Skip to main content
NCBI home page
Medicine logoLink to Medicine
. 2020 May 8;99(19):e20251. doi: 10.1097/MD.0000000000020251

No association between abortion and risk of breast cancer among nulliparous women

Evidence from a meta-analysis

Huazhang Tong a, Yifan Wu b, Yin Yan c, Yonghai Dong d, Xihong Guan e, Yun Liu f, ZhiHui Lu g,∗
Editor: Martin S Staege
PMCID: PMC7220471  PMID: 32384520

Abstract

Background:

Various epidemiological studies have demonstrated the association between abortion and risk of breast cancer among nulliparous women; however, results remain inconclusive. This meta-analysis assessed the association based on previous studies.

Methods:

PubMed, EMBase, China National Knowledge Infrastructure, Chongqing VIP, and Wanfang databases were searched for relevant articles until February 2018. In this meta-analysis, fixed-effects models were used to estimate the combined effect size and the corresponding 95% confidence interval (CI). All statistical data were analyzed using STATA 12.0.

Results:

A total of 14 articles consisting of 6 cohort studies and 8 case-control studies were included in this review. All articles were of high quality, as determined based on the Newcastle Ottawa Scale assessment. The combined risk ratio (RR) indicated no significant association between abortion and breast cancer among nulliparous women (RR = 1.023, 95%CI = 0.938–1.117; Z = 0.51, P = .607). Subgroup analyses revealed no significant associations between risk of breast cancer and induced abortion or between risk of breast cancer and spontaneous abortion (SA) among nulliparous women (RR = 1.008, 95% CI = 0.909-1.118 and RR = 1.062, 95%CI = 0.902-1.250, respectively). Neither 1 nor >2 abortions increased the risk of breast cancer among nulliparous women. Sensitivity analysis showed that our results were reliable and stable.

Conclusion:

Current evidence based on epidemiological studies showed no association between abortion and risk of breast cancer among nulliparous women.

Keywords: induced abortion, meta-analysis, nulliparous, spontaneous abortion

1. Introduction

Breast cancer is the most commonly occurring cancer among women regardless of race or ethnicity[1] and is the sixth leading cause of cancer-related deaths.[2,3] World Cancer Research Fund International reported nearly 1.7 million new breast cancer cases in 2012.[4] Breast cancer in women represented 12% of all new cancer cases and 25% of all cancer cases.[4] In the United States, about 1/8 of the female population suffer from breast cancer during their lifetime, and 40,000 women die from breast cancer annually.[5] Breast cancer has become a major public health problem worldwide.

Previous studies reported that several factors could increase the risk of breast cancer, such as early menarche,[6] family history,[7] number of births,[8] obesity,[9] increasing age,[10] and genetic mutation.[11] Several epidemiological studies also revealed the relationship between abortion and breast cancer. Paoletti[12] performed a large-scale cohort study through a ten-year period, including 100,000 females aged 40 to 65 years; no relationship was found between breast cancer and induced abortion (IA), but a potential association between breast cancer and spontaneous abortion (SA) depending on menopausal status was suggested. In Denmark, a cohort study among women aged >50 years conducted through a twelve-year period showed that IA exerted no long-term effects on the risk of breast cancer (HR 0.95, 95% confidence intervals [CI] 0.83–1.09).[13] Another prospective cohort study[8] with 970,437 person-years among predominantly premenopausal population in Scandinavia found that breast cancer risk was not associated with either IA or SA.

In addition, several comprehensive meta-analysis described the relationship between breast cancer and abortion. Guo[14] pooled 15 prospective studies that revealed no sufficient evidence to support a positive association between either IA and breast cancer risk or SA and breast cancer risk. However, another meta-analysis involving Chinese females suggested that IA could significantly increase the risk of breast cancer.[15]

Thus far, analysis of the risk of abortion among nulliparous women with breast cancer is rarely conducted. In addition, previous studies have led to inconsistent results,[16–18] and no systematic comprehensive review has been performed to examine the relation between breast cancer and abortion. Thus, the current study is the first meta-analysis to determine the relation between breast cancer and abortion among nulliparous women.

2. Materials and methods

2.1. Literature search

The meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis recommendations. Two investigators independently searched several large databases, including PubMed Medline (1966 to Feb 2018), EMBase (1966 to Feb 2018), China National Knowledge Infrastructure (1994 to Feb 2018), Wanfang Data (1980 to Feb 2018), and Chongqing VIP (1989 to Feb 2018) for all relevant studies. The following terms were searched in the article titles, abstracts, and keywords: “induced abortion”/“spontaneous abortion”, “nulliparous”/“nonparous” and “breast cancer”/“breast carcinoma”/“breast tumor.” Relevant studies were also selected based on the reference list of identified articles.

2.2. Inclusion and exclusion criteria

Two reviewers (M Li and L Xu) independently selected the potential articles based on inclusion and exclusion criteria. In case of a divergence of opinion between the 2, a third reviewer was selected to evaluate whether the article in question was eligible. All selected studies were subjected to the following inclusion criteria:

  • (1)

    An observational design based on human population is adopted in the studies;

  • (2)

    The types of abortion include SA and IA;

  • (3)

    Sufficient information is provided to calculate the effect size;

  • (4)

    The articles are written in English or Chinese.

Any study that failed to meet these criteria was excluded.

2.3. Data extraction

Data on the study were independently extracted by 2 review authors using a standard form for each study. The extracted information included the first author, year of publication, country, study design, odds ratio (OR) or risk ratio (RR), and 95%CI. If required, information that had been omitted was retrieved via communication with the authors of the studies. During data collection, the third reviewer adjudicated on any divergence.

2.4. Quality assessment

In this review, Y Liu and X Guan independently evaluated the quality of the included studies by using the Newcastle Ottawa Scale (NOS) in accordance with the Cochrane Collaboration. The NOS checklist included 3 items: selection, comparability, and outcomes. The scale ranged from 0 to 9 stars with NOS score > = 7 stars considered as high quality, 4 to 6 stars as moderate quality, and < = 3 stars as low quality.

2.5. Ethical approval

Ethical approval was not required because all data were extracted from previously published articles.

2.6. Statistical analyses

The results of the meta-analysis were analyzed using Stata 12.0 (Stata Corporation, College Station, TX). Although different studies provided different measures of effects like RR or OR, as the incidence of breast cancer was low, these 2 effect measures should give similar estimates.[19,20] The pooled outcomes were assessed using the RR, along with the 95% CI. To assess the heterogeneity among the included studies, Cochran Q test and I2 were applied. I2>50% indicated heterogeneity, and the random-effects model was applied. Conversely, the fixed-effect model was applied. In this meta-analysis, the funnel plot and Egger test were performed to detect possible publication bias. To further explore the source of heterogeneity, subgroup analysis was conducted. In addition, sensitivity analysis was carried out by omitting each study and recalculating the pooled effect size for the remaining studies to validate the credibility of outcomes. When P ≤.05, the difference was considered statistically significant.

3. Results

3.1. Search results

A total of 656 articles were initially generated based on the search strategy. Removal of 107 duplicate articles ultimately resulted in 549 articles for further assessment. After screening the abstracts, 87 articles were assessed based on their full texts. Further screening led to the exclusion of 73 articles from the meta-analysis. Ultimately, 14 articles[8,12,16–18,21–29] met the rigid screening criteria (Fig. 1).

Figure 1.

Figure 1

Open in a new tab

Flow diagram of the screening process.

Table 1 lists basic information on the articles included in this review. Two types of abortion, IA and SA, were deduced from 7 articles in which case each of them was considered as 2 distinct articles. Thus, 21 studies were included in this meta-analysis.

Table 1.

Characteristics of the included studies.

3.1.

Open in a new tab

3.2. Characteristics of included studies

All selected articles were published in English. The field sites of all included articles covered over 6 countries, including United States,[8,16,18,21,22,24,25,27,28] Greece,[23] China,[17] Denmark,[26] France,[12] and Slovenia.[29] A cohort study design was adopted in 6 studies,[8,12,16,21,26,28] and a case-control study design was used in 8 studies.[17,18,22–25,27,29] All included studies reported on the association of IA with breast cancer risk among nulliparous women. Seven studies[8,12,16,23,24,27,29] reported on the relationship between SA and risk of breast cancer among nulliparous women.

3.3. Methodological quality

Using the NOS checklist, we assessed the methodological quality of the included studies and identified 4 moderate-quality articles[24,26,27,29] and 10 high-quality articles[8,12,16–18,21–23,25,28] (Table 1). Therefore, the included 14 articles were of good quality.

Table 2

Table 2.

Results of subgroup analysis.

3.3.

Open in a new tab

3.4. Association of breast cancer risk with abortion

No significant heterogeneity was found (Q = 15.29, P = .759 > .05; I2 = 0.1%). Thus, we used a fixed-effects model to pool the effect size (Fig. 2). The pooled effect size indicated no significant association between abortion and breast cancer among nulliparous women (pooled RR = 1.023, 95%CI = 0.938–1.117; Z = 0.51, P = .607).

Figure 2.

Figure 2

Open in a new tab

Forest plot of association between abortion and risk of breast cancer among nulliparous women.

3.5. Publication bias

The funnel plot, which was used to visually identify publication bias, was generally symmetric (Fig. 3). In addition, neither Begg test nor Egger test showed a significant risk of publication bias (z = 0.75, P = .450 > .05; t = 0.41, P = .686 > .05).

Figure 3.

Figure 3

Open in a new tab

Funnel plot of the included studies in this meta-analysis.

3.6. Subgroup analysis

A total of 14 studies reported on the risk of breast cancer among nulliparous women with IA, and the pooled effect size revealed no significant difference (pooled RR = 1.008, 95%CI: 0.909–1.118); 7 studies analyzed the risk of breast cancer among nulliparous women with SA, and the pooled RR was 1.062 (95%CI: 0.902–1.250).

On the basis of the cohort studies, nulliparous women with a history of abortion had a pooled risk of breast cancer equal to 1.041 (95%CI: 0.942–1.150). With regard to the case-control studies, the relationship between abortion and risk of breast cancer among nulliparous women was not determined, and the combined RR was 0.966 (95%CI: 0.804–1.159).

With regard to the influence of the number of abortions on the risk of breast cancer among nulliparous women, 10 studies reported on the relationship between 1-time abortion and risk of breast cancer, and the pooled RR was 1.081 (95%CI: 0.958-1.221). Moreover, 10 studies showed indicated the relationship between abortion occurring ≥2 times and risk of breast cancer, and the pooled RR was 0.942 (95%CI: 0.779–1.139). To ensure that similar results were obtained among nulliparous women with different types of abortion, we further conducted a subgroup analysis. Indeed, similar results were found for the women with only 1-time abortion or ≥2 times abortion, provided that they had neither IA nor SA

3.7. Sensitivity analysis

Sensitivity analysis was performed to assess the reliability and stability of the results. Fig. 4 showed that no significant change in the pooled effect size was observed when any 1 study was removed.

Figure 4.

Figure 4

Open in a new tab

Sensitivity analysis of the included studies in this meta-analysis.

4. Discussion

Nulliparous women are those who have experienced IA or SA before the midpoint of pregnancy but not those who have experienced pregnancy loss after 20 weeks. This occurrence has thus far been observed in a growing number of women worldwide, particularly in the developed countries. Considerable attention has been devoted toward the health of nulliparous women, for instance, breast cancer.[30–34] Deng[3] performed a meta-analysis to explore the association of IA with breast cancer, with no finding of significant difference between breast cancer and IA among nulliparous women. However, several deficiencies should be considered. This meta-analysis was based on case-control studies but did not include cohort studies. In addition, the study did not reveal the relationship between breast cancer and SA among nulliparous women. Thus, we performed this current more comprehensive meta-analysis to examine the relationship between abortion and breast cancer among nulliparous women on the basis of the case-control and cohort studies.

For this review, we analyzed 14 articles with 21 studies meeting the rigid screening criteria. All included articles employed proper methodology, which were assessed using the NOS checklists. The pooled OR of all studies was 1.023 [95%CI: 0.938–1.117], which suggested that no association was found between abortion and breast cancer among nulliparous women.

To further explore the different relationships between breast cancer and IA or breast cancer and SA, subgroup analysis was conducted. No significant difference was found in either IA or SA among nulliparous women. Our findings with regard to IA were consistent with the meta-analyses by Deng[3] and Guo[14] but not with the review by Huang.[15] The difference could be attributed to the variation in population of the different studies. Deng[3] and Guo[14] did not limit the countries where the participants came from, whereas Huang[15] included only Chinese females. This observation led us to conclude that race was the main reason.

In the United States, the number of new cases of females with breast cancer is 126.0 per 100,000 women annually. Compared with any chronic disease, breast cancer has a lower incidence. The OR in case-control studies should be close to the RR. Thus, we pooled the total RR based on the effect size (OR) of case-control studies and the effect size (RR) of cohort studies. Considering the influence of study design on the results, we also calculated the relationship between breast cancer and abortion for the 2 types of study design. On the basis of the 6 cohort studies, the pooled RR was 1.041 [95%CI: 0.942–1.150], which suggested that there was not sufficient proof to support the risk of breast cancer for abortion. To date, a cohort study[16] with the largest sample size (29,390 individuals), which used the California State Teachers Retirement System, found no statistically significant association between abortion and breast cancer risk among nulliparous women after Cox multivariable regression. This strong evidence proves that abortion and breast cancer risk are not associated. On the basis of the 8 case-control studies, the pooled effect size indicated that no relationship between abortion and breast cancer risk among nulliparous women was determined. The result was consistent with a previous systematic review.[3]

Several studies have suggested that 2 or more abortions can increase the risk of breast cancer among women.[27,35] However, other studies hold the opposite view. The subgroup analysis in our study revealed no association between breast cancer risk and the number of abortions among nulliparous women who had neither SA nor IA.

Although our meta-analysis demonstrated that no association was found between the risk of breast cancer and abortion among nulliparous women on the basis of the case-control studies and cohort studies, several flaws should be considered. First, some studies identified age as an important factor affecting the risk of breast cancer. However, owing to the variation in age in all included studies, the association of age with the risk for breast cancer was not analyzed in this review. Second, the sample size was small in all included studies, except for the study by Henderson.[16] Therefore, studies with a larger sample size need to be conducted in the future to further determine the relationship between abortion and the risk of breast cancer among nulliparous women.

Acknowledgments

Z Lu designed the study, H Tong, Y Wu, and Y Yan wrote the manuscript. Y Dong and X Guan performed the statistical analysis. Y Liu discussed the results.

Author contributions

Conceptualization: Zhihui Lu.

Formal analysis: Yonghai Dong.

Methodology: Yun Liu.

Project administration: Yun Liu, Zhihui Lu.

Resources: Xihong Guan.

Software: Yonghai Dong.

Supervision: Xihong Guan.

Writing – original draft: Huazhang Tong, Yifan Wu, Yin Yan, Yonghai Dong, Zhihui Lu.

Writing – review and editing: Huazhang Tong, Yifan Wu, Zhihui Lu.

Footnotes

Abbreviations: CI = confidence interval, IA = induced abortion, NOS = Newcastle Ottawa scale, OR = odds ratio, RR = risk ratio, SA = spontaneous abortion.

How to cite this article: Tong H, Wu Y, Yan Y, Dong Y, Guan X, Liu Y, Lu Z. No association between abortion and risk of breast cancer among nulliparous women: evidence from a meta-analysis. Medicine. 2020;99:19(e20251).

HT and YW contributed equally to this manuscript.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

The authors have no conflicts of interest to disclose.

References

  • [1].Centers for Disease Control and Prevention. Breast Cancer Statistics. Available at: https://www.cdc.gov/cancer/breast/statistics/. July 11, 2018. [Google Scholar]
  • [2].Fan L, Strasser-Weippl K, Li JJ, et al. Breast cancer in China. Lancet Oncol 2014;15:e279–89. [DOI] [PubMed] [Google Scholar]
  • [3].Deng Y, Xu H, Zeng X. Induced abortion and breast cancer: an updated meta-analysis. Medicine (Baltimore) 2018;97:e9613. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].World Cancer Research Fund International. Breast cancer statistics. Available at: https://www.wcrf.org/int/cancer-facts-figures/data-specific-cancers/breast-cancer-statistics. July 10, 2018. [Google Scholar]
  • [5].Mambou SJ, Maresova P, Krejcar O, et al. Breast cancer detection using infrared thermal imaging and a deep learning model. Sensors (Basel) 2018;18:e2799. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Leung AW, Mak J, Cheung PS, et al. Evidence for a programming effect of early menarche on the rise of breast cancer incidence in Hong Kong. Cancer Detect Prev 2008;32:156–61. [DOI] [PubMed] [Google Scholar]
  • [7].Printz C. Older women with a family history of breast cancer face increased risk of the disease. Cancer 2018;124:2673. [DOI] [PubMed] [Google Scholar]
  • [8].Michels KB, Xue F, Colditz GA, et al. Induced and spontaneous abortion and incidence of breast cancer among young women: a prospective cohort study. Arch Intern Med 2007;167:814–20. [DOI] [PubMed] [Google Scholar]
  • [9].Nattenmuller CJ, Kriegsmann M, Sookthai D, et al. Obesity as risk factor for subtypes of breast cancer: results from a prospective cohort study. BMC Cancer 2018;18:616. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Schoemaker MJ, Nichols HB, Wright LB, et al. Association of body mass index and age with subsequent breast cancer risk in premenopausal women. JAMA Oncol 2018;4:e181771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Lim J, Macluran M, Price M, et al. Short- and long-term impact of receiving genetic mutation results in women at increased risk for hereditary breast cancer. J Genet Couns 2004;13:115–33. [DOI] [PubMed] [Google Scholar]
  • [12].Paoletti X, Clavel-Chapelon FO. Induced and spontaneous abortion and breast cancer risk: results from the E3N cohort study. Int J Cancer 2003;106:270–6. [DOI] [PubMed] [Google Scholar]
  • [13].Brauner CM, Overvad K, Tjonneland A, et al. Induced abortion and breast cancer among parous women: a Danish cohort study. Acta Obstet Gynecol Scand 2013;92:700–5. [DOI] [PubMed] [Google Scholar]
  • [14].Guo J, Huang Y, Yang L, et al. Association between abortion and breast cancer: an updated systematic review and meta-analysis based on prospective studies. Cancer Causes Control 2015;26:811–9. [DOI] [PubMed] [Google Scholar]
  • [15].Huang Y, Zhang X, Li W, et al. A meta-analysis of the association between induced abortion and breast cancer risk among Chinese females. Cancer Causes Control 2014;25:227–36. [DOI] [PubMed] [Google Scholar]
  • [16].Henderson KD, Sullivan-Halley J, Reynolds P, et al. Incomplete pregnancy is not associated with breast cancer risk: the California Teachers Study. Contraception 2008;77:391–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Wu J, Li Y, Ren J, et al. Induced abortion and breast cancer: results from a population-based case control study in China. Asian Pac J Cancer Prev 2014;15:3635–40. [DOI] [PubMed] [Google Scholar]
  • [18].Ma H, Ursin G, Xu X, et al. Reproductive factors and the risk of triple-negative breast cancer in white women and African-American women: a pooled analysis. Breast Cancer Res 2017;19:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Zhang J, Yu KF. What's the relative risk?. A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690–1. [DOI] [PubMed] [Google Scholar]
  • [20].Alfaiate D, Clement S, Gomes D, et al. Chronic hepatitis D and hepatocellular carcinoma: a systematic review and meta-analysis of observational studies. J Hepatol 2020;S168–8278. [DOI] [PubMed] [Google Scholar]
  • [21].Palmer JR, Wise LA, Adams-Campbell LL, et al. A prospective study of induced abortion and breast cancer in African-American women. Cancer Causes Control 2004;15:105–11. [DOI] [PubMed] [Google Scholar]
  • [22].Newcomb PA, Mandelson MT. A record-based evaluation of induced abortion and breast cancer risk (United States). Cancer Causes Control 2000;11:777–81. [DOI] [PubMed] [Google Scholar]
  • [23].Lipworth L, Katsouyanni K, Ekbom A, et al. Abortion and the risk of breast cancer: a case-control study in Greece. Int J Cancer 1995;61:181–4. [DOI] [PubMed] [Google Scholar]
  • [24].Rosenberg L, Palmer JR, Kaufman DW, et al. Breast cancer in relation to the occurrence and time of induced and spontaneous abortion. Am J Epidemiol 1988;127:981–9. [DOI] [PubMed] [Google Scholar]
  • [25].Rookus MA, van Leeuwen FE. Induced abortion and risk for breast cancer: reporting (recall) bias in a Dutch case-control study. J Natl Cancer Inst 1996;88:1759–64. [DOI] [PubMed] [Google Scholar]
  • [26].Melbye M, Wohlfahrt J, Olsen JH, et al. Induced abortion and the risk of breast cancer. N Engl J Med 1997;336:81–5. [DOI] [PubMed] [Google Scholar]
  • [27].Mahue-Giangreco M, Ursin G, Sullivan-Halley J, et al. Induced abortion, miscarriage, and breast cancer risk of young women. Cancer Epidemiol Biomarkers Prev 2003;12:209–14. [PubMed] [Google Scholar]
  • [28].Daling JR, Brinton LA, Voigt LF, et al. Risk of breast cancer among white women following induced abortion. Am J Epidemiol 1996;144:373–80. [DOI] [PubMed] [Google Scholar]
  • [29].Robertson C, Van Den Donk M, Primic-Zakelj M, et al. The association between induced and spontaneous abortion and risk of breast cancer in Slovenian women aged 25-54. Breast 2001;10:291–8. [DOI] [PubMed] [Google Scholar]
  • [30].Lundberg FE, Iliadou AN, Rodriguez-Wallberg K, et al. The risk of breast and gynecological cancer in women with a diagnosis of infertility: a nationwide population-based study. Eur J Epidemiol 2019;34:499–507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Minami Y, Nishino Y, Kawai M, et al. Reproductive history and breast cancer survival: a prospective patient cohort study in Japan. Breast Cancer 2019;26:687–702. [DOI] [PubMed] [Google Scholar]
  • [32].Nguyen B, Venet D, Lambertini M, et al. Imprint of parity and age at first pregnancy on the genomic landscape of subsequent breast cancer. Breast Cancer Res 2019;21:25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].Da CD, Zelkowitz P, Nguyen TV, et al. Mental health help-seeking patterns and perceived barriers for care among nulliparous pregnant women. Arch Womens Ment Health 2018;21:757–64. [DOI] [PubMed] [Google Scholar]
  • [34].Ramezani S, Khosravi A, Motaghi Z, et al. The effect of cognitive-behavioural and solution-focused counselling on prevention of postpartum depression in nulliparous pregnant women. J Reprod Infant Psychol 2017;35:172–82. [DOI] [PubMed] [Google Scholar]
  • [35].Reeves GK, Kan SW, Key T, et al. Breast cancer risk in relation to abortion: results from the EPIC study. Int J Cancer 2006;119:1741–5. [DOI] [PubMed] [Google Scholar]

Articles from Medicine are provided here courtesy of Wolters Kluwer Health

RESOURCES