Menstrual cycles may affect cancer risk.

Several reproductive factors have been linked to increased risk of breast cancer -- from early menses and late menopause to childlessness and late age at first pregnancy. A new study would add yet another: menstrual cycles that are typically either shorter or longer than average.

Five years ago, epidemiologist Elizabeth A. Whelan, then at the National Institute of Environmental Health Sciences (NIEHS) in Research Triangle Park, N.C., began mailing questionnaires to nearly 1,000 early entrants in the ongoing Menstruation and Reproductive History Study (MRHS).

This project had recruited most of its nearly 4,000 volunteers between 1934 and 1939 or 1960 and 1964. The women reported not only such information as age at first menstrual cycle, but also the dates of menstrual bleeding, hormone use, and medical conditions (including pregnancies).

Although epidemiological studies have linked lifetime estrogen exposures to breast cancer -- with higher exposures increasing risk -- Whelan and many others suspect that hormones in general (including estrogen and progesterone), and their peaks during an ovulatory cycle, collectively affect cancer risk.

For instance, breast cells divide more rapidly during the last 14 or so days of each menstrual cycle -- the lowerestrogen, luteal phase. Because cells become more susceptible to damage when they divide, a factor that could increase their vulnerability to cancerfostering changes, Whelan suspected shorter cycles might elevate breast cancer risk. Her reason? Even among women who remain fertile for the same number of years, those with short cycles undergo more luteal phases.

In the just-released Dec. 15, 1994 AMERICAN JOURNAL OF EPIDEMIOLOGY, she and her coworkers report that, compared to women with menstrual cycles lasting 26 to 29 days during the most stable period of their reproductive lives (age 25 to 29), those with shorter cycles face roughly twice the risk of breast cancer. This trend held even after accounting for such potentially confounding variables as age, family history of breast cancer, weight, and reproductive factors that boost the risk of breast malignancies.

Her team also found a near doubling of breast cancer risk in women whose cycles exceeded the average length. "Our first thought was that this was just a chance observation," says coauthor Dale P. Sandler of NIEHS, because long cycles would reduce the time a woman spends in the potentially high-risk, luteal phases.

But "we're now willing to stick our necks out and say that this [trend] looks like something real," Sandler says, because cycles in the 25- to 29-year-old group that her team focused on are the ones "least affected" by possibly confounding influences.

Whelan, now at the National Institute of Occupational Safety and Health in Cincinnati, concludes that women with extremes in cycle length probably possess "some hormonal disturbance that not only gives them a wacky cycle ... but also increases their risk of breast cancer." If true, however, entirely different disturbances probably contribute to the cancer risk of long versus short cycles, she says.

Breast cancer researcher Dimitrios Trichopoulos of the Harvard School of Public Health in Boston describes the new study as "very well done" and perhaps "more important than three-quarters of the ones I authored." But from a research perspective, he suspects it's "a dead end." Why? It leads one to conclude "only what we already know -- that hormones are important for breast cancer."