Stress and the menstrual cycle: relevance of cycle quality in the short- and long-term response to a 5-day endotoxin challenge during the follicular phase in the rhesus monkey.

The notion that stress activates central and peripheral pathways to inhibit the menstrual cycle is well accepted, but the initial processes through which this occurs have not been investigated. This study uses a relevant nonhuman primate model to document the cyclic endocrine effects imposed by a moderate short-term stress episode in the follicular phase. The stress paradigm is a 5-day inflammatory/immune-like challenge produced by the administration of bacterial endotoxin [lipopolysaccharide (LPS)], which, through the release of endogenous cytokines and other mediators, induces a physiopathological response similar to a bacterial infection. LPS was administered iv twice daily for 5 days starting on days 2-8 of the follicular phase. The stress challenge resulted in a significant lengthening of the follicular phase in all monkeys. Two distinct groups were observed. In group 1 (n = 5), the mean (+/- SE) length of the follicular phase in the LPS-treated cycle was significantly increased, from 10.2 +/- 0.2 in control cycle 2 to 30.8 +/- 4.3 days (except in one monkey that had a 4-month amenorrheic interval). In group 2 (n = 5), the length of the follicular phase significantly increased but not to exceed the duration of the LPS treatment (9.7 +/- 1.1 vs. 13.6 +/- 1.2). Estradiol concentrations decreased significantly after LPS in group 1 (34.8 +/- 5.5 vs. 16.2 +/- 6.5 pg/mL) and remained suppressed after the challenge. In group 2, estradiol levels remained stationary throughout the 5-day LPS treatment (26.0 +/- 6.5 vs. 25.6 +/- 3.9). Compared with control values at a similar stage of the follicular phase, most LH and FSH values during LPS treatment were higher than controls. Estradiol and gonadotropin surges were delayed by LPS treatment for a varying length of time according to each grp. Significant differences in integrated luteal progesterone concentrations characterized control cycles of groups 1 and 2 (group 1: 36.5 +/- 1.5, group 2: 47.5 +/- 2.6). In group 1, there were no further effects of LPS on luteal progesterone during the treatment and two post-LPS cycles. In contrast, in group 2, integrated luteal progesterone concentrations were significantly decreased in post-LPS cycle 1 (to 36.0 +/- 4.4). Cortisol significantly increased at hour 3 after each morning LPS injection but the amplitude of the response decreased over the 5-day period. Progesterone increased significantly by hour 3 after the first LPS injection but remained unchanged after subsequent LPS administration. Our data demonstrate that a 5-day inflammatory-like episode during the follicular phase can delay folliculogenesis and that damage to this process is intensified in individuals who already demonstrate a subtle cyclic degradation, in the form of decreased progesterone secretion in the luteal phases preceding the stress episode. Long-term endocrine effects, in the form of decreased luteal secretory activity in the first poststress cycle, are observed in normally cycling individuals, suggesting that inadequacy of the luteal phase may represent the first stage in the damage that a stress episode can inflict upon the normal menstrual cycle.