{"paper_id":"6588874f-1cb5-4ba3-8174-c324eed218a7","body_text":"Vol.:(0123456789)\nEuropean Journal of Clinical Pharmacology (2024) 80:855–867 \nhttps://doi.org/10.1007/s00228-024-03656-y\nRESEARCH\nUse of statins and risks of ovarian, uterine, and cervical diseases: \na cohort study in the UK Biobank\nXue‑Feng Jiao1,2,3,4 · Hailong Li1,2,3,4 · Linan Zeng1,2,3,4 · Huazhen Yang5,6 · Yao Hu5,6 · Yuanyuan Qu5,6 · \nWenwen Chen5,6 · Yajing Sun5,6 · Wei Zhang5,6 · Xiaoxi Zeng5,6,7 · Lingli Zhang1,2,3,4,8\nReceived: 22 January 2024 / Accepted: 22 February 2024 / Published online: 28 February 2024 \n© The Author(s) 2024\nAbstract\nPurpose To examine the associations between use of statins and risks of various ovarian, uterine, and cervical diseases, \nincluding ovarian cancer, endometrial cancer, cervical cancer, ovarian cyst, polycystic ovarian syndrome, endometriosis, \nendometrial hyperplasia, endometrial polyp, and cervical polyp.\nMethods We conducted a cohort study among female participants in the UK Biobank. Information on the use of statins \nwas collected through verbal interview. Outcome information was obtained by linking to national cancer registry data and \nhospital inpatient data. We used Cox proportional hazards regression to examine the associations.\nResults A total of 180,855 female participants (18,403 statin users and 162,452 non-users) were included. Use of statins was \nsignificantly associated with increased risks of cervical cancer (adjusted hazard ratio (HR), 1.55; 95% confidence interval \n(95% CI), 1.05–2.30) and polycystic ovarian syndrome (adjusted HR, 4.39; 95% CI, 1.68–11.49). However, we observed no \nsignificant association between use of statins and risk of ovarian cancer, endometrial cancer, ovarian cyst, endometriosis, \nendometrial hyperplasia, endometrial polyp, or cervical polyp.\nConclusion Our findings suggest that use of statins is associated with increased risks of cervical cancer and polycystic ovar-\nian syndrome, but is not associated with increased or decreased risk of ovarian cancer, endometrial cancer, ovarian cyst, \nendometriosis, endometrial polyp, or cervical polyp.\nKeywords Statins · Cohort study · Risk · Cervical cancer · Polycystic ovarian syndrome\nIntroduction\nStatins, as inhibitors of 3-hydroxy-3-methyl-glutaryl coen -\nzyme A (HMG-CoA) reductase, function by impeding the \nbiosynthesis of cholesterol through the inhibition of the \nconversion of HMG-CoA to mevalonate. Consequently, they \nare primarily used in the treatment of hypercholesterolemia \nand for the secondary prevention of coronary artery diseases. \nStatins are among the most widely prescribed drugs world-\nwide [1, 2]. For example, in the United States, an estimated \n * Wei Zhang \n zhangwei@wchscu.cn\n * Xiaoxi Zeng \n zengxiaoxi@wchscu.cn\n * Lingli Zhang \n zhanglingli@scu.edu.cn\n1 Department of Pharmacy, West China Second University \nHospital, Sichuan University, Chengdu, Sichuan, China\n2 Evidence-Based Pharmacy Center, West China Second \nUniversity Hospital, Sichuan University, Chengdu, Sichuan, \nChina\n3 NMPA Key Laboratory for Technical Research On Drug \nProducts In Vitro and In Vivo Correlation, Chengdu, \nSichuan, China\n4 Key Laboratory of Birth Defects and Related Diseases \nof Women and Children, Sichuan University, Ministry \nof Education, Chengdu, Sichuan, China\n5 West China Biomedical Big Data Center, West China \nHospital, Sichuan University, Chengdu, Sichuan, China\n6 Medical Big Data Center, Sichuan University, Chengdu, \nSichuan, China\n7 Division of Nephrology, Kidney Research Institute, West \nChina Hospital, Sichuan University, Chengdu, Sichuan, \nChina\n8 Chinese Evidence-Based Medicine Center, West China \nHospital, Sichuan University, Chengdu, Sichuan, China\n\n856 European Journal of Clinical Pharmacology (2024) 80:855–867\n38.7 million persons, about 12% of the population, were tak-\ning a statin [3].\nIn addition to their lipid-lowering effect, statins exhibit \nother pleiotropic effects. For example, some experimental \nstudies of human cell lines and animal models suggest that \nstatins may have beneficial effects in the prevention and treat-\nment of several ovarian and uterine diseases, such as ovarian \ncancer, endometrial cancer, polycystic ovary syndrome, and \nendometriosis [4, 5]. However, clinical studies regarding this \nissue are scarce and have yielded inconsistent results [5].\nOn the other hand, the pleiotropic effects of statins are not \nalways considered beneficial in previous studies. For exam-\nple, some other experimental studies of human cell lines and \nanimal models have reported the toxic effects of statins on \nthe ovary and uterus. These toxic effects include antiprolif-\nerative and pro-apoptotic effects on ovarian and endometrial \ncells, inhibition of ovarian steroidogenesis, morphological and \nhistological changes in the ovary, antiangiogenic effects, and \nreduced fertility [6, 7]. Moreover, in our prior pharmacovigi-\nlance study, by disproportionality analyses using the FDA \nAdverse Event Reporting System (FAERS) database, we found \nthat use of statins might be associated with increased risks of \novarian cancer, endometrial cancer, cervical cancer, ovarian \ncyst, polycystic ovarian syndrome, endometriosis, endometrial \nhyperplasia, endometrial polyp, and cervical polyp [8]. How-\never, the results of disproportionality analyses could only dem-\nonstrate statistical associations and not causations and should \nbe verified by further cohort studies [9].\nThe UK Biobank is a large-scale database containing exten-\nsive sociodemographic, lifestyle, and clinical data on half a \nmillion participants. Leveraging this database, we conducted \na cohort study to comprehensively examine the associations \nbetween use of statins and risks of ovarian cancer, endometrial \ncancer, cervical cancer, ovarian cyst, polycystic ovarian syn-\ndrome, endometriosis, endometrial hyperplasia, endometrial \npolyp, and cervical polyp.\nMethods\nData source\nThe UK Biobank comprises 502,507 volunteer participants \naged 37–73 from England, Scotland, and Wales who were \nrecruited between 2006 and 2010. Details of the design \nand survey methods for UK Biobank have been described \nin previous studies [10, 11]. At baseline assessment visit \nand repeat assessment visits, participants completed a \ntouchscreen questionnaire and a verbal interview, which \ncollected information on sociodemographic characteris-\ntics, lifestyle, medical history, medication history, and \nreproductive factors. Repeat assessment visits were con-\nducted every 2–3 years during the follow-up period, at \nwhich participants underwent a repetition of the baseline \nassessment visit. Thus, repeat assessment visits could \nenrich, confirm, and calibrate the data collected at base-\nline assessment visit. Moreover, touchscreen questionnaire \nvalidation was performed in two ways. First, some ques -\ntions (especially medical questions) in the touchscreen \nquestionnaire would be asked again and confirmed in \nthe subsequent verbal interview. Second, the touchscreen \nquestionnaire incorporated a number of logic checks on \nthe data that were entered, such as checking for contra-\ndictory answers and impossible or improbable numeric \nvalues [12].\nIn addition, the collected data were linked to hospital \ninpatient data, national cancer registry data, and national \ndeath registry data, which enabled long-term follow-up of \nparticipants and their health-related outcomes. Hospital \ninpatient data on participants in England, Scotland, and \nWales were received from their respective databases: the \nHospital Episode Statistics for England (HES), the Scot-\ntish Morbidity Record (SMR), and the Patient Episode \nDatabase for Wales (PEDW) [9 ]. National cancer registry \ndata and national death registry data were acquired from \nthe National Health Service (NHS) Digital (for partici -\npants in England or Wales) and the NHS Central Register \n(for participants in Scotland) [13].\nStudy design and population\nWe conducted a cohort study of female participants in \nthe UK Biobank. We excluded females who had a his-\ntory of cancer (except for non-melanoma skin cancer) \n[14], ovarian cyst, polycystic ovarian syndrome, endo-\nmetriosis, endometrial hyperplasia, endometrial polyp, \ncervical polyp, ovariectomy, hysterectomy, or cervice-\nctomy at baseline, or who had withdrawn from the UK \nBiobank. The required information was collected through \ntouchscreen questionnaire/verbal interview and linkage to \nhospital inpatient data and national cancer registry data. \nDetails of the variable name, data field, and data coding \nin the UK Biobank are given in Supplemental Table 1.\nExposure\nInformation on the use of statins was self-reported and col-\nlected through verbal interview. If the participant indicated \nin the touchscreen that they were taking cholesterol-lowering \ndrugs, then the interviewer was prompted to record the name \nof the drug. Use of statins was defined as continuous use \nof statins for months or years. It did not include the use of \nstatins for a few days or a week, or prescribed statins that \nwere not taken [15]. Based on treatment with a statin or not, \nthe participants were divided into statin users and non-users. \n\n857European Journal of Clinical Pharmacology (2024) 80:855–867 \nThe statins recorded in the UK Biobank included simvasta-\ntin, atorvastatin, rosuvastatin, and pravastatin (Supplemental \nTable 2). Specific data on usage, dosage, and duration were \nnot recorded.\nOutcome\nThe outcomes were first diagnoses of ovarian cancer, endo-\nmetrial cancer, cervical cancer, ovarian cyst, polycystic \novarian syndrome, endometriosis, endometrial hyperplasia, \nendometrial polyp, and cervical polyp during the follow-up \nperiod. Cases of incident ovarian cancer, endometrial cancer, \nand cervical cancer were ascertained by linking to national \ncancer registry data and hospital inpatient data, and incident \novarian cyst, polycystic ovarian syndrome, endometriosis, \nendometrial hyperplasia, endometrial polyp, and cervical \npolyp were ascertained by linking to hospital inpatient data. \nWe also obtained the first diagnosis date from national can-\ncer registry data and hospital inpatient data. The correspond-\ning variable name, data field, and data coding in the UK \nBiobank are presented in Supplemental Table 3.\nFollow‑up time\nWhen assessing cancer outcomes, female participants were \nfollowed from baseline visit until the first diagnosis of the \noutcome, the diagnosis of other cancer (except for non-\nmelanoma skin cancer), death, or the last linkage date with \nnational cancer registry data and hospital inpatient data (31 \nDecember 2016 for national cancer registry data, 31 March \n2017 for HES, 31 October 2016 for SMR, or 29 February \n2016 for PEDW), whichever came first [13, 16]. In addition, \nwhen assessing non-cancer outcomes, female participants \nwere followed from baseline visit until the first diagnosis \nof the outcome, death, or the last linkage date with hospital \ninpatient data (31 March 2017 for HES, 31 October 2016 for \nSMR or 29 February 2016 for PEDW), whichever came first \n[13, 16]. The required information was obtained by linking \nto national cancer registry data, hospital inpatient data, and \nnational death registry data.\nCovariates\nThe covariates included age, race (white or others), \nTownsend deprivation index (quintiles), smoking status \n(never, past, or current), alcohol use (daily or almost daily, \nthree or four times a week, once or twice a week, one to \nthree times a month, special occasions only, or never), vig-\norous physical activity (low, moderate, or high), number of \nchildbirth, number of abortion, comorbidities at baseline \n(hyperlipidemia, ischemic heart disease, ischemic cerebro-\nvascular disease, hypertension, diabetes, obesity, or pelvic \ninflammatory disease), and oral contraceptive. These covari-\nates were factors known to be correlated with risks of all \noutcomes according to previous literatures, or indications for \nuse of statins [4]. Moreover, for each outcome, we included \nextra related covariates which were correlated solely with \nrisk of this outcome according to previous literatures (Sup-\nplemental Table 4). All these covariates were collected \nthrough touchscreen questionnaire/verbal interview and link-\nage to hospital inpatient data. Details of the variable name, \ndata field and data coding in the UK Biobank are given in \nSupplemental Table 5. The Townsend deprivation index was \nwidely used as a measure of socioeconomic deprivation, \nwith higher scores indicating greater deprivation [17]. The \nnumber of childbirth was derived from the number of live \nbirths and stillbirths. In addition, the number of abortion was \nderived from the number of spontaneous miscarriages and \npregnancy terminations. Furthermore, obesity was defined \nas body mass index (BMI) ≥ 30. Missing data were coded \nas a missing indicator category for categorical variables and \nwith mean values for continuous variables.\nStatistical analysis\nBaseline analysis\nComparisons were made between satin users and non-users \nfor the following variables at baseline: age, race, Townsend \ndeprivation index, smoking status, alcohol use, vigorous \nphysical activity, number of childbirth, number of abor -\ntion, comorbidities (hyperlipidemia, ischemic heart disease, \nischemic cerebrovascular disease, hypertension, diabetes, \nobesity, and pelvic inflammatory disease), and oral con-\ntraceptive. Continuous variables were presented as mean \n(standard deviation (SD)) and analyzed by using the Stu-\ndent’s t-test or median (interquartile range (IQR)) and by \nWilcoxon rank-sum test, as appropriate. Categorical vari-\nables were presented as counts and percentages and evalu-\nated by chi-square test, Fisher’s exact test, or rank-sum test \nas appropriate.\nMain analysis\nWe used Cox proportional hazards regression to analyze  \nthe associations between use of statins and risks of  \novarian, uterine, and cervical diseases, with results \nexpressed as hazard ratios (HRs) and 95% confidence \nintervals (95% CI). Time since baseline visit was used as \nthe underlying timescale. We developed a multivariable \nmodel with adjustment for age, race, Townsend deprivation  \nindex, smoking status, alcohol use, vigorous physical  \nactivity, number of childbirth, number of abortion, any \ncomorbidity at baseline (hyperlipidemia, ischemic heart \n\n858 European Journal of Clinical Pharmacology (2024) 80:855–867\ndisease, ischemic cerebrovascular disease, hypertension, \ndiabetes, obesity, or pelvic inflammatory disease), and \noral contraceptive. Moreover, for better control of some \noutcome-specific confounders, we included extra related \ncovariates in the Cox proportional hazards model for  \neach outcome (Supplemental Table 4). Furthermore, the \nanalyses of ovarian cyst, polycystic ovarian syndrome, and \nendometriosis were restricted to the premenopausal female \ncohort because these diseases are less likely to develop \nafter menopause.\nSubgroup analysis\nWe used Schoenfeld residuals to test the proportional haz-\nards assumption and found that the assumption was vio-\nlated for age. Thus, we performed subgroup analysis strati-\nfied by age to assess if change in result was noteworthy. We \nperformed subgroup analysis stratified by the median age \n(age (≤ 56 or > 56 years) for ovarian cancer, endometrial \ncancer, cervical cancer, endometrial hyperplasia, endome-\ntrial polyp, and cervical polyp; age (≤ 46 or > 46 years) for \novarian cyst, polycystic ovarian syndrome, and endome-\ntriosis). In addition, to assess the potential modification \neffects by statin type, we performed subgroup analysis \namong different statins.\nSensitivity analysis\nWe conducted several sensitivity analyses to confirm the \nrobustness of the results. First, to minimize the potential \nfor reverse causality, we performed a sensitivity analysis by \nexcluding the first year of follow-up (for all individuals). \nSecond, to minimize indication bias, we performed a sensi-\ntivity analysis by restricting the study population to females \nwith hyperlipidemia, ischemic heart disease, ischemic cer -\nebrovascular disease, hypertension, diabetes, or obesity (all \nthese diseases are indications for use of statins or common \ncomorbidities in statin users). Third, as the average age of \nmenopause in UK women is 51 years [18], we performed a \nsensitivity analysis by censoring the follow-up at age 51 for \nthe outcomes of ovarian cyst, polycystic ovarian syndrome, \nand endometriosis.\nAll data analyses were conducted using R version 3.6.3. \nStatistical significance was set at P  < 0.05 using two-sided \ntests. However, as the threshold of P < 0.05 is conventional \nand arbitrary, it does not convey any meaningful evidence \nof clinical significance or the size of the effect. Thus, we \ncomprehensively examined the precise P  values, the esti-\nmates of the effect sizes, and the confidence intervals, to \ninterpret the statistical analyses and evaluate the clinical \nsignificances [19].\nResults\nOur study identified 273,314 female participants in the UK \nBiobank. Among these, 92,459 were excluded because of \nhaving a history of cancer (except for non-melanoma skin \ncancer), ovarian cyst, polycystic ovarian syndrome, endome-\ntriosis, endometrial hyperplasia, endometrial polyp, cervical \npolyp, ovariectomy, hysterectomy, or cervicectomy at base-\nline. In total, 180,855 female participants were included in \nanalysis (18,403 statin users and 162,452 non-users) (Fig. 1). \nThe median age of the included participants was 56 years \n(IQR, 49–62) at baseline. Among them, 54,359 participants \n(1510 statin users and 52,849 non-users) were premenopau-\nsal females, and their median age was 46 years (IQR, 43–49) \nat baseline. Table  1 describes the baseline characteristics \nof participants according to use of statins. Compared with \nnon-users, statin users were more likely to be older, socio-\neconomically deprived, and smokers. They also had higher \nnumber of childbirth and more comorbidities. Moreover, sta-\ntin users were less likely to be white and physical active, yet \nhad fewer alcohol consumption, lower number of abortion, \nand less use of oral contraceptives. In addition, when we \nrestricted the study population to premenopausal females, \nthere were no significant differences between statin users \nand non-users in smoking status, number of childbirth, or \nnumber of abortion, while the characteristics of other covari-\nates were similar to the whole study population (Supplemen-\ntal Table 6).\nTable  2 shows the results of main analysis. During a \nmedian follow-up of 8–9 years, the numbers of female par -\nticipants with a first diagnosis of ovarian cancer, endometrial \ncancer, cervical cancer, ovarian cyst, polycystic ovarian syn-\ndrome, endometriosis, endometrial hyperplasia, endometrial \npolyp, and cervical polyp were 599, 849, 363, 601, 32, 528, \n397, 3166, and 814, respectively. The crude incidence per \n1000 person-years among non-users and statin users was 0.41 \ncompared to 0.56 for ovarian cancer, 0.55 compared to 1.09 \nfor endometrial cancer, 0.26 compared to 0.24 for cervical \ncancer, 1.37 compared to 1.39 for ovarian cyst, 0.06 compared \nto 0.49 for polycystic ovarian syndrome, 1.20 compared to \n1.23 for endometriosis, 0.26 compared to 0.39 for endometrial \nhyperplasia, 2.15 compared to 2.58 for endometrial polyp, and \n0.56 compared to 0.52 for cervical polyp. After adjustment for \nthe covariates, use of statins was significantly associated with \nincreased risks of cervical cancer (adjusted HR, 1.55; 95% \nCI, 1.05–2.30) and polycystic ovarian syndrome (adjusted \nHR, 4.39; 95% CI, 1.68–11.49). However, we observed no \nsignificant association between use of statins and risk of ovar-\nian cancer (adjusted HR, 0.94; 95% CI, 0.73–1.22), endome-\ntrial cancer (adjusted HR, 1.06; 95% CI, 0.88–1.28), ovarian  \ncyst (adjusted HR, 0.92; 95% CI, 0.56–1.52), endometriosis  \n\n859European Journal of Clinical Pharmacology (2024) 80:855–867 \n(adjusted HR, 0.84; 95% CI, 0.49–1.42), endometrial hyper-\nplasia (adjusted HR, 1.04; 95% CI, 0.77–1.40), endometrial \npolyp (adjusted HR, 0.99; 95% CI, 0.88–1.11), or cervical \npolyp (adjusted HR, 0.99; 95% CI, 0.76–1.28).\nFigure  2 shows stratified analyses by statin type. The \nnumbers of simvastatin, atorvastatin, rosuvastatin, and \npravastatin users in the subgroups were 13,426, 3873, \n905, and 664, respectively. When we restricted the study \npopulation to premenopausal females, the numbers of sim-\nvastatin, atorvastatin, rosuvastatin, and pravastatin users \nin the subgroups were 1107, 337, 67, and 47, respectively. \nFor cervical cancer, use of pravastatin was significantly \nassociated with increased risk of cervical cancer (adjusted \nHR, 4.31; 95% CI, 1.36–13.63), use of simvastatin was \nborderline associated with increased risk of cervical can-\ncer (adjusted HR, 1.55; 95% CI, 0.98–2.42), whereas use \nof atorvastatin was not significantly associated with risk of \ncervical cancer. For polycystic ovarian syndrome, uses of \nsimvastatin (adjusted HR, 3.90; 95% CI, 1.27–11.94) and \natorvastatin (adjusted HR, 7.00; 95% CI, 1.55–31.58) were \nall significantly associated with increased risk of polycys-\ntic ovarian syndrome. For endometrial hyperplasia, use of \npravastatin (adjusted HR, 2.50; 95% CI, 1.03–6.10) was \nsignificantly associated with increased risk of endometrial \nhyperplasia, whereas use of other types of statins was not \nsignificantly associated with risk of endometrial hyper -\nplasia. For other outcomes, use of simvastatin, atorvasta-\ntin, rosuvastatin, or pravastatin was all not significantly \nFig. 1  Flow chart of study population\n\n860 European Journal of Clinical Pharmacology (2024) 80:855–867\nassociated with risk of ovarian cancer, endometrial cancer, \novarian cyst, endometriosis, endometrial polyp, or cervi-\ncal polyp.\nFigure  3 shows stratified analyses by the median age. \nThere remained no significant association between use \nof statins and risk of ovarian cancer, endometrial cancer, \nTable 1  Baseline characteristics \nof participants by use of statins\nIQR interquartile range. Data are n (%) unless otherwise indicated\nCharacteristics Non-users (n = 162,452) Statin users (n = 18,403) P value\nAge, years, median (IQR) 55 (48–61) 62 (57–66)  < 0.001\nRace  < 0.001\n    White 152,657 (94.0) 17,118 (93.0)\n    Others 8972 (5.5) 1202 (6.5)\n    Missing 823 (0.5) 83 (0.5)\nTownsend deprivation index (quintiles)  < 0.001\n    1 (least deprived) 33,061 (20.4) 3104 (16.9)\n    2 32,757 (20.2) 3333 (18.1)\n    3 32,536 (20.0) 3588 (19.5)\n    4 32,352 (19.9) 3773 (20.5)\n    5 (most deprived) 31,537 (19.4) 4589 (24.9)\n    Missing 209 (0.1) 16 (0.1)\nSmoking status  < 0.001\n    Never 99,523 (61.3) 10,158 (55.2)\n    Past 48,284 (29.7) 6384 (34.7)\n    Current 13,802 (8.5) 1752 (9.5)\n    Missing 843 (0.5) 109 (0.6)\nAlcohol use  < 0.001\n    Daily or almost daily 26,988 (16.6) 2657 (14.4)\n    Three or four times a week 35,612 (21.9) 3055 (16.6)\n    Once or twice a week 42,661 (26.3) 4206 (22.9)\n    One to three times a month 20,931 (12.9) 2360 (12.8)\n    Special occasions only 22,209 (13.7) 3539 (19.2)\n    Never 13,571 (8.4) 2528 (13.7)\n    Missing 480 (0.3) 58 (0.3)\nVigorous physical activity  < 0.001\n    Low 22,797 (14.0) 2911 (15.8)\n    Moderate 55,190 (34.0) 5849 (31.8)\n    High 49,685 (30.6) 4655 (25.3)\n    Missing 34,780 (21.4) 4988 (27.1)\nNumber of childbirth, median (IQR) 2 (1–2) 2 (1–3)  < 0.001\nNumber of abortion, median (IQR) 0 (0–1) 0 (0–1)  < 0.001\nComorbidities\n    Hyperlipidemia 3944 (2.4) 12,767 (69.4)  < 0.001\n    Ischemic heart disease 1827 (1.1) 2680 (14.6)  < 0.001\n    Ischemic cerebrovascular disease 860 (0.5) 1153 (6.3)  < 0.001\n    Hypertension 29,333 (18.1) 11,024 (59.9)  < 0.001\n    Diabetes 2408 (1.5) 3719 (20.2)  < 0.001\n    Obesity 32,385 (19.9) 6992 (38.0)  < 0.001\n    Pelvic inflammatory disease 1759 (1.1) 161 (0.9) 0.010\nOral contraceptive  < 0.001\n    Yes 134,183 (82.6) 13,259 (72.0)\n    No 27,398 (16.9) 5031 (27.3)\n    Missing 871 (0.5) 113 (0.6)\n\n861European Journal of Clinical Pharmacology (2024) 80:855–867 \novarian cyst, endometriosis, endometrial hyperplasia, \nendometrial polyp, or cervical polyp in all age groups. \nA tendency toward increased risk of cervical cancer was \nobserved in statin users aged > 56 years (adjusted HR, \n1.62; 95% CI, 0.94–2.79), but this tendency was not \nobserved in users aged ≤ 56 years (adjusted HR, 1.39; \n95% CI, 0.75–2.55). Moreover, increased risk for poly -\ncystic ovarian syndrome from use of statins was seen in \npremenopausal females aged ≤ 46 years (adjusted HR, \n7.74; 95% CI, 2.52–23.79), whereas no significant asso-\nciation was seen in premenopausal females aged  > 46 \nyears (adjusted HR, 1.46; 95% CI, 0.18–12.02).\nIn our sensitivity analyses, the associations between use \nof statins and risks of all outcomes remained: (1) when \nwe excluded the first year of follow-up (for all individu-\nals) (Fig.  4A); (2) when we restricted the study population \nto females with hyperlipidemia, ischemic heart disease, \nischemic cerebrovascular disease, hypertension, diabetes, \nor obesity (Fig.  4B); and (3) when we censored the follow-\nup of premenopausal females at age 51 (Fig.  4C).\nDiscussion\nPrincipal findings\nIn this large-scale cohort study, we found that use of statins \nwas significantly associated with increased risks of cervi-\ncal cancer and polycystic ovarian syndrome, but was not \nsignificantly associated with risk of ovarian cancer, endo-\nmetrial cancer, ovarian cyst, endometriosis, endometrial \nhyperplasia, endometrial polyp, or cervical polyp. We also \nnoticed the potential modifying effects of statin type and \nage on the aforementioned associations. For instance, use \nof simvastatin was significantly associated with increased \nrisk of polycystic ovarian syndrome and was borderline \nassociated with increased risk of cervical cancer; use of \natorvastatin was significantly associated with increased \nrisk of polycystic ovarian syndrome; use of pravastatin was \nsignificantly associated with increased risks of cervical \ncancer and endometrial hyperplasia. Moreover, when we \nstratified by the median age, increased risk for polycystic \nTable 2  The associations between use of statins and risks of ovarian cancer, endometrial cancer, cervical cancer, ovarian cyst, polycystic ovarian \nsyndrome, endometriosis, endometrial hyperplasia, endometrial polyp, and cervical polyp\nHR hazard ratio, CI confidence interval\n*Adjusted for age, race, Townsend deprivation index, smoking status, alcohol use, vigorous physical activity, number of childbirth, number of \nabortion, any comorbidity at baseline (hyperlipidemia, ischemic heart disease, ischemic cerebrovascular disease, hypertension, diabetes, obesity, \nor pelvic inflammatory disease), oral contraceptive, and extra outcome-specific covariates. Moreover, the analyses of ovarian cyst, polycystic \novarian syndrome, and endometriosis were restricted to the premenopausal female cohort\nOutcome Non-users Statin users Adjusted*\nNo. of  \nparticipants\nNo. of outcome Incidence per \n1000 person-\nyears\nNo. of  \nparticipants\nNo. of outcome Incidence per \n1000 person-\nyears\nHR (95% CI) P\nOvarian cancer 162,452 519 0.41 18,403 80 0.56 0.94 (0.73–\n1.22)\n0.665\nEndometrial \ncancer\n162,452 693 0.55 18,403 156 1.09 1.06 (0.88–\n1.28)\n0.541\nCervical cancer 162,452 329 0.26 18,403 34 0.24 1.55 (1.05–\n2.30)\n0.028\nOvarian cyst 52,849 584 1.37 1510 17 1.39 0.92 (0.56–\n1.52)\n0.758\nPolycystic \novarian syn-\ndrome\n52,849 26 0.06 1510 6 0.49 4.39 (1.68–\n11.49)\n0.003\nEndometriosis 52,849 513 1.20 1510 15 1.23 0.84 (0.49–\n1.42)\n0.503\nEndometrial \nhyperplasia\n162,452 340 0.26 18,403 57 0.39 1.04 (0.77–\n1.40)\n0.819\nEndometrial \npolyp\n162,452 2790 2.15 18,403 376 2.58 0.99 (0.88–\n1.11)\n0.883\nCervical polyp 162,452 738 0.56 18,403 76 0.52 0.99 (0.76–\n1.28)\n0.926\n\n862 European Journal of Clinical Pharmacology (2024) 80:855–867\n\n\n863European Journal of Clinical Pharmacology (2024) 80:855–867 \novarian syndrome from use of statins was only seen in \npremenopausal females aged ≤ 46 years.\nCompared with previous studies\nThe relationship between use of statins and risks of ovar -\nian cancer and endometrial cancer is an intensely disputed \ntopic. Some case–control studies found that use of statins \nwas associated with reduced risks of ovarian cancer and \nendometrial cancer, which suggests that statins might have \npreventive effects on ovarian cancer and endometrial cancer \n[20, 21]. However, in recent years, more and more cohort \nand case–control studies showed that use of statins was not \nassociated with risk of ovarian cancer or endometrial cancer \n[22–24]. Our study also found no association between use \nof statins and risk of ovarian cancer or endometrial cancer \nand does not support that use of statins may prevent ovarian \ncancer or endometrial cancer.\nOur study indicated that use of statins was associated with \nincreased risk of cervical cancer, which is inconsistent with \na prior cohort study conducted by Kim et al. Kim et al.’s \nstudy is the only clinical study to date exploring the asso-\nciation between use of statins and risk of cervical cancer. \nThat study used health insurance claims data and found that \nuse of statins was associated with reduced risk of cervical \ncancer [22]. We cannot completely explain the discrepancies \nFig. 2  The associations between use of statins and risks of ovarian \ncancer, endometrial cancer, cervical cancer, ovarian cyst, polycystic \novarian syndrome, endometriosis, endometrial hyperplasia, endome-\ntrial polyp, and cervical polyp stratified by statin type. HR hazard \nratio, CI confidence interval; —, the sample size was too small to \nenable statistical analysis. *Adjusted for age, race, Townsend depri-\nvation index, smoking status, alcohol use, vigorous physical activity, \nnumber of childbirth, number of abortion, any comorbidity at base-\nline (hyperlipidemia, ischemic heart disease, ischemic cerebrovascu-\nlar disease, hypertension, diabetes, obesity, or pelvic inflammatory \ndisease), oral contraceptive, and extra outcome-specific covariates. \nMoreover, the analyses of ovarian cyst, polycystic ovarian syndrome, \nand endometriosis were restricted to the premenopausal female cohort\n◂\nFig. 3  The associations between use of statins and risks of ovarian \ncancer, endometrial cancer, cervical cancer, ovarian cyst, polycystic \novarian syndrome, endometriosis, endometrial hyperplasia, endome-\ntrial polyp, and cervical polyp stratified by the median age. HR haz-\nard ratio, CI confidence interval. *Adjusted for race, Townsend depri-\nvation index, smoking status, alcohol use, vigorous physical activity, \nnumber of childbirth, number of abortion, any comorbidity at base-\nline (hyperlipidemia, ischemic heart disease, ischemic cerebrovascu-\nlar disease, hypertension, diabetes, obesity, or pelvic inflammatory \ndisease), oral contraceptive, and extra outcome-specific covariates. \nMoreover, the analyses of ovarian cyst, polycystic ovarian syndrome \nand endometriosis were restricted to the premenopausal female cohort\n\n864 European Journal of Clinical Pharmacology (2024) 80:855–867\nbetween Kim et al.’s study and our study, but it should be \nnoted that some differences in study design exist. Due to \nthe limited information contained in health insurance claims \ndata, Kim et al.’s study only analyzed the potential con-\nfounding effects of age, comorbidities, and co-medication \nand was unable to analyze the potential confounding effects \nof other sociodemographic, lifestyle, and clinical factors. \nIn our study, UK Biobank contains extensive sociodemo-\ngraphic, lifestyle, and clinical information. Thus, compared \nwith Kim et al.’s study, we further analyzed the potential \nconfounding effects of Townsend deprivation index, smok-\ning status, alcohol use, vigorous physical activity, num-\nber of childbirth, number of abortion, lifetime number of \nsexual partners, age first had sexual intercourse, and oral \nFig. 4  Sensitivity analyses for the associations between use of statins \nand risks of ovarian cancer, endometrial cancer, cervical cancer, ovar-\nian cyst, polycystic ovarian syndrome, endometriosis, endometrial \nhyperplasia, endometrial polyp, and cervical polyp by excluding the \nfirst year of follow-up (A), restricting the study population to females \nwith hyperlipidemia, ischemic heart disease, ischemic cerebrovascu-\nlar disease, hypertension, diabetes, or obesity (B), and censoring the \nfollow-up of premenopausal females at age 51 (C). HR hazard ratio, \nCI confidence interval. *Adjusted for age, race, Townsend depriva-\ntion index, smoking status, alcohol use, vigorous physical activity, \nnumber of childbirth, number of abortion, any comorbidity at base-\nline (hyperlipidemia, ischemic heart disease, ischemic cerebrovascu-\nlar disease, hypertension, diabetes, obesity, or pelvic inflammatory \ndisease), oral contraceptive, and extra outcome-specific covariates. \nMoreover, the analyses of ovarian cyst, polycystic ovarian syndrome, \nand endometriosis were restricted to the premenopausal female cohort\n\n865European Journal of Clinical Pharmacology (2024) 80:855–867 \ncontraceptive. All these factors have been reported to be cor-\nrelated with the occurrence of cervical cancer. For example, \nsocioeconomic deprivation, smoking, alcohol use, multiple \nsexual partners, early age at first intercourse, and use of \noral contraceptives are important risk factors for cervical \ncancer [25–27], while multiple childbirth is a protective fac-\ntor for cervical cancer [28]. By adjusting for these potential \nconfounding factors, our study might provide more reliable \nresults than Kim et al.’s study. Several possible mechanisms \nmight explain the increased risk of cervical cancer associ-\nated with use of statins. First, inhibition of serum cholesterol \nlevels by statins may be associated with increased risk of \ncancer [ 29]. Second, statins could enhance mitotic abnor -\nmalities, which may interfere with centromere development \nand function, leading to increased risk of mutations and \ncancer [30]. Third, statins could increase regulatory T cell \nnumbers, which may impair the antitumor immune response \nof the host [31].\nOur study also found that use of statins was associated \nwith increased risk of polycystic ovarian syndrome and \nwas not associated with risk of endometriosis. These find-\nings are inconsistent with previous experimental studies of \nhuman cell lines and animal models. For example, previ-\nous experimental studies suggest that statins might prevent \npolycystic ovarian syndrome by reducing steroid hormone \nsynthesis and inhibiting the growth of theca-interstitial \ncells in ovaries [4 , 5]. In addition, previous experimen-\ntal studies also suggest that statins might prevent endo-\nmetriosis due to their antiproliferative and pro-apoptotic \neffects on endometrial and endometriotic cells, their abil-\nity to reduce cell viability and migration, the inhibition \nof angiogenesis, and anti-inflammatory activities [5 , 6]. \nAs it is possible that the effects of statins in patients may \nbe different from those observed in cell culture or animal \nmodels, our cohort study provides more credible results \nthan previous experimental studies.\nExplain unexpected findings\nFor cervical cancer, when we performed subgroup analysis \nstratified by statin type or the median age, the association \nwas attenuated in most subgroups, which may be due to \nthe decreased sample size. However, the association with \ncervical cancer risk was enhanced in the pravastatin sub-\ngroup. Similarly, some previous clinical studies also found \nthat pravastatin was more likely to increase cancer risk than \nother types of statins. For example, a cohort study by Desai \net al. indicated that use of pravastatin was associated with \nincreased risk of ovarian cancer, whereas use of other types \nof statins was not [32]. In addition, a record-linkage study by \nHaukka et al. showed that use of pravastatin was associated \nwith increased risk of non-melanoma skin cancer, whereas \nuse of other types of statins was not [ 33]. The mechanism \nwhy pravastatin is more likely to increase cancer risk than \nother types of statins is unclear, but may be related to the \nhighly hydrophilic property of pravastatin. Based on their \nsolubility, statins can be chemically classified as lipophilic \nstatins and hydrophilic statins. Lipophilic statins enter cells \nthrough passive diffusion, whereas hydrophilic statins enter \ncells through active transport. It is postulated that the cel -\nlular uptake pattern of statins might be related to their effect \non tumor growth [32].\nFor polycystic ovarian syndrome, we found that \nincreased risk for polycystic ovarian syndrome from use \nof statins was seen in premenopausal females aged ≤ 46 \nyears, whereas no significant association was seen in pre-\nmenopausal females aged > 46 years. Polycystic ovarian \nsyndrome mainly occurs in reproductive aged females \n(12–45 years) and is less likely to occur in females \naged > 46 years [34]. Similarly, in our study, only 10 poly-\ncystic ovarian syndrome cases occurred in premenopausal \nfemales aged > 46 years, and the result for this subgroup \nwas imprecise (with wide confidence interval) and may be \na false negative. Thus, further studies are needed to verify \nthis finding and explore the underlying mechanism.\nFor endometrial hyperplasia, we found that use of \npravastatin was significantly associated with increased risk \nof endometrial hyperplasia, whereas use of other types of \nstatins was not. Currently, there is no clear explanation \nfor this finding. Besides, as the sample size of pravastatin \nusers (664) was relatively small, we could not rule out the \npossibility that the increased risk for endometrial hyper -\nplasia from use of pravastatin was due to chance. Thus, \nfurther studies are also needed to confirm this finding and \nexplore the possible mechanism.\nStrengths and limitations\nOur study has several strengths. First, the UK Biobank con-\ntains extensive sociodemographic, lifestyle, and clinical \ninformation, which enabled us to adjust for a wide range of \nconfounders and conduct multiple subgroup analyses. Sec-\nond, most subgroup and sensitivity analyses showed consist-\nent results with the main analysis, which further confirmed \nthe robustness of our results. Third, the prospective design \nlimited recall bias on the assessment of statins.\nOur study also has some limitations. First, use of statins \nand some covariates were assessed by self-report, which \nmight induce misclassification. Such misclassification is \nlikely non-differential between individuals with and with-\nout outcome events, which would attenuate the association \ntoward null. However, this cannot flip a protective effect \n(HR < 1) to a harmful one (HR > 1). Second, we did not \nhave information on duration or dosage of statins, and it \nmay take time for statins to have effects on outcome events. \nFurther studies are needed to evaluate the impacts of these \n\n866 European Journal of Clinical Pharmacology (2024) 80:855–867\nfactors on results. Third, the diagnosis information of ovar-\nian cyst, polycystic ovarian syndrome, endometriosis, endo-\nmetrial hyperplasia, endometrial polyp, and cervical polyp \nwas obtained by linking to hospital inpatient data. That \nsaid, these diseases diagnosed at the outpatient clinic and \nasymptomatic/undiagnosed ones were not captured in our \ndata. This misclassification might be differential between \nstatin users and non-users because users have more fre-\nquent healthcare visits and are subject to surveillance bias. \nFourth, potential reverse causality may exist in our study \nas it takes years for outcome events to develop. However, \nthe results remained unchanged when we excluded the first \nyear of follow-up. Fifth, although we adjusted for all main \nindications for statins in the statistical model and further \nperformed sensitivity analysis, indication bias could not be \ncompletely avoided.\nConclusions and clinical and  \nresearch implications\nIn conclusion, in this cohort study of UK Biobank female \nparticipants, use of statins was associated with increased \nrisks of cervical cancer and polycystic ovarian syndrome, \nbut was not associated with increased or decreased risk of \novarian cancer, endometrial cancer, ovarian cyst, endome -\ntriosis, endometrial polyp, or cervical polyp. Unlike some \nprevious studies, our findings do not support that use of \nstatins may prevent ovarian cancer, endometrial cancer, cer-\nvical cancer, polycystic ovarian syndrome, or endometriosis. \nMoreover, according to our findings, the potential risks of \ncervical cancer and polycystic ovarian syndrome associated \nwith use of statins are of great importance and should be \nclosely monitored in future clinical practice. However, our \nfindings should be interpreted with cautions due to indica-\ntion and surveillance biases.\nAbbreviations HMG-CoA: 3-Hydroxy-3-methyl-glutaryl coenzyme \nA; FAERS: FDA Adverse Event Reporting System; HES: Hospital \nEpisode Statistics for England; SMR: Scottish Morbidity Record; \nPEDW: Patient Episode Database for Wales; NHS: National Health \nService; SD: Standard deviation; IQR: Interquartile range; HRs: Hazard \nratios; 95% CI: 95% Confidence intervals\nSupplementary Information The online version contains supplemen-\ntary material available at https:// doi. org/ 10. 1007/ s00228- 024- 03656-y.\nAcknowledgements This work uses data provided by patients and \ncollected by the NHS as part of their care and support. This research \nused data assets made available by National Safe Haven as part of \nthe Data and Connectivity National Core Study, led by Health Data \nResearch UK in partnership with the Office for National Statistics and \nfunded by UK Research and Innovation (grant ref: MC_PC_20029 and \nMC_PC_20058). Moreover, this research has been conducted using \nthe UK Biobank Resource under Application 54803. We thank the \nteam members involved in West China Biomedical Big Data Center \nfor their support.\nAuthor contribution X.F.J., L.Z., X.Z., and W.Z. were responsible for \nthe study concept and design. H.Y., Y.H., Y.Q., and W.C. did the data \nand project management. X.F.J. did the data cleaning and analysis. \nX.F.J. and H.L. made the figures and tables. L.Z. and Y.S. interpreted \nthe data. X.F.J. drafted the manuscript. X.Z. and L.L.Z. revised the \nmanuscript. All the authors approved the final manuscript as submitted \nand agree to be accountable for all aspects of the work.\nFunding This study was supported by Natural Science Foundation of \nSichuan Province (grant number 2022NSFSC0644).\nData availability Data from the UK Biobank (http:// www. ukbio bank. \nac. uk/) are available to all researchers upon making an application. \nPart of this research was conducted using the UK Biobank Resource \nunder Application 54803.\nDeclarations \nEthics approval and informed consent The UK Biobank has full ethical \napproval from the NHS National Research Ethics Service (16/NW/0274), \nand this study was approved by the biomedical research ethics committee \nof West China Hospital (2019.1171). All the UK Biobank participants \nprovided written informed consent before data collection.\nCompeting interests The authors have no competing interests.\nDisclaimer The funders had no role in the development of this article \n(i.e., in the study design; collection, analysis, and interpretation of \ndata; report writing; or decision to submit the paper for publication).\nOpen Access This article is licensed under a Creative Commons Attri-\nbution 4.0 International License, which permits use, sharing, adapta-\ntion, distribution and reproduction in any medium or format, as long \nas you give appropriate credit to the original author(s) and the source, \nprovide a link to the Creative Commons licence, and indicate if changes \nwere made. The images or other third party material in this article are \nincluded in the article’s Creative Commons licence, unless indicated \notherwise in a credit line to the material. If material is not included in \nthe article’s Creative Commons licence and your intended use is not \npermitted by statutory regulation or exceeds the permitted use, you will \nneed to obtain permission directly from the copyright holder. To view a \ncopy of this licence, visit http://creativecommons.org/licenses/by/4.0/.\nReferences\n 1. Wong ND, Young D, Zhao Y, Nguyen H, Caballes J, Khan I et al \n(2016) Prevalence of the American College of Cardiology/Ameri-\ncan Heart Association statin eligibility groups, statin use, and \nlow-density lipoprotein cholesterol control in US adults using the \nNational Health and Nutrition Examination Survey 2011–2012. J \nClin Lipidol 10(5):1109–1118\n 2. Vancheri F, Backlund L, Strender LE, Godman B, Wettermark B \n(2016) Time trends in statin utilisation and coronary mortality in \nWestern European countries. BMJ Open 6(3):e010500\n 3. Adedinsewo D, Taka N, Agasthi P, Sachdeva R, Rust G, Onwuanyi \nA (2016) Prevalence and factors associated with statin use among \na nationally representative sample of US adults: National Health \n\n867European Journal of Clinical Pharmacology (2024) 80:855–867 \nand Nutrition Examination Survey, 2011–2012. Clin Cardiol \n39(9):491–496\n 4. De La Cruz JA, Mihos CG, Horvath SA, Santana O (2019) The \npleiotropic effects of statins in endocrine disorders. Endocr Metab \nImmune Disord Drug Targets 19(6):787–793\n 5. Zeybek B, Costantine M, Kilic GS, Borahay MA (2018) Thera-\npeutic roles of statins in gynecology and obstetrics: the current \nevidence. Reprod Sci 25(6):802–817\n 6. Vitagliano A, Noventa M, Quaranta M, Gizzo S (2016) Statins as \ntargeted “magical pills” for the conservative treatment of endome-\ntriosis: may potential adverse effects on female fertility represent \nthe “dark side of the same coin”? A systematic review of litera-\nture. Reprod Sci 23(4):415–428\n 7. William GP, Suhail M, Jaferi F, Nasim M (2014) Effects of simv-\nastatin 20mg on the histology of albino rat ovary. Pakistan Journal \nof Medical and Health Sciences 8(4):1091–1099\n 8. Jiao XF, Li HL, Jiao XY, Guo YC, Zhang C, Yang CS et al (2020) \nOvary and uterus related adverse events associated with statin use: \nan analysis of the FDA Adverse Event Reporting System. Sci Rep \n10(1):11955\n 9. Duggirala HJ, Tonning JM, Smith E, Bright RA, Baker JD, Ball R \net al (2016) Use of data mining at the Food and Drug Administra-\ntion. J Am Med Inform Assoc 23(2):428–434\n 10. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J et al \n(2015) UK biobank: an open access resource for identifying the \ncauses of a wide range of complex diseases of middle and old age. \nPLoS Med 12(3):e1001779\n 11. Jiao X, Li H, Zeng L, Han L, Yang H, Hu Y et al (2023) Use of \nstatins and risk of uterine leiomyoma: a cohort study in the UK \nBiobank. J Evid Based Med 16(4):424–427\n 12. UK Biobank (2021) Touch Screen Questionnaire Version \n1.0.  https:// bioba nk. ndph. ox. ac. uk/ showc ase/ refer. cgi? id= 100247. \nAccessed 18 Sep 2022\n 13. UK Biobank (2020) Data providers and dates of data availability.  \nhttps:// bioba nk. ndph. ox. ac. uk/ showc ase/ exinfo. cgi? src= Data_  \nprovi ders_ and_ dates.    Accessed 18 Sep 2022\n 14. Liu Z, Luo Y, Ren J, Yang L, Li J, Wei Z et al (2022) Associa -\ntion between fish oil supplementation and cancer risk according \nto fatty fish consumption: a large prospective population-based \ncohort study using UK Biobank. Int J Cancer 150(4):562–571\n 15. UK Biobank. The verbal interview within ACE centres. 2012. \nhttps:// bioba nk. ndph. ox. ac. uk/ showc ase/ ukb/ docs/ Inter view. pdf. \nAccessed 18 Sep 2022).\n 16. Knuppel A, Papier K, Fensom GK, Appleby PN, Schmidt JA, \nTong TYN et al (2020) Meat intake and cancer risk: prospective \nanalyses in UK Biobank. Int J Epidemiol 49(5):1540–1552\n 17. Townsend P, Phillimore P, Beattie A (1988) Health and Depriva-\ntion - Inequality and the North. Routledge, London\n 18. Sarri G, Davies M, Lumsden MA, Guideline DG (2015) Diagno-\nsis and management of menopause: summary of NICE guidance. \nBMJ 351:h5746\n 19. Yaddanapudi LN (2016) The American Statistical Association \nstatement on P-values explained. J Anaesthesiol Clin Pharmacol \n32(4):421–423\n 20. Akinwunmi B, Vitonis AF, Titus L, Terry KL, Cramer DW (2019) \nStatin therapy and association with ovarian cancer risk in the New \nEngland Case Control (NEC) study. Int J Cancer 144(5):991–1000\n 21. Lavie O, Pinchev M, Rennert HS, Segev Y, Rennert G (2013) The \neffect of statins on risk and survival of gynecological malignan-\ncies. Gynecol Oncol 130(3):615–619\n 22. Kim DS, Ahn HS, Kim HJ (2022) Statin use and incidence and \nmortality of breast and gynecology cancer: a cohort study using \nthe National Health Insurance claims database. Int J Cancer \n150(7):1156–1165\n 23. Sperling CD, Verdoodt F, Friis S, Dehlendorff C, Kjaer SK \n(2017) Statin use and risk of endometrial cancer: a nationwide \nregistry-based case-control study. Acta Obstet Gynecol Scand \n96(2):144–149\n 24. Baandrup L, Dehlendorff C, Friis S, Olsen JH, Kjaer SK (2015) \nStatin use and risk for ovarian cancer: a Danish nationwide case-\ncontrol study. Br J Cancer 112(1):157–161\n 25. Huang J, Deng Y, Boakye D, Tin MS, Lok V, Zhang L et al \n(2022) Global distribution, risk factors, and recent trends for cer-\nvical cancer: a worldwide country-level analysis. Gynecol Oncol \n164(1):85–92\n 26. Zhang S, Xu H, Zhang L, Qiao Y (2020) Cervical cancer: epi-\ndemiology, risk factors and screening. Chin J Cancer Res \n32(6):720–728\n 27. Srivastava S, Shahi UP, Dibya A, Gupta S, Roy JK (2014) Dis-\ntribution of HPV genotypes and involvement of risk factors in \ncervical lesions and invasive cervical cancer: a study in an Indian \npopulation. Int J Mol Cell Med 3(2):61–73\n 28. Makuza JD, Nsanzimana S, Muhimpundu MA, Pace LE, Ntaganira \nJ, Riedel DJ (2015) Prevalence and risk factors for cervical cancer \nand pre-cancerous lesions in Rwanda. Pan Afr Med J 22:26\n 29. Gonyeau MJ, Yuen DW (2010) A clinical review of statins and \ncancer: helpful or harmful? Pharmacotherapy 30:177–194\n 30. Lamprecht J, Wójcik C, Jakóbisiak M, Stoehr M, Schrorter D, \nPaweletz N (1999) Lovastatin induces mitotic abnormalities in \nvarious cell lines. Cell Biol Int 23(1):51–60\n 31. Fujimoto M, Higuchi T, Hosomi K, Takada M (2015) Associa-\ntion between statin use and cancer: data mining of a spontane-\nous reporting database and a claims database. Int J Med Sci \n12(3):223–233\n 32. Desai P, Wallace R, Anderson ML, Howard BV, Ray RM, Wu C \net al (2018) An analysis of the association between statin use and \nrisk of endometrial and ovarian cancers in the Women’s Health \nInitiative. Gynecol Oncol 148(3):540–546\n 33. Haukka J, Sankila R, Klaukka T, Lonnqvist J, Niskanen L,  \nTanskanen A et al (2010) Incidence of cancer and statin usage–\nrecord linkage study. Int J Cancer 126(1):279–284\n 34. Lamba P, Sharma D, Sinnarkar VV (2022) Polycystic ovarian \nsyndrome treated with individualized homeopathy: a case report. \nAltern Ther Health Med 28(6):60–64\nPublisher's Note Springer Nature remains neutral with regard to \njurisdictional claims in published maps and institutional affiliations.","source_license":"CC0","license_restricted":false}