The Role of Pregnancy, Perinatal Factors, and Hormones in Maternal Cancer Risk: A review of the evidence

Abstract

An understanding of the origin of cancer is critical for cancer prevention and treatment. Complex biological mechanisms promote carcinogenesis, and there is increasing evidence that pregnancy-related exposures influence fetal growth cell division and organ functioning and may have a long-lasting impact on health and disease susceptibility in the mothers and offspring. Nulliparity is an established risk factor for breast, ovarian, endometrial, and possibly pancreatic cancer, while kidney cancer is elevated in parous compared with nulliparous women. For breast, endometrial and ovarian cancer, each pregnancy provides an additional risk reduction. The associations of parity with thyroid and colorectal cancers are uncertain. The timing of reproductive events is also recognized to be important. Older age at first birth is associated with an increased risk of breast cancer and older age at last birth is associated with a reduced risk of endometrial cancer. The risks of breast and endometrial cancers increase with younger age at menarche and older age at menopause. The mechanisms, and hormone profiles, that underlie alterations in maternal cancer risk are not fully understood and may differ by malignancy. Linking health-registries and pooling of data in the Nordic countries have provided opportunities to conduct epidemiologic research of pregnancy exposures and subsequent cancer. We review the maternal risk of several malignancies, including those with a well-known hormonal etiology and those with less established relationships. The tendency for women to have fewer pregnancies and at later ages, together with the age-dependent increase in the incidence of most malignancies, are expected to affect the incidence of pregnancy-associated cancer.

Introduction

In recent years, the influence of pregnancy on subsequent maternal health has become an important focus of research. Changes in reproductive patterns (i.e., having fewer children and giving birth at older ages), pregnancy complications, and exposures occurring earlier in life are being examined. This review focuses on the relation of hormones, pregnancy and other reproductive factors on subsequent maternal cancer risk (summarized in Table 1 and 2). We outline prominent hypotheses concerning biological mechanisms that may underlie the influence of pregnancy on cancer and review epidemiologic studies on several types of cancer. Finally, we describe the resources available in the Nordic countries to facilitate epidemiologic research on maternal risk of cancer.

Table 1.

Summary of pregnancy related biological mechanisms thought to confer risk for developing various maternal cancers

	Breast cancer	Colorectal cancer	Endometrial cancer	Epithelial ovarian cancer	Thyroid cancer
Pregnancy exposure to
excess estradiol	Increased risk	Decreased risk /uncertain	Increased risk	Increased risk	Increased risk/ uncertain
excess progesterone	Increased risk	Unknown	Decreased risk /uncertain	Decreased risk	Unknown
elevated levels of hCG	Decreased risk	Unknown	Unknown	Unknown	Increased risk/ uncertain
elevated levels of IGFs	Increased risk	Unknown	Unknown	Decreased risk	Increased risk/ uncertain
elevated levels of leptin (weight gain)	Increased risk	Unknown	Unknown	Unknown	Unknown

Abbreviations: hCG=human chorionic gonadotropin, IGF=Insulin-like growth factors

Table 2.

Overview of pregnancy-related risk factors for maternal breast, colorectal, endometrial, ovarian and thyroid cancers.

	Breast cancer	Colorectal cancer	Endometrial cancer	Ovarian cancer	Thyroid cancer
Nulliparous*	↑	?	↑	↑	?
Parity	↓	?	↓	↓	?
Lactation	↓	-	-	↓	?
Older age at menarche	↓	?	↓	?	?
Older age at menopause	↑	?	↑	?	?
Older age at first birth	↑	?	?	?	?
Older age at last birth	?	-	↓	?	?
Short pregnancy Length	?	?	-	?	?
Fetal growth/Birth weight	?	?	?	?	?
Twin birth	?	?	?	?	?
Infant sex	-	?	?	?	?
Preeclampsia	↓	?	?	?	?

↓ decreased risk of maternal cancer, ↑ increased risk of maternal cancer, - no association, ? uncertain/conflicting findings

^*never vs. ever

Biological mechanisms

Estrogens and progesterone

Most insights into potential biological mechanisms underlying associations between pregnancy characteristics and cancer risk have been gained by studying breast cancer. Although some associations between pregnancy factors and breast cancer risk are well established, the biologic mechanisms are not fully understood. Several hypotheses have been proposed, all of which involve pregnancy hormones (reviewed in [1, 2] and summarized in table 1. A prominent hypothesis is that women who had relatively high circulating estrogen levels during pregnancy or were exposed to the synthetic estrogen diethylstilbestrol are at greater risk of developing breast cancer (reviewed in [3]). However, the relationship is complex, given the transient increase in breast cancer risk during and after pregnancy (Figure 1), followed by a protective effect of pregnancy after some years, depending on the age of the mother [4, 5].

Figure 1.

Possible mechanisms and exposures mediating effects of pregnancy on subsequent breast cancer risk the mother

Pregnancy-associated breast cancer (PABC) is rare and often diagnosed at an advanced stage with poor prognosis for the mother [5]. PABC is thought to be related to tumor progression due to the large increases in levels of estrogens and progesterone during pregnancy. Recent studies have shown that epithelia from these tumors are enriched with expression of major hormone-regulated genes involved in cell proliferation, metabolism, tumor aggressiveness, and recurrence [6]. Compared to normal epithelia, a significant number of genes associated with cell cycle processes, many of which are hormonally regulated, are also enriched in PABC. Women with a family history of breast cancer have the same breast cancer risk during and following pregnancy as women without a family history [7].

There is consensus that a full-term pregnancy differentiates the breast epithelium and reduces breast cancer risk, despite few human data to support this (Figure 1). Experimental models suggest that the hormones of pregnancy will block any future damage caused by carcinogens or endocrine disruptors through the induction of mammary gland stem cell differentiation (reviewed in [2]).

Human chorionic gonadotropin (hCG)

Synthesis of hCG, initiated shortly after conception, is necessary for maintaining early pregnancy and promoting normal breast cell differentiation. Administration of hCG reduces risk of carcinogen-induced breast cancer in rodents, and elevated hCG levels during the first trimester of pregnancy are associated with reduced long-term risk of breast cancer in women [8]. Preclinical studies have shown that the hCG-induced differentiation of the mammary gland epithelial cells during pregnancy involves morphological, physiological, and molecular changes, resulting in increased DNA repair capabilities of the mammary epithelium. Furthermore, activation of genes controlling programmed cell death, as well as activation of tumor suppressor activity, are mediated through upregulation of inhibin A and B (reviewed in [2]).

Insulin-like growth factors (IGFs)

The role of IGFs in development and progression of a broad range of epithelial cancers is well established [9]. Evidence of an adverse effect of elevated IGF-I during early pregnancy on maternal breast cancer risk, thought to be mediated by the mitogenic and antiapoptotic effects of the hormone on both normal breast epithelial and tumor cells [10], was reported in one study. However, this finding was not confirmed in subsequent studies [11].

Malignant melanoma is the most common malignancy diagnosed during pregnancy [12]. IGF-I seems to play a role in the disease process through its concerted action with pregnancy-associated plasma protein-A (PAPPA), a pregnancy-associated metalloproteinase produced by the placenta that increases the bioavailability of IGF-I. The PAPPA/IGF-I axis has been shown to accelerate melanoma progression during pregnancy [13].

Leptin

Serum leptin concentrations increase during pregnancy, particularly in women who gain an excessive amount of weight [14]. These women also are more likely to develop breast cancer after menopause compared to women who do not exceed pregnancy weight recommendations [15]. Leptin-treated parous rats do not exhibit the pregnancy-induced protective genomic signature observed in parous control rats. Specifically, leptin-treated rats exhibit downregulation of genes involved in cell differentiation, immune functions, and upregulation of genes involved in angiogenesis, growth, and epithelial-to-mesenchymal transition. Interestingly, a similar response also has been observed after estradiol treatment [3].

Preeclampsia

A prevailing feature of preeclampsia, a hypertensive disorder of pregnancy, is anti-angiogenesis, which is also essential in restricting tumor growth. Thus, these disorders are thought to be linked to a reduced risk of solid cancers later in life. Several studies have shown a reduced risk of breast cancer among women with a history of preeclampsia, compared with those with normotensive pregnancies.[1] However, the results are not completely consistent [16, 17] and it is unclear whether the anti-angiogenic imbalance in women who experience preeclampsia is pregnancy-induced, or if they have an innate tendency toward an anti-angiogenic response to biological challenge.

Epigenetics

Epigenetic mechanisms, such as DNA methylation and histone modifications, have emerged as links between hormonal and other types of exposure and disease risk later in life. Long-term epigenetic changes in genes that are important in determining the risk for cancer development are likely to be involved in the various associations described in this review [18]. Recent developments in microarray technology with increased genome coverage of regulatory regions have provided valuable tools for high-throughput human methylome analyses, which are essential for increasing our mechanistic insights in this field [19].

Pregnancy Characteristics and Maternal Cancer Risk

The following sections review the epidemiology of pregnancy risk factors that have been studied in relation to subsequent development of several cancers in women (Table 2).

Breast cancer

Breast cancer is one of the most studied cancers regarding the influence of pregnancy (Figure 1). It has been long known that pregnancy is associated with reduced breast cancer risk. Over 300 years ago, Bernardino Ramazzini noted that “… cancerous tumors are very often generated in the woman’s breast, and tumors of this sort are found in nuns more often than in other women. Now these are not caused by suppression of the menses but rather, in my opinion, by their celibate life” [20]. Only in the mid-20^th century was nulliparity, parity among parous women, age at first birth, and lactation identified as important reproduction-associated risk factors [21–24]. Indeed, early age at first birth is associated with one of the largest risk reductions, and each subsequent pregnancy confers an additional though smaller benefit [25]. Breastfeeding duration is inversely associated with breast cancer risk, with a stronger and more consistent relation for receptor-negative tumors [26]. Whether twin pregnancies are a risk factor for breast cancer is unclear, although several Scandinavian cohort studies, mainly in younger women, have found an approximate 10–30% risk reduction in risk in mothers of twins compared with mothers of singletons (reviewed in [1]).

The relationship of other pregnancy characteristics with breast cancer have been studied less often (reviewed in [17]). To provide some benefit, a full-term pregnancy appears necessary, as miscarriages do not provide equivalent protection [25]. Overall, data are inconsistent regarding whether gestational length is associated with breast cancer risk [4]. The association of maternal weight gain with breast cancer risk is complicated by the strong correlation between pregnancy and later adult weight, and studies for the most part have been inconsistent. However, a Finnish cohort study found that pregnancy weight gain was positively associated with subsequent breast cancer risk independent of weight at diagnosis [27].

Fetal and placental characteristics and breast cancer

The effects of fetal and placental characteristics on maternal breast cancer risk also have been investigated. Offspring sex was hypothesized to affect maternal breast cancer risk through differences in maternal hormones, but results of studies have been inconsistent [1]. Birth weight has been positively associated with maternal breast cancer risk, but there have been few studies, with findings limited to subgroups [28]. Despite the paucity of information on placental size, two studies have demonstrated associations between placental weight, [29] diameter, maternal floor infarctions, [30] and maternal breast cancer.

Pregnancy complications and breast cancer

Certain pregnancy complications also may be associated with subsequent maternal breast cancer risk. As mentioned above, a history of preeclampsia has been linked to an approximate 20% reduction in this risk [1]. Moreover, in one study a marked reduction in breast cancer risk was observed with elevated mean arterial pressure (MAP) [31] and systolic blood pressure increases from mid- to late pregnancy below the diagnostic criterion for hypertension (i.e., in normotensive pregnancies) [30]. Other complications have been studied less because of methodological issues or rarity of occurrence [32]. Studies of gestational diabetes, for example, are challenging because of potentially strong confounding by body mass index and treatment.

Future directions

Because of associations detected with placental characteristics, additional investigation of abnormalities such as placenta previa, accrete, increta, and percreta would be of interest. These investigations will require large studies to reach the statistical precision required for accuracy. In addition, investigations would benefit from availability of information on the hormone receptor status of breast tumors, and their genetic characteristics, as associations with risk factors may vary by these variables.

Colorectal cancer

Sex differences in rates of colorectal cancer

Several epidemiological observations suggest the involvement of steroid hormones in colorectal cancer etiology. In the Nordic countries during 2010–2014, age-standardized incidence rates (based on the world standard population) were 35.7 per 100,000 population in men and 27.8 per 100,000 population in women (NORDCAN) [33]. The lower incidence rates found among women compared with men have yielded speculation that estrogen exposure may confer protection. Estrogen’s influence could be mediated by the anti-proliferative effect of estrogen receptor (ER)-beta in colonic epithelium, by the decline in secondary bile acid production, and/or by lower IGF-I (insulin-like growth factor-I) levels [34–38]. The incidence of colorectal cancer by location also varies by gender. Right-sided (proximal) colon tumors are more common in women, while left-sided (distal) tumors are more common in men [39]. Depending on location, colorectal cancers exhibit differences in pathological and molecular characteristics, including genetic and epigenetic alterations [39].

Exogenous hormones and colorectal cancer

A substantial body of epidemiological research suggests a 20–40% protective effect of menopausal hormone replacement therapy on development of colorectal cancer [40, 41]. A 2012 meta-analysis found that ever use of both estrogen-progestogen therapy [RR (relative risk) 0.74, 95% CI (confidence interval) 0.68–0.81] and estrogen-only therapy (RR 0.79, 95% CI 0.69–0.91) were associated with decreased risks [41]. The effects of duration and recency exposure of hormone therapy are less clear. Oral contraceptives have been studied extensively over the past few decades, with most reporting an inverse association, although results have been conflicting. A 2015 meta-analysis of 17 case-control studies and 12 cohort studies reported a summary RR forever use of oral contraceptives of 0.82 (95% CI 0.76–0.88) [42]. In contrast, relatively recent large cohort studies have not confirmed these associations [43, 44].

Endogenous hormones /reproductive factors and colorectal cancer

Several prospective studies have evaluated women’s reproductive history, as a surrogate measure of lifetime exposure to endogenous sex hormones, in relation to colorectal cancer risk, with conflicting results. An EPIC study (European Prospective Investigation into Cancer and Nutrition) found little evidence for associations between reproductive factors and colorectal cancer [45]. In contrast, a US study (NIH-AARP Diet and Health Study) found that age at menopause and age at first childbirth were positively associated with colorectal cancer, while parity and age at menarche were inversely associated [44]. A Swedish registry-based study reported that parity was positively associated with adenocarcinoma of the proximal colon [46], In a recent report from the Women’s Health Initiative Observational Study in the US, parity was inversely associated with colorectal cancer risk, but no associations were found between other reproductive and menstrual factors and colorectal cancer [47]. The Million Women Study in the UK reported lower colorectal cancer risk in parous than in nulliparous women, but no trend was observed in risk by parity and no associations were identified with other reproductive factors [48]. Furthermore, prospective studies investigating associations between circulating oestrogens and colorectal cancer have reached inconclusive results [47].

A large Nordic population-based case-control study including more than 22,000 cases found no evidence for associations between reproductive history (parity, age at first and last birth, and time since first and last birth) and colorectal adenocarcinoma in parous women overall [49], by specific subsites (proximal and distal colon and rectum), or in analyses stratified by mother’s year of birth, parity, and proxies for menopausal status. The cases and controls were relatively young (mean age at diagnosis 57 years), as it was restricted to women with a prior birth recorded in national birth registries; inclusion of older cases may have changed the results. Future studies would benefit from robust analyses of possible confounders, such as use of exogenous hormones, obesity, alcohol, smoking, and aspirin and NSAID (nonsteroidal anti-inflammatory drug) use.

Future directions / opportunities

Results of studies evaluating associations between reproductive history and colorectal cancer have been conflicting. Most have demonstrated inverse associations between oral contraceptive use, menopausal hormone therapy and colorectal cancer. However, it is difficult to quantify a woman’s total lifetime hormonal exposure and still unclear if this is the relevant risk factor. Furthermore, reproductive behavior, as well as use and content of exogenous hormones, are changing over time [50]. Larger studies may provide clarification by allowing comparison of risks across anatomical subsites and across morphologically and molecularly defined subtypes, and by incorporating information on colorectal cancer screening.

Endometrial cancer

Much of what is known about the epidemiology of uterine cancer relates to endometrial cancer, as uterine sarcomas comprise only 3–7% of uterine malignancies [51]. The risk of endometrial cancer rises sharply among women in their late forties to mid-sixties and is strongly dependent on lifetime hormonal exposures. Various aspects of reproduction and use of exogenous hormones have been explored extensively in regard to endometrial cancer.

Estrogens and endometrial cancer

Obesity and menopausal hormone therapy with unopposed estrogen are two established risk factors for endometrial cancer [52–56]. Given the risks of unopposed estrogens, they are prescribed with progestins for women who have not had a hysterectomy. Research has shown that the number of days on progestin each month is important for endometrial cancer risk [57, 58]. Long duration use of sequential estrogen plus progestins (<10 days’ progestin per month) is associated with modest increases in risk, while continuous (>15 days per month to daily) estrogen plus progestin has been associated with null or decreased endometrial cancer risk, suggesting complete protection against adverse effects of exogenous estrogen on the endometrium [57, 58]. Clinical trials of tamoxifen-treated breast cancer patients have reported increased risk of endometrial cancer, consistent with tamoxifen’s estrogenic effects on the endometrium. Elevated risks specifically have been observed within short periods following exposure and among women receiving high cumulative doses of tamoxifen [59]. Regarding obesity, very heavy women have exceptionally high risks. It is estimated that obesity may account for up to 25% of endometrial cancer cases [55]. Further, the relation between endometrial cancer and obesity/increased body mass index (BMI) appears stronger among women not exposed to exogenous estrogens [56]. Increased physical activity has been associated with reduced risk of endometrial cancer, independent of BMI [60]. While sedentary behavior or prolonged sitting has been associated with increased risk, it remains to be determined if this is independent of physical activity or BMI [60]. The role of dietary factors and endometrial cancer risk remains inconclusive, despite extensive research.

Past use of combination oral contraceptives is associated with marked reductions in endometrial cancer risk, with the greatest risk reductions for use of long duration [61]. Risk reductions persist for more than 30 years after last use. Hypothesized mechanisms underlying this risk reduction include reduced exposure to unopposed estrogen during the follicular phase of the menstrual cycle and the ability of progestins to oppose estrogen-induced cellular proliferation.

Pregnancy and pregnancy characteristics and endometrial cancer

Pregnancy is known to confer long-term protection against endometrial cancer. Conversely, nulliparity is associated with elevated endometrial cancer risk [62, 63]. Primary and secondary infertility are also associated with increased endometrial cancer risk, with independent effects.[64] The hormonal milieu of pregnancy is characterised by elevated levels of estrogen, progesterone, and intrauterine growth factors, almost exclusively produced by the placenta.

The pregnancy history of women with endometrial cancer has been examined in depth, including timing of births, [62, 63, 65–67], twin births, and sex of offspring.[68] Several studies have reported decreased risks with either older age or shorter time since last birth. Investigators have hypothesized that this reflects a protective effect of the mechanical clearance of initiated cells [62, 63, 65–67]. However, existing studies have been unable to evaluate timing of pregnancy associations by histologic subtype [67]. Birth of twin boys, but not twin girls or non-sex-concordant twins, appears to put women at increased risk of endometrial cancer [68]. Pregnancy complications such as preeclampsia also have been explored in relation to endometrial cancer, with inconclusive results [69, 70].

Diabetes and polycystic ovarian syndrome (PCOS) have been linked to increased endometrial cancer risks. However, the extent to which these relations are independent of obesity remains unclear [71, 72]. It also has been hypothesized that diabetes’ association with endometrial cancer risk may be related to increased exposure to proliferative effects of circulating insulin, [73] which also has been associated with endometrial hyperplasia, a precursor to type I endometrial cancer. Given the positive association between diabetes and endometrial cancer, it is also plausible that gestational diabetes may be related to maternal risk of endometrial cancer later in life. Gestational diabetes has been associated with increased risk of both endometrial hyperplasia and endometrial cancer [74].

Future directions

Many identified risk factors (Table 1) are hypothesized to operate through alterations in endogenous hormones, such as exposure to increased bioavailable estrogen unopposed by progesterone/progestins. However, the mechanisms mediating many of the exposure-disease relationships have not been studied fully. Recent data indicate that the associations between reproductive factors and endometrial cancer differ by sub-type (types I and II), [67, 75], but further studies are needed.

Ovarian cancer

Epithelial ovarian cancer has the worst prognosis of all gynecological malignancies with an approximately 45% overall 5-year relative survival rate [76]. Ovarian cancer is divided into five subgroups: high-grade serous (70%), low-grade serous (5%), clear cell (10 %), endometrioid (10%), and mucinous (5%). Risk of epithelial ovarian cancer is strongly associated with reproductive factors (Table 2). Parity and oral contraceptive use are well known to be associated with decreased risk of ovarian cancer, while menopausal hormone therapy is linked to higher risk.

Biological hypotheses for ovarian cancer risk

The underlying mechanism for ovarian cancer remains unclear. Several hypotheses have been proposed (Figure 2) [77]. Among the most cited is the incessant ovulation hypothesis, which posits that repeated trauma and subsequent repair of the ovarian epithelium, caused by ovulation, increases the risk of mutations. However, this hypothesis cannot explain why nine months of pregnancy has a stronger protective effect than nine months of anovulation due to other causes. Similarly, polycystic ovarian syndrome (PCOS), which causes anovulatory cycles, is not associated with a lower risk of ovarian cancer [76]. In addition, because the serous subtype is thought to originate from the distal fallopian tube rather than the surface epithelium of the ovary, the protective role of anovulation is unclear for this subtype.

Figure 2.

There are several proposed hypotheses for ovarian cancer development-although none fully explain the mechanisms

Other postulated explanations include increasing malignant transformation of the ovarian epithelium through high levels of pituitary gonadotropins, either directly through luteinizing hormone and follicle-stimulating hormone, or through estrogen stimulation. The androgen/progesterone hypothesis posits that increased levels of androgens stimulate the ovarian epithelium, while high levels of progesterone decrease ovarian cancer risk [77]. According to the inflammatory hypothesis, inflammatory conditions increase ovarian cancer risk, [76] and according to the cell clearance hypothesis, high levels of progesterone (or possibly other hormones during pregnancy) clear precancerous cells from the epithelium of the ovary [78]. However, none of these hypotheses fully explains the mechanism behind the risk of ovarian cancer associated with pregnancy factors and endocrine therapies.

Pregnancy-related risk factors and ovarian cancer

Parous women have a 30–40% lower risk of developing ovarian cancer, and an additional protective effect is seen with increasing parity [76, 79]. Lactation is also protective, and the effect size increases with duration of breastfeeding [80]. Abortion does not seem to influence the risk of ovarian cancer, [81] suggesting that a longer period of hormone exposure or anovulation is needed. Studies on pre- and post-term delivery have had inconsistent results, [82, 83] and further studies are needed to understand fully the effect of pregnancy length. Other pregnancy-related factors associated with differing levels of reproductive hormones, such as high or low birth weight, [82] preeclampsia, [69, 84] gender of offspring, [83, 85] twin birth, [83] and placental weight, [86] have not been conclusively associated with ovarian cancer. Pregnancies at older ages seem to provide stronger protection against ovarian cancer than pregnancies at younger ages [76].

Other reproductive risk factors and ovarian cancer

Oral contraceptives have been found to reduce the risk of epithelial ovarian cancer by about 20–30% after 5 years of use [87]. Parity confers a similar level of protection. Intrauterine devices have been studied less, but seem to be protective, especially with long-term use [88]. Tubal ligation and hysterectomy each confer about a 30% risk reduction [89].

A 2013 Cochrane analysis found no convincing evidence of an association between infertility or infertility treatment and risk of ovarian cancer. However, the complexity of infertility makes it difficult to study. It has been speculated that both infertility and ovarian cancer may have a common underlying cause [90].

The effect of ages at menarche and menopause has been extensively studied with inconsistent results and any association with ovarian cancer seems to be weak [91]. Postmenopausal hormone therapy is associated with an increased risk of epithelial ovarian cancer. A stronger association has been observed with longer duration of use [61, 79].

Future directions

For decades, a history of childbirth has been recognised as a strong protective factor against epithelial ovarian cancer. However, the underlying mechanism remains unclear. As well, differences in disease mechanisms recently have been observed for subtypes of ovarian cancer. Further investigation of subgroup-specific risk factors may be informative. Pregnancy leads to a period of anovulation, reduced gonadotropin secretion, and increased estrogen and progesterone levels. The importance of each factor needs to be explored further, possibly through mechanistic or pooled studies focusing on women with specific pregnancy and offspring characteristics (i.e., premature delivery, twin birth). A better understanding of the biology behind the risk associations possibly could provide a strategy for mimicking the protective mechanism in women at high risk of ovarian cancer.

Thyroid cancer

Thyroid cancer, the incidence of which is increasing rapidly in most countries worldwide, is one of the most common malignancies diagnosed among women of reproductive age [92]. Apart from childhood exposure to ionizing radiation and obesity [93, 94] few modifiable risk factors for thyroid cancer have been identified. Sex steroid hormones, including estrogen, have been hypothesized to play a role in thyroid carcinogenesis due to the higher female-to-male incidence, which begins in early adolescence, peaks during the reproductive ages, and diminishes around the ages when most women experience menopause [95–97]. However, the “classical” estrogen-related risk factors for breast, ovarian, and endometrial cancers (e.g., age at menarche, age at first birth, parity, number of live births, use of oral contraceptives, and use of menopausal hormone therapy) have been much less consistently associated with female thyroid cancer risk [93, 97]. However, such factors may not reflect the types or timing of hormonal changes that are most relevant for thyroid cancer development.

A limited number of studies have evaluated other indicators of sex steroid hormone exposure. Greater duration of breastfeeding, characterized by delayed ovulation and suppression of gonadotropins and estradiol, has been associated with a non-significant reduction in thyroid cancer risk in two prospective studies [98, 99]. Longer menstrual cycle length, irregular menstrual cycles, and greater estimated lifetime number of ovulatory cycles have each been associated with an increased risk of thyroid cancer, but findings across studies are inconsistent [100–102]. A medical history of uterine fibroids and breast cancer have each been linked to an increased risk of thyroid cancer [102, 103], possibly reflecting shared genetic, immunologic, or hormonal etiologies.

Some studies have shown an excess risk of thyroid cancer in the first five years after pregnancy but not subsequently [99, 101] suggesting a transient effect of certain pregnancy-related hormones or other exposures. Increasing levels of endogenous estrogen throughout gestation may promote thyroid carcinogenesis via indirect stimulation of the pituitary-thyroid axis or direct stimulation of normal and malignant thyroid cell growth and proliferation; these effects may be more pronounced in the context of iodine deficiency [104]. Hyperemesis gravidarum, an early-pregnancy complication related to altered pregnancy hormone levels, including hCG and estrogen, has been associated with increased risk of thyroid cancer in both the mothers and their offspring [105, 106]. Other placental hormones and growth factors that are increased during pregnancy may also contribute to thyroid cancer development. For instance, findings from a recent Swedish study showing an increased risk of maternal thyroid cancer associated with higher fetal growth and birth weight support a possible role of placental IGF-I [107]. A history of infertility and use of fertility drugs have been associated with thyroid cancer risk in some studies [108, 109], albeit not entirely consistently [99]. It remains unclear whether some of the observed findings, including the increased risk shortly after pregnancy, simply reflect greater likelihood of thyroid cancer detection among women who undergo more frequent medical examinations or who have a medical condition (e.g., infertility) that may be linked to an underlying thyroid disorder [100]. Future investigations of this topic should attempt to minimize detection-related biases, for instance, by restricting thyroid cancers to more clinically relevant or aggressive subtypes.

Pregnancy factors and risk of leukemia/lymphoma, sarcomas, and other solid tumors

The relations between pregnancy characteristics and other malignancies have been much less investigated than breast, endometrial, and ovarian cancers. This is likely due to the less pronounced hormonal etiology of these malignancies and their lower incidence.

While malignant melanoma is the most common cancer arising during pregnancy, accounting for about a third of malignancies among expectant mothers, parous women are not at higher risk of subsequently developing melanoma than nulliparous women [110]. Similarly, despite lower incidence of leukemias and lymphomas in women than in men, there is little evidence of associations with pregnancy factors and lymphoma. Parity is unlikely to play an important role in the etiology and disease progression of Hodgkin lymphoma [111, 112]. Pregnancy- related hormonal or immunological changes seem to have only a minor influence in the etiology of leukemias. However, one study did find a small tendency toward reduced risk of chronic myeloid leukemia with higher parity [113] and another reported short-term protection against acute myeloid leukemia with pregnancy [114]. Risk of sarcoma was not associated with parity and number of abortions in one study, [115] but another suggested an increased risk in women who were older at first birth [116].

Meta-analyses of pancreatic cancer show a risk reduction in parous women compared with nulliparous women [117], with two children being most protective [118]. Higher risk of pancreatic cancer has been associated with older age at first birth [119]. In contrast, a meta-analysis demonstrated an increased risk of kidney cancer in parous compared with nulliparous women, and an increase in risk with each subsequent birth [120].

Parity as a cofactor

Parity has been evaluated as a cofactor for cancer risk. In high-risk HPV-infected women, higher parity and younger age at first birth was associated with increased risk of cervical cancer [121]. In never smokers, increased parity was associated with decreased risk of lung cancer among parous women, [85, 122] while in current smokers, older age at first birth was associated with increased risk [85].

With the finding that hypertensive disorders of pregnancy are protective for breast cancer, there has been interest in evaluating their association with other tumors. In a large linked-registry study, hypertensive disorders of pregnancy were not associated overall with a meaningful reduction in risk of solid tumors. While there was a general tendency toward risk reductions across tumor sites, only a decreased risk of breast and lung cancer and an increased risk of endometrial and urinary tract cancers were statistically significant [16].

Methodological challenges

Understanding the role of pregnancy on subsequent maternal cancer risk is challenging due to the relatively long latent period between exposure and disease, the possibility of bias in recall of information about pregnancy in case-control studies based on interviews or questionnaires, and the large numbers of cases required for stable estimates when studying rare exposures or cancer types. The next section describes the advantages of using linked registry studies for this purpose.

Use of Nordic registers in research

The Nordic countries have a long tradition of collecting administrative health and welfare data for management, organisation, planning, evaluation, and quality control purposes, [123, 124] through legislation that allows the collection of nationwide health data without explicit informed consent of registered individuals (Figure 3). All health care providers, both public and private, are required to submit requested data to the registry custodians. The compilation and maintenance of health registry data and their use in research have been widely accepted by the population of the Nordic countries. For example, the Eurobarometer [125] reported that between 88% and 90% of Danish, Finnish, and Swedish respondents trust health and medical institutions to protect their personal information, while only 74% of persons surveyed in the rest of EU shared this point of view.

Figure 3.

Main Health Registries in Denmark, Finland, Norway, and Sweden with year of establishment.

Use of existing registry data in studies of pregnancy characteristics and subsequent cancer risk significantly reduces total study costs and time spent on data collection compared with primary data collection [126], provides long-term follow-up data, and eliminates some selection bias. Other countries also have established registries for administration and surveillance, such as the Surveillance Epidemiology, and End Results Program and Medicare linked registries in the United States [127]. However, registry-based research in the Nordic countries has a major unique advantage – the Personal Identity Code (PIC) system, making it possible to link various data sources, including all other registers, but also for example medical records and biobank samples.

A PIC is assigned to all residents at birth or upon immigration and is recorded during all contacts with health care systems and other public authorities [128–130]. Personal identity codes, maintained by Central Population Registers, allow for tracking of health information virtually from birth to death [131]. Linkage of registry data across the Nordic countries permits the study of real-world populations who are medically managed using standard clinical practice. It allows the study of rare exposures, such as eclampsia or other pregnancy complications recorded in medical birth registries or rare malignancies recorded in cancer registries. As medical birth registries were established in Nordic countries beginning in the late 1960s, and cancer registries even earlier, cancer incidence can be followed up for over 40 years after birth (Figure 1). The scientific value of these registries will increase over time since the incidence of most cancers increase with age.

As all residents of the Nordic countries are guaranteed unfettered access to health care and are captured in registries if they had a birth or a cancer diagnosis, selection bias introduced by including patients treated by specific hospitals or with specific insurance plans is virtually eliminated. Moreover, all countries have hospitals registries with data on dates of admission and discharge, emergency room visits, outpatient clinic visits and discharge diagnoses from all non-psychiatric hospitals.

Conclusions

An understanding of the origin of cancer is crucial for cancer prevention and treatment. Complex biological mechanisms promote carcinogenesis, and there is increasing evidence that pregnancy-related exposures influence fetal growth cell division and organ functioning and may have a long-lasting impact on health and disease susceptibility in the mothers and offspring. In addition, understanding the role of pregnancy in the subsequent health of the mother and offspring is important as more recently, women have delayed pregnancy until older ages and have had smaller families. However, our knowledge is still very limited. This review has provided evidence that hormones and pregnancy-related factors are involved in the carcinogenesis of several types of cancer. The rapid development in molecular biology and technology will most likely lead to better understanding of the complex interactions between risk factors and genetic predispositions. The abilities to link various health registries and to pool data across the Nordic countries provide further opportunities to conduct high-quality research of pregnancy exposures and subsequent maternal cancer risk and cancer in the offspring.

The value of these resources will grow in the future with the increasing number of pregnancy and (cancer-related) biobanks [132, 133], where information on molecular and cellular events can be linked to health information from registries. The coming generations in the Nordic countries can be followed from cradle to grave. In addition to biobank data, many other important health registries can be added to the cancer and birth registry data, such as the prescription and in-vitro fertilization (IVF) registries, and registers on social benefits and services, to improve our understanding of potential preventable causes of cancer.