Abstract
Chronic obstructive pulmonary disease (COPD) is no longer considered to affect only the lungs and airways but also the rest of the body. The systemic manifestations of COPD include a number of endocrine disorders, such as those involving the pituitary, the thyroid, the gonads, the adrenals and the pancreas.
The mechanisms by which COPD alters endocrine function are incompletely understood but likely involve hypoxaemia, hypercapnia, systemic inflammation and glucocorticoid administration. Altered endocrine function can worsen the clinical manifestations of COPD through several mechanisms, including decreased protein anabolism, increased protein catabolism, nonenzymatic glycosylation and activation of the rennin–angiotensin–aldosterone system.
Systemic effects of endocrine disorders include abnormalities in control of breathing, decreases in respiratory and limb-muscle mass and function, worsening of respiratory mechanics, impairment of cardiac function and disorders of fluid balance.
Research on endocrine manifestations of COPD embraces techniques of molecular biology, integrative physiology and controlled clinical trials. A sound understanding of the various disorders of endocrine function associated with COPD is prudent for every physician who practices pulmonary medicine.
Patients with chronic obstructive pulmonary disease (COPD) have a chronic illness. They are commonly hypoxaemic, hypercapnic or both, have increased levels of systemic inflammatory markers, and receive glucocorticoids. Each of these factors can affect endocrine function. In turn, altered endocrine function can worsen COPD through several mechanisms. Systemic effects of endocrinopathies triggered by COPD (or therapies used in COPD) include abnormalities in control of breathing, decreases in respiratory and limb-muscle mass and function, worsening of respiratory mechanics, impaired cardiac function, and abnormal fluid balance. In this review, we discuss endocrinopathies triggered by COPD, such as those involving the pituitary, the thyroid, the gonads, the adrenals and the pancreas. Areas of active research and controversial topics are specifically highlighted.
COPD AND THE SOMATOTROPIC AXIS
Weight loss and muscle wasting are present in about 20% of stable outpatients with COPD 1, 40% of patients with acute respiratory failure and 70% of patients requiring mechanical ventilation 2. A decrease in fat-free mass is accompanied by a decrease in muscle mass (sarcopenia) 2, 3. The control of muscle mass is complex and includes the action of inflammatory cytokines, mechanical load on the muscles and anabolic hormones 4–6. There are four main anabolic axes: somatotropic, gonadal, adrenal and insulin 4–8.
Central to the somatotropic axis are growth hormone (GH) and insulin-like growth factor (IGF)-I. GH is released by the hypophysis and stimulates production of IGF-I. The major source of GH-dependent (and insulin-dependent) IGF-I production is the liver (circulating isoforms) 9; however, other tissues, including the skeletal muscles and the adipose tissue, produce both local and circulating isoforms of IGF-I 6.
Although the GH/IGF-I axis is often considered a major regulator of muscle mass, there is accumulating evidence that IGF-I also acts independently of GH 10. The role of locally versus systemically produced IGF-I in determining muscle mass is not clear 11, 12. IGF-I stimulates muscle protein synthesis and hypertrophy, and inhibits protein catabolism via the phosphatidylinositol 3-kinase (PI3K)/Akt (also called protein kinase B) pathway (fig. 1⇓) 16.
Insulin-like growth factor (IGF)-I mediates many of the actions of growth hormone. IGF-I can originate from the liver (endocrine function of IGF-I) or locally (autocrine/paracrine function). a) IGF-I stimulates muscle protein synthesis and hypertrophy via the phosphatidylinositol 3-kinase (PI3K)-Akt pathway. (PI3K can also be activated by insulin and other extracellular signals 13.) Akt, activated by PI3K-induced phosphorylation, stimulates several hypertrophic pathways including glycogen synthase kinase-3β (not shown) and the mammalian target of rapamycin (mTOR). mTOR itself can promote protein synthesis through activation of 70-kD ribosomal S6 protein kinase and by inhibiting eukaryotic translation initiation factor 4E binding protein-1 (not shown) 14. IGF-I also reduces net protein breakdown by inactivation of the forkhead box class O (FoxO) transcription factors 1, 3 through Akt-mediated phosphorylation. Inactivation of FoxO factors 1, 3 decreases the expression of atrophy-related genes or “atrogenes” including atrogin-1 (also known as muscle atrophy F-box or MaF-bx) and muscle ring finger 1 (MuRF-1) both E3-ubiquitin ligases in the ubiquitin-proteasome proteolytic pathway 15 and other atrophy-related genes such as poly-ubiquitin, metallothionein, and cathepsin-L (not shown) 16. In addition to its anabolic and anti-catabolic effects, the IGF-I activation of Akt and inhibition of FoxO increases the resistance of many cells to apoptosis (programmed cell death) 17. b) When IGF-I levels are low, as during administration of glucocorticoids, the protein synthesis triggered by the IGF-I/PI3K/Akt pathway is reduced and protein degradation is increased. Not shown is a second IGF-I intracellular signaling pathway that potentiates proliferation of myoblasts and satellite cells through the activation the mitogen-activated protein kinase cascade 17.
COPD and depression of the somatotropic axis: risk factors
Advanced age, malnutrition, inactivity and administration of glucocorticoids are associated with downregulation of the GH/IGF-I system. In contrast, hypoxaemia and hypercapnia probably increase GH levels in patients with COPD 18, 19.
Ageing
Most patients with COPD are old, and ageing is associated with progressive decrease in muscle mass 20. The mechanisms responsible for age-related decrease in muscle mass probably include decreases in GH/IGF-I 21–23, testosterone 21, 23, 24 and dehydroepiandrosterone sulfate (DHEAS) 7.
Inactivity
In animal models of inactivity, loss of muscle mass has been associated with a decrease in IGF-I mRNA 25. In patients with COPD, inactivity has been implicated in the reduction of quadriceps muscle expression of IGF-I and of myogenic differentiation protein (MyoD): a transcription factor required to maintain stable skeletal muscle differentiation and to induce proliferation and repair by satellite cells in response to muscle injury 10.
Malnutrition
Malnutrition can decrease serum levels of IGF-I 11, 26, 27. In laboratory animals, malnutrition can also decrease IGF-I mRNA expression in the liver and limb muscles but not in the diaphragm 11. Nutritional supplements can increase levels of IGF-I 27.
Glucocorticoids
Steroid myopathy results, at least in part, from a perturbation of the GH/IGF-I system 28, 29. Steroid myopathy can affect respiratory and limb muscles and produce a wide range of manifestations 15. On one end of the spectrum, usually associated with chronic moderate doses of glucocorticoids, is a mild-to-moderate weakness, mild myosin loss, and atrophy of type II fibres and less-so of type I fibres (table 1⇓). On the opposite end, usually associated with acute administration of large doses of glucocorticoids in critically ill patients, is severe paresis, sarcomeric disarray, profound atrophy and acute myonecrosis 15.
Characteristics of the different types of muscle fibres
Studies suggest that glucocorticoid-induced depression of the GH/IGF-I system contributes to the development of steroid myopathy 11, 16, 30–32 (fig. 1⇑). In healthy subjects, the combination of prednisone (0.8 mg·kg−1·day−1) plus recombinant human GH for 7 days increased protein synthesis but did not decrease proteolysis 33. Whether administration of recombinant human GH prevents myopathy in patients is unknown. In laboratory animals, administration of recombinant human GH may 34 or may not 35 attenuate steroid myopathy. In contrast to the conflicting effects of the administration of recombinant human GH on myopathy, administration of IGF-I in laboratory animals with 32 or without emphysema 36 attenuates 36 or totally prevents 32 steroid myopathy.
Hypoxaemia
In patients with COPD and moderate-to-severe hypoxaemia, muscle wasting is associated with depressed protein synthesis 37. Whether or not the depressed protein synthesis is caused by impairment in the somatotropic axis is controversial. Scalvini et al. 18 recorded GH levels in 17 stable patients with COPD (eight of whom were hypoxaemic). In nonhypoxaemic patients, GH levels were 3-fold greater than GH levels recorded in healthy subjects. In hypoxaemic patients, GH levels were 5-fold greater than GH levels recorded in healthy subjects. The investigators speculated that increased levels of GH in COPD might reflect a nonspecific response of the body to stress. Such upregulation of GH could play a role in pulmonary vasculature remodelling and right-sided hypertrophy of patients with pulmonary hypertension 18. That hypoxaemia can increase GH levels is supported by a recent study of Benso et al. 38, which recorded an increase in GH concentration, as well as an increase of total IGF-I, in athletes after 7 weeks of high-altitude exposure to hypoxia. Despite higher levels of IGF-I, hypoxia was associated with a decrease in body weight 38. In contrast to the data in healthy subjects, hypoxaemic children with cyanotic congenital heart diseases have depressed IGF-I 39. Depressed IGF-I is independent from nutritional status 39. In animal models, IGF-I administration only partially reverses the catabolic effects of hypoxaemia 40.
Hypercapnia
In nine patients with COPD who had hypoxaemia, hypercapnia and peripheral oedema, GH concentrations were 11-fold greater than normal 41. In that study, it was not possible to dissect the potential effects of hypoxaemia 18, 38 from those of hypercapnia on the circulating levels of GH. An independent effect of hypercapnia on GH was suggested by the observation that, acute normoxic hypercapnia in mechanically ventilated patients increased levels of GH 19.
Inflammatory cytokines
Circulating and local levels of tumour necrosis factor (TNF)-α and other inflammatory cytokines can be increased in patients with COPD 42. In cultured myoblasts, TNF-α at low concentrations induces a state of IGF-I receptor resistance 43. Such resistance is responsible for inhibition of protein synthesis stimulated by IGF-I 43 and IGF-I-stimulated expression of myogenin, a key myogenic transcription factor 43.
COPD and depression of the somatotropic axis: prevalence
Limited information is available on the interaction between COPD and the somatotropic axis. In 87 patients with COPD enrolled in a rehabilitation program, Casaburi et al. 44 reported that levels of IGF-I were about half the values recorded in healthy subjects. In a follow-up study, they reported that levels of IGF-I were low in 47 men with COPD and low testosterone (mean total testosterone concentration 320 ng·dL−1) 45. As in the studies by Casaburi and colleagues 44, 45, other investigators have reported a high prevalence of decreased IFG-I in stable patients with COPD 46, 47 and in patients experiencing an exacerbation 47. The high prevalence of decreased IGF-I in several investigations 44–48 raises the possibility that a depression of the somatotropic axis may contribute to the decreased muscle mass in some patients with COPD.
In contrast to the preceding studies 44–48, other investigators 14, 49 found that IGF-I in patients with COPD and muscle atrophy is equivalent to that in healthy subjects. These contrasting results may be reconciled by the variable concentration of ghrelin, a novel GH-releasing peptide synthesised by the stomach, anterior pituitary and hypothalamus 22. Levels of ghrelin can be higher in underweight patients with COPD than in normal-weight patients with COPD and healthy volunteers 46. Decreased levels of IGF-I in patients with COPD and muscle atrophy 14, 49, despite higher levels of ghrelin 46, might indicate that the somatotropic axis in these underweight patients has developed resistance to the biological functions of ghrelin. The aforementioned studies 14, 49 are limited by the failure to determine whether the local expression and function of IGF-I, and its downstream pathways (fig. 1⇑), are affected by COPD.
The pathways downstream of IGF-I have been recently assessed by Doucet et al. 14. Patients with COPD and muscle atrophy had greater expression of some components of the catabolic pathways in the quadriceps than healthy subjects, including increased levels of atrogin-1 and muscle ring finger protein (MuRF)1 mRNA and of forkhead box class O protein (FoxO)-1 (fig. 1⇑). What role the IGF-I system has in determining muscle atrophy is difficult to assess. First, serum levels of IGF-I in patients with COPD and muscle atrophy were similar to the corresponding levels in healthy subjects 14. Yet, muscle IGF-I mRNA expression in hospitalised and in stable patients with COPD measured by other investigators 10 was lower than in healthy elderly subjects. Secondly, compared with patients with COPD who have muscle atrophy, patients with preserved muscle mass showed similar activation of the atrophying factors and increased expression of some components of the anabolic pathways. It is possible that patients with muscle atrophy are not able to respond to the upregulated hypertrophy signalling pathways 14. Possible mechanisms include ageing and dysfunction of translational efficiency or capacity 14. This mechanism is also supported by the observation that levels of GH are higher in cachectic patients with COPD than in noncachectic patients 50. The latter finding suggests that acquired GH resistance contributes to the cachexia 50.
Upregulation of the IGF-I system within the fibres of the diaphragm has been reported within 4 days after lung-volume reduction surgery in hamsters 51. This upregulation is likely to be an anabolic response to passive stretch of the diaphragm and ultrastructural sarcomeric injury 51.
COPD and depression of the somatotropic axis: functional consequences
Decreased muscle mass
In patients with COPD, decreased muscle mass reduces respiratory muscle function, limb muscle function, exercise capacity and life expectancy 15. As already stated, whether a depression of the somatotropic axis contributes to decreased muscle mass in COPD, however, is unclear, and administration of recombinant human GH has produced mixed results.
In seven patients with COPD and ideal body weight (Metropolitan height–weight tables for adults) of <90%, 3 weeks of recombinant GH increased weight, improved nitrogen balance and maximal inspiratory pressure 52. These results contrast with the negative results of Burdet et al. 48, the only randomised-controlled trial in malnourished patients with COPD. Several factors may explain the negative results of Burdet et al. 48: first, redistribution of protein toward central organs rather than toward the muscles 48; secondly, small dose and short duration of therapy (3 weeks) 48; thirdly, uncoupling of Akt (activated by IGF-I) and the downstream anabolic and catabolic pathways (fig. 1⇑) 14. Lack of beneficial effect of recombinant GH on muscle function has also been reported in patients requiring prolonged mechanical ventilation, half of whom had COPD 53; recombinant GH can increase mortality of critically ill patients 54.
In the study of Burdet et al. 48, maximal power output and oxygen consumption at the conclusion of a symptom-limited exercise test were not affected by recombinant GH administration. The response to six-minute walk was decreased in the treated patients 48. A potential mechanism for this decrease in submaximal tests is an increase in oxygen consumption and carbon dioxide production induced by GH 48. This possibility, however, is not supported by data recorded in young patients with COPD caused by cystic fibrosis in whom recombinant GH administration produced a significant increase in exercise capacity 55, 56.
In adults, chronic recombinant GH therapy can be associated with significant side effects including arthralgia, carpal tunnel syndrome, oedema and abnormal glucose tolerance 57, 58. The potential risk of development and progression of common cancers including colon, breast and prostate cancers with chronic use of IGF-I 59 raises important questions on selection criteria, monitoring and safety of any study of recombinant GH/IGF-I supplementation in COPD.
Recently, Nagaya et al. 60 investigated a novel pharmacological strategy to modulate the somatotropic axis in patients with severe COPD. Administration of ghrelin for 3 weeks produced increased appetite, increased circulating GH, increased peripheral and respiratory muscle strength, increased six-minute walking distance and decreased plasma norepinephrine; levels of IGF-I were unaffected by ghrelin. These positive findings 60 must be interpreted with caution because the study was open labelled and lacked a control arm.
In patients with severe COPD, nonpharmacological upregulation of anabolic signals in the vastus lateralis muscle, including IGF-I and MyoD proteins and mRNA of IGF-I, and its load-sensitive splice variant (mechanogrowth factor), can be induced by pulmonary rehabilitation 61. This favourable response is associated with a reduction in the proportion of type IIb muscle fibres (table 1⇑), increases in the mean cross-sectional area of all fibre types and improved peak exercise work rate 61. The anabolic effect occurred despite the fact that rehabilitation did not decrease plasma levels of inflammatory mediators 61, 62 and it did not decrease mRNA expression of TNF-α 61, 62 and interleukin (IL)-6 in the muscle 61.
COPD and depression of the somatotropic axis: treatment
Despite the plausibility of linking derangements of the somatotropic axis and decreased muscle mass in COPD, it is premature to recommend the use of recombinant GH, IGF-I or ghrelin in patients with COPD and decreased muscle mass.
COPD AND GONADAL AXIS
In men and women, the gonadal axis is a complex network of hormones that includes testosterone, an important anabolic hormone 63. In men, the main production site of androgens is the testicle 63; in women, the adrenals and the ovaries 64. Secretion is stimulated by luteinising hormone (LH), one of the pituitary gonadotropins 65. Under physiological conditions, about 44% of circulating testosterone is strongly bound to the sex hormone binding globulin; 54% is weakly bound to albumin and other lower affinity, high-capacity binding sites, such as α1 acid glycoprotein and transcortin 66. About 2% of circulating testosterone is nonprotein bound, i.e. free testosterone. Together, weakly bound testosterone and free testosterone are referred to as “bioavailable testosterone” 66. Women have much lower levels of circulating testosterone than men. In both men and women, testosterone is responsible for libido, sexual hair and anabolic functions that affect muscle and bone 64, 67.
In patients with COPD, testosterone supplementation can increase leg muscle mass 45 and vastus lateralis mRNA for slow/β-myosin heavy chain 68. These anabolic responses are accompanied by increases in vastus lateralis IFG-I protein and myogenin mRNA 68. In these patients, circulating levels of IGF-I 45 and intramuscular expression of MyoD mRNA and myostatin mRNA 68 are not affected by testosterone supplementation. Taken together, these data suggest that androgens produce an enhanced local anabolic milieu, driven in large part by the muscle’s IGF system. Increased size of motor neurons 69 is an additional mechanism through which androgens may exert their anabolic action.
The concentration of serum testosterone 70, 71 and other androgens, such as the adrenal steroid dehydroepiandrosterone (DHEA), a testosterone precursor 72, decline in middle-aged and elderly men. When excessive, this decline may contribute to various signs and symptoms, including decreased energy level, libido, bone density and muscle mass and, possibly, reductions in cognitive function and memory 73. This constellation of signs and symptoms has been termed late-onset hypogonadism, symptomatic late-onset hypogonadism, androgen deficiency in the ageing male or andropause 67. The observation that many men with COPD fit the profile of late-onset hypogonadism 74–76 has spurred a flurry of research on the incidence 76–78, functional impact 49, 76, 78 and possible treatment 45 of this abnormality.
As in men, levels of testosterone and of other androgens, such as DHEA, decline in women with ageing 24. This process starts before, and continues after, the menopause 24. Little is known about the impact of declining levels of testosterone and other androgens in women with COPD. Accordingly, most of our discussion will focus on late-onset hypogonadism in men with COPD.
COPD and late-onset hypogonadism: risk factors
Several risk factors that may decrease testosterone in patients with COPD have been reported. These include ageing, chronic disease, hypoxaemia, hypercapnia, smoking, administration of glucocorticoids, systemic inflammation and obesity.
Ageing
Purported mechanisms by which old age causes hypogonadism in men include decreased responsiveness of the testes to human chorionic gonadotropin (which acts as LH), increased testosterone feedback at the hypothalamic-pituitary level, decreased responsiveness of the pituitary to gonadotropin releasing hormone, and asynchronous (chaotic) release of gonadotropin releasing hormone from the hypothalamus 79. In men, sex hormone binding globulin increases with age 23 and reduces the biological availability of testosterone 23.
The mechanisms responsible for age-related decrease in testosterone in women include decreased adrenal production of DHEA and DHEAS 80, decreased ovarian production of androgens 80 and increased specific activity of aromatase in the adipose tissue 81.
Chronic illness
Chronic illnesses, including diabetes, cardiovascular disease and hypertension, have been associated with a decline in serum testosterone 82. An important contributor to this decline is a decrease in LH (hypogonadotrophic hypogonadism) 82, the most common presentation of late-onset hypogonadism in men with COPD 45, 76. Whether COPD is a “chronic illness” that can cause, by itself, hypogonadism is controversial considering that among ambulatory men with COPD severity of lung disease does not predict the hormonal abnormality 76, 78. Also, the number of comorbid conditions in hypogonadal and eugonadal men with COPD is similar 78.
While the prevalence of hypogonadism seems to be unaffected by severity of COPD when patients are in a stable clinical condition 76, 78, acute exacerbations 83 and, possibly, the need for prolonged mechanical ventilation 84, can (further) decrease testosterone.
Hypoxaemia/Hypercapnia
In 1980, Semple et al. 85 assessed serum levels of testosterone and gas exchange in 22 men with COPD. When the partial pressure of arterial oxygen (Pa,O2) dropped below 55 mmHg, there was a strong correlation between the degree of hypoxia and the degree of testosterone reduction. Lack of association between testosterone and hypoxia (when Pa,O2 is >55 mmHg) has been confirmed by more recent studies 49, 76, 78. Semple et al. 85 also reported a correlation between the degree of hypercapnia and testosterone.
Smoking
In men, both testosterone and sex hormone-binding globulin are greater in smokers than in nonsmokers, yet, levels of bioavailable testosterone are similar 86. These results suggest that smoking modulates total testosterone through changes in the concentration of sex hormone-binding globulin but has no significant impact on the biologically active fraction of the hormone.
Glucocorticoids
The effect of glucocorticoids on testosterone concentration is controversial 49, 76–78. In 36 men with COPD, 16 of whom were taking oral glucocorticoids, Kamischke et al. 77 found that free testosterone was inversely related to glucocorticoid dosage. Mechanisms that may explain the decrease in levels of testosterone include a reduction in gonadotropin releasing hormone, decrease in adrenal precursors and reduction of testosterone biosynthesis 77. In addition, glucocorticoids may compete with testosterone for the same receptor sites in the peripheral tissues 87. Despite these results 77, three groups of investigators have more recently reported no correlation between testosterone and glucocorticoid therapy in men with COPD 49, 76, 78.
Systemic inflammation
In laboratory animals 88 and healthy men 89, TNF-α can lead to decreased testosterone. Mechanisms through which TNF-α and other inflammatory cytokines, such as IL-1 and IL-6, could decrease testosterone include increased expression of aromatase, the enzyme that irreversibly converts testosterone to oestradiol (via the testosterone–oestradiol shunt) 90, a reduction in gonadotropin releasing hormone 65 and a cytokine-mediated testicular defect 88, 91. Despite the biological plausibility, a recent investigation in more than 100 men with COPD 83 suggested that the increased levels of inflammatory cytokines in patients with stable COPD were insufficient to decrease testosterone concentrations.
Obesity
Body mass index of men with COPD and late-onset hypogonadism is greater than that of eugonadal men with COPD with similar age and severity of lung disease 76. Increased body mass index could result from a positive feedback between obesity and hypogonadism. Hypogonadism is associated with a preferential deposition of adipose tissue in the abdomen 92. A state of leptin resistance may contribute to this increased adipose deposition 93; in men, levels of leptin have been reported to be inversely related to testosterone 94. A second mechanism is a decrease in serum concentrations of ghrelin 95 resulting in less fat utilisation 96. Irrespective of the mechanism, greater adipose tissue accumulation increases aromatase activity and, thus, the conversion of testosterone to oestradiol (testosterone–estradiol shunt) 90. Such an upregulation could produce a vicious cycle where depression of testosterone concentration causes deposition of abdominal fat that, in turn, worsens the hypogonadal state 97.
COPD and late-onset hypogonadism: prevalence
Laghi et al. 76 tested 101 men with stable COPD for late-onset hypogonadism. Late-onset hypogonadism was present in 38% of the men. This value falls between the prevalence of 69% reported by Kamischke et al. 77 and 22% reported Debigare et al. 49.
The 47% difference in the prevalence of hypogonadism between the studies of Kamischke et al. 77 and Debigare et al. 49 cannot be ascribed to age or severity of COPD. It may be related to patient selection: 84% of the patients in the first study had concurrent chronic diseases 77, whereas patients with chronic diseases were excluded in the second study 49. A difference in comorbidities can explain only 15% of the 47% difference 49. Sample size is a more likely explanation. When data from the two investigations 49, 77 are combined, 35 (43%) of 84 patients exhibit hypogonadism; similar to that in the study of Laghi et al. 76.
Van Vliet et al. 78 reported a prevalence of late-onset hypogonadism in 51% of 78 men with COPD and 26% of 21 age-matched controls. These results do not necessary mean that late-onset hypogonadism is more prevalent in men with COPD than in the general population. The prevalence of hypogonadism in the small control group 78 was much lower than the 34–40% prevalence of hypogonadism for subjects in their 60s and the near 70% prevalence for subjects in their 70s, reported in large population studies 23, 70. Levels of testosterone did not correlate with severity of airway obstruction 76, 78 or with a Pa,O2 >55 mmHg or glucocorticoid therapy 49, 76, 78.
COPD and late-onset hypogonadism: functional consequences
Sexual function
In a study of 36 men with COPD, 25 of whom were hypogonadal, the prevalence of sexual difficulties was not associated with the level of free testosterone 77.
Quality of life
COPD 98–100 and late-onset hypogonadism 101 can decrease quality of life. To determine whether the occurrence of late-onset hypogonadism had an additive effect on reduced quality of life in COPD, Laghi et al. 76 compared the results of the St George's Respiratory Questionnaire (SGRQ), a respiratory-specific questionnaire, and of the Veterans Short Form (SF)-36 Health Survey, a general health questionnaire, in 38 men with COPD and late-onset hypogonadism and 63 eugonadal men with similar severity of airway obstruction 76. The questionnaire scores were similar in the two patient groups. Free testosterone was not associated with the perceived impact of COPD on health and well being 76.
Women with COPD report worse health-related quality of life than men 102. Whether menopause and hormone supplementation in women with COPD could affect the health-related quality of life is untested.
Depression
Decreased testosterone levels 103, 104 and COPD 105 have been associated with depressed mood. Yet, the SGRQ impact score 106 and the SF-36 mental health domain 107, which are affected the presence of depression are equivalent in eugonadal and hypogonadal men with COPD 76.
Respiratory muscles
Recently, Laghi et al. 108 reported that diaphragmatic contractility was similar in hypogonadal and eugonadal men with COPD (fig. 2⇓). When performing a progressive inspiratory threshold loading protocol, hypogonadal and eugonadal patients had similar respiratory muscle endurance times, perceived inspiratory effort and perceived air hunger 108. Neither group of patients developed diaphragmatic fatigue after loading (fig. 2⇓). The lack of fatigue was similar to its absence after failed weaning from mechanical ventilation 109 and under more extreme conditions in laboratory animals 110.
Potentiated (⨐, ▪) and nonpotentiated (○, □) transdiaphragmatic twitch pressure (twitch Pdi) in 10 hypogonadal (⨐, ○) and 10 eugonadal (▪, □) men with COPD recorded before (baseline) and 20 min after conclusion of inspiratory threshold loading to task-failure. Potentiated and nonpotentiated twitch Pdi values were not affected by hormonal status or by threshold loading. Date are presented as mean±se. Reproduced and modified from 108 with permission from the publisher.
The equivalent performance of the respiratory muscles in hypogonadal and eugonadal men 108 is consistent with the observations of Casaburi et al. 45, who found that inspiratory muscle strength was not altered by the administration of testosterone enanthate to patients with COPD who had variable serum testosterone levels.
Several mechanisms may protect the respiratory muscles against the deleterious effects of low testosterone levels. First, a conditioning effect of the increased inspiratory load 111 may counterbalance the catabolic effects 112, 113. Secondly, in the investigation of Laghi et al. 108, most patients were not underweight. Thirdly, respiratory muscles may be less sensitive to testosterone (and lack thereof) than the limb muscles considering that administration of massive doses of testosterone (20 mg·kg−1·day−1 for 14 days) to rabbits did not produce increased diaphragmatic strength or endurance 110.
Limb muscles
It is unclear whether late-onset hypogonadism in men with COPD affects limb muscle strength and endurance. In 78 men with COPD, Van Vliet et al. 78 recorded a correlation between quadriceps strength and testosterone concentrations. Conversely, Laghi et al. 108 recorded no difference in quadriceps strength (and endurance) between hypogonadal and eugonadal men with COPD. This equivalent limb muscle performance in the two patients groups is in line with the observation that testosterone has no effect on leg-muscle endurance when administrated at physiological doses in otherwise healthy hypogonadal men 114 or in healthy eugonadal men receiving a long-acting gonadotropin-releasing hormone agonist 115. Likewise, the equivalent limb-muscle performance in the two groups is consistent with the observation that physiological doses of testosterone had no effect on leg-muscle strength in some hypogonadal men 116–118. Similarly, when Bhasin et al. 119 administered a long-acting gonadotropin-releasing hormone agonist to induce hypogonadism in elderly men and then supplemented the men with sub-therapeutic doses of testosterone enanthate, they recorded no difference in fat free mass or muscle strength. When they doubled the circulatory levels of total and free testosterone using higher doses of testosterone enanthate, the increases in fat free mass and strength were within the noise of the measurement. They, however, reported a 17% increase in quadriceps strength (and endurance) when administering 100 mg·week−1 of testosterone enanthate for 10 weeks in men with COPD who had variable testosterone levels 45. It is not known why low testosterone should cause a decrease in strength in some studies and not in others, nor whether the statistical differences in quadriceps strength, when present, are clinically important. The same investigators reported preliminary results on testosterone supplementation (12.5 and 25 mg testosterone enanthate weekly for 10 weeks) in women with COPD 120. Only modest changes in body composition and strength without evidence of virilisation were observed.
Exercise capacity
Exercise capacity in men with COPD with and without late-onset hypogonadism has been found to be equivalent 76, 78, 108. During cycle exercise to exhaustion, exercise performance, gas exchange and respiratory muscle recruitment (estimated by oesophageal and gastric pressure swings during tidal breathing) were similar in both groups 108. Equivalent exercise capacity has been recorded even when quadriceps strength among hypogonadal patients with COPD was less than in eugonadal patients with COPD 78.
The similar exercise performance in patients with COPD with or without late-onset hypogonadism 76, 78, 108 is consistent with the observation that administration of anabolic steroids to unselected patients with COPD 121, 122 or testosterone enanthate to men with COPD and variable testosterone levels 45 produced no improvement in whole body exercise capacity.
Risk for cardiovascular disease
Several case–control studies suggest an association between low testosterone levels and excess cardiovascular risk in men 123–125, while the opposite may be true in women 126. It is not known whether late-onset hypogonadism has an additive effect on the already increased risk of cardiovascular disease associated with COPD 127–131.
Bone density
Reduced bone mineral density is common in COPD 74. Reduced bone mineral density has not been associated with low testosterone levels in two studies 132, 133, possibly because the risk of osteoporosis is already so high in COPD that the additional effect of late-onset hypogonadism does not further affect bone mineral density.
COPD and late-onset hypogonadism: therapy
If patients are deficient in a hormone, it might seem self-evident to replenish normal levels. Before recommending hormone supplementation, however, it is critical to weigh potential risks and benefits. Stated differently, does it matter if a man with COPD develops late-onset hypogonadism? The question entails whether clinically important functional differences between hypogonadal and eugonadal patients do exist, which, in turn, could have important therapeutic implications. For some clinical aspects, the accumulating evidence suggests that men with COPD and late-onset hypogonadism are not different from their nonhypogonadal counterparts. Therefore, if testosterone is administered, it would be unclear what should be the therapeutic effects to monitor.
The caution about whether to advocate testosterone replacement in men with late-onset hypogonadism and COPD needs to reflect the ongoing controversy regarding indications for testosterone replacement in older men in general 67, 73, 134. Long-term testosterone supplementation can be associated with side effects, including increase in haematocrit, sleep apnoea, prostatic hypertrophy (fig. 3⇓) 67, 73. Long-term effects of testosterone administration on the risk of atherosclerotic heart disease and prostate cancer remain unknown 67, 73. More than 6,000 elderly hypogonadal men, randomly assigned to receive testosterone or placebo for 6 years, would be necessary to determine whether testosterone treatment increases the risk of prostate cancer by 30% 73. The Institute of Medicine (Washington DC, USA) recently concluded that, unless more convincing studies are published, there is currently insufficient evidence to support testosterone therapy in older men 135.
Schematic representation of the mechanisms of action of testosterone, of the tissues affected by testosterone, and the potential effects of testosterone-replacement therapy. Testosterone has protean effects because it can act as three different hormones. In some tissues, testosterone binds to the androgen receptor and acts directly as an androgen. In tissues expressing the enzyme 5α-reductase, it is converted to dihydrotestosterone. Dihydrotestosterone binds to the androgen receptor with greater affinity than testosterone. Consequently, dihydrotestosterone has effects that a physiological concentration of testosterone does not have. Finally, in tissues expressing the enzyme aromatase, testosterone is terminally metabolised to oestradiol, which acts through the oestrogen receptor. Reproduced and modified from 73 with permission from the publisher.
COPD AND ADRENAL AXIS
The adrenal glands produce a vast array of hormones with protean metabolic effects; among these are cortisol, DHEA and its metabolite, DHEAS, the most abundant steroid present in the blood 49. The action of DHEAS at the tissue level is both direct, through specific binding sites of the hormone in peripheral tissues 136, and indirect. The indirect action occurs as a result of the conversion DHEAS to androstenedione or androstenediol, and finally to testosterone 137. Reduced levels of DHEA and DHEAS and high cortisol/DHEA or cortisol/DHEAS ratios are thought to create an imbalance between protein synthesis and degradation favouring catabolism over anabolism 49.
In contrast to the anabolic properties of DHEA and DHEAS, cortisol (when in excess) has catabolic properties 138. Cortisol mobilises glucose, free fatty acids and amino acids from endogenous stores 138. In addition, cortisol increases appetite and induces insulin resistance 138.
COPD and imbalance of the adrenal axis: risk factors
COPD and imbalance of the adrenal axis: prevalence
The limited available information suggests circulating levels of DHEAS are decreased in COPD 49, 83. Debigare et al. 49 found lower DHEAS levels in patients with COPD and decreased muscle mass than in those without decreased muscle mass or in control subjects. Similarly, Karadag et al. 83 found lower DHEAS levels in patients with COPD than in controls. During acute exacerbations DHEAS levels decreased even further 83.
Whether cortisol levels are altered in patients with stable COPD is unclear 18, 41, 141, 142. Systemic and possibly (high-dose) inhaled corticosteroids increase the risk of adrenal insufficiency 143.
During acute exacerbations, levels of cortisol are increased in patients not treated with systemic corticosteroids 144. In these patients, the cortisol response to adrenocorticotropic hormone (ACTH) is normal 144. In contrast, patients treated with corticosteroids during an exacerbation have decreased circulating cortisol levels and commonly limited cortisol response to ACTH 145. The limited response can last more than 3 weeks after corticosteroid withdrawal 145. Neither glucocorticoid dose nor duration of treatment can be used to predict adrenal insufficiency 146.
COPD and imbalance of the adrenal axis: functional consequences
Decreased muscle mass/decreased muscle function
In men with severe COPD, the cortisol/DHEAS ratio was greater among patients with decreased muscle mass than among those without decreased muscle mass 49. Nevertheless, administration of DHEA, for 2 years, to elderly men and women without COPD and reduced DHEAS levels had no effect on body composition, exercise capacity, muscle strength or quality of life 147.
COPD and imbalance of the adrenal axis: therapy
In elderly patients without COPD, DHEA administration had no significant benefit 147.
COPD AND DIABETES
Insulin is an anabolic hormone that exerts its effects by binding to its cognate receptor 148. Many cells throughout the body, including liver, lung, skeletal muscle, adipose tissue and vascular endothelial cells, express this transmembrane receptor 149, 150. When insulin binds to the receptor in myocytes, it activates the intrinsic tyrosine protein kinase activity 8. The activated receptor phosphorylates the insulin receptor substrates 1 and 2, which, in turn, activate PI3K and, with it, the mTOR signal transduction pathway (fig. 1⇑) 8. Insulin is a mitogen and growth factor of airway smooth muscle cells and of bronchial and alveolar epithelial cells 149. In vitro, insulin improves hypoxia-induced vasoconstriction and causes pulmonary artery vasodilation 149.
Diabetes mellitus can result from destruction of the pancreatic β cells, which leads to absolute insulin deficiency (type 1) or, more often, from insulin resistance coupled with relative impairment in insulin secretion (type 2) 151. Abnormalities of the insulin receptor and abnormalities downstream to the receptor contribute to insulin resistance in patients with type 2 diabetes 150.
COPD and diabetes: prevalence/risk factors
The prevalence of diabetes in patients with COPD is 10–14% 78, 152–154. The relationship between impaired pulmonary function and increased risk of developing diabetes is controversial 131, 154–156. Among almost 8,000 participants in the Framingham Heart Study 154 and the National Health and Nutrition Examination Survey 155, there was no association between COPD and the development of diabetes. In contrast to these studies, in which about half the participants were men, COPD was found to increase the risk of diabetes in the Nurses' Health Study (involving nearly 100,000 female nurses; multivariate relative risk 1.8, 95% CI 1.1–2.8) 156. Likewise, among more than 20,000 participants (56% female) of the Atherosclerosis Risk in Communities Study (ARIC) and the Cardiovascular Health Study (CHS), the presence of moderate-to-severe COPD was shown to increase the risk of diabetes (multivariate relative risk 1.5, 95% CI 1.1–1.9) 131. It is not possible to reconcile the contrasting results. The mechanisms through which COPD might induce type 2 diabetes include systemic inflammation, oxidative stress, smoking and administration of glucocorticoids.
Smoking/systemic inflammation/oxidative stress
Cigarette smoking can increase the risk of insulin resistance and diabetes by triggering systemic inflammation 154, 157 and oxidative stress 158; both of which are common in COPD 154, 157–159. In nonhypoxaemic nondiabetic patients with COPD, insulin resistance was linked to systemic inflammation (via IL-6) and body mass index 160.
Glucocorticoids
In elderly patients, administration of systemic glucocorticoids more than doubles the risk of developing diabetes 161. Potential mechanisms include direct inhibition of insulin release, reduction in the hepatic levels of insulin receptor substrate-1 and PI3K, decrease in glucose uptake and increase in hepatic glucose production 162, 163.
Systemic and topical glucocorticoids are used in patients with COPD during exacerbations and as maintenance therapy. Systemic corticosteroids can interfere with glycaemic control 161, 164 and poor glycaemic control in patients hospitalised for COPD exacerbation has been associated with increased length of hospital stay and increased mortality 165. Diabetes is a risk factor for mortality, even after hospital discharge for COPD exacerbations 166. It is unclear whether short- 165 and long-term survival 166 following an acute COPD exacerbation can be improved by tight glycaemic control.
In contrast to systemic glucocorticoids, administration of inhaled glucocorticoids is probably safe in patients with COPD and diabetes 161, 167. In 10 patients with type 2 diabetes and asthma or COPD, inhaled glucocorticoids for 6 weeks caused no change in glycosylated haemoglobin 167.
COPD and diabetes: functional consequences
In patients who do not necessarily have COPD, diabetes can negatively affect respiratory muscle function, pulmonary mechanics, gas exchange and respiratory drive. Patients with diabetes are also at increased risk of infections and cardiovascular complications.
Respiratory muscles
Patients with diabetes exhibit a 20% decrease in global inspiratory strength 168 and a 30–50% decrease in diaphragmatic strength 168, 169. Weakness might result from nonenzymatic glycosylation (glycation) of muscle fibres (fig. 4⇓) 170, 171. Rarely, diaphragmatic weakness results from diabetic neuropathy of the phrenic nerves 172, 173, which can occur despite the absence of peripheral neuropathy 173.
Effect of nonenzymatic glycosylation on the speed of actin filaments sliding over a) slow myosin (soleus muscles) and b) fast myosin (extensor digitorum longus) in control muscle fibres of rats incubated with low-salt buffer (○) and with 6 mM glucose (⨐). Incubation of muscle fibres for 30 min in low-salt buffer had no effect on motility speed in the slow (a) and fast (b) myosin preparations. In contrast, incubation with glucose for 30 min caused a progressive decrease in motility speed both in the slow (a) and fast (b) myosin preparations (p<0.05). Data are presented as mean±sd. Reproduced and modified from 170 with permission from the publisher.
Pulmonary mechanics/gas exchange
Lung specimens in patients with diabetes demonstrate several abnormities that are reminiscent of those described in the glomeruli. These abnormities include intraseptal nodular fibrosis, microangiopathy, thickening of the basal laminae of pulmonary capillaries and loss of alveolar microvascular bed associated with increased extracellular matrix and connective tissue 149. Thickening of the basal laminae of pulmonary capillaries and loss of alveolar microvascular bed 149 may contribute to increased work of breathing, decreased diffusing capacity 149 and dyspnoea in patients with diabetes 174.
Respiratory drive
The ventilatory response to hypercapnia is normal in diabetic patients without autonomic neuropathy, reduced in patients with parasympathetic neuropathy and increased in patients with combined parasympathetic and sympathetic neuropathy 175. The differences have been attributed to proportional variation in cerebrovascular reactivity (velocity of blood flow in the middle cerebral artery) to hypercapnia 175. Parasympathetic neuropathy, with secondary depression of vagal tone, may contribute to diminished bronchodilator response to anticholinergic drugs in some patients with diabetes 149.
Infections
Diabetes increases the risk of lower respiratory tract infections 176. In patients with COPD exacerbations requiring noninvasive ventilation, hyperglycaemia is associated with pulmonary infections and is an independent risk factor for failure of noninvasive ventilation 177. In euglycaemic subjects with airway inflammation or in hyperglycaemic patients, glucose concentrations in airway secretions are increased 178. Mechanically ventilated patients with detectable glucose in bronchial aspirates have increased likelihood of pathogenic bacteria, particularly Staphlococcus aureus 179. Airway glucose may promote bacterial growth or interfere with local immune defences 165.
Osteoporosis
Diabetes has been associated with higher bone mineral density in both women and men 180, 181. The higher bone mineral density could result from the hyperinsulinaemia that precedes overt type 2 diabetes 181. In turn, hyperinsulinaemia could exert an anabolic effect on bone tissue 181. Hyperinsulinaemia has been reported in nondiabetic patients with COPD 160. Despite a higher bone mineral density, patients display an increased risk of fractures, mainly attributable to the increased risk of falling 182.
COPD and diabetes: therapy
The goal of diabetic care is to achieve glucose levels close to normal without inducing significant hypoglycaemia 183. With the exception of inhaled insulin 184, strategies to reach this goal in COPD should be the same as for diabetic patients without COPD 183, 185, 186. Detailed reviews of this topic are available elsewhere 183, 185, 186.
In 2006, US and European Drug Agencies (the US Food and Drug Agency and the European Medical Association) approved inhaled insulin for clinical use in nonsmoking adults with type 1 and type 2 diabetes 187. Inhaled insulin can cause small but significant decreases in diffusing capacity and FEV1 184. In addition, absorption of inhaled insulin in COPD is unpredictable 184 and, at least in ex-smokers, inhaled insulin might increase the rate of bronchogenic carcinomas 187. Whether inhaled insulin should be even considered in diabetic patients with COPD has become a moot point: in 2007 inhaled insulin was discontinued as a result of poor market acceptance 187.
The oral anti-hyperglycaemic agent, metformin, is thought to increase the risk of lactic acidosis, and is considered contraindicated in chronic hypoxaemic conditions 188. Despite these concerns, a recent meta-analysis of 206 trials revealed no cases of fatal or nonfatal lactic acidosis in 47,846 patient-yrs of metformin use or in 38,221 patients-yrs in the non-metformin group 188.
COPD AND THYROID DISEASES
The thyroid hormone regulates the metabolism of proteins, lipids and carbohydrates, and controls the activity of membrane-bound enzymes 189, 190. This hormone can also regulate the transcription of numerous genes encoding both myofibrillar and calcium-regulatory proteins in myofibres (table 2⇓) 191, 192. The thyroid hormone enhances mitochondrial oxidation and, thus, augments metabolic rate 193. This effect on metabolic rate is probably responsible for the association between the thyroid hormone and respiratory drive 194.
Muscle genes responsive to thyroid hormone
Limited data on the prevalence of thyroid diseases among patients with COPD are available 195, 196. Yet, several characteristics of patients with COPD could potentially increase their likelihood of developing hypothyroidism and hyperthyroidism.
COPD and hypothyroidism: prevalence/risk factors
Impaired thyroid function can present as subclinical hypothyroidism, overt hypothyroidism and nonthyroidal illness syndrome. Of these, nonthyroidal illness syndrome is the most common in COPD 195, being reported in 20% of stable patients 195 and 70% of patients experiencing an exacerbation 195. Patients with nonthyroidal illness syndrome have decreased levels of the biologically active hormone tri-iodothyronine (T3) and normal or decreased levels of the prohormone thyroxine (T4) 197. Serum levels of the TSH are usually normal, though they can be low in the most severely ill patients 197. In the past, patients with nonthyroidal illness syndrome were said to have euthyroid sick syndrome. The latter nomenclature, however, is now considered inappropriate because it is not clear whether these patients are in fact euthyroid in those tissues that are targets for the action of the thyroid hormone 197. To date it is impossible to state which patient with nonthyroidal illness syndrome, if any, might benefit from treatment, nor what that treatment should be 198. It is not known if nonthyroidal illness syndrome is a favourable compensatory mechanism to offset catabolism and protein breakdown or if it is an unfavourable adaptation with resultant biochemical hypothyroidism 195. Severity of airway obstruction, hypoxaemia, systemic inflammation and glucocorticoids may predispose to develop subclinical hypothyroidism, overt hypothyroidism and nonthyroidal illness syndrome.
Severity of airway obstruction
Compared with moderate COPD, patients with severe COPD (FEV1 <50%) demonstrate reductions in total T3 and in the T3/T4 ratio (a marker of peripheral conversion of the prohormone (T4) to the active cellular form of the thyroid hormone (T3)) 195. In addition, low FEV1 is associated with low basal and stimulated levels of thyroid stimulating hormone (TSH) 199.
Chronic hypoxaemia
It is controversial whether hypoxaemia impairs the TSH response to exogenous thyroid releasing hormone (TRH) in COPD 141, 200. In patients with stable severe COPD, chronic hypoxaemia is associated with a decrease in the peripheral conversion of T4 to the biologically active hormone T3 196.
Systemic inflammation
Patients with COPD can present with increased systemic levels of inflammatory cytokines such as IL-6, IL-1 and TNF-α 42, 195. These cytokines can inhibit the synthesis or secretion of TSH, T3 and thyroid hormone-binding proteins, and can decrease the mRNA for the hepatic enzyme iodothyronine deiodinase type 1; an enzyme that converts T4 to T3 197, 198. Notwithstanding these observations, the impact of inflammatory cytokines on thyroid function in patients with COPD is controversial 195, 196.
Glucocorticoids
In 25 patients with stable COPD, serum T4 levels were inversely related to the dose of oral prednisone 201. Glucocorticoids may decrease circulating thyroid hormone by decreasing serum TSH 202–205, by redistributing T4 and T3 in the vascular and tissue compartments 206 and by decreasing peripheral conversion of T4 to T3 207. Hypothyroidism, when present, can interfere with the metabolism of glucocorticoids by increasing the area-under-the-curve and by delaying the time-to-peak of plasma concentrations of oral prednisone 208. In contrast, inhaled glucocorticoids do not appear to affect thyroid function 196.
COPD and hypothyroidism: functional consequences
When present, hypothyroidism can decrease respiratory drive, respiratory muscle function, exercise capacity, and increase the risk for sleep disordered breathing in COPD.
Respiratory muscle function/respiratory drive
Hypothyroidism can cause inspiratory and expiratory muscle weakness in patients with 209 and without COPD 210. Weakness is proportional to the severity of hypothyroidism 210 and is reversed by replacement therapy 209, 210. Weakness might result from decreased expression of Type IIb myosin heavy chains (table 1⇑) 192, phrenic nerve neuropathy 211 or decreased neuromuscular transmission secondary to a decrease in the planar areas of nerve terminals and end-plates of type I and IIa fibres 212. Although neuromuscular transmission is decreased at rest, transmission is increased during repetitive contractions 212.
It is unclear whether the respiratory muscle weakness can cause chronic or acute-on-chronic alveolar hypoventilation 209, or even need for prolonged mechanical ventilation 213, 214. In the few series of hypothyroid patients who developed ventilatory failure 213, 214 it is impossible to differentiate the contributions of respiratory muscle weakness and decreased respiratory drive 215. Several mechanisms have been postulated as risk factors for the development of alveolar hypoventilation in hypothyroidism (table 3⇓) 213, 216.
Potential risk factors for the development of alveolar hypoventilation in hypothyroidism
Exercise capacity
Exercise tolerance and maximal oxygen uptake are decreased in patients with COPD and overt 209 or subclinical hypothyroidism 217. Impaired muscle energy metabolism, resulting from a defect in glycogen breakdown or mitochondrial function 217, hypothyroid myopathy with the characteristic type II fibre atrophy and large type I fibres 218, and impaired gas exchange 209, systolic function 219 and ventricular relaxation 219, probably contributes to the reduced exercise capacity in these patients.
Sleep-disordered breathing
High prevalence of sleep-disordered breathing (mainly obstructive in nature) has been reported in hypothyroid patients with 209 and without COPD 220–222. In these patients, sleep-disordered breathing is largely reversible with replacement therapy 222. Potential mechanisms of sleep-disordered breathing include obesity, mucoprotein deposition in the upper airway, upper airway myopathy and decreased ventilatory drive 220–224.
COPD and hypothyroidism: treatment
Hypothyroidism in patients with COPD should be treated in the same manner as in patients without COPD 225. Patients who also have obstructive sleep apnoea and coronary artery disease require particular attention. In this situation, thyroid replacement should be implemented with care because it causes an increase in myocardial oxygen consumption 226. During an apnoeic episode, hypoxaemia is further increased in patients with COPD. In the presence of coronary artery disease, these changes can trigger ischaemic events 224.
COPD and hyperthyroidism: prevalence/risk factors
The prevalence of subclinical 219 and overt hyperthyroidism among patients with COPD is unknown. That hyperthyroidism is more prevalent among former and current smokers 227 raises the possibility that hyperthyroidism may be more prevalent in COPD than in the general population 228.
COPD and hyperthyroidism: functional consequences
Hyperthyroidism may impair respiratory muscle function, respiratory mechanics and exercise capacity in COPD.
Respiratory muscle function
Hyperthyroidism causes inspiratory and expiratory muscle weakness in COPD 210, 229–231. The weakness is proportional to the severity of hyperthyroidism 210 and it can be reversed by 3–9 months of antithyroid therapy 210, 229, 231–233.
Enhanced proteolysis through the activation of a proteasome-dependent pathway (fig. 5⇓) 234 and oxidative modification of myofibrillar proteins contribute to diaphragmatic atrophy 235 and weakness in hyperthyroidism 235, 236. Proteolysis triggered by excess T3 levels 234, 236 can be enhanced by administration of glucocorticoids 16. β-Adrenergic blockade can rapidly reverse much of (limb) muscle weakness in hyperthyroidism 237, suggesting that reversible changes in calcium handling and cyclic adenosine monophosphate-mediated contractile function contributes to hyperthyroid-associated myopathy 191. Lastly, hyperthyroid-associated neuropathy may contribute to the weakness 191, 238.
Ubiquitin–proteasome degradation of contractile proteins. The first step in degradation of actin and myosin is activation of ubiquitin (Ub) by a first enzyme, E1; a process requiring ATP. Activated Ub interacts with a second enzyme, E2, a carrier protein. Ub and E2 join a third enzyme, E3. E3 transfers activated Ub to actin and myosin. The cycle is repeated until a chain of Ub is bound to the contractile proteins. The chain of Ub binds to one end of a proteasome complex in a process requiring ATP. The Ub chain is subsequently removed (allowing reuse of Ub), and actin and myosin are unfolded and pushed into the core of the proteasome. Multiple enzymes within the core degrade actin and myosin into small peptides. The peptides are extruded from the proteasome and degraded to amino acids by peptidases in the cytoplasm. The ubiquitin–proteasome pathway degrades many proteins other than actin and myosin. Reproduced and modified from 15 with permission from the publisher.
Respiratory mechanics
Some 233, 239 but not all investigators 229, 240 have reported decreased lung compliance in hyperthyroidism. The mechanisms are unclear 233, 239. Decreased lung compliance 233, 239 and muscle strength 210, 229–231 probably contribute to the mild reduction in vital capacity seen in some patients 210, 229, 230, 233, 239–242. The reduction in vital capacity is reversed by antithyroid therapy 210, 229, 230, 233, 241. It is not known if hyperthyroid patients with COPD experience worsening in airway obstruction, as reported in some patients with severe asthma 243–245.
Respiratory failure
Hyperthyroid patients are at increased risk for respiratory failure 246 because of decreased respiratory muscle performance 210, decreased lung compliance 233, 239, increased ventilatory requirements 241, increased peripheral and central chemoreceptor sensitivity 231, increased respiratory effective impedance 231, and impaired cardiovascular performance 247.
Exercise capacity
Exercise capacity is decreased in subclinical 248 and overt hyperthyroidism 229, 242, and can be improved by treatment 241, 242, 248. Potential mechanisms include decreased limb muscle strength 230, 232, reduced muscle mass 232, decreased fatigue-resistant myofibres 249, reduced skeletal muscle l-carnitine (a carrier of fatty acids in the mitochondria for subsequent oxidation 250), decreased efficiency of oxygen utilisation 241, reduction in cardiac functional reserve 247, 251, excessive increase in respiratory drive 242, decreased lung volumes 230, 240 and excessive lactate production 247.
COPD and hyperthyroidism: treatment
Hyperthyroidism in patients with COPD should be treated in the same manner as in patients without COPD 225.
COPD AND RENIN–ANGIOTENSIN–ALDOSTERONE SYSTEM
The renin–angiotensin–aldosterone system is central to the control of sodium reabsorption by the kidney 252. Decreases in glomerular filtration rate and sodium delivery to the distal nephron cause release of renin. Renin cleaves angiotensinogen to form angiotensin I 252. Angiotensin I is then cleaved into angiotensin II by the angiotensin-converting enzyme. Angiotensin II can increase sodium retention through three mechanisms. First, it enhances sodium reabsorption in the proximal tubule. Secondly, it decreases the filtered load of sodium. Thirdly, it stimulates the adrenal cortex to secrete aldosterone. Aldosterone, in turn, increases sodium reabsorption through multiple mechanisms. While activation of the renin–angiotensin–aldosterone system causes sodium retention, sodium excretion is controlled by several natriuretic factors 252. These factors include a number of natriuretic peptides such as the atrial natriuretic peptide, and the B-type and C-type natriuretic peptides 253. Natriuretic peptides are primarily synthesised in the heart and brain. Atrial natriuretic peptide and B-type natriuretic peptide increase in response to overload of the extracellular volume. They cause natriuresis, vasodilation and suppress activity of the renin–angiotensin–aldosterone system. The C-type natriuretic peptide is primarily vasodilatory.
In the 1970s and 1980s, the group of Farber and Manfredi was among the first to demonstrate that clinically stable hypercapnic patients with COPD often exhibited impaired excretion of sodium and water 254, which was aggravated when oedema was present 255. They reported upregulation of the renin–angiotensin–aldosterone system and vasopressin (which increases water reabsorption in the distal nephron) in COPD 255, 256. Increased renin–angiotensin–aldosterone may contribute to sodium retention and increased vasopressin may contribute to hyponatraemia and water retention 255–257. (Sodium and fluid homeostasis in COPD is reviewed in detail by de Leeuw and Dees 252.)
COPD and imbalance of the renin–angiotensin–aldosterone system: risk factors/prevalence
Most studies of the renin–angiotensin–aldosterone system in COPD have focused on patients who already have evidence of fluid retention. Therefore, there are limited data on the frequency with which this system is activated in the general COPD population (table 4⇓). Development of sodium and water retention in COPD implies poor prognosis 258.
Abnormalities of the renin–angiotensin–aldosterone system in chronic obstructive pulmonary disease (COPD)
COPD and imbalance of the renin–angiotensin–aldosterone system: functional consequences
Patients with COPD can develop fluid retention when stable 255, 257 or during exacerbations 41, 259. Fluid retention can cause peripheral oedema, ascites and pleural effusions 260. Right heart pressures can be normal or increased 259. Cardiac output is often preserved or even increased 41, 259, 261.
The mechanisms responsible for development of fluid retention are complex 252. The traditional view, that volume overload occurs as a result of right-ventricular failure, itself caused by hypoxia-induced pulmonary vasoconstriction, probably holds true only for certain patients. Growing evidence suggests that renal vasoconstriction is central in the development of fluid retention in patients with COPD 41, 252.
Anand et al. 41 reported increased pulmonary artery resistance and decreased peripheral vascular resistance in nine hypoxaemic, hypercapnic patients experiencing an exacerbation of COPD and new-onset fluid retention. Cardiac output was normal and mean arterial pressure was reduced. Renal plasma flow and glomerular filtration rate were decreased. Plasma norepinephrine, renin activity, circulating vasopressin and atrial natriuretic peptide concentrations were all increased 41. The investigators 41 argue that hypercapnia plays a central role by decreasing systemic vascular resistance through a direct effect on systemic arterioles 262, 263. The increased vascular capacitance will decrease the effective circulating volume and renal blood flow (severe hypoxaemia can also contribute to reduced renal flow 264–267). A second mechanisms by which hypercapnia may decrease effective circulating volume is through a reduction in precapillary tone 252. The reduced tone will move the point of filtration equilibrium distally in the capillaries and, consequently, increase extravasation and the loss of plasma volume 252. Reduction of effective circulating volume will stimulate the sympathetic nervous system, the renin–angiotensin–aldosterone system and vasopressin 260. To restore intravascular volume and tissue perfusion, the kidney will respond with vasoconstriction and sodium retention (fig. 6⇓) 260. Hypercapnia may also increase sodium retention by accelerating the renal reabsorption of sodium bicarbonate 255. The retention of salt and water, with the attendant expansion of the extracellular volume, may be sufficient to increase intracardiac pressures and, with it, it may increase circulatory levels of atrial natriuretic peptide 41. The atrial natriuretic peptide, with its vasodilator properties, could further reduce systemic vascular resistance. Two additional factors may contribute to fluid retention: GH, which can be increased in patients with COPD 18, 41, can itself activate the renin–angiotensin–aldosterone system 255. Second, the carbon dioxide might directly or indirectly turn on the sodium–hydrogen antiporter in the luminal membrane of proximal tubules 252. The sodium–hydrogen antiporter, also known as sodium–hydrogen exchanger, is primarily responsible for maintaining the balance of sodium and it is involved in the buffering of respiratory acidosis at the expense of sodium gain (fig. 6⇓) 252. Anand et al. 41 argue that persistence of hypercapnia will produce vasodilatation and retention of sodium and water. Administration of diuretics may aggravate the vicious cycle by promoting sodium loss and, thus, further activation of the renin–angiotensin system 260.
Hypothetical pathophysiology of sodium and water retention in chronic obstructive pulmonary disease (see text for details). Not shown are pathways through which atrial natriuretic peptide can modulate fluid homeostasis including dilation of the afferent glomerular arteriole and constriction of the efferent glomerular arteriole with consequent increased glomerular filtration (and thus greater excretion of sodium and water) and inhibition of renin and aldosterone secretion. SNS: sympathetic nervous system; RAAS: renin–angiotensin–aldosterone system; AVP: arginine vasopressin; ANP: atrial natriuretic peptide. Reproduced and modified from 252 with permission from the publisher.
The model proposed by Anand et al. 41 has limitations. First, the investigators fail to reconcile the observation that, after the resolution of the congestive state and normalisation of peripheral vascular resistances, levels of renin, angiotensin and aldosterone remained the same as those recorded during the exacerbation 41. Secondly, the vasoconstrictor effect of hypercapnia-associated activation of the sympathetic nervous system predominates over the direct vasodilatation effect of hypercapnia (indirect vasoconstriction is absent when the vasomotor centre cannot respond to hypercapnia-induced sympathetic activation or when the vasomotor centre is disconnected from peripheral parts of the sympathetic nervous system) 268. Thirdly, although most patients with COPD and fluid retention are hypercapnic 269, 270, hypercapnia is not a necessary prerequisite for the development of fluid retention in these patients 257, 259. Lastly, the investigators 41 did not control for sodium intake, and all medications, including diuretics, were continued. Therefore, it is difficult to decipher whether fluid retention was a primary event or secondary to relative cardiac impairment; in the study of Anand et al. 41, patients did not compensate for the decreased blood pressure with supraphysiologic cardiac output.
Additional observations raise questions about the contributions of hypercapnia, aldosterone and vasopressin to fluid retention. In a study of saline loading of patients with COPD 257, only severity of parasympathetic neuropathy and renal blood flow independently influenced sodium excretion. In patients with stable COPD and fluid retention, normalisation of aldosterone by administration of angiotensin-converting enzyme inhibitors did not improve sodium excretion 271. In seven mechanically ventilated patients with COPD, changes in plasma hormones during acute hypercapnia were associated with the haemodynamic changes induced by respiratory failure and not to acute hypercapnia per se 272.
A fall in renal blood flow and development of a subclinical parasympathetic neuropathy 273, 274, reported in >35% of patients with COPD 273, might be more important than aldosterone and vasopressin in causing fluid retention in nonhypoxaemic patients with COPD 257. Interruption of the parasympathetic afferents can increase vasopressin 275 and, consequently, increase tubular reabsorption of solute-free water. Interruption of vagal pathways can increase intrarenal sympathetic activity with abolition of the reflex fall in sympathetic tone associated with an increase in effective blood volume 257. A failure to correctly interpret this increase could partly explain the reported impaired sodium excretion 257.
The presence of sympathetic neuropathy 274, 276, 277 may explain why hypercapnia causes a (direct) decrease in peripheral vascular resistance rather than an (indirect) increase through hypercapnia-mediated sympathetic stimulation. A correlation between severity of autonomic neuropathy and hypoxaemia reported by some 273, 274, but not all 276 investigators raises the possibility that autonomic neuropathy is caused by intraneural hypoxaemia 273.
COPD and imbalance of the renin–angiotensin–aldosterone system: therapy
Few investigators have assessed treatments aimed at reducing fluid retention in COPD 271, 278, 279. Some suggest postponing diuretics as long as possible 252, 260. Diuretics can aggravate retention of sodium and water 252, 260 through several pathways including hypoventilation-induced hypocloraemic metabolic alkalosis 280, 281. Others suggest the use of angiotensin-converting enzyme inhibitors to increase sodium excretion 256. These agents, however, have inconsistent effects and may increase 282 or have no effect on sodium excretion 271. Vasopressin levels vary inversely with arterial oxygen tension 252 suggesting that oxygen supplementation may help.
CONCLUSION
Despite plausibility and increased research, little information is available whether patients with COPD are at increased risk for specific endocrinopathies. The role of the renin–angiotensin–aldosterone system in contributing to oedema in COPD has been studied for more than 30 years, yet our understanding of its contribution is rudimentary. Even less is known about the indications and contraindications for hormone-replacement therapy. No study to assess whether any specific replacement therapy has some effect on COPD progression has been conducted. Cross-fertilisation is source of much creativity in science: Pasteur was a skilled chemist and Darwin a published geologist before they embarked on biology; and Crick was a physicist and Watson a zoologist when they started to research DNA. The paucity of knowledge about the interactions between COPD and the endocrine system indicates that this would be a fertile territory for cross-fertilising collaboration.
Support statement
The present study was supported by grants from the Veterans Administration Research Service.
Statement of interest
None declared.
Acknowledgments
The authors would like to thank N. Emanuele and D. J. Leehey for their helpful comments, and J. LaMarre for her help with the biographical search.
- Received July 7, 2008.
- Accepted May 1, 2009.
- © ERS Journals Ltd
References
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
- ↵
