Abstract
Sleep-disordered breathing (SDB) is not only a problem of the upper airway but is a systemic condition with endocrine and metabolic interactions. The accumulating body of evidence shows that SDB induces changes in the serum levels or secretory patterns of several hormones. Conversely, various endocrine disorders and hormone therapies may induce, exacerbate or alleviate SDB.
Much of the understanding of the interactions between hormones and sleep-disordered breathing derive from intervention studies with nasal continuous positive airway pressure therapy. Better understanding of hormones and breathing may open new perspectives in developing strategies to prevent, alleviate or cure sleep-disordered breathing and its systemic consequences.
This study was supported by the Finnish Anti-Tuberculosis Association Foundation, Emil Aaltonen Foundation and Paulo Foundation.
Sleep-disordered breathing (SDB) is an extremely common condition that compromises the vital functions of respiration and circulation. There is a myriad of adaptive physiological responses that are activated when cellular gas exchange and acid-base balance are endangered. Therefore, SDB has widespread systemic effects, which are, unfortunately, rarely considered by medical professionals other than those specialised in diagnosing and treating this disorder. The many adaptive endocrine alterations associated with SDB are an example of how a seemingly local upper airway dysfunction induces systemic consequences, affecting every cell of the organism. Conversely, manifestation of sleep apnoea is critically linked with control of breathing. All endocrine changes that increase the tendency for periodic breathing will also increase the episodes of sleep apnoea. The present review focuses on SDB secondary to various abnormal endocrine states and on the physiological endocrinological responses to primary SDB. The specific aspects of treatment of SDB in endocrine disorders are also discussed and some treatment strategies, based on the literature, are suggested.
The concept of SDB has markedly evolved during the past decade. The episodes of sleep apnoea and hypopnoea result from periodic total or partial closure of the upper airway. These episodes are often accompanied by hypoxaemia and terminated with cortical electroencephalogram arousals. The severity of SDB is commonly expressed as the apnoea/hypopnoea index (AHI), which indicates the frequency of the apnoea/hypopnoea episodes per hour of sleep. Some authors also include the respiratory effort-related arousals and express the severity of SDB as the respiratory disturbance index (RDI). Most studies referred to in the present review define SDB in terms of AHI or RDI. Prolonged episodes of obstructive hypoventilation (upper airway flow limitation, partial obstruction) are acknowledged, but commonly not entered into the severity indices.
Sleep apnoea seems like an epidemic, which spreads rapidly with obesity, another major health problem in Western societies. In the USA, 24% of male and 9% of female government workers present with episodes of sleep apnoea or hypopnoea of five or more per hour 1. Daytime symptoms of sleep apnoea appear in 4% of males and 2% of females 1. Similar prevalence rates of symptomatic sleep apnoea in adults aged 20–100 yrs are reported in a community-based study: 3.9% in males and 1.2% in females 2.
A number of hormones interact with sleep 3 and breathing 4. SDB affects hormones via a number of mechanisms. Conversely, hormones and endocrine states induce, aggravate or alleviate SDB. Finally, nasal continuous positive airway pressure (CPAP) therapy influences hormone secretion.
SDB and sleep disturbances may interact with hormones in several ways. Episodes of apnoea or hypopnoea cause sleep fragmentation and disturb sleep cycles and stages. Arousals may induce stress response resulting in increased levels of stress hormones 5. Hypoxia may also have direct effects on central neurotransmitters 6, which result in alterations in the hypothalamo-pituitary axis and in secretion of the peripheral endocrine glands 7. Hypercapnia alone or combined with hypoxia may increase levels of renin, adrenocorticotrophic hormone, corticosteroids, aldosterone and vasopressin 8, 9. Finally, disorganisation of sleep, sleep loss and naps disturb sleep-controlled endocrine rhythms resulting in endocrine and metabolic abnormalities.
The direct and indirect effects of hormones and endocrine disorders on sleep and breathing are mediated via several pathways. Male sex and postmenopausal state, as risk factors 10, 11, link sex hormones to the pathophysiology of SDB. Sleep apnoea is common in acromegaly 12–17, hypothyroidism 18–21 or Cushing's syndrome 22, 23 (table 1⇓). The most recent studies suggest that SDB may not only complete the clinical picture but play a central role in the pathophysiology of obesity 32, leptin resistance 33–35 or the metabolic syndrome 32, 36, 37. The prevalence estimates of sleep apnoea among various endocrine states and disorders are shown in table 1⇓.
Prevalence of sleep-disordered breathing in some endocrine disorders and states
Unfortunately, there is a lack of well-documented epidemiological studies and thus most prevalence estimations are based on small study populations. Prevalence estimates are also limited because of different definitions of SDB. Many studies investigating the effects of SDB or nasal CPAP therapy on hormone levels have only assessed single morning levels of hormones, and therefore the effects on the 24‐h secretory profile are poorly known.
Metabolic syndrome and diabetes
Sleep fragmentation resulting in sleep deprivation is likely to have an impact on hormones that regulate glucose tolerance. Partial sleep restriction (4‐h sleep·night−1 for 6 days) resulted in increased cortisol levels and impaired glucose tolerance even in healthy nonobese young males 38. These metabolic and endocrine alterations were recuperated during recovery sleep. There is an accumulating body of evidence that SDB is linked with insulin resistance and metabolic syndrome independently of body mass index (BMI) and other known risk factors 32, 36, 37 (fig. 1⇓). Oxygen desaturation index (drops of oxygen saturation of ≥4%·h−1) is a better predictor of insulin resistance than BMI 39. In females with polycystic ovary syndrome, insulin resistance is a stronger risk factor for sleep apnoea than BMI or serum testosterone levels 30.
Index of insulin sensitivity calculated with a homeostasis model assessment (HOMA=G0xI0/22.5, where G0 and I0 represent fasting serum glucose and insulin, respectively) in different apnoea/hypopnoea index (AHI) categories (n=150 males). Subjects with increasing AHI are increasingly resistant to insulin. p-Value is significant for trend across AHI categories (p<0.05). Based on data from 37.
The prevalence of SDB in type‐1 diabetes remains to be confirmed. Some authors have reported a prevalence rate of sleep apnoea as high as 42% 25, whereas others have not observed a difference from the general population 28. Small sample sizes and different diagnostic criteria for sleep apnoea may explain some of the discrepancy. Diabetic children (n=25) have more episodes of apnoea during sleep and the duration of apnoeic events is longer than in healthy controls 40. Further, the degree of severity of sleep apnoea correlates with the glucose control and the duration of diabetes.
Among ∼13,000 Japanese hospital inpatients, the prevalence of sleep apnoea was 0.3% 26. In a subgroup of ∼600 male type‐2 diabetics, the prevalence of sleep apnoea was higher than in nondiabetics (1.9 versus 0.3%, respectively). In a Swedish 10-yr follow-up study, snoring was a risk factor for diabetes, independent of other risk factors 41. Among hypertensive diabetics, the prevalence of sleep apnoea, defined as AHI ≥20, was 36% compared with 14.5% in nondiabetics 27. Autonomic diabetic neuropathy may be associated with sleep apnoea. Among 23 diabetics with autonomic neuropathy (one had type‐1 diabetes), six had sleep apnoea, whereas none of the diabetics without autonomic neuropathy were affected 29.
Leptin
Besides its best known function as a satiety hormone, leptin is also a powerful respiratory stimulant 42. Plasma leptin levels are higher in sleep apnoeics than in controls matched for BMI 43. Furthermore, hypercapnic patients with obstructive sleep apnoea syndrome (OSAS) have higher leptin levels than eucapnic BMI-matched controls with sleep apnoea 44. Leptin secretion could provide an adaptive mechanism to enhance ventilation in patients with severe respiratory impairment. Conversely, high circulating leptin levels suggest leptin resistance at the level of the central nervous system. Elevated leptin levels are likely to contribute to comorbidity of OSAS because high leptin levels are associated with coronary heart disease 45, insulin resistance 46, impaired fibrinolysis 47, development of obesity 48, or type-2 diabetes 49, which are all highly prevalent in patients with OSAS.
Catecholamines
In blood and urine, high levels of catecholamines and their metabolites reflect increased sympathetic activity. Muscle sympathetic nerve activity is greater in obese than in normal-weight subjects 50, and greater in sleep apnoeics than in age- and BMI-matched controls 51, 52. Hypoxia and hypercapnia induce sympathetic nervous system overactivity 53. The sympathetic responses to hypoxia and hypercapnia are further potentiated during apnoea, when the inhibitory influence of the thoracic afferent nerves is eliminated 50, 52. Nocturnal noradrenaline levels correlate with OSAS severity and oxygen saturation 54, 55. Also sleep fragmentation leading to chronic partial sleep loss is likely to contribute to the increased sympathoadrenal activity and increased circulating catecholamine levels encountered in OSAS. This assumption is supported by observations in healthy male volunteers 56. One night of partial sleep deprivation resulted in increases in circulating noradrenaline and adrenaline levels 56. Most studies report a positive relationship between episodes of obstructive apnoea and noradrenaline levels, whereas a minority of studies have found a relationship between adrenaline and episodes of obstructive apnoea 57, 58.
Hypothyroidism
A link between SDB and hypothyroidism is suggested by the high prevalence of sleep apnoea among hypothyroid patients, particularly in rare myxoedematous patients (7.7–100%) 18–21 (table 1⇑). Therefore, symptoms of SDB should be routinely asked in all hypothyroid patients and sleep studies should be considered when symptoms present.
The increased prevalence of SDB appears to be related to obesity and male sex rather than hypothyroidism per se 21. However, decreased ventilatory responses 59, extravasation of albumin and mucopolysaccharides in the tissues of the upper airway 60, 61 and hypothyroid myopathy 19 have been suggested as possible contributing factors for SDB in hypothyroidism.
The decreased ventilatory responses increase with thyroxin replacement 59, 62, and episodes of apnoea may disappear 18–20, 63. After initiation of thyroxin replacement therapy, patients may snore more 20, suffer from nocturnal chest pain and ventricular arrhythmia 19. A temporary worsening of SDB after onset of thyroxin therapy could be due to an increase in basal metabolic rate, increased oxygen consumption and increased respiratory drive, which could promote periodic breathing and upper airway instability. Prolonged episodes of apnoea and lower oxyhaemoglobin saturation could be risky in patients with pre-existing coronary heart disease. To avoid the possible complications, hypothyroid patients with SDB should be, at least, initially treated with nasal CPAP. When the steady state has been achieved and the patient no longer has symptoms of hypothyroidism, the need for nasal CPAP therapy has to be re-evaluated.
In patients with OSAS, the prevalence of hypothyroidism is 1–3% 20, 64, 65, which does not essentially differ from that in the general population. Screening for hypothyroidism in patients with sleep apnoea does not seem necessary unless the patient is symptomatic or belongs to a risk group (i.e. females aged ≥60 yrs) 65.
Acromegaly
The association of snoring and daytime sleepiness and acromegaly was first reported more than a century ago 66. Macroglossia and pharyngeal swelling are the most probable reasons for the high incidence of SDB in acromegaly 67–70 (table 1⇑). Accordingly, sleep apnoea alleviates when tissue hypertrophy decreases with somatostatin analogue treatment 71–73. Growth hormone and insulin-like growth factor (IGF)‐I may also have a direct role in the pathogenesis of sleep apnoea but the observations are controversial 12–14, 16, 74. Some investigators report an association between the presence of sleep apnoea and high growth hormone and IGF‐I levels 12, 16, 74, whereas the others fail to show any association between obstructive sleep apnoea and biochemical activity of acromegaly 13, 14. One study found an association between the biochemical activity of acromegaly and central sleep apnoea 14. The high IGF‐I levels in acromegaly may drive breathing and result in increased hypercapnic ventilatory response measured during wakefulness 71, and increased frequency of central apnoea 71 or periodic breathing with symmetric waxing and waning respiratory efforts 15 during sleep.
Treatment of acromegaly with adenomectomy 16 or octreotide 71 may cure acromegaly related OSAS. The operative team should be aware of the risks of performing the trans-sphenoidal adenoma resection in acromegalic patients with sleep apnoea in whom upper airway oedema could potentially further aggravate gas exchange postoperatively 67. Octreotide treatment may promptly alleviate OSAS, and thus its preoperative administration is recommended 72, 75–77. Preoperative nasal CPAP therapy could also reduce the perioperative risks 77. Sedatives have to be avoided and monitoring of breathing should be extended beyond the immediate postoperative period. Perioperative tracheostomy is the safest and sometimes the only alternation to secure breathing after surgery.
After adenomectomy, sleep apnoea persists in every fifth patient, in particular, in those whose growth hormone levels remain high 16. In addition to endocrine factors, the high prevalence of residual SDB after adenomectomy could be related to soft tissue hypertrophy, which remains unaltered. However, uvulopalatopharyngoplasty is not feasible in the treatment of acromegaly related OSAS 78. Nasal CPAP with new pressure titration is often needed after surgery 77.
Growth hormone deficiency
Not only excessive growth hormone production, but also growth hormone deficiency could link with sleep apnoea. Syndromes with hereditary growth hormone deficiency are often associated with obesity, craniofacial and pharyngeal abnormalities predisposing to SDB. However, in lack of comprehensive studies only anecdotal case reports about Laron dwarfism 79 and Turner syndrome 80 would support this. In Prader-Willi syndrome, severe growth hormone deficiency occurs in 38% of adults 81. The pathological somnolence in Prader-Willi patients could be due to nonapnoeic breathing abnormalities rather than episodes of sleep apnoea 82.
Sleep apnoea patients have low growth hormone levels without any specific causes of growth hormone deficiency 83. Growth hormone secretion occurs mostly during sleep, and 70% of nocturnal growth hormone pulses are associated with slow-wave sleep 84, 85. In OSAS, growth hormone secretion is decreased not only due to obesity 86–88, but also because of sleep fragmentation resulting in decreased amount of slow-wave sleep 89. In addition, repetitive hypoxaemia may affect growth hormone secretion. In animals, hypoxia inhibits growth hormone release or biosynthesis 90. Growth hormone deficiency in adults is associated with impaired psychological well-being, insulin resistance, endothelial dysfunction, increased visceral fat, increased cardiovascular mortality and accelerated ageing 91, 92. Similar features are typical in OSAS, which raises the question of a possible link between OSAS-related growth hormone deficiency and the comorbidity seen in OSAS. Indeed, patients with severe OSAS have similar levels of IGF‐I to adult patients with growth hormone deficiency 83. Low IGF‐I may contribute to an increased risk for cardiovascular diseases among sleep apnoeics. Vascular effects of IGF‐I are endothelium-dependent 93, and endothelial cells have IGF‐I receptors 94. IGF‐I increases endothelial cell nitric oxide production 95. Nitric oxide is an important paracrine mediator of vasodilatation and inhibition of vascular smooth muscle cell growth 96.
Two recent reports suggest that growth hormone replacement therapy may also affect sleep and breathing 97, 98. Among 145 children on growth hormone replacement, four developed sleep apnoea; in three cases this was associated with tonsillar and adenoidal hypertrophy 97. Sleep apnoea improved in one patient after cessation of growth hormone therapy, and in all patients following tonsillectomy and adenoidectomy. In five, male, middle-aged patients with postoperative pituitary insufficiency, cessation of growth hormone replacement for 6 months resulted in a decrease of obstructive apnoeic events but in an increase of central apnoeic events 98. Following cessation of growth hormone replacement, slow-wave sleep decreased markedly 98.
At least in theory, an unfortunate coexistence of growth hormone deficiency and SDB would result in a potentially vicious interaction between two altered physiological functions, resulting in severe anatomical abnormalities. A primary growth hormone deficiency could predispose to SDB through short stature, craniofacial growth retardation and low respiratory drive. SDB would further aggravate growth hormone deficiency through sleep disturbance. A primary SDB could aggravate itself by affecting craniofacial and upper airway soft tissue growth through induction of secondary growth hormone deficiency.
In patients with growth hormone deficiency and with predisposing anatomical abnormalities for SDB, a systematic screening for SDB is encouraged. Nasal CPAP treatment and maxillomandibular surgery are feasible therapeutical approaches in these patients. Treatment of SDB may result in normalisation of growth hormone secretion and normal growth in children 99–101. Conversely, symptoms of SDB should also be monitored during growth hormone replacement therapy because of increased risk of SDB.
Cushing's syndrome and Cushing's disease
Shipley and co-workers 22, 23 found sleep apnoea in 45% of their 22 patients with Cushing's disease or Cushing's syndrome. Long-term, high-dose corticosteroid therapy may also contribute to SDB 77. This is of importance especially in patients with juvenile rheumatoid arthritis, whose craniofacial abnormalities (micrognathia) also predispose to SDB.
Pregnancy
Pregnancy has a marked impact on breathing, which is largely mediated through hormones. The levels of the female sex hormones, progesterone and oestrogen, increase markedly. Progesterone increases ventilation 102 and may cause hypocapnia and respiratory alkalosis, and result in respiratory instability and episodes of central apnoea during nonrapid‐eye movement sleep 103. Pharyngeal dimensions decrease during pregnancy 104, nasal congestion and rhinitis are frequent 105, and the enlarging uterus compromises the performance of the diaphragm. Increased oestrogen levels may cause oedema in the upper airway mucosa, and, thereby, be responsible for the upper airway symptoms 106. Conversely, increased female hormone levels may protect the upper airway patency, assuming that upper airway dilators are capable of responding appropriately 107.
Despite marked “central obesity”, neither normal 108, 109 nor multiple pregnancy 110 seems to predispose to SDB. However, in obese females pre-existing SDB may deteriorate during pregnancy 109. In pre-eclampsia, partial upper airway obstruction during sleep is common 111, 112. The long periods of partial upper airway obstruction are associated with increased systemic arterial blood pressure, which can be lowered with nasal CPAP therapy 111.
Snoring is frequent among pregnant females (12–23 versus 4% in nonpregnant women) 113–115. Snoring or OSAS during pregnancy have been suggested to cause intrauterine foetal growth restriction and lower Apgar scores at birth 113, 114, 116. Nasal CPAP therapy also seems safe and effective during pregnancy 111, 117, and early intervention may improve the outcome of the mother and baby.
Polycystic ovary syndrome
A high prevalence rate of SDB in females with polycystic ovary syndrome is a recent observation 30, 31. Previously, SDB in polycystic ovary syndrome was correlated entirely with obesity, but Vgontzas et al. 30 showed that insulin resistance was a stronger determinant of SDB than BMI or serum testosterone levels. The AHI correlates with waist-to-hip ratio and serum total and free testosterone concentrations 31. Suspicion and verification of SDB should not be ignored when females with polycystic ovary syndrome present with compatible symptoms.
Menopause
In clinical studies, male:female ratio of OSAS is ∼10:1 118–120. In community-based populations, the prevalence of OSAS is higher, and the male:female ratio ranges from 2:1–4:1 121–124. Female hormones are thought to protect them from SDB until menopause 125. Among females referred to the sleep clinic, 47% of the postmenopausal and 21% of the premenopausal females had sleep apnoea 126. The observations from community-based studies on the impact of menopause on the prevalence of SDB are not consistent, although most studies show increased prevalence estimates of sleep apnoea after menopause 2, 122, 123, 127. Much of the discrepancy could be attributed to variation in the definition of SDB. Episodes of sleep apnoea seem to grossly underestimate SDB in females, since partial upper airway obstruction is far more common. Of 62 healthy postmenopausal females, 17% had a significant amount of partial upper airway obstruction during sleep 11. In a large community-based study, 1.9% of postmenopausal females and 0.6% of premenopausal females had OSAS, defined as an AHI of ≥10 and occurrence of daytime symptoms 2.
Postmenopausal hormone replacement therapy (HRT) may prevent SDB. In a cross-sectional study, the prevalence estimates of sleep apnoea were almost similar in postmenopausal females without HRT than in males, whereas in HRT users they were compatible with those in premenopausal females 2 (table 2⇓). No significant difference was found between the effect of oestrogen alone or oestrogen plus progestin users.
Effect of sex, menopause and hormone replacement therapy on prevalence of sleep-disordered breathing (SDB)
Short-term administration of progestin alone 128, 129 (fig. 2⇓) or in combination with oestrogen 130, 131 has shown only slight, if any, improvement in SDB in postmenopausal females. However, it is not excluded that long-term HRT may be beneficial in terms of improving SDB. Increasing evidence suggests that menopause is an independent risk factor for SDB. Indeed, SDB partial obstruction, in particular, should be considered in the differential diagnostics of depression, insomnia or restless legs syndrome to explain excessive sleepiness, and fatigue among postmenopausal females.
Decrease in apnoea-hypopnoea index (AHI) with a 2-week medroxyprogesterone acetate (MPA; 60 mg daily) or with 1-night nasal continuous positive airway pressure (CPAP) treatment in six postmenopausal females. Values are expressed as median with the bars representing interquartile range. Based on data from 129.
Androgens
The male predominance of OSAS has been attributed to testosterone-mediated aggravation of SDB or to the lack of a protective effect of female hormones. Androgens do not affect oestradiol or progesterone levels but may reduce their effect by downregulating oestrogen and progesterone receptors 132. Oestrone levels increase with androgens 132. Among seven obese males, all except one with hypogonadism presented with sleep apnoea 133. In males, exogenous testosterone may suppress 134 or augment 135, 136 hypoxic respiratory responses and lead to periodic breathing and sleep apnoea 133, 135, 136. Exogenous testosterone does not affect the upper airway dimensions in males 136.
Few studies have systemically evaluated the effects of exogenous androgen replacement therapy on SDB. Testosterone replacement therapy induced OSAS in one of five males and aggravated pre-existing SDB in another 134. In 11 hypogonadal males, testosterone replacement increased apnoeic events but only in three subjects was the increase considered clinically significant 136. In a placebo-controlled study of 17 elderly males with partial androgen deficiency, testosterone replacement therapy decreased total sleep time and sleep efficiency, and aggravated sleep apnoea 137.
The few available data from females with SDB support the link between androgens and SDB. Irrespective of the menopausal state, obese females have higher androgen levels than nonobese females 138, 139. The prevalence of snoring plateaus or decreases in males after the age of 60 yrs 140, 141. Contrary to the observations in males, females continue to increase their snoring even beyond the age of 60 yrs 140. Those observations suggest that decrease of androgens in ageing males alleviates snoring, whereas menopause-induced oestrogen and progesterone deficiency and increased androgenicity continue to aggravate snoring in postmenopausal females. In a lean, 70-yr-old female, a testosterone-producing tumour caused sleep apnoea, which disappeared after removal of the tumour 142. Exogenous testosterone induces sleep apnoea and even alters the upper airway dimensions in females 143.
However, it is still somewhat controversial whether testosterone contributes to the development or aggravation of SDB. In males with OSAS, androgen blockade with flutamide did not have any effects on ventilatory responses or SDB 144. However, basal testosterone levels may be decreased in such patients and thus the therapeutic response to blockade may be lessened.
After discontinuing testosterone therapy for 2 months in haemodialysis patients, no change in AHI occurred 145. Conversely, 75% of the patients with a clinical history of OSAS were on testosterone therapy, compared with only 35% of those without history of SDB. It is not clear whether these observations reflect the effect of testosterone per se, or possibly the severity of underlying renal disease and disturbances in the acid/base balance leading to respiratory changes.
In OSAS, both morning and nocturnal testosterone concentrations may be decreased 83, 146, 147, but increase after uvulopalatal resection 146. There are several mechanisms in which OSAS may impair testosterone levels. First, in obesity, total testosterone is decreased and in massively obese patients the free testosterone levels may also decrease 148, 149. Secondly, sleep apnoeics are sleep-deprived. Testosterone concentrations fall with prolonged physical stress, sleep deprivation and sleep fragmentation in normal young males 150, 151, including internal medicine residents 152. Thirdly, repetitive episodes of hypoxaemia is typical for OSAS. Hypoxia decreases luteinising hormone (LH) and testosterone levels and alters the circadian rhythm of testosterone secretion 7, 153, 154. Depression of testosterone levels correlates with the severity of hypoxaemia in patients with chronic obstructive pulmonary disease (COPD) or sleep apnoea 7, 153, 154. Testosterone levels rise with oxygen therapy in COPD 155 and with weight reduction in obesity hypoventilation syndrome 156. Fourthly, decreased testosterone levels may be part of an adaptive homeostatic mechanism to reduce SDB assuming that testosterone aggravates SDB.
Androgen replacement therapy is likely to become more common in the treatment of andropausal symptoms in ageing males. With the availability of preparations developed specifically for females, androgen replacement therapy is also likely to become more widespread as a treatment of fatigue, decreased libido or osteoporosis in postmenopausal females 157. Apparition of symptoms suggesting SDB should be monitored in males and in females during androgen replacement therapy.
Effect of nasal continuous positive airway pressure on hormones
Much of the current knowledge on the interactions between hormones and OSAS is based on intervention studies with nasal CPAP. Nasal CPAP is the most efficient therapy to maintain the upper airway patency during sleep. Its efficacy to control sleep apnoea and hypopnoea starts from the very first night of therapy 158. Changes at the levels of several hormones (table 3⇓) are interpreted to be related to SDB or associated sleep disturbance, if they consistently respond to on/off nasal CPAP interventions. Hormonal changes are potential mediators to link SDB with various comorbidities.
Various hormones in obstructive sleep apnoea syndrome (OSAS) and the effect of nasal continuous positive airway pressure (CPAP) therapy on hormones
Diabetes
Among morbidly obese (average BMI 42.7 kg·m−2) patients with sleep apnoea and type‐2 diabetes, nasal CPAP treatment for 4 months improved insulin responsiveness 168. The study population was highly selected, and therefore these results cannot be extrapolated to all type‐2 diabetics. Another study in sleep apnoeics found no effect of a 2‐month nasal CPAP therapy on glucose and insulin metabolism 169. The duration of nasal CPAP therapy was only for 2 months, and thus it cannot be excluded that a longer treatment would improve glucose tolerance.
Leptin
Serum leptin levels decrease with nasal CPAP therapy 33–35 (fig. 3⇓) without weight loss 34, 35. The decrease in leptin levels is already observed after the first night on nasal CPAP 170. Nasal CPAP does not affect the secretory profile of leptin. The nocturnal increase in serum leptin levels is observed both on and off nasal CPAP 170. CPAP-induced reduction in leptin level is likely due to both improved sleep and breathing. Nasal CPAP therapy increases slow-wave sleep and increases growth hormone secretion 160, 162, 171, which in turn inhibits leptin secretion 172, 173. While normalising nocturnal breathing, hypoxic and hypercapnic stimuli may no longer increase leptin secretion 44, 174–177. These findings suggest that nasal CPAP therapy either improves the leptin resistance in obese patients, or with improved ventilation, less leptin is needed to stimulate breathing. It is also possible that nasal CPAP therapy, by normalising sleep structure and growth hormone secretion, may in turn normalise leptin. Moreover, by reducing leptin levels, nasal CPAP therapy is likely to decrease the comorbidity related to OSAS.
Serum leptin levels in 30 patients with obstructive sleep apnoea syndrome (OSAS; └) and in 30 matched controls (▓). The effects of 6-month nasal CPAP treatment on OSAS patients is also shown. Pre-CPAP: □; Post-CPAP: 
. Reproduced from 35 with permission.
Catecholamines
In most studies, nasal CPAP treatment decreases plasma or urinary noradrenaline levels, whereas adrenaline levels in most cases remain unchanged 55, 178, 179. In contrast, in noninsulin-dependent diabetics with OSAS, neither fasting adrenaline nor noradrenaline levels change on CPAP 169.
Thyrotrophin
Although the prevalence of hypothyroidism is not essentially increased, thyrotrophin (TSH) levels may be low in OSAS. In male patients with OSAS, the decrease in serum TSH was most pronounced in patients with the most severe pretreatment nocturnal hypoxaemia. The response to TSH-releasing hormone challenge was normal before and after treatment and was not affected by CPAP treatment 55. However, TSH levels decreased even further after 7 months of CPAP therapy 55.
Growth hormone and insulin-like growth factor‐I
Nasal CPAP therapy increases slow-wave sleep and normalises growth hormone secretion without changes in body weight 160, 162, 171 (fig. 4⇓). Increases in IGF‐I concentrations with CPAP 83 are most probably mediated via increased growth hormone secretion. However, the possible effects of improved nocturnal breathing on growth hormone release cannot be excluded 90. Increased production of IGF‐I 83 and circulating nitric oxide 180 are plausible mediators of the beneficial effect of nasal CPAP on cardiovascular disorders.
Plasma growth hormone (GH) levels measured with frequent sampling during different sleep stages in eight patients with severe obstructive sleep apnoea syndrome (OSAS) before (└) and after (□) nasal continuous positive airway pressure (CPAP) treatment. The relationship between slow-wave sleep (SWS) and GH concentrations becomes significant on CPAP treatment. Wake: stage wake; S1–S2: sleep stage 1–2; REM: rapid-eye movement sleep. Reproduced from 171 with permission.
Testosterone
Both LH and testosterone secretion increase during sleep 181, 182. In young healthy males, there is a sleep-controlled rise in serum testosterone concentration that is linked with the first rapid-eye movement sleep period 183. Since CPAP therapy improves sleep quality, it is also logical that the decreased testosterone levels 83, 146, 147 are normalised on nasal CPAP therapy 83, 146.
Conclusions
Sleep-disordered breathing is still underdiagnosed 184, also when appearing in connection with hormone disorders. There are complex interactions between hormones and sleep-disordered breathing. Nasal continuous positive airway pressure is currently the treatment of choice in mild-to-severe obstructive sleep apnoea 158, 185, but expanding understanding of hormone interactions could provide the tools for powerful alternative therapeutic approaches. Sleep apnoea brings together pulmonologists, endocrinologists, paediatricians, gynaecologists, epidemiologists and many other specialists to combat the sleep apnoea epidemic.
- Received July 10, 2002.
- Accepted February 25, 2003.
- © ERS Journals Ltd
References
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
-
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
