Ovarian cancer risk factors

The aetiology of ovarian cancer is not yet completely clear. The strongest known risk factors are increasing age (Figure 4.1) and the presence of certain gene mutations, the latter accounting for around 10% of cases. As more research is carried out into the histological diversity and origin of ovarian cancers, so it may become more fruitful to examine risk factors by histological subtype.1,2 A summary of the most well-researched factors which may raise or lower risk is given below.

Figure 4.1: Ovarian Cancer (C56-C57), Number of New Cases Per Year and Age-Specific Incidence Rates, UK, 2006-2008

Download this chart (62KB)

section updated 08/03/11

Women who have a first-degree relative diagnosed with ovarian cancer have a three to four-fold increased risk of developing the disease compared with women with no family history, although only about 10% of ovarian cancer cases occur in women with a family history3, 4 The known susceptibility genes (e.g. BRCA1 and BRCA2) explain less than 40% of the excess risk of familial ovarian cancer.3 These estimates suggest that more research is needed.

section updated 08/03/11

Ovarian cancer risk tends to be reduced by factors which interrupt ovulation such as pregnancy, breastfeeding, and oral contraceptive use, while those that prolong exposure to ovulation such as nulliparity and infertility increase risk.1, 5-7 While the epidemiological evidence is less consistent for some of these factors, there is good evidence that both pregnancy and oral contraceptive use lower risk.

Pregnancy

Women who have given birth have a lower risk of ovarian cancer than women who have not.8 There is a dose response relationship between increasing risk and a lower number of children (Table 4.1).4 Studies have also shown a risk reduction for incomplete pregnancies.9, 10

Table 4.1: Relative Risk for Ovarian Cancer by Parity and Duration of Oral Contraceptive Use

* Relative risk stratified by study, age, parity, and hysterectomy.

• Download this table (27.5KB)

Breastfeeding

Evidence for a protective effect of breastfeeding is conflicting. Results from a combined analysis of two cohorts of parous women showed that breastfeeding for 18 or more months reduced the women’s risk by 34%.11 However, a large case-control study showed no effect of breastfeeding after parity when other potential confounders were taken into account.12 One study found little evidence of reduced risk for those who breast-fed some children when the last born child was not breast-fed.12, 13 Based on the protective association shown in the combined analysis, a study published in December 2011 estimated that around 18% of ovarian cancers in the UK in 2010 were linked to insufficient breastfeeding.63

Infertility

There is some evidence to suggest that infertility increases risk. Two cohort studies have shown a 36-46% risk increase for ovarian cancer in infertile women which was not the effect of fertility drugs.14, 15 Research suggests that neither assisted reproductive technology nor fertility drugs has an impact on risk of ovarian cancer overall, although one study found a 67% risk increase for serous tumours after use of clomifene citrate (a drug used to treat infertility which blocks the effect of oestrogen in the body).16,17

Exogenous hormones

• Oral contraceptives

Oral contraceptives (OCs) are an established protective factor for ovarian cancer. A re-analysis of 45 separate studies conducted in 21 countries showed that the longer a woman has used OCs, the greater her reduction in risk (Table 4.1).18 Women who have used OCs for 15 years or more halved their risk of ovarian cancer. The risk reduction was shown to be long-term, persisting for 30 or more years after OC use had ceased.18 Use of OCs has also been shown to reduce the risk of ovarian cancer in women with a BRCA1 or BRCA2 mutation.44 A study published in December 2011 estimated that, in 2010, there were around 700 fewer cases of ovarian cancer in the UK linked to the protective effect of OCs.64

• Hormone replacement therapy

A systematic review of published case-control and cohort studies and randomised trials has studied the effect of oestrogen-only and combined oestrogen-progestin hormone replacement therapy (HRT) in relation to ovarian cancer risk. It reported that five years’ use of oestrogen-only HRT increased the risk by 22% - significantly more than the 10% risk increase with use of oestrogen-progestin HRT.19 According to the UK Million Women Study, risk was increased for current users of HRT and the risk increased with duration of use becoming significant after seven or more years of use. Past or short-term use of HRT was unlikely to increase the risk of ovarian cancer.20 A study published in December 2011 estimated around 50 cases of ovarian cancer in the UK in 2010 were linked to HRT, equivalent to around 1% of all ovarian cancers.64

section updated 08/03/11

Smoking

IARC recently stated that there is sufficient evidence that smoking causes ovarian cancer.21 A systematic review showed a doubling in risk of mucinous tumours in current smokers, no effect on serous and endometroid cancers and a 40% reduction in risk of clear cell tumours.22 A similar risk increase for mucinous tumours for current and past smoking was subsequently shown in the Nurses’ Health Study.24 A study published in December 2011 estimated that, in 2010, more than 2% of ovarian cancers in the UK were caused by smoking.65

Physical activity

Evidence is mixed. A meta-analysis of case-control studies showed a 21% risk reduction for women with the highest versus the lowest levels of recreational physical activity, but the combined results of cohort studies did not show any risk reduction.24 Subsequently, the European Prospective Investigation into Cancer and Nutrition (EPIC) found no risk reduction for women with the highest levels of total, occupational, recreational or household physical activity,25while a case-control study showed a 60% risk reduction for serous tumours for women doing the most recreational physical activity, but an increase in risk of clear cell and endometroid tumours.26

section updated 08/03/11

Height

Studies show a risk increase of approximately 40% for women measuring 1.7m or over compared to women of less than 1.6m in height.27,28

Body mass index

The evidence points to a probable link between body mass index (BMI) and ovarian cancer. A pooled analysis of 12 prospective studies showed a 75% increase in risk of ovarian cancer in premenopausal women who were obese (BMI of 30 or higher) compared to women of a healthy weight (BMI of 18.5-23). There was no risk increase with the same comparison in postmenopausal women.27 However, results from EPIC and the UK Million Women Study indicate that being obese after the menopause may also increase the risk of ovarian cancer.29,30

section updated 08/03/11

The epidemiological evidence is not strong enough to make any dietary recommendations as a means of reducing the risk of developing ovarian cancer. Fruit and vegetables do not appear to affect risk31-33 nor does alcohol.34 The evidence for other dietary factors, such as meat, fat, fish, dairy products, tea and phytoestrogens, is inconclusive.

section updated 08/03/11

Previous cancer

Studies have shown a doubling in ovarian cancer risk for women with a previous breast cancer.35 For women whose breast cancer was diagnosed before the age of 40, a four-fold risk increase has been shown. Risk is even higher for women in this group with a family history of ovarian or breast cancer.35 Long term risk from radiotherapy is an issue - women treated for cervical cancer 30-39 years ago had a 73% higher risk, and those treated 40 or more years ago had a 172% higher risk of ovarian cancer.36

Endometriosis

Endometriosis is a common condition in which endometrial tissue is found outside the uterus, for instance, on the fallopian tubes and ovaries. Women with endometriosis have been shown to have a 30-66% increased risk of ovarian cancer. 37,38

Ovarian cysts

Young women (15-29 years old) with ovarian cysts and functional cysts (harmless, short-lived cysts that are formed as a part of the menstrual cycle) have been shown to have a doubling in ovarian cancer risk later in life, and women who had cysts surgically removed, or unilateral oophorectomy, have a nine-fold risk increase.39

Hysterectomy

Hysterectomy may reduce ovarian cancer risk, with case-control studies reporting a 30-40% risk reduction regardless of age at time of surgery, and a 50% risk reduction for women whose hysterectomy was 15 or more years before the study.40,41

Tubal sterilisation

Results from the Nurses’ Health Study showed a 34% risk reduction for ovarian cancer in women reporting a history of tubal ligation.14 A recent meta analysis showed a similar result, though other cohort studies have not shown an effect and results of case-control studies have been conflicting.42-45

Intrauterine device

The Nurses’ Health Study showed a 76% increased risk for women reporting use of an intrauterine device, compared to women who had not used an intrauterine device. Results were adjusted for duration of OC use.14

Non-steroidal anti-inflammatory drugs

Studies of anti-inflammatory drugs in relation to ovarian cancer are conflicting. A 2005 meta-analysis showed no effect of aspirin and other non-steroidal anti-inflammatory drugs (NSAIDS) on risk.46 A recent large cohort study of almost 200,000 women found no effect of regular use of NSAIDs, or aspirin specifically, on risk.47 A subsequent, much smaller cohort study showed a 39% risk reduction for women taking aspirin six or more times per week.48 Results of recent case-control studies have varied between showing a reduction in risk, no effect on risk or an increase in risk in relation to use of NSAIDs.49-51 Because of the potential adverse consequences of high intake of aspirin, such as gastrointestinal haemorrhage, it would not be recommended as a prophylactic measure.

Paracetamol

Evidence is conflicting. A meta-analysis showed a 30% risk reduction with regular use of paracetamol.52 However, a recent large cohort study of almost 200,000 women found no effect of regular use of paracetamol on risk,47 while a case-control study showed an 80% increase in risk with long-term use.51

section updated 08/03/11

A 2003 meta-analysis of 16 individual studies showed a 33% risk increase for ovarian cancer in relation to perineal talcum powder application.53 One study, which looked at use of talc both in the perineal and non-perineal area, showed a doubling in risk for long-duration (>20 years), with at least daily use, compared to women who never used talc.50

Before the mid-1970s, contamination of talc with asbestos fibres was known to occur, and in 1975 guidelines were introduced to prevent this.54One study, which examined year of talc use, showed that use before 1975 was associated with an increase in risk, whereas use after 1975 was not.50 This may explain some of the risk increase shown (see below).

section updated 08/03/11

The International Agency for Research on Cancer (IARC) classifies asbestos exposure as an ovarian carcinogen.55 Studies from the 1970s and 1980s have shown that risk of ovarian cancer death is increased by around three–five times in women with “severe” occupational asbestos exposure, compared with background mortality rates.56-58 More recent studies of women employed in the asbestos industry before the 1980s also show an association with long-term occupational exposure to asbestos.59-60 Asbestos fibres have been found in ovarian tissue, and at higher rates among women living with men with documented asbestos exposure than those living with men with no documented asbestos exposure.61

section updated 08/03/11