This study was a retrospective study involving patients who reported to Adıyaman University Gynaecology and Family Planning polyclinics from January 2017 to December 2019. The study was approved by the Adıyaman University Non-Interventional Clinical Research Ethics Committee (No: 2019 / 9-1).

The minimum sample size for each of the two groups was calculated using the formula [11]. n=D (z)² p (1-p)/d² where, n=minimum sample size at the 95% confidence level; Z=the standard deviation, D: design effect usually set at 2 [12]; d=precision: the difference between the true population rate and acceptable sample rate, set at 0.05; p=population prevalence from a previous study: 0.058 [13]. The minimum sample size (n) was calculated to be 168. In this study, considering the error rate of 20%, the sample finally taken in each group was 202.

Inclusion criteria: the use of Cu-IUD or COC as a contraceptive method, and accessibility to sociodemographic and obstetric data, cervical cytological reports, HPV test and colposcopic biopsy reports.

Exclusion criteria: Subjects under the age of 21, having received corticosteroid and/or other immunosuppressive treatment in the past one year, having diabetes mellitus, history of malignancy, history of chemotherapy/radiotherapy, history of cervical pathology (colposcopic biopsy, Loop Electrosurgical Excision Procedure (LEEP) konisation, cold ablation etc.,) were excluded. Patients who did not give information about the contraceptive methods they used or who did not have regular follow-up data were also excluded from the study.

In the light of these criteria, 1373 patients whose data were fully accessed were included in the study. Eight hundred thirty five patients included in the study used Cu-IUD (Group A) as contraceptive use method and 538 patients were using COC (Group B). Demographic characteristics such as age, obstetric data (parity), education level, occupation, smoking history, BMI, and duration of contraceptive use were noted. The genital examination findings of the patients were noted. The presence of genital wart, cervical cytology results, HPV tests results and colposcopic biopsy results were also noted.

Cervical cytology results were classified according to the Bethesda system. The corresponding epithelial cellular abnormalities, most notably: Atypical Squamous Cells of Undetermined Significance (ASCUS), Low Grade Squamous Intraepithelial Lesions (LSIL), High-Grade Squamous Intraepithelial Lesions (HSIL), Atypical Squamous Cells cannot exclude HSIL (ASC-H) and Atypical Glandular Cells (AGC) were reported according to the Bethesda system [14]. Colposcopic biopsy was performed on those whose cytology results were abnormal and/or infected with high-risk HPV types [3]. Biopsy results were classified into three grades: CIN I, CIN II and CIN III equivalent to mild, moderate and severe dysplasia respectively; and carcinoma in situ [4]. Here, CIN II+ is used for patients with moderate to severe dysplasia and carcinoma insutu.

The data were analysed with SPSS v22 (SPSS Inc., Chicago, USA). Qualitative data were expressed as frequency, percentage, median (minimum-maximum), and mean±standard deviation. Pearson chi-square test and binary logistic regression test were used for data analysis. A p-value of <0.05 was considered statistically significant.

The omnibus test was performed to determine the compatibility of the binary logistic regression test model and its value in predicting the analysis. Omnibus test showed the compatibility of the model (p<0.001).

All women participating in the study were married, monogamous and sexually active. The age distribution of the participants ranged from 21-51 years, and the mean age was 39.19±14.9 years for the group using Cu-IUD and 31.83±12.71 years for the group using COC [Table/Fig-1]. The prevalence of cervical atypia was 17.96% (150) in those using Cu-IUD and 17.65% (95) in those using COC.

While comparing the women using COC and women using Cu-IUD, positivity of high-risk HPV types was common among women using COC p<0.001. In addition, when the colposcopic biopsy results were examined, the incidence of CIN I was higher in women using Cu-IUD, and the frequency of CIN II+ was higher in women using COC (p=0.006) [Table/Fig-2].

Frequency of CIN I and CIN II and above (CIN II+) lesions were increased in women aged 40-51 years, women with positive high-risk HPV test results, multiparous women, urban residents (p<0.001) [Table/Fig-3].

When the relationship between the contraceptive method and cervical atypia was investigated, the frequency of CIN I lesions increased in women using Cu-IUD (p=0.038), while the frequency of CIN II+ lesions increased in COC users (p=0.014) [Table/Fig-3].

In this study, compared to women using Cu-IUD and using COC as a contraceptive method, the risk of CIN I in women using Cu-IUD and CIN II+ in women using COC was found to be increased. Moreover, being in the age group of 40-51 years for the development of CIN lesions, positivity of high-risk HPV types, secondary education level, multiparity, being unemployed and urban location were identified as risk factors.

It has been reported in previous publications that there is a negative relationship between the use of Cu-IUD and the development of cervical cancer [6,15]. In a new meta-analysis examining the relationship between Cu-IUD use and cervical cancer; cervical cancer has been reported to be less common in women using Cu-IUD [8]. In a study between a group of women using Cu-IUD and two women who did not use contraceptive methods; there was no significant relationship between abnormal cytological results and Cu-IUD [16]. This study is different from the earlier studies. Firstly, the study was done at single centre and includes homogeneous group of patients. Earlier studies have included large group of patients, patients who had multiple partners and a different age at first intercourse. All women in this study were married and monogamous. This homogeneity may have implications on the results. A recent large-scale study has shown to increase the risk of CIN II+ in women using LNG-IUD, and there was a tendency to diminish the risk of CIN II+ in women using Cu-IUD [10]. The results of this study are coherent with the previous study, as the risk of CIN II+ is shown to decrease in women using Cu-IUD.

In previous studies, Bagchi S et al., and Nayak SR et al., the increase in the frequency of ASCUS and LSIL in Cu-IUD users has been shown to be 2 and 3 years of use, respectively [17,18]. Similarly, the frequency of CIN I lesions increased in those using Cu-IUD in this study (the average usage time of Cu-IUD in the present study was 5 years).

A recent meta-analysis of 33 substances including COC, Depot Medroxyprogesterone Acetate (DMPA), Cu-IUD, LNG-IUD and other methods; could not ascertain a clear relationship between COC and HPV infection [19]. Nielsen A et al., has reported that high-risk HPV infections are more common in COC users, and have shown that COC change susceptibility or immune response to HPV infections [20]. A recent study in Nigeria reported that the incidence of CIN II lesions increased in COC users [21]. Moreover, in a recent study by Xu H et al., COC use has been shown to increase the risk of CIN II and III [22]. In a study by Loopik DL et al., it was shown that the risk of CIN III increased in COC users compared to IUD users [23]. Moreover, in another study, Adhikari I et al., stated that the risk of ASCUS and LSIL decreases the risk of CIN I and cytological atypia with the use of COC, and that COC use may be protective against CIN I lesions [9]. In this study, the risk of CIN I was decreased in COC users.

The strength of the study lies in evaluating not only the contraceptive methods, but also the other risk factors associated with cervical atypia; and it is among the few studies from Turkey.

This research had some limitations. First, the present study was retrospective and single-centered study. There may be deficiencies in the retrospective analysis of the data. When the sexual behaviours of the patient groups were questioned, they all stated that they were married and monogamous. This homogenicity of the group should be considered before applying this data to larger homogeneous population.

It has been shown that the use of Cu-IUD decreases the risk of CIN II+ lesions by increasing the risk of CIN I, while the use of COC decreases the risk of CIN I by increasing the risk of CIN II+. In addition, although the most important risk factor for the development of cervical atypia is the positivity of high-risk HPV types, it should be remembered that sociodemographic features are also important risk factors.