In rats with 5-day estrous cycles, the surgical stress of laparotomy on the morning of proestrus leads to release of adrenal estradiol and progesterone preceding an early surge of LH from the pituitary 38. Moreover, intact 4- and 5-day cycling rats exhibit increased HPA sensitivity, releasing higher amounts of ACTH and corticosterone into the blood stream, immediately after a 20-min restraint on the morning of proestrus than animals at either estrus or diestrus (combined diestrus-1, 2 and 3) 39. Notwithstanding, Carey et al. 40 found higher plasma levels of ACTH and corticosterone immediately after a 20-min restraint in 4-day cycling rats but did not evidence a significant effect of the estrous cycle on plasma concentrations of these hormones

In cycling female rhesus monkeys, a 30-min intracerebroventricular administration of interleukin-1α, which simulates an inflammatory/immune-like challenge, induces a higher increase in plasma levels of cortisol and progesterone if administered during the mid-follicular phase compared to the early-follicular phase 41.

Women with normal cycles undergoing bilateral ovariectomy plus total hysterectomy under general anesthesia during the mid- to late-follicular phase of the menstrual cycle respond with a small increase in plasma levels of progesterone 12 h after surgery preceding a small rise of LH. Likewise, after the surgical stress of cholecystectomy for benign conditions of the gallbladder under general anesthesia during the early- to mid-follicular phase, women exhibit a small surge-like increase in plasma levels of progesterone 12 h after the operation 42. These increases, however, are not observed in women undergoing bilateral ovariectomy plus total hysterectomy during the early- to mid-luteal phase 4243.

Other studies show that women in the mid-luteal phase of the menstrual cycle display enhanced plasma levels of ACTH in response to a 20-min progressive treadmill exercise compared to women in the early follicular phase 44. Mid-luteal phase women also display reduced sensitivity to glucocorticoid feedback (lower suppression of plasma cortisol in response to a low dose of synthetic glucocorticoid dexamethasone) and decreased glucocorticoid receptor (type II) mRNA expression in lymphocytes compared to women in the early follicular phase 45. In addition, the cortisol response to a 90-min submaximal bicycle 46 or treadmill 47 exercise is higher in the mid-luteal phase than in the early-follicular 47, mid-follicular 46 and late-follicular phases 47. Likewise, a lower adrenocortical reactivity (decreased plasma cortisol levels) to the psychological stress of self-evaluation has been evidenced during the ovulatory period (between 12-16 days from the onset of the previous menstruation) than during the premenstrual phase (between 1-3 days prior to the onset of the next menstruation) 48. There are, however, other studies that either do not find menstrual cycle-related differences in the HPA axis responsivity to intense and moderate physical exercise 495051 or to the psychological stress of remembering stressful situations in their lives and self-evaluation 52. Interestingly, although Galliven et al. 51 did not find a significant effect of the menstrual cycle on plasma cortisol levels after a 20-min progressive submaximal treadmill, the analysis of the net integrated area under the curve revealed a marginal (P = 0.056) lower response of cortisol in the periovulatory (between 10-16 days after the start of menses) phase compared with the early-mid-follicular (between 3-9 days after the start of menses) and mid-late-luteal (between 18-26 days after the start of menses) phases.

From the studies analyzed in this section, it seems that the ovarian cycle modulates the HPA axis response to acute stressors. In rats, it appears that proestrus morning is the more sensitive period. In rhesus monkeys, the HPA axis response is higher during the mid-follicular phase than during the early-follicular phase. And in women, the period more responsive seems to be during the mid-follicular and mid- and late-luteal phases. The absence of effect of the menstrual cycle on HPA axis responsivity to physical stress found in several studies may be explained by the presence of uncontrolled variables related to the training status of women and possibly to the training regimen followed in the days previous to the exercise test (cited by Williams et al. 53).

It is known that immediately after exposure of ovariectomized rats to a 20-min restraint stress the plasma levels of ACTH are higher in estradiol-treated rats compared to the oil-treated control and to estradiol plus progesterone-treated rats. Moreover, during the 20-min period of restraint, the plasma levels of ACTH and corticosterone are always higher in the estradiol-treated group than in the oil-treated control and estradiol plus progesterone-treated rats 39. Likewise, during the first 60 min after exposure to stress from a novel environment, ovariectomized estradiol- and estradiol plus progesterone-treated rats display an enhanced ACTH response compared to ovariectomized non-treated control and progesterone-treated rats. Ovariectomized estradiol plus progesterone-treated rats also display a higher corticosterone response than non-treated control and progesterone-treated rats during the same period of time 40.

Plasma levels of ACTH and corticosterone during the 2 h after a 5-sec footshock are higher in ovariectomized estradiol-treated rats than in ovariectomized non-treated control rats. At 20 min after 1-min exposure to ether vapors, plasma concentration of corticosterone is also higher in the group of estradiol-treated rats. Moreover, the ACTH and corticosterone secretory responses are less effectively suppressed by RU 28362 (a specific glucocorticoid receptor agonist) in estradiol-treated animals than in non-treated control rats 54. This finding endorses other studies showing an inhibitory effect of estradiol on glucocorticoid receptor-mediated negative feedback 55.

Studies in ovariectomized estrogen-replaced rhesus monkeys show that ACTH, cortisol and progesterone responses to a 30-min intracerebroventricular infusion of interleukin-1α are increased compared to the responses found in ovariectomized estrogen-replaced control monkeys infused intracerebroventricularly with a physiological saline solution. Experimental females received estrogen replacement therapy during 5 days at doses that elevated plasma estradiol concentration to 378 pmol/L (level that reproduces the typical high estradiol concentrations of the late follicular phase) 56. The cortisol response to this immune-inflammatory challenge is also increased in ovariectomized 5-day-estrogen-treated monkeys with plasma concentrations of estradiol ≤ 73 pmol/L 57. However, in 5-day-replaced rhesus monkeys with plasma levels of 114 pmol/L estradiol (concentration similar to that of the early-mid follicular phase), the cortisol response is inhibited 57.

Studies in estrogen-replaced menopausal or ovariectomized women provide intriguing results. Some reports evidence no effect of estradiol treatment on (1) the increases of plasma cortisol and progesterone observed in response to a single intravenous injection of endotoxin after 4 weeks of treatment [mean ( ± SEM) plasma estradiol levels: 23.5 ± 3.3 pmol/L in unreplaced women and 315.7 ± 36.7 pmol/L after estradiol replacement] 35; or (2) the cortisol response to psychological stressors (speech and math tasks) after at least 2 years of estrogen alone (a mixture of 6 estrogenic substances) or estrogen and progestogen treatment (authors did not specify the plasma levels of estradiol of replaced and unreplaced women) 58. Another study shows no changes in plasma ACTH concentration or even decreased peak levels of plasma cortisol and androstenedione in response to psychological stressors (speech and math tasks) after 6 weeks of treatment with estradiol compared to the increases evidenced in these women before estradiol replacement (plasma estradiol levels: ≤ 73.4 pmol/L before estradiol replacement and 587.4 ± 58.7 pmol/L after estradiol replacement) 59. In perimenopausal women (within 2 years of their last period and actively experiencing vasomotor symptoms of menopause), estradiol supplementation for 8 weeks results in decreased plasma levels of ACTH and cortisol after a psychological stress (mental arithmetic tasks accompanied by a repetitive annoying background noise to induce difficulty in concentration) compared to placebo-treated perimenopausal women (plasma estradiol levels: 115.0 ± 19.0 pmol/L in placebo-treated women and 1118.0 ± 111.0 pmol/L in estradiol-treated women) 60.

As far as we know, there are no studies reporting increases in HPA axis reactivity to acute stressors in estrogen-treated perimenopausal, menopausal or ovariectomized women. Only one study in young men has found increased peak ACTH and cortisol responses to a brief psychosocial stress (free speech and mental arithmetic in front of an audience) after 24-48 h of estradiol treatment compared to a placebo group 61.

The studies focused on the effects of hormone replacement on HPA axis response suggest that estrogen treatment in ovariectomized females or perimenopausal and menopausal women may modify the HPA axis response to acute stressors. In rats, it appears that estradiol treatment increases the responsivity of the HPA axis to different acute stressors. In rhesus monkeys, the HPA axis response to a 30-min intracerebroventricular infusion of interleukin-1α depends on the plasma level of estradiol after estrogen replacement. In particular, low and high plasma concentrations increase the response whereas intermediate concentrations are inhibitory. In contrast, estrogen replacement in women does not change or even may decrease the typical response of the HPA axis to acute stressors. Although it is difficult to compare among studies the absolute plasma levels of estradiol exhibited by women before being exposed to acute stressors because different measurement methods were used, it is worth mentioning that, in all the studies analyzed in this section, independently of the replacement regimen followed, the HPA axis always displayed a positive response when plasma levels of estradiol were relatively low (≤ 73.4 pmol/L) or intermediate (115.0-315.7 pmol/L) whereas the response was decreased when plasma levels of estradiol were relatively high (587.4-1118.0 pmol/L). Thus, it can be hypothesized that the response of the HPA axis to inflammatory or psychological stressors evidenced in perimenopausal, menopausal or ovariectomized women may depend on the range of plasma estradiol concentration present in women when exposed to these stressors.

Brown-Grant et al. 62 reported that ≈ 50% of light-induced persistent-vaginal-estrus rats, which show elevated levels of estradiol (similar to the levels on the morning of proestrus in rats maintained under a controlled photoperiod 63), can ovulate when placed in a mating cage without a male or when subjected to a more severe stress such as laparotomy. The laparotomy-associated stress can also advance the time of LH surge (zenith: ≈ 0.130 IU/L) in 5-day cycling rats if performed on the morning of proestrus 38. On the contrary, in 4-day cyclers, laparotomy delays pituitary LH discharge and ovulation if carried out on the morning of diestrus-2 64. An inhibition of the primary LH surge and ovulation has also been found in 4-day cycling rats after intracerebroventricular injection of interleukin-1α or β on the morning of proestrus 65. Furthermore, it has been reported that a rapid blood volume depletion from the external jugular veins under constant ether anesthesia is able to increase plasma LH concentrations (from ≈ 0.017 IU/L to 0.043 IU/L) during the first 10 min after the initial blood loss in 4-day cycling rats at diestrus (combined diestrus-1 and 2) 66.

In cycling rhesus monkeys, Norman et al. 67 found that a prolonged (6 h) chair restraint resulted in suppression of LH release, but 1 out of the 9 females tested exhibited a 2-5-fold increase in plasma levels of LH within 30 min of the initiation of restraint in both the mid-follicular and mid-luteal phases, as well as an increase in frequency of LH pulses over that normally evidenced in the mid-luteal phase. Furthermore, a 30-min intracerebroventricular administration of interleukin-1α induces a 3-fold increase in LH by 5 h after interleukin-1α infusion if administered during the mid-follicular phase, but there is no LH response during the early follicular phase 41. We should mention that 1 out of the 9 mid-follicular monkeys exhibited a sustained surge-like release of LH after interleukin-1α treatment reaching a maximum (≈ 0.130 IU/L) at 13 h after treatment 41. These findings contrast with the inhibition of the primary LH surge in cycling rats after intracerebroventricular injection of interleukin-1α or β on the morning of proestrus (see above). Discrepancies between rhesus monkeys and rats may be explained by the higher capacity of the rat adrenal glands to synthesize and release progesterone compared to the primate adrenal glands 68. Thus, administration of interleukin in rats may induce the release of adrenal progesterone at concentrations that become inhibitory to LH secretion, whereas in rhesus monkeys the activation of the HPA axis by interleukin administration may result in the release of adrenal progesterone at concentrations equivalent to those observed during the preovulatory midcycle rise (cited by Xiao and Ferin 68).

The surgical stress of bilateral ovariectomy plus total hysterectomy performed under general anesthesia during the mid- to late-follicular phase of the menstrual cycle induces a small increase in plasma levels of LH 36 h after surgery (from ≈ 6 IU/L to 8 IU/L), preceded by an elevation of adrenal progesterone 43. Likewise, women undergoing cholecystectomy under general anesthesia during the early- to mid-follicular phase exhibit a small but significant increase in plasma levels of LH (from ≈ 6 IU/L to 12 IU/L) and FSH (from ≈ 6 IU/L to 9 IU/L) 12 h after the operation, coinciding with a peak in plasma progesterone 42. However, no wave-like changes in plasma LH levels are observed in women undergoing bilateral ovariectomy plus total hysterectomy during the early- to mid-luteal phase 43.

Although some studies report that light, heavy or exhaustive exercise does not alter 69 or even decreases 46 plasma levels of LH at the mid-follicular or mid-luteal phases, other studies evidence small increases in plasma levels of LH in response to acute physical exercise. In fact, 60 min of progressive submaximal treadmill exercise in the mid-follicular phase produces a small surge-like increase in plasma levels of LH (from ≈ 10 to 14 IU/L) in 24-h fasted eumenorrheic women 49 as well as a significant stimulatory effect on maximal peak amplitude of LH pulses (from 8.8 to 9.5 IU/L) in sedentary eumenorrheic women 53. In contrast, the same stress applied in the mid-luteal phase 70 does not affect the LH pulse characteristics.

The studies analyzed in this section show conflicting results on the effects of ovarian cycle on acute-stress-induced LH release in rats. Some studies seem to indicate that the adenohypophysis is much more responsive to acute stressors on the morning of proestrus than at diestrus. In contrast, other studies suggest the opposite. It can be hypothesized that the degree of adrenal-progesterone response depends on the strength of the stressor applied. Thus, strong stressors such as intracerebroventricular injection of interleukin-1α or β or rapid blood volume depletion may induce an adrenal-progesterone response higher than that elicited by more moderate stressors such as laparotomy. Such an adrenal-progesterone response would inhibit the release of LH if plasma levels of estradiol are elevated (proestrus morning) but it would stimulate the release of LH under low estradiol backgrounds (e.g. the diestrus stage). Likewise, a moderate rise in adrenal-progesterone concentrations equivalent to that observed during the preovulatory midcycle rise would be stimulatory if it takes place under high plasma levels of estradiol (proestrus morning) but it would be unable to elicit a release of LH under low estradiol backgrounds (e.g. the diestrus stage).

In rhesus monkeys, females at the mid-follicular and mid-luteal phases, but not at the early-follicular stage, display appropriate plasma levels of estradiol for acute stressors to induce a release of LH.

Although there are some conflicting reports in women, it seems that the mid-follicular phase gathers more favorable conditions than the mid-luteal phase for the release LH by the adenohypophysis in response to acute stressors.

In contrast to ovariectomized non-estrogen-primed rats, ovariectomized estrogen-primed rats respond with a small but significant LH peak (≈ 0.017 IU/L) either 1 h after the onset of a prolonged (11 h) immobilization 71 or 10 min after a short (15 min) immobilization 71. Likewise, ovariectomized rats treated with estradiol, progesterone and thyroxine respond with an LH surge (zenith: ≈ 0.052 IU/L) 10 min after undergoing a rapid blood volume depletion from the external jugular veins. This response, however, is not observed in ovariectomized non-treated rats 66. We should note, however, that acute stressors such as the transfer of rats to a novel environment followed by a return to their original quarters 30 min later, or 15-min exposure to a strobe light have no effects on serum LH in 2-week ovariectomized estradiol-primed rats, but significantly enhance LH release (zenith: ≈ 0.015 IU/L and ≈ 0.020 IU/L, respectively) in 2-week ovariectomized non-estrogen-primed rats 72.

In ovariectomized rhesus monkeys after 5 days of estrogen replacement therapy at doses simulating the late follicular phase, a 30-min intracerebroventricular infusion of interleukin-1α results in a small but significant increase in LH release during the first 5 h after treatment (from 0.014 IU/L to 0.019 IU/L) 56. However, the response of LH is prevented in replaced monkeys with plasma levels similar to those of the early-mid follicular phase 57.

Estrogen-treated menopausal or ovariectomized women respond to an inflammatory/immune-like stress, induced by single intravenous injection of endotoxin, with an ovulatory-like LH peak (50 IU/L) 7 h after injection. In contrast, non-estrogen treated women do not modify their basal LH levels after endotoxin injection 35.

Studies in ovariectomized rats show paradoxical results as far as the response of the adenohypophysis to different acute stressors is concerned. Whereas 2 studies 6671 found a stimulatory effect of estradiol replacement on acute-stress-induced LH release, Briski and Sylvester 72 evidenced that several types of acute stress exerted different effects on pituitary LH release and that the steroid environment modulated in a different way (inhibiting or stimulating) the pattern of response of the HPG axis induced by these stressors.

Literature focused on rhesus monkeys shows a stimulatory effect of estradiol replacement on acute-stress-induced LH release at doses that result in plasma estradiol levels reproducing the late follicular phase. This stimulatory effect, however, is not found after estradiol treatment at doses that result in plasma estradiol levels simulating the early-mid follicular phase.

Finally, estrogen-treated menopausal or ovariectomized women respond with a clear ovulatory-like LH peak after a single intravenous injection of endotoxin 35.