Academic journal article Demographic Research

Cohort Fertility Patterns and Breast Cancer Mortality among U.S. Women, 1948-2003

Academic journal article Demographic Research

Cohort Fertility Patterns and Breast Cancer Mortality among U.S. Women, 1948-2003

Article excerpt


Epidemiological research has shown that women who have early and numerous births have reduced risks of being diagnosed with breast cancer. We use U.S. Vital Statistics and Census data and age-period-cohort models to examine whether cohort fertility patterns are associated with breast cancer mortality rates among women aged 40 and older in 1948-2003. Cohorts marked by higher proportions childless at ages 15-24 and lower cumulative second birth rates at ages 15-29 have higher rates of breast cancer mortality. This is the first demonstration that cohort fertility patterns have left a clear imprint on trends in U.S. breast cancer mortality rates.

1. Introduction and background

Epidemiological studies using cohort and case-control designs have shown that reproductive factors - especially a woman's age at first birth and her total number of births - affect the risk of death from breast cancer. Kelsey, Gammon, and John (1993) provide an early review of the vast amount of literature on the topic. A great deal of subsequent research has been conducted and reviewed by Althius et al. (2004) and Merrill et al. (2005). Nearly all studies find that giving birth at younger ages and giving birth to numerous children are protective against breast cancer mortality. The accumulation of evidence led the National Cancer Institute's (2003) Workshop on Early Reproductive Events and Breast Cancer to conclude that the relationships between breast cancer and age at first birth and parity are "well established."

Ma et al. (2006) perform a meta-analysis and conclude that age at first birth and parity are significant factors in estrogen and progesterone receptor positive breast cancers but not in other breast cancers, suggesting that the hormonal changes associated with fertility play a strong role in breast cancer incidence (see also Althuis et al. 2004). Although the duration of breastfeeding is also inversely associated with all types of breast cancers (Collaborative Group on Hormonal Factors in Breast Cancer 2002), it does not account for the relationship between early or numerous births and breast cancer. Further, the relationship between breast cancer and reproductive factors is stronger among older (e.g., postmenopausal) women, because pregnancy is a short-term risk factor for breast cancer (Kelsey, Gammon, and John 1993; Ursin et al. 2004).

The relationships between breast cancer and reproductive factors, combined with dramatic changes in cohort fertility among American women, should have left an imprint on U.S. breast cancer mortality patterns. One example of such a population-level imprint has been demonstrated among a half-million women in Norway. A 25% excess mortality from breast cancer among well-educated women in Norway was statistically reduced to a non-significant 8% in regressions by adjusting for the earlier childbearing among more poorly educated women (Strand et al. 2005).

In the United States, the main efforts to relate fertility trends to breast cancer mortality at the national level have been mounted by Tarone and Chu (2000) and Tarone, Chu, and Gaudette (1997). Tarone and Chu (2000) fit an age-period-cohort model to data for 1950-1995. Because of the collinearity among age, period, and cohort (i.e., the value of any one of the variables can be uniquely identified by knowledge of the other two), they employ a model that can only detect non-linear cohort and period effects (see also Tarone and Chu 1996). Although they cite reproductive factors to interpret cohort trends in breast cancer mortality, they do not include measures of fertility directly in their models, and the cohort patterns they identify are weakly related to fertility patterns. Their estimates show decreasing breast cancer mortality rates among baby boomers, whereas the lower fertility among baby boomers would predict increasing rates.

Given the persistent finding that the timing of early births and total parity are predictive of breast cancer incidence and mortality, it is reasonable to hypothesize that those factors shape U. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.