Irregular Menses Linked to High Risk of Ovarian Cancer

Barbara A. Cohn, Ph.D. Director, Child Health and Development Studies A Project of the Public Health Institute Berkeley, CA Interview with:
Barbara A. Cohn, Ph.D.
Director, Child Health and Development Studies
A Project of the Public Health Institute
Berkeley, CA 94709 What are the main findings of the study?

  • Women with irregular menses had a statistically significant 2.4 fold increase in risk of death due to any form ovarian cancer, and a statistically significant 3-fold increase in risk of death due to late stage serous disease. Consistent with these findings, the incidence of late stage disease at diagnosis, and late stage serous cancer was increased about 2-fold in women with irregular menses.
  • Irregular menses was defined as irregular cycles (variation of 10 days or more) or infrequent cycles (>35 days) or history of annovulatory cycles identified during an in-person interview with women at an average age of 26 years or mentioned in their medical records. What are the strengths of the study?

  • Irregular menses was defined by women’s own report of usual cycle length when women were an average age of 26 years, during an in-person interview. Errors of recall would be minimal at that age and interviewers could probe for correct classification of irregular menses.
  • The study design was a prospective 50+ year follow-up of 14,403 pregnant women recruited from 1959-1967 to the Child Health and Development Studies (  There were 103 incident cases and 65 ovarian cancer deaths in this study.
  • We ruled out the contribution of infertility, the use of fertility drugs, or the use of birth control bills as explanations of study findings.  All women in this study had a live birth, and medical records recorded pharmaceuticals prescribed 6 months prior to pregnancy. Clomiphene was rarely used prior to 1967, nor were current methods of assisted reproduction in use in the 1960’s. During this period, birth control pill use was also infrequent, and pill formulations were very different from those currently in use.
  • Women with irregular menses constituted 13% of this large pregnancy cohort, and so findings are relevant to many women. Given the lack of information about risk factors, screening or biomarkers for ovarian cancer in young life, our findings offer an opportunity for prevention and for understanding the 90% of ovarian cancer cases that occur in women who do not have rare heritable germline mutations or family history in a first degree relative. What are the weaknesses of the study?

  • This study does not include infertile women (approximately 10 percent of all women) and therefore findings are not relevant to risk of ovarian cancer in infertile women.
  • There were some tumors with missing data on histology and stage, although there was no evidence that missing data was correlated with irregular menses.
  • Other than serous tumors, the most common form of ovarian cancer, the sample size for other tumor types was too small for study. However, there was a suggestive finding that risk of death due to endometroid tumors was also elevated (p=0.14).
  • We estimate that about 80% of women with irregular menses may have polycystic ovarian syndrome (PCOS).  However women without significant clinical symptoms (e.g. hirsutism, infertility, obesity) may never be diagnosed with PCOS, even today. Still we cannot determine with certainty whether it is predominantly women with PCOS who were at increased risk of ovarian cancer in our study. Were any of the findings unexpected?

Findings were unexpected. The “incessant ovulation hypothesis” predicts that women with less frequent cycles or more annovulatory cycles would be at lower risk of ovarian cancer. We observed the opposite.  What should clinicians and patients take away from your report?

  • Physicians may wish to obtain and record a history of usual menstrual cycle length; preferably during women’s early reproductive years, when recall is more accurate. Even though there are no known effective screening strategies for early detection of ovarian cancer, results of clinical trials in progress, or other advances could provide an opportunity to offer women with irregular cycles screening options in the future.
  • Physicians may wish to begin a conversation with women who have irregular menses about both the personal benefits and personal risks of oral contraceptives/hormonal contraception. There is good evidence that oral contraceptive use correlates with lower risk of ovarian cancer.  It is important to emphasize that there is no known effective screening strategy for ovarian cancer at this time, so this strategy may be a reasonable opportunity for prevention until there are alternatives. What recommendations do you have for future research as a result of this study?

  • It would be of interest to replicate our findings with studies of similar design:  prospective, with assessment of cycle length during the reproductive years when errors of classification are less likely. I am hopeful that our findings will stimulate interest in more research among those with access to such data.
  • Human pathology studies of tissues available from women with a history of irregular cycles (but not infertility) in comparison to controls, and similarly designed endocrine studies could yield clues about the etiology of ovarian cancer, and opportunities for prevention, early detection and treatment for the 90% of ovarian cancers that occur in women without evidence of heritable risk.
  • Experimental studies in appropriate animal models where phenotypes correlated to “long or irregular cycles” could be simulated would also advance understanding, similar to the strategy of investigating the functional consequences of BRCA1 mutations in animal models.


Abstract Presented at the AACR 2014  Irregular Menses and Ovarian Cancer