Cancer Causes & Control

, Volume 17, Issue 10, pp 1275–1280

Induced abortions and the risk of all cancers combined and site-specific cancers in Shanghai

Authors

    • Department of Kinesiology and Community HealthUniversity of Illinois at Urbana Champaign
  • Dao L. Gao
    • Department of EpidemiologyZhong Shan Hospital Cancer Center
  • Roberta M. Ray
    • Program in EpidemiologyFred Hutchinson Cancer Research
  • Michelle R. Rowland
    • Department of Kinesiology and Community HealthUniversity of Illinois at Urbana Champaign
  • Zakia C. Nelson
    • Program in EpidemiologyFred Hutchinson Cancer Research
  • Karen J. Wernli
    • Program in EpidemiologyFred Hutchinson Cancer Research
    • Department of EpidemiologyUniversity of Washington
  • Wenjin Li
    • Program in EpidemiologyFred Hutchinson Cancer Research
  • David B. Thomas
    • Program in EpidemiologyFred Hutchinson Cancer Research
    • Department of EpidemiologyUniversity of Washington
Original Paper

DOI: 10.1007/s10552-006-0067-x

Cite this article as:
Rosenblatt, K.A., Gao, D.L., Ray, R.M. et al. Cancer Causes Control (2006) 17: 1275. doi:10.1007/s10552-006-0067-x
  • 105 Views

Abstract

Although some previous case–control studies found an increased risk of breast cancer in women who had an induced abortion, the evidence from prospective studies suggests that induced abortions do not cause breast cancer. We have assessed risks of 12 types of cancer in women who have had induced abortions in a prospective study in China. Female textile workers (n = 267,400) completed a baseline questionnaire (1989–1991) that ascertained information on the major risk factors for breast cancer, contraceptive use, and induced abortions and were actively followed until July 2000. Cox Proportional Hazards analysis was used to calculate incidence rate ratios for specific types of cancer in women who ever had an induced abortion and by number of induced abortions. Women who had had an abortion were not at increased risk of cancer. There was a significant reduction in risk of uterine corpus cancer in women who had ever had an induced abortion, and a significant decreasing trend in risk with increasing number of induced abortions. No convincing associations with other cancers were observed. Women who have induced abortions after a live birth are not at increased risk of cancer and induced abortions may reduce risk of cancer of the corpus uteri.

Keywords

NeoplasmsAbortions, InducedCohort studies

Introduction

Although an increased risk of breast cancer in women who have had an induced abortion has been observed in some case–control studies [1], these observations may be biased results of more accurate and complete reporting of induced abortion by cases than controls. A pooled analysis of data from 13 prospective studies, in which information on abortion was ascertained before diagnosis, showed no association between breast cancer risk and induced abortion [1]. Few prospective studies of induced abortions have addressed possible risks of cancers other than those of the breast.

The purpose of this study is to update (by 5 years) a previous report [2] on the relationship between induced abortion and breast cancer in a cohort of Chinese women, and to explore possible relationships of induced abortions to other types of cancer.

Methods

Study design

Between October 1989 and October 1991, 267,400 current and retired employees of the Shanghai Textile Industry Bureau (STIB), who were born between 1925 and 1958, and who worked in one of 519 factories of the STIB, were enrolled in a randomized trial of breast self-examination (BSE) [3]. All women received their primary care from medical workers in clinics located in their factories. After receiving training from study workers, approximately 500 factory clinic medical workers orally administered a four-page optically-scanned questionnaire to all eligible women to ascertain information on induced abortions, the major risk factors for breast cancer, and contraceptive use. The question on induced abortions ascertained whether a woman had had 0, 1, 2, or 3+ induced abortions and the question on parity ascertained whether a woman had had 0, 1, 2, 3, 4, or 5+ live births. Comparisons between the baseline questionnaire (administered between 1989 and 1991) and a subsequent breast cancer nested case–control study questionnaire with a detailed pregnancy history (data collection stopped in 1999) showed a 97.5% agreement for ever/never having been pregnant, and, among women who had ever been pregnant, a 99.8% agreement of ever/never having a live birth and an 86.8% agreement for ever/never having an induced abortion (unpublished data, not shown).

Women in the cohort initially consulted medical workers in their factory when they became ill. If the clinic medical worker suspected cancer, the woman was referred to one of three hospitals operated by the STIB, or to other hospitals with contracts with individual factories. Surveillance for breast cancer was accomplished as part of the BSE trial through active reporting by the medical workers in each factory. For other cancers, case finding was through a Tumor and Death Registry operated by the STIB, which received annual reports from each factory of all cases of cancer and deaths that had occurred in the cohort members during the previous year. This information was supplemented by periodic reviews of the records of the Shanghai Cancer Registry, which meets the standards for inclusion of data in Cancer Incidence in Five Continents [4].

Diagnoses of cancers diagnosed prior to 31 December 1998 were confirmed by computer matching to the Shanghai Cancer Registry, and by review of medical records if the diagnosis differed from that in the STIB Tumor and Death Registry, or if the cancer could not be found in the Shanghai Cancer Registry. Hospital records were also reviewed for tumors coded as “uterine cancer, unspecified” and “intestinal cancer, unspecified.” The percentage of initially reported diagnoses that was corrected was 5.4% for all cancers combined, and varied from 0.3% (for breast cancer) to 13.5% (for gallbladder cancer). Because we were not able to validate the diagnoses from 1 January 1999 onward, the cancer site as reported by the STIB registry was used for cancers that were identified from that date to the end of follow-up on 31 July 2000. Cancers selected for specific analyses included those that might have a hormonal etiology (breast, colon, gallbladder, liver, ovary, thyroid, uterine cervix, and uterine corpus) and those that occurred commonly in this population (lung, pancreas, rectum, and stomach). Information on censored observations was obtained from personnel records (i.e., whether the woman left the STIB) and from the STIB Tumor and Death Registry (i.e., deaths). Only 7.5% of the subjects transferred out of the STIB and were lost to follow-up [3].

Statistical analyses

Cox Proportional Hazards models [5] were used to calculate cumulative risk ratios with the number of months of follow-up being the time scale variable. Observations with missing values for questions included in the analysis (less than 0.1% of any data item) were excluded from the analysis. Analyses were restricted to women who had ever been pregnant, since they would be the only ones at risk of having had an induced abortion. All models were controlled for the potentially confounding effects of age (using linear splines with knots at 5-year periods) and parity. Age at first live birth, duration of breast feeding, the use of various contraceptive methods (oral contraceptives, intrauterine devices, and monthly injectable contraceptives), spontaneous abortions, use of injections during pregnancy, tobacco use, and alcohol drinking were considered as potential confounders. Variables that altered the rate ratio (RR) of cancer in relation to ever having had an induced abortion by more than 5% were retained as confounding variables in the model in addition to age and parity.

This study was approved by the Institutional Review Boards of the Fred Hutchinson Cancer Research Center, the University of Illinois at Urbana-Champaign, and the Station for Prevention and Treatment of Cancer of the STIB and was approved by the Office for the Protection from Research Risks (OPRR) of the National Institutes of Health.

Results

Of women in the cohort who did not develop cancer, 53.2% had ever had an induced abortion, and 37.3%, 13.1%, and 2.7% had had one, two, and three or more induced abortions, respectively.

Women who had ever had an induced abortion were at slightly lower risk of developing any type of cancer than women who had never had an abortion (Table 1). There was a statistically significant reduction in risk of uterine corpus cancer in women who ever had an induced abortion. Risks of cancers of the colon and stomach were also lower in women with than without a prior induced abortion, but these observations were only of borderline statistical significance. No association was observed between ever having had an induced abortion and the development of breast, gallbladder, liver, lung, ovarian, pancreatic, rectal, thyroid, or cervical cancers.
Table 1

Rate ratio (RR) and 95% confidence intervals (CI) for cancer by site as associated with abortion

Cancer site

Abortion

No. of cancer cases

Person-years

RR (95% CI)

Alla,b

No

3,969

1,075,269

 

Yes

3,296

1,225,741

0.95 (0.90, 0.99)

Breasta,b,c

No

771

1,064,919

 

Yes

872

1,219,044

1.01 (0.92, 1.12)

Colona,b

No

372

1,072,259

 

Yes

256

1,225,834

0.84 (0.71, 1.00)

Gallbladdera,b

No

78

1,072,606

 

Yes

49

1,226,126

0.99 (0.69, 1.43)

Livera,b

No

265

1,073,511

 

Yes

183

1,226,746

0.95 (0.78, 1.15)

Lunga,b

No

457

1,074,148

 

Yes

329

1,227,142

1.02 (0.88, 1.18)

Ovarya,b,d

No

140

1,050,404

 

Yes

145

1,211,612

0.97 (0.76, 1.24)

Pancreasb,e

No

136

1,072,935

 

Yes

96

1,226,379

1.04 (0.79, 1.37)

Rectuma,b

No

192

1,072,258

 

Yes

163

1,225,800

1.02 (0.81, 1.27)

Stomacha,b

No

463

1,073,335

 

Yes

333

1,226,427

0.87 (0.75, 1.01)

Thyroida,b

No

67

1,072,024

 

Yes

86

1,225,491

1.09 (0.78, 1.51)

Uterine cervixa,b,f,g

No

36

1,035,184

 

Yes

23

1,199,431

0.93 (0.54, 1.62)

Uterine Corpusa,b,f

No

148

1,034,917

 

Yes

97

1,199,169

0.70 (0.53, 0.91)

a Restricted to women with one or more pregnancies

b Adjusted for parity (0, 1, 2, 3, 4, and 5+) and age with splines with knots at 5-year intervals

c Restricted to women with no prior mastectomy

d Restricted to women with no prior bilateral oophorectomy

e Restricted to women with one or more live births

f Restricted to women with no prior hysterectomy

g Adjusted by oral contraceptive use

There was a significant decreasing trend (p = 0.007) in risk of uterine corpus cancer with increasing numbers of induced abortions (Table 2). There was no trend in risk with number of induced abortions and all cancers combined or for any of the other cancers considered.
Table 2

Rate ratio (RR) and 95% confidence intervals (CI) for cancer by site as associated with number of previous abortions

Cancer site

Number of abortions

Number of cancer cases

Person-years

RR (95% CI)

Alla,b

0

3,969

1,075,269

1.00 (referent)

1

2,290

857,317

0.94 (0.89, 0.99)

2

823

304,595

0.96 (0.89, 1.03)

3+

183

63,829

0.96 (0.83, 1.12)

   

p = 0.09*

Breasta,b,c

0

771

1,064,919

1.00 (referent)

1

630

852,787

1.06 (0.95, 1.18)

2

192

302,921

0.87 (0.74, 1.02)

3+

50

63,336

1.06 (0.80, 1.42)

   

p = 0.52

Colona,b

0

372

1,072,259

1.00 (referent)

1

172

857,633

0.82 (0.68, 0.98)

2

66

304,459

0.89 (0.68, 1.16)

3+

18

63,742

1.01 (0.62, 1.65)

   

p = 0.21

Gallbladdera,b

0

78

1,072,606

1.00 (referent)

1

36

857,775

1.01 (0.68, 1.50)

2+

13

367,394

0.94 (0.52, 1.70)

   

p = 0.89

Livera,b

0

265

1,073,511

1.00 (referent)

1

124

858,193

0.91 (0.73, 1.13)

2

49

304,709

1.06 (0.77, 1.45)

3+

10

63,843

0.97 (0.51, 1.83)

   

p = 0.89

Lunga,b

0

457

1,074,148

1.00 (referent)

1

239

858,505

1.04 (0.89, 1.22)

2

80

304,830

1.07 (0.84, 1.36)

3+

10

63,807

0.58 (0.31, 1.09)

   

p = 0.73

Ovarya,b,d

0

140

1,050,404

1.00 (referent)

1

102

847,177

0.99 (0.76, 1.29)

2+

43

364,434

0.93 (0.65, 1.32)

   

p = 0.73

Pancreasb,e

0

214

1,072,935

1.00 (referent)

1

122

857,970

1.06 (0.79, 1.42)

2

40

304,592

0.87 (0.53, 1.42)

3+

10

63,817

1.55 (0.75, 3.18)

   

p = 0.69

Rectuma,b

0

192

1,072,258

1.00 (referent)

1

116

857,592

1.04 (0.82, 1.32)

2

40

304,446

1.03 (0.72, 1.96)

3+

7

63,761

0.68 (0.30, 1.54)

   

p = 0.81

Stomacha,b

0

463

1,073,335

1.00 (referent)

1

221

857,950

0.82 (0.69, 0.97)

2

91

304,662

0.99 (0.78, 1.25)

3+

21

63,815

1.04 (0.67, 1.61)

   

p = 0.43

Thyroida,b

0

67

1,072,024

1.00 (referent)

1

58

857,325

1.04 (0.72, 1.50)

2+

28

367,209

1.20 (0.76, 1.89)

   

p = 0.47

Uterine cervixa,b,f,g

0

36

1,035,184

1.00 (referent)

1

18

838,648

0.99 (0.55, 1.79)

2+

5

360,782

0.75 (0.29, 1.97)

   

p = 0.65

Uterine Corpusa,b,f

0

148

1,034,917

1.00 (referent)

1

69

838,446

0.72 (0.54, 0.97)

2+

28

360,723

0.63 (0.41, 0.96)

   

p = 0.007

Restricted to women with one or more pregnancies

b Adjusted for parity (0, 1, 2, 3, 4, 5+) and age with splines with knots at 5 year intervals

c Restricted to women with no prior mastectomy

d Restricted to women with no prior bilateral oophorectomy

e Restricted to women with one or more live births

f Restricted to women with no prior hysterectomy

g Adjusted by oral contraceptive use (ever, never)

* p-values are for the test of trend

Discussion

In this study, the risk of all cancers combined was no greater in women who had had an induced abortion than in women who had not. In China, abortions are generally performed after a first live birth [6]. Although such abortions are common in China, they are unlikely to have an appreciable impact on the occurrence of cancer in that country. A limitation of this study is our inability to assess the risk of cancer in relation to abortions performed before a live birth.

The only statistically significant relationships that we observed were a reduction in risk of uterine corpus cancer (which can be assumed to be largely endometrial cancer) in women who ever had an induced abortion, and a significant decreasing trend in risk of this cancer with increasing numbers of induced abortions. These results are in contrast to the only other cohort study (Iowa Women’s Health Study) that investigated this relationship [7], which found a significant increase in risk of endometrial cancer with ever having had an induced abortion (RR = 2.50, 95% CI = 1.10–5.66). That association is based only on six exposed cases, and less than 2% of the unaffected women reported an induced abortion.

Most case–control studies of endometrial cancer and induced abortions have shown a reduction in risk with ever having had an induced abortion [8, 9] or a reduction in risk with increasing numbers of induced abortions [1012], including two studies conducted in Shanghai [9, 10], however, studies in Greece [13] and the United States [14] did not observe an association with induced abortions. Since induced abortions have a social stigma attached to them in the United States and Greece (but not in China), this could have led to underreporting of induced abortions in the United States cohort study [7] and reporting bias (cases more accurately reporting induced abortions than controls) in the case–control studies in the United States [14] and Greece [13]. These potential biases would tend to obscure true reductions in relative risk. The preponderance of the evidence, including the results of our study, suggests that induced abortions reduce the risk of endometrial cancer.

Abortions tend to be performed after childbearing and at a later age in China [15] than in the United States, where they are usually performed before age 30 [16]. Because the incidence of endometrial cancer increases with age, having had an induced abortion at a later age could hypothetically be more likely result in removal of pre-malignant endometrial tissue than an induced abortion at a younger age. Also, induced abortions in China are most frequently performed using dilation and curettage (D and C) [17], but in the United States most abortions are performed by suction, which removes less endometrial tissue than a D and C. It would be useful for future studies to investigate the possible association between different types of induced abortions and endometrial cancer.

We did not observe an increased risk of breast cancer in women who ever had an induced abortion, and we did not see an increasing trend in risk with an increasing number of induced abortions. These results are similar to those from the large pooled analysis of data from 13 cohort studies [1], which included earlier data from this study. All positive relationships between induced abortions and breast cancer have come from case–control studies, which are prone to reporting bias, and it can be concluded with reasonable confidence that induced abortions do not cause breast cancer. This conclusion is in accordance with results of a National Cancer Institute sponsored workshop conducted on 24–25 February 2003 [18].

We have previously reported a reduction in risk of colon cancer, of borderline statistical significance, in women in this study cohort who ever had an induced abortion, but no trend in risk with number of induced abortions was found [19]. Two case–control studies in Italy [20, 21] also found significant reductions in risk of colon cancer in women who had induced abortions, and this possible relationship warrants additional investigation.

Previous studies have found no evidence for abortions causing an increase in risks of cancers of the gallbladder [20], liver [20, 22], ovary [20, 2336], pancreas [20, 37], rectum [20, 21], stomach [20, 38, 39], or thyroid [40]; and no such evidence was observed in this study. We also found no association between induced abortion and lung cancer. To our knowledge, results for lung cancer in relation to induced abortion have not been reported previously. One prior case–control study in Italy [20] found an increased risk of cervical cancer in women who had had abortions. We did not confirm this observation, which may be biased due to uncontrolled confounding by sexual behavior, or more complete reporting of abortions by cases than controls in that study.

A limitation of the present study is the inability to control for the potentially confounding effects of risk factors associated with some of the site-specific cancers. Most of the risk factors ascertained in the baseline questionnaire were breast cancer risk factors. However, many of these factors are also risk factors for ovarian, endometrial, and some other cancers. We also had information on alcohol and tobacco use, although the prevalence of tobacco use (2.8% ever use) and alcohol drinking (4.2% once a week or more) was rather infrequent and neither one was associated with ever having an induced abortion, making them unlikely to be confounders. Although we did not ascertain information on other risk factors, such as dietary factors and family history of cancers other than cancers of the breast, there is no reason to suspect that these or other missing risk factors would be related to having had an induced abortion in China, where abortions are readily available, encouraged to limit family size to one child, and socially acceptable. It is therefore unlikely that confounding by unmeasured risk factors obscured true relationships between induced abortions and the risk of any specific cancer.

The frequency of induced abortions in our study cohort (53.2%) is similar to controls in case–control studies performed in Shanghai: an endometrial cancer study published in 1991 (56.6%) [10], a breast cancer study published in 2001 (66.4%) [41], and an endometrial cancer study published in 2003 (63.1%) [9]. This observation, plus the high level of concordance between induced abortions reported on our baseline questionnaire and a case–control study nested in the cohort, suggest that reporting of induced abortions by women in our study was sufficiently complete and accurate so as not to have appreciably reduced relative risk estimates and obscured true relationships to cancer risk.

Induced abortions, as they have generally been performed in China, do not increase a woman’s risk of cancer. Induced abortions may protect against endometrial cancer.

Acknowledgments

This work was supported by National Cancer Institute grants R03-CA80637, R01-CA46823, and R01-CA80180. Earlier findings from this study were presented at the November 2002 American Public Health Association meeting.

Copyright information

© Springer Science+Business Media B.V. 2006