Androgen Responses to Stress

Androgen Responses to Stress /. Psychoendocrine Relationships and Assessment of Androgen Activity ROBERT M. ROSE, MD There is good evidence to support extention of the psychoendocrine model beyond the adrenal to include the hypothalamic-pituitary-gonadal system. However, androgen secretion and metabolism is more complex than that of cortisol. Total 17-ketosteroids (17-KS) have been used historically to estimate changes in androgen activity. Because they are derived from glucocorticoids, the adrenal androgens as well as testosterone, they are an unreliable index of changes in androgen activity. To properly evaluate alterations in testosterone secretion, which is the most potent of androgens, it is best to measure blood or urinary production rates. It is also possible to measure the excretion of testosterone glucuronide to evaluate the possible inhibition of testosterone secretion during exposure to potential stress. A HE INFLUENCE of psychological state on adrenal cortical activity has been well established. Mason recently reviewed some 200 studies completed during the past 15 years, which, taken together, establish the sensitivity of the pituitaryadrenal cortical system to a wide variety of psychological stimuli. 25 The adrenal cortical response relates not only to characteristics of the stimulus eg, electric shock, combat, noval situations but also is determined by the past history of the individual, his unique perception of the events, as well as constitutional factors such as weight or body size. 33 We have come to view changes in the secretion of cortisol (man and monkey) or corti- From the Department of Psychiatry, Walter Reed Army Institute of Research, Walter Reed Army Medical Center, Washington, DC. Received for publication May 26, 1969; revision received July 22, 1969. costerone (rat) as a rather sensitive barometer of the organism's response to changes in his environment. When the experience is one of threat or challenge, adrenal cortical activity increases. When the environment is no longer so arousing, when one "copes" or "adapts," cortisol secretion falls to more basal levels, or even lower than normal levels. Low or suppressed levels are seen during hypnosis, 35 or in response to distracting activities such as watching Disney nature movies. 41 The firmly established relationship between environmental stimuli and adrenal cortical activity suggests a corresponding interaction between psychological state and other endocrine systems. There are several lines of evidence which point to potential psychological influences on androgen secretion. The role of the central nervous system medi- 405

406 ANDROGEN RESPONSES TO STRESS. I ating gonadal function and other endocrine activity rests on firm anatomical and physiological evidence. This has been brought together in a recent twovolume work on neuroendocrinology edited by Martini and Ganong. 24 A model similar to that for ACTH has been postulated for the control of luteinizing hormone (LH), also termed interstitial cell stimulating hormone (ICSH). Material classified as releasing-factors, possibly of polypeptide composition, is synthesized in hypothalamic neurosecretory cells and carried via a pituitary portal system to the anterior pituitary (adenohypophysis). 1 ' These releasing-factors act in an unknown manner to promote the synthesis and/or release of the various tropic hormones. With respect to the hypothalamic-pituitary-gonadal system, luteinizing hormone-releasing factor (LH-RF), elaborated in various hypothalamic nuclei, acts on cells located in the anterior pituitary to release increased amounts of LH or ICSH, which in males stimulates secretion of testosterone from the testis. Recent work by Eleftheriou and Church 10 has demonstrated that repeated exposure of mice to aggression with resultant defeat produced a significant decline in LH-RF activity in the hypothalamus. Eleftheriou and Pattison also demonstrated that bilateral amygdaloid lesions in the female deer mouse result in a marked decrease in hypothalamic LH-RF activity. Figure 1 summarizes the proposed neuroendocrine relationships for all the major endocrine systems, including the hypothalamic-pituitary-gonadal system. Other lines of investigation have sought to demonstrate the impact of environmental stimuli on androgen activity by measuring changes in testosterone or its major metabolites. These studies may be viewed as supporting the significance of changes in hypothalamic LH-RF or in plasma or pituitary LH concentration following various experimental manipulations. Earlier work, prior to the development of adequate chemical methods, utilized changes in gonadal size as reflecting potential psychological influences. Christian, 7 in summarizing endocrine adaptive mechanisms in response to population growth in mice, reported a decrease in testis and seminal vesicle weights after crowding. More direct evidence of inhibition of testosterone secretion following ether anesthesia stress in rats was recently reported by Bardin and Peterson. 1 Bliss 3 has also found a fall in plasma testosterone levels in rats following the onset of shock avoidance. In a series of experiments which initially demonstrated central nervous system control of ACTH function, Mason 2827 has obtained evidence of psychological influences depressing androgen secretion. Male rhesus monkeys which were required to avoid an electric shock by appropriate lever pressing over a continuous 72-hr period, showed marked increases in plasma and urinary levels of 17-OHCS. During this period of increased adrenal cortical activity, there was a concomitant fall in urinary testosterone and urinary androsterone and etiocholanolone, the two major 17-ketosteroids. This suggests that during this period of arousal or stress, there may have been a fall in the pituitary secretion of LH, during increased secretion of ACTH. Many animals showed a rebound rise in testosterone following termination of avoidance sessions when the 17-OHCS levels fell to normal. In previous work on psychological influences on cortisol activity, it has been found that individuals differed significantly in their response to apparently similar environments. Not all parents were equally distressed during the terminal phase of their child's illness with leukemia; 43 not all men in a platoon engaged in basic combat training were equally threatened by this often taxing experience; 33 not all men were equally aroused or challenged anticipating an imminent attack from the Vietcong; 4 not PSYCHOSOMATIC MEDICINE

ROSE 407 ANTERIOR HYPOTHALAMIC NEUROSECRETORY NUCLEI -SUPRAOPTIC NUCLEUS.HYPOTHALAMO-NEUROHYPOPHYSEAL TRACT ACTH TSH GH (STH) LH (ICSH) FSH PROLACTIN (LTH) FIG 1. Schematic drawing of the relationship between hypothalamus, median eminence of infundibular stalk, adenohypophysis, and neurohypophysis. One set of neurosecretory cells elaborates releasing factors, which are picked up by the superior capillary plexus of the long pituitary portal system. These substances are carried to various cells in the adenohypophysis where they stimulate the production/release of the trophic hormones. One exception is prolactin-inhibiting-factor (PIF), which exerts a tonic inhibition of prolactin synthesis in the adenohypophysis. Another set of neurosecretory cells originating in the supraoptic and paraventricular hypothalamic nuclei elaborates oxytocin and antidiuretic hormone (ADH). As the neurohypophysis is continuous with the hypothalamus, these hormones are released directly in the posterior pituitary, but they may be carried anterior by short portal vessels. (Abbreviations: CRF, corticotrophinreleasing factor; TRF, thyrotrophin-releasing factor; GHRF, growth hormone-releasing factor; LHRF, luteinizing hormone-releasing factor; FSH-RF, follicule stimulating hormone releasing factor; PIF, prolactin-inhibiting-factor). VOL. XXXI. NO. 5, 1969

ADRENAL TESTIS neutral I7 - KETOSTEROIDS FIG 2. General outline of androgen metabolism. Thickness of arrows denotes approximate quantitative importance of metabolic pathway. Positions of Carbon 11 and Carbon 17 on steroid nucleus are labelled along with C-20 and C-21 of side chain. These are crucial determinants of various classes of steroid metabolites.

ROSE 409 all rhesus monkeys maintained elevated levels of 17-OHCS excretion when forced to remain vigilant and avoid an electric shock by bar pressing. 23 Individuals differ, and their different adrenal cortical responses can be predicted and may provide a method of assessment that is independent of what individuals say they are experiencing, independent of how "stressed" they may appear to others. By grouping individuals in terms of their 17-OHCS response, we have been able to corroborate clinical impressions that individuals differ with respect to how they manifest their distress. For example, some tend to 'suffer in silence,' while others dramatize and exaggerate their difficulties. Some individuals who exemplify this latter characteristic have been shown to have low 17-OHCS excretion, which may be interpreted as their experiencing less distress than their overt behavior might indicate. 44 These differences may reflect the operation of two contrasting psychological modes of response, as described by Weinstein et al 42 as repressors versus sensitizers. Repressors tend to show more evidence of physiological arousal than their verbal reports of stress might indicate. They tend to deny difficulties. Sensitizers are discrepant in the opposite direction, more reported evidence of distress with diminished physiological evidence of such arousal. Most of the work to date reports a suppression of androgen activity following exposure to potentially stressful stimuli. Because individuals do not show uniform cortisol or 17-OHCS responses, it is likely that there will be important individual differences in the degree of testosterone suppression during exposure to potentially threatening or challenging environments. Indeed, this does appear to be the case as reported in the companion paper. 34 However, because androgens are secreted by both the testis and adrenals and form a diverse class of hormones, androgen secretion and me- VOL XXXI, NO. 5, 1969 tabolism is more complex than that seen with the glucoeorticoids. There have been many recent advances in techniques for the measurement of testosterone and the other androgens, and it would be helpful to review these methods and present a brief summary of androgen metabolism. Assessment of Androgen Activity 17-Ketosteroida Although investigators have been aware for some time of possible psychological influences on androgen activity, research in the area has been hampered by the lack of specific techniques for measuring the various androgenic hormones. When methods for the specific and precise measurement of glucoeorticoids cortisol or its metabolites became available, in turn it became possible to investigate issues of psychoendocrine control of the ACTH-adrenal cortical system. Up to most recently, parallel work in the androgens has been limited primarily to the measurement of 17- ketosteroids. Because of the complexity of the 17-ketosteroid story, it is important to digress somewhat and discuss in some detail the reasons which essentially invalidate the use of 17-ketosteroids as a specific measure of changes in androgen secretion. An outline of androgen metabolism is shown in Fig 2. For the sake of the present discussion, let us consider only two of the major classes of secretory products of the adrenal, the glucoeorticoids cortisol and cortisone primarily in humans and the adrenal androgens, which are hormones with relatively weak androgenic activity when compared to testosterone. The two major adrenal androgens are A 4 androstenedione and dehydroepiandrosterone (DHEA), the latter secreted unconjugated and as the sulfate. We will exclude progesterone, es-

410 ANDROGEN RESPONSES TO STRESS. I trogens, and aldosterone as not crucially related to the present discussion. The testis, under normal conditions, is effectively the sole source of testosterone in males and also contributes to a minor degree to circulating A 4 androstenedione and DHEA. 19 17-hydroxycorticosteroids are characterized by the presence of an additional two-carbon side chain (C-21) over the usual 19-carbon structure of the androgens (C-19). This carbon side chain attached to the ring at the C-17 position is one of the essential differences between the glucocorticoids and the androgens. The presence of two hydroxy groups at the C-21 and C-17 positions with a keto group between them yields a positive color reaction with phenylhydrazine, the Porter Silber reaction, and this reaction is characteristic of the class designated as 17-hydroxycorticosteroids (17-OHCS). The androgens do not possess this side chain, but usually have a 17-keto group, hence the designation 17-ketosteroids. The major exception is testosterone with a 17/3-hydroxy group which is usually metabolized to a 17- keto group. The major difficulty in utilizing 17- ketosteroids as an index of androgen activity relates to their diverse origins. The neutral 17-ketosteroids, those compounds with the 17-keto configuration minus the phenolic estrogens, are derived from three major sources glucocorticords, adrenal androgens, and testosterone. The 17-ketosteroids may be subdivided into two groups, those with a hydroxy or a keto group at the C-ll carbon the ll-oxy-17-ketosteroids and those without this additional oxygen the 11-desoxy-17-ketosteroids. The 11- oxy-17-ketosteroids are derived from two major sources. One is from cortisol and cortisone, although less glucocorticoid is metabolized to ll-oxy-17-ketosteroids than to the 17-OHCS, 13 the latter composed primarily of tetrahydrocortisol (THF), tetrahydrocortisone (THE), and allo-tetrahydrocortisol (allo-thf). The other source of the ll-oxy-17-ketosteroids is from A 4 androstenedione, via 11/J-hydroxylation. The ll-desoxy-17-ketosteroids comprise the major portion of urinary 17-ketosteroids and include androsterone, etiocholanolone, and dehydroepiandrosterone (DHEA). The 11- desoxy compounds are derived from A 4 androstenedione, DHEA, as well as testoterone itself. In summary, taken as a group, the neutral 17-ketosteroids may be affected by changes in cortisol secretion, and adrenal androgen secretion as well as changes in testosterone secretion, and provide no clue as to specifically what has occurred. In addition to representing metabolites of several quite different steroid hormones, problems with the chemical determination of the 17-ketosteroids further diminish their usefulness. Goldzieher and Axelrod 14 analyzed a large number of urine samples for their 17-ketosteroid content. They compared the values obtained by measuring the 17-ketosteroids as a group, utilizing the usual Zimmerman color reaction, with that determined by estimating the individual ketosteroids and adding these individual values to get total 17-ketosteroid content. If the method for analyzing the 17-ketosteroids as a group was specific, it should compare closely to that obtained by adding the individual components. They found that in many cases the value obtained by measuring 17-ketosteroids as a group far exceeded the sum of individual compounds and concluded that nonspecific interfering substances often comprise a significant proportion of the total 17- ketosteroid value. In other words, methods that attempt to estimate total urinary 17-ketosteroids often measure substances that are not 17-ketosteroids, and this is not a constant source of error. The fact that 17-ketosteroids are composed of metabolites from cortisol, ad- PSYCHOSOMATIC MEDICINE

ROSE 411 renal androgens, and testosterone, as well as the relative unreliability of the methods used to estimate total 17-ketosteroids, minimize their usefulness as a precise index of androgen secretion, although they may reflect diseases of endocrine excess or deficiency. Testosterone Production Testosterone is believed to be the most potent androgen secreted. There is recent evidence that it may have to be reduced to dihydrotestosterone intracellularly to exert its endocrine effects. 28 Besides stimulating the development of male secondary sex characteristics, testosterone is known to have potent effects on nitrogen balance and protein metabolism. Recent work has also demonstrated the importance of testosterone in modifying social and sexual behavior in animals, not only at the time of administration, but later in life if given during a certain crucial period in neonatal development. 5 The ideal method of measuring testosterone should provide information on how much active hormone is available to the tissues during anv specified period of time. This availability is determined by two major factors. The first is how much testosterone is secreted or converted from various precursors: The amount secreted plus that which is converted equals the production rate of the hormone; the second factor relates to the binding of testosterone in plasma: 30 Approximately 90-953! of testosterone in normal males is bound to alpha and beta globulins. 12 As it is felt that only unbound testosterone is available for cellular uptake, the functional role of bound steroids in plasma is still unclear. 8 There are two methods of estimating the oroduction rate of testosterone, one in blood and the other in urine. Neither attempt to assess any alteration in free versus bound testosterone. As the percentage-binding appears to be relatively constant, this may not be a crucial factor in determining the influence of potential psychoendocrine stimuli. However, this remains to be determined. The blood production rate (PR) is estimated by measurement of the metabolic clearance rate (MCR) and determination of the mean plasma concentration of testosterone (MPC) and calculated as follows: PR (mg/day) = MCR (I/day) X MPC (mg/1). Estimation of the MCR is usually accomplished by the intravenous administration of a priming dose of labelled testosterone followed by the constant continuous infusion of the radioactive steroid. Blood samples are drawn after the administered testosterone is in equilibrium in the body pool at approximately 2 hr after the start of infusion. 18 The radioactivity of the plasma testosterone in those samples is then determined. The MCR is calculated according to the following expressions: 37 MCR (i/day) = r (muc/day) -H- c (muc/1), where r = the rate of infusion, and c = the radioactivity present as plasma testosterone. This method provides the most precise estimate of the testosterone produced over a relatively short period (1-2 hr). Because the metabolic clearance rate has been shown to vary during the day, 37 this method may have limitations in assessing production over a several-day period, unless repeated blood production measurements are made. Urinary production rate may provide a more reliable and practical estimate of testosterone production over a 2 or 3 day period. This technique does involve the continuous and complete collection of urine during this period, which poses considerable technical problems. Selection of the blood or urinary method is therefore determined by the time period over which testosterone production is to be studied as well as the feasibility of continuous intravenous infusion with multiple blood samples versus the continuous collection of urine over 2-3 days. Urinary production rate involves mcas- VOL. XXXI, NO. 5, 1969

ADRENAL TESTIS 90-95% 5-10s '' DEHTDROEPlANDROSTEROIO*L (OSO3l (DHEA + DHEA sulfate) EPWESTOSTE~)O~~E FIG 3. Major sources of urinary testosterone, androsterone, and etiocholanolone. The lower set of arrows represents the major precursors of these urinary metabolites and do not indicate quantitative importance of the pathway. Thus, for example, although only a small percent of produced testosterone is excreted as testosterone glucuronide, approximately 95% of testosterone glucuronide is derived from this source. TLSTOSTE(KWIE ANDROSTERONE ETIOCHOLANOLONE kkmw Wcw-IdeJ ~*oclyi) (~tucuronldd DHEA t DHEA Sultot* ADRENAL TESTIS 80-05% I5-20%

ROSE 413 urement of excreted testosterone glucuronide changed, save for conjugation with glu- as well as estimating the specific curonic acid. 2 Approximately 25-50$ of activity of this metabolite. However, secreted testosterone is metabolized to this technique is based on the assumption androsterone and etiocholanolone, two that testosterone glucuronide is major ll-desoxy-lt-ketosteroids.^1 How- derived only from testosterone. It has been shown by several groups of investigators ever, there is evidence that the percentage of secreted testosterone that is ex- that this is not the case. 8 ' 18 ' 22 In creted as testosterone glucuronide may females, blood testosterone is derived from A 4 androstenedione, and this falsely vary from individual to individual, thus making comparison between individuals elevates estimates of testosterone production more difficult. 8 It does appear, however, by the urinary method. 18 In males, that the percentage of testosterone ex- the secretion of A 4 androstenedione is creted as the glucuronide remains constant much less than females, and this is not for the individual, even in the face so important a source of circulating testosterone. of large increases in secretion rate 17 or Korenman and Lipsett 22 have following exogenously administered tes- shown that small amounts of A* androstenedione tosterone. 2 This would tend to support are conjugated directly in the liver to testosterone glucuronide, and demonstrated that urinary testosterone in males has more than one source. However, the use of testosterone glucuronide for measuring changes in testosterone secretion over time, utilizing the individual as his own control. as they, along with other investiga- There is considerable evidence that the tors, generally find close agreement between blood rnd urinary production rates excretion of urinary testosterone glucuronide reflects changes in the amount of in males, it is felt that the urinary method is useful for physiological studies. testosterone produced in the body. Males excrete 10-20 times more testosterone One exception to this may be that when glucuronide than females. 40 Testosterone the adrenal responds to ACTH stimulation with a large increase in A 4 glucuronide is very low to undetectable androstenedione secretion, the urinary method in castrates, hypogonadal males, 20 and boys before puberty. 15 Testosterone glucuronide is suppressed by substances may give values in excess of that estimated by blood production methods. 38 known to decrease testosterone secretion, However, this is not a constant or totally predictable response to ACTH. 32 such as estrogen, 11 and elevated after the Figure administration of human chorionic gonadotropin. In the study reported in 3 summarizes the relationship between the secretion of testosterone, A 4 androstenedione and DHEA, and the excre- the companion paper, 34 androgen activity was assessed by the excretion of testosterone, epitestosterone, androsterone, tion of testosterone, androsterone and etiocholanolone in the urine. and etiocholanolone. Although the origin of epitestosterone, the alpha epimer of testosterone, is unclear, it may reflect alterations in endogenous testosterone Urinary Testosterone The measurement of urinary testosterone glucuronide itself, without estimating production rate by isotope dilution techniques, is possibly the last compromise from the ideal that may still provide valid information about changes in testosterone production. Approximately 1% of secreted testosterone is excreted un- VOL. XXXI, NO. 5, 1969 activity. 38 As mentioned, androsterone and etiocholanolone are derived from both testosterone and A 4 androstenedione and DHEA. However, it was felt that if changes in excretion of these two major androgen metabolites paralleled those seen in urinary testosterone, it would

414 ANDROGEN RESPONSES TO STRESS. I provide additional evidence for altered testosterone production. Plasma Testosterone Until recently, techniques for the measurement of plasma testosterone have been extremely complex and time consuming. This is due primarily to the fact that the concentration of testosterone in males averages 0.2-1.0 fig/100 ml compared to 10-20 ^g/100 ml for cortisol. Within the last 17 or 18 months, several investigators- 1 ' 29 have published new methods utilizing competitive protein-binding techniques. These promise to be much simpler and adequately sensitive, specific, and reliable. As it appears that plasma testosterone does not exhibit much diurnal variation, it may be that the study of potential psychological influences on testosterone secretion may be accomplished by longitudinal determinations of plasma testosterone. However, this technique would still not take into account possible alterations in metabolic clearance rate (MCR). Summary It is apparent that the secretion of cortisol provides a sensitive index of the degree of arousal or threat the individual experiences. In assessing the potential stress of a particular stimulus, one must take into account the individual's characteristic style of interaction and characteristics of the stimulus itself. There is good evidence to support extending the psychoendocrine model to the hypothalamic-pituitary-gonadal system. In turning our attention to the assessment />f androgen activity, it is clear that the secretion and metabolism of various androgens present a more complex story than that for cortisol. It is therefore important to clarify the usefulness of various indices of androgen or testosterone production before using these to interpret responses in this system to potential stress. Preliminary evidence obtained in studies of hypothalamic LH-releasing factor, along with changes in plasma and urinary testosterone, suggest that production of testosterone may be inhibited during exposure to poetntial stress. However, historically, 17-ketosteroids have been used almost exclusively to assess changes in androgen production. 17- ketosteroids are a class of metabolites that potentially reflect changes in the secretion of glucocorticoids (cortisol and cortisone), adrenal androgens (A 4 androstenedione and DHEA), and testosterone. In addition, techniques for the measurement of total 17-ketosteroids are relatively nonspecific and yield significant nonsystematic errors. Therefore, measurement of this class of metabolites is an unreliable index of change in testosterone production. Assessment of alterations in the production of testosterone, the most potent androgen, is best accomplished by measurement of blood or urinary secretion (production) rates. Although there are both theoretical and technical limitations to these techniques, they represent the best currently available. It is also possible to measure the excretion of testosterone glucuronide, a minor metabolite of secreted testosterone. In males, this is almost entirely derived from endogenous testosterone and may accurately reflect longitudinal changes in the amount of testosterone produced. Measurement of excreted testosterone glucuronide along with androsterone and etiocholanolone, the two major ll-desoxy-17-ketosteroids, may better reflect diminished testosterone production. References 1. BARDIN, C. W., and PETERSON, R. E. Studies of androgen production by the rat: Testosterone and androstenedione PSYCHOSOMATIC MEDICINE

ROSE 415 content of blood. Endocrinology 80: 38, 1967. 2. BAULIEU, E. E., and MAUVAIS-JAHVIS, P. Studies on testosterone metabolism. II. Metabolism of testosterone-4-14 C and androst-4-ene-3, 17-dione-l, 2- s H. / Bio Chem 239:1578, 1964. 3. BLISS, E. Personal communication. 4. BOURNE, P. G., ROSE, R. M., and MA- SON, J. W. 17-OHCS levels in combat: Special forces 'A' team under threat of attack. Arch Gen Psychiat 19.135, 1968. 5. BROWN-GRANT, K. The action of hormones on the hypothalamus. Brit Med Bull 22:273, 1966. 6. CAMACHO, A. M., and MIGEON, C. J. Studies on the origin of testosterone in the urine of normal adult subjects and patients with various endocrine disorders. J Clin Invest 43.1083, 1964. 7. CHRISTIAN, J. J. "Endocrine Adaptive Mechanisms and the Physiologic Regulation of Population Growth." In Physiological Mammology (Vol 1), Mayer, W. V., and Van Gelder, R. G., Eds. Acad Press, New York, 1963, p 189. 8. DE MOOR, P., STEENO, O., and HEYNS, W. Possible role or significance of protein-steroid binding in plasma. Ann Endocr (Pans) 29(Suppl.): 119, 1968. 9. ELEFTHERIOU, B. E., and PATTISON, M. L. Effect of amygdaloid lesions in hypothalamic follicle-stimulating hormone-releasing factor in the female deermouse. / Endocrin 39:613, 1967. 10. ELEFTHERIOU, B. E., and CHURCH, R. L. Levels of hypothalamic luteinizing hormone-releasing factor after exposure to aggression (defeat) in C57BL/ 6J mice. / Endocrin 42:347, 1968. 11. FORCHIELLI, E., RAO, G. S., SARDA, I. R., GIBREE, N. B., Poem, P. E., STRAUSS, J. S., and DORFMAN, R. I. Effect of ethinyloestradiol on plasma testosterone levels and urinary testosterone excretion in man. Ada Endocrin 50:51, 1965. 12. FOREST, M. G., RIVAROLA, M. A., and MIGEONT, C. J. Percentage binding of testosterone, androstenedione, and dehydroisoandrosterone in human plasma. Steroids 12:323, 1968. VOL. XXXI, NO. 5, 1969 13. FUKUSHIMA, D. K., BRADLOW, H. L., HELLMAN, L., ZUMOFF, B., and GAL- LACJHER, T. F. Metabolic transformation of hydrocortisone-4-c 14 in normal men. / Bio Chem 235:2246, 1960. 14. GOLDZIEHER, J. W., and AXELROD, L. R. A study of methods for the determination of total, grouped and individual urinary 17-ketosteroids. / Clin Endocrin Metab 22:1234, 1962. 15. GUPTA, D. Separation and estimation of testosterone and epitestosterone in the urine of pre-pubertal children. Steroids 10:457, 1967. 16. HARRIS, G. W., REED, M., and FAW- CETT, C. P. Hypothalamic releasing factors and the control of anterior pituitary function. BrU Med Bull 22:266, 1966. 17. HORTON, R., SHINSAKO, J., and FOR- SHAM, P. H. Testosterone production and metabolic clearance rates with volumes of distribution in normal adult men and women. Ada Endocrin 48: 446, 1965. 18. HORTON, R., and TAIT, J. F. Androstenedione production and interconversion rates measured in peripheral blood and studies on the possible site of its conversion to testosterone. / Clin Invest 45:301, 1966. 19. HUDSON, B., COGHLAN, J. P., and DUL- MANIS, A. "Testicular Function in Man." In Endocrinology of the Testis (Vol 16), Ciba Colloquia, Little Brown, 1967, p 140. 20. ISMAIL, A. A. A., and HARKNESS, R. A. Urinary testosterone excretion in men in normal and pathological conditions. Ada Endocrin 56:469, 1967. 21. KATO, T., and HORTON, R. A rapid method for the estimation of testosterone in female plasma. Steroids 12:631, 1968. 22. KORENMAN, S. G., and LIPSETT, M. B. Is testosterone glucuronoside uniquely derived from plasma testosterone? / Clin Invest 43:2125, 1964. 23. LEVINE, M. D., GORDON, T. P., and ROSE, R. M. "Behavioral and Endocrine Correlates of Adaptation to Chronic Shock Avoidance." In Proceedings of the Second International Congress of Primatology, Karger, Basel. In press.

416 ANDROGEN RESPONSES TO STRESS. I 24. MARTINI, L., and GANONG, W. F., Eds. Neuroendocrinology (Vol. I and II), Acad Press, New York, 1967. 25. MASON, J. W. "A Review of Psychoendocrine Research on the Pituitary- Adrenal Cortical System." In Organization of Psychoendocrine Mechanisms. (Suppl.) Psychosom Med 30: 576, 1968. 26. MASON, J. W., TOLSON, W. W., ROBIN- SON, J. A., BRADY, J. V., TOLLIVEH, G. A., and JOHNSON, T. A. "Urinary Androsterone, Etiocholanolone, and Dehydroepiandrosterone Responses to 72-hr Avoidance Sessions in the Monkey." In Organization of Psychoendocrine Mechanisms. (Suppl.) Psychosomat Med 30:710, 1968. 27. MASON, J. W., KENION, C. C, COLLIN.J, D. R., MOUGEY, E. H., JONES, J. A., DRIVER, G. C, BRADY, J. V., and BEER, B. "Urinary Testosterone Response to 72-hr Avoidance Sessions in the Monkey." In Organization of Psychoendocrine Mechanisms. (Suppl.) Psychosomat Med 30:721, 1968. 28. MAUVAIS-JARVIS, P., FLOCH, H. H., and BERCOVICI, J. P. Studies on testosterone metabolism in human subjects with normal and pathological sexual differentiation. / Clin Endocrin Metab 28: 460, 1968. 29. MAYES, D., and NUGENT, C. A. Determination of plasma testosterone by the use of competitive protein binding. / Clin Endocrin Metab 28:1169, 1968. 30. MERCIER, A., ALFSEN, A., and BAULIEU, E, E. "A Testosterone Binding Globulin." In Androgens in Normal and Pathological Conditions. Proceedings of the Second Symposium on Steroid Hormones, International Congress Series No. 101. Excerpta Medica Foundation, 1966, p 212. 31. PRUNTY, F. T. G. Androgen metabolism in man some current concepts. Brit Med ] 2:605, 1966. 32. RIVAROLA, M. A., SAEZ, J. M., MEYER, W. J., JENKINS, M. E., and MIGEON, C. J. Metabolic clearance rate and blood production rate of testosterone and androst-2-ene-3, 17-dione under basal conditions, ACTH and HCG stimulation. Comparison with urinary production rate of testosterone. ] Clin Endocrin Metab 26:1208, 1966. 33. ROSE, R. M., POE, R. O., and MASON, J. W. Psychological state and body size as determinants of 17-OHCS excretion. Arch Intern Med 121:406, 1968. 34. ROSE, R. M., BOURNE, P. G., POE, R. O., MOUGEY, E. H., COLLINS, D. R., and MASON, J. W. Androgen responses to stress: II. Excretion of testosterone, epitestosterone, androsterone and etiocholanolone during basic combat training and under threat of attack. Psychosom Med 31:418, 1969. 35. SACHAR, E. J., COBB, J. C, and SHOR, R. E. Plasma cortisol changes during hypnotic trance. Arch Gen Psychiat 14:484, 1966. 36. SAEZ, J. M., and MIGEON, C. J. Problems related to the determination of the secretion and interconversion of androgens by "urinary" methods. Steroids 10:441, 1967. 37. SOUTHREN, A. L., GORDON, G. G., and TOCHIMOTO, S. Further study of factors affecting the metabolic clearance rate of testosterone in man. / Clin Endocrin Metab 28:1105, 1968. 38. TAMM, J., VOLKWEIN, U., and STAR- CEVIC, Z. The urinary excretion of epitestosterone, testosterone and androstenedione following intravenous infusions of high doses of these steroids in human subjects. Steroids 8:659, 1966. 39. TAMM, J., APOSTOLAKIS, M.,andVoiCT, K. D. The effects of ACTH and HCG on the urinary excretion of testosterone in male patients with various endocrine disorders. Ada Endocrin 53:61, 1966. 40. VERMEULEN, A. "Urinary Excretion of Testosterone." In Androgens in Normal and Pathological Conditions. Proceedings of the Second Symposium on Steroid Hormones, International Congress Series No. 101. Excerpta Medica Foundation, 1966, p 71. 41. WADESON, R. W., MASON, J. W., HAM- BURG, D. A., and HANDLON, J. H. Plasma and urinary 17-OHCS responses to motion pictures. Arch Gen Psychiat 9:146, 1963. 42. WEINSTEIN, J., AVERILL, J. R., OPTON, E. M., and LAZARUS, R. S. Defensive PSYCHOSOMATIC MEDICINE

ROSE 417 style and discrepancy between self- of parents of fatally ill children. Psyreport and physiological indexes of stress. / Personality Soc Psychol 10: chosom Med 26:576,1964. 44. WOLFF, C. T., HOFER, M. A., and 43. 406, 1968. MASON, J. W. Relationship between WOLFF, C. T., FRIEDMAN, S. B., psychological defenses and mean uri- HOFEH, M. A., and MASON, J. W. Re- nary 17-OHCS excretion rates: Part II. lationship between psychological defenses and mean urinary 17-OHCS ex- Methodological and theoretical con- siderations. Psychosom Med 26:592, cretion rates: Part I. A predictive study 1964. VOL XXXI. NO. 5, 1969