10.1.1 Growth hormone and ageing

Ageing is characterized by undesirable changes in body composition and a decline in many physiological functions, leading to reduced physical fitness and increased susceptibility to illness. With the projected growth of the elderly population worldwide, the ageing process is likely to give rise to increasing demands on health and welfare service budgets. The WHO projects that between the years 2000 and 2050, the world’s population of persons aged 60 and over will more than triple, from 600 million to 2 billion (1). The proportion of the EU population aged 65 years and over is predicted to rise from 17.1% in 2008 to 30.0% in 2060, and the proportion aged 80 and over to rise from 4.4% to 12.1% over the same period (2).

Ageing is a complex and poorly understood process. In recent years, there has been considerable interest in the role of the growth hormone/insulin-like growth factor 1 (GH/IGF-1) axis. Prior to 1985, supplies of GH were limited as it was obtainable only from human pituitary tissue, largely restricting its use to the treatment of childhood short stature. The development of recombinant GH has made available theoretically infinite supplies of GH, and allowed exploration of the role of GH in adult pathophysiology.

While GH is best recognized for its stimulation of longitudinal bone growth in childhood, recent evidence has demonstrated that GH continues to play a central role in adulthood in the regulation of fat and protein metabolism, body composition, and many physiological functions. The steady decline in GH secretion through adulthood, termed the ‘somatopause,’ raises the possibility of involvement of the GH/IGF-1 axis in the structural and functional changes that accompany advancing age.

This chapter explores the role of the somatopause and reviews the evidence for GH as a strategy for modifying age-related deterioration.

The progressive changes in body composition and physiological function that occur with advancing age are shown in Table 10.1.1.1. Pathological studies have demonstrated an age-related reduction in the size of the kidneys, liver, and spleen. Loss of lean body mass (comprising body cell mass, extracellular water, and bone mineral mass) and accumulation of adipose tissue are characteristic consequences of ageing. Cohn and colleagues derived skeletal muscle mass, body cell mass, fat mass, and bone mineral mass from measurements of total body nitrogen, potassium, water and calcium in 135 normal male and female subjects aged 20–80 years (3). Over this age range they reported a mean 45% reduction in skeletal muscle mass, a 23% reduction in body cell mass, a 10% fall in bone mineral mass, and a 12% increase in fat mass in male subjects. Similar changes were found in female subjects, with greater reductions in bone mineral mass reflecting accelerated postmenopausal bone loss.

In addition to an increase in fat mass with age, there is a parallel change in the distribution of body fat, resulting in central and visceral adiposity. This pattern of body fat distribution is associated with lipid abnormalities, insulin resistance, and cardiovascular disease. Plasma total and low-density lipoprotein (LDL) cholesterol concentrations increase progressively from an age of around 20 years until the sixth decade. Triglyceride concentrations also increase with age, reaching peak values in men between 40 and 50 years, and continuing to rise throughout life in women. This age-related change in lipid profile predisposes to atherogenesis.

The changes in body composition that occur with advancing age are undesirable, and are accompanied by deterioration in many physiological functions. Cardiovascular mortality increases progressively with age in men and postmenopausal women. There is a 25% reduction in aerobic work performance from age 20 to 50 years (4). Muscle strength reaches a peak between the second and third decade and declines from around the fifth decade at a rate of 12–15% per decade (5). Fracture risk increases with age in both men and women and is attributed to age-related reduction in bone density and postural instability (6).

There is a steady deterioration in renal function with age, including a decrease in glomerular filtration rate and renal blood flow. Age-related changes also occur in renal tubular function, which limit renal concentrating ability and impair solute conservation. This decline in renal function reduces the capacity of the elderly to withstand stress including hypotension, fluid deprivation, and electrolyte or acid–base changes, and predisposes to drug toxicity.

Impaired hepatic function in the elderly results in decreased clearance of many commonly prescribed medications; this is attributable to reductions in liver volume and hepatic blood flow rather than decreased hepatic enzyme activity.

There are clear similarities between the clinical features of ageing and those of the now well-described syndrome of adult growth hormone deficiency (Table 10.1.1.1). The clinical features of this syndrome are reviewed in Chapter 2.3.7. Adult GH deficiency (GHD) is characterized by changes in body composition including increased fat mass, and reduced lean body soft tissue and bone mass (7). As in the elderly, the adiposity of GHD has a central, visceral distribution. Lean body mass is reduced in GHD adults by approximately 7–8% compared with age and gender-matched normal subjects, representing similar reductions in extracellular water (ECW) and body cell mass (BCM), the metabolically active component of lean body mass (7).

The structural changes of GHD are accompanied by a decline in strength and exercise capacity. Reduced muscle mass in GHD subjects is associated with reduced isometric and isokinetic muscle strength (7). It remains uncertain whether reduced strength is entirely accounted for by the reduction in muscle mass, or whether there is also intrinsic muscle weakness associated with GHD. Exercise performance is impaired in GHD adults, with maximum oxygen consumption (VO_2max, aerobic capacity or the maximum ability to take in and use oxygen) consistently shown to be reduced by estimates ranging from 17% to 27% compared to values predicted for age, gender, and height (8).

GHD is also associated with clinically relevant effects. Fracture frequency is increased in adults with GHD on standard replacement therapy for other pituitary hormone deficiencies, compared to a healthy control population. Elevated concentrations of total and LDL cholesterol, reduced high-density lipoprotein (HDL) cholesterol, and raised triglyceride concentrations, occur in GHD adults compared with healthy control subjects, and probably contribute to premature atherosclerosis. Cardiovascular mortality is increased twofold in patients with hypopituitarism, possibly attributable to GH deficiency. Echocardiographic studies have also demonstrated a reduction in left ventricular mass and impairment of systolic function in these patients. Glomerular filtration rate and renal plasma flow are reduced in GHD adults.

Secretion of GH is pulsatile, and therefore it is not surprising that isolated basal plasma GH concentrations are not age-dependent. Studies employing frequent sampling over a 24-hour period to produce integrated GH concentrations (IGHC) have clearly demonstrated that GH secretion is age-related. IGHC have been shown to increase at the onset of puberty, peak at mid to late puberty and gradually decline thereafter with advancing age (Fig. 10.1.1.1) (9, 10). The reduction in GH concentration with age is related to a decrease in the area under the curve as well as diminution in the amplitude of the pulses. Both GH production and clearance decline with age, each decade of advancing age resulting in reduction of the GH production rate by 14% and GH half-life by 6% (11). Although there is substantial evidence for hyposomatotropism in the elderly, the degree of GH deficiency in the elderly is less than that of patients with organic GHD.

The influence of age on stimulated GH production is less certain than for spontaneous secretion (12). The GH response to insulin-induced hypoglycaemia has been reported to be unchanged or decreased with age. The GH response to arginine, a GH secretagogue, and to GH-releasing hormone (GHRH) does not change significantly with age (12). In contrast exercise-induced GH release is reduced with age and it has been demonstrated that even in early middle age (mean age 42 years), the GH response to exhaustive exercise is greatly attenuated compared to younger (mean age 21 years) subjects (13).

IGF-1 levels follow a similar ontogeny to GH, increasing two- to threefold at puberty in both sexes, falling to adult levels by the third decade and progressively declining with advancing age (14) (Fig. 10.1.1.2). Interestingly, two recent cross-sectional studies of elite athletes demonstrated an age-related decline in IGF-1 levels that was at least as marked as previous reports of the age-related decline in sedentary subjects (15, 16).

Physical activity, sleep patterns, adiposity, and gonadal steroid status all change with age and have all been shown to regulate GH secretion (12). Age, body composition, and physical fitness are independent predictors of IGHC (11, 17). GH secretion in response to arginine and clonidine in healthy adults are determined by body composition and physical fitness rather than by age (18) while the GH response to exercise has been reported to be determined by age and physical fitness (VO_2max) but not by body fat (19). These findings suggest that maintenance of physical fitness throughout life might attenuate the decline in GH secretion rates, although, notably, training programmes that improve physical fitness do not appear to increase the GH response to exercise (13).

The age-related decline in GH secretion could be mediated at the adenohypophyseal, hypothalamic or a higher level. Studies in animals have indicated that enhanced hypothalamic production of somatostatin, an inhibitor of GH secretion, is the principal mechanism for age-related hyposomatotropism, with reduced growth hormone releasing-hormone (GHRH) production involved to a lesser degree (20). Human studies using indirect approaches, however, have demonstrated an important role for reduced hypothalamic GHRH secretion in ageing. Firstly, the GH response to withdrawal of somatostatin infusion, which provides an estimate of GHRH release, is reduced in elderly compared to young women, with a similar trend occurring in men (21). Secondly, the GH pulse amplitude, a function of GHRH secretion, is reduced in elderly compared to young subjects (22). The suppressive effects of exogenous IGF-1 on GH are reduced rather than enhanced in the elderly, implying that increased sensitivity to negative feedback by endogenous IGF-1 is not a mechanism through which this effect occurs (23).

GH replacement exerts beneficial effects on body composition including increased lean body and skeletal muscle mass, increased extracellular water, and reduced total body fat, effects which can be demonstrated within months of commencement of treatment. An increase in bone mass is observed after treatment for 12 to 18 months (7). These effects have also been demonstrated in studies limited to older adults with organic GH deficiency.

These favourable effects on body composition translate to improvements in functional performance. Most studies in which the effects of GH replacement on exercise capacity were investigated have reported improvement, and a recent meta-analysis of placebo-controlled trials supports improvement in both maximal power output and VO_2max (24). The evidence to support a beneficial effect of GH replacement on muscle strength is less strong. However, data from a cohort of GH-treated patients from Sweden followed up continuously over 10 years indicate that GH replacement results in a transient increase (of up to 5 years) in absolute values for most measures of isometric and isokinetic muscle strength, and a sustained increase in absolute values for isometric knee flexor strength (25). By the end of 10 years of follow-up, all measures of muscle strength were comparable to an age-related reference population.

GH replacement increases glomerular filtration rate and renal plasma flow to levels comparable to age-matched controls. Improvements in cardiac function occur in parallel with normalization of ventricular size and systolic function (7).

The overlap of the clinical features of adult GHD and ageing, evidence of hyposomatotropism in the elderly, and the unequivocal beneficial effects of GH replacement in GHD adults have raised the question of whether GH treatment can reverse undesirable age-related changes. A number of studies of GH treatment in the elderly have been reported, alone and in combination with sex steroids (Table 10.1.1.2).

The first major study to explore a possible beneficial effect of GH in ageing was reported by Rudman et al. (26), who demonstrated increased lean body mass, skin thickness, and bone mineral density, and reduced total body fat, following administration of GH for 6 months to older men. This study provoked major interest among scientific researchers and the general public that GH might be an effective anti-ageing therapy. Indeed there remains widespread off-license use of GH for this indication. However, despite confirming these potentially beneficial changes in body composition, subsequent studies demonstrated little or no improvement in strength or functional ability following administration of GH alone or in combination with exercise training, to elderly subjects (Table 10.1.1.2). The findings of these studies have been considered in a recent meta-analysis totaling 220 participants (37). Overall, GH increased lean body mass and reduced body fat mass, more markedly in men compared to women. GH also reduced LDL cholesterol, but did not influence VO_2max or fasting glucose or insulin levels.

While most trials of GH treatment in the elderly have involved healthy subjects, others have investigated the effects of GH treatment on specific age-related health problems (Table 10.1.1.3). The impact of GH replacement in 10 malnourished elderly patients was investigated in a randomized placebo-controlled trial of 3 weeks duration (38). The treatment group showed significant weight gain, improved anthropometric measures for muscle mass and urinary nitrogen retention suggesting that GH may be an effective therapeutic agent in this situation.

Initial studies investigating treatment of osteoporosis with GH were not very promising, but more recent evidence supports a beneficial effect of more prolonged GH administration on bone density. In the study by Rudman et al., there was a small but significant improvement in bone density after 6 months of GH treatment, but this was not sustained after 12 months (26). Aloia et al. reported a reduction in bone mineral content following 12 months of GH treatment in an uncontrolled trial of 8 osteoporotic patients (43). In a larger double-blind placebo controlled study of elderly women with low bone mass, 12 weeks of GH treatment increased biochemical markers of bone formation and resorption, but did not influence bone density (39). Holloway et al. reported small but significant increases in bone density following 2 years of cyclical treat replacement (42). Bone mineral content continued to increase in the open-label follow-up phase of the trial (3 years GH treatment in total) in women who remained on GH, and surprisingly a further increase of 14% was seen in the year following discontinuation of GH (Fig. 10.1.1.3). These results suggest that GH might represent a useful agent for treatment of osteoporosis although long-term treatment is likely to be necessary.

Trials of GH treatment in the elderly have revealed an unexpectedly high incidence of side effects. These were systematically evaluated in the meta-analysis described above (Table 10.1.1.4) (37). Higher rates of soft tissue oedema, carpal tunnel syndrome, arthralgias, and gynaecomastia were all observed in subjects receiving GH. Notably, rates of oedema were greater in women. Higher rates of new diagnoses of diabetes or pre-diabetic conditions also occurred. These side effects likely occur because doses of GH used in trials in the elderly to date (Table 10.1.1.2) are supraphysiological. Studies of GH production rates indicate that the average 70 kg adult secretes 3–10 μg/kg per day of GH. Cohn et al. found an increased frequency of side effects to GH treatment occurred with higher intratreatment IGF-1 levels, while attaining lower IGF-1 levels did not result in attenuation of the beneficial changes in body composition with GH treatment (27). The elderly are particularly susceptible to carpal tunnel syndrome in comparison to younger GH deficient adults and children. The exclusion of subjects with early indications of carpal tunnel syndrome, along with use of lower dosages of GH, should substantially reduce the adverse event rate.

Three recent prospective studies have demonstrated that high IGF-1 levels within the normal range are predictive of cancer. In a meta-analysis of hormonal predictors of prostate cancer, it was found that men with either serum testosterone or IGF-1 levels in the upper quartile of the population had an approximately twofold higher risk of developing prostate cancer (44). In other prospective trials among premenopausal women in the Nurse’s Health Study, there was a 4.5-fold relative risk of breast cancer in the highest quartile of serum IGF-1 as compared with the lowest quartile (45). Similar results were also found for colorectal cancer in men in the Physician’s Health Study (46). One interpretation of these observations is that elevation of IGF-1, which occurs with GH treatment, may increase the risk of developing these cancers. If this is true, patients with acromegaly who have sustained elevated IGF-1 levels should have a higher incidence of these malignancies. There is no strong evidence, however, of an increase in the incidence of cancer in acromegalic patients, with studies to date yielding conflicting results. A major limitation of studies of cancer incidence in acromegaly is the potential bias introduced from greater vigilance and regular medical attendance likely with this chronic disease. Nevertheless, this theoretical risk is an important issue to consider in all future trials concerned with GH supplementation in the elderly.

In addition to reduced rates of GH secretion, it is likely that in men the age-related reduction in total and bioavailable testosterone also contributes to the age-related reduction in lean body mass and increase in body fat mass. In GHD subjects, GH and testosterone exert additive effects to increase protein anabolism, fat oxidation and extracellular water (47), and it has been hypothesized that in elderly subjects these two hormones in combination might be more efficacious than either hormone alone. Three recent studies have addressed this question. In a 26-week double blind, placebo controlled trial, increases in muscle strength and VO_2max that correlated with increases in lean body mass were demonstrated in men treated with combined GH and testosterone (34). Notably, deterioration in glucose tolerance occurred in a significant number of subjects. A crossover study compared the effect of administration of testosterone, GH, and combined testosterone and GH in doses chosen to approximate physiologic production rates for one month each to elderly men. Improvements were seen in some indices of physical function, including walking and climbing stairs, following administration of either hormone alone or in combination, and improvement in balance was seen following treatment with GH alone. The effects of administration of GH and testosterone alone and in combination for 6 months to healthy elderly men were studied in a more recent double-blind, placebo-controlled trial (36). The dose of GH was titrated to achieve plasma IGF-1 levels in the upper half of the normal range, and a transdermal preparation of testosterone was administered daily, resulting in plasma testosterone levels within the normal range. Lean body mass increased with GH alone, while there was an increase in muscle mass and a reduction in total body fat following combined treatment. The VO_2max also increased significantly in patients who received combined treatment, compared to those who received placebo and those who received either treatment alone. Overall, the combined effect of the two hormones was additive rather than synergistic.

In contrast to testosterone, there are theoretical reasons why oestrogen administration might attenuate some of the effects of GH. Oestrogen has a major effect on GH action, which is dependent on its route of delivery. When compared to the transdermal route, oral oestrogen reduces IGF-1 and suppresses lipid oxidation, causing a loss of lean body mass and a gain in fat mass after six months treatment in postmenopausal women (48). This phenomenon occurs as a result of a first-pass hepatic effect of oestrogen on the endocrine and metabolic function of the liver. The biological effects are opposite to those of growth hormone and induce detrimental changes in body composition, which are already occurring in ageing. Addition of transdermal oestrogen to GH in female subjects in the study reported by Blackman et al. did not significantly alter the effects observed with GH treatment alone (34).

Regular GHRH administration to the elderly results in augmentation of pulsatile release of GH and allows negative feedback by IGF-1 on the pituitary gland. This has an important theoretical advantage over GH treatment, as by allowing normal regulatory mechanisms to operate it should not be associated with the high incidence of side effects reported with GH. GH-releasing peptides, which have been available for more than a decade, and which are now known to act through the ghrelin receptor, (49) also stimulate pulsatile GH release, probably at both hypothalamic and pituitary levels. The drugs MK-677 and capromorelin are orally active ghrelin mimetics, which potentially circumvent the practical difficulties of daily injections in the elderly.

Initial short-term studies of GHRH and ghrelin mimetic administration in the elderly (Table 10.1.1.5) mostly reported an increase in GH and IGF-1 concentrations with few adverse events. In a 5 month trial of a GHRH analogue in 19 healthy subjects over 55 years, Khorram et al. reported increased skin thickness following the treatment phase in both sexes and increased lean body mass in men only (50). There were no changes in fat mass or bone mineral density. Men also reported an increase in general wellbeing and libido, changes not described by women. There were no serious adverse effects. In an extension of this study, the authors reported enhancement of the immune system within 4 weeks of treatment in both elderly men and women. Two recent double blind, placebo controlled trials of ghrelin mimetics have provided further information regarding the potential therapeutic use of these agents. Nass et al. administered MK-677 for 1 year to healthy older adults (51). Mean 24-h GH and IGF-1 levels increased by 1.8- and 1.5-fold, respectively, and by the end of the study IGF-1 levels were within the normal young adult range. Following 1 year of treatment, lean body increased in treated subjects compared to placebo by 1.6 kg, which appeared to reflect an increase in both body cell mass and extracellular water, with no change in body fat. No differences were observed in muscle strength, physical function or quality of life. White et al. demonstrated similar increases in IGF-1 and lean body mass following administration for up to one year of capromorelin, but additionally demonstrated improvements in certain functional measures including tandem walk and stair climb (52). Side effects associated with fluid retention did not emerge during these studies, but fatigue and insomnia were reported following capromorelin treatment, and there was evidence of increased insulin resistance following administration of both agents.

Taken together, these studies have demonstrated that regular administration of GHRH and ghrelin mimetics can restore normal GH secretion in the elderly without evidence of the adverse effects associated with fluid retention that have limited the use of GH, and with some evidence of functional improvement. The possible effect of these agents to increase insulin resistance needs further investigation.

The structural and physiological changes that accompany ageing mimic those of adult GHD. These changes along with the age-related decline in GH secretion have become known as the somatopause. The unequivocal benefits of GH replacement in adults with organic GHD have led to interest in the role of GH treatment in reversing or arresting the ageing process.

There is substantial evidence that GH treatment in the elderly restores body soft tissue composition towards more youthful proportions with an increase in lean body mass and a reduction in fat mass. There is now evidence that GH treatment improves osteoporosis in older women. There is little evidence, however, of any beneficial effect of GH treatment alone on physiological functions in the elderly, although the studies published to date have a number of limitations, which may have prevented the demonstration of positive effects of GH (Table 10.1.1.6).

Two potential approaches to optimize the benefit of GH replacement while reducing side effects are the administration of GH in combination with testosterone, and the use of ghrelin mimetic agents. Combining GH with testosterone enables similar changes in body composition to be achieved using lower doses of GH and hence minimizing side effects. There is preliminary evidence from small studies of relatively short duration that this approach might result in clinically relevant effects, although larger studies of longer duration are necessary. The recently developed ghrelin mimetics are also promising alternatives for the enhancement of GH production, as they allow feedback regulation of GH secretion and are theoretically less likely to cause side effects from overtreatment. Two recent studies have demonstrated an effect of ghrelin mimetic treatment to increase lean body mass while one of these studies has also demonstrated functional benefits. Further studies are needed to determine a whether these agents result in a clinically significant increase in insulin resistance.

It is unlikely that GH or any other hormone can ever be a ‘cure’ for the complex, multifactorial process of ageing. However, GH plays a central role in the regulation of metabolism, body composition, and physiological function, and has the potential to modify the structural and functional deterioration that accompanies ageing. Recognition of factors that affect the biological action of endogenous GH is an important avenue of future research. The benefits of GH or GH secretagogue treatment may be more evident if subjects with specific age-related problems such as frailty or those convalescing from surgery or acute catabolic illness were studied. Further long-term studies addressing these issues are needed before the efficacy and safety of GH treatment in the elderly can be defined.