- Email: [email protected]
Sleep apnea and hypothyroidism: mechanisms and management
I
I
Sleep Apnea and Hypothyroidism: Mechanisms and Management RONALD R. GRUNSTEIN, M.D., COLIN E. SULLIVAN, M.D., Ph.D. Sydney, Australia
PURPOSE: T h e r e is a high incidence of sleep apnea in patients with untreated hypothyroidism. Thyroxine treatment is said to significantly reduce the apnea index and length and sleep apnea symptoms. We undertook a review of 10 consecutive hypothyroid patients with sleep apnea to investigate mechanisms and m a n a g e m e n t of these two disorders. P A T I E N T S A N D METHODS: Polysomnograms were obtained in l0 consecutive hypothyroid patients referred to our sleep disorders unit. All patients were studied while hypothyroid. Eight patients were restudied later when euthyroid. Lung function, blood gas values, and awake supraglottic resistance were also assessed in each patient. RESULTS: All 10 patients had sleep apnea and were treated with thyroxine. In one patient, hypothyroid myopathy involving the upper airway was demonstrated to be a potential mechanism of sleep apnea in hypothyroidism. Nocturnal angina and ventricular arrhythmias developed in two patients, despite the use of low thyroxine doses. Nasal continuous positive airways pressure (CPAP) was begun in eight patients. Initiation of CPAP prevented f u r t h e r angina or a r r y t h m i a in the patients with these cardiac complications. Six of the eight patients who were available for follow-up studies had persistent sleep apnea despite an euthyroid status (apnea index before thyroxine, 51 + 6; apnea index after thyroxine, 45 + 8), and CPAP therapy was continued in these patients. CONCLUSION: O u r experience suggests that the apnea index does not decrease significantly in all patients with hypothyroidism and sleep apnea when euthyroidism is achieved. T r e a t m e n t of hypothyroidism in the presence of sleep apnea is potentially hazardous and may lead to cardiovascular complications. M a n a g e m e n t by a combination of CPAP and low-dose thyroxine is helpful in this situation.
case reports have shown that hypothyroidism can produce either obstructive [1,2] or central Ssleepeveral apnea [3]. More recently, of 11 successive hypothyroid patients presenting to two institutions, nine were found to have obstructive sleep apnea [4]. Following thyroxine treatment, the patients in these studies had either complete reversal [1-3] or marked reduction in the number of apneas [4]. Narrowing of the upper airway (by mucoprotein deposition in the tongue and oropharynx) or abnormalities of ventilatory control are two proposed mechanisms of obstructive apnea in these patients [1-4]. Since a sleep disorders service was started at our institution, 10 patients admitted for investigation of a sleep disorder have been found to be hypothyroid at the time of their initial sleep study. Contrary to previous studies, we found that thyroxine replacement alone was inadequate to significantly reduce the frequency of apnea in some patients. Nasal continuous positive airways pressure (CPAP), a therapy first devised by our group [5,6], was used in six patients in addition to thyroxine replacement. In one patient, the clinical course and subsequent investigations, including upper airway mechanics, suggest that hypothyroid myopathy of the upper airway musculature is another mechanism leading to obstructive sleep apnea in hypothyroidism.
PATIENTS AND METHODS All 10 patients presented between 1981 and 1985. They were all significantly hypothyroid (thyroid-stimulating hormone [TSH] level greater than 25 mIU/ liter) except Patient 3 (TSH 16 mIU/liter) at the time of their initial polysomnogram. Five patients had not started thyroxine therapy at the time of the initial study, and the other five had received only several days' treatment. All were euthyroid (TSH less than 5 mIU/liter) at the time of their follow-up study. Body mass index (BMI) was calculated by dividing the patient's mass (kg) by the square of the height (meters). A value greater than 30 was considered obese.
Sleep Studies
From the Sleep Unit, Department of Thoracic Medicine, Royal Prince Alfred Hospital, and Department of Medicine, University of Sydney, Sydney, Australia. This work was supported in part by grants from the NH and Medical Research Counsel. Requests for reprints should be addressed to Dr. Ronald R. Grunstein, Department of Medicine, University of Sydney, N.S.W. 2006, Australia. Manuscript submitted May 5, 1988, and accepted in revised form September 26, ]988.
Polysomnograms were performed at the Sleep Unit, Royal Prince Alfred Hospital, between 11 P.M. and 6 A.M. Sleep state was recorded with two channels of electroencephalogram (C1/A3, C2/A4), two channels of electrooculogram, and one channel of submental electromyogram. Breathing variables included chest wall and abdominal motion (Respitrace, Ambulatory Monitoring Inc., Ardsley, New York), nasal airflow with a pressure transducer, and arterial oxyhemoglobin saturation (Hewlett-Packard 47201A, Mountain View, California). An electrocardiogram was recorded continuously. During institution of nasal CPAP therapy, nasal airway pressure was monitored continuously December 1988
The American Journal of Medicine
Volume 85
775
SLEEP APNEA IN HYPOTHYROIDISM / GRUNSTEIN AND SULLIVAN TABLE I Patient Data and Sleep-Breathing Variables Minimum Apnea Duration Patient Number
1
Sex
Age
M
51
Used Home CPAP
2
M
62
+
3
M
51
+
4
F
54
-
5
F
64
+~
6
F
63
+
7
M
58
+
8
M
47
+
9 10
F F
68 52
+t +t
BMI
Study Time (months)
AI
Maximum
Mean (seconds)
SEM
Lowest
Mean (percent)
SEM
22.5 22.0 30.1 26.8 43.3 43.3
0 5 0 13 0 13
33 0 57 73 52 58
130
72.00
+5.00
68
82.0
+1.50
19.8
0
46
17.5 31.3 34.8 41.2 28.6 35.3 32.7 34.2 34.2 38.0 46.6
2 0 10 0 5 0 5 0 5 0 0
0 26 39 55 51 46 14 70 35 45 66
.
-
Oxyhemoglobin Saturation
-78
41 83 100 58 -84
88 41 32 54 21 30 25
29 36
--
44.85 22.35 43.05 30.00 22.30
--
+2.33 4.1.22 -1-2.54 =1=3.55 +1.91
94
67 82 38 66 88
--
--
80.05 89.45 69.50 83.55 92.20
4.0.94 +0.46 4-1.61 4-0.99 +0.45
--
--
96
--
--
47.80 47.50 27.20 19.40 18.90 15.30 18.90 15.65 22.10 18.60
4-5.34 4-6.26 4-2.15 4-0.94 4-1.78 4-0.94 4-2.06 4-1.68 4-1.11 4-1.30
30 30 62 81 49 81 74
30 53.20 71.20 86.95 63.10 84.95 83.40 90.10 78.50 86.70
-4-2.36 4-1.45 4-0.45 4-1.90 4-0.82 4-0.68 4-1.60 4-0.58 +0.93
82
71 79
AI = apnea index. Used home CPAP with oxygen added. TNo follow-up sleep study.
with a pressure transducer (Statham PM 131 ETc, Dobbi Instruments, Canterbury, Australia). All variables were recorded continuously on a 16-channel electroencephalograph (Grass Instruments, Quincy, Massachusetts). All patients were studied while hypothyroid prior to any nasal CPAP therapy. Eight patients were restudied later when euthyroid. Six of these patients had regular home CPAP treatment in the intervening period. CPAP therapy was started as previously described [6]. Calculated respiratory variables were apnea index (number of apneas/hours of sleep), apnea duration, and minimal arterial oxygen saturation during apneas. Sleep recordings were scored in 20-second epochs and staged according to standard criteria [5] (Table I). No patient was receiving any sedating medications or alcohol prior to study. Other Studies
Lung function was measured by spirometry (Vitalograph, Ltd., Buckingham, England) and arterial blood was sampled for blood gas estimation. Awake supraglottic resistance was measured by the technique of posterior rhinometry [7], as adapted for supraglottic TABLE II Supraglottic Resistance Values* Patient Number
Status
Erect
Supine
I
H E H E H E H E
6.0 5.8 1.0 0.9 4.0 2.5 2.0 2.0
12.5 6.0 3.0 3.0 8.0 6.5 2.0 2.0
2 4 I0
= hypothyroid; E = euthyroid. ° Values expressed in cm HzO/liter/second. 776
December 1988
The American Journal of Medicine
resistance by Anch et al [10], in Patients 1, 2, 4, and 10 when hypothyroid and again in the euthyroid state (Table II). Patients breathed through a moulded nosemask and airflow was measured with a Fleisch pneumotachograph (Vacumed Inc., Ventura, California) with simultaneous pressure differentiation from nasal to supraglottic space (via a catheter placed at the tip of the epiglottis) using a Validyne (Validyne Engineering Corp., Northbridge, California) pressure transducer. Multiple pressure flow plots were taken using a storage oscilloscope. Measurements were made with the patient first seated upright and then lying supine. The pressure-flow relationship was measured over a large range of flows from tidal breathing to maximal voluntary ventilation. The reported values were those obtained during tidal breathing. Most subjects demonstrated complete closure of the upper airway (and therefore infinite "resistance") with maximal breathing efforts. All studies were done initially and at the time of the follow-up sleep study. CASE REPORTS Patient 1
Progressive lethargy and muscle weakness developed in a 50-year-old man, an active sportsman in excellent health. Being a chorister, he noticed deepening of his voice and occasional minimal dysarthria. His weakness progressed to having difficulty combing his hair and standing up from a seated position. At the time his muscle weakness developed, his very mild chronic snoring increased markedly in intensity and was associated with apneas, noted by his wife to be up to two minutes in length. Cold intolerance, dry skin, and constipation developed prior to his admission to Royal Prince Alfred Hospital for exclusion of sleep apnea. There was no weight gain. Examination revealed slow speech but normal mentation. Apart from skin dryness, there were no myxedematous features on examination. His temperature was 36.4°C. Bradycardia of 46 beats/minute was present. Neurologic examination demonstrated mild bilateral palatal weakness
Volume 85
SLEEPAPNEAIN HYPOTHYROIDISM/ GRUNSTEINAND SULLIVAN
as well as mil d proximal myopathy. Pseudomyotonia was present. There was no goiter. Investigations confirmed primary hypothyroidism. Sleep studies demonstrated obstructive sleep apnea in the supine position only (Table I). Most notable were the length of his apneas with relatively minimal desaturation. Findings on the lateral airway xerogram were normal. Supraglottic resistance was measured (Table II). Thyroxine replacement therapy, 50 ug, was started, and the dose was gradually increased, achieving a documented euthyroid status in six months. At that time, repeat sleep studies showed his sleep apnea had completely resolved. Supraglottic resistance was again measured and found to be reduced compared with pretreatment levels (Table II).
and hypertensive (160/100 mm Hg). Hypothyroidism was confirmed, and therapy with thyroxine, 50 pg daily, was started. Five days later, this was increased to 75 ~g of thyroxine, and severe nocturnal angina developed that awakened her from sleep. The pain persisted, and a rise in the creatinine kinase level to 1,176 U (MB fraction of 7 percent) was noted as were electroencephalographic changes of anterior myocardial infarction. Thyroxine was halted, but she continue'd to experience nocturnal angina, despite nitrate therapy. A sleep study was performed. CPAP at 10 cm H20 was started, and thyroxine was reintroduced without any recurrence of chest pain. RESULTS Patients
Patient 9 This 72-year-old woman was transferred to Royal Prince Alfred Hospital after initially being admitted to a district hospital because her relatives were unable to manage her at home. She was continually somnolent and loud snoring with apneas had been noticed by the patient's son who looked after her. This somnolence and snoring had developed six months previously and progressively worsened. Also, the patient had been unable to perform household chores, especially hanging up washing, and in the weeks prior to hospital admission had to be assisted in standing from a sitting position. TWO years previously, she had been diagnosed as being hypothyroid and thyroxine therapy was begun. She had ceased taking medications one year later. There was a long-standing history of hypertension but ischemic chest pain was denied. Other symptoms were progressive voice hoarseness, dysphagia, dry skin, ankle swelling, and mild cold intolerance. Examination revealed a grossly myxedematous obese woman with delayed ankle jerks and proximal myopathy clinically. She was bradycardic but not hypothermic. There was some cyanosis and she was extremely somnolent. Thyroid function tests confirmed marked hypothyroidism, and because of her clinical state, therapy with thyroxine 12.5 ug was initiated. A sleep study confirmed obstructive sleep apnea. Ventricular arrhythmias were present and these were closely associated with periods of sleep, being minimal during awake periods as confirmed by continuous electrocardiographic telemetry. In addition, angina at rest developed that was resistant to increasing nitrate doses, after thyroxine was begun. Therapy with CPAP was started, and apnea was prevented at a pressure of 8 cm H20. Following initiation of CPAP therapy, her angina and nocturnal ventricular arrhythmias were abolished. She was discharged with increased thyroxine and CPAP therapy arranged for her at home. A follow-up study was not performed at the patient's request. Patient 6 This 61-year-oid woman was admitted after a oneyear history of increasing weight, hoarse voide, dry skin, constipation, and cold intolerance. Three weeks before admission, proximal weakness and symptoms of carpal tunnel syndrome developed. There was a maternal history of hypothyroidism, and her father had had ischemic heart disease. She had no complaints of chest pain. She smoked 20 cigarettes/day. A history of recent loud snoring and daytime sleepiness was elicited. On examination, she was frankly myxedematous
All patients were heavy snorers and hypersomnolent. Two patients (Patients 2 and 3), who were studied because of sleep apnea symptoms, had abnormal thyroid function values that were noted only after outpatient follow-up following the commencement of CPAP therapy. They had no delayed ankle jerks or other classic signs of hypothyroidism. Their history of somnolence, snoring, and slow mentation was thought to be due to sleep apnea. Another two patients were referred because a thorough history and examination suggested hypothyroidism and this was confirmed with specific blood tests. Pre-sleep study classic myxedematous symptoms were present in the other six. 0nly two patients were non-obese at presentation (BMI less than 30). 0nly Patient 5 had abnormal lung function and blood gas values with a forced expiratory volume in one second of 49 percent of predicted, partial pressure of oxygen of 47 mm Hg, and partial pressure of carbon dioxide of 69 mm Hg. Sleep Studies All patients had obstructive apnea. Eight patients had studies done when they were hypothyroid and again when they were euthyroid. In two patients (both non-obese), complete cures were achieved by thyroxine replacement. Neither received treatment with nasal CPAP, Patient 1 because he presented prior to the advent of this therapy and Patient 4 because she had minimal oxyhemoglobin desaturation. In all other patients (including the two who did not have follow-up studies), nasal CPAP was started on the night following their diagnostic study, at the pressures determined individually for each patient. In each case, nasal CPAP prevented apnea, allowed quiet breathing, and maintained oxyhemoglobin saturation at greater than 92 percent with one exception (Patient 5 who required a combination of CPAP and 2 liters/minute oxygen to achieve a saturation greater than 90 percent). Changes with Treatment The follow-up studies were done as soon as practicable after T S H levels returned to normal. Of the eight patients who had follow-up studies, five became euthyroid five months or less after the initial sleep study, one became euthyroid after eight months, and in the other two patients, in whom the diagnosis of hypothyroidism was missed clinically, a euthyroid state was achieved five to six months after hormone replacement was started. Therefore, in all but one of the patients, follow-up studies were performed less than six months after hormone replacement was started. Thyroxine achieved complete cures in the two non-
December 1988 The American Journal of Medicine Volume 85
777
SLEEPAPNEAIN HYPOTHYROIDISM/ GRUNSTEIN AND SULLIVAN
obese patients, with no apneas noted on follow-up study. The six other patients with follow-up data still had persisting apneas of variable severity (Table I). The apnea index decreased in three patients and increased in three patients (pretreatment mean 51 4- 6, follow-up mean 45 4- 8). Mean apnea duration fell in all but one patient (34 4- 6 tQ 25 4- 5 seconds). Mean minimum oxygen saturation rose in all patients (66 4- 8 percent to 81 4- 6 percent). Patient 5, with both an increase in the apnea index and an unchanged mean duration of apnea, had coexisting chronic lung disease. At follow-up, there was no change in lung function or arterial blood gas values except in Patient 5 in whom there was an improvement in arterial blood gas tensions following treatment without change in the lung function. The partial pressure of oxygen increased to 59 mm Hg and the partial pressure of carbon dioxide decreased to 48 mm Hg. COMMENTS
This series of 10 patients confirms the association between hypothyroidism and obstructive sleep apnea. The clinical features of hypothyroidism should be sought in all patients with sleep apnea. Although eight of our patients had obvious clinical myxedema, two presented with sleep apnea with no clinical evidence of myxedema and were found to be hypothyroid only because of our practice of routine thyroid function testing in all patients with sleep apnea. Our results demonstrate two broad responses of sleep apnea to thyroxine treatment. One group (Patients 1 and 4) had complete resolution of apnea, whereas the other group continued to have persisting evidence of sleep apnea both clinically and polysomnographically. Mean apnea duration and mean minimum oxyhemoglobin desaturation improved, but not to the degree where these patients could be advised to cease CPAP therapy. In three patients, the apnea index increased. The improvement in blood gas values in Patient 5 is likely to be due to increased ventilatory drive secondary to CPAP treatment [19] or thyroxine
[4].
The failure of sleep apnea to resolve after thyroxine treatment may indicate that sleep apnea and hypothyroidism coexist by chance in some patients. Both are common disorders. In our series, all our patients with persisting apnea were obese at the time of the initial study. Obesity is a recognized precipitant of sleep apnea [18]. However, in the majority of our patients (except Patients 2 and 3), the clinical history of loud snoring and somnolence was temporally related to the development of symptoms of hypothyroidism and unlikely to be coincidental. The clinical history was difficult to assess in Patients 2 and 3 since hypothyroidism was not suspected, and in these two patients the diagnoses may well have been coincidental. An alternative explanation may be that hypothyroidism induces long-term changes in upper airway mechanics or breathing control that do not resolve immediately after a euthyroid state is achieved. We have now been able to follow up some of our patients for almost four years after diagnosis and all continue to require nasal CPAP. Without treatment, loud snoring, apneas, and somnolence recur. Therefore, the most likely hypothesis is that the two disorders are coincident in a subgroup of patients. Our results following treatment of hypothyroidism 778
differ from previous reports. Orr et al [1] described three obese patients with myxedema and sleep apnea. Sleep apnea was cured in all three patients when they became euthyroid. Other case reports [2,3] describe similar cures. In a recent series of nine hypothyroid patients [4], thyroxine therapy caused significant reduction in the apnea index for obese and non-obese patients. The index fell from 99.5 to less than 20 in the six obese patients in that series, and in all patients there was a marked decrease in the duration of apnea. The three non-obese patients reduced their apnea indices to less than 5 after treatment. In our series, cures were achieved in only two patients. It is clear that patients with sleep apnea and hypothyroidism are a heterogeneous group and that certainly in some patients thyroxine alone will not be adequate treatment. Nasal CPAP is an ideal therapy in patients with sleep apnea triggered by hypothyroidism. First, because sleep apnea will resolve in some patients when they become euthyroid, treatment is only administered for the few months required for normal thyroid status to return; clearly, invasive therapies such as tracheostomy would be inappropriate. The second more compelling reason for its use is to prevent cardiovascular complications in the initial stages of replacement therapy. It is well recognized that rapid restoration of the euthyroid state in hypothyroid patients may entail significant cardiovascular morbidity and mortality [16]. This is particularly so in the elderly or those with pre-existing cardiovascular disease. Certainly, there is an increased risk in initiating treatment in those patients with coexisting sleep apnea. A striking example of this problem was seen in Patient 1. He had extremely long apneas lasting up to 140 seconds in rapid-eye-movement sleep, yet the oxyhemoglobin desaturation only fell to 64 percent. Undoubtedly, his low metabolic rate and oxygen cons u m p t i o n ( r e d u c e d to 50 p e r c e n t of n o r m a l ) contributed to his ability to maintain such saturation despite long apneas. After starting thyroxine treatment, there may be a more rapid increase in basal metabolic rate and oxygen consumption than in clearance of abnormal mucoprotein from the upper airway and normalization of depressed ventilatory responses. Long apneas may then be associated with a lower oxyhemoglobin saturation as the oxygen consumption rate increased, therefore posing a major risk of dangerous hypoxemia for a patient with compromised coronary blood supply. SIeep apnea may be an important factor in worsening coronary artery disease or sudden death in some patients early in their course of treatment for hypothyroidism. Two of our patients had cardiac complications after starting thyroxine therapy prior to a sleep study and to the use of nasal CPAP therapy. Patient 6 had a myocardial infarction with residual nocturnal angina after her thyroxine dosage was increased. Her nocturnal angina resolved after nasal CPAP was begun. Patient 9 had nocturnal ventricular arrythmias and unstable angina after thyroxine was commenced. Both complications resolved with CPAP therapy. We suspect that because of Patient 9's severe daytime somnolence, episodes of angina were occurring during daytime sleep that awakened the patient. By abolishing daytime somnolence, CPAP prevented these episodes of angina. Orr et al [!] described a myxedematous patient with obstructive sleep apnea and cardiac
December1988 The American Journal of Medicine Volume85
SLEEP APNEA IN HYPOTHYROIDISM / GRUNSTEIN AND SULLIVAN
arrhythmias who had a tracheostomy performed resolving both the obstructive sleep apnea and arrhythmias. In our experience, nasal CPAP is a preferable alternative to tracheostomy. The combination of hypothyroidism and ischemic heart disease is a difficult clinical management problem [16] even without the further complication of sleep apnea, let alone the performance of tracheostomy. It has been suggested that patients with hypothyroidism have advanced atherosclerosis and, as in our patient series, that some experience their first episode of angina after receiving thyroxine therapy [14]. Exercise electrocardiograms seem to have a low predictive value in this situation [17]. We found nasal CPAP valuable in managing cardiovascular complications in two patients. The role of sleep apnea in provoking cardiac complications when initiating thyroid replacement therapy warrants further study. In at least three of our patients (Patients 1, 4, and 6), proximal myopathy was suggested historically and confirmed clinically. Others complained of weakness and slowness of movement but did not have convincing signs of myopathy. Myopathy is a well-recognized feature of hypothyroidism and is preseht in approximately one third of patients [8,9]. Anch and co-workers [10] have found elevated supraglottic resistance in patients with obstructive sleep apnea compared with resistance in normal subjects. Both normal subjects and patients had larger values supine compared with erect, suggesting narrowing of the pharyngeal aperture by posterior movement of the tongue and soft palate in the supine position. Others have demonstrated an increased supraglottic resistance (especially the pharyngeal component) in sleep [11]. The defense mechanism against airway occlusion is increased tone in the upper airway muscles, particularly the genioglossus [12,13]. Occlusion may occur if the upper airway muscles are hypotonic and cannot overcome the narrowing of the pharyngeal aperture in the supine position. We postulate that such hypotonia secondary to hypothyroid myopathy occurred in at least one patient (Patient 1) and represents another mechanism for producing obstructive sleep apnea in the hypothyroid patient. Support for this hypothesis is suggested by the presence of apneas in the supine position only in this patient, as well as the simultaneous full recovery of myopathy and obstructive sleep apnea with thyroxine alone and supraglottic resistance studies. These studies showed a markedly increased resistance in the supine versus erect position. The absence of any myxedema in this thin patient makes upper airway myopathy the most likely cause of his sleep apnea. Notably the supraglottic resistance in the supine position was halved after the patient became euthyroid, while the erect value did not change. There was no weight change in this non-obese patient after he was euthyroid. This response to thyroxine was also noted in the supraglottic
resistance of another patient (Patient 4). Therefore, sleep apnea may be precipitated by hypothyroid myopathy. Our results demonstrate that some patients with hypothyroidism and coexisting sleep apnea require further treatment of their sleep apnea even after a euthyroid state is achieved. Nasal CPAP is an effective therapy in these patients and useful in those who have coexisting cardiac disease, complicating thyroid 'hormone replacement. In addition, hypothyroid myopa: thy with upper airway involvement may be an additional precipitant for the development of sleep apnea in a hypothyroid patient.
ACKNOWLEDGMENT We wish to acknowledge the help of the nursing staff of the Sleep Unit, the technical assistanceof Ms. Lucy Costasand Mr. Peter Donnelly, and the secretarial assistance of Ms. Janet Bevan and Ms. Farida Bolano.
REFERENCES 1. Orr WC, MalesJL, lmes NK: Myxedemaand obstructive sleep apnea.Am J Med 1981; 70: 1061-1066. 2. Skatrud J, Iber C, Ewart R, Thomas G, Rasmussen H, Schultze B: Disordered breathingduring sleep in hypothyroidism. Am Rev Respir Dis 1981; 124: 325-329. 3. Milman RP, BevilacquaJ, PetersonDD, PackAI: Central sleep apneain hypothyroidism. Am Rev Respir Dis 1983; 127: 504-507. 4. Rajagobal KR, Abbrecht PH, Derderias SS, et al: Obstructive sleep apnea in hypothyroidism. Ann Intern Med 1984; 101: 491-494. 5. Sullivan CE, Issa FG, Berthon-Jones M, Eves L: Reversal of obstructive sleep apnea by continuous positive airway pressure applied through the nares. Lancet 1981; I: 862-865. 6. Sullivan CE, Issa FG, Berthon-JonesM, McCauleyVB, Costas LJV: Home treatment of obstructive sleep apnoeawith continuous positive airway pressureapplied through a nose mask. Bull Eur Physiopathol Respir 1984; 20: 49-54. 7. Cockroft DW, MacCormack DW, TarIo SM, Hargreave FE, Pengelly LD: Nasal airway inspiratory resistance. Am Rev Respir Dis 1979; 119: 921-926. ~]. Ramsay I: Thyroid diseaseand muscle dysfunction. London: Heinemann, 1974: 127-171. 9. Kendall-Taylor P, Turnbull DM: Endocrine myopathies. Br Med J 1983; 287: 705-708. 10. Anch AM, Returners JE, Bunce H: Supraglottic airway resistance in normal subjects and patients with occlusive sleep apnea. J Appl Physio11982; 53:11581163. 11. HodgesDW, Martin RJ, Johnson B, Robertson D: Increasein pharyngeal resistance during sleep in normal man (abstr). Am Rev Respir Dis 1983; 127: 4(part 2): 235. 12. Remmers JE, DeGroot WJ, Sauerland EK, Anch AM: Pathogenesisof upper airway occlusion during sleep. J Appl Physiol 1978; 44: 931-938. 13. Brouillette R, Thach B: Control of genioglossus inspiratory activity. J Appl Physiol 1980; 49: 801-808. 14. Keating FR, Parkin TW, Selby JB, Dickinson LS: Treatment of heart disease associated with myxoeclema. Prog Cardiovasc Dis 1961; 3: 364-387. 15. Rechtschaffen A, Kales A, eds. A manual of standardized terminology, techniques and scoring system for sleep stagesin human subjects. Los Angeles: Brain Information Service, UCLA; 1968. 16. Becker C: Hypothyroidism and atherosclerotic heart disease: pathogenesis, medical management,and the role of coronary artery bypasssurgery. Endocr Rev 1985; 6: 432-440. 17. Cohen RD, Lloyd-Thomas HG: Exerciseelectrocardiogram in myxedema. Br Med J 1966; 2: 327-330. 18. Guilleminault C, van den Hoed J, Mitler M: Clinical overview. Guilleminault C, Dement WC, eds. Sleepapnea syndromes. New York: Alan R. Liss, Inc., 1978; 112. 19. Berthon-JonesM, Sullivan CE: Time course of change in ventilatory response to COz with long-term CPAP therapy for obstructive sleep apnea. Am Rev Respir Dis 1987; 135: 144-147.
December ]988
The American Journal of Medicine
Volume 85
779
I
Sleep Apnea and Hypothyroidism: Mechanisms and Management RONALD R. GRUNSTEIN, M.D., COLIN E. SULLIVAN, M.D., Ph.D. Sydney, Australia
PURPOSE: T h e r e is a high incidence of sleep apnea in patients with untreated hypothyroidism. Thyroxine treatment is said to significantly reduce the apnea index and length and sleep apnea symptoms. We undertook a review of 10 consecutive hypothyroid patients with sleep apnea to investigate mechanisms and m a n a g e m e n t of these two disorders. P A T I E N T S A N D METHODS: Polysomnograms were obtained in l0 consecutive hypothyroid patients referred to our sleep disorders unit. All patients were studied while hypothyroid. Eight patients were restudied later when euthyroid. Lung function, blood gas values, and awake supraglottic resistance were also assessed in each patient. RESULTS: All 10 patients had sleep apnea and were treated with thyroxine. In one patient, hypothyroid myopathy involving the upper airway was demonstrated to be a potential mechanism of sleep apnea in hypothyroidism. Nocturnal angina and ventricular arrhythmias developed in two patients, despite the use of low thyroxine doses. Nasal continuous positive airways pressure (CPAP) was begun in eight patients. Initiation of CPAP prevented f u r t h e r angina or a r r y t h m i a in the patients with these cardiac complications. Six of the eight patients who were available for follow-up studies had persistent sleep apnea despite an euthyroid status (apnea index before thyroxine, 51 + 6; apnea index after thyroxine, 45 + 8), and CPAP therapy was continued in these patients. CONCLUSION: O u r experience suggests that the apnea index does not decrease significantly in all patients with hypothyroidism and sleep apnea when euthyroidism is achieved. T r e a t m e n t of hypothyroidism in the presence of sleep apnea is potentially hazardous and may lead to cardiovascular complications. M a n a g e m e n t by a combination of CPAP and low-dose thyroxine is helpful in this situation.
case reports have shown that hypothyroidism can produce either obstructive [1,2] or central Ssleepeveral apnea [3]. More recently, of 11 successive hypothyroid patients presenting to two institutions, nine were found to have obstructive sleep apnea [4]. Following thyroxine treatment, the patients in these studies had either complete reversal [1-3] or marked reduction in the number of apneas [4]. Narrowing of the upper airway (by mucoprotein deposition in the tongue and oropharynx) or abnormalities of ventilatory control are two proposed mechanisms of obstructive apnea in these patients [1-4]. Since a sleep disorders service was started at our institution, 10 patients admitted for investigation of a sleep disorder have been found to be hypothyroid at the time of their initial sleep study. Contrary to previous studies, we found that thyroxine replacement alone was inadequate to significantly reduce the frequency of apnea in some patients. Nasal continuous positive airways pressure (CPAP), a therapy first devised by our group [5,6], was used in six patients in addition to thyroxine replacement. In one patient, the clinical course and subsequent investigations, including upper airway mechanics, suggest that hypothyroid myopathy of the upper airway musculature is another mechanism leading to obstructive sleep apnea in hypothyroidism.
PATIENTS AND METHODS All 10 patients presented between 1981 and 1985. They were all significantly hypothyroid (thyroid-stimulating hormone [TSH] level greater than 25 mIU/ liter) except Patient 3 (TSH 16 mIU/liter) at the time of their initial polysomnogram. Five patients had not started thyroxine therapy at the time of the initial study, and the other five had received only several days' treatment. All were euthyroid (TSH less than 5 mIU/liter) at the time of their follow-up study. Body mass index (BMI) was calculated by dividing the patient's mass (kg) by the square of the height (meters). A value greater than 30 was considered obese.
Sleep Studies
From the Sleep Unit, Department of Thoracic Medicine, Royal Prince Alfred Hospital, and Department of Medicine, University of Sydney, Sydney, Australia. This work was supported in part by grants from the NH and Medical Research Counsel. Requests for reprints should be addressed to Dr. Ronald R. Grunstein, Department of Medicine, University of Sydney, N.S.W. 2006, Australia. Manuscript submitted May 5, 1988, and accepted in revised form September 26, ]988.
Polysomnograms were performed at the Sleep Unit, Royal Prince Alfred Hospital, between 11 P.M. and 6 A.M. Sleep state was recorded with two channels of electroencephalogram (C1/A3, C2/A4), two channels of electrooculogram, and one channel of submental electromyogram. Breathing variables included chest wall and abdominal motion (Respitrace, Ambulatory Monitoring Inc., Ardsley, New York), nasal airflow with a pressure transducer, and arterial oxyhemoglobin saturation (Hewlett-Packard 47201A, Mountain View, California). An electrocardiogram was recorded continuously. During institution of nasal CPAP therapy, nasal airway pressure was monitored continuously December 1988
The American Journal of Medicine
Volume 85
775
SLEEP APNEA IN HYPOTHYROIDISM / GRUNSTEIN AND SULLIVAN TABLE I Patient Data and Sleep-Breathing Variables Minimum Apnea Duration Patient Number
1
Sex
Age
M
51
Used Home CPAP
2
M
62
+
3
M
51
+
4
F
54
-
5
F
64
+~
6
F
63
+
7
M
58
+
8
M
47
+
9 10
F F
68 52
+t +t
BMI
Study Time (months)
AI
Maximum
Mean (seconds)
SEM
Lowest
Mean (percent)
SEM
22.5 22.0 30.1 26.8 43.3 43.3
0 5 0 13 0 13
33 0 57 73 52 58
130
72.00
+5.00
68
82.0
+1.50
19.8
0
46
17.5 31.3 34.8 41.2 28.6 35.3 32.7 34.2 34.2 38.0 46.6
2 0 10 0 5 0 5 0 5 0 0
0 26 39 55 51 46 14 70 35 45 66
.
-
Oxyhemoglobin Saturation
-78
41 83 100 58 -84
88 41 32 54 21 30 25
29 36
--
44.85 22.35 43.05 30.00 22.30
--
+2.33 4.1.22 -1-2.54 =1=3.55 +1.91
94
67 82 38 66 88
--
--
80.05 89.45 69.50 83.55 92.20
4.0.94 +0.46 4-1.61 4-0.99 +0.45
--
--
96
--
--
47.80 47.50 27.20 19.40 18.90 15.30 18.90 15.65 22.10 18.60
4-5.34 4-6.26 4-2.15 4-0.94 4-1.78 4-0.94 4-2.06 4-1.68 4-1.11 4-1.30
30 30 62 81 49 81 74
30 53.20 71.20 86.95 63.10 84.95 83.40 90.10 78.50 86.70
-4-2.36 4-1.45 4-0.45 4-1.90 4-0.82 4-0.68 4-1.60 4-0.58 +0.93
82
71 79
AI = apnea index. Used home CPAP with oxygen added. TNo follow-up sleep study.
with a pressure transducer (Statham PM 131 ETc, Dobbi Instruments, Canterbury, Australia). All variables were recorded continuously on a 16-channel electroencephalograph (Grass Instruments, Quincy, Massachusetts). All patients were studied while hypothyroid prior to any nasal CPAP therapy. Eight patients were restudied later when euthyroid. Six of these patients had regular home CPAP treatment in the intervening period. CPAP therapy was started as previously described [6]. Calculated respiratory variables were apnea index (number of apneas/hours of sleep), apnea duration, and minimal arterial oxygen saturation during apneas. Sleep recordings were scored in 20-second epochs and staged according to standard criteria [5] (Table I). No patient was receiving any sedating medications or alcohol prior to study. Other Studies
Lung function was measured by spirometry (Vitalograph, Ltd., Buckingham, England) and arterial blood was sampled for blood gas estimation. Awake supraglottic resistance was measured by the technique of posterior rhinometry [7], as adapted for supraglottic TABLE II Supraglottic Resistance Values* Patient Number
Status
Erect
Supine
I
H E H E H E H E
6.0 5.8 1.0 0.9 4.0 2.5 2.0 2.0
12.5 6.0 3.0 3.0 8.0 6.5 2.0 2.0
2 4 I0
= hypothyroid; E = euthyroid. ° Values expressed in cm HzO/liter/second. 776
December 1988
The American Journal of Medicine
resistance by Anch et al [10], in Patients 1, 2, 4, and 10 when hypothyroid and again in the euthyroid state (Table II). Patients breathed through a moulded nosemask and airflow was measured with a Fleisch pneumotachograph (Vacumed Inc., Ventura, California) with simultaneous pressure differentiation from nasal to supraglottic space (via a catheter placed at the tip of the epiglottis) using a Validyne (Validyne Engineering Corp., Northbridge, California) pressure transducer. Multiple pressure flow plots were taken using a storage oscilloscope. Measurements were made with the patient first seated upright and then lying supine. The pressure-flow relationship was measured over a large range of flows from tidal breathing to maximal voluntary ventilation. The reported values were those obtained during tidal breathing. Most subjects demonstrated complete closure of the upper airway (and therefore infinite "resistance") with maximal breathing efforts. All studies were done initially and at the time of the follow-up sleep study. CASE REPORTS Patient 1
Progressive lethargy and muscle weakness developed in a 50-year-old man, an active sportsman in excellent health. Being a chorister, he noticed deepening of his voice and occasional minimal dysarthria. His weakness progressed to having difficulty combing his hair and standing up from a seated position. At the time his muscle weakness developed, his very mild chronic snoring increased markedly in intensity and was associated with apneas, noted by his wife to be up to two minutes in length. Cold intolerance, dry skin, and constipation developed prior to his admission to Royal Prince Alfred Hospital for exclusion of sleep apnea. There was no weight gain. Examination revealed slow speech but normal mentation. Apart from skin dryness, there were no myxedematous features on examination. His temperature was 36.4°C. Bradycardia of 46 beats/minute was present. Neurologic examination demonstrated mild bilateral palatal weakness
Volume 85
SLEEPAPNEAIN HYPOTHYROIDISM/ GRUNSTEINAND SULLIVAN
as well as mil d proximal myopathy. Pseudomyotonia was present. There was no goiter. Investigations confirmed primary hypothyroidism. Sleep studies demonstrated obstructive sleep apnea in the supine position only (Table I). Most notable were the length of his apneas with relatively minimal desaturation. Findings on the lateral airway xerogram were normal. Supraglottic resistance was measured (Table II). Thyroxine replacement therapy, 50 ug, was started, and the dose was gradually increased, achieving a documented euthyroid status in six months. At that time, repeat sleep studies showed his sleep apnea had completely resolved. Supraglottic resistance was again measured and found to be reduced compared with pretreatment levels (Table II).
and hypertensive (160/100 mm Hg). Hypothyroidism was confirmed, and therapy with thyroxine, 50 pg daily, was started. Five days later, this was increased to 75 ~g of thyroxine, and severe nocturnal angina developed that awakened her from sleep. The pain persisted, and a rise in the creatinine kinase level to 1,176 U (MB fraction of 7 percent) was noted as were electroencephalographic changes of anterior myocardial infarction. Thyroxine was halted, but she continue'd to experience nocturnal angina, despite nitrate therapy. A sleep study was performed. CPAP at 10 cm H20 was started, and thyroxine was reintroduced without any recurrence of chest pain. RESULTS Patients
Patient 9 This 72-year-old woman was transferred to Royal Prince Alfred Hospital after initially being admitted to a district hospital because her relatives were unable to manage her at home. She was continually somnolent and loud snoring with apneas had been noticed by the patient's son who looked after her. This somnolence and snoring had developed six months previously and progressively worsened. Also, the patient had been unable to perform household chores, especially hanging up washing, and in the weeks prior to hospital admission had to be assisted in standing from a sitting position. TWO years previously, she had been diagnosed as being hypothyroid and thyroxine therapy was begun. She had ceased taking medications one year later. There was a long-standing history of hypertension but ischemic chest pain was denied. Other symptoms were progressive voice hoarseness, dysphagia, dry skin, ankle swelling, and mild cold intolerance. Examination revealed a grossly myxedematous obese woman with delayed ankle jerks and proximal myopathy clinically. She was bradycardic but not hypothermic. There was some cyanosis and she was extremely somnolent. Thyroid function tests confirmed marked hypothyroidism, and because of her clinical state, therapy with thyroxine 12.5 ug was initiated. A sleep study confirmed obstructive sleep apnea. Ventricular arrhythmias were present and these were closely associated with periods of sleep, being minimal during awake periods as confirmed by continuous electrocardiographic telemetry. In addition, angina at rest developed that was resistant to increasing nitrate doses, after thyroxine was begun. Therapy with CPAP was started, and apnea was prevented at a pressure of 8 cm H20. Following initiation of CPAP therapy, her angina and nocturnal ventricular arrhythmias were abolished. She was discharged with increased thyroxine and CPAP therapy arranged for her at home. A follow-up study was not performed at the patient's request. Patient 6 This 61-year-oid woman was admitted after a oneyear history of increasing weight, hoarse voide, dry skin, constipation, and cold intolerance. Three weeks before admission, proximal weakness and symptoms of carpal tunnel syndrome developed. There was a maternal history of hypothyroidism, and her father had had ischemic heart disease. She had no complaints of chest pain. She smoked 20 cigarettes/day. A history of recent loud snoring and daytime sleepiness was elicited. On examination, she was frankly myxedematous
All patients were heavy snorers and hypersomnolent. Two patients (Patients 2 and 3), who were studied because of sleep apnea symptoms, had abnormal thyroid function values that were noted only after outpatient follow-up following the commencement of CPAP therapy. They had no delayed ankle jerks or other classic signs of hypothyroidism. Their history of somnolence, snoring, and slow mentation was thought to be due to sleep apnea. Another two patients were referred because a thorough history and examination suggested hypothyroidism and this was confirmed with specific blood tests. Pre-sleep study classic myxedematous symptoms were present in the other six. 0nly two patients were non-obese at presentation (BMI less than 30). 0nly Patient 5 had abnormal lung function and blood gas values with a forced expiratory volume in one second of 49 percent of predicted, partial pressure of oxygen of 47 mm Hg, and partial pressure of carbon dioxide of 69 mm Hg. Sleep Studies All patients had obstructive apnea. Eight patients had studies done when they were hypothyroid and again when they were euthyroid. In two patients (both non-obese), complete cures were achieved by thyroxine replacement. Neither received treatment with nasal CPAP, Patient 1 because he presented prior to the advent of this therapy and Patient 4 because she had minimal oxyhemoglobin desaturation. In all other patients (including the two who did not have follow-up studies), nasal CPAP was started on the night following their diagnostic study, at the pressures determined individually for each patient. In each case, nasal CPAP prevented apnea, allowed quiet breathing, and maintained oxyhemoglobin saturation at greater than 92 percent with one exception (Patient 5 who required a combination of CPAP and 2 liters/minute oxygen to achieve a saturation greater than 90 percent). Changes with Treatment The follow-up studies were done as soon as practicable after T S H levels returned to normal. Of the eight patients who had follow-up studies, five became euthyroid five months or less after the initial sleep study, one became euthyroid after eight months, and in the other two patients, in whom the diagnosis of hypothyroidism was missed clinically, a euthyroid state was achieved five to six months after hormone replacement was started. Therefore, in all but one of the patients, follow-up studies were performed less than six months after hormone replacement was started. Thyroxine achieved complete cures in the two non-
December 1988 The American Journal of Medicine Volume 85
777
SLEEPAPNEAIN HYPOTHYROIDISM/ GRUNSTEIN AND SULLIVAN
obese patients, with no apneas noted on follow-up study. The six other patients with follow-up data still had persisting apneas of variable severity (Table I). The apnea index decreased in three patients and increased in three patients (pretreatment mean 51 4- 6, follow-up mean 45 4- 8). Mean apnea duration fell in all but one patient (34 4- 6 tQ 25 4- 5 seconds). Mean minimum oxygen saturation rose in all patients (66 4- 8 percent to 81 4- 6 percent). Patient 5, with both an increase in the apnea index and an unchanged mean duration of apnea, had coexisting chronic lung disease. At follow-up, there was no change in lung function or arterial blood gas values except in Patient 5 in whom there was an improvement in arterial blood gas tensions following treatment without change in the lung function. The partial pressure of oxygen increased to 59 mm Hg and the partial pressure of carbon dioxide decreased to 48 mm Hg. COMMENTS
This series of 10 patients confirms the association between hypothyroidism and obstructive sleep apnea. The clinical features of hypothyroidism should be sought in all patients with sleep apnea. Although eight of our patients had obvious clinical myxedema, two presented with sleep apnea with no clinical evidence of myxedema and were found to be hypothyroid only because of our practice of routine thyroid function testing in all patients with sleep apnea. Our results demonstrate two broad responses of sleep apnea to thyroxine treatment. One group (Patients 1 and 4) had complete resolution of apnea, whereas the other group continued to have persisting evidence of sleep apnea both clinically and polysomnographically. Mean apnea duration and mean minimum oxyhemoglobin desaturation improved, but not to the degree where these patients could be advised to cease CPAP therapy. In three patients, the apnea index increased. The improvement in blood gas values in Patient 5 is likely to be due to increased ventilatory drive secondary to CPAP treatment [19] or thyroxine
[4].
The failure of sleep apnea to resolve after thyroxine treatment may indicate that sleep apnea and hypothyroidism coexist by chance in some patients. Both are common disorders. In our series, all our patients with persisting apnea were obese at the time of the initial study. Obesity is a recognized precipitant of sleep apnea [18]. However, in the majority of our patients (except Patients 2 and 3), the clinical history of loud snoring and somnolence was temporally related to the development of symptoms of hypothyroidism and unlikely to be coincidental. The clinical history was difficult to assess in Patients 2 and 3 since hypothyroidism was not suspected, and in these two patients the diagnoses may well have been coincidental. An alternative explanation may be that hypothyroidism induces long-term changes in upper airway mechanics or breathing control that do not resolve immediately after a euthyroid state is achieved. We have now been able to follow up some of our patients for almost four years after diagnosis and all continue to require nasal CPAP. Without treatment, loud snoring, apneas, and somnolence recur. Therefore, the most likely hypothesis is that the two disorders are coincident in a subgroup of patients. Our results following treatment of hypothyroidism 778
differ from previous reports. Orr et al [1] described three obese patients with myxedema and sleep apnea. Sleep apnea was cured in all three patients when they became euthyroid. Other case reports [2,3] describe similar cures. In a recent series of nine hypothyroid patients [4], thyroxine therapy caused significant reduction in the apnea index for obese and non-obese patients. The index fell from 99.5 to less than 20 in the six obese patients in that series, and in all patients there was a marked decrease in the duration of apnea. The three non-obese patients reduced their apnea indices to less than 5 after treatment. In our series, cures were achieved in only two patients. It is clear that patients with sleep apnea and hypothyroidism are a heterogeneous group and that certainly in some patients thyroxine alone will not be adequate treatment. Nasal CPAP is an ideal therapy in patients with sleep apnea triggered by hypothyroidism. First, because sleep apnea will resolve in some patients when they become euthyroid, treatment is only administered for the few months required for normal thyroid status to return; clearly, invasive therapies such as tracheostomy would be inappropriate. The second more compelling reason for its use is to prevent cardiovascular complications in the initial stages of replacement therapy. It is well recognized that rapid restoration of the euthyroid state in hypothyroid patients may entail significant cardiovascular morbidity and mortality [16]. This is particularly so in the elderly or those with pre-existing cardiovascular disease. Certainly, there is an increased risk in initiating treatment in those patients with coexisting sleep apnea. A striking example of this problem was seen in Patient 1. He had extremely long apneas lasting up to 140 seconds in rapid-eye-movement sleep, yet the oxyhemoglobin desaturation only fell to 64 percent. Undoubtedly, his low metabolic rate and oxygen cons u m p t i o n ( r e d u c e d to 50 p e r c e n t of n o r m a l ) contributed to his ability to maintain such saturation despite long apneas. After starting thyroxine treatment, there may be a more rapid increase in basal metabolic rate and oxygen consumption than in clearance of abnormal mucoprotein from the upper airway and normalization of depressed ventilatory responses. Long apneas may then be associated with a lower oxyhemoglobin saturation as the oxygen consumption rate increased, therefore posing a major risk of dangerous hypoxemia for a patient with compromised coronary blood supply. SIeep apnea may be an important factor in worsening coronary artery disease or sudden death in some patients early in their course of treatment for hypothyroidism. Two of our patients had cardiac complications after starting thyroxine therapy prior to a sleep study and to the use of nasal CPAP therapy. Patient 6 had a myocardial infarction with residual nocturnal angina after her thyroxine dosage was increased. Her nocturnal angina resolved after nasal CPAP was begun. Patient 9 had nocturnal ventricular arrythmias and unstable angina after thyroxine was commenced. Both complications resolved with CPAP therapy. We suspect that because of Patient 9's severe daytime somnolence, episodes of angina were occurring during daytime sleep that awakened the patient. By abolishing daytime somnolence, CPAP prevented these episodes of angina. Orr et al [!] described a myxedematous patient with obstructive sleep apnea and cardiac
December1988 The American Journal of Medicine Volume85
SLEEP APNEA IN HYPOTHYROIDISM / GRUNSTEIN AND SULLIVAN
arrhythmias who had a tracheostomy performed resolving both the obstructive sleep apnea and arrhythmias. In our experience, nasal CPAP is a preferable alternative to tracheostomy. The combination of hypothyroidism and ischemic heart disease is a difficult clinical management problem [16] even without the further complication of sleep apnea, let alone the performance of tracheostomy. It has been suggested that patients with hypothyroidism have advanced atherosclerosis and, as in our patient series, that some experience their first episode of angina after receiving thyroxine therapy [14]. Exercise electrocardiograms seem to have a low predictive value in this situation [17]. We found nasal CPAP valuable in managing cardiovascular complications in two patients. The role of sleep apnea in provoking cardiac complications when initiating thyroid replacement therapy warrants further study. In at least three of our patients (Patients 1, 4, and 6), proximal myopathy was suggested historically and confirmed clinically. Others complained of weakness and slowness of movement but did not have convincing signs of myopathy. Myopathy is a well-recognized feature of hypothyroidism and is preseht in approximately one third of patients [8,9]. Anch and co-workers [10] have found elevated supraglottic resistance in patients with obstructive sleep apnea compared with resistance in normal subjects. Both normal subjects and patients had larger values supine compared with erect, suggesting narrowing of the pharyngeal aperture by posterior movement of the tongue and soft palate in the supine position. Others have demonstrated an increased supraglottic resistance (especially the pharyngeal component) in sleep [11]. The defense mechanism against airway occlusion is increased tone in the upper airway muscles, particularly the genioglossus [12,13]. Occlusion may occur if the upper airway muscles are hypotonic and cannot overcome the narrowing of the pharyngeal aperture in the supine position. We postulate that such hypotonia secondary to hypothyroid myopathy occurred in at least one patient (Patient 1) and represents another mechanism for producing obstructive sleep apnea in the hypothyroid patient. Support for this hypothesis is suggested by the presence of apneas in the supine position only in this patient, as well as the simultaneous full recovery of myopathy and obstructive sleep apnea with thyroxine alone and supraglottic resistance studies. These studies showed a markedly increased resistance in the supine versus erect position. The absence of any myxedema in this thin patient makes upper airway myopathy the most likely cause of his sleep apnea. Notably the supraglottic resistance in the supine position was halved after the patient became euthyroid, while the erect value did not change. There was no weight change in this non-obese patient after he was euthyroid. This response to thyroxine was also noted in the supraglottic
resistance of another patient (Patient 4). Therefore, sleep apnea may be precipitated by hypothyroid myopathy. Our results demonstrate that some patients with hypothyroidism and coexisting sleep apnea require further treatment of their sleep apnea even after a euthyroid state is achieved. Nasal CPAP is an effective therapy in these patients and useful in those who have coexisting cardiac disease, complicating thyroid 'hormone replacement. In addition, hypothyroid myopa: thy with upper airway involvement may be an additional precipitant for the development of sleep apnea in a hypothyroid patient.
ACKNOWLEDGMENT We wish to acknowledge the help of the nursing staff of the Sleep Unit, the technical assistanceof Ms. Lucy Costasand Mr. Peter Donnelly, and the secretarial assistance of Ms. Janet Bevan and Ms. Farida Bolano.
REFERENCES 1. Orr WC, MalesJL, lmes NK: Myxedemaand obstructive sleep apnea.Am J Med 1981; 70: 1061-1066. 2. Skatrud J, Iber C, Ewart R, Thomas G, Rasmussen H, Schultze B: Disordered breathingduring sleep in hypothyroidism. Am Rev Respir Dis 1981; 124: 325-329. 3. Milman RP, BevilacquaJ, PetersonDD, PackAI: Central sleep apneain hypothyroidism. Am Rev Respir Dis 1983; 127: 504-507. 4. Rajagobal KR, Abbrecht PH, Derderias SS, et al: Obstructive sleep apnea in hypothyroidism. Ann Intern Med 1984; 101: 491-494. 5. Sullivan CE, Issa FG, Berthon-Jones M, Eves L: Reversal of obstructive sleep apnea by continuous positive airway pressure applied through the nares. Lancet 1981; I: 862-865. 6. Sullivan CE, Issa FG, Berthon-JonesM, McCauleyVB, Costas LJV: Home treatment of obstructive sleep apnoeawith continuous positive airway pressureapplied through a nose mask. Bull Eur Physiopathol Respir 1984; 20: 49-54. 7. Cockroft DW, MacCormack DW, TarIo SM, Hargreave FE, Pengelly LD: Nasal airway inspiratory resistance. Am Rev Respir Dis 1979; 119: 921-926. ~]. Ramsay I: Thyroid diseaseand muscle dysfunction. London: Heinemann, 1974: 127-171. 9. Kendall-Taylor P, Turnbull DM: Endocrine myopathies. Br Med J 1983; 287: 705-708. 10. Anch AM, Returners JE, Bunce H: Supraglottic airway resistance in normal subjects and patients with occlusive sleep apnea. J Appl Physio11982; 53:11581163. 11. HodgesDW, Martin RJ, Johnson B, Robertson D: Increasein pharyngeal resistance during sleep in normal man (abstr). Am Rev Respir Dis 1983; 127: 4(part 2): 235. 12. Remmers JE, DeGroot WJ, Sauerland EK, Anch AM: Pathogenesisof upper airway occlusion during sleep. J Appl Physiol 1978; 44: 931-938. 13. Brouillette R, Thach B: Control of genioglossus inspiratory activity. J Appl Physiol 1980; 49: 801-808. 14. Keating FR, Parkin TW, Selby JB, Dickinson LS: Treatment of heart disease associated with myxoeclema. Prog Cardiovasc Dis 1961; 3: 364-387. 15. Rechtschaffen A, Kales A, eds. A manual of standardized terminology, techniques and scoring system for sleep stagesin human subjects. Los Angeles: Brain Information Service, UCLA; 1968. 16. Becker C: Hypothyroidism and atherosclerotic heart disease: pathogenesis, medical management,and the role of coronary artery bypasssurgery. Endocr Rev 1985; 6: 432-440. 17. Cohen RD, Lloyd-Thomas HG: Exerciseelectrocardiogram in myxedema. Br Med J 1966; 2: 327-330. 18. Guilleminault C, van den Hoed J, Mitler M: Clinical overview. Guilleminault C, Dement WC, eds. Sleepapnea syndromes. New York: Alan R. Liss, Inc., 1978; 112. 19. Berthon-JonesM, Sullivan CE: Time course of change in ventilatory response to COz with long-term CPAP therapy for obstructive sleep apnea. Am Rev Respir Dis 1987; 135: 144-147.
December ]988
The American Journal of Medicine
Volume 85
779