Incidence of ovarian cancer of grand multiparous women—A population-based study in Finland

Gynecologic Oncology 103 (2006) 207 – 211 www.elsevier.com/locate/ygyno

Incidence of ovarian cancer of grand multiparous women—A population-based study in Finland Marianne Hinkula a,⁎, Eero Pukkala b , Pentti Kyyrönen b , Antti Kauppila a a

b

Department of Obstetric and Gynaecology, University of Oulu, PL 24, FIN-90029 OYS, Finland Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research, Helsinki, Finland Received 24 November 2005 Available online 3 April 2006

Abstract Objectives. Parity is known to induce protective effects on ovarian cancer. This study aimed to evaluate how far upon births the protection reaches, the effect of age at first birth, the interval between births in the whole population and the length of time from the first to the last birth and from the last birth to cancer among postmenopausal women. Method. The population-based cohort consisted of 87,929 grand multiparous (GM) women, i.e. women with at least 5 deliveries. Standardised incidence ratios (SIRs) were calculated by dividing the number of observed cancer cases by the expected number based on the national incidence rates, both extracted from the population-based Finnish Cancer Registry. Conditional logistic regression for the case–control design nested in the GM cohort was used to estimate proportional hazards by different factors. Results. The SIR for ovarian cancer among GM women was low (418 cases; SIR 0.64, 95% confidence interval 0.58–0.69). Further births over five did not give additional protection. The relative risk did not vary significantly by age at first birth or interval between the births in any histological subtype. Conclusion. The risk of ovarian cancer was low in all GM women no matter how many children and at which ages they had delivered or contracted cancer. © 2006 Elsevier Inc. All rights reserved. Keywords: Ovarian cancer; Parity

Introduction

Material and methods

Previous studies on the role of reproductive risk factors in the aetiology of ovarian cancer have shown that increasing parity decreases the risk of ovarian cancer, especially epithelial ovarian cancer [1–4]. However, it is not exactly known how many pregnancies would give maximal protection. We designed this study on grand multiparous (GM) women (at least five deliveries) to explore this problem. The roles of age at first birth, the interval between the births and the length of the time between the first and last births and the time between the last birth and diagnosis of cancer were also investigated.

Population-based registers The study cohort was comprised of all 87,929 GM women registered in the computerised files of the Finnish Population Register from the year 1974 to 2002. Ovarian cancer cases in the cohort were identified from the files of the national, population-based Finnish Cancer Registry with personal identifiers (unique identifiers given to all residents of Finland and used in all main personal registers in Finland). Because the historical standard Cancer Registry classification of histological types did not separate the cases into modern categories needed in our analyses, all 452 ovarian cancer cases by the end of year 2003 were reclassified using free-text data on cancer topography and morphology stored in the database of the Finnish Cancer Registry.

Statistical methods ⁎ Corresponding author. Fax: +358 8 315 4310. E-mail address: [email protected] (M. Hinkula). 0090-8258/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2006.02.025

The standardised incidence ratios (SIR) were calculated to compare the incidence of ovarian cancer among the GM women to that of the entire Finnish female population. The follow-up started from the birth of the fifth child or from

208

M. Hinkula et al. / Gynecologic Oncology 103 (2006) 207–211

Table 1 Relative risks (RR) with 95% confidence interval (95% CI) in nested case– control study of ovarian cancer cases (n = number of cases) among GM women Variables

n

RR

95% CI

Parity 5 6 7 8+

230 127 38 57

1 1.02 0.86 1.08

Ref 0.81–1.28 0.60–1.24 0.78–1.50

Each histology category was analysed separately. Variables included in the analyses were parity (5, 6, 7, 8+ children), age at first birth (<20, 20–24, 25–29, 30+ years) and birth interval (the average interval between first and fifth deliveries; <2.0, 2.0–2.9, ≥3.0 years). For women 50 years or older – who were assumed to have already delivered their last child – we also analysed the significance of the birth period (the time between first and last birth; <10, 10– 14.9, 15–19.9, 20+ years) and the time between last birth and cancer (<10, 10– 19.9, 20–29.9, 30+ years).

Results Age at first birth (years) <20 69 20–24 225 25–29 128 ≥30 30

1.00 0.90 1.03 1.02

Ref 0.67–1.21 0.72–1.46 0.60–1.72

Birth interval (years) a <2.0 2.0–2.9 ≥3.0

1.00 1.05 0.93

Ref 0.83–1.34 0.69–1.25

a

21 328 103

Average interval between the first five deliveries.

January 1974, whichever was later, and ended at emigration, death or on 31 December 2002, whichever was earliest. The number of person years was 2.0 million. The numbers of observed cases of ovarian cancer and person–years at risk were counted by 5-year age groups and by calendar periods. The expected numbers of cases were calculated by multiplying the number of person–years in each stratum by the corresponding incidence rate in Finland. The SIR was calculated by dividing the number of observed cases by the number of expected cases. The 95% confidence intervals (95% CI) for the SIRs were based on the assumption that the number of observed cases presents in the Poisson distribution. Conditional logistic regression (SAS, 2000) for case–control design nested within the cohort was used to estimate proportional hazards by different factors in this cohort. These are referred to as relative risks (RR) in the tables. Controls were matched with tolerance of ±1 year on the date and age at entry in the cohort. For each ovarian cancer case, 50 controls were randomly selected among the cohort members who were at risk for ovarian cancer at the time of the cancer onset of the case and fulfilled the matching criteria. The follow-up of this cohort continued until the end of the year 2003.

The number of ovarian cancers cases diagnosed in the GM cohort by the end of year 2002 was significantly smaller than expected (SIR 0.64, 95% CI 0.58–0.69; n = 418). The incidence was significantly decreased in each epithelial subtype (based on the rough histopathological classification of the Finnish Cancer Registry), the SIR for serous tumours was 0.69 (95% CI 0.59– 0.78; n = 184), endometrioid tumours 0.43 (95% CI 0.34–0.53; n = 80) and mucinous tumours 0.82 (95% CI 0.62–1.05; n = 59). The SIR for granulous-stromal cell tumours was 1.03 (95% CI 0.52–1.85; n = 11) and for borderline tumours 0.74 (95% CI 0.58–0.92; n = 76). During the follow-up, the SIRs were equal in premenopausal (SIR 0.63, 95% CI 0.46–0.82) and postmenopausal women (SIR 0.65, 95% CI 0.58–0.72). After histopathological reclassification, epithelial ovarian cancers represented 78% (n = 352), granulous-stromal cell tumours 8% (n = 36), other ovarian tumours 4% (n = 20) and borderline tumours 10% (n = 44). Thirty-four percent (n = 155) of all ovarian cancer cases were serous, 8% (n = 35) endometrioid, 13% (n = 59) mucinous, 2% (n = 9) clear cell tumours and 21% (n = 94) unclassified epithelial ovarian tumours. Twelve percent of ovarian cancer patients were less than 50 years old. In conditional logistic regression analysis with case–control design, the increase in parity from 5 to 8+ did not change the protective effect of parity in ovarian cancer cases (Table 1) or in any of the subtypes of ovarian tumours (Tables 2 and 3). In the

Table 2 Relative risks (RR) with 95% confidence interval (95% CI) in nested case–control study of epithelial ovarian cancer cases (n = number of cases) among GM women, clear cell tumour (n = 9) and non-epithelial tumours are excluded Epithelial ovarian cancer (n = 352)

Serous tumour (n = 155)

Endometrioid tumour (n = 35)

Mucinous tumour (n = 59)

Unclassified epithelial tumour (n = 94)

Variable

n

RR

95% CI

n

RR

95% CI

n

RR

95% CI

n

RR

95% CI

n

RR

95% CI

Parity 5 6 7 8+

179 102 25 46

1.00 1.04 0.73 1.11

Ref 0.80–1.35 0.47–1.13 0.77–1.60

82 44 12 17

1.00 0.90 0.71 0.90

Ref 0.61–1.32 0.38–1.32 0.50–1.59

21 8 1 5

1.00 0.73 0.27 0.89

Ref 0.31–1.74 0.03–2.10 0.28–2.79

28 18 4 9

1.00 1.21 0.68 1.29

Ref 0.65–2.26 0.23–2.03 0.53–3.11

41 31 8 14

1.00 1.49 1.24 1.72

Ref 0.91–2.45 0.59–2.75 0.85–3.45

Age at first birth (years) <20 51 1.00 20–24 176 0.96 25–29 101 1.10 ≥30 24 1.14

Ref 0.68–1.34 0.73–1.65 0.63–2.06

32 71 45 7

1.00 0.67 0.92 0.66

Ref 0.43–1.07 0.52–1.63 0.25–1.74

4 18 8 5

1.00 1.13 1.11 4.14

Ref 0.36–3.58 0.28–4.41 0.81–21.0

8 35 13 3

1.00 1.23 0.83 0.75

Ref 0.55–2.77 0.30–2.30 0.16–3.57

7 47 31 9

1.00 1.40 1.60 1.84

Ref 0.61–3.25 0.63–4.04 0.57–5.98

Birth interval (years) a <2.0 97 1.00 2.0–2.9 153 1.06 ≥3.0 102 0.90

Ref 0.81–1.40 0.64–1.26

40 77 38

1.00 1.32 0.80

Ref 0.88–1.98 0.47–1.35

11 13 11

1.00 0.80 0.78

Ref 0.33–1.94 0.27–2.21

17 25 17

1.00 0.95 0.71

Ref 0.48–1.85 0.31–1.64

26 35 33

1.00 0.91 1.30

Ref 0.53–1.57 0.69–2.44

a

Average interval between the first five deliveries.

M. Hinkula et al. / Gynecologic Oncology 103 (2006) 207–211 Table 3 Relative risks (RR) with 95% confidence interval (95% CI) in nested case– control study of granulous-stromal cell and borderline ovarian tumours (n = number of cases) among GM women Variable

Granulous-stromal cell tumour (n = 36)

Borderline tumour (n = 44)

n

RR

95% CI

n

RR

95% CI

16 10 3 7

1.00 1.99 0.99 1.99

Ref 0.52–2.72 0.27–3.64 0.71–5.60

24 11 6 3

1.00 0.95 1.23 0.55

Ref 0.45–2.00 0.47–3.23 0.14–2.08

Age at first birth (years) <20 6 1.00 20–24 21 1.03 25–29 8 0.86 ≥30 1 0.47

Ref 0.39–2.71 0.25–2.97 0.05–4.71

12 21 9 2

1.00 0.55 0.42 0.35

Ref 0.25–1.22 0.14–1.27 0.06–2.12

Birth interval (years) a <2.0 10 1.00 2.0–2.9 15 1.07 ≥3.0 11 0.83

Ref 0.45–2.52 0.28–2.47

8 21 15

1.00 1.64 1.58

Ref 0.69–3.91 0.58–4.32

Parity 5 6 7 8+

a

Average interval between first five deliveries.

group of unclassified epithelial tumours, of whom most were anaplastic, the increasing parity seemed to increase the risk of cancer, but the trend was not significant (Table 2). The risk of endometrioid tumours and unclassified epithelial ovarian tumours tended slightly to increase with increasing age at first birth while the risk of serous tumours was highest in women with younger than 20 years at first birth (Table 2). The birth interval did not present any significant trend in any analysis (Tables 1–3). A long period from the first to last birth and a long interval between the last birth and diagnosis of cancer tended to increase the risk of epithelial cancer risk in GM women older than 50 years (Table 4). Discussion It is well known, that multiparity reduces the risk of ovarian cancer, especially for epithelial ovarian cancer [2,4,5]. Our results support this observation. The incidence of ovarian cancer was 36% smaller than among average Finnish women with about 2 children. In earlier case–control studies, the risk reduction of ovarian cancer was about 50% in parous women compared to nulliparous women and each consequent pregnancy decreased the risk by about 15% [5,6]. Using this approximation, the SIR among our GM women would be about 0.50, if the reference would be nulliparous women. The increase in parity from five births upwards does not provide any additional protection from that obtained by 5 births. This finding is in line with the results from a Norwegian study, in which the risk reduction after third delivery was 44% (SIR 0.56, 95% CI 0.48–0.67) and after the fifth delivery 46% (SIR 0.54, 95% CI 0.39–0.76) compared to uniparous women [7]. Most GM women belong to the Laestadian religious movement within the Lutheran church. High parity is common

209

within this movement, as any kind of contraception is forbidden, but the living habits do not differ markedly from those of other Finns—smoking is permitted, but alcohol consumption is rare. The use of oral contraceptives significantly reduces the risk of ovarian cancer [8,9]. Because their usage is common in the reference population, our SIR estimates may be a bit too high. Women with a family history of ovarian cancer have a substantially higher risk of developing ovarian cancer [10–14]. The frequency of such cancers, representing less than 10% of all ovarian cancers [10–14], might be smaller among GM women than in reference group, which might slightly diminish our SIR estimates. We did not have any data about hysterectomies and sterilisations in our population, which have conveyed some protection for ovarian cancer [3,15]. There might be no any bias, because the frequencies of hysterectomies and sterilisations among GM women may not differ much from those in our reference population. Ovarian cancer belongs, together with breast and endometrial cancers, to the category of hormone dependent cancers [16]. They, however, differ from each other in their relationships to different reproductive factors [17–22]. Increasing parity from the 5th birth onwards had no effect on ovarian cancer risk. In the same population, it significantly improved protection against breast [21] and endometrial cancer [22]. Our present study suggests that the age at first birth as a risk factor is without any significance in ovarian cancer. Other previous studies on the importance of age at first [4,6,28,29] or last birth [7,29–32] as a risk factor of ovarian cancer have also yielded conflicting results. On the other hand, early age at first birth was protective Table 4 Relative risks (RR) with 95% confidence interval (95% CI) in nested case– control study of epithelial ovarian cancer cases (n = number of cases) (n = 320, total number of cases) among GM women aged 50+ years at cancer onset Variable

n

RR

95% CI

Parity 5 6 7 8+

161 96 19 44

1.00 1.31 0.66 1.31

Ref 0.87–1.47 0.40–1.07 0.90–1.89

Age at first birth (years) <20 42 20–24 159 25–29 96 ≥30 23

1.00 0.94 1.05 1.04

Ref 0.64–1.37 0.65–1.68 0.52–2.06

Birth period (years) a <10 10–14.9 15–19.9 ≥20

1.00 0.78 0.87 0.67

Ref 0.57–1.06 0.58–1.30 0.34–1.32

0.68 1.37 1.26 1.00

0.08–5.93 0.68–2.77 0.83–1.90 Ref

81 128 93 18

Years between last birth and cancer <10 2 10–19.9 54 20–29.9 131 ≥30 133 a

Average interval between the first and last deliveries.

210

M. Hinkula et al. / Gynecologic Oncology 103 (2006) 207–211

in breast cancer [21,23,24], whereas in endometrial cancer it increased the risk [22,25–27]. Population-based studies, comprising all birth categories, have demonstrated that in Norway high parity was protective against different subtypes of epithelial ovarian cancer, except mucinous ones and non-epithelial cancers [6] and in Sweden against borderline and invasive ovarian cancers [2]. Our study could not find any relationship of parity or other study variables with the RR of any histological subtype. Incessant ovulations have been regarded as a primary cause of epithelial ovarian cancer [33]. Because ovulations during the 20–29 years age carry the greatest risk [34], this theory well fits to the markedly lowered ovarian cancer risk in GM women. In our population, 65% of GM women delivered the first child at an age younger than 25 years, whereas the average age of Finnish women undergoing first birth has recently been 28 years [35]. Absence of ovulations owing to pregnancies after 29 years of age is associated with weakened efficacy against ovarian cancer [34], which partly explains why the risk remains similar after 3– 5 births as seen here and previously [7]. However, pregnancy at older ages reduces ovarian cancer risk [7,32]. Pregnancy is associated with clearance of a fraction of the genetically modified (premalignant) cells from the ovaries [36]. Apoptosis of such epithelial cells in the ovary may be under the control of fetoplacental hormones [32], possibly progesterone [37]. High ovulatory levels of FSH and LH might stimulate malignant transformation of ovarian epithelial cells [38]. Maternal immunoassayable LH and FSH levels are virtually undetectable throughout pregnancy [39]. Thus, pregnancies may also counteract against ovarian cancer development by saving ovarian epithelial cells from the action of pituitary gonadotropins. Our study confirms that pregnancies diminish the ovarian cancer risk, whereas age at first/last birth and the interval between births remained insignificant in GM women. These findings are different from the respective findings in breast and endometrial cancers and demonstrate that ovarian cancer behaves in a unique way in its hormonal relationships. Acknowledgment This study has been supported by Finnish Cultural Foundation. References [1] Whittemore AS. Characteristics relating to ovarian cancer risk: implications for prevention and detection. Gynecol Oncol 1994:55. [2] Adami HO, Hsieh CC, Lambe M, et al. Parity, age at first childbirth, and risk of ovarian cancer. Lancet 1994;344:1250–4. [3] Riman T, Persson I, Nilsson S. Hormonal aspects of epithelial ovarian cancer: review of epidemiological evidence. Clin Endocrinol 1998;49: 695–707. [4] Riman T, Nilsson S, Persson IR. Review of epidemiological evidence for reproductive and hormonal factors in relation to the risk of epithelial ovarian malignancies. Acta Obstet Gynecol Scand 2004;83: 783–95. [5] Negri E, Franceschi S, Tzonou A, et al. Pooled analysis of 3 European case–control studies: I. Reproductive factors and risk of epithelial ovarian cancer. Int J Cancer 1991;49:50–6.

[6] Kvale G, Heuch I, Nilssen S, Beral V. Reproductive factors and risk of ovarian cancer: a prospective study. Int J Cancer 1988;42:246–51. [7] Albrektsen G, Heuch I, Kvale G. Reproductive factors and incidence of epithelial ovarian cancer: a Norwegian prospective study. Cancer Causes Control 1996;7:421–7. [8] La Vecchia C, Franceschi S, Decarli A. Oral contraceptive use and the risk of epithelial ovarian cancer. Br J Cancer 1984;50:31–4. [9] Risch HA, Marrett LD, Howe GR. Parity, contraception, infertility, and the risk of epithelial ovarian cancer. Am J Epidemiol 1994;140:585–97. [10] Kerlikowske K, Brown JS, Grady DG. Should women with familial ovarian cancer undergo prophylactic oophorectomy? Obstet Gynecol 1992;80:700–7. [11] Amos CI, Struewing JP. Genetic epidemiology of epithelial ovarian cancer. Cancer 1993;71:566–72. [12] Sarantaus L, Auranen A, Nevanlinna H. BRCA1 and BRCA2 mutations among Finnish ovarian carcinoma families. Int J Oncol 2001;18:831–5. [13] Eerola H, Pukkala E, Pyrhonen S, Blomqvist C, Sankila R, Nevanlinna H. Risk of cancer in BRCA1 and BRCA2 mutation-positive and -negative breast cancer families (Finland). Cancer Causes Control 2001;12:739–46. [14] Auranen A, Pukkala E, Makinen J, Sankila R, Grenman S, Salmi T. Cancer incidence in the first-degree relatives of ovarian cancer patients. Br J Cancer 1996;74:280–4. [15] Green A, Purdie D, Bain C, et al. Tubal sterilisation, hysterectomy and decreased risk of ovarian cancer. Survey of Women's Health Study Group. Int J Cancer 1997;71:948–51. [16] Key TJ. Hormones and cancer in humans. Mutat Res 1995;333:59–67. [17] Elwood JM, Cole P, Rothman KJ, Kaplan SD. Epidemiology of endometrial cancer. J Nat Cancer Inst 1977;59:1055–60. [18] Franceschi S. Reproductive factors and cancers of the breast, ovary and endometrium. Eur J Cancer Clin Oncol 1989;25:1933–43. [19] Sutton G. Hormonal aspects of endometrial cancer. Curr Opin Obstet Gynecol 1990;2:69–73. [20] Helzlsouer KJ, Alberg AJ, Gordon GB, et al. Serum gonadotropins and steroid hormones and the development of ovarian cancer. JAMA 1995;274:1926–30. [21] Hinkula M, Pukkala E, Kyyrönen O, Kauppila A. Grand multiparity and risk of breast cancer; population-based study in Finland. Cancer Causes Control 2001;12:491–500. [22] Hinkula M, Pukkala E, Kyyrönen P, Kauppila A. Grand multiparity and incidence of endometrial cancer: a population-based study in Finland. Int J Cancer 2002;98:912–5. [23] MacMahon B, Cole P, Lin TM, et al. Age at first birth and breast cancer risk. Bull World Health Organ 1970;43:209–21. [24] Ewertz M, Duffy SW, Adami HO, et al. Age at first birth, parity and risk of breast cancer: a meta-analysis of 8 studies from the Nordic countries. Int J Cancer 1990;46:597–603. [25] Parazzini F, La Vecchia C, Negri E, Fedele L, Balotta F. Reproductive factors and risk of endometrial cancer. Am J Obstet Gynecol 1991;164:522–7. [26] Lesko SM, Rosenberg L, Kaufman DW, et al. Endometrial cancer and age at last delivery: evidence for an association. Am J Epidemiol 1991;133: 554–9. [27] Lambe M, Wuu J, Weiderpass E, Hsieh CC. Childbearing at older age and endometrial cancer risk (Sweden). Cancer Causes Control 1999;10: 43–9. [28] Booth M, Beral V, Smith P. Risk factors for ovarian cancer: a case–control study. Br J Cancer 1989;60:592–8. [29] Merill RM, Fugal S, Novilla LB, Raphael MC. Review. Cancer risk associated with early and late maternal age at first birth. Gynecol Oncol 2005;96:583–93. [30] Cooper GS, Schildkraut JM, Whittemore AS, Marchbanks PA. Pregnancy recency and risk of ovarian cancer. Cancer Causes Control 1999;10: 397–402. [31] Titus-Ernstoff L, Perez K, Cramer DW, Harlow BL, Baron JA, Greenberg ER. Menstrual and reproductive factors in relation to ovarian cancer risk. Br J Cancer 2001;84:714–21. [32] Whiteman DC, Siskind V, Purdie DM, Green AC. Timing of pregnancy and the risk of epithelial ovarian cancer. Cancer Epidemiol, Biomarkers Prev 2003;12:42–6.

M. Hinkula et al. / Gynecologic Oncology 103 (2006) 207–211 [33] Fathalla MF. Incessant ovulation—A factor in ovarian neoplasia? Lancet 1971;2:163. [34] Purdie DM, Bain CJ, Siskind V, Webb PM, Green AC. Ovulation and risk of epithelial ovarian cancer. Int J Cancer 2003;104:228–32. [35] Stakes. http://www.stakes.info 2002. [36] Rostgaard K, Wohlfahrt J, Andersen PK, et al. Does pregnancy induce the shedding of premalignant ovarian cells. Epidemiology 2003;14: 168–73.

211

[37] Risch H. Hormonal etiology of epithelial ovarian cancer, with a hypothesis concerning the role of androgens and progesterone. J Nat Cancer Inst 1998;90:1774–86. [38] Cramer DW, Welch WR. Determinants of ovarian cancer risk: II. Inferences regarding pathogenesis. J Natl Cancer Inst 1983;71: 717–21. [39] Yen S, Jaffe R, Barbieri R. Neuroendocrine metabolic regulation of pregnancy. Reprod Endocrinol 1999;4:768–70.