Cancer Science
  • Open Access

Higher parity and earlier age at first birth are associated with lower risk of death from colon cancer

Authors


Abstract

This study was undertaken to examine whether there is an association between parity and age at first birth and risk of colon cancer. The study cohort consisted of 1 292 462 women who had a first and singleton childbirth between 1978 and 1987. We tracked each woman from the time of their first childbirth to December 31, 2009, and their vital status was ascertained by linking records with the computerized mortality database. We used the Cox proportional hazards model with time-dependent covariates to estimate the hazard ratios (HR) of death from colon cancer associated with parity and age at first birth. We limited eligible colon cancer deaths to those who were 45 years old or more to exclude possible heredity colon cancer cases, which usually occur at an early age. There were 670 colon cancer deaths during 34 980 246 person-years of follow-up. The colon cancer death rate was 1.96 cases per 100 000 person-years. The adjusted HR was 2.76 (95% CI = 1.60–4.75) for women who gave birth between 20 and 24 years and 7.35 (95% CI = 4.28–12.62) for women who gave birth after 24 years of age when compared with women who gave birth at younger than 20 years. A rising risk of colon cancer was seen with increasing age at first birth. The adjusted HR were 0.81 (95% CI = 0.65–1.02) among women with two live births, 0.93 (95% CI = 0.74–1.18) among women with three live births and 0.72 (95% CI = 0.51–1.00) for women with four or more births when compared with women who had given birth to only one child. The present study provides evidence that reproductive factors (parity and early age at first birth) may confer a protective effect on the risk of colon cancer. (Cancer Sci, doi: 10.1111/j.1349-7006.2012.02336.x, 2012)

In Taiwan, colorectal cancer is the third leading cause of cancer death for men and women.[1] In 2010, The age-adjusted death rate for colorectal cancer was 23.4 per 100 000 among men and 17.0 per 100 000 among women. There is substantial geographic variation in colorectal cancer death within the country.

Men tend to have a higher incidence of colon cancer than women of similar age.[2] An excess not only of some reproductive cancers but also of colon cancer was observed among nuns.[3] It has been hypothesized that sex hormones play a role in the protective effect against colon cancer.[4] The influence of sex hormones on colon cancer risk is supported by the presence of estrogen receptors in the normal colonic epithelium and neoplastic colonic tissue.[5-7] In an experimental study of rats, there is also a greater preponderance of colon cancer in men compared with women.[8] In addition, estradiol has been shown to inhibit chemically-induced murine colon carcinogenesis.[9] The recent Women's Health Initiative trial reported a connection between hormonal status and colorectal cancer risk, where postmenopausal hormone use (estrogen-plus-progestin) is associated with a 40% reduction in colorectal cancer risk,[10] further suggesting possible protective effects of estrogen and/or progesterone in the development of colorectal cancer. However, two prospective studies report a positive association between endogenous estrogen level and the risk of colorectal cancer in postmenopausal women.[11, 12] The findings from these observational studies, when considered together with the data on hormone therapy use and colorectal cancer, suggest that endogenous and exogenous sex hormones play different roles in colorectal tumorigenesis.[13]

The role of reproductive factors (parity and age at first birth) in the etiology of colon cancer in women has drawn interest in the published literature, but findings from previous prospective cohort studies have been inconsistent. Some studies describe inverse associations between increasing parity and the risk of colon cancer[14-16] and others report no association.[13, 17-25] Likewise, reported associations between late age at first birth and colon cancer risk have been inconsistent, with two prospective studies reporting decreased[13, 25] and another prospective study reporting increased risk.[20] Other published studies have failed to find any association.[14, 16-19, 21, 23, 24]

The majority of epidemiologic studies investigating the relationship between reproductive factors and colorectal cancer have been conducted in high-incidence populations in Europe and North America and study results have not been consistent. Incidence and death rates for female colorectal cancer have been increasing in some Asian countries, including Taiwan. The present study was carried out because only limited studies investigating the relationship between reproductive factors and colorectal cancer have been conducted in Asian countries.[16, 22, 25] Epidemiologic evidence suggests that cancers of the colon and rectum display similar but not identical etiologies and need to be examined as separate sites in analyses.[26] It is common to discuss relevant differences between these two tumors in terms of causation. For simplicity, however, in the present study, colon and rectal cancer are not compared in terms of their association with reproductive factors (parity and age at first birth). We studied a cohort of women who experienced a first and singleton childbirth between January 1, 1978 and December 31, 1987 to explore further the association between parity, age at first birth and the risk of colon cancer in Taiwan.

Materials and Methods

Data source

In Taiwan, all births are compulsorily reported to the Taiwan Birth Registry within 15 days of delivery. The Birth Registration System, which is managed by the Department of the Interior, has released computerized data on live births since 1978. The birth registration form, which requests information on maternal age, education, parity, gestational age, date of delivery, infant gender and birth weight, is completed by the physician attending the delivery. Because most deliveries in Taiwan take place in either a hospital or clinic, the birth certificates are completed by physicians attending the delivery and it is mandatory to register all live births at local household registration offices, the birth registration data are considered complete, reliable and accurate.[27]

Study population

The study cohort consisted of all women with a record of a first and singleton childbirth in the Birth Registry between January 1, 1978 and December 31, 1987. Over this period, there were 1 333 312 first and singleton births in Taiwan. Information on any subsequent births was also retrieved from the Birth Registry. Of the 1 399 312 primiparous women, 106 850 subjects were excluded because data were missing on at least one variable, such as maternal age (n = 100 099), years of schooling (n = 382), marital status (n = 2665) or birth place (n = 4535). This left 1 292 462 subjects with complete information for the analysis. Their details are described in an earlier publication.[28]

Follow-up

Each woman has her own unique personal identification number. Using this number, we tracked each woman from the time of their first childbirth to December 31, 2009, and their vital status was ascertained by linking records with the computerized mortality database, identifying the date of any deaths occurring in this cohort. Of the 1 292 462 women followed, none had a missing personal identification number; therefore, all cohort members were followed up. Because it is mandatory to register death certificates at local household registration offices, the death statistics in Taiwan were considered to be highly accurate and complete.[28]

Statistics

The person-years of follow-up for each woman was calculated from the date of first childbirth to the date of death or December 31, 2009. Death rates were calculated by dividing the number of deaths from colon cancer by the number of person-years of follow-up. We limited eligible colon cancer deaths to those who were ≥45 years old to exclude possible heredity colon cancer cases, which usually occur at an early age. We used the Cox proportional hazards model with time-dependent covariates to estimate the hazard ratios (HR) of death from colon cancer associated with parity (the number of children recorded in the last childbirth record of each woman registered during follow-up) and age at first childbirth. Time-dependent indicator variables were created for parity = 2, parity = 3 and parity = 4 or more, with parity = 1 serving as the reference group. Thus, for example, for a woman entering the study with parity = 1, her baseline values for the indicators will all equal 0. If she has her second child at time = 10, her value for the indicator for parity = 2 will change to 1 and stay at 1 until she has her third child, at which time her parity = 2 indicator will change back to 0 and her parity = 3 indicator will change to 1. The 95% confidence intervals (CI) for the relative risk were also calculated. Colon cancer is defined according to the International Classification of Disease, Injury, and Causes of Death (9th revision) (ICD code 153). The variables in the final model included age at first childbirth (<20, 20–24 and >24 years), parity (1, 2, 3 and 4 or more), marital status (married or unmarried), years of schooling (≤9 or >9 years) and birth place (hospital/clinic or home/other). The proportional hazards assumption was assessed for all the abovementioned variables, and no violations were observed. Analyses were performed using the sas statistical package (version 8.02; SAS Institute). All statistical tests were twosided; P-values of less than 0.05 were considered to be statistically significant.

Results

Altogether, 1 292 462 primiparous women with complete information were included in the analysis. A total of 34 980 246 person-years were observed during the follow-up period from the time of their first childbirth to December 31, 2009. There were 670 colon cancer deaths, yielding a death rate of 1.92 cases per 100 000 person-years.

Table 1 presents the numbers of person-years of follow-up and colon cancer deaths by age at recruitment (age at first birth), parity, marital status, years of schooling and birth place. The death rate was 2.59 among women who had given birth to one child, 1.87 among those who had had two children, 1.91 among those who had had three children and 1.36 among those who had given birth to four or more children.

Table 1. Demographic characteristics of the study cohort
VariablesNumber of subjectsFollow-up person-yearsNumber of deaths from colon cancerMortality rate (per 100 000 person-years)
Age at recruitment (first birth)
<20 years87 6842 543 858.58140.55
20–24 years629 18817 204 144.832121.23
>24 years575 59015 232 242.584442.91
Parity
1157 2074 170 772.331082.59
2564 72715 124 112.332831.87
3436 25011 925 297.252281.91
4+134 2783 760 064.08511.36
Marital status
Married1 260 61534 115 479.256511.91
Not married31 847864 766.75192.20
Years of schooling
≤9 years722 51819 850 938.173541.78
>9 years569 94415 129 307.833162.09
Birth place
Hospital/clinic1 245 92533 638 862.836411.91
Home/other46 5371 341 383.17292.16

The multivariate-adjusted HR and 95% CI are shown in Table 2. Older age at first birth was associated with increased colon cancer risk. The adjusted HR was 2.76 (95% CI = 1.60–4.75) for women who gave birth between 20 and 24 years and 7.35 (95% CI = 4.28–12.62) for women who gave birth after 24 years when compared with women who gave birth younger than 20 years. A trend of increasing risk of colon cancer was seen with increasing age at first birth (P for trend <0.0001).

Table 2. Association between parity age at first birth and relative risk of death from colon cancer over a 32-year follow-up period
VariablesCrude RR (95% confidence interval)Multivariate-adjusted RR (95% confidence interval)a
  1. a

    Mutually adjusted. RR, relative risk.

Age at recruitment (first birth)
<20 years1.001.00
20–24 years2.74 (1.59–4.70)2.76 (1.60–4.75)
>24 years7.28 (4.27–12.39)7.35 (4.28–12.62)
 P < 0.0001 for linear trendP < 0.0001 for linear trend
Parity
11.001.00
20.72 (0.58–0.90)0.81 (0.65–1.02)
30.68 (0.54–0.86)0.93 (0.74–1.18)
4+0.45 (0.32–0.62)0.72 (0.51–1.00)
 P < 0.0001 for linear trendP = 0.3623 for linear trend
Marital status
Married1.001.00
Not married1.12 (0.71–1.76)1.20 (0.76–1.91)
Years of schooling
≤9 years1.001.00
>9 years1.34 (1.15–1.56)0.97 (0.83–1.14)
Birth place
Hospital/clinic1.001.00
Home/other0.91 (0.63–1.32)1.19 (0.81–1.73)

After adjustment for age at first birth, marital status, years of schooling and birth place, the adjusted HR were 0.81 (95% CI = 0.65–1.02) among women with two live births, 0.93 (95% CI = 0.74–1.18) among women with three live births and 0.72 (95% CI = 0.51–1.00) for women with four or more births when compared with women who had given birth to only one child.

Discussion

To our knowledge, this is the largest study (n = 1 292 462 women) published to date to examine the relationship between reproductive factors (parity and age at first birth) and colon cancer risk. In this prospective cohort study, we found a possible inverse association of parity and colon cancer. For example, compared with women who had given birth to only one child, women with four or more births had 0.72 times the risk of colon cancer. Our finding of a reduced risk of colon cancer being associated with higher parity is in agreement previous published studies.[14-16] However, other studies report no association with parity.[13, 17-25]

The mechanism by which increased parity may confer protection against the future development of colon cancer in women remains unclear. Pregnancy elevates serum estrogen levels approximately 100-fold.[29] Increasing parity is associated with an overall increase in lifetime estrogen exposure. There is experimental evidence that estrogen regulates the cell growth of colonic mucosa and inhibits cell proliferation of colonic tumors through binding to estrogen receptors.[5-7, 30] Estrogen has also been shown to reduce bile acids, to decrease the growth enhancement of insulin-like growth factors and to maintain the transcription and expression of estrogen and progesterone receptors.[31] Thus, if estrogens are associated with a reduced risk of colon cancer, we would expect pregnancy to offer some protection from colon cancer. Our data provide support for this hypothesis.

In the present prospective cohort study, we found that risk of colon cancer rose with increasing age at first birth after adjusting for parity. Our finding of a positive association between age at first birth and colon cancer risk is in accordance with two previous studies,[13, 20] but not with other studies that report no association with age at first birth.[14, 16-19, 21, 23, 24] The reasons for the positive association between age at first birth and colon cancer risk are unknown. It may be that a younger age at first birth played a protective role in colon cancer risk through elevated levels of certain hormones (including estrogen and progesterone) during pregnancy. Moreover, it has been found that age at menarche (and, therefore, exposure to period estrogen stimulation) is related to the age at which a women delivers her first child.[32] Our finding of an increased risk of colon cancer associated with older age at first birth is in keeping with the hypothesis that estrogen exposure is protective with respect to colon cancer risk. However, because, to date, there is only limited evidence for a positive association between age at first birth and risk of death from colon cancer, the possibility that this is a chance finding must considered. Clearly, more work is needed before the influence of age at first birth on the risk of colon cancer is understood.

It is notable that both nulliparity and later age at first birth are established risk factors for breast cancer.[33] These factors are believed to influence the risk of breast malignancy by increasing lifetime exposure to endogenous sex hormones, such as estrogens.[33, 34] Our study found that these same risk factors are associated with the risk of colon cancer, which suggests that similar endocrinologic mechanisms play a role in colon carcinogenesis.[13]

Mortality data have been widely used to generate epidemiologic hypotheses, despite their inherent limitations. The completeness and accuracy of the death registration system should be evaluated before any conclusions based on mortality analysis are made. In the event of a death in Taiwan, the deceased's family is required to obtain a death certificate from the hospital or local community clinic. The death certificate must then be submitted to the household registration office to cancel the deceased's household registration. The death certificate is required in order to have the deceased's body buried or cremated. Death certificates must be completed by physicians in Taiwan. It is also mandatory to register all deaths at local household registration offices; death registrations are accurate, reliable and complete. The complete population coverage and follow-up made possible by the national identification number means that the study is without selection bias. Information bias is also unlikely to be significant for parity and age at first birth.

Taiwan inaugurated its National Health Insurance (NHI) program in 1995 to finance health care for all citizens of Taiwan. As of 2007, the NHI program provided coverage for 22.60 million out of 22.96 million Taiwanese (coverage rate 98.4%).[35] It is believed that all colon cancer sufferers had access to medical care. The 5-year survival rate of colon cancer in Taiwan was approximately 54–58% in the 1990s.[36, 37] However, there is no reason to believe that there would be any correlation between parity, age at first birth and 5-year survival rate of colon cancer. Therefore, we believe that using mortality data rather than data on inpatient admissions would not have a substantive effect on the associations we observed.

Our data takes into account the effect of the number of children on the risk of death from colon cancer. The risk progressively declines with each additional birth. The birth registration system in Taiwan covers only live births. We were unable to examine the possible relationship of nulliparity to colon cancer risk. Therefore, the generalizability of our findings is limited. Further study in independent cohorts with the inclusion of non-parous women for follow-up studies of the association between parity and the risk of colon cancer is necessary.

Several studies have found that the use of oral contraceptives (OC) is associated with a reduced risk of colon cancer.[20, 23, 24, 38] Other studies, however, have failed to find statistically significant association between the use of OC and colon cancer.[14, 17, 19, 25] Published studies have largely reported a decreased risk between hormone replacement therapy (HRT) and colon cancer risk.[17, 19, 39, 40] We were unable to adjust for these two hormonal factors in the current study due to the lack of available data. The prevalence of OC use (approximately 10%) and HRT use (approximately 7.4%) are low in Taiwan compared with developed countries (approximately 59% and 31%, respectively).[41-45] The confounding effect resulting from these two factors should be small, if any exists at all. Furthermore, if the association between these two potential confounding variables and the risk of colon cancer is not as strong as that which has been observed for parity and age at first birth, adjustment of these variables will not qualitatively change the conclusion.

Physical activity and dietary habits are the most important risk factors for colon cancer.[26] There is no information available on these variables for individual study subjects and, therefore, they could not be adjusted for in the analysis. However, there is no reason to believe that there would be any correlation between these variables and parity and age at first birth.

In conclusion, we observed a possible inverse association of parity and colon cancer risk. In addition, we found that risk of colon cancer increased with increasing age at first birth. This study provides evidence that reproductive factors (parity and early age at first birth) may confer a protective effect on the risk of colon cancer.

Acknowledgments

This study was partly supported by a grant from the National Science Council, Executive Yuan, Taiwan (NSC-100-2314-B-037-023-MY2).

Disclosure Statement

The authors have no conflict of interest to declare.

Ancillary