Reproductive biology
The Two Clocks Inside the Menstrual Cycle
Pick any menstrual calendar. Day one is the first day of bleeding. Somewhere in the middle, an ovary releases an egg: that moment is called ovulation. What follows it is an almost fixed countdown, about twelve days long. What comes before it has no fixed deadline at all. Across 612,613 real cycles, tracked scientifically, only 13% lasted exactly twenty-eight days, precisely because of this difference.
Source: Bull, Rowland, Scherwitzl et al., 2019, npj Digital Medicine: 612,613 cycles, from 124,648 people, with ovulation confirmed objectively through basal body temperature and luteinising-hormone testing.
Twenty-Eight Days Is Just a Rounded Average
The table below comes straight from the study. Researchers split the 612,613 cycles into five groups, by total length. For each group, they measured, separately, how long the part before ovulation lasted and how long the part after it lasted. The 25–30 day group is the most common: 65% of all cycles. But even inside it, exactly twenty-eight days is only one of many possible values.
| Cycle length | Of 612,613 cycles | Before ovulation (days) | After ovulation (days) | How the two parts split |
|---|---|---|---|---|
| 15–20 days | 0.6% | 10.4 | 8.0 | |
| 21–24 days | 8% | 12.4 | 11.0 | |
| 25–30 days | 65% | 15.2 | 12.4 | |
| 31–35 days | 19% | 19.5 | 12.9 | |
| 36–50 days | 7% | 26.8 | 12.9 |
The five groups cover 99.1% of the 612,613 cycles. The remaining 0.9% fell outside the study's tabulated 15–50 day range and aren't shown. A cycle of exactly twenty-eight days averaged 15.4 days before ovulation and 12.6 days after.
The Part That Doesn't Move
After ovulation, the released egg leaves behind a small, temporary structure: the corpus luteum. The corpus luteum produces progesterone, the hormone that prepares the uterus for a possible pregnancy. It runs on an almost fixed clock. Without a pregnancy signal, it shuts down around twelve days after ovulation. It doesn't matter whether ovulation came early or late that month: the countdown that follows it runs to nearly the same length for almost everyone. That's why the part after ovulation stays so tight in the data. Its average moves from 8.0 to 12.9 days across all five groups in the study, a shift of only 4.9 days. Its standard deviation stays between 2.1 and 2.8 days throughout.
The Part That Can Run Late
The part before ovulation has no clock that strict. Each month, the ovary starts with a batch of immature follicles: small sacs, each holding one unripe egg. Only one has to reach full maturity before ovulation can happen. How long that takes depends on age, on how many follicles remain in reserve, on weight, and on other factors that change the speed of the dominant follicle's growth. That's why the part before ovulation varies far more. Its average climbs from 10.4 to 26.8 days across the same five groups, a shift of 16.4 days. That's more than three times the shift on the other side. In practice, almost the entire difference between a short cycle and a long one is absorbed by this first part.
The Age Effect, Between Twenty-Five and Forty-Five
The same study also tracks how the two parts change with age, between twenty-five and forty-five. Total cycle length falls slightly with age: the observed decline over that span was 2.9 days in total, at an average slope of about 0.18 days per year. The part before ovulation falls at a similar pace: 3.2 days over the same span. The part after ovulation doesn't follow that slope. It stays essentially unchanged.
Even the same person doesn't repeat exactly the same cycle length from one month to the next. Average cycle-to-cycle variation for a given person was 2.6 days. It's a little higher at younger ages (2.9 days between eighteen and twenty-four) and a little lower toward the end of the reproductive span (2.4 days between forty and forty-five).
Who Was Tracked in This Study
The data come from people using a fertility-awareness app. Ovulation was confirmed objectively, through basal body temperature and luteinising-hormone testing. That's a more accurate method than a memory-based survey. The population, though, isn't a random sample of everyone who menstruates: they are reproductive-age adults who had already chosen to track their cycle with a digital tool. The study excludes adolescents at the start of their reproductive lives and people nearing menopause, periods when cycles often become irregular or anovulatory. The figures above describe a pattern observed across the whole studied population, calculated over hundreds of thousands of cycles. Any one person, in any one month, can easily fall outside these ranges without it meaning anything concerning. Large, repeated deviations are still worth discussing with a doctor.
Full source: Bull JR, Rowland SP, Scherwitzl EB, Scherwitzl R, Danielsson KG, Harper J. "Real-world menstrual cycle characteristics of more than 600,000 menstrual cycles." npj Digital Medicine 2, 83 (2019). Full text: pmc.ncbi.nlm.nih.gov/articles/PMC6710244.