuger), ruud.bekkers@radboudumc.nl (R.L.M. Bekkers), bert.siebers@
radboudumc.nl (A.G. Siebers).
1
Present address: GROW, School for Oncology & Developmental Biology, Maastricht University Medical Centre, PO Box 616, 6200MD,
Maastricht, the Netherlands.

https://doi.org/10.1016/j.ejca.2019.10.009
0959-8049/ª 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
 D.L. Loopik et al. / European Journal of Cancer 124 (2020) 102e109 103

(CIN3þ).
Follow-up studies;
Methods: A retrospective population-based cohort study including women aged 29e44 years
Intrauterine devices;
attending the cervical cancer screening program with normal cytology between 2005 and 2009
Risk assessment;
identified from the Dutch Pathology Registry. Subgroups with at least 5 years registered use of
Retrospective studies;
an oral contraceptive (OC) or IUD were compared with non-users. Risk ratios of CIN3þ were
Uterine cervical
estimated per contraceptive type.
neoplasms
Results: 702,037 women were included with a median follow-up of 9.7 years, of which 6705
(0.96%) and 559 (0.08%) women developed CIN3 and cervical cancer, respectively. IUD use
was associated with an increased risk of developing CIN3þ (risk ratio (RR) 1.51, 95% confi-
dence interval (CI) 1.32e1.74), and OC use was associated with an increased risk of devel-
oping CIN3þ (RR 2.77, 95%CI 2.65e3.00) and cervical cancer (RR 2.06, 95%CI 1.52
e2.79). The risk of developing CIN3þ and cervical cancer was higher for OC users compared
with IUD users (RR 1.83, 95%CI 1.60e2.09 and RR 1.70, 95%CI 1.00e2.90, respectively).
Conclusions: Both OC use and IUD use were associated with an increased risk of developing
CIN3þ. However, for women with a contraceptive wish, an IUD seems safer than an OC as
the risk of developing CIN3þ and cervical cancer was higher for OC users.
ª 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction development of cervical intraepithelial neoplasia grade

Cervical cancer is caused by persistence of a common
sexually transmitted high-risk human papillomavirus
(hrHPV) infection [1]. The lifetime risk to acquire an
HPV infection is ~80%, but the majority are transient
infections. Low-grade cervical lesions generally repre-
sent productive hrHPV infection with a low risk of
progression to cervical cancer. Of the persistent in-
fections, a minority becomes a transforming infection
that may develop into a high-grade cervical lesion and
cervical cancer, especially after integration of hrHPV
into the host genome [2,3]. Prophylactic vaccination in
conjunction with cervical screening, followed by
removal of premalignant cervical lesions that are likely
to become malignant, are preventive measures for
developing cervical cancer.
HrHPV is a necessary, although not sufficient,
cause of cervical cancer. Tobacco smoking, infection
with other sexually transmitted diseases, and hor-
monal factors, including hormonal contraception and
multiparity, are identified as cofactors for the devel-
opment of high-grade cervical lesions and cervical
cancer [4,5]. Although previous studies have shown an
association with the use of hormonal contraceptives
and risk of cervical cancer, inconsistent results are
reported on the association with intrauterine device
(IUD) use [6e12]. A pooled analysis of 26 epidemio-
logical studies from Castellsagué et al. even suggests
that IUD use is a protective cofactor in the risk of
cervical cancer [8]. However, these data are mostly
from older studies, and more recent epidemiological
studies on the effect of different contraceptives on
cervical carcinogenesis are missing.
The aim of this study was to evaluate the association
between the type of contraceptive use and the
III or worse in women attending the nationwide cervical
cancer screening program in the Netherlands.
2. Material and methods
Women aged 30e60 years are invited at 5-year intervals
for the Dutch nationwide cervical cancer screening
program. This population-based retrospective cohort
study included women aged 29e44 years who partici-
pated in this screening program between January 1,
2005, and December 31, 2009 in the Netherlands with a
normal cervical smear and at least 5 years of follow-up.
These data were obtained from the Dutch nationwide
registry of histopathology and cytopathology (PALGA;
Houten, the Netherlands). The contraceptive use status
is a standard question on the cervical screening form,
with ‘no contraceptive use’, ‘oral contraceptive (OC)
use’, ‘IUD use’ (copper or levonorgestrel-releasing IUDs
not further specified), ‘other’ (for example the contra-
ceptive skin patch or subdermal contraceptive implant)
or ‘unknown’ as answer options.
All subsequent results from cervical cytology and
histology were retrieved during follow-up until
December, 31, 2017, including the date and the contra-
ceptive use status if noted, resulting in up to 13 years of
follow-up. Follow-up ended earlier if women received an
excisional procedure of the cervix. Women with a pre-
ceding low-grade or worse cytologic result (atypical
squamous cells of undetermined significance) or histo-
logic result (cervical intraepithelial neoplasia (CIN)1)
between January 1, 2000 and December 31, 2004, or
within 1 year of follow-up, were excluded from analysis.
Women with a high-grade cytologic result (high-grade
squamous intraepithelial lesion) or histologic result
(CIN2) within the first 5 years of follow-up were also
104 D.L. Loopik et al. / European Journal of Cancer 124 (2020) 102e109
excluded to really evaluate the effect of the use of a
contraceptive for at least 5 years. Three subgroups were
selected: (I) women who reported no use of any con-
traceptive type, (II) women who reported the use of an
OC and (III) women who reported the use of an IUD
for all subsequent follow-up visits for at least 5 unin-
terrupted years from date of inclusion. The subgroups
OC and IUD users were primarily compared with non-
users, but a comparison with the total group of included
women was also made. The highest cervical abnormality
during follow-up was identified per subgroup. The
highest histologic result was used as primary outcome.
The cytological classification was performed according
to the Dutch CISOE-A classification. For analysis, the
CISOE-A classification system was translated into the
Bethesda nomenclature [13]. The histologic specimens
were classified according to the CIN histologic grading
system.
The Chi-squared test was used to compare the dis-
tribution of categorical variables. Binary logistic
regression with adjustment for age was used to estimate
risk ratios (RRs) and 95% confidence intervals (CIs) for
the risk of developing CIN3þ and cervical cancer ac-
cording to contraceptive type. All statistical analyses
were performed with SAS, version 9.2 (SAS Institute,
Cary, NC).
The study was approved by the scientific committee
of PALGA. The study was exempt from institutional
review board approval because data were gathered
retrospectively and analysed anonymously. There was
no patient or public involvement in this study.
Fig. 1. Description of study population; inclusions and exclusions.
3. Results
From the 1,101,365 identified women, 702,037
women were included in the study. The exclusions
and distribution between contraceptive type use in
the included women is shown in Fig. 1. 25% re-
ported explicitly no use of any type of contraceptive,
12% reported the use of an OC and 4% the use of an
IUD for at least 5 uninterrupted years. The baseline
characteristics and highest cervical abnormality
diagnosed during follow-up of the three subgroups
and the total group of included women are listed in
Table 1. The age distribution between the subgroups
was different, as non-users of contraceptives were
older than women who used an OC or IUD (median
39 versus 35 versus 35 years). The number of follow-
up visits and the length of follow-up did not notably
differ between the subgroups, with a median of two
follow-up visits and a median follow-up period of
9.7 years. The majority of women (91.4%) did not
develop any cervical abnormality during follow-up,
whereas 6.6% developed a low-grade lesion and
1.9% a high-grade lesion. The majority of high-grade
lesions were histologically proven (94.6%), with 5547
(0.79%), 6705 (0.96%) and 559 (0.08%) of all
included women developing histologically proven
CIN2, CIN3 and cervical cancer, respectively. Fig. 2
shows the distribution between the cumulative inci-
dence of CIN2, CIN3 and cervical cancer per con-
traceptive type. The main difference between the
IUD and OC group was found between the
 D.L. Loopik et al. / European Journal of Cancer 124 (2020) 102e109 105

Table 1
Baseline characteristics and highest cervical abnormality during follow-up per type of contraceptive.
Parameters No contraceptive usej Oral contraceptive usek Intrauterine device usel Total included womenm
for 5 years n Z 178,545 for 5 years n Z 85,823 for 5 years n Z 27,509 N Z 702,037 (100%)
(25.4%) (12.2%) (3.9%)
Age, years
Median (range) 39 (29e44) 35 (29e44) 35 (29e44) 35 (29e44)
Age group, years, No. (%)
29 9213 (5.2) 9222 (10.8) 1708 (6.2) 54,498 (7.8)
30e34 47,572 (26.6) 31,519 (36.7) 9362 (34.0) 230,447 (32.8)
35e39 52,064 (29.2) 24,265 (28.3) 8860 (32.2) 196,419 (28.0)
40e44 69,696 (39.0) 20,817 (24.3) 7579 (27.6) 220,673 (31.4)
Follow-up, years
Median (range) 9.6 (5.0e13.1) 9.7 (5.0e13.2) 9.4 (5.0e13.1) 9.7 (5.0e13.2)
Cytologya, No. (%)
Normal b 162,624 (91.1) 76,405 (89.0) 24,822 (90.2) 618,145 (88.1)
Low-gradec 8451 (4.7) 4884 (5.7) 1438 (5.2) 38,375 (5.5)
High-graded 196 (0.1) 91 (0.1) 29 (0.1) 728 (0.1)
Unknowne 184 (0.1) 90 (0.1) 15 (0.1) 531 (0.1)
Total 171,455 (96.0) 81,470 (94.9) 26,304 (95.6) 657,779 (93.7)
Histologyf, No. (%)
Normal 4042 (2.3) 1220 (1.4) 410 (1.5) 23,479 (3.3)
Low-gradeg 1210 (0.7) 858 (1.0) 292 (1.1) 7865 (1.1)
High-gradeh 1823 (1.0) 2270 (2.6) 502 (1.8) 12,811 (1.8)
Unknowni 15 (0.0) 5 (0.0) 1 (0.0) 103 (0.0)
Total 7090 (4.0) 4353 (5.1) 1205 (4.4) 44,258 (6.3)
ASC-H: atypical squamous cellsdcannot rule out high-grade squamous intraepithelial lesion; ASC-US: atypical squamous cells of undetermined
significance; CIN: cervical intraepithelial neoplasia; hrHPV: high-risk human papillomavirus; HSIL: high-grade squamous intraepithelial lesion;
LSIL: low-grade squamous intraepithelial lesion; NILM: negative for intraepithelial lesion or malignancy; NOS: not otherwise specified.
a
Highest cytologic result during follow-up in case of absent histologic results.
b
Normal: including NILM and negative hrHPV result.
c
Low-grade: including hrHPV positive result, ASC-US and LSIL.
d
High-grade: including ASC-H and HSIL.
e
Unknown: including inadequate cytology.
f
Highest histologic result during follow-up.
g
Low-grade: including CINeNOS and CIN1.
h
High-grade: including CIN2, CIN3 and cervical cancer.
i
Unknown: including missing data.
j
Women who did not use any type of contraceptive for at least five uninterrupted years during follow-up.
k
Women who used an oral contraceptive for at least five uninterrupted years during follow-up.
l
Women who had an intrauterine device for at least five uninterrupted years during follow-up.
m
All women included in the study, independent from contraceptive use.

cumulative incidence of CIN2 and CIN3. The cu-
mulative incidence of CIN2 was similar (0.94% and
0.92%, respectively; p0.41), whereas the cumulative
incidence of CIN3 was significantly higher in the OC
group (0.83% and 1.62%, respectively; p < 0.01).
The RR of CIN3þ stratified by age and contraceptive
type is shown in Fig. 3. Older age in the IUD group was
associated with a higher risk of developing CIN3þ,
therefore the age-adjusted RR per contraceptive type
has been estimated (Fig. 4). The use of an IUD was
associated with an increased risk of developing CIN3þ
(RR 1.51, 95%CI 1.32e1.74) but did not significantly
influence the risk of cervical cancer (RR 1.21, 95%CI
0.71e2.07). The use of an OC was associated with an
increased risk of developing CIN3þ (RR 2.77, 95%CI
2.65e3.00) and cervical cancer (RR 2.06, 95%CI
1.52e2.79). The risk of developing CIN3þ and cervical
cancer was higher for OC users compared with IUD
users (RR 1.83, 95%CI 1.60e2.09 and RR 1.70, 95%CI
1.00e2.90, respectively). Women who did not use any
type of contraceptive had a significantly lower risk of
developing CIN3þ (RR 0.61, 95%CI 0.58e0.64) and
cervical cancer (RR 0.63, 95%CI 0.53e0.75) compared
with the total group of included women. When strati-
fying cervical cancer by histological type, 347 women
(62.1%) had a squamous cell carcinoma, 183 women
(32.7%) an adenocarcinoma and 29 women (5.2%) had
another type of cervical cancer.
4. Discussion
This study, including 702,037 screened women with a
lowerisk profile for the development of cervical ab-
normalities, confirms that OC use is associated with an
increased risk of developing CIN3þ, but in contrast to
previous studies, we also found this increased risk in
IUD users compared with non-users. The risk of
106 D.L. Loopik et al. / European Journal of Cancer 124 (2020) 102e109
Fig. 2. Cumulative incidence and 95%CI of CIN2, CIN3 and
cervical cancer stratified by contraceptive type with at least five
years of use. Note: All p-values were <0.01 between any subgroup,
unless otherwise specified. CI: confidence interval; CIN: cervical
intraepithelial neoplasia.
developing CIN3þ was higher for OC users compared
with IUD users.
Two meta-analyses did find a protective effect of
IUD use; however, they could only compare ever with
never users [8,11]. The group of never users consists of a
heterogeneous group, including non-users and other
type of contraceptive users, which makes it hard to
interpret the results. We compared IUD users with non-
Fig. 3. Risk ratio and 95%CI of CIN3þ stratified by age. CI: confidence interval; CIN: cervical intraepithelial neoplasia; RR: risk ratio.
users to rule out the effect of other hormonal contra-
ceptives, as the total group of women is a mixture of
women with all types and durations of contraceptive use
and women without the use of a contraceptive. A po-
tential confounder could be a difference in sexual
behaviour between non-users and contraceptive users. A
study about potential confounding effects did find a
difference in sexual behaviour in non-users compared
with contraceptive users. Interestingly enough though,
the hrHPV status did not differ between the groups with
or without the use of contraceptives [14]. If contracep-
tive use was only a surrogate marker for sexual behav-
iour and this behaviour is the true risk factor for
developing CIN, one would expect to see a difference in
the acquisition or persistence of hrHPV infection.
Only IUD use in women aged 30 years was asso-
ciated with a significantly higher risk of developing
CIN3þ. This could be based on the duration of IUD
use; however, the observed increased risk in IUD users
30 years might be because of the lingering effects of
previous use of OCs.
The studies included in the meta-analyses were mostly
older studies, which suggests these women were more
likely copper-IUD users. One study has suggested that
the effect on hrHPV biology may be different, in which
copper-IUD users were more likely to clear hrHPV in-
fections than levonorgestrel-releasing IUD users [15]. A
caseecontrol study with recent IUD use also found a
trend toward a protective effect of copper-IUDs, but an
increased risk of CIN2þ for levonorgestrel-releasing
IUDs [12]. Most Dutch IUD users >2005 were likely
using levonorgestrel-releasing IUDs.
 D.L. Loopik et al. / European Journal of Cancer 124 (2020) 102e109
Fig. 4. Risk ratio and 95%CI of CIN3þ and cervical cancer adjusted for age. CI: confidence interval; CIN: cervical intraepithelial
neoplasia; RR: risk ratio.
A possible mechanism that could explain the poten-
tial protective effect of an IUD is through a device-
related inflammatory response in the endocervical canal
[8]. Copper-IUDs release ions which increase prosta-
glandin levels causing chronic inflammation, while
levonorgestrel-releasing IUDs decrease prostaglandin
levels suppressing the local immunity [15]. In contrast to
combined hormonal contraceptive use, progesterone
only may have a protective effect by increasing the
number of Langerhans cells, which are important for
immunosurveillance [16]. However, progesterone has
also been shown to increase E6/E7 oncogene transcrip-
tion in cell lines with integrated HPV-16 [17].
A meta-analysis of 28 studies found a two-fold risk of
developing cervical cancer in women with long duration
of hormonal contraceptives, also after adjustment for
HPV status, smoking status and sexual behaviour [6].
Another study with 17,032 women found a persistent
increased risk for cervical cancer for many years after
cessation of use [18]. However, a study about potential
confounding effects concluded that sexual behaviour is
different among OC users and non-users, and they could
not confirm that the use of OCs was an independent risk
factor for developing CIN [14]. The decision to use or
107
not to use a type of contraceptive could be a reflection of
adopted sexual behaviour. Unfortunately, we could not
adjust for sexual behaviour, hrHPV status or other po-
tential confounders, such as smoking status, as this was
a retrospective quantitative study. Therefore, contra-
ceptive use could be a surrogate marker for a certain
sexual behaviour pattern.
In our study, the risk of developing CIN3þ was
higher for OC users compared with IUD users. A similar
sexual behaviour pattern has been found in women with
OC use and the use of another type of contraceptive [14].
The main difference between the IUD and OC group
was found between the cumulative incidence of CIN2
and CIN3. The cumulative incidence of CIN2 was
similar, whereas the cumulative incidence of CIN3 was
significantly higher in the OC group (Fig. 2). These data
may show that IUD use is more associated with pro-
ductive and transient hrHPV infections and OC use with
transforming and integrated hrHPV infections. Com-
bined hormonal contraceptives may stimulate the inte-
gration of hrHPV-DNA into the host genome, which
enhances the expression of the E6 and E7 HPV onco-
proteins. These oncoproteins stimulate the degradation
of P53 tumour suppressor genes and enhance the ability
108 D.L. Loopik et al. / European Journal of Cancer 124 (2020) 102e109
of the viral DNA to transform cells and induce carci-
nogenesis [16,19]. This could explain the similar preva-
lence of hrHPV between the different subgroups, as a
DNA hrHPV test does not differ between integrated and
not integrated hrHPV.
HrHPV positive women with a contraceptive wish
should be counseled about the possible positive and
negative side-effects of the different contraception op-
tions and may be counseled against long-term use of an
OC and may benefit from an IUD. On the other hand,
OCs may have beneficial effects on other type of can-
cers, such as ovarian, endometrial and colorectal cancer
[20,21]. These known health benefits should be weighed
against the increased risk of developing cervical cancer.
Nonetheless, all women should be encouraged to
participate in the cervical screening program and further
research about the benefits of intensified screening in
women with long-term use of an OC is warranted.
Beside the lack of information about other behav-
ioural factors, the selection of subgroups with at least 5
years of reported use of one type of contraception is not
a hundred percent certain as most women participate in
the screening program every 5 years, and women may
have changed their use of contraceptive type tempo-
rarily. However, the differences between the subgroups
were statistically significant, regardless of potential
misreporting.
The strengths of our study includes its population-
based design with a large sample size of women with a
lowerisk profile for the development of cervical ab-
normalities with a median follow-up of 9.7 years, the
differentiation between the high-grade outcomes: CIN2,
CIN3 and cervical cancer and the comparison of women
with OC use and IUD use to non-users.
Studies focusing on behavioural patterns (including
contraceptive use), the incidence and persistence of hrHPV
infection and progression of CIN are needed. Also studies
exploring the biological role of OCs and IUDs on acqui-
sition, persistence and integration of hrHPV infections are
important to fully understand the association.
5. Conclusions
Both OC use and IUD use were associated with an
increased risk of developing CIN3þ compared with
non-users. However, for women with a contraceptive
wish, an IUD seems safer than an OC as the risk of
developing CIN3þ and cervical cancer was higher for
OC users.
Submission declaration and verification
This manuscript has not been published previously and
is not under consideration for publication elsewhere,
and this publication is approved by all authors and the
responsible authorities where the work has been carried
out. This manuscript will not be published elsewhere in
the same form.
Role of the funding source
This research did not receive any specific grant from
funding agencies in the public, commercial or not-for-
profit sectors.
Ethics approval
All the authors report adherence to ethical standards in
the conception of the work, data collection and writing
of the manuscript. The study was approved by the sci-
entific committee of the Dutch Pathology Registry
(PALGA). The study was exempt from institutional
review board approval because data were gathered
retrospectively and analysed anonymously.
Conflict of interest statement
None declared.
Acknowledgements
None.
References
[1] Bosch FX, Lorincz A, Munoz N, Meijer CJ, Shah KV. The causal
relation between human papillomavirus and cervical cancer. J
Clin Pathol 2002;55(4):244e65.
[2] Einstein MH, Schiller JT, Viscidi RP, Strickler HD, Coursaget P,
Tan T, et al. Clinician’s guide to human papillomavirus immu-
nology: knowns and unknowns. Lancet Infect Dis 2009;9(6):
347e56. https://doi.org/10.1016/s1473-3099(09)70108-2.
[3] Pett M, Coleman N. Integration of high-risk human papilloma-
virus: a key event in cervical carcinogenesis? J Pathol 2007;212(4):
356e67. https://doi.org/10.1002/path.2192.
[4] Castellsague X, Bosch FX, Munoz N. Environmental co-factors
in HPV carcinogenesis. Virus Res 2002;89(2):191e9.
[5] Munoz N, Castellsague X, de Gonzalez AB, Gissmann L. Chapter
1: HPV in the etiology of human cancer. Vaccine 2006;24(Suppl
3). https://doi.org/10.1016/j.vaccine.2006.05.115. S3/1-10.
[6] Smith JS, Green J, Berrington de Gonzalez A, Appleby P, Peto J,
Plummer M, et al. Cervical cancer and use of hormonal contra-
ceptives: a systematic review. Lancet (London, Engl) 2003;
361(9364):1159e67.
[7] Averbach SH, Ma Y, Smith-McCune K, Shiboski S,
Moscicki AB. The effect of intrauterine devices on acquisition and
clearance of human papillomavirus. Am J Obstet Gynecol 2017;
216(4):386. https://doi.org/10.1016/j.ajog.2016.11.1053. e1 - e5.
[8] Castellsague X, Diaz M, Vaccarella S, de Sanjose S, Munoz N,
Herrero R, et al. Intrauterine device use, cervical infection with
human papillomavirus, and risk of cervical cancer: a pooled
analysis of 26 epidemiological studies. Lancet Oncol 2011;12(11):
1023e31. https://doi.org/10.1016/s1470-2045(11)70223-6.
[9] Roura E, Travier N, Waterboer T, de Sanjose S, Bosch FX,
Pawlita M, et al. The influence of hormonal factors on the risk of
developing cervical cancer and pre-cancer: results from the EPIC
 D.L. Loopik et al. / European Journal of Cancer 124 (2020) 102e109
cohort. PLoS One 2016;11(1):e0147029. https://doi.org/10.1371/
journal.pone.0147029.
[10] Jensen KE, Schmiedel S, Norrild B, Frederiksen K, Iftner T,
Kjaer SK. Parity as a cofactor for high-grade cervical disease
among women with persistent human papillomavirus infection: a
13-year follow-up. Br J Canc 2013;108(1):234e9. https:
//doi.org/10.1038/bjc.2012.513.
[11] Cortessis VK, Barrett M, Brown Wade N, Enebish T, Perrigo JL,
Tobin J, et al. Intrauterine device use and cervical cancer risk: a
systematic review and meta-analysis. Obstet Gynecol 2017;130(6):
1226e36. https://doi.org/10.1097/aog.0000000000002307.
[12] Averbach S, Silverberg MJ, Leyden W, Smith-McCune K, Raine-
Bennett T, Sawaya GF. Recent intrauterine device use and the
risk of precancerous cervical lesions and cervical cancer.
Contraception 2018. https://doi.org/10.1016/j.contraception.
2018.04.008.
[13] Bulk S, Van Kemenade FJ, Rozendaal L, Meijer CJ. The Dutch
CISOE-A framework for cytology reporting increases efficacy of
screening upon standardisation since 1996. J Clin Pathol 2004;
57(4):388e93.
[14] Syrjanen K, Shabalova I, Petrovichev N, Kozachenko V,
Zakharova T, Pajanidi J, et al. Oral contraceptives are not an
independent risk factor for cervical intraepithelial neoplasia or
high-risk human papillomavirus infections. Anticancer Res 2006;
26(6c):4729e40.
[15] Lekovich JP, Amrane S, Pangasa M, Pereira N, Frey MK,
Varrey A, et al. Comparison of human papillomavirus infection
and cervical cytology in women using copper-containing and
levonorgestrel-containing intrauterine devices. Obstet Gynecol
109
2015;125(5):1101e5. https:
//doi.org/10.1097/aog.0000000000000760.
[16] Delvenne P, Herman L, Kholod N, Caberg JH, Herfs M,
Boniver J, et al. Role of hormone cofactors in the human
papillomavirus-induced carcinogenesis of the uterine cervix. Mol
Cell Endocrinol 2007;264(1e2):1e5. https:
//doi.org/10.1016/j.mce.2006.10.014.
[17] Yuan F, Auborn K, James C. Altered growth and viral gene
expression in human papillomavirus type 16-containing cancer
cell lines treated with progesterone. Cancer Invest 1999;17(1):
19e29.
[18] Vessey M, Painter R. Oral contraceptive use and cancer. Findings
in a large cohort study, 1968-2004. Br J Canc 2006;95(3):385e9.
https://doi.org/10.1038/sj.bjc.6603260.
[19] Moodley M, Moodley J, Chetty R, Herrington CS. The role of
steroid contraceptive hormones in the pathogenesis of invasive
cervical cancer: a review. Int Gynecol Cancer Soc 2003;13(2):
103e10.
[20] Iversen L, Sivasubramaniam S, Lee AJ, Fielding S,
Hannaford PC. Lifetime cancer risk and combined oral contra-
ceptives: the royal college of general practitioners’ oral contra-
ception study. Am J Obstet Gynecol 2017;216(6):580. https:
//doi.org/10.1016/j.ajog.2017.02.002. e1 - e9.
[21] Michels KA, Pfeiffer RM, Brinton LA, Trabert B. Modification
of the associations between duration of oral contraceptive use and
ovarian, endometrial, breast, and colorectal cancers. JAMA
oncology 2018;4(4):516e21. https://doi.org/10.1001/jamaoncol.
2017.4942.