Kaplan, J. R. PhD; Manuck, S. B. PhD; Anthony, M. S. PhD; Clarkson, T. B. DVM
Ovarian hormones are widely thought to protect premenopausal women against the development of atherosclerosis and coronary heart disease.1 The well‐known increased incidence of heart disease in peri‐ and postmenopausal women supports this suggestion.2 Evidence also links premenopausal reductions in endogenous estrogen to premature acceleration of cardiovascular disease. Hence, increased risk of cardiovascular disease is observed in relatively young women in association with bilateral oophorectomy and early menopause.1,3 Furthermore, premenopausal women with angiographically confirmed coronary disease have significantly lower plasma estradiol concentrations than do controls.4
To the extent that premenopausal estrogen deficiency increases cardiovascular risk, factors correlated with such deficiency may themselves be relevant predictors of cardiovascular disease. Notably, certain behavioral phenomena, including psychologic stress and related affective states such as depression, are associated with both ovarian dysfunction and heart disease.5–7 In addition to women, many species of female mammals are susceptible to behaviorally induced ovarian dysfunction and reproductive suppression.8 Observations in cholesterolfed cynomolgus macaques (Macaca fascicularis) illustrate the potential influence of behavioral and hormonal factors on cardiovascular disease. In this species, females of chronically low social status (subordination to dominant animals within social groups) reliably become estrogen deficient and develop pathobiologic changes in the coronary arteries, including accelerated atherogenesis and abnormalities of coronary vasomotion.9–11 Subordinate females are also hypercortisolemic and often display behavioral withdrawal and inactivity (ie, depression), characteristics shared with women having psychogenic amenorrhea.12
The current study—a pre‐ and postmenopausal intervention trial—explored the life course implications of the foregoing associations on behavior, ovarian hormones, and atherosclerosis in cholesterolfed, group‐housed cynomolgus monkeys. We hypothesized that if ovarian impairment accelerates atherogenesis, exogenous estrogen should be protective premenopausally, particularly among the animals behaviorally predisposed by the greatest degree of impairment (ie, socially subordinate monkeys). Half of the animals were thus randomly assigned to treatment with oral contraceptives (OCs). An iliac artery was removed from each monkey at the end of the premenopausal phase and atherosclerosis evaluated as a “surrogate” for estimating the extent of coronary artery atherosclerosis then present.13 All animals were then oophorectomized and entered the postmenopausal phase of the experiment, which involved treatment with soy phytoestrogens or conjugated equine estrogens. Finally, coronary artery atherosclerosis was assessed on necropsy of the animals.
Previously, we reported findings from each phase separately, as related to the hypotheses tested in that phase.13,14 The premenopausal data revealed the predicted ovarian dysfunction and exacerbation of iliac artery atherosclerosis in untreated subordinate monkeys in comparison with their dominant counterparts; exposure to OC, however, completely prevented the exacerbation of iliac artery atherosclerosis in subordinates.13 The data from the postmenopausal phase were evaluated using statistical procedures designed to account for premenopausal differences in atherosclerosis extent and test only postmenopausal treatment effects; social status was not considered in this analysis. Conjugated equine estrogens significantly inhibited the development of coronary artery atherosclerosis; soy phytoestrogens provided weaker protection at that site.14
The premenopausal results relating social status and contraceptive steroid exposure to atherosclerosis leave four unresolved questions: 1) Does the preferential protection of socially subordinate monkeys afforded by OC exposure as seen in the iliac arteries extend also to the coronary arteries, sites of significant clinical relevance? 2) Do these premenopausal effects establish a trajectory of atherogenesis that projects into the postmenopausal period, the time during which most clinical events in women are manifested? 3) Do postmenopausal hormone treatments modify the influence of premenopausal status and OC treatment on postmenopausal atherosclerosis? And finally, 4) Are premenopausal influences on postmenopausal atherosclerosis independent of concomitant variation in plasma lipids?
MATERIALS AND METHODS
The original animals used in this study were 213 female cynomolgus monkeys imported from Indonesia (Institut Pertanian Bogor, Bogor, Indonesia) as fully mature, unrelated, premenopausal adults (average age 6 years, estimated by dental examination) and subsequently housed in social groups of five or six animals each. One hundred seventy‐seven monkeys completed the pre‐ and postmenopausal phases of this experiment (Figure 1); their data are reported here. The remaining 36 monkeys either died before the end of the entire (5.5 years) study period or were removed from their social groups for significant periods of time due to illness. Premenopausal females of this species are susceptible to diet‐induced atherosclerosis, resemble women in menstrual cycle length and hormone profile, and have been used in numerous studies evaluating behavioral and hormonal influences on coronary artery atherosclerosis.15 All procedures involving animals were conducted in compliance with State and Federal laws, standards of the U.S. Department of Health and Human Services, and guidelines established by the Wake Forest University Animal Care and Use Committee.
For the premenopausal phase of the study, before any treatment was given, animals were randomly assigned to social groups, which remained unmanipulated for an average of 6 months. All groups then entered a 26‐month treatment period, during which animals consumed an atherogenic diet containing 17% of calories from protein, 45% from fat, 38% from carbohydrates, and 0.28 mg cholesterol/cal. Additionally, half of the social groups were assigned randomly to receive, in the diet, a triphasic OC containing varying proportions of ethinyl estradiol (0.03–0.04 mg/1800 cal) and levonorgestrel (0.05–0.125 mg/1800 cal) for the initial 21 days of each cycle, with a placebo given for the final 7 days—the same regimen used by women. The dose for the treatments was adjusted on a caloric basis, assuming an average human consumption of 1800 cal/day. Experimenters and technicians were blinded to the treatment assignments.
The social status of each animal relative to the others in her social group was based on data collected during weekly, 30‐minute observations beginning after social group formation and before provision of the atherogenic diet or contraceptive steroids. Dominance and subordination were determined by the outcomes of fights, which are highly asymmetric in this species and yield clear winners and losers as judged by specific facial expressions, postures, and vocalizations.15 The female in each group that defeated all others was designated the first‐ranking monkey. The female that defeated all but the first‐ranking monkey was designated second‐ranking monkey, and so forth. For purposes of analysis, animals ranking first or second in groups of five were considered dominant, as were animals ranking 1, 2, or 3 in groups of six; the remainder of monkeys were subordinate.13
Determinations of plasma lipids (high‐density lipoprotein cholesterol, low‐density plus very low‐density lipoprotein cholesterol) were made nine times, using previously published methods.14 Oral conraceptive–treated monkeys were also sampled on several occasions for plasma concentrations of ethinyl estradiol and levonorgestrel (day 21 of the pill cycle), revealing blood concentrations comparable to those observed in women using the same compound. Plasma samples from a randomly chosen subset of untreated monkeys were collected three times per week for 10 weeks and assayed for progesterone to determine menstrual cycle quality.16 Finally, all untreated animals were vaginally swabbed (to detect menses) three times per week for 10 weeks, on four occasions in the premenopausal period.
After 26 months (the premenopausal phase of this experiment), a segment of the left common iliac artery (approximately 3 cm in length) was removed, perfusion‐fixed in neutral buffered formalin, sectioned, and then stained in preparation for measurement of atherosclerosis extent. The monkeys were then oophorectomized and entered the postmenopausal phase.
The postmenopausal phase of the study lasted for 36 months. During this time, animals remained in their original social groups and consumed an atherogenic diet containing 19% of calories from protein, 37% from carbohydrates, and 44% from fat; cholesterol was added at 0.28 mg/cal. The protein in this diet was derived entirely from isolated soy protein that was either alcohol washed to remove phytoestrogens or left with phytoestrogens intact. Stratifying by OC treatment, social groups were randomized to one of three postmenopausal treatment conditions: 1) alcohol washed; 2) phytoestrogens intact at 129 mg/1800 cal (approximately 91 mg genistein, 31 mg daidzein, and 7 mg glycitein); and 3) alcohol‐washed plus conjugated equine estrogens at 0.625 mg/1800 cal. Blood samples for determination of plasma lipids were collected biannually. Dominance status was determined weekly, as in the premenopausal phase.
Monkeys were necropsied at the end of the postmenopausal phase. The remaining common iliac artery (the right, approximately 3 cm) was removed and perfusion fixed. The heart was removed, and the coronary arteries were perfused with neutral buffered formalin at physiologic pressure.
To assess coronary artery atherosclerosis, 15 blocks (each 3 mm in length) were cut perpendicular to the long axis of the arteries. Five of these were serial blocks from the left circumflex artery, five were from the left anterior descending artery, and five were from the right coronary artery. Sections from each block were stained, projected, and then quantified with a computer‐assisted digitizer. The extent of atherosclerosis in each arterial section was measured as the area (in mm2) between the internal elastic lamina and the lumen. Iliac artery atherosclerosis was measured from sections of three serial blocks using the procedures described for the coronary arteries. The same approach was used in evaluating iliac atherosclerosis from biopsies taken at the end of the premenopausal phase.13
Before statistical analysis was performed, lesion data were subjected to square‐root transformation to yield distributions acceptable for parametric analysis. All figures depict untransformed values. We had previously shown in the premenopausal phase that untreated subordinate animals were estrogen deficient and developed more atherosclerosis than untreated dominants; subordinates treated with OCs had significantly less atherosclerosis than untreated subordinates, and did not differ from either OC‐treated or untreated dominants. To determine whether these same factors (social status, OC exposure) predicted postmenopausal atherosclerosis or interacted with postmenopausal treatments to influence outcomes, the coronary data were subjected to three‐way analyses of variance; the three between‐subject factors were social status (dominant, subordinate), premenopausal treatment (OC, no OC), and postmenopausal treatment (conjugated equine estrogens, soy phytoestrogens, control). Possible mediation by concomitant variation in plasma lipids was evaluated by analysis of covariance. Pairwise comparisons with Dunn's procedure17 were used to confirm the significance of a priori group contrasts hypothesized on the basis of the premenopausal outcomes identified above. Also, atherosclerosis extent measured in the premenopausal iliac artery biopsies was compared with lesion extent measured in the contralateral iliac artery sampled at necropsy. This was done by repeated‐measure analysis of variance using time of assessment (premenopausal, postmenopausal) in addition to the same between‐subject factors as used for the coronary arteries.
For plasma lipid data, treatment period averages were used in the foregoing analyses. Categorical status rankings (dominant, subordinate) were based on the outcomes of competitive interactions observed during the months animals were in social groups prior to the beginning of the premenopausal treatment. The possibility that social groups may have influenced outcomes was tested by means of analysis of variance, entering each social group as a dummy variable. Social groups were not significant predictors of atherosclerosis in this analysis, suggesting that social group identification did not account for appreciable variability in the analysis of status and hormonal influences on lesions.
A two‐tailed significance level of .05 was chosen for all comparisons except those involving multiple contrasts. Analyses were done using the SAS statistical package (SAS Institute, Cary, NC).
The monkeys were relatively stable in social status. Across the 26‐month premenopausal phase, seven monkeys (4%) changed categoric status from dominant to subordinate or the reverse; four of these changes occurred in groups from which animals had been removed due to extended illness or death. Sixteen additional changes (9%) occurred over the 36‐month postmenopausal phase; eight of these changes occurred in groups affected by illness or death. To be conservative, atherosclerosis and plasma lipid analyses were conducted first on the entire data set, and again after removing the 23 animals that had changed categoric ranks; the patterns of significance and the mean values were almost indistinguishable between the analyses.
The mean extent of coronary artery atherosclerosis for dominant and subordinate animals in each pre‐ and postmenopausal treatment condition is shown in Figure 2. Analysis of variance revealed main effects for all three factors (status: F1,165 = 6.82, P < .01; premenopausal treatment: F1,165 = 4.27, P = .04; postmenopausal treatment: F2,165 = 4.00, P = .02), as well as a significant status × premenopausal treatment interaction (F1,165 = 5.20, P = .02). There were no other significant interactions. As shown in Figure 2, the postmenopausal treatment effect reflected the significant inhibition of atherosclerosis by conjugated equine estrogens.14 The status × premenopausal treatment interaction was evaluated by planned comparisons—depicted in Figure 3 —which confirmed three study predictions: 1) untreated dominant monkeys had less coronary atherosclerosis than untreated subordinates (P = .001); OC‐treated subordinates had less atherosclerosis than untreated subordinates (P < .01); and 3) OC‐treated dominant and subordinate animals did not differ in lesion extent (NS). Hence, premenopausal subordinate monkeys develop significantly more atherosclerosis as measured postmenopausally than dominant monkeys, an effect that can be mitigated by OC treatment and that persists regardless of postmenopausal hormone treatment. To determine whether the effects of premenopausal conditions on coronary artery atherosclerosis were mediated by concomitant variation in plasma lipids, we reanalyzed the coronary data by analysis of covariance, entering the premenopausal plasma lipids as covariates. There were no significant changes in outcome, suggesting that the behavioral and hormonal influences on atherosclerosis were not lipid mediated.
Repeated measures analysis of variance showed significant lesion progression in the iliac artery between the pre‐ and postmenopausal assessments (F1,165 = 20.3, P < .001). There were also significant main effects of status (F1,165 = 4.42, P < .04) and postmenopausal treatment (F2,165 = 3.60, P < .03), though interpretation of each of these effects is qualified by significant interaction terms. First, there was a significant status × premenopausal treatment interaction (F1,165 = 4.70, P = .03). As depicted in Figure 4 (which also shows lesion progression across all conditions), subordinate monkeys had exacerbated iliac artery atherosclerosis, an effect mitigated by OC treatment and obtained irrespective of time of assessment (either pre‐ or postmenopausally). These effects closely resemble the outcome observed in the coronary arteries, which may relate in part to the significant correspondence between atherosclerosis extent measured in these two sites at necropsy (r177 = 0.75, P < .001). Additionally, the postmenopausal treatment × time of assessment interaction was also significant (F1,165 = 20.3, P < .001). This latter effect (not shown, but described in Clarkson et al14) indicates that conjugated equine estrogens significantly inhibited postmenopausal atherosclerosis progression.
The number of menstrual cycles identified through 40 weeks of vaginal swabbing of all premenopausal monkeys not treated with OCs did not differ by social status (n = 41 dominants: 5.4 ± 0.36 [SEM] cycles; n = 46 subordinates: 5.9 ± 0.36 cycles, t85 = 0.91, NS). However, in a randomly chosen subset of animals (n = 18), luteal phase plasma progesterone concentrations were reliably higher among dominants than subordinates (n = 9 dominants: 10.58 ± 1.6 ng/mL; n = 9 subordinates: 5.96 ± 0.97 ng/mL, t16 = 2.47, P < .03).
The data reported here provide experimental evidence that premenopausal social status and OC exposure influence postmenopausal coronary artery atherosclerosis measured in the same individuals. Social subordination was associated with an acceleration of coronary artery atherosclerosis, whereas premenopausal OC treatment (or dominant social status) completely inhibited the lesion exacerbation occurring in subordinates. Furthermore, these effects were not explained by concomitant variation in plasma lipids, and persisted irrespective of postmenopausal treatment with soy phytoestrogens or conjugated equine estrogens.
Ancillary evidence suggests that the subordinate monkeys may have been relatively hypoestrogenic, possibly accounting for the exacerbation of atherosclerosis seen in subordinate animals not treated with OCs. Insofar as the depression of luteal phase progesterone concentrations associated with low social status, seen here in a subset of animals, is representative of the full cohort, it is conceivable that the greater atherosclerosis of untreated subordinate animals stemmed from relative estrogen deficiency.16 This suggestion is consistent also with the observation that exogenous estrogen exposure (OCs) inhibited atherogenesis among subordinate animals alone, rendering them equivalent in extent of lesion to both treated and untreated dominants.
The influence of the premenopausal hormonal and behavioral conditions was sustained postmenopausally despite oophorectomy of all animals (which effectively rendered dominants and subordinates equivalent with respect to gonadal hormones) and the termination of OC treatment. This outcome suggests that the conditions prevailing premenopausally caused coronary artery atherosclerosis to develop more rapidly in some animals (ie, untreated subordinates) than in others (ie, all dominants and treated subordinates), with significant differences probably already established by the beginning of the postmenopausal phase. This suggestion is supported by the iliac artery data showing that dominants and subordinates did not differ in extent of atherosclerosis progression over the postmenopausal period. Rather, subordinates had more extensive atherosclerosis at the end of the premenopausal phase, and this difference continued unabated. The persistence of the premenopausal effects despite a prolonged period of postmenopausal hormone replacement emphasizes the potential importance of early events in the life course development of coronary artery atherosclerosis.
The foregoing observations may be relevant to lesion development in women. First, atherosclerosis takes decades to progress from initial lesions to clinical events.18 Also, one‐third of all women already have raised lesions in their coronary arteries by age 35.19 This pattern suggests that coronary disease, the largest cause of death in perimenopausal and older women, originates premenopausally. Furthermore, as observed in female monkeys, a substantial number of young women may experience ovarian compromise at some time during their reproductive years.20–22 One particular expression of compromise—functional hypothalamic amenorrhea (sometimes called psychogenic amenorrhea)—offers perhaps the closest parallel between women and the subordinate monkeys described in this report. Functional hypothalamic amenorrhea is an important and reversible cause of clinically apparent ovulatory dysfunction; it is associated with abnormal luteinizing hormone pulse generator activity and is accompanied by hypercortisolemia and other neuroendocrine and behavioral indicators of stress.23–25 Less severe, subclinical ovarian impairment may affect an even larger number of women. For example, a year‐long study of 66 women thought to have normal menstrual cycles revealed that 80% experienced some degree of disturbed luteal phase function.22 Decreases in spinal bone density over the subsequent 5 years were associated significantly with this degree of abnormality, emphasizing that even subclinical ovarian dysfunction is pathobiologically relevant to estrogen‐sensitive tissues.26 The subordinate monkeys in our studies also experience similarly moderate, reversible ovarian impairment, which is not reflected in an increased incidence of amenorrhea.16,27 However, subordinates are reliably characterized by hypercortisolemia and behavioral indices of stress.12,28
If ovarian hormones are cardioprotective premenopausally, women with functional hypothalamic amenorrhea and subclinical ovarian dysfunction could share with the subordinate monkeys in this report a vulnerability to atherogenesis and an increased risk of coronary disease. The relatively high incidence of subclinical ovarian dysfunction described above suggests that the percentage of premenopausal women experiencing accelerated atherosclerosis may be much larger than the number clinically diagnosed with functional hypothalamic amenorrhea and may contribute substantially to the coronary disease observed postmenopausally.
Finally, the protective effect of estrogen suggested by the current report must be reconciled with the results of recent studies conducted on postmenopausal women, which indicate that estrogen may have neutral or adverse effects on advanced atherosclerosis. For example, in the Estrogen Replacement and Atherosclerosis Trial, neither unopposed conjugated equine estrogens or equine estrogens plus medroxyprogesterone acetate slowed the progression of coronary artery atherosclerosis in women with preexisting disease.29 These results mirror those of the Heart and Estrogen/Progestin Replacement Study, which found that combined hormone replacement therapy (conjugated equine estrogens and medroxyprogesterone acetate) was ineffective in secondary prevention of coronary heart disease.30 In contrast, large prospective investigations, such as the Nurses Health Study, have repeatedly shown hormone replacement to be effective in the primary prevention of coronary disease.31,32 Similarly, studies with monkeys indicate that conjugated equine estrogens inhibit atherogenesis in oophorectomized females initially free of lesions, but provide no additional benefit to dietary therapy in females with preexisting atherosclerosis.33 Together, the results of epidemiologic and experimental investigations are consistent with the hypothesis that estrogen inhibits atherogenesis, but is less effective or even detrimental in the face of established disease.29 This possibility implies, in turn, that efforts directed at primary prevention of coronary disease in women should include premenopausal considerations, especially the quality of ovarian function during this time, rather than focus solely on postmenopausal interventions.
We were able to analyze luteal phase plasma progesterone concentrations of 12 additional animals randomly chosen from the non‐OC treated portion of this experiment. When combined with the data from the original 18 animals (see Results, last paragraph), the significant progesterone difference between dominant and subordinate monkeys across this larger and more representative sample became stronger (n = 15 dominants: 10.93 ± 1.23 ng/mL; n = 15 subordinates: 6.58 ± 0.71 ng/mL, t28 = 3.04, P < .01). These data further support our suggestion (in the Discussion) that the depression of luteal phase progesterone concentrations associated with low social status was representative of the entire cohort, indicating that the non‐OC treated subordinates were relatively estrogen deficient.
1. Manson JE. Postmenopausal hormone therapy and atherosclerotic disease. Am Heart J 1994;128:1337–43.
2. Douglas PS. Cardiovascular health and disease in women. 1st ed. Philadelphia: W.B. Saunders Company, 1993.
3. Hu FB, Grodstein F, Hennekens CH, Colditz GA, Johnson M, Manson JE, et al. Age at natural menopause and risk of cardiovascular disease. Arch Intern Med 1999;159:1061–6.
4. Hanke H, Hanke S, Ickrath O, Lange K, Bruck B, Muck AO, et al. Estradiol concentrations in premenopausal women with coronary heart disease. Coron Artery Dis 1997;8:511–5.
5. Laughlin GA, Dominguez CE, Yen SSC. Nutritional and endocrine-metabolic aberrations in women with functional hypothalamic amenorrhea. J Clin Endocrinol Metab 1998; 83:25–32.
6. Rozanski A, Blumenthal JA, Kaplan J. Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation 1999;99:2192–217.
7. Young EA, Midgley AR, Carlson NE, Brown MB. Alteration in the hypothalamic-pituitary-ovarian axis in depressed women. Arch Gen Psychiatry 2000;57:1157–62.
8. Wasser SK, Barash DP. Reproductive suppression among female mammals: Implications for biomedicine and sexual selection theory. Q Rev Biol 1983;58:513–38.
9. Adams MR, Kaplan JR, Koritnik DR. Psychosocial influences on ovarian endocrine and ovulatory function in Macaca fascicularis. Physiol Behav 1985;35:935–40.
10. Williams JK, Shively CA, Clarkson TB. Determinants of coronary artery reactivity in premenopausal female cynomolgus monkeys with diet-induced atherosclerosis. Circulation 1994;90:983–7.
11. Kaplan JR, Adams MR, Clarkson TB, Koritnik DR. Psychosocial influences on female “protection” among cynomolgus macaques. Atherosclerosis 1984:53;283–95.
12. Shively CA, Laber-Laird K, Anton RF. The behavior and physiology of social stress and depression in female cynomolgus monkeys. Biol Psychiatry 1997;41:871–82.
13. Kaplan JR, Adams MR, Anthony MS, Morgan TM, Manuck SB, Clarkson TB. Dominant social status and contraceptive hormone treatment inhibit atherogenesis in premenopausal monkeys. Arterioscler Thromb Vasc Biol 1995;15:2094–100.
14. Clarkson TB, Anthony MS, Morgan TM. Inhibition of postmenopausal atherosclerosis progression: A comparison of the effects of conjugated equine estrogens and soy phytoestrogens. J Clin Endocrinol Metab 2001;87:41–7.
15. Kaplan JR, Adams MR, Clarkson TB, Manuck SB, Shively CA, Williams JK. Psychosocial factors, sex differences, and atherosclerosis: Lessons from animal models. Psychosom Med 1996;58:598–611.
16. Adams MR, Kaplan JR, Clarkson TB, Koritnik DR. Ovariectomy, social status, and atherosclerosis in cynomolgus monkeys. Arteriosclerosis 1985;5:192–200.
17. Sokal RR, Rohlf FJ, eds. Biometry. 2nd edition. San Francisco: WH Freeman and Company, 1981.
18. Strong JP, Eggen DA, Oaldmann NC. The natural history, geographic pathogenesis and epidemiology of atherosclerosis. In: Wissler RW, Geer JC, eds. The pathogenesis of atherosclerosis. Baltimore: Williams and Wilkins, 1972: 20–40.
19. Strong JP, Malcom GT, McMahan CA, Tracy RE, Newman WP III, Herderick EE, et al. Prevalence and extent of atherosclerosis in adolescents and young adults. Implications for prevention from the Pathobiological Determinants of Atherosclerosis in Youth Study. JAMA 1999;281:727–35.
20. Reindollar RH, Novak M, Tho SP, McDonough PG. Adult-onset amenorrhea: A study of 262 patients. Am J Obstet Gynecol 1986;155:531–43.
21. Suh BY, Liu JH, Berga SL, Quigley ME, Laughlin GA, Yen SS. Hypercortisolism in patients with functional hypothalamic-amenorrhea. J Clin Endocrinol Metab 1988;66:733–9.
22. Prior JC, Vigna YM, Schechter MT, Burgess AE. Spinal bone loss and ovulatory disturbances. N Engl J Med 1990;323:1221–7.
23. Berga SL, Girton LG. The psychoneuroendocrinology of functional hypothalamic amenorrhea. Psychiatr Clin North Am 1989;12:105–16.
24. Berga SL, Daniels TL, Giles DE. Women with functional hypothalamic amenorrhea but not other forms of anovulation display amplified cortisol concentrations. Fertil Steril 1997;67:1024–30.
25. Judd SJ, Wong J, Saloniklis S, Maiden M, Yeap B, Filmer S, et al. The effect of alprazolam on serum cortisol and luteinizing hormone pulsatility in normal women and in women with stress-related anovulation. J Clin Endocrinol Metab 1995;80:818–23.
26. Prior JC, Vigna YM, Barr SI, Kennedy S, Schulzer M, Li DKB. Ovulatory premenopausal women lose cancellous spinal bone: A five-year prospective study. Bone 1996;18:261–7.
27. Walker ML, Gordon TP, Wilson ME. Menstrual cycle characteristics of seasonally breeding rhesus monkeys. Biol Reprod 1983;29:841–8.
28. Kaplan JR, Adams MR, Koritnik DR, Rose JC, Manuck SB. Adrenal responsiveness and social status in intact and ovariectomized Macaca fascicularis. Am J Primatol 1986;11:181–93.
29. Herrington DM, Reboussin DM, Brosnihan KB, Sharp PC, Shumaker SA, Snyder TE, et al. Effects of estrogen replacement on the progression of coronary-artery atherosclerosis. N Engl J Med 2000;343:522–9.
30. Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. JAMA 1998;280:605–13.
31. Nabel EG. Coronary heart disease in women—an ounce of prevention. N Engl J Med 2000;343:572–4.
32. Hu FB, Stampfer MJ, Manson JE, Grodstein F, Colditz GA, Speizer FE, et al. Trends in the incidence of coronary heart disease and changes in diet and lifestyle in women. N Engl J Med 2000;343:530–7.
33. Williams JK, Anthony MS, Honore EK, Herrington DM, Morgan TM, Register TC, et al. Regression of atherosclerosis in female monkeys. Arterioscler Thromb Vasc Biol 1995;15:827–36.