Factors Influencing Induced Abortion in Women Aged 18–45 Years: A Nationwide Web-Based Questionnaire Survey in Japan

Abstract

Abortion remains a significant public health concern and often results from unintended pregnancies. Although abortion rates have declined, approximately 40% are still due to unplanned pregnancies. We conducted a subanalysis of the FIKA study, a survey on sexuality among Japanese women aged 18–45 years, to examine factors correlated with abortion. We analyzed responses obtained from 3228 sexually experienced women, who were categorized based on abortion history. Sources of knowledge about intercourse and contraceptives were assessed using a standardized questionnaire. A decision tree analysis was performed to identify associated risk factors. Among the participants, 410 women (12.7%) had undergone an abortion. Women who had undergone an abortion were significantly more likely to rely on friends as a source of contraceptive information than women who had not had an abortion (p = 0.0024). Participants with abortion experience showed greater variation in knowledge about contraceptive scores, with most of them being in the lowest knowledge group. Decision tree analysis revealed two key factors: sexual debut before the age of 17 years and only compulsory education. The abortion rate was 39.5% among women with both factors, compared to 7.9% among those with none. Early sexual initiation and limited education were strongly associated with abortion experience. Reliance on informal sources may also contribute to a poor understanding of reproductive health. These findings underscore the need for early, comprehensive, and formal sex education to improve sexual literacy and reduce unintended pregnancies.

INTRODUCTION

Unintended pregnancies remain a significant global public health concern, with an estimated 121 million cases occurring annually, and account for approximately half of all pregnancies worldwide.¹⁾ Of these unintended pregnancies, 61% result in induced abortion, amounting to approximately 73 million abortions performed each year globally.^2,3)

While abortion rates have declined in developed countries since the 1990s, they remain high in developing countries. Between 2010 and 2014, 45.1% of all abortions globally were considered unsafe, with 97% of these occurring in developing countries.^4,5)

In Japan, the Ministry of Health, Labour and Welfare reported approximately 130000 induced abortions in 2021; this figure was substantially lower than that reported two decades earlier (330000 cases).⁶⁾ A study using data from the 2021 National Fertility Survey found that among 5861 couples, 1050 (17.9%) had experienced an unintended pregnancy, 296 (5.1%) had undergone an induced abortion, and 228 (3.9%) had an abortion resulting from an unintended pregnancy.⁷⁾

Induced abortion carries both physical and psychological risks. Physical risks include short-term complications, such as hemorrhage, infection, and injury to the uterus or surrounding organs, as well as long-term consequences, such as infertility and menstrual irregularities.⁸⁾ Psychological risks may involve stress, anxiety, and postabortion depression.⁹⁾ In rare cases, maternal death may also occur as a result of induced abortion.¹⁰⁾

Our previous research, “Questionnaire for Fact-finding survey to know about emergency contraceptives, pregnancy, and childbirth of women aged 18 to 45 (FIKA study),” a survey on sexuality conducted among women aged 18–45 years, revealed that nearly half of sexually active women lacked adequate knowledge about sexual and reproductive health and that a notable proportion had experienced abortion.¹¹⁾ To effectively prevent unintended pregnancies and support women’s health, it is essential to conduct a comprehensive analysis of the factors associated with induced abortion and develop evidence-based, multifaceted interventions.

However, previous studies have not sufficiently examined the systematic risk factors associated with induced abortion. Therefore, the present study aimed to investigate the contributing factors of induced abortion through a subanalysis of data from the FIKA study.

MATERIALS AND METHODS

Data Source and Participant Identification

This study is a secondary analysis of data from the nationwide “FIKA study,” a web-based survey designed to assess emergency contraceptive use, pregnancy, and childbirth among women aged 18–45 years in Japan. The reason for setting the survey target age range to 18–45 years was that over 99% of childbirths occur at this age in Japan.¹²⁾ The survey included 5752 responses to 30 structured questions across 4 domains, with the sample stratified by age and geographic region to ensure representativeness. In our study, “primary caregiver” referred to the person(s) who mainly provided care for the participant during childhood (under 17 years of age). In addition, “sexual intercourse with an unknown partner” was defined as sexual intercourse with someone the participant did not have an ongoing or continuing relationship with after the sexual encounter. Furthermore, we categorized educational attainment as up to compulsory education in Japan (junior high school) or higher. For this subanalysis focusing on abortion, the following exclusion criteria were applied: (1) incomplete responses, (2) potentially unreliable responses, and (3) participants without a history of sexual intercourse. Potentially unreliable responses were identified through independent review by two researchers, with disagreements resolved through consensus. Responses were excluded if they met either of the following criteria: (1) inconsistencies between answers to related questions (e.g., logically incompatible responses) and (2) clearly noncredible or meaningless content in the free-text comments (e.g., nonsensical character strings). After applying these criteria, 3228 responses were included in the final analysis. We constructed a classification tree using recursive partitioning. We developed a unique sexual knowledge score for emergency contraception knowledge (Supplementary Table 1). It consisted of multiple-choice items between two options, and participants earned 1 point for each correct answer (total score range: 0–5). For the main analysis, we categorized the total score as 0–1 (low knowledge) or 2 or higher (high knowledge) to match the distribution of the observed data and improve interpretability. To support content validity, the items were developed through repeated discussions within a multidisciplinary expert panel (male and female physicians, including an obstetrician–gynecologist, male and female pharmacists, a dentist, an ethics specialist, a statistician, and an epidemiologist), and the final content was refined through consensus.

Study Endpoints and Statistical Analysis

The primary endpoint of this study was to identify risk factors associated with abortion, while the secondary endpoint was to evaluate the association between abortion history and sexual knowledge. Participants were categorized based on abortion history, and group comparisons were conducted using the chi-squared test or Wilcoxon’s rank-sum (Mann–Whitney U) test, as appropriate. To identify factors associated with abortion, a decision tree model was developed using the rpart package in R. Candidate variables included age at first sexual intercourse, educational attainment, primary caregiver during adolescence, and level of knowledge about contraceptives. In addition, associations between sources of sexual education, such as sources of information about sexual intercourse, contraception, and abortion history, were evaluated using the chi-squared or Fisher’s exact test, as appropriate. We constructed a classification tree using recursive partitioning. An unpruned tree was first grown, and cost–complexity pruning was guided by 10-fold cross-validation to minimize the cross-validated misclassification error. To ensure interpretability a priori, we additionally required that the final model contain at least one split. For the primary analysis, we prespecified a minimum terminal node size of 50 observations and collapsed any split that produced terminal nodes smaller than this threshold. Sensitivity analyses used an uncollapsed tree (Supplementary Fig. 1) and a smaller minimum terminal node size of 20 observations (Supplementary Fig. 2). The generational distribution of abortion history and knowledge about contraceptive scores, based on five structured questions, was visualized using 3-dimensional bar plots. Data management and analysis were conducted using JMP 17 (SAS Institute Inc., Cary, NC, U.S.A.) and R version 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria).

Ethics Approval

Data were collected anonymously. Informed consent was obtained from all participants using an appropriate web-based method. The study protocol was approved by the institutional review board of Showa Medical University on November 9, 2022 (Approval No. 22-191-A).

RESULTS

Participants’ Characteristics

Of the 3228 women with a history of sexual intercourse in the FIKA study, 410 (12.7%) had undergone at least one abortion (Fig. 1). Women with a history of abortion had a higher median age than those without it (37 vs. 33 years, p < 0.001) (Table 1). The age at first sexual intercourse was significantly younger among women with an abortion history than among those without it (17 vs. 19 years, p < 0.001). In addition, women with an abortion history were more likely to report having sexual intercourse with an unknown partner than those without it (64.1 vs. 47.0%, p < 0.001).

Fig. 1. Participant Identification Flow

Table 1. Patient Characteristics

	With abortion experience (n = 410)	Without abortion experience (n = 2818)	p-Value
Age (year), median [IQR]	37 [31–42]	33 [25–39]	<0.001
Sexually transmitted diseases, n (%)	107 (26.1)	282 (10.0)	<0.001
Age at the time of first intercourse, median [IQR]	17 [16–19]	19 [17–21]	<0.001
Experience of sexual intercourse with unknown person(s), n (%)	263 (64.1)	1324 (47.0)	<0.001

Primary Endpoint: Factors Associated with Abortion

Decision tree analysis revealed several key factors associated with abortion history (Fig. 2). The most significant predictor was age at the time of first sexual intercourse. Participants who initiated sexual activity before the age of 17 years were at a substantially higher prevalence of abortion than those who began at the age of ≥18 years. Among those who initiated sexual activity early, those who had only compulsory education (e.g., less than high school) had the highest prevalence of abortion. Among women who met both conditions, that is, early sexual initiation and only compulsory education, the abortion rate was 39.5% (n = 81), which was significantly higher than that observed among those meeting none of the criteria (7.9% [n = 2268]). In the primary analysis enforcing a minimum terminal node size of 50, the tree first split on age at the time of first intercourse, followed by education level among participants who had first intercourse before the age of 17 years (Fig. 2). The frequency of abortion history across terminal nodes was 180/2268 (7.9%) among those who had first intercourse at the age of ≥18 years, 198/879 (22.5%) among those who had first intercourse before the age of 17 years and attained education at high school/technical school/university or higher, and 32/81 (39.5%) among those who had first intercourse before the age of 17 years and attained compulsory education. No further splits satisfied the minimum terminal node size criterion (Fig. 2). The uncollapsed tree (Supplementary Fig. 1) included terminal nodes with small sample sizes, with a minimum terminal node size of 9. In the sensitivity analysis using a minimum terminal node size of 20 (Supplementary Fig. 2), an additional split by primary caregiver appeared within the subgroup of participants who had first intercourse before the age of 17 years and attained compulsory education (n = 81). In this subgroup, the frequency of abortion history was 22/52 versus 10/29 across caregiver categories (Fisher’s exact test p = 0.636). Overall, the decision tree revealed that early sexual initiation and lower educational attainment were strongly associated with abortion history.

Fig. 2. Decision Tree Analysis Identifying Factors Associated with Abortion History

Secondary Endpoint: Association between Abortion and Sexual Knowledge

We compared sources of knowledge about intercourse and contraceptives between participants with and without abortion experience (Fig. 3). Those with abortion history were more likely to have acquired sources of knowledge about intercourse from friends (p = 0.0470) and sources of knowledge about contraceptives from friends (p = 0.0024), as evidenced by significant between-group differences. In addition, a higher proportion of abortion-experienced participants reported not having received formal education on sexual intercourse (p = 0.052). Approximately 20% of participants with an abortion history reported learning about sexual intercourse through school lectures.

Fig. 3. Comparison of Information on (a) Sources of Knowledge about Sexual Intercourse and (b) Sources of Knowledge about Contraceptives between Participants with and without Abortion History

The distribution of scores on the knowledge about contraceptive questionnaire (range: 0–5), stratified by age group and abortion history, is shown in Fig. 4. Knowledge about contraceptive scores was generally consistent across age groups in both cohorts. Participants with abortion experience showed greater variation in knowledge of contraceptive scores, with most being in the lowest knowledge group.

Fig. 4. Three-Dimensional Distribution of Contraceptive Knowledge Scores by Age, Comparing Participants with and without Abortion History

(a) Scores on contraceptive knowledge questionnaire with abortion experience. (b) Score on the contraceptive knowledge questionnaire without abortion experience.

DISCUSSION

In this study, we examined the association between induced abortion history and several social and educational factors among women who participated in the FIKA study. Our findings evidenced that early sexual initiation (before the age of 17 years) and low educational attainment were significant predictors of abortion experience.

Women who initiated sexual activity before the age of 17 years had a significantly higher rate of abortion history than those who initiated sexual activity later (Fig. 2). This is consistent with the results of a Japanese study that showed a positive correlation between first sexual intercourse at 10–19 years and abortion history.¹³⁾ National statistics have indicated that abortion incidence and trends vary across high-income countries. In the United States, the abortion rate was 11.2 per 1000 women aged 15–44 years in 2022,¹⁴⁾ whereas in England and Wales, the age-standardized abortion rate was 18.6 per 1000 women aged 15–44 years in 2021, with increases since 2012 and heterogeneous age-specific trends.¹⁵⁾ Importantly, several European studies have shown that socioeconomic inequalities persist even when overall rates are low or changing.¹⁶⁾ For example, in Denmark, the abortion rate peaked in the mid-1970s and declined by the mid-2000s, yet social vulnerability remained associated with induced abortion.¹⁷⁾ In Finland, the overall abortion rate has remained comparatively low since the mid-1990s, but the proportion of repeat abortions has increased over time and educational inequalities have become evident.¹⁸⁾ In Barcelona (Spain), induced abortion rates increased between the 1990s and early 2000s, with the increase being most pronounced among women with lower educational qualifications.¹⁹⁾ These findings support the plausibility that abortion history can disproportionately accumulate in socially disadvantaged groups, even within broadly comparable high-income settings. Early sexual initiation may be associated with an increased likelihood of a history of unintended pregnancy and may limit opportunities to acquire accurate information about sexual behavior and contraception beforehand.

Women with an abortion history were more likely to report having sexual intercourse with an unknown partner. Previous studies have suggested that sexual intercourse outside of ongoing partnerships is not uncommon in Japan. Ghaznavi et al. reported that sexual activity tends to be relatively inactive at the population level in Japan, while a substantial proportion of men reported use of commercial sex services (approximately 48%), suggesting that sexual intercourse outside of continuous partner relationships occurs more frequently than that in some other countries.²⁰⁾ In addition, the Japan Family Planning Association (JEX Japan Sex Survey 2020) reported that 41.1% of men and 31.4% of women had experienced sexual intercourse with someone other than a partner (e.g., a boyfriend/girlfriend or spouse).²¹⁾ These figures are comparable to those observed in our study (with abortion: 64.1%; without abortion: 47.0%), supporting the plausibility of our findings.

We also noted that the recruitment strategy of the FIKA study may have influenced the participant characteristics, potentially resulting in a sample with greater sexual experience or higher engagement in reproductive and sexual health–related issues than the general population. This may partly explain the relatively high prevalence observed in this study.

Educational attainment was also strongly associated with abortion experience. Women whose education was limited to compulsory schooling had significantly higher abortion rates than those with higher education (Fig. 2). This finding supports the results of a previous study that reported that lower education was correlated with increased abortion prevalence, whereas higher education was associated with a greater use of modern contraceptive methods and reduced risk of unintended pregnancies.²²⁾ Another study reported that women with some college education had the highest abortion rate (26 per 1000), whereas college graduates had the lowest abortion rate (13 per 1000).²³⁾ In addition, repeated abortions were more frequently observed among women with lower educational attainment than among those with a university degree.²⁴⁾ These findings suggest that education alone may not linearly predict abortion risk. Our results help clarify this complexity and support the broader notion that lower educational levels are linked to greater abortion experience.

Knowledge about intercourse and contraceptives was often acquired through informal sources, such as friends, rather than formal education (Fig. 3). Analysis of contraceptive knowledge by age group showed that, compared with women without abortion experience, women with abortion history had greater variability in correct response rates across age groups. Notably, among women with abortion experience, the lowest knowledge group was more prevalent (Fig. 4). Women with no knowledge about contraceptive methods were at particularly high prevalence, highlighting the importance of accurate and reliable information. These findings underscore the necessity of delivering comprehensive sex education from formal sources.²⁵⁾ Our survey supports the implementation of structured educational programs to improve sexual literacy and may reduce abortion prevalence among women in Japan.

Limitations

This study had some limitations. First, the cross-sectional design limited the ability to establish causal relationships between the identified factors and abortion history. Second, the data were self-reported, which may have introduced recall bias, particularly in responses related to sexual behavior and abortion. Third, the study population was limited to the participants of the FIKA study, which may not be fully representative of all women of reproductive age in Japan. Despite these limitations, our findings contribute to a better understanding of the multifactorial risk factors associated with abortion among young women. Notably, cross-national evidence was used only for contextual background, and differences in legal, healthcare, and sociocultural environments limit direct comparability. In particular, this study underscores the critical importance of accessible and accurate information about sexual activity and contraception as well as the influence of educational background on reproductive decision-making.

CONCLUSION

This study identified early sexual initiation and limited education as key factors associated with induced abortion among Japanese women. These findings highlight the urgent need for comprehensive, early sex education to improve reproductive health literacy and may reduce the unintended pregnancies in this population.

DECLARATIONS

Funding

This study was supported by a Grant from the OTC Self-Medication Promotion Foundation (Grant Nos. 58-3-2 and 36-3-1).

Author Contributions

All authors meet the ICMJE recommendations. HI and K Momo contributed to the study conception, drafted the manuscript, and collected raw data; HI, YK, HH, NT, YO, and K Momo created and confirmed the definitions of the analysis; HI, HT, HM, and K Momo conducted data management. KM, KR, NH, TS, and NS performed clinical interpretation and provided the final approval. All authors participated in the discussions during the manuscript preparation. All authors have agreed to publish the final version of the manuscript.

Conflict of Interest

K Momo received funding from Hisamitsu and honoraria from Hisamitsu, Nippon-Kayaku, Sawai, and AbbVie for presentations. The Department of Hospital Pharmaceutics, School of Pharmacy, Showa Medical University received funding from Ono for a contract research project under a collaborative research agreement.

As a potential conflict of interest, the department also received research grants from Nippon-Kayaku, Ono, Shionogi, Bayer, Daiichi Sankyo, Eisai, Mochida, and Taiho. The remaining authors declare no conflicts of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1)  United Nations Population Fund. 2022. “Seeing the Unseen: the case for action in the neglected crisis of unintended pregnancy.”: <https://www.unfpa.org/sites/default/files/pub-pdf/EN_SWP22%20report_0.pdf>, accessed 14 July, 2025.
• 2)  World Health Organization. 2022. “Abortion care guideline. World Health Organization.”: <https://www.who.int/publications/i/item/9789240039483>, accessed 14 July, 2025.
• 3)  Bearak J, Popinchalk A, Ganatra B, Moller AB, Tunçalp Ö, Beavin C, Kwok L, Alkema L. Unintended pregnancy and abortion by income, region, and the legal status of abortion: estimates from a comprehensive model for 1990-2019. Lancet Glob. Health, 8, e1152–e1161 (2020).
• 4)  Sedgh G, Bearak J, Singh S, Bankole A, Popinchalk A, Ganatra B, Rossier C, Gerdts C, Tunçalp Ö., Johnson BR Jr, Johnston HB, Alkema L. Abortion incidence between 1990 and 2014: global, regional, and subregional levels and trends. Lancet, 388, 258–267 (2016).
• 5)  Ganatra B, Gerdts C, Rossier C, Johnson BR Jr, Tunçalp Ö, Assifi A, Sedgh G, Singh S, Bankole A, Popinchalk A, Bearak J, Kang Z, Alkema L. Global, regional, and subregional classification of abortions by safety, 2010–14: estimates from a Bayesian hierarchical model. Lancet, 390, 2372–2381 (2017).
• 6)  Ministry of Health. Labour and Welfare, 2023. “Report on public health administration and services.”: <https://www.mhlw.go.jp/toukei/saikin/hw/eisei_houkoku/23/dl/gaikyo.pdf>, accessed 14 July, 2025.
• 7)  Okui T. Association of socioeconomic status with unintended pregnancy and induced abortion in married couples: an analysis of the 2021 national survey data in Japan. Clin. Epidemiol. Glob. Health, 30, 101814 (2024).
• 8)  Bridwell RE, Long B, Montrief T, Gottlieb M. Post-abortion complications: a narrative review for emergency clinicians. West. J. Emerg. Med., 23, 919–925 (2022).
• 9)  Major B, Cozzarelli C, Cooper ML, Zubek J, Richards C, Wilhite M, Gramzow RH. Psychological responses of women after first-trimester abortion. Arch. Gen. Psychiatry, 57, 777–784 (2000).
• 10)  Bartlett LA, Berg CJ, Shulman HB, Zane SB, Green CA, Whitehead S, Atrash HK. Risk factors for legal induced abortion-related mortality in the United States. Obstet. Gynecol., 103, 729–737 (2004).
• 11)  Momo K, Maeda E, Hattori H, Isozaki H, Takita H, Morohoshi H, Ryu K, Hida N, Sambe T, Shirato N. Descriptive study on a nationwide exploratory questionnaire survey of emergency contraceptive pills and their sexual history and knowledge in Japan. Biol. Pharm. Bull., 46, 1296–1303 (2023).
• 12)  Ministry of Health, Labour and Welfare. “Vital Statistics of Japan 2024: Monthly Report (Annual Total, Preliminary),” accessed 6 January, 2026.
• 13)  Kojo T, Ae R, Tsuboi S, Nakamura Y, Kitamura K. Differentials in variables associated with past history of artificial abortion and current contraception by age: results of a randomized national survey in Japan. J. Obstet. Gynaecol. Res., 43, 516–522 (2017).
• 14)  Ramer S, Nguyen AT, Hollier LM, Rodenhizer J, Warner L, Whiteman MK. Abortion surveillance—United States, 2022. MMWR Surveill. Summ., 73, 1–28 (2024).
• 15)  Office for Health Improvement & Disparities. “Abortion statistics, England and Wales: 2021. [updated 2025 Nov 4].”: <https://www.gov.uk/government/statistics/abortion-statistics-for-england-and-wales-2021/>, accessed 7 January, 2026.
• 16)  Scott RH, Bajos N, Slaymaker E, Wellings K, Mercer CH. Understanding differences in conception and abortion rates among under-20 year olds in Britain and France: examining the contribution of social disadvantage. PLOS ONE, 17, 648 (2017).
• 17)  Rasch V, Gammeltoft T, Knudsen LB, Tobiassen C, Ginzel A, Kempf L. Induced abortion in Denmark: effect of socio-economic situation and country of birth. Eur. J. Public Health, 18, 144–149 (2008).
• 18)  Väisänen H. Educational inequalities in repeat abortion: a longitudinal register study in Finland 1975–2010. J. Biosoc. Sci., 48, 820–832 (2016).
• 19)  Pérez G, García-Subirats I, Rodríguez-Sanz M, Díez E, Borrell C. Trends in inequalities in induced abortion according to educational level among urban women. J. Urban Health, 87, 524–530 (2010).
• 20)  Ghaznavi C, Ueda P, Okuhama A, Sakamoto H. Sexual behaviors among individuals aged 20–49 in Japan: initial findings from a quasi-representative national survey, 2022. J. Sex Res., 60, 317–324 (2023).
• 21)  Japan Family Planning Association. “The 4th JEX Japan Sex Survey 2020: Summary of Survey Results.” Family Planning Research Center. accessed 6 January, 2026.
• 22)  Weitzman A. The effects of women’s education on maternal health: evidence from Peru. Soc. Sci. Med., 180, 1–9 (2017).
• 23)  Jones RK, Darroch JE, Henshaw SK. Patterns in the socioeconomic characteristics of women obtaining abortions in 2000-2001. Perspect. Sex. Reprod. Health, 34, 226–235 (2002).
• 24)  Tang L, Wu S, Liu D, Temmerman M, Zhang WH. Repeat induced abortion among Chinese women seeking abortion: two cross-sectional studies. Int. J. Environ. Res. Public Health, 18, 4446 (2021).
• 25)  Klutsey EE, Ankomah A. Factors associated with induced abortion at selected hospitals in the Volta Region. Ghana Int. J. Womens Health, 6, 809–816 (2014).