Background
It is known that the percentage of pregnancies resulting from single episodes of forced penile-vaginal intercourse (rape) is significantly higher (8.0% in a sample of 405 women from a national random-digit dialing sample of households in USA) than the percentage of pregnancies resulting from single episodes of consensual, unprotected intercourse (3.1% in a sample of 221 women with no fertility problems planning to become pregnant in USA) [
1]. It is worth mentioning that data from the study by Gottschall and Gottschall [
1] were adjusted for the use of oral contraception and intra-uterine devices (IUDs). Furthermore, Gottschall and Gottschall [
1] elegantly ruled out the possibility that in their study higher rape-pregnancy rates may result from (1) women being more likely to report to medical or law-enforcement authorities rapes resulting in conception; (2) women sometimes attributing paternity to rapists when they were fertilized by a consensual partner; and (3) a high number of rape victims coming from the most fecund age cohorts of the population, i.e. rapists disproportionately would target young women in their most fecund years. These points were the main grounds used in previous studies to reject the fact that per-incident rape-pregnancy risk had been reiteratively reported to be higher than per-incident consensual pregnancy risk.
It is also known that, in fertile women planning to become pregnant, ovulation and conception may occur on any day of the menstrual cycle, although the maximum probability is reached at the middle of the cycle [
2]. Furthermore, ovulations and conceptions may arise during periods of amenorrhea associated with oral contraceptive use, drug addiction, pregnancy (superfetation; for review, see Pape et al. [
3]) and lactation. In addition, fertilization of ≥ 2 oocytes from the same menstrual cycle by sperm from separate acts of sexual intercourse has been also reported (superfecundation [
4,
5]).
All this epidemiological evidence, together with the fact that copulation can trigger or hasten (facilitate) ovulation in otherwise spontaneous ovulating species such as the rat (for reviews, see Gibson et al. [
6], Milligan [
7], Bakker and Baum [
8] and Nagy et al. [
9]), led Zarrow et al. [
10] and Jöchle [
11,
12] to propose that women may be facultative coitus-induced ovulators.
This attractive and stimulating hypothesis is supported by the fact that 68% of women display 2 follicular waves and the remaining 32% exhibit 3 waves of ovarian follicular development during an interovulatory interval [
13]. It is likely that women exhibit waves of follicular development during pregnancy and lactation as it occurs in cattle, sheep, goats and mares [
14] (for review, see Evans [
15]).
The presence of multiple follicular waves in women may provide an extra source of oocytes to be ovulated if an appropriate non-midcycle luteinizing hormone (LH) surge took place. We should note that follicular waves can be either ovulatory (the final wave of follicular development) or anovulatory (all the preceding waves) [
13,
16]. Twenty one percent of the anovulatory follicular waves are major, i.e. those in which one follicle is selected to be dominant over other follicles of the wave, and the remaining 79% minor, i.e. those in which no selection of a dominant follicle is evidenced. Although, anovulatory follicles do not develop as large as ovulatory follicles, they grow to an ostensible preovulatory diameter before undergoing atresia (maximum diameter of the dominant follicle of an anovulatory major wave: 12.9 ± 0.8 mm, range 10-17 mm; maximum diameter of the largest follicle of an anovulatory minor wave: 8.1 ± 0.1 mm, range 7-9 mm) [
17]. The maximum diameters reached by dominant follicles from anovulatory major waves (range 10-17 mm) are compatible with potential ovulations if follicles are correctly stimulated. In fact, spontaneous ovulations in normal menstrual cycles have been reported to occur in follicles ≥ 15 mm in diameter [
16].
Despite the fact that women display multiple waves of follicular development during an interovulatory interval, and likely during pregnancy and lactation, the hypothesis that women may be coitus-induced ovulators is directly refuted by studies showing that neither coitus nor orgasm induces a surge in LH secretion (and thus ovulation) in women [
18‐
20] and rhesus monkeys [
21]. Results from these classical studies have been recently endorsed by Baerwald et al. [
13,
16]. These authors tracked daily the follicle diameter and follicle number present in the ovaries of 50 healthy women of reproductive age (28.0 ± 6.9, range 19-43 years) displaying clinically normal menstrual cycles and not taking medications known to interfere with reproductive function. The ovarian ultrasonographic examinations only evidenced ovulations in follicles from the last wave of the interovulatory cycle, which emerges at the early follicular phase. The preceding waves, emerging at the luteal phase, were anovulatory in all 50 women entered into the study. This fact contradicts the assumption that coitus may induce ovulation in women. We should bear in mind that, in addition to a periovulatory peak, human beings display no changes at all or even rises in male- and female-initiated sexual activity, woman's sexual desire, autosexual activity and sexual arousability, and interpersonal sexual activities, including sexual intercourse, during the mid-follicular and late-luteal phases (for review, see Tarín and Gómez-Piquer [
22]). Thus, it is expected that a non-negligible number of the 50 women analyzed by Baerwald et al. [
13,
16] was presumably engaged in sexual intercourse during the period of ultrasonographic evaluations (one interovulatory cycle).
If we consider the laboratory rat as a paradigm of facultative coitus-induced ovulation, it can be noted that there are notable differences between women and rats in the neuroendocrine mechanisms controlling the ovarian cycle. For instance, in contrast to women, both spontaneous and induced ovulatory mechanisms are integrated in the rat. Female rats have no functional corpora lutea and must receive vulval, vaginal and/or cervical intromissive stimulation in order for the ovaries to develop fully functional corpora lutea. Mating stimulates the release of prolactin from the anterior pituitary, which is required for activation of the corpora lutea and progesterone biosynthesis (for review, see Bakker and Baum [
8]). Furthermore, although technical difficulties have precluded determining the ovarian dynamics in rats by transcutaneous ultrasound bio-microscopy [
23], it is expected that rats exhibit a single wave of follicular development such as evidenced in the mouse [
23] and the coitus-induced ovulating species analyzed as of today, including cats, llamas and camels (for review, see Evans [
15]).
Interaction of the hypothalamic-pituitary-adrenal (HPA) and hypothalamic-pituitary-gonadal (HPG) axes
While the woman's reproductive physiology does not fit the rat paradigm of facultative coitus-induced ovulation, the HPA and HPG axes interact with each other in a similar way in both species. Indeed, in rats and women, there is a positive coupling between the HPA axis and the HPG axis at the time of the preovulatory LH surge. In women, initiation of the LH surge takes place either at 04.00 a.m. in 20% of the cases or at 08.00 a.m. in the remaining 80% of the cases. The onset of the LH surge is strongly coupled with the time at which the peak (acrophase) of the cortisol circadian rhythm occurs, i.e. maximal cortisol plasma concentrations take place at 04.00 a.m. when the LH preovulatory surge initiates at 04.00 a.m. and at 08.00 a.m. when the LH preovulatory surge starts at 08.00 a.m. [
24]. Likewise, in the laboratory rat, both the preovulatory LH surge and the maximal plasma values of the corticosterone circadian rhythm are observed during the transition from the light to the dark phase (2 h preceding and 2 h following) on the day of proestrus [
25,
26].
Moreover, it is well-known that prolonged or chronic stress in rats and women may block, inhibit or delay the preovulatory LH surge and therefore disrupt the estrous or menstrual cycle (for reviews, see Rivier and Rivest [
27] and Kalantaridou et al. [
28]). On the contrary, women and rats exposed to acute stress may respond with an adrenal-progesterone-induced LH surge (for review, see Mahesh and Brann [
29]).
There is considerable evidence in rats and women showing an elevation in the levels of serum progesterone mainly from the adrenal glands prior to the onset of the LH surge. This increase in progesterone levels serves to initiate, synchronize, potentiate and limit the preovulatory LH surge to a single day (for review, see Mahesh and Brann [
29]). In order for progesterone to exert its facilitatory role on gonadotropin secretion, the presence of a background of high estradiol is essential [
30]. This is due to the fact that estradiol induces the expression of anterior pituitary, hypothalamic and extrahypothalamic progesterone receptors, which function as transcriptional regulators that prompt alterations in gene expression needed for facilitation of neurosecretion of gonadotropin-releasing hormone (GnRH) surges and release of periovulatory gonadotropin surges (for reviews, see Mahesh and Brann [
29] and Levine et al. [
31]).
It is not surprising, therefore, that progesterone and natural mineralocorticoids, such as deoxycorticosterone, and synthetic glucocorticoids, such as triamcinolone acetonide, which possess "progestin-like" activity, stimulate the release of LH and follicle-stimulating hormone (FSH) when administered acutely to pregnant mare serum gonadotropin (PMSG)-primed immature [
32] or estrogen-primed ovariectomized immature rats [
33]. In estrogen-treated menopausal women, it has been demonstrated that progesterone administration is able to induce a pre-ovulatory-type surge of LH and FSH (for review, see Mahesh and Brann [
29]. These findings have led to propose that the improvement in menstrual rhythm and ovulatory activity following glucocorticoid therapy in women suffering from polycystic ovarian syndrome or other syndromes of androgen excess may be due to its direct effects on the release of gonadotropins, in addition to the ability of glucocorticoids to suppress adrenal overproduction of androgens [
32,
33].
Moreover, it has been reported that a single injection of adrenocorticotropic hormone (ACTH) to estrogen-primed intact and ovariectomized immature rats causes a significant elevation in serum LH and FSH levels. However, this ACTH treatment fails to induce a surge of gonadotropins in non-estrogen-primed intact immature rats. It is worth mentioning that of the 2 major adrenal steroids secreted as result of ACTH administration, i.e. progesterone and corticosterone, only progesterone is able to stimulate LH release in estrogen-primed ovariectomized immature rats on the day of its administration [
34]. Likewise, a 3-h intravenous infusion of ACTH to postmenopausal or ovariectomized women with estrogen replacement raises the plasma levels of corticosterone and progesterone 3-4 h after injection, accompanied by a significant stimulation of LH release 2-3 h after the initial raise in progesterone [
35].
Aim of the study
The aim of this bioessay is to gather information either supporting or rejecting the hypothesis that acute stress may induce ovulation in women. The formulation of this hypothesis is based on 2 facts: 1) estrogen-primed postmenopausal or ovariectomized women display an adrenal-progesterone-induced ovulatory-like LH surge in response to exogenous ACTH administration; and 2) women display multiple follicular waves during an interovulatory interval, and likely during pregnancy and lactation. Thus, acute stress may induce ovulation in women displaying appropriate serum levels of estradiol and one or more follicles large enough to respond to a non-midcycle LH surge.
Summary conclusions
The studies reviewed in this bioessay indicate that the ovarian cycle and, in particular, the female's estradiol background modulates the response of the HPA and HPG axes to acute stressors. The pattern of response, however, differs between the HPA and HPG axes. In fact, although the highest responses of the HPA axis within the ovarian cycle are observed when females display relatively high plasma levels of estradiol (proestrus morning in rats, mid-follicular phase in rhesus monkeys and mid-follicular and mid- and late-luteal phases in women), the HPA axis exhibits positive responses in practically all phases of the ovarian cycle. In contrast, it seems that there is only one specific period of time within the ovarian cycle during which the HPG axis response is possible. In particular, positive responses of the HPG axis are found under relatively high plasma levels of estradiol on the morning of proestrus in rats, during the mid-follicular and mid-luteal phase in rhesus monkeys and during the mid-follicular phase in women. This conclusion is endorsed by the fact that the HPG axis of estrogen-treated ovariectomized females and perimenopausal or menopausal women also exhibit a positive response to acute stressors.
Two studies [
66,
72] in the rat show that several types of acute stress may exert different effects on pituitary LH release and that the steroid environment may modulate in a different way (inhibiting or stimulating) the pattern of response of the HPG axis induced by acute stressors. In rhesus monkeys and women, the steroid environment may also modulate in a different way the pattern of response of the HPA axis (and therefore the response of HPG axis including release of LH) elicited by acute stressors. In particular, in estrogen-treated ovariectomized females and perimenopausal or menopausal women, some ranges of plasma estradiol concentration seem to prevent whereas others allow a response of the HPA axis to acute stressors.
In women, the pattern of LH release elicited by acute stressors may vary from small non-ovulatory rises in plasma concentrations and/or slight changes in pulse characteristics to the typical ovulatory surge that spontaneously takes place at the middle of the menstrual cycle. As women present waves of follicular development during an interovulatory interval (and likely during pregnancy and lactation), they may be induced to ovulate at any point of the menstrual cycle or even during periods of amenorrhea associated with pregnancy and lactation if exposed to an appropriate acute stressor under a right estradiol environment. We should bear in mind that the maximum diameters (range of 10-17 mm) of the dominant follicles from anovulatory major waves, which account for 21% of the anovulatory follicular waves, are compatible with potential ovulations if follicles were correctly stimulated by an ovulatory-like LH surge. Follicles large enough (≥ 15 mm of diameter) to ovulate may be found in the ovaries of all the regular-cycling women at the late-mid follicular phase (growing antral follicles from the ovulatory follicular wave). In addition, they may be present in the ovaries of women exhibiting (1) a major-major 2-wave pattern at the mid-luteal phase (10% of the overall population of regular-cycling women); (2) a minor-major-major 3-wave pattern at the early-mid-follicular phase (6% of the overall population of regular-cycling women); and (3) a major-major-major 3-wave pattern at the late-mid-luteal and early-mid follicular phases (6% of the overall population of regular-cycling women) [
13]. We should note, however, that ovulation during the luteal phase or during pregnancy and lactation are extremely rare due to the presence of negative feedbacks that prevent the LH surge.
We should bear in mind that an acute-stress-induced surge of LH is shortly preceded by an elevation of serum progesterone from the adrenal glands. This fact suggests that such an elevation of serum progesterone may advance the secretory transformation of the endometrium resulting in embryo-endometrium asynchrony and consequently reduced chances of implantation and pregnancy if ovulation and fertilization took place. However, a recent systematic review and meta-analysis [
73] has not detected detrimental effects of elevated levels of progesterone on the day of human chorionic gonadotropin (hCG) administration on the probability of clinical pregnancy in women undergoing ovarian stimulation with GnRH analogues and gonadotropins for in vitro fertilization. Although this may not be the case for women treated with GnRH antagonists (for review, see Venetis et al. [
73]), we should take into account that an elevation of plasma progesterone on the day of hCG administration should be sustained in order to impair endometrial receptivity. This is not the case in estrogen-treated menopausal or ovariectomized women after being exposed to single intravenous injection of endotoxin [
35]. These women display a wave-like adrenal release of progesterone. In particular, plasma progesterone concentrations rise from ≈ 0.2 nmol/L 1-2 h after injection to ≈ 5 nmol/L 4-5 h after injection. Thereafter, plasma levels of progesterone drop to reach their baseline levels of ≈ 0.2 nmol/L 7 h after injection.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JJT has been involved in conception and design, acquisition, analysis and interpretation of data and drafting the article. TH and AC have been involved in analysis and interpretation of data, and revising the article critically for important intellectual content. All authors read and approved the final manuscript.