Androgen Blockade Does Not Affect Sleep-disordered Breathing or Chemosensitivity in Men with Obstructive Sleep Apnea 1 •3
DEIRDRE A. STEWART, RONALD R. GRUNSTEIN, MICHAEL BERTHON-JONES, DAVID J. HANDELSMAN, and COLIN E. SULLIVAN
Introduction
Sleep apnea is significantly more prevalent in men (1), although the reason for this is unclear. There are several studies demonstrating that testosterone or other androgen therapy may worsen or induce sleep apnea, suggesting that the male predisposition in this illness may be hormonal in origin (2, 3). Recent interest in the level of ventilatory chemosensitivity in patients with sleep apnea has emerged from studies demonstrating that high drives may destabilize breathing during sleep and increase apnea frequency (4, 5). In addition, increasing inspiratory effort lowers intrapharyngeal pressures, which may cause upper airway collapse. Men have higher ventilatory sensitivity than do women (6, 7), and testosterone therapy has an additional stimulatory influence on hypoxicbreathing responses (8).These data suggest that androgen-induced changes in respiratory control may be partially responsible for the sex differences in sleep-apnea prevalence and the induction of sleep-disordered breathing in some patients treated with testosterone. If androgen activity is pivotal in the pathogenesis of sleep-disordered breathing, withdrawal of this hormonal influence might significantly reduce the severity of the disorder. Peripheral and central androgen blockade is now achievable using specific receptor uptake and nuclear binding inhibitors. Accordingly, we designed a controlled trial evaluating the influence of androgen blockade on the frequency and the severity of sleepdisordered breathing found in men with the sleep-apnea syndrome. For this study we used flutamide (Eulexin; ScheringPlough, NSW,Australia), a nonsteroidal antiandrogen that acts as a competitive inhibitor of androgen binding to the androgen receptor both centrally and peripherally. Flutamide fully blocks exoge-
SUMMARY As sleep apnea Is more prevalent in men and testosterone has known effects on sleep apnea and Chemosensitivity, reduction of androgen activity may influence sleep-disordered breathing and respiratory control. Westudied the effect of 1 wk of treatment with flutamide, a nonsteroidal antlandrogen, on sleep, respiration, and ventilatory control in eight men with sleep apnea. Results on flutamlde were compared with two baseline studies performed before and after the drug treatment period. Although effective androgen blockade was achieved as evidenced by Increased hormone levels, flutamide had no effect on sleep architecture or chemoresponsiveness to hypoxia and hypercapnia. There was a trend towards a reduction in respiratory disturbance index In both NREM and REM sleep (41 ± 4 baseline versus 34 ± 3 f1utamlde, p = 0.09 NREM; 53 ± 4 baseline versus 48 ± 3 flutamide, p = 0.16 REM), but this was not significant. Our results Indicate that androgen blockade had no clinically significant effect on sleep, sleep-disordered breathing, or chemosensitivity In patients with moderate to severe sleep apnea. More specific blockers such as gonadotrophlnreleasing hormone analogs may have more clinical effect or, alternatively, androgen blockade may be more beneficial In patients with milder sleep apnea. AM REV RESPIR DIS 1992; 146:1389-1393
nous testosterone administered to castrated animals, and androgen blockade is virtually complete by I wk (9). Methods Subjects Eight male patients with sleep-apnea were recruited from the Sleep Disorders Centre at the Royal Prince Alfred Hospital. They were selected because of the absence of coexisting chronic lung or endocrine disease. These men demonstrated normal gonadal function, and they gaveinformed consent to the study, which was approved by the Ethics Committee of the Royal Prince Alfred Hospital. Those patients receiving steroids or other drugs affecting endocrine or respiratory function were excluded. The mean age of patients was 52 yr (range, 35 to 65 yr) and mean body mass index (BMI) was 32 (normal, 20 to 25) (BMI = weight in kilograms/height in meters').
Studies All subjects were studied on three occasions: baseline (Bl), after I wk of treatment with flutamide 250 mg three times a day (F), and again after allowing a 2-wk washout period after cessation of flutamide (B2). The study design was open, with both patients and investigators being aware of which week the drug was taken. On each occasion all subjects had full overnight sleep studies in the Royal Prince Alfred
Hospital Sleep Disorders Centre. Sleep was monitored with two channels of electroencephalogram (EEG) (C3/A2, 02lAI), two channels of electrooculogram (EOG) (ROC/ AI, WC/A2), and one channel of submental electromyogram (EMG). Sleep was staged according to standard criteria (10): Stages I and II, slow-wavesleep (NREM), and rapideye-movement (REM) sleep, and it was expressed as a percentage of total sleep time (TST). Sleep latency (time from lights out to sleeponset) and REM latency(time from sleep onset to first episode of REM sleep) werealso calculated. Airflow was measured using nasal prongs attached to a pressure transducer, rib cage and abdominal wall movement were recorded using inductive plethysomnography (Respitrace," Ambulatory Monitoring, Ardesley,
(Received in original form January 17, 1992 and in revised form July 1. 1992) 1 From the Sleep Disorders Centre, Royal Prince Alfred Hospital, Camperdown, and the David Read Laboratory and the Departments of Obstetrics and Gynecology, Department of Medicine, University of Sydney, New South Wales, Australia. 2 Supported by the National Health and by the Medical Research Council of Australia. 3 Correspondence and requests for reprints should be addressed to Dr. Ron Grunstein, David Read Laboratory, Department of Medicine, University of Sydney, Sydney, NSW 2006 Australia.
1389
1390
NY), and diaphragm EMG was measured using two electrodes placed in the right subcostal region, one at the midclavicular line and one 2 em lateral to it. Oxygen saturation was measured with an ear oximeter (Biox III; Ohmeda, Boulder, CO). All sleep and respiratory variables were recorded continuously on a 16-channel polygraph (Model 78; Grass Instruments, Quincy, MA). Apneas were defined as cessation of airflow for 10 s or longer. Hypopneas were defined as a decrease in the thoracoabdominal activity by at least 500/0 of baseline amplitude associated with a 40/0 decrease in oxygen saturation. The number of these episodes per hour of sleepwas calculated and expressed as a respiratory disturbance index (RDI) for both NREM and REM. For convenience all respiratory events will be called apneas. Apneas wereclassified as either central (absence of airflow associated with cessation of breathing as measured by inductive plethysmography and diaphragm EMG) or obstructive (absence of airflow associated with recorded breathing efforts and phasic diaphragm EMG activity). Mixed apneas (where absence of airflow associated with cessation of breathing effort in the first part of the event is followed by a period of chest wallmovement and phasic EMG activity) were separated into their central and obstructive components, and the time of each component measured. The duration of all obstructive and central apneas was measured and expressed as percentage of total sleep time, either central apnea time or obstructive apnea time. Oxygen saturation data were expressed in terms of mean minimum oxygen saturation (MMOS) in both NREM and REM, and the nadir oxygen saturation (the lowest oxygen saturation value recorded during the study).
Ventilatory Responses Ventilatory response testing to hypoxia and hypercapnia was performed after each overnight sleep study. Patients were seated comfortably in a quiet room and connected to the circuit using a standard mouthpiece with the nose occluded. The circuit used has been described previously (11). Arterial oxygen saturation (Sao,) was measured using a pulse oximeter and an ear probe, and end-tidal CO, (PETeo,) was measured using an infrared carbon dioxide analyzer (Hewlett-Packard, Waltham, MA). The hypoxic ventilatory response (HVR) was performed using the method originally described by Rebuck and Campbell (12)with a modification allowing a 2- to 3-min baseline stabilization period while the patient breathed 50% oxygen. Pure nitrogen wasthen added to the circuit at 4 L/min. There was a gradual decrease in Sao, over the following 3 to 5 min while PETeo, was maintained at resting values throughout the test by adjusting the amount of expired gas directed through the soda-lime absorber. The test was terminated when Sao, reached 80%. The breath-by-breath ventilation versus Sao,line was calculated by linear regression.
STEWART, GRUNSTEIN, BERTHON.JONES, HANDELSMAN, AND SULLIVAN
The hypercapnic ventilatory response (HCVR) was performed using the method originally described by Read (13). The test was terminated when PETeo, had risen 10mm Hg above the oxygenated mixed venous plateau (PVeo,). The HCVR is quantitated as the slope of the minute ventilation versus PETeo, relationship. The filtered flow, PETeo" and Sao, signals were recorded both on a 16channel polygraph (Model 78; Grass Instruments) and an IBM compatible AT computer with a 12-bit AID converter sampling at 125 Hz. From the flow signal the inspiratory and expiratory time, tidal volume, and minute ventilation were calculated for analysis.
the two baseline studies for each variable. All results are expressed as the mean value ± within-subject standard error of the mean (SEM) calculated from the ANOVA.
Results
Baseline to Baseline Variability We found no difference between the two baseline studies for any hormonal, sleep, sleep-breathing, or ventilatory response variables. There was no change over the study period in weight, alcohol consumption, or medications for any of the patients.
Endocrine Measurements On each study day we also measured baseline endocrine function at 6:30 A.M. The following hormones weremeasured; luteinizing hormone (LH), follicle-stimulating hormone (FSH), total and free testosterone, sexhormone-binding globulin (SHBG), cortisol, and dehydroepiandrosterone sulphate (DHAS) using previously described assay techniques (14).
Effect of Flutamide Endocrine data. All results are shown in table 1. There was a marked increase in plasma LH and FSH (p < 0.01)and total and free testosterone (p < 0.002) during treatment with flutamide. Other endocrine parameters (SHBG, DHAS, and cortisol) were unchanged with flutamide administration. Sleep. Changes in sleep architecture are shown in table 2. Flutamide had no effect on any sleep variables measured. Sleep-disordered breathing. Changes in respiratory parameters are shown in table 3. In seven of the eight patients studied there was a small decrease in the NREM RDI with flutamide (41 ± 4 baseline versus 34 ± 3 flutamide, p = 0.09) (figure 1), but this was not significant. The RDI in REM sleep was also de-
Data Analysis The ventilatory response data were analyzed using linear regression by plotting ventilation against oxygen saturation or end-tidal carbon dioxide levels. Effects of flutamide were analyzed using repeated measures ANOVA followed (where the omnibus F-test was significant) by Fisher's least significant difference for multiple planned comparisons. The mean of the two baseline values were calculated and compared with the value obtained after a week of flutamide. We also compared
TABLE 1 CHANGES IN ENDOCRINE MEASUREMENTS WITH FLUTAMIDE' Baseline t LH, lUlL FSH, lUlL Testosterone, nmollL Free testosterone, nmollL SHBG, nmol/L DHAS, umollL Cortisol, nmollL
4.4 5.2 15 334 21 5.4 264
± ± ± ± ± ± ±
0.5 0.5 2.3 63 1.1 0.7 26
Flutamide 7.5 6.5 24 556 22 4.31 282
± ± ± ± ± ± ±
0.4 0.4 1.6 50 1.6 0.5 19
p Value 0.001 0.01 0.002 0.002 0.51 0.09 0.43
Definition of abbreviations: LH = leutinizing hormone; FSH = follicle-stimulating hormone; SHBG sex-hormone-binding globulin; DHAS = dehydroepiandrosterone sulphate. • Values are mean ± within-subject SEM. t Baseline values are the mean values of the two baseline studies.
=
TABLE 2 CHANGES IN SLEEP ARCHITECTURE WITH AND WITHOUT FLUTAMIDE' Baseline Total sleep time, min Stage I sleep, min Stage II sleep, min Slow-wave sleep, min REM sleep, min Sleep latency, min REM latency, min
388 81 216 38 52 24 175
• Values are mean ± within-subject SEM.
± 18 ± 17
± 28 12
± ± ± ±
7 11 32
Flutamide 387 79 231 27 49
± 13 ± 12 ± 19 ± 8
± 5
6 ± 8 141 ± 32
p Value 0.92 0.88 0.53 0.31 0.60 0.08 0.24
1391
ANDROGEN BLOCKADE DOES NOT AFFECT SLEEP-DISORDERED BREATHING IN OSA
TABLE 3 CHANGES IN SLEEP·BREATHING VARIABLES WITH AND WITHOUT FLUTAMIDE* Baseline NREM RDI, nlhr REM RDI, nlhr Central apnea time, min Obstructive apnea time, min MMOS NREM, % MMOS REM, % Nadir Sao" %
41 53 16 129 86 79 73
± ± ± ± ± ± ±
4 4 5 22 2 2 2
Flutamide 34 48 11 116 87 81 74
± ± ± ± ± ± ±
p Value
3 3 3 16 1 2 2
0.09 0.16 0.31 0.52 0.63 0.57 0.62
Definition of abbreviations: ADI = respiratory disturbance index; MMOS = mean minimum oxygen saturation. • Values are mean ± within-subject SEM.
creased (53 ± 4 baseline versus 48 ± 3 flutamide, p = 0.16), with six subjects showing a decrease, two with no change in RDI, and one with an increase in RDI (figure 2). Despite these results there was no change in the total amount of time spent in obstructive and central apnea, mean minimal oxygen saturation in
70
•
60 s: 0
c
40
"a:z
30
-
~
50
0
•
a:
UJ
20
10
61
62
Fig. 1. Individual changes in NREM RDI during the two baseline studies (B1 and B2) and when patients were receiving f1utamide (F).
80
70
s:
NREM and REM sleep, and nadir oxygen saturation.
Ventilatory Responses to Hypoxia and Hypercapnia Results are shown in table 4. There was no change in the HVR with flutamide. The PETC0 2 at which the test was done was slightly lower (37 ± 1.7mm Hg baseline versus 35 ± 1.2 mm Hg flutamide), but this was not significant. The HCVR was unchanged with flutamide as was the PVC02'
Discussion
Our results indicate that androgen blockade with flutamide had no clinically significant effect on sleep-disordered breathing. Androgen blockade was demonstrated by a rise in total and free testosterone caused by interference of testosterone central negative feedback. Tho studies (15, 16) have demonstrated that testosterone therapy in hypogonadal men leads to a significant increase in sleep apnea in some of them. Others (2, 3) have observed increased sleep apnea in male and female patients with normal gonadal function receiving supplemental androgens that was reversed with discontinuation of the drug. However, the reasons for these observations as well as the increased prevalence of sleep apnea in men are unknown.
Our hypotheses, based on these studies, was that blockade of androgen receptors would reduce sleep-disordered breathing in men with sleep apnea. A number of reasons may explain the lack of effect of flutamide on sleep apnea despite the presence of an underlying hormonal effect. Sleep apnea is a disorder that typically evolves over many years. The male preponderance may be due to changes in upper airway structure and function induced by male sex steroids. Clearly, the time of life when the greatest changes occur is at puberty, and it may be against this background of pubertal change that sleep apnea subsequently develops in the male. The deepening of the male voice at puberty is evidence of a major change in upper airway structure and function in at least one upper airway location in response to androgens. Thus, the absence of an effect of flutamide on sleep apnea may simply reflect the fact that the dominant androgenic effect in sleep apnea occurs at puberty. This is consistent with the finding that pharyngeal resistance is not measurably influenced by androgen replacement in hypogonadal male adults (16). Androgens have direct neuronal effects in brain regions not related to reproductive function. For example, testosterone also accelerates the regeneration of the crushed hypoglossal nerve in the young adult and the 4-wk-old rat but not in prepubertal rats (17, 18). Androgen receptors have been found in rat hypoglossal and facial nuclei as wellas in tongue muscles (19). Further, there are certain regions of the brain in which neurones take up both estrogen and testosterone hormones. Notably, testosterone is selectively concentrated in neurons that have a somatomotor function, whereas estrogen concentration prevails in neurons known to modulate sensory perception (20). Therefore, it might be predicted that sex steroids or their receptor blockade would have a relativelyimmediate effect on neu-
60
0
c
TABLE 4 50
CHANGES IN VENTILATORY RESPONSE DATA WITH AND WITHOUT FLUTAMIDE*
0
a:
40
"a:
Baseline
UJ
30
HCVR slope, Umin/mm Hg Pvco" mm Hg HVR, Umin/% PETCO" mm Hg
20
a l' 61
2.5 57 0.9 37
± ± ± ±
1.1 6.4 0.4 1.7
Flutamide 3.3 56 1.0 35
± ± ± ±
0.7 4.4 0.3 1.2
p Value 0.39 0.88 0.74 0.15
62
Fig. 2. Individual changes in the REM RDI during the two baseline studies (Bl and B2)and when patientswere receiving f1utamide (F).
Definition of abbreviations: HCVA = hypercapnic ventilatory response; Plica, = mixed venous carbon dioxide pressure; HVA = hypoxic ventilatory response; PETco, = end-tidal carbon dioxide pressure . • Values are mean ± within-subject SEM.
1392
ronal function. Our observation that flutamide, an androgen receptor blocker, did not alter the degree of sleep apnea strongly implies that androgens do not significantly alter, for example, the neural output to the upper airway. It may be that structural changes, or effects on neurons take longer than I wk to occur. Thus, flutamide may have had an effect if administered for a longer period than 1 wk. However, we did not use flutamide for a longer period because of potential side effects. Another possibility is that the androgen blockade produced by an antiandrogen such as flutamide was inadequate or potentially overcome by the reflex rise in gonadotrophins (and consequently testosterone) that we observed (21). This is due to operation of the closed loop negative feedback system and limits the degree of androgen withdrawal experienced. However, the dose of flutamide we used was similar to that currently used in treating advanced prostatic cancer. Flutamide has a rapid onset of action, and there is a complete central hormonal effect with maximal rise in LH and FSH within 1 wk (22). The existence of high LH, FSH, and testosterone levels at the end of the drug phase of the study indicated that the reflex testosterone increase to flutamide did not fully compensate for the androgen deprivation and that the patients did experience persistent androgen deficiency in this phase of the study. It is possible that the lack of effect of flutamide may be due to its incomplete degree of androgen blockade. Interruption of androgen secretion at other points in the hypothalamic-pituitary-testicular axis such as by blockade of gonadotrophin-releasing hormone (GnRH) action would result in a greater degree of androgen deprivation. As a result, experimental studies of GnRH analogs in sleep apnea would be of interest. The individual reports of testosterone induction of sleep apnea therefore require another explanation. Although testosterone use in hypogonadal men produced worsening of sleep apnea in a relatively short period of time, it typically occurred in men who were either heavy snorers or had preexisting sleep apnea. Our patients all had at least moderately severe sleep apnea. It is possible that men with only heavy snoring or mild forms of sleep apnea may have been more responsive to flutamide, Of interest is the fact that we observed a consistent but small decrease in the number of respiratory events per hour. However, this change was statistically insignificant (table 3) and almost
STEWART, GRUNSTE1N, SERTHON.JONES, HANDELSMAN, AND SULLIVAN
certainly of no clinical relevance. Even by pooling our baseline data and hence reducing our statistical power, our study with eight subjects had a 950/0 chance of detecting a 40% reduction in RDI (twotailed alpha = 0.05). Clinical response to therapy in sleep apnea is usually at least a 50% reduction in RDI with surgical therapy or complete abolition of apnea with CPAP therapy. . Interestingly, recent work (14, 23) has demonstrated that testosterone levels are lower in patients with sleep apnea than in control subjects, which may be consistent with the suggestion that hypogonadism may be a protection against sleep apnea (22). However, in these previous studies of testosterone levels in sleep apnea (14,24) and in this current study, the testosterone levels of the patients were still within the physiologic range well above castrate levels. Therefore, there would have been adequate levels of testosterone in these patients to allow effective androgen blockade. Another potential mechanism by which sex steroids may affect sleep apnea is by influencing ventilatory control. For example progesterone increases ventilation and chemosensitivity (25,26). In patients with alveolar hypoventilation and sleep apnea, progesterone can improve ventilation (27). Despite these ventilatory effects and the reduction in testosterone levels after progesterone use (28), its effects on the degree of apnea in men are relatively minor. However, progesterone combined with estrogen ameliorates sleepdisordered breathing in postmenopausal women (29, 30). Testosterone administration in hypogonadal men increases minute ventilation and the ventilatory response to hypoxia, but not the ventilatory response to hypercapnia (8, IS). Any drug that increases ventilatory responsiveness may worsen or induce apnea by driving arterial CO 2 below the apnea threshold (31). Androgen blockade with flutamide did not alter hypercapnic or hypoxic chemosensitivity. This is in marked contrast to the changes in ventilatory control induced by testosterone in hypogonadal men. Thus, the effect of testosterone on ventilatory control may not have acted through the receptor mechanisms blocked by flutamide. Whatever the mechanism, the lack of effect of flutamide on ventilatory control and the degree of sleep apnea suggests that the reported worsening of apnea induced by testosterone may be due to as yet unknown changes in upper air-
way function or effects on ventilatory control in sleep rather than wakefulness. Other forms of androgen blockade or withdrawal are required to fully exclude testosterone effects on sleep and breathing. Acknowledgment
The writers thank the staff of the Sleep Disorders Centre and the Endocrinology Laboratory at Royal Prince Alfred Hospital for their assistance, and Professor Clifford Zwillich for his help in preparation of the manuscript. References I. Block AJ, Boysen PG, Wynne JW, Hunt LA. Sleep apnea, hypopnea and oxygen desaturation in normal subjects. N Engl J Med 1979;300:513-7. 2. Johnson MW, Anch AM, Remmers JE. Induction of the obstructive sleep apnea syndrome in a woman by exogenous androgen administration. Am Rev Respir Dis 1984; 129:1023-5. 3. Sandblom RE, Matsumoto AM, Schoene RB, et at. Obstructive sleep apnea syndrome induced by testosterone administration. N Engl J Med 1983; 308:508-10. 4. Sullivan C, Scheibner T, Parker S, Grunstein R, Ho K. Pathophysiology of sleep apnea; a search for neurochemical defects. Prog Clin Bioi Res 1990; 345:325-34. 5. Cherniack NS. Sleep apnea and its causes. J Clin Invest 1984; 73:1501-5. 6. White DP, Douglas NJ, Pickett CK, Weil JV, Zwillich CWO Sexual influence on the control of breathing. J Appl Physiol 1983; 54:874-9. 7. McCauley VB, Grunstein RR, Sullivan CEo Ethanol-induced depression of hypoxic drive and reversal by naloxone: a sex difference. Am Rev Respir Dis 1988; 137:1406-10. 8. White DP, Schneider BK, Santen RJ, et al. Influence of testosterone on ventilation and chemosensitivity in male subjects. J Appl Physiol1985; 59:1452-7. 9. Brogden RN, Clissold SP. Flutamide. A preliminary reviewof its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in advanced prostatic cancer. Drugs 1989; 38: 185-203. 10. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Bethesda: National Institute for Neurological Disease and Blindness, 1968 (NIH publication no. 204). 11. Berthon-Jones M, SullivanCEoVentilatoryand arousal responses to hypoxia in sleeping humans. Am Rev Respir Dis 1982; 125:632-9. 12. Rebuck AS, Campbell EJ. A clinical method for assessing the ventilatory response to hypoxia. Am Rev Respir Dis 1973; 109:345-50. 13. Read DJC. A clinical method for assessing the ventilatory response to carbon dioxide. Australas Ann Med 1967; 16:20-32. 14. Grunstein RR, Handelsman DJ, Lawrence SJ, Blackwell C, Caterson ID, Sullivan CEo Neuroendocrine dysfunction in sleep apnea: reversalby continuous positive airway pressure therapy. J Clin Endocrinol Metab 1989; 68:352-8. 15. Matsumoto AM, Sandblom RE, Schoene RB, et at. Testosterone replacement in hypogonadal males: effects on obstructive sleep apnea, respiratory drives and sleep. Clin Endocrinol (Oxf) 1985; 22:713-21.
ANDROGEN BLOCKADE DOES NOT AFFECT SLEEP-DISORDERED BREATHING IN OSA
16. Schneider BK, Pickett CK, Zwillich CW, et al. Influence of testosterone on breathing during sleep. J Appl Physiol 1986; 61:618-23. 17. Yu WH, Yu MC. Acceleration of the regeneration of the crushed hypoglossal nerve by testosterone. Exp Neurol 1983; 80:349-60. 18. Yu WHo Responsiveness of hypoglossal neurons to testosterone in pre-pubertal rats. Brain Res Bull 1984; 13:667-72. 19. Yu WH, McGinnis MY. Androgen receptor levels in cranial nerve nuclei and tongue muscles in rats. J Neurosci 1986; 6:1302-7. 20. Strumpf WE, Sar M. Steroid hormone target cells in the periventricular brain: relationship to peptide hormone producing cells. Fed Proc 1977; 36:1973-7. 21. Mowszowicz I. Antiandrogens: mechanisms and paradoxical effects. Ann Endocrinol (Paris) 1989; 50:189-99.
22. Knuth VA, Hano R, Nieschlag E. Effect of flutamide or cyproterone on pituitary and testicular hormones in normal men. J Clin Endocrinol Metab 1984; 59:963-9. 23. Harman E, Wynne JW, Block AJ, MalloyFisher L. Sleep disordered breathing and oxygen desaturation in obese patients. Chest 1981; 70: 256-60. 24. Santamaria JD, Prior JC, Fleetham JA. Reversible reproductive dysfunction in men with obstructive sleep apnea. Clin Endocrinol 1988; 28:461-70. 25. Zwillich CW, Natalino MR, Sutton FD, Weil JV. Effects of progesterone on chemosensitivity in normal men. J Lab Clin Med 1978; 92:262-9. 26. Morikawa T, Tanaka Y, Maruyama R, Nishibayashi Y, Honda Y. Comparison of two synthetic progesterones on ventilation in normal males: CMA vs. MPA. J Appl Physiol 1987; 63:1610-5.
1393
27. Lyons HA, Huang CT. Therapeutic use of progesterone in alveolar hypoventilation associated with obesity. Am J Med 1968; 44:881-8. 28. Smith KD, Rodriguez-Rigau J. Laboratory evaluation of testicular function. In: De Groot L, ed. Endocrinology. Philadelphia: W. B. Saunders, 1989; 2169. 29. Pickett CK, Regensteiner JG, Woodard WD, Hagerman DD, Weil JV, Grindlay Moore L. Progestin and estrogen reduce sleep-disordered breathing in postmenopausal women. J Appl Physiol 1989; 66:1656-61. 30. Franklin K, Lundgren R, Rabben T. Sleep apnea syndrome treated with oestradial and cyclic medroxyprogesterone. Lancet 1991; 338:252-3. 31. Dempsey JA, Skatrud JB. A sleep-induced apneic threshold and its consequences. Am Rev Respir Dis 1986; 133:1163-70.
DEIRDRE A. STEWART, RONALD R. GRUNSTEIN, MICHAEL BERTHON-JONES, DAVID J. HANDELSMAN, and COLIN E. SULLIVAN
Introduction
Sleep apnea is significantly more prevalent in men (1), although the reason for this is unclear. There are several studies demonstrating that testosterone or other androgen therapy may worsen or induce sleep apnea, suggesting that the male predisposition in this illness may be hormonal in origin (2, 3). Recent interest in the level of ventilatory chemosensitivity in patients with sleep apnea has emerged from studies demonstrating that high drives may destabilize breathing during sleep and increase apnea frequency (4, 5). In addition, increasing inspiratory effort lowers intrapharyngeal pressures, which may cause upper airway collapse. Men have higher ventilatory sensitivity than do women (6, 7), and testosterone therapy has an additional stimulatory influence on hypoxicbreathing responses (8).These data suggest that androgen-induced changes in respiratory control may be partially responsible for the sex differences in sleep-apnea prevalence and the induction of sleep-disordered breathing in some patients treated with testosterone. If androgen activity is pivotal in the pathogenesis of sleep-disordered breathing, withdrawal of this hormonal influence might significantly reduce the severity of the disorder. Peripheral and central androgen blockade is now achievable using specific receptor uptake and nuclear binding inhibitors. Accordingly, we designed a controlled trial evaluating the influence of androgen blockade on the frequency and the severity of sleepdisordered breathing found in men with the sleep-apnea syndrome. For this study we used flutamide (Eulexin; ScheringPlough, NSW,Australia), a nonsteroidal antiandrogen that acts as a competitive inhibitor of androgen binding to the androgen receptor both centrally and peripherally. Flutamide fully blocks exoge-
SUMMARY As sleep apnea Is more prevalent in men and testosterone has known effects on sleep apnea and Chemosensitivity, reduction of androgen activity may influence sleep-disordered breathing and respiratory control. Westudied the effect of 1 wk of treatment with flutamide, a nonsteroidal antlandrogen, on sleep, respiration, and ventilatory control in eight men with sleep apnea. Results on flutamlde were compared with two baseline studies performed before and after the drug treatment period. Although effective androgen blockade was achieved as evidenced by Increased hormone levels, flutamide had no effect on sleep architecture or chemoresponsiveness to hypoxia and hypercapnia. There was a trend towards a reduction in respiratory disturbance index In both NREM and REM sleep (41 ± 4 baseline versus 34 ± 3 f1utamlde, p = 0.09 NREM; 53 ± 4 baseline versus 48 ± 3 flutamide, p = 0.16 REM), but this was not significant. Our results Indicate that androgen blockade had no clinically significant effect on sleep, sleep-disordered breathing, or chemosensitivity In patients with moderate to severe sleep apnea. More specific blockers such as gonadotrophlnreleasing hormone analogs may have more clinical effect or, alternatively, androgen blockade may be more beneficial In patients with milder sleep apnea. AM REV RESPIR DIS 1992; 146:1389-1393
nous testosterone administered to castrated animals, and androgen blockade is virtually complete by I wk (9). Methods Subjects Eight male patients with sleep-apnea were recruited from the Sleep Disorders Centre at the Royal Prince Alfred Hospital. They were selected because of the absence of coexisting chronic lung or endocrine disease. These men demonstrated normal gonadal function, and they gaveinformed consent to the study, which was approved by the Ethics Committee of the Royal Prince Alfred Hospital. Those patients receiving steroids or other drugs affecting endocrine or respiratory function were excluded. The mean age of patients was 52 yr (range, 35 to 65 yr) and mean body mass index (BMI) was 32 (normal, 20 to 25) (BMI = weight in kilograms/height in meters').
Studies All subjects were studied on three occasions: baseline (Bl), after I wk of treatment with flutamide 250 mg three times a day (F), and again after allowing a 2-wk washout period after cessation of flutamide (B2). The study design was open, with both patients and investigators being aware of which week the drug was taken. On each occasion all subjects had full overnight sleep studies in the Royal Prince Alfred
Hospital Sleep Disorders Centre. Sleep was monitored with two channels of electroencephalogram (EEG) (C3/A2, 02lAI), two channels of electrooculogram (EOG) (ROC/ AI, WC/A2), and one channel of submental electromyogram (EMG). Sleep was staged according to standard criteria (10): Stages I and II, slow-wavesleep (NREM), and rapideye-movement (REM) sleep, and it was expressed as a percentage of total sleep time (TST). Sleep latency (time from lights out to sleeponset) and REM latency(time from sleep onset to first episode of REM sleep) werealso calculated. Airflow was measured using nasal prongs attached to a pressure transducer, rib cage and abdominal wall movement were recorded using inductive plethysomnography (Respitrace," Ambulatory Monitoring, Ardesley,
(Received in original form January 17, 1992 and in revised form July 1. 1992) 1 From the Sleep Disorders Centre, Royal Prince Alfred Hospital, Camperdown, and the David Read Laboratory and the Departments of Obstetrics and Gynecology, Department of Medicine, University of Sydney, New South Wales, Australia. 2 Supported by the National Health and by the Medical Research Council of Australia. 3 Correspondence and requests for reprints should be addressed to Dr. Ron Grunstein, David Read Laboratory, Department of Medicine, University of Sydney, Sydney, NSW 2006 Australia.
1389
1390
NY), and diaphragm EMG was measured using two electrodes placed in the right subcostal region, one at the midclavicular line and one 2 em lateral to it. Oxygen saturation was measured with an ear oximeter (Biox III; Ohmeda, Boulder, CO). All sleep and respiratory variables were recorded continuously on a 16-channel polygraph (Model 78; Grass Instruments, Quincy, MA). Apneas were defined as cessation of airflow for 10 s or longer. Hypopneas were defined as a decrease in the thoracoabdominal activity by at least 500/0 of baseline amplitude associated with a 40/0 decrease in oxygen saturation. The number of these episodes per hour of sleepwas calculated and expressed as a respiratory disturbance index (RDI) for both NREM and REM. For convenience all respiratory events will be called apneas. Apneas wereclassified as either central (absence of airflow associated with cessation of breathing as measured by inductive plethysmography and diaphragm EMG) or obstructive (absence of airflow associated with recorded breathing efforts and phasic diaphragm EMG activity). Mixed apneas (where absence of airflow associated with cessation of breathing effort in the first part of the event is followed by a period of chest wallmovement and phasic EMG activity) were separated into their central and obstructive components, and the time of each component measured. The duration of all obstructive and central apneas was measured and expressed as percentage of total sleep time, either central apnea time or obstructive apnea time. Oxygen saturation data were expressed in terms of mean minimum oxygen saturation (MMOS) in both NREM and REM, and the nadir oxygen saturation (the lowest oxygen saturation value recorded during the study).
Ventilatory Responses Ventilatory response testing to hypoxia and hypercapnia was performed after each overnight sleep study. Patients were seated comfortably in a quiet room and connected to the circuit using a standard mouthpiece with the nose occluded. The circuit used has been described previously (11). Arterial oxygen saturation (Sao,) was measured using a pulse oximeter and an ear probe, and end-tidal CO, (PETeo,) was measured using an infrared carbon dioxide analyzer (Hewlett-Packard, Waltham, MA). The hypoxic ventilatory response (HVR) was performed using the method originally described by Rebuck and Campbell (12)with a modification allowing a 2- to 3-min baseline stabilization period while the patient breathed 50% oxygen. Pure nitrogen wasthen added to the circuit at 4 L/min. There was a gradual decrease in Sao, over the following 3 to 5 min while PETeo, was maintained at resting values throughout the test by adjusting the amount of expired gas directed through the soda-lime absorber. The test was terminated when Sao, reached 80%. The breath-by-breath ventilation versus Sao,line was calculated by linear regression.
STEWART, GRUNSTEIN, BERTHON.JONES, HANDELSMAN, AND SULLIVAN
The hypercapnic ventilatory response (HCVR) was performed using the method originally described by Read (13). The test was terminated when PETeo, had risen 10mm Hg above the oxygenated mixed venous plateau (PVeo,). The HCVR is quantitated as the slope of the minute ventilation versus PETeo, relationship. The filtered flow, PETeo" and Sao, signals were recorded both on a 16channel polygraph (Model 78; Grass Instruments) and an IBM compatible AT computer with a 12-bit AID converter sampling at 125 Hz. From the flow signal the inspiratory and expiratory time, tidal volume, and minute ventilation were calculated for analysis.
the two baseline studies for each variable. All results are expressed as the mean value ± within-subject standard error of the mean (SEM) calculated from the ANOVA.
Results
Baseline to Baseline Variability We found no difference between the two baseline studies for any hormonal, sleep, sleep-breathing, or ventilatory response variables. There was no change over the study period in weight, alcohol consumption, or medications for any of the patients.
Endocrine Measurements On each study day we also measured baseline endocrine function at 6:30 A.M. The following hormones weremeasured; luteinizing hormone (LH), follicle-stimulating hormone (FSH), total and free testosterone, sexhormone-binding globulin (SHBG), cortisol, and dehydroepiandrosterone sulphate (DHAS) using previously described assay techniques (14).
Effect of Flutamide Endocrine data. All results are shown in table 1. There was a marked increase in plasma LH and FSH (p < 0.01)and total and free testosterone (p < 0.002) during treatment with flutamide. Other endocrine parameters (SHBG, DHAS, and cortisol) were unchanged with flutamide administration. Sleep. Changes in sleep architecture are shown in table 2. Flutamide had no effect on any sleep variables measured. Sleep-disordered breathing. Changes in respiratory parameters are shown in table 3. In seven of the eight patients studied there was a small decrease in the NREM RDI with flutamide (41 ± 4 baseline versus 34 ± 3 flutamide, p = 0.09) (figure 1), but this was not significant. The RDI in REM sleep was also de-
Data Analysis The ventilatory response data were analyzed using linear regression by plotting ventilation against oxygen saturation or end-tidal carbon dioxide levels. Effects of flutamide were analyzed using repeated measures ANOVA followed (where the omnibus F-test was significant) by Fisher's least significant difference for multiple planned comparisons. The mean of the two baseline values were calculated and compared with the value obtained after a week of flutamide. We also compared
TABLE 1 CHANGES IN ENDOCRINE MEASUREMENTS WITH FLUTAMIDE' Baseline t LH, lUlL FSH, lUlL Testosterone, nmollL Free testosterone, nmollL SHBG, nmol/L DHAS, umollL Cortisol, nmollL
4.4 5.2 15 334 21 5.4 264
± ± ± ± ± ± ±
0.5 0.5 2.3 63 1.1 0.7 26
Flutamide 7.5 6.5 24 556 22 4.31 282
± ± ± ± ± ± ±
0.4 0.4 1.6 50 1.6 0.5 19
p Value 0.001 0.01 0.002 0.002 0.51 0.09 0.43
Definition of abbreviations: LH = leutinizing hormone; FSH = follicle-stimulating hormone; SHBG sex-hormone-binding globulin; DHAS = dehydroepiandrosterone sulphate. • Values are mean ± within-subject SEM. t Baseline values are the mean values of the two baseline studies.
=
TABLE 2 CHANGES IN SLEEP ARCHITECTURE WITH AND WITHOUT FLUTAMIDE' Baseline Total sleep time, min Stage I sleep, min Stage II sleep, min Slow-wave sleep, min REM sleep, min Sleep latency, min REM latency, min
388 81 216 38 52 24 175
• Values are mean ± within-subject SEM.
± 18 ± 17
± 28 12
± ± ± ±
7 11 32
Flutamide 387 79 231 27 49
± 13 ± 12 ± 19 ± 8
± 5
6 ± 8 141 ± 32
p Value 0.92 0.88 0.53 0.31 0.60 0.08 0.24
1391
ANDROGEN BLOCKADE DOES NOT AFFECT SLEEP-DISORDERED BREATHING IN OSA
TABLE 3 CHANGES IN SLEEP·BREATHING VARIABLES WITH AND WITHOUT FLUTAMIDE* Baseline NREM RDI, nlhr REM RDI, nlhr Central apnea time, min Obstructive apnea time, min MMOS NREM, % MMOS REM, % Nadir Sao" %
41 53 16 129 86 79 73
± ± ± ± ± ± ±
4 4 5 22 2 2 2
Flutamide 34 48 11 116 87 81 74
± ± ± ± ± ± ±
p Value
3 3 3 16 1 2 2
0.09 0.16 0.31 0.52 0.63 0.57 0.62
Definition of abbreviations: ADI = respiratory disturbance index; MMOS = mean minimum oxygen saturation. • Values are mean ± within-subject SEM.
creased (53 ± 4 baseline versus 48 ± 3 flutamide, p = 0.16), with six subjects showing a decrease, two with no change in RDI, and one with an increase in RDI (figure 2). Despite these results there was no change in the total amount of time spent in obstructive and central apnea, mean minimal oxygen saturation in
70
•
60 s: 0
c
40
"a:z
30
-
~
50
0
•
a:
UJ
20
10
61
62
Fig. 1. Individual changes in NREM RDI during the two baseline studies (B1 and B2) and when patients were receiving f1utamide (F).
80
70
s:
NREM and REM sleep, and nadir oxygen saturation.
Ventilatory Responses to Hypoxia and Hypercapnia Results are shown in table 4. There was no change in the HVR with flutamide. The PETC0 2 at which the test was done was slightly lower (37 ± 1.7mm Hg baseline versus 35 ± 1.2 mm Hg flutamide), but this was not significant. The HCVR was unchanged with flutamide as was the PVC02'
Discussion
Our results indicate that androgen blockade with flutamide had no clinically significant effect on sleep-disordered breathing. Androgen blockade was demonstrated by a rise in total and free testosterone caused by interference of testosterone central negative feedback. Tho studies (15, 16) have demonstrated that testosterone therapy in hypogonadal men leads to a significant increase in sleep apnea in some of them. Others (2, 3) have observed increased sleep apnea in male and female patients with normal gonadal function receiving supplemental androgens that was reversed with discontinuation of the drug. However, the reasons for these observations as well as the increased prevalence of sleep apnea in men are unknown.
Our hypotheses, based on these studies, was that blockade of androgen receptors would reduce sleep-disordered breathing in men with sleep apnea. A number of reasons may explain the lack of effect of flutamide on sleep apnea despite the presence of an underlying hormonal effect. Sleep apnea is a disorder that typically evolves over many years. The male preponderance may be due to changes in upper airway structure and function induced by male sex steroids. Clearly, the time of life when the greatest changes occur is at puberty, and it may be against this background of pubertal change that sleep apnea subsequently develops in the male. The deepening of the male voice at puberty is evidence of a major change in upper airway structure and function in at least one upper airway location in response to androgens. Thus, the absence of an effect of flutamide on sleep apnea may simply reflect the fact that the dominant androgenic effect in sleep apnea occurs at puberty. This is consistent with the finding that pharyngeal resistance is not measurably influenced by androgen replacement in hypogonadal male adults (16). Androgens have direct neuronal effects in brain regions not related to reproductive function. For example, testosterone also accelerates the regeneration of the crushed hypoglossal nerve in the young adult and the 4-wk-old rat but not in prepubertal rats (17, 18). Androgen receptors have been found in rat hypoglossal and facial nuclei as wellas in tongue muscles (19). Further, there are certain regions of the brain in which neurones take up both estrogen and testosterone hormones. Notably, testosterone is selectively concentrated in neurons that have a somatomotor function, whereas estrogen concentration prevails in neurons known to modulate sensory perception (20). Therefore, it might be predicted that sex steroids or their receptor blockade would have a relativelyimmediate effect on neu-
60
0
c
TABLE 4 50
CHANGES IN VENTILATORY RESPONSE DATA WITH AND WITHOUT FLUTAMIDE*
0
a:
40
"a:
Baseline
UJ
30
HCVR slope, Umin/mm Hg Pvco" mm Hg HVR, Umin/% PETCO" mm Hg
20
a l' 61
2.5 57 0.9 37
± ± ± ±
1.1 6.4 0.4 1.7
Flutamide 3.3 56 1.0 35
± ± ± ±
0.7 4.4 0.3 1.2
p Value 0.39 0.88 0.74 0.15
62
Fig. 2. Individual changes in the REM RDI during the two baseline studies (Bl and B2)and when patientswere receiving f1utamide (F).
Definition of abbreviations: HCVA = hypercapnic ventilatory response; Plica, = mixed venous carbon dioxide pressure; HVA = hypoxic ventilatory response; PETco, = end-tidal carbon dioxide pressure . • Values are mean ± within-subject SEM.
1392
ronal function. Our observation that flutamide, an androgen receptor blocker, did not alter the degree of sleep apnea strongly implies that androgens do not significantly alter, for example, the neural output to the upper airway. It may be that structural changes, or effects on neurons take longer than I wk to occur. Thus, flutamide may have had an effect if administered for a longer period than 1 wk. However, we did not use flutamide for a longer period because of potential side effects. Another possibility is that the androgen blockade produced by an antiandrogen such as flutamide was inadequate or potentially overcome by the reflex rise in gonadotrophins (and consequently testosterone) that we observed (21). This is due to operation of the closed loop negative feedback system and limits the degree of androgen withdrawal experienced. However, the dose of flutamide we used was similar to that currently used in treating advanced prostatic cancer. Flutamide has a rapid onset of action, and there is a complete central hormonal effect with maximal rise in LH and FSH within 1 wk (22). The existence of high LH, FSH, and testosterone levels at the end of the drug phase of the study indicated that the reflex testosterone increase to flutamide did not fully compensate for the androgen deprivation and that the patients did experience persistent androgen deficiency in this phase of the study. It is possible that the lack of effect of flutamide may be due to its incomplete degree of androgen blockade. Interruption of androgen secretion at other points in the hypothalamic-pituitary-testicular axis such as by blockade of gonadotrophin-releasing hormone (GnRH) action would result in a greater degree of androgen deprivation. As a result, experimental studies of GnRH analogs in sleep apnea would be of interest. The individual reports of testosterone induction of sleep apnea therefore require another explanation. Although testosterone use in hypogonadal men produced worsening of sleep apnea in a relatively short period of time, it typically occurred in men who were either heavy snorers or had preexisting sleep apnea. Our patients all had at least moderately severe sleep apnea. It is possible that men with only heavy snoring or mild forms of sleep apnea may have been more responsive to flutamide, Of interest is the fact that we observed a consistent but small decrease in the number of respiratory events per hour. However, this change was statistically insignificant (table 3) and almost
STEWART, GRUNSTE1N, SERTHON.JONES, HANDELSMAN, AND SULLIVAN
certainly of no clinical relevance. Even by pooling our baseline data and hence reducing our statistical power, our study with eight subjects had a 950/0 chance of detecting a 40% reduction in RDI (twotailed alpha = 0.05). Clinical response to therapy in sleep apnea is usually at least a 50% reduction in RDI with surgical therapy or complete abolition of apnea with CPAP therapy. . Interestingly, recent work (14, 23) has demonstrated that testosterone levels are lower in patients with sleep apnea than in control subjects, which may be consistent with the suggestion that hypogonadism may be a protection against sleep apnea (22). However, in these previous studies of testosterone levels in sleep apnea (14,24) and in this current study, the testosterone levels of the patients were still within the physiologic range well above castrate levels. Therefore, there would have been adequate levels of testosterone in these patients to allow effective androgen blockade. Another potential mechanism by which sex steroids may affect sleep apnea is by influencing ventilatory control. For example progesterone increases ventilation and chemosensitivity (25,26). In patients with alveolar hypoventilation and sleep apnea, progesterone can improve ventilation (27). Despite these ventilatory effects and the reduction in testosterone levels after progesterone use (28), its effects on the degree of apnea in men are relatively minor. However, progesterone combined with estrogen ameliorates sleepdisordered breathing in postmenopausal women (29, 30). Testosterone administration in hypogonadal men increases minute ventilation and the ventilatory response to hypoxia, but not the ventilatory response to hypercapnia (8, IS). Any drug that increases ventilatory responsiveness may worsen or induce apnea by driving arterial CO 2 below the apnea threshold (31). Androgen blockade with flutamide did not alter hypercapnic or hypoxic chemosensitivity. This is in marked contrast to the changes in ventilatory control induced by testosterone in hypogonadal men. Thus, the effect of testosterone on ventilatory control may not have acted through the receptor mechanisms blocked by flutamide. Whatever the mechanism, the lack of effect of flutamide on ventilatory control and the degree of sleep apnea suggests that the reported worsening of apnea induced by testosterone may be due to as yet unknown changes in upper air-
way function or effects on ventilatory control in sleep rather than wakefulness. Other forms of androgen blockade or withdrawal are required to fully exclude testosterone effects on sleep and breathing. Acknowledgment
The writers thank the staff of the Sleep Disorders Centre and the Endocrinology Laboratory at Royal Prince Alfred Hospital for their assistance, and Professor Clifford Zwillich for his help in preparation of the manuscript. References I. Block AJ, Boysen PG, Wynne JW, Hunt LA. Sleep apnea, hypopnea and oxygen desaturation in normal subjects. N Engl J Med 1979;300:513-7. 2. Johnson MW, Anch AM, Remmers JE. Induction of the obstructive sleep apnea syndrome in a woman by exogenous androgen administration. Am Rev Respir Dis 1984; 129:1023-5. 3. Sandblom RE, Matsumoto AM, Schoene RB, et at. Obstructive sleep apnea syndrome induced by testosterone administration. N Engl J Med 1983; 308:508-10. 4. Sullivan C, Scheibner T, Parker S, Grunstein R, Ho K. Pathophysiology of sleep apnea; a search for neurochemical defects. Prog Clin Bioi Res 1990; 345:325-34. 5. Cherniack NS. Sleep apnea and its causes. J Clin Invest 1984; 73:1501-5. 6. White DP, Douglas NJ, Pickett CK, Weil JV, Zwillich CWO Sexual influence on the control of breathing. J Appl Physiol 1983; 54:874-9. 7. McCauley VB, Grunstein RR, Sullivan CEo Ethanol-induced depression of hypoxic drive and reversal by naloxone: a sex difference. Am Rev Respir Dis 1988; 137:1406-10. 8. White DP, Schneider BK, Santen RJ, et al. Influence of testosterone on ventilation and chemosensitivity in male subjects. J Appl Physiol1985; 59:1452-7. 9. Brogden RN, Clissold SP. Flutamide. A preliminary reviewof its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in advanced prostatic cancer. Drugs 1989; 38: 185-203. 10. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Bethesda: National Institute for Neurological Disease and Blindness, 1968 (NIH publication no. 204). 11. Berthon-Jones M, SullivanCEoVentilatoryand arousal responses to hypoxia in sleeping humans. Am Rev Respir Dis 1982; 125:632-9. 12. Rebuck AS, Campbell EJ. A clinical method for assessing the ventilatory response to hypoxia. Am Rev Respir Dis 1973; 109:345-50. 13. Read DJC. A clinical method for assessing the ventilatory response to carbon dioxide. Australas Ann Med 1967; 16:20-32. 14. Grunstein RR, Handelsman DJ, Lawrence SJ, Blackwell C, Caterson ID, Sullivan CEo Neuroendocrine dysfunction in sleep apnea: reversalby continuous positive airway pressure therapy. J Clin Endocrinol Metab 1989; 68:352-8. 15. Matsumoto AM, Sandblom RE, Schoene RB, et at. Testosterone replacement in hypogonadal males: effects on obstructive sleep apnea, respiratory drives and sleep. Clin Endocrinol (Oxf) 1985; 22:713-21.
ANDROGEN BLOCKADE DOES NOT AFFECT SLEEP-DISORDERED BREATHING IN OSA
16. Schneider BK, Pickett CK, Zwillich CW, et al. Influence of testosterone on breathing during sleep. J Appl Physiol 1986; 61:618-23. 17. Yu WH, Yu MC. Acceleration of the regeneration of the crushed hypoglossal nerve by testosterone. Exp Neurol 1983; 80:349-60. 18. Yu WHo Responsiveness of hypoglossal neurons to testosterone in pre-pubertal rats. Brain Res Bull 1984; 13:667-72. 19. Yu WH, McGinnis MY. Androgen receptor levels in cranial nerve nuclei and tongue muscles in rats. J Neurosci 1986; 6:1302-7. 20. Strumpf WE, Sar M. Steroid hormone target cells in the periventricular brain: relationship to peptide hormone producing cells. Fed Proc 1977; 36:1973-7. 21. Mowszowicz I. Antiandrogens: mechanisms and paradoxical effects. Ann Endocrinol (Paris) 1989; 50:189-99.
22. Knuth VA, Hano R, Nieschlag E. Effect of flutamide or cyproterone on pituitary and testicular hormones in normal men. J Clin Endocrinol Metab 1984; 59:963-9. 23. Harman E, Wynne JW, Block AJ, MalloyFisher L. Sleep disordered breathing and oxygen desaturation in obese patients. Chest 1981; 70: 256-60. 24. Santamaria JD, Prior JC, Fleetham JA. Reversible reproductive dysfunction in men with obstructive sleep apnea. Clin Endocrinol 1988; 28:461-70. 25. Zwillich CW, Natalino MR, Sutton FD, Weil JV. Effects of progesterone on chemosensitivity in normal men. J Lab Clin Med 1978; 92:262-9. 26. Morikawa T, Tanaka Y, Maruyama R, Nishibayashi Y, Honda Y. Comparison of two synthetic progesterones on ventilation in normal males: CMA vs. MPA. J Appl Physiol 1987; 63:1610-5.
1393
27. Lyons HA, Huang CT. Therapeutic use of progesterone in alveolar hypoventilation associated with obesity. Am J Med 1968; 44:881-8. 28. Smith KD, Rodriguez-Rigau J. Laboratory evaluation of testicular function. In: De Groot L, ed. Endocrinology. Philadelphia: W. B. Saunders, 1989; 2169. 29. Pickett CK, Regensteiner JG, Woodard WD, Hagerman DD, Weil JV, Grindlay Moore L. Progestin and estrogen reduce sleep-disordered breathing in postmenopausal women. J Appl Physiol 1989; 66:1656-61. 30. Franklin K, Lundgren R, Rabben T. Sleep apnea syndrome treated with oestradial and cyclic medroxyprogesterone. Lancet 1991; 338:252-3. 31. Dempsey JA, Skatrud JB. A sleep-induced apneic threshold and its consequences. Am Rev Respir Dis 1986; 133:1163-70.
