What’s Normal?
The Important Issue of Cycle Length
Jerilynn C. Prior BA, MD, FRCPC
A recent app-based assessment of pre- versus post-COVID-19 vaccine menstrual cycle lengths 1 reminded me that we are, again, facing a fundamental issue in women’s reproductive health. The issue is: What is a normal cycle length?
The Scientific Method requires us to observe and accurately document and record what is occurring. This becomes what is accepted, reproducible and therefore “normal” plus the usual variability (range) around that normal value. The definition of “normal” may change over time. This depends on the population studied, the geographical location, and culture and economic characteristics. Those granted the status of “scientist” or “expert” in the culture are allowed to declare what is normal. However, there is a growing trend to include the public and patients in all aspects of research 2. Therefore, it is important that persons who have a lived experience of a condition or disease help to decide what is normal.
Here’s an example of change in what is “normal.” In the early 1990s a once-in-3-months blood test for glycosylated hemoglobin (HbA1c) was discovered. It tells us the percentage of time we have had high blood sugar levels. Initially, anything 7% or lower was normal (i.e. no diagnosis of diabetes mellitus). Now a HbA1c higher than 6.4% is considered abnormal. Although this change was potentially to prevent diabetes by identifying risks earlier, it may also advance profitable enterprises. Today what is considered “normal” changes through the process of clinical practice guidelines that are written by experts and reviewed by their peers before publication.
The definition of a normal cycle length was 21-36 days long in the 1980s 3. Normal cycle length has been 21-35 days for many years, but when that happened is not clear. What I do know is that the definition of a normal cycle length as 21-35 days was based on good evidence. That was from a random population of women in Copenhagen, Denmark, aged 15-44 years, who recorded and reported 1-year of records per woman taking into account age, different amounts of cycle record, and whether women were well off and educated or not.4 Almost 4,000 women were invited and nearly 80% agreed; a third gave researchers their 1-year menstrual cycle record for 1988, another third had kept the record and recalled what they had found, and the rest relied mostly on memory 4. For all cycles that year, only half a percent had a usual length less than 21 days; less than one percent had a cycle longer than 35 days. Almost 90% had regular cycles but 13% of these reported one cycle longer or shorter by 15 days during that year 4.
The International Federation of Gynecology and Obstetrics (FIGO), in 2007, under the leadership of Professor Ian Fraser, of the University of Sydney, one of the most important clinician-scientists in women’s reproduction, a man whom I know and respect, began the process leading to cycle length change 5. Dr. Malcolm Munro is also a key player in this FIGO process. He is a physician originally from Vancouver who started the Osteoporosis Clinic at the BC Women’s Health Centre that I later led.
The intent of these FIGO leaders was a good one—to simplify language about menstrual flow from old-fashioned Latin such as “menometrorrhagia” to words meaning the same thing, “irregular, heavy flow” 5. FIGO used a consultative, interactive committee process with representative experts from many countries, and specified that the new normal cycle length was 24 to 38 days long 6. If the cycle were shorter than 24 days, they called it “frequent bleeding”. If the cycle were longer than 38 days, they called it “infrequent bleeding”.6 Finally, if the flow lasted longer than 8 days they called it “prolonged bleeding” 6.
I puzzled over why these terms, “frequent” and “infrequent” flow, that are simple and easily understood, just did not feel right to me. What I now see is that this language focuses on the bleeding, rather than on what I believe is most important, the particular menstrual cycle. Instead, I think the new and simpler terminology should call cycles less than 24 or 21 days a “short cycle” and longer than 35 or 38 days a “long cycle.” Why is that important? Because, each cycle is a unique, complexly regulated, ovarian-endocrine phenomenon created by a different follicle with its individual egg, with characteristics adjusted by hypothalamic (brain) feedback control to be appropriate for that individual woman’s immediate situation.
I agree with simplification of language around women’s menstrual cycles; plain, unambiguous language has long been better than “doctor talk”. However, these new terms have been created by experts, most of whom have no lived experience of menstruation, since only one of the three authors of the 2018 peer-reviewed publication is a woman 6. I also have not found evidence that representative community women have been involved in this decision-making process. Both of the normal cycle length ranges are very English- and White-centric; I wonder whether they will be accurate for women of different races/ethnicities and when translated into other languages. It also bothered me that the recent paper refers to the women experiencing/reporting the duration of flow as “patients” (Figure 1 6). The term “patient” implies a person who has a disease and who is dependent on the skills and knowledge of experts/doctors.
I do not agree with changing the normal cycle length. Why?
It must be based on good evidence collected over a full year, in non-biased groups of women, about whom we also know other important things: age, education and how well off they are, race or ethnicity, and body mass index (BMI, weight in kg divided by height in meters squared) 7. The FIGO Menstrual Flow group refers to two papers supporting the new definition 8 9 and neither is a random sample nor has all the information we now know that we need to accurately describe cycle lengths to create a new normal. (They also reference the 1983 book, Patterns and Perceptions of Menstruation by the World Health Organization [WHO], eds R. Snowden, B. Christian, that provides regional menstruation feelings and attitudes, not cycle lengths.) The first cycle length study started in 1934 by asking first year women students at the University of Minnesota to keep 1-year cycle length records as the PhD research of Esther Doerr 8. (I have always known these data as the “Treloar Study” after the last name of her supervisor! Dr. Esther Doerr, never appeared as an author on the subsequent publications that used her data.) This was an important and unique study, especially for its time, but it was in a highly unrepresentative group of women. They were likely mostly or all White, usually well off and highly intelligent women who were some of the few who were able to attend university in the 1930s, just after the Depression. The other cycle length study is from Switzerland by Dr. Rudolph Vollman (with extensive support from his wife, Ursula) that began when he asked his patients to record their cycles 9. These important data add new information about ovulation as well as cycle lengths. (Vollman also pioneered a quantitative way of assessing basal temperature data to determine egg release/ovulation 9 that we have since validated 10). Starting with patients and their friends and later daughters makes for knowledge that is not appropriate for many women. And the Vollmans never reported average cycle lengths over time for each woman, but only for all women sharing an age or other characteristic 9. Likely few non-White women participated in either of these two studies of cycle length that FIGO references.
Cycle lengths are most reliably obtained with records that average over a full year. We learned this from a study over 1-year (prospective) but without information about how the 400+ women were recruited, or anything about their race/ethnicity or BMI11. We also know from about 400 women ages 24-45 in population-based data from the Michigan Women’s Bone Health Study, that a higher BMI is statistically associated with longer cycle lengths 7. Thus, any new reference standard should be obtained from women of a normal BMI (that is 18.5-24.9 for White women or 18.5-22.9 for women from the Asia-Pacific region).
Does the normal cycle length matter? For a start, 5-year menstrual cycle length records in 900 women ages 28-32 from the Doerr/Treloar cohort showed that those women whose cycle length averaged 31 days or longer (compared to 26-30 days) had more broken bones when they were menopausal 12. In the 500 or so premenopausal women in the population-based Canadian Multicentre Osteoporosis Study (CaMos) lower spinal bone mineral density was significantly related to both absent cycles for a year (amenorrhea) and to androgen excess (by physician-treated increased pimples or unwanted facial hair) 13. There are also complex relationships between cycle lengths and cardiovascular health; both overly short and too long cycles may be associated with a higher risk for strokes and heart attacks, including, in those with longer cycles, the risk of polycystic ovary syndrome (PCOS) 14.
In summary, it is important that we apply Scientific Principles when changing what is considered normal. We need to engage participants with lived experiences, study a year or more of records in large, well-described and representative groups of women of varying ethnicities and BMI values as well as ages. CeMCOR also believes it is better to describe cycles less than 21 days long as “shorter” and those that are longer than 35 days as “longer” cycles. It is the menstrual cycle in all its variability, not menstrual bleeding, that is important for women’s health.
Reference List
1. Edelman A, Boniface ER, Benhar E, et al. Association Between Menstrual Cycle Length and Coronavirus Disease 2019 (COVID-19) Vaccination: A U.S. Cohort. Obstet Gynecol 2022 doi: 10.1097/AOG.0000000000004695 [published Online First: 2022/01/07]
2. Skovlund PC, Nielsen BK, Thaysen HV, et al. The impact of patient involvement in research: a case study of the planning, conduct and dissemination of a clinical, controlled trial. Res Involv Engagem 2020;6(43).
3. Abraham GE. The normal menstrual cycle. In: Givens JR, ed. Endocrine causes of menstrual disorders. Chicago: Year Book Medical Publishers, Inc 1978:15-44.
4. Munster K, Schmidt L, Helm P. Length and variation in the menstrual cycle--a cross-sectional study from a Danish county. British Journal Obstetrics Gynaecology 1992;99(5):422-29.
5. Fraser IS, Critchley HO, Munro MG, et al. Can we achieve international agreement on terminologies and definitions used to describe abnormalities of menstrual bleeding? Human Reproduction 2007;22(3):635-43.
6. Munro MG, Critchley HOD, Fraser IS, et al. The two FIGO systems for normal and abnormal uterine bleeding symptoms and classification of causes of abnormal uterine bleeding in the reproductive years: 2018 revisions. Int J Gynaecol Obstet 2018;143(3):393-408. doi: 10.1002/ijgo.12666 [published Online First: 2018/09/11]
7. Symons JP, Sowers MF, Harlow SD. Relationship of body composition measures and menstrual cycle length. AnnHumBiol 1997;24(2):107-16.
8. Treloar AE, Boyton RE, Behn BG, et al. Variations of the human menstrual cycle through reproductive life. International Journal of Fertility 1967;12:77-126.
9. Vollman RF. The menstrual cycle. In: Friedman EA, ed. Major Problems in Obstetrics and Gynecology, Vol 7. Toronto: W.B. Saunders Company 1977:11-193.
10. Prior JC, Vigna YM, Schulzer M, et al. Determination of luteal phase length by quantitative basal temperature methods: validation against the midcycle LH peak. Clinical & Investigative Medicine 1990;13:123-31.
11. Marshall J. PREDICTING LENGTH OF MENSTRUAL CYCLE. Lancet 1965;191:263-65.
12. Cooper GS, Sandler DP. Long-term effects of reproductive-age menstrual cycle patterns on peri- and postmenopausal fracture risk. American Journal Epidemiology 1997;145(9):804-09.
13. Goshtasebi A, Berger C, Barr SI, et al. Adult Premenopausal Bone Health Related to Reproductive Characteristics-Population-Based Data from the Canadian Multicentre Osteoporosis Study (CaMos). Int J Env Res Pub He 2018;15(5)
14. Bozdag G, Mumusoglu S, Zengin D, et al. The prevalence and phenotypic features of polycystic ovary syndrome: a systematic review and meta-analysis. Hum Reprod 2016;31(12):2841-55. doi: 10.1093/humrep/dew218 [published Online First: 2016/09/25]
