Parity and Risk of Thyroid Cancer: a Population-Based Study in Lithuania

Abstract

An association between parity and thyroid cancer risk has been investigated in a number of independent studies but yielded contradictory findings. The aim of this study was to explore the association between parity and thyroid cancer risk. The population-based cohort study in Lithuanian was conducted. The study dataset based on the linkages between all records from the 2001 population census, all cancer incidence records from the Lithuanian Cancer Registry, and all death and emigration records from Statistics Lithuania for the period between 6 April 2001 and 31 December 2009. Cox’s proportional hazards regression models were used to estimate the hazard ratios (HRs) for parity, age at first birth, number of children, place of residence, education, and age at census. The cohort of 868,105 women was followed for 8.6 years, and 1775 thyroid cancer cases were diagnosed during the study period. The significantly higher thyroid cancer risk was observed among parous women (HR = 1.45, 95% CI: 1.20, 1.75) and in women with 1, 2, and 3 children, after adjusting for the possible confounding effects of relevant demographic variables. The findings of this study are consistent with the hypothesis that parity might be associated with the risk of thyroid cancer in women.

Introduction

Thyroid cancer is the most common endocrine cancer, accounting for more than 95% of these malignancies [1]. The incidence of thyroid cancer varies considerably across geographic locations, populations, ethnic groups, and gender, and accounts for about 1% of all cancers [2]. In Lithuania, thyroid cancer amounts 0.6 and 3.7% of all new malignant tumors in men and women, respectively (male:female ratio 1:3) [3]. The incidence of thyroid cancer has significantly increased in many countries, including Lithuania, over the last few decades, and the most notable changes have been observed for papillary thyroid cancer [3, 4]. The causes of the increasing incidence rates are not fully understood and may be due to screening, use of newer diagnostic technologies allowing early detection of thyroid cancer, changes in diagnostic criteria, or a true increase due to unidentified multifactorial environmental or lifestyle factors [2, 5].

With the exception of exposure to ionizing radiation during childhood and adolescent, only a few other potential risk factors such as sex hormones and iodine deficiency for thyroid cancer have been identified. However, to date, a substantial share of the causes of thyroid cancer remains unknown or not fully indentified [2, 5, 6]. Gender differences in thyroid cancer are systematically found in different continents and regions. Generally, thyroid cancer among women is three times more frequent than among men, whereas ionizing radiation and nutritional factors seem not to be involved in this gender discrepancy [1, 6].

The marked predominance of thyroid cancers among women tends to be the highest at the reproductive ages peaking at 40–49 years indicating that reproductive and hormonal factors may contribute to the risk in tumor development [6]. An association between the parity and thyroid cancer risk has been investigated in numerous independent studies but yielded contradictory findings. Several recently published reviews suggest that there is a significantly increased risk of thyroid cancer after childbirth [7, 8]. Although many studies examining the relationship between thyroid cancer and number of live-born children (parity) among women point to a significant positive association with thyroid cancer risk, not all studies on the subject have reached the same conclusion [9–14]. Such heterogeneity of the results might be attributed to differences in thyroid cancer incidence across countries and regions, different study design, insufficient statistical power, and sample sizes, and therefore, the exact relationship needs clarification by further studies. To explore the association between parity and thyroid cancer risk, the present population-based cohort study in Lithuanian was conducted.

Materials and Methods

The study uses a dataset based on the linkages between all records from the 2001 population census, all cancer incidence records from the Lithuanian Cancer Registry, and all death, and emigration records from Statistics Lithuania for the period between 6 April 2001 and 31 December 2009. All the linkage procedures were implemented by Statistics Lithuania following data protection rules. For the purposes of this study, an aggregated frequency dataset combining cancer diagnoses, population exposures for every possible combination of relevant sociodemographic, and epidemiological variables was provided. For the analysis, the frequency records representing two or more cases were split into individual records.

The cohort includes all women aged 40–89 from the official census organized by Statistics Lithuania on the 6th of April 2001. Information about parity, number of children, age at first birth, education, and place of residence was taken from the census records. Women, for which information about parity and/or number of children was not available in the census data, were excluded from the study (2.3%). Information about diagnosed thyroid cancers comes from the Cancer Registry records. Person years were calculated for the period from the 6th of April 2001 until the date of thyroid cancer diagnosis, death, emigration, or the end of the follow-up (December 31, 2009), whichever came first.

Cox’s proportional hazards regression models were used to estimate the adjusted hazard ratio (HR) and 95% confidence interval (CI), as well as to estimate the effect of parity while adjusting for the possible confounding effects of relevant demographic variables. STATA 11 statistical software was used to carry out the analyses (StataCorp. 2009. Stata Statistical Software: Release 11.0. College Station, TX, USA).

Results

Selected demographic and childbearing characteristics of the study population are presented in Table 1.

Table 1.

Distribution of the baseline characteristics of the study population

Characteristics		Number of women	Percentage
Total		868,105	100.0
Parity	Nulliparous	87,942	10.1
	Parous	780,163	89.9
Number of children	1	179,556	20.7
	2	366,451	42.2
	3	138,026	15.9
	4	49,785	5.8
	5 +	46,345	5.4
Place of residence	Urban	573,228	66.0
	Rural	294,877	34.0
Education	Higher	131,095	15.1
	Secondary	392,814	45.3
	Lover than secondary	344,196	39.6

Selected demographic and childbearing characteristics of the study population

During the study period, 1775 thyroid cancer cases were diagnosed. In the univariate analysis parity, number of children, place of residence, education and age at census affected thyroid cancer risk significantly, age at first birth was an insignificant variable. After adjustment for age at census, place of residence, and education, a significantly higher thyroid cancer risk was observed in parous women (HR = 1.45, 95% CI: 1.20, 1.75) and in women with one, two, and three children (Table 2). The highest risk was found among women with two children (HR = 1.60, 95% CI: 1.32, 1.93). For women with four and five and more children, thyroid cancer risk also exceeded that found among nulliparous women, but the observed differences were not significant.

Table 2.

Hazard ratios (HR) of thyroid carcinoma according to reproductive variables

Characteristics		Person years	Cancers	HR (95% CI)
				Unadjusted		Adjusteda
Parity	Nulliparous	669,552.30	119	ref.	–	ref.	–
	Parous	6,252,016.15	1656	1.48	1.23, 1.79 b	1.45	1.20, 1.75
Number of children	1	1,442,607.45	357	1.39	1.13, 1.70	1.27	1.03, 1.57
	2	2,996,507.76	918	1.71	1.41, 2.07	1.60	1.32, 1.93
	3	1,094,621.74	262	1.34	1.08, 1.67	1.44	1.16, 1.79
	4	379,098.81	67	0.99	0.74, 1.34	1.21	0.89, 1.64
	5 +	339,180.39	52	0.87	0.63, 1.20	1.14	0.82, 1.59

^aAdjusted for age at the census, residence, and education

^bStatistically significant differences are marked in bold

Discussion

The cohort of 868,105 women was followed for 8.6 years, and 1775 thyroid cancer cases were diagnosed during the study period. The findings of our study showed a 45% higher risk of thyroid cancer risk in parous women vs. nulliparous women. In contrast to our study, some other cohort studies have failed to confirm these associations. A cohort study conducted in Norway, for example, did not find any increase in the risk of thyroid cancer among parous women [9]. No significant association as related to parity, number of births, or age at first full-term pregnancy was either observed in a large prospective cohort study carried out in ten countries of Western Europe (Denmark, France, Germany, Greece, Italy, Norway, Spain, Sweden, The Netherlands, and UK) [11]. The US Radiologic Technologists Study also found no clear association between parity and the incidence of thyroid cancer [12]. In the studies conducted in Japan and Thailand, the thyroid cancer risk was even lower in those who had experienced pregnancy or a live birth than those for nulliparous; however, the conclusions were based on 86 and 17 cases of thyroid cancer, respectively [15, 16]. Furthermore, no significant association between papillary thyroid cancer risk and parity was found in the analysis of nine individual epidemiological studies with a total of 833,852 subjects [17]. Our findings were supported by the results of the prospective California Teachers Study, where nulliparous women had not significantly lower risk of thyroid cancer when compared to women with one or two children [13]. Likewise, not significantly increased risk for parous women was reported in several other cohort studies [18–20]. The same pattern was confirmed in a pooled analysis of 14 case-control studies, including data on 2247 female cases of thyroid cancer [21]. A meta-analysis of 25 independent studies on the association between hormonal and reproductive factors and the risk of thyroid cancer by Cao et al. suggests that parity was associated with an increased risk of thyroid cancer and that parous women were more susceptible to thyroid cancer than nulliparous women [7]. The recent systematic review and pooled meta-analysis of 23 reports including 8860 patients by Zhu et al. demonstrated that ever giving a birth was significantly associated with an increased risk of developing thyroid cancer [8].

Regardless of the findings indicating that thyroid cancer risk shows no significant differences between parous and nulliparous women, thyroid cancer risk was reported to be higher in women with high parity. An increase in the risk with each additional childbirth was reported by Galanti et al. suggesting that this effect was mainly due to an excess risk in the groups of women with very high parity (four children or more) [22]. Similar results were presented by Braganza et al. in a cohort of 70,047 women, where a greater number of live births were associated with an increased risk of thyroid cancer [14]. These findings are in line with the results of two case-control studies showing that the risk of thyroid cancer was significantly affected by the number of pregnancies [23, 24]. The New Caledonia study showed a statistically significant 8% increase in the risk of thyroid cancer with each additional pregnancy and a twofold increase of this risk among women having eight or more pregnancies [25]. The incidence ratio of thyroid cancer was more than twice as high in Finnish women with at least ten childbirths [26]. In the French Polynesian study, the risk of thyroid cancer increased with the number of births compared with that for nulliparous women, while eight or more births almost doubled the risk [27]. A case-control study in Kuwait revealed that the risk for thyroid cancer increased approximately twofold in women experienced 11 or more pregnancies [28]. Our study showed that in women with four and five and more children, the risk of the thyroid cancer was also higher than in nulliparous women; however, the observed differences were not significant.

The methodological strengths of this study include a cohort design based on the follow-up of the entire population of Lithuania, a large sample size, a substantial number of women with thyroid cancer, and a reliable identification of the cases over the study period. In our study, 868,105 women were followed for 8.6 years for a total of 6,921,568.45 person years. During the study period, 1775 thyroid cancer cases were diagnosed. To our knowledge, this is the largest cohort study assessing the relationship between parity and thyroid cancer risk in the last two decades, the limitations of the study stem mainly from the retrospective nature of the cohort study; therefore, we could not take into account exposure to ionizing radiations during childhood or previous thyroid disorders such as goiter or to control for confoundig from other hormonal and reproductive factors which may have influenced the results (miscarriages and abortions, hormone use, stillbirths, etc.).

The new and reliable evidence on the importance of parity in the risk of thyroid cancer in Lithuania calls for a more in-depth research on potential mechanisms and confounders behind this relationship. The role of hormonal factors may explain the relation between parity, number of births, and risk of thyroid cancer, and although exact biological mechanisms of the cancer have not been completely established, plausible explanations have been suggested. Estrogens are sex steroids that play a key role in the regulation of growth, differentiation, and functioning of a broad range of target tissues, the male and female reproductive organs, mammary gland, musculoskeletal, cardiovascular, and the immune and central nervous system. They are involved in cellular processes such as growth and cell motility and in the functioning of the thyroid gland mainly at the puberty, early pregnancy, delivery, puerperium, and breastfeeding [29–32]. During early pregnancy in healthy women, estrogens and beta-human chorionic gonadotropin are elevated. There is a progressive increase in thyroid-stimulating hormone levels, and all of this may promote thyroid growth leading to the development of benign thyroid tumors and thus to increasing the risk of malignant transformation [32, 33]. Experimental studies indicate that estradiol, the most potent form of estrogen, stimulates proliferation of papillary thyroid cancer cells in vitro and is involved in the regulation of angiogenesis. It also affects the metastatic phenotype in normal, cancer thyroid cell, and cancer stem cell line enhancing adhesion, migration, and invasiveness of malignant thyroid cells [29, 33, 34].

Conclusion

Our findings are consistent with the hypothesis that parity might be associated with the increased risk of thyroid cancer among women. Findings from other studies also suggest that several factors related to effects of pregnancy may be particularly important in influencing thyroid cancer risk. Future studies are needed, however, in order to establish more precisely the magnitude of these associations and to examine more complex relationships using more detailed information on other reproductive factors.

Compliance with Ethical Standards

Conflict of Interest

The authors declare that they have no conflict of interest.