Ovarian follicle loss in humans and mice: lessons from statistical model comparison

Menopause is triggered by the number of ovarian follicles falling below a threshold number and is irreversible because oogonial stem cells disappear after birth. Since it is the result of programmed disappearance of a limited store of follicles, menopause can be predicted using mathematical models based on total follicle counts at different ages. Our model shows follicle numbers decline bi-exponentially rather than as a simple exponential function of age, as had been assumed, with a first exponential rate parameter of −0.097 and a second of −0.237. The change occurred when numbers had fallen to the critical figure of 25 000 at age 37.5 years. The unexpectedly faster rate of ovarian ageing afterwards lowers the follicle population to 1000 at ∼51 years, and was adopted as the menopausal threshold because it corresponds to the median age of menopause in the general population. Had the earlier rate persisted menopause would not be expected until 71 years. The impact of step reductions of follicle numbers on the prospective span of menstrual life was predicted by the model. A reduction by 50% before age 30 years resulted in the threshold being reached at 44 years and 0.6 year later for every subsequent year until age 37.5 years after which it is reached at 48 years. A reduction of 90% in childhood before age 14 years could result in menopause as early as 27 years, with increments of 0.6 year per year afterwards until after 37.5 years when it is expected at age 41 years. The predictions are reassuring insofar as they confirm anecdotal evidence that long-term ovarian function is not substantially comprised by reducing as much as one-half of the mass.