Abstract and Introduction
Abstract
The timing of natural menopause has implications for several health endpoints; in particular, it is a risk factor for breast cancer. The authors investigated factors influencing the timing of natural menopause among 95,704 women with a mean age of 59.7 years (10th-90th percentile range, 47.0-71.0) in five racial/ethnic groups in the Multiethnic Cohort Study, including non-Latina Whites, Japanese Americans, African Americans, Native Hawai'ians, and Latinas. The authors investigated whether race/ethnicity and several lifestyle and reproductive characteristics were associated with the timing of natural menopause. Race/ethnicity was a significant independent predictor of the timing of natural menopause. Other factors, including smoking, age at menarche, parity, and body mass index, did not significantly alter the race/ethnicity-specific hazard ratios. Relative to non-Latina Whites, natural menopause occurred earlier among Latinas (US-born Latinas: hazard ratio (HR) = 1.10, 95% confidence interval (CI): 1.07, 1.14; non-US-born Latinas: HR = 1.25, 95% CI: 1.21, 1.30) and later among Japanese Americans (HR = 0.93, 95% CI: 0.90, 0.95). These results support the hypothesis that the timing of natural menopause is driven by a combination of genetic, reproductive, and lifestyle factors.
Introduction
The timing of natural menopause is an important risk factor for breast cancer. Women who experience later natural menopause have higher rates of breast cancer than do women with earlier natural menopause. In 1972, Trichopoulos et al. reported that women who had experienced natural menopause at age 55 years or older had a twofold greater risk of breast cancer than women who had experienced natural menopause before age 45 years. In 1983, Pike et al. proposed a breast cancer risk model that reflected the exposure of breast tissue to hormones over the life span; according to this model, the age-related increase in breast cancer risk begins to decelerate distinctly at menopause. This age-incidence curve and the role of the timing of natural menopause as a breast cancer risk factor have since been firmly established (see reviews by Bernstein and Hankinson et al.).
The biologic mechanism that accounts for the association between the relative timing of menopause and breast cancer risk is hypothesized to be hormonal. Circulating levels of estradiol and progesterone, hormones produced by the ovary during ovulatory menstrual cycles which have been associated with breast cancer risk, decline markedly during the menopausal transition. The breast tissue of a woman who experiences menopause later will have greater cumulative exposure to these hormones than the breast tissue of a woman who experiences menopause earlier.
Lifestyle and reproductive factors such as smoking and parity have been associated with the timing of natural menopause; however, the variance in age at natural menopause that is explained by such factors is estimated to be low, less than 10 percent. This finding, in combination with observational studies implicating familial factors in the timing of natural menopause and twin studies showing significant heritability for age at natural menopause, has supported the hypothesis that the timing of natural menopause is a heritable trait.
Despite a growing body of evidence showing that the timing of natural menopause is likely to be a heritable trait, the relative timing of natural menopause across racial/ethnic groups has not been established. Some investigators have reported that American women of African ancestry experience natural menopause earlier than American non-Latina Whites; others have reported that American women of Asian ancestry experience natural menopause later than American non-Latina Whites. The Multiethnic Cohort Study is a population-based study of non-Latina Whites, Latinas, African Americans, Native Hawai'ians, and Japanese Americans in Hawai'i and Los Angeles, California, that provides an excellent resource for examining the effect of race/ethnicity in the timing of natural menopause while controlling for a variety of other factors.