The quality of ovulation is strained in normal women

Abstract

‘Ovulatory dysfunction’ bears the imprimatur of a succinct medical diagnosis, but what does it really mean and above all, how do you diagnose it, especially in women with a normal menstrual history? Although we have multiple measures in our modern toolbox: ultrasound for daily follicular monitoring, hormonal assays of serum, urine and saliva (some for instance LH available over the counter to our patients), and even, heaven forbid, an endometrial biopsy; these are expensive, invasive and labor intensive for both physician and patient to document, and often lacking in evidence-based cutoffs. The bottom line is that the most common clinical practice to screen for ovulatory dysfunction is a menstrual history. Much of our knowledge of the normal length of the menstrual cycle and the effects of age on the menstrual cycle comes from the classic studies of Treloar et al. (1967) which were based on menstrual diaries collected on the back of a postcard, beginning in young women recruited during their college years and then collected over their lifetimes. We are likely to assume that a woman is ovulating regularly if her menstrual cycle length is 21–35 days (thanks to Treloar). Therefore, the BioCycle study by Hambridge et al. (2013) reported in this issue of the journal offers important and novel data about ovulatory function in normal women, including the frequency of anovulation (12%), and the influence of anovulation upon ovulatory cycles (adversely in terms of integrated serum estrogen and progestin measures). The BioCycle study investigators prospectively enrolled a cohort of 259 healthy, menstruating women, 18 –44 years of age (excluding confounders such as hormonal use, history of ovulatory disorders, etc.). Women were asked to provide eight timed serum specimens during each of two menstrual cycles, and these were timed in relation to self-monitoring for ovulation using urinary LH and estrogen kits, to allow for comparison at similar phases of the cycle. The current study includes 250 women who provided (at least some) data from two menstrual cycles. A significant strength of the study is the retention and compliance in the protocol for such a large number of women. Over 94% of the women completed seven or more of the planned visits per cycle (far exceeding the average dropout in randomized clinical trials!) and all women had at least five visits per cycle. An additional strength is that the investigators studied more than one menstrual cycle in each subject, and carefully collected confounders of the menstrual cycle in the cohort. The investigators used a serum progesterone cut-off point (.5 ng/ml) to categorically determine whether ovulation took place in the cycle, analogous to our own clinical practice of documenting ovulation with a mid-luteal serum progesterone level. Overall 12% of the women with a normal menstrual history had at least one anovulatory cycle and of these 3% had two anovulatory cycles. Women with two ovulatory cycles were older (mean age of this group was 28 years) and women with anovulatory cycles were significantly younger (the mean age of women with one anovulatory cycle was 5 years younger and with two anovulatory cycles 8 years younger). Women with anovulatory cycles were also more likely, not surprisingly, to be unmarried and nulliparous, and anovulation was not associated with smoking, former OCP use or physical activity. However, the most important and novel observation of this study is the effect of anovulation on subsequent ovulation, and hence the title of this editorial. The investigators utilized complex modeling to construct the menstrual cycle (and also created some elegant figures) based on their repeated serum hormonal measures, and divided the subjects categorically on the basis of their ovulatory status. They found diminished integrated levels of serum sex steroids in the ovulatory cycles of women who also had one anovulatory cycle (and this was independent of the temporal relation of the ovulatory cycle to the anovulatory cycle), compared with those women who had two ovulatory cycles. In regard to integrated progesterone levels there was a dose– response stepdown in integrated progesterone levels with highest in the women with two ovulatory cycles, followed by those with one ovulatory/one anovulatory cycle followed by those with two anovulatory cycles. This may be (in the periodic ovulators) the clearest depiction of the long hypothesized luteal phase defect (Jones, 2008) that I have seen. While these hormone levels represent only a surrogate of oocyte quality or endometrial function, and ultimately fecundity, they offer evidence, that documentation of an anovulatory cycle in a women suggests altered and likely diminished ovulatory quality (with potential downstream endometrial effects) in these women when they do ovulate. In other words, the anovulatory