Trends and factors associated with previous induced abortions among young women in Foshan, China: insights from a retrospective cross-sectional study

Abstract

Objectives

This study aims to explore the history of prior abortions and the factors influencing them among young women seeking abortion services in Foshan, Guangdong, China.

Design

This is a retrospective, cross-sectional study of young women seeking abortion care.

Setting

Gynaecological outpatient clinics at the Department of Gynaecology, Foshan Women and Children Hospital Affiliated to Guangdong Medical University, Foshan, Guangdong, China, between 2013 and 2023.

Participants

A total of 7361 young women aged 12–25 years seeking abortion services.

Measures

Data on abortion history, sociodemographic factors, contraceptive use and postabortion contraceptive choices were collected and analysed, with special emphasis on the incidence of repeat abortions and the factors associated with them.

Results

Of the 7361 participants, 34.2% reported at least one previous abortion, underscoring a notable public health concern. The mean age of the participants was 22.30 years (SD=2.13). Women with a history of abortion were significantly older than those without (22.57 vs 22.08 years, p<0.001). Logistic regression analyses suggested higher, but not statistically significant, odds of prior abortion among women aged 20–25 years compared with those aged 12–19 years (adjusted OR (aOR) 1.23, 95% CI 0.93 to 1.63, p=0.151). There was no significant difference in the odds of prior abortion between urban and rural residents (rural vs urban: aOR 0.94, 95% CI 0.67 to 1.32, p=0.732), and single or divorced women had higher odds of reporting a prior abortion than married women (aOR 1.39, 95% CI 1.12 to 1.73, p=0.003). Importantly, while 13.92% of women reported using no contraceptive method at the time of conception, postabortion counselling reduced non-acceptance to only 3.65%, demonstrating the effectiveness of intervention but highlighting remaining gaps in prepregnancy contraceptive access. Women with a prior abortion history demonstrated higher postabortion contraceptive acceptance (80.30% vs 77.25%, p<0.01) and increased uptake of long-acting reversible contraception, particularly intrauterine devices and systems (18.96% vs 10.76%, p<0.001).

Conclusions

This study underscores the need for targeted interventions to address the multifaceted factors leading to repeat abortions among young women in China. The results offer valuable insights for improving reproductive health outcomes in this vulnerable population and highlight the importance of expanding access to contraceptive education and services in China.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Large, consecutive hospital-based dataset spanning 10 years (2013–2023).

• Focus on a defined, policy-relevant age group (12–25 years).

• Standardised case collection forms and trained data collectors across the study period.

• Retrospective single-centre design, which may limit generalisability.

• Reliance on self-reported history and clinical records may introduce information bias.

Background

Abortion, defined as the termination of a pregnancy, whether induced or spontaneous, before 28 weeks of gestation, represents a significant global health concern.¹ It is estimated that approximately 55.9 million unsafe abortions occur annually worldwide, with 49.3 million of these occurring in developing countries.^{2 3} Unsafe abortion remains a leading cause of maternal mortality and morbidity,⁴ accounting for 13% of global maternal deaths⁵ and resulting in 5 million disabilities each year.⁶ Enhancing education on sexual behaviour, family planning and ensuring access to safe abortion services could significantly mitigate the incidence of unsafe abortions.

Abortion, a widespread medical procedure, constitutes a critical aspect of reproductive health and carries substantial public health implications, particularly among young women.^{7 8} In China, where access to abortion services is relatively straightforward, the prevalence of induced abortions among adolescents and young adults has garnered considerable attention in recent years.⁹ Specifically, young women aged 12 to 25 years constitute a vulnerable population concerning unintended pregnancies and subsequent abortions, as they often engage in exploratory sexual behaviours while lacking sufficient knowledge of contraception and reproductive health.

Research in this domain has underscored several factors contributing to the elevated risk of induced abortion among young women, including limited access to comprehensive sexuality education, inadequate contraceptive practices and societal pressures about childbearing and sexuality.^{10 11} Given the potentially adverse consequences associated with recurrent abortions, such as physical complications, psychological distress and social stigma,¹²,¹⁴ it is imperative to attain a comprehensive understanding of the sociodemographic and reproductive characteristics of this population to devise effective interventions.

To this end, the present study aims to conduct a cross-sectional analysis of young women aged 12 to 25 years seeking abortion-related care in China, with a particular emphasis on their history of previous induced abortions. By investigating the demographic and reproductive factors associated with prior abortion experiences, we seek to identify key predictors of recurrent abortions and to elucidate the underlying mechanisms driving these behaviours. Furthermore, this investigation emphasises the importance of addressing the unmet need for contraception among young women, particularly those residing in urban areas and those who are married or have given birth. By elucidating the relationships between sociodemographic variables, reproductive characteristics and previous abortion history, we aim to inform the development of targeted interventions aimed at reducing the burden of recurrent abortions and improving the overall reproductive health outcomes of this vulnerable population.

Methods

Data sources and sampling strategy

This study was conducted in Foshan City, Guangdong Province, China, which is administratively divided into five districts: Chancheng, Nanhai, Shunde, Sanshui and Gaoming. As reported in the Foshan Statistical Bulletin, the city’s resident population at the end of 2024 was 9 698 900 (approximately 9.70 million), with an urbanisation rate of 95.6%. For comparison, the resident population at the end of 2023 was 9 615 400.

Participant recruitment occurred at Foshan Women and Children Hospital Affiliated to Guangdong Medical University, a tertiary (Class A) maternal–child speciality hospital that also functions as the municipal guidance centre for maternal and child health. The hospital offers comprehensive medical services, referrals and technical guidance to all five districts of Foshan, serving primarily its residents while also accepting referrals from neighbouring regions. Multiple campuses and specialised functions are outlined in the hospital profile and municipal health authority documents.

For clarity, “RPL” in this study denotes recurrent spontaneous pregnancy loss, which is distinct from induced abortion (including medication-induced abortion). Foshan is served by multiple providers of abortion care, including family-planning and postabortion care clinics in public hospitals, as well as medication abortion services in private facilities. As such, our single-centre, hospital-based cohort does not represent the sole or exhaustive source of abortion services within the city.

Foshan Women and Children Hospital Affiliated to Guangdong Medical University, Foshan, Guangdong, China, is the city’s main tertiary (Grade-3A) maternal and child health centre, operating multiple branches and routinely accepting referrals from all five Foshan districts. These features support its role as an important source of abortion care in the municipal health system. At the same time, we acknowledge that private clinics, family-planning clinics and smaller hospitals also provide abortion services in Foshan; therefore, our hospital-based cohort may not capture women who exclusively attend non-hospital providers.

Potential selection bias may arise because our single-centre, hospital-based sample may not represent women who exclusively seek abortion care at private clinics or other hospitals. Additionally, the primary outcome (history of prior induced abortion) is self-reported and may be subject to recall and social-desirability bias despite efforts to collect data confidentially and cross-check records where possible.

Data collection

Throughout the decade, the research team systematically gathered sociodemographic characteristics, reproductive history, clinical records and physical examination findings of all women receiving abortion-related treatment. To ensure data accuracy and consistency, all healthcare professionals involved underwent rigorous training on administering standardised questionnaires. The data was recorded using a standardised abortion case collection form. To mitigate reporting bias, trained clinicians administered standardised, private interviews and responses were cross-referenced with existing medical records when available; however, residual information bias may remain.

Study population

The analysis focuses on 7361 young women aged 12 to 25 who sought abortion care during the study period, forming the final cohort for analysis.

Variables

The primary outcome was participants’ self-reported history of previous abortions. Each participant was asked whether she had undergone any abortions before the current treatment. In addition, the study assessed sociodemographic and reproductive factors associated with multiple abortions, including age, ethnicity, residential area (urban vs rural) and marital status.^{14 15} Sociodemographic factors comprised age, ethnicity, residence (rural vs urban) and marital status. Reproductive factors included parity, contraceptive method used at the time of the index pregnancy and pregnancy intention. Our multivariable analysis incorporated all sociodemographic characteristics, along with three reproductive factors: parity, contraceptive method and pregnancy intention, previously found to be associated with the likelihood of reporting multiple abortions.^{16 17}

The exclusion criteria were aged >25 years, presentation for spontaneous abortion (miscarriage) rather than induced abortion care, and incomplete demographic or reproductive data.

Of the 73 862 records initially screened, 66 501 were excluded, where 52 391 were excluded due to age over 25 years, 1246 due to spontaneous abortion and 12 864 due to incomplete demographic or fertility-related data, resulting in a final sample of 7361 participants included in the analysis. Because this represents a substantial proportion of screened records, these exclusions could have introduced selection bias if the excluded individuals differed systematically from those included. These criteria ensured a focused analysis on young women at risk of unintended pregnancies and repeat abortions.

Sample size determination and sampling technique

As this is a retrospective cross-sectional study using all available hospital records over 10 years, a formal sample size calculation was not performed. The final cohort of 7361 participants represents the complete population of eligible young women who sought abortion services during the study period, which provides sufficient statistical power for descriptive and regression analyses.

A consecutive sampling approach was applied, including all eligible patients presenting to the hospital’s gynaecology outpatient clinics for abortion care during the study period. This approach minimised selection bias and ensured comprehensive coverage of the target population within the hospital setting.

Statistical analysis

Descriptive statistics were employed to analyse the sociodemographic and reproductive characteristics of the study population. Bivariate and multivariate logistic regression models were used to examine factors associated with having a history of previous induced abortion among women seeking abortion care. Unadjusted ORs were calculated for each sociodemographic and reproductive variable in the binary analyses. Subsequently, multivariate models were developed to evaluate the adjusted correlations. Initially, a model containing solely sociodemographic variables was constructed to generate ORs adjusted for these characteristics. This was followed by a combined model incorporating both sociodemographic and reproductive variables to produce fully adjusted ORs. Statistically significant correlations were identified using the adjusted Wald test, with p values <0.05 considered significant. Data processing and analysis were performed using R V.4.5.1, along with Zstats V.1.0 (www.zstats.net) and Free Statistics analysis platform (V.2.2, Beijing, China).

Patient and public involvement

Patients and the public were not directly involved in setting the research agenda, designing the study or interpreting the results. The study was based on a retrospective analysis of routinely collected hospital data from young women seeking abortion services. Informed consent was obtained from all participants and their families before data collection. Findings from this study will be disseminated through the Foshan Women and Children Hospital, affiliated to Guangdong Medical University’s reproductive health education programmes, to inform contraceptive education and public health interventions for young women in the region.

Result

Of the 73 862 records initially screened, 66 501 were excluded, where 52 391 were excluded due to age over 25 years, 1246 due to spontaneous abortion, and 12 864 due to incomplete demographic or fertility-related data, resulting in a final sample of 7361 participants included in the analysis. (Online supplemental table 1) presents the baseline sociodemographic and reproductive characteristics of young women (12–25 years) seeking abortion care, stratified by abortion history. The overall mean age was 22.30±2.13 years, with women having a history of abortion being significantly older. The mean age increased progressively from 22.08 years among participants with no prior abortions (Group A) to 23.29 years among those with three prior abortions (Group D; p<0.001). The majority of participants (87.90%) were aged 20–25 years, while 12.10% were aged 12–19 years. Han ethnicity predominated (93.06%), and minority groups comprised 6.94%, with no statistically significant difference between groups (p=0.055). Urban residents accounted for 88.40% of the cohort, with higher proportions observed among women with repeated abortions compared with first-time abortion seekers (p<0.001). Marital status varied significantly across groups: 33.97% of women without prior abortions were married, compared with 47.71% among those with one previous abortion and 50.60% among those with two (p<0.001).

Significant differences were observed in reproductive history. More than half of the participants (54.27%) reported a gravida of two or higher, with the proportion exceeding 97% among women with one to four prior abortions (Groups B–E), though it decreased in groups with five or more abortions. Multiparity was more frequent among women with repeated abortions (p<0.001), and a history of caesarean delivery was also more common in this group (p<0.001). Surgical procedures accounted for 97.95% of abortions, whereas medical abortions represented 2.05%, with no significant difference between groups (p=0.587). Uterine morphology was normal in 99.48% of cases. Dysmenorrhoea was reported by 15.51% overall and was more frequent among women without prior abortions (17.22%) compared with those with prior abortions (ranging from 9.92% to 20.00%; p<0.001). Anaemia (1.86%) and menorrhagia (4.85%) were infrequent and did not differ significantly across groups.

Variations were also observed in fertility intentions and contraceptive practices. Only a small proportion of women (4.71%) intended to conceive within the next year, while the vast majority (95.29%) planned to postpone pregnancy (p=0.023). At the time of the index pregnancy, 13.92% of women reported using no contraceptive method, 2.23% had relied on the rhythm method, 3.56% on coitus interruptus and 2.45% on emergency contraception. Importantly, postabortion counselling resulted in only 3.65% declining any contraceptive method, with the majority accepting combined oral contraceptives (79.36%) or long-acting reversible contraception. Postabortion, the vast majority received at least one contraceptive method, most commonly combined oral contraceptives and intrauterine devices (IUDs) (online supplemental table 1) (p<0.001). The choice of contraceptive method varied significantly by abortion history (p<0.001). Combined oral contraceptives were the most frequently used method overall (79.36%); however, their use was less common among women with multiple prior abortions (eg, 67.63% in Group C, 64.89% in Group D and as low as 16.67% in Group F) compared with those with no previous abortions (82.72%). In contrast, the use of intrauterine systems (IUSs) and IUDs increased with a history of repeated abortions (eg, IUS: 6.69% in Group A vs 15.84% in Group C and 18.32% in Group D; IUD: 4.07% in Group A vs 11.07% in Group C; see online supplemental table 1).

Results from the logistic regression analyses of factors associated with a history of induced abortion are presented in online supplemental table 2. In the unadjusted model, women aged 20–25 years were more likely to report a prior abortion than those aged 12–19 years (OR=1.93, 95% CI 1.63 to 2.28); however, this association was attenuated and became non-significant after adjustment (Model 1: OR=1.23, 95% CI 0.93 to 1.63). Ethnic minority women had significantly higher odds of prior abortion compared with Han Chinese women in both the unadjusted model (OR=1.33, 95% CI 1.10 to 1.59) and the adjusted model (OR=1.33, 95% CI 1.02 to 1.73). Conversely, rural residence was associated with lower odds of prior abortion in the unadjusted analysis (OR=0.48, 95% CI 0.40 to 0.57), but this association was not significant in the adjusted model (OR=0.94, 95% CI 0.67 to 1.32). Marital status exhibited a shifting pattern: single or divorced women had lower odds of prior abortion in the unadjusted model (OR=0.55, 95% CI 0.49 to 0.60), but higher odds after adjustment (OR=1.39, 95% CI 1.12 to 1.73). With respect to fertility history, gravidity ≥2 was strongly associated with prior abortion in both models (Model 1: OR=40.90, 95% CI 27.77 to 60.24). In contrast, parity ≥1 was associated with significantly lower odds in the adjusted model (parity 1: OR=0.19; parity ≥2: OR=0.23). In the unadjusted analyses, use of less effective contraceptive methods (such as withdrawal and emergency contraception) was associated with higher odds of reporting a previous induced abortion. In the fully adjusted model, some of these associations remained statistically significant, indicating that contraceptive behaviours are closely linked to repeat abortion risk (online supplemental table 2).

Our retrospective data show that the overall proportion of young women reporting at least one previous abortion remained relatively stable, with minor year-to-year fluctuations. No clear, consistent increase or decrease was observed across the entire period; however, specific subgroups, particularly urban residents and women aged 20–25 years, showed a slightly upward trend in repeat abortions. These findings underscore persistent public health challenges and the need for targeted interventions to reduce unintended pregnancies among young women in Foshan.

Discussion

This study reveals a substantial prevalence of prior induced abortions among young women in Foshan, Guangdong Province, with 34.2% reporting at least one previous abortion.^{7 11 18} This indicates that abortion is often recurrent in this population, reflecting complex social, cultural and reproductive health determinants.¹⁹ Understanding these predictors is crucial for designing effective interventions aimed at reducing unintended pregnancies and improving reproductive outcomes among young women.

Women aged 20–25 years were more likely than adolescents to report previous abortions; however, this association attenuated and lost statistical significance after adjustment.^{20 21} This trend may reflect greater sexual activity, delayed childbearing or accumulated exposure to contraceptive failure over time. These findings emphasise the importance of age-specific sexual education and targeted counselling to promote consistent and correct contraceptive use in this age group. Women from ethnic minority groups exhibited higher abortion rates than Han women.²² This disparity likely stems from barriers to sexual education and limited access to contraceptive services within minority communities. Culturally tailored interventions are therefore essential to ensure equitable reproductive health education and service delivery across diverse populations.

In our crude comparisons, urban residents had higher proportions of prior abortion than rural residents, consistent with previous studies; however, this association was not statistically significant after adjustment for other factors.²³ While urban areas may offer better access to abortion services, this finding suggests that availability alone does not prevent unintended pregnancies. Effective contraceptive education and behavioural interventions remain critical in urban settings to address recurrent abortion risks. Although the proportion of married women increased with the number of previous abortions in descriptive analyses, the multivariable model indicated higher adjusted odds of prior abortion among single or divorced women compared with married women.²⁴ This may reflect sociocultural and familial pressures related to childbearing and family planning, highlighting the need for marital-status-specific counselling and reproductive health support. Interventions targeting married women should address both contraceptive uptake and social pressures influencing reproductive decisions.

In crude analyses, multiparity was more frequent among women with repeated abortions. However, in the fully adjusted model that simultaneously accounted for gravidity and other reproductive factors, higher parity (≥1) was associated with lower odds of reporting a prior induced abortion, suggesting complex interactions between pregnancy history and abortion patterns. Women with prior children may experience additional reproductive pressure, while those not using contraception before the index abortion indicate persistent gaps in knowledge and access. Postabortion uptake of long-acting reversible contraceptives (LARCs) such as IUDs demonstrates the potential of provider-led interventions to reduce recurrent abortions.²⁵,²⁹ Prior non-use of contraception increased the risk of repeated abortions, whereas women with previous abortion experiences were more likely to use condoms.²⁵ The preference for IUDs postabortion highlights the opportunity to reduce recurrence through targeted promotion of effective, long-term contraceptive methods.

The high prevalence of unintended pregnancies underscores ongoing deficiencies in sexual education and contraceptive access, particularly in rural and minority populations.^{30 31} Cultural perceptions framing abortion as corrective rather than preventive further contribute to recurrent abortions.^{21 32 33} These findings reinforce the need for comprehensive sexual education, accessible contraceptive services and postabortion counselling focused on LARCs.³⁴,³⁹ Public health strategies should address sociocultural pressures, integrate psychological support and tailor interventions according to age, marital status, residence, ethnicity and reproductive history.^{40 41} By emphasising targeted education and counselling, these strategies can help reduce repeated abortions and improve overall reproductive health outcomes.

This study highlights reproductive health challenges among young women (12–25 years old) in urban China, showing how sociodemographic factors, such as urbanisation, ethnicity and marital status, affect repeat abortions. It underscores disparities in access to contraceptives and sexual education and calls for targeted, culturally sensitive interventions beyond contraception. Without focused policies, repeated abortions and negative reproductive outcomes will continue. These findings provide a foundation for future research and strategies to reduce unintended pregnancies and improve reproductive health services for vulnerable populations. While Foshan Women and Children Hospital Affiliated to Guangdong Medical University is the city’s principal tertiary maternal and child centre and receives referrals from all districts, findings may not fully generalise to women who access private clinics or services outside the hospital network; comparisons with regional and national studies are therefore provided to contextualise our results.

Limitations

While our study offers valuable insights, it is important to acknowledge its limitations. First, our sample was drawn from a single tertiary maternal and child health hospital in Foshan, Guangdong, which may limit generalisability to other regions of China, as it may not fully capture the diversity of experiences across different geographic and cultural contexts. Future studies should strive to broaden the scope of the sample to include a more diverse population. Second, this research did not delve deeply into the cultural and environmental factors that could impact reproductive health behaviours and attitudes. Future research should investigate social pressures, prevailing gender norms and the accessibility of comprehensive sexuality education to obtain a more holistic understanding of the factors shaping young Chinese women’s abortion experiences. Such exploration will provide a more nuanced perspective on the mechanisms influencing reproductive health decisions and outcomes.

Conclusion

This study highlights the critical issue of recurrent abortions among young Chinese women within abortion-related care and underscores the necessity of enhancing sexual education and contraceptive services. By formulating and implementing more effective public health policies and interventions, we can significantly reduce abortion rates and improve women’s reproductive health.

Data availability statement

No data are available.