Evaluation and Treatment of
Adult-Onset Growth Hormone Deficiency

Overview

What makes me grow?  When do I stop growing?  Why is my sister taller than I am?  Why am I shrinking? All through life, children, parents, and adults compare the amount and speed of growth. What makes a pre-pubescent male look like the Incredible Hulk?  Who better to answer these questions than The Doctor?  You are the fountain of knowledge about growth. So what do you know? 

This module provides an overview and rationale for the reasoned use of growth hormone modulation therapy for adult onset growth hormone deficiency in the clinical setting when a comprehensive evaluation reveals an adult deficiency.  The goals for this module are to enable a clinician’s ability (1) to evaluate a patient in the proper clinical context so informed treatment decisions can be made, regarding the use of growth hormone in adult deficiency; (2) to be able to measure appropriate objective markers of disease risk and outcome, including IGF-1; (3) to act appropriately upon the obtained data; and (4) to use objective criteria for monitoring disease risk markers and outcomes, both laboratory and clinical.  This process is not a set-it-and-forget-it phenomenon—it requires continued diligence in a broad clinical setting.   

This module is really not about growth hormone alone, so much as about evaluation of growth hormone levels and their consideration as one clinical factor among many in a comprehensive adult patient-oriented diagnosis, prevention, and life-style alteration program. 

Growth hormone is a product of anterior pituitary gland.  It is a 191-amino acid peptide hormone, which is pleiotropic in its effect.  Growth hormone is technically a “pro hormone” since in liver and other peripheral tissue, it causes the production of the Insulin-like Growth Factor family of molecules.  The predominant metabolically significant final hormone is Insulin-like Growth Factor 1, or IGF-1.  In addition to IGF-1, other Insulin-like Growth Factor family molecules are produced.  slide01

The feedback loop for growth hormone secretion is multi-factorial. Growth hormone itself provides strong negative feedback on both the hypothalamus and the anterior pituitary to inhibit further release of growth hormone.  As growth hormone is metabolized to IGF-1, IGF-1 provides only minor negative feedback on the hypothalamic pituitary axis; subsequent pulses of growth hormone can be generated readily in the presence of IGF-1.  Additionally, growth hormone (strongly) and IGF-1 (weakly) exert positive feedback on the hypothalamus to generate the release of somatostatin, which also has inhibitory effect on the anterior pituitary with regard to growth hormone release.  Homeostatsis of hormone production is maintained via both positive and negative feedback loops. slide02  

Growth hormone is released in pulsatile fashion.  Each secretion pulse can last between 20 and 60 minutes; the average person will produce four to six of these pulses per day.  While the majority of growth hormone secretion pulses are random, there are relatively predictable growth hormone release pulses occurring after vigorous exercise and early in a subject’s nightly sleep pattern.  The graph demonstrates the pulsatile nature of growth hormone release, pointing out one of the reasons that measuring growth hormone levels directly has low utility when evaluating a patient.  Between pulses, growth hormone levels are essentially nil. If blood were drawn during this period, an incorrect assumption regarding low growth hormone levels would be made. Conversely, blood drawn during a growth hormone peak would give the impression of adequate or even high-growth hormone levels.  IGF-1, however, is present at stable levels over the course of an average patient day. Given the relationship between growth hormone and IGF-1 noted in the previous slide—and with stable IGF-1 levels present at any given time—it has become the growth hormone marker of choice, presenting the clinician with a broad window regarding the examination timing for blood draws.   

Adult Growth Hormone Deficiency (AGHD)

The following slides will address growth hormone physiology and the impact of replacement therapy. slide03 slide04  After its youthful peak, growth hormone production declines with age—approximately 1.3% per year after age 29.  For a healthy individual at age 40, IGF-1 levels are still maintained at approximately 275 ng/ml. 

The most commonly seen signs and symptoms associated with decline in growth hormone level in adulthood are the same as those found in patients with a lifelong deficiency state.  As growth hormone production drops, especially in patients whose growth hormone drops abnormally quickly with age, we see declines in measures of mood and well-being; loss of lean mass and corresponding strength; gains in fat mass, especially central adiposity, which is its own cardiac risk factor; diminished skin thickness and elasticity; derangements in glucose and insulin metabolism, which (as mentioned with central adiposity) are independent markers of disease risk. There also is a decline in quality of sleep, an increase in undesirable lipid values and a concomitant decline in desirable lipid levels; and a decrease in cardiac contractility.   

The most interesting thing to note from an actuarial point of view is that declines in growth hormone are associated with increases in multiple independent risk markers of disease and premature mortality.   

Subjects who show an abnormal rate of decline in growth hormone production with age show much more rapid than usual declines in cardiac endurance, left ventricular thickness and ejection fraction, declines in B and T-cell immune markers, declines in antibody response to vaccine as well as alterations in other quality of life markers.  When thinking of IGF-1 levels and growth hormone replacement to correct proven adult deficiencies and restore endocrine balance in our subjects, the most important long-term goal is reduction in disease risk and improvement and maintenance of independent function.  As such, the rate of decline in IGF-1 levels/growth hormone secretion is an important marker for long-term health outcomes.   

As discussed earlier, a significant body of literature exists, addressing hormones as markers of disease risk and as outcome markers of therapy. Our approach to patient evaluation and treatment, apart from the current mainstream, is the importance of incorporating these markers in the same context and with the same statistical requirements, as done for other standardized disease risk markers.  

As such, the measurement of IGF-1 alone is, in fact, not an adequate single measurement in determining who may or may not be a candidate for growth hormone therapy.  As part of the comprehensive health management for adult patients, we also stress the importance of these evaluations before making treatment decisions: pretreatment cancer screening, dementia screening, evaluation measures that may be indicative of early cognitive impairment, bone metabolism measurement, mood evaluation and other factors. IGF-1 is a useful tool, but it must be properly considered and applied in a complete clinical context. 

Signs and Symptoms of AGHD

AGHD is characterized by changes in body composition, carbohydrate and lipid metabolism, bone mineral density, CV risk profile and quality of life (QOL). ¹²⁰ Also, GHD may contribute to CV morbidity and mortality.  Other signs of GHD include the decreased mass and strength of muscles, increase in body fat (especially visceral fat),¹⁴⁷  reduced mass and impaired contractility of the heart, CV  mortality, abnormally low bone density, unfavorable changes in lipid profile and some psychiatric (emotional labiality, feeling of social isolation, sometimes anxiety and depression)  and sexual problems.⁹⁶ GH in most patients reduced total cholesterol and LDL cholesterol level,  as well as  fibrinogen and homocysteine, independent predictors of  CV events.¹³¹   

Pandian and Carroll as well as other authors have assembled clinical signs and symptoms, associated most frequently with suboptimal levels of growth hormone.  These clinical signs and symptoms are consistent between studies. ⁹ slide05

Growth Hormone and Slow Sleep Wave

With regard to specific effects of growth hormone, there is a consistent relationship that has been demonstrated between growth hormone and slow wave sleep quality.

Slow wave sleep (SWS) is a term used to describe the combination of stage 3 and stage 4 sleep.  The longest periods of SWS take place in the early parts of the night and are seen in greatest amount in children and young adults. SWS diminishes with age; many elderly patients do not achieve SWS at all during their nightly sleep cycles. In younger adults who are deprived of SWS, a lack of “refreshing” sleep is reported; after deprivation, there is a sharp rebound increase in SWS for the next one to several sleep cycles. 

It has been shown that slow wave sleep quality diminishes in conjunction with growth hormone levels. In two studies, Van Cauter demonstrated the correlation between retaining slow wave sleep quality and well-maintained growth hormone values. ^{22 23} In growth hormone replacement studies, SWS length and frequency have improved.

Growth Hormone and the Brain

Nyberg, in 2000, summarized growth hormone’s impact on other brain processes, as well.  It was reported that growth hormone replacement has a positive impact on measures of memory, cognitive capabilities (especially processing speed), alertness and self-reports of motivation and work capacity.  Additionally, it was reported that growth hormone receptors had been identified in the hypothalamus, choroid plexus and the hippocampus, demonstrating a potential direct role for growth hormone in neurophysiology, neurotransmission and maintaining neural tissue. ²⁴

Aberg reported growth hormone increases expression of Connexin-43, which is a molecule forming tight gap junctions between cells that facilitates more efficient cell-to-cell communication. In experimental models, Aberg found cells cultured in the presence of growth hormone expressed increased levels of Connexin-43 in both the cortex and hypothalamus. The retention of tight gap junctions is associated with better neuron-to-neuron communication and better maintenance of cell populations. ⁹⁷ 

In a physiologically related study, Niikura found stimulation of IGF-1 receptors in the hippocampus by IGF-1 decreased the amount of apoptosis caused by beta amyloid protein, whose production is associated with Alzheimer’s disease. ²⁶

Additionally, both Riddle and Caro (both in 1999) reported IGF-1 stimulates dendrite formation in cortical neurons. Greater dendrite formation in the cortex is associated with lower dementia risk and less anatomically demonstrated cortical atrophy on imaging studies. ^{27 28}

Also in 1999, Thornton demonstrated that age-related deficits in type 2 Dopamine receptors were reversed in association with replacement. ²⁹

A 1999 study by Aleman, demonstrated direct correlation between IGF-1 levels and perceptual motor and mental processing speed—and that low IGF-1 values correlated with poorer reports of a patient’s well-being and social function. ³¹

Therefore, growth hormone has been associated with better maintenance of cell/cell communication channels, decreased sensitivity to the negative impact of beta amyloid, better development of axonal dendrites and more advantageous neuronal receptor expression. Declines in growth hormone levels have been associated with declines in cognitive function and well-being. This set of findings would predict the demonstration of benefit in clinical trials, which it has: 

The Van Dam study demonstrated adult growth hormone deficiency was associated with more rapidly advancing declines in cognitive performance than in normal subjects and that these deficits were improved after instituting growth hormone therapy. ³⁰

Aleman, in 2000, found that circulating IGF-1 levels were directly correlated with measures of cognitive function quality and processing speed in “normal” community dwelling subjects. Growth Hormone modulation has now been linked to positive responses to modulation within normal range lab values in otherwise clinically “normal” subjects. ³¹

Growth Hormone and Bone Density

With regard to adult growth hormone and bone density, a clinically significant study published by Longobardi, in 1999, looked at the effect of two years of growth hormone replacement therapy on bone metabolism and bone mineral density. ³²

It was found that growth hormone deficient patients have decrease bone mineral density compared to normal population and had lower indices of bone formation measurements than normal controls. But adult growth hormone deficient patients showed no decrease in the measurements of bone resorption. It seems osteoblasts are at a functional disadvantage in association with diminished growth hormone levels, but osteoclasts are not. It was found that growth hormone deficient patients absorb bone normally, but do not form or replace bone as well as normal controls.   

When growth hormone therapy was instituted, measures of both bone formation and bone resorption were increased almost equally over the first three to six months of therapy.  This is of clinical interest with regard to patient care because in the bone formation/resorption equilibrium, if both index values were increased, then during that period, no change in bone density would be expected—or, given the equilibrium had favored resorption at the start of therapy, a transient decline in bone mineral density may actually be seen.   

As the study progressed, this response was their initial finding. However, between 6 and 24 months of follow-up, the equilibrium had shifted to favor bone formation; bone density increased substantially. During the last six months of the study, bone resorption had returned to baseline, while bone formation remained improved, yielding a significant and substantial net positive change in bone density. 

Regarding growth hormone replacement therapy, an important part of physician and patient education is that bone mineral density should not be expected to change significantly for the better in the first three to six months of treatment. By month 24, a substantial benefit is expected. Growth hormone therapy in adult growth hormone deficient patients should be viewed as a long-term intervention, which favors a healthier physiologic profile, not act as a “quick fix.” 

A graphic recap: slide06

At baseline, bone resorption (red line) takes place at a higher rate than bone formation (smell, weak, green line). During the first six months of therapy, both resorption and formation are increased. During months 6-24, resorption falls to baseline levels, while formation continues to be increased, creating a prolonged equilibrium shift toward bone formation. 

Additional references regarding the impact of growth hormone replacement and bone are found at the end or the references.

Growth Hormone and the Heart 

From a general perspective, it would appear growth hormone does not have a great positive impact on cardiac function, but within published studies, those looking at treating ischemic heart failure have shown no benefit—while studies looking at the impact of growth hormone on dilative cardiomyopathies have had positive results. ³⁶

These findings make sense clinically in that just as positive outcomes shouldn’t be expected from “beating a dead horse,” minimal or no benefit should be expected from “feeding a dead horse.”  Ischemic cardiomyopathy is associated with a decline in cardiac function as a result of lost tissue—once tissue is dead, no amount of intervention can bring it back. In patients with dilative cardiomyopathy who still have viable myocardium, growth hormone therapy does appear to offer clinical benefit, even in these relatively short trials. ³⁷

Cho et al ⁸⁸ investigated the effect of GH replacement and it persistence on cardiac function assessed by two-dimensional or tissue Doppler strain echocardiography in adult-onset GH deficiency with normal global systolic function.⁸⁸ Ten patients (six women) were evaluated.  All patients received GH (somatotropin ) at a starting dose of 1.0 IU/day injected subcutaneously (6 times/week),with titration according to serum insulin-like growth factor-I (IGF-I) level. ⁸⁸

Among conventional echocardiographic parameters,⁸⁸ LV ejection fractions were  nonsignificantly increased afer 6 months of GH replacement.  Average peak strain and strain  rate were significantly decreased in patients with GH deficiency.  Peak strain and strain rate  were increased significantly after 6 months of GH replacement, with values improved up to the levels of healthy control, but returned to baseline after 12 months off therapy. ⁸⁸

Cho et al ⁸⁸ demonstrated that average peak strain and peak strain rate were significantly  decreased in patients with GH deficiency despite normal LV systolic function measured to conventional echocardiography and decreased cardiac contractility was improved after 6 months of GH replacement. 

In AGHD, the goals are to relieve symptoms associated with the  deficiency rather to increase height.  Vasan et al ¹⁷⁰ prospectively studied the association  between serum IGF-I levels and the incidence of congestive heart failure.  These investigatiors showed that there was a decrease in the risk for heart failure with increments in IGF-I.⁸⁸

Although echocardiography is the most widely used modality for the evaluation of systolic function, the LV ejection fraction may  be not sensitive enough to detect subclinical myocardial disease.  Tissue Doppler-derived strain and strain rate have been validated (Neilan, Abrasham) and can be used for noninvasive indexes for contractility ¹⁰⁹ which appears to be extremely effective for the identification for subclinical LV dysfunction in diabetic or obese patients.^{139 174 175 177} 

In CHO et als study, ⁸⁸ the positive effect of short-term GH replacement on cardiac Contractility disappeared after the discontinuation of GH replacement.  This finding agrees with a previous study in that the positive effect of replacement is not persistent.  LV mass was not changed after GH replacement and significantly decreased after withdrawal.  However, interpretation should be made with caution because of small sample size. 

GH treatment in AGHD is associated with a correction of body composition and the abnormalities in lipid profile ^{74 169} but effects on cardiac parameters are less consistent.¹³⁸  Ozdogru et al’s¹⁵¹ study included a small number of participants. All participants received GHRT for 12 months.  GH was self-administered at night subcutaneously and drug compliance was assessed by vial count.  Treatment with GH (Genotropin) was started at a dose 0.45 IU (0.15 mg)/l in month 1, was increased to 0.9 IU (0.30 mg)/d in month 2 and was maintained at 2 IU (0.66 mg)/d until the end of month 12.¹⁵¹     

Each participant underwent echocardiographic evaluation before and after 12 months of treatment.¹⁵¹ The results of this study are represented by the ability of pulse DTI to identify early myocardial markers of systolic dysfunction when ejection fraction (EF) is still normal in GHD.  .  Adults with were GHD were demonstrated for the first time to have a reduced IV longitudinal systolic function which was corrected with GHDT.  Measurement of myocardial velocities by Doppler tissue imaging (DTI) is a promising modality for quantitative assessment of longitudinal systolic and diastolic ventricular performance. ¹⁵¹ The study ¹⁵¹ showed the usefulness of pulse DTI in the assessment of patients with severe GHD.  Pulsed DTI highlights the possibility of identifying early markers of LV myocardial systolic and diastolic dysfunction.  Adult-onset GH deficiency is associated with normal left ventricular sizes and preserved ejection fraction, whereas longitudinal systolic function obtained by Doppler tissue imaging is deteriorated, which improved with 12 months of GHRT.¹⁵¹ 

Growth Hormone and Immune Function 

Growth hormone has been evaluated in terms of impact on immune function as well.  The effect of growth hormone has been measured on cytokine profiles, revealing growth hormone replacement decreased levels of disadvantageous levels of tumor necrosis factor interleukin-6 factor levels in patients with dilative cardiomyopathy, but was not found to have influence on other inflammatory markers, such as interleukin-1, 2 or 8, and did not have an impact on NK (natural killer) T cells.   

It has also been demonstrated that rising levels of these cytokines can have a negative effect on growth hormone levels. 

At present, there has not been a standardization of results with regard to the specific impact of growth hormone on cytokine levels and vice versa. This will be of great interest because there are currently several postulated Immune Risk Phenotypes, which may have an association with health and mortality risk: 

1) Pawelec, G, Immunology Clinics of North America, Feb 2003:

Poor T-Cell proliferative response, Increase CD28-/CD8+ numbers, Increase in CD57 cells, CD4/CD8 ratio less than 1, Low B Cell number, Increased TNF-alpha, and IL-6, CMV seropositivity. ⁶²

2) Wikby, JM, Mech Aging Dev, 1998:

Poor T-Cell proliferative response, Decreased CD4/CD8 ratio, Decline in CD19+ cells (B Cells). ⁶³

3) Franceschi, C, Immunol Today, 1995, and Wayne, SJ, J Geront, 1990:

T-cell anergy (loss of in vivo delayed type hypersensitivity - DTH), associated with RR of 1.89 for mortality over 5-year study, subjects aged 60 and above. ^{64 65}

Tapson, in 1988, demonstrated the presence of IGF-1 receptors on human T-lymphocytes; in 1992, Koojiman demonstrated beneficial effect in terms of T-cell antigen responsiveness in cell cultures. ^{38 39}

A study by Geffner, in 1992, demonstrated similar findings. Also in 1992, Timsit demonstrated growth hormone stimulated proliferation and function of thymic epithelial cells, both directly and by stimulating epithelial cell signaling pathways. ^{40 41}

Clark (1997) and Burgess (1999) also demonstrated that growth hormone therapy had a measurable beneficial effect on function of thymic tissue with replacement, reverting the age-related decline in thymulin, an intra-thymic communication protein that is part of maintaining generation, proper maturation and T-cell function. ^{42 43}

Franco et al¹⁰⁴ investigated the effect of 12-month GH treatment on inflammatory serum markers and vascular adhesion molecule (VAMS)  in postmenopausal women with abdominal obesity.  Forty women aged 51-63 years received GH (.67 mg/d) in a randomized, double-blind, placebo-controlled 12-month trial.  Measurements of inflammatory markers C-reactive protein (CRP) and markers of endothelial dysfunction were performed at baseline and after 6 and 12 months of treatment.  The subjects were treated either with GH (n=19) or placebo (n=2), administered SC before bedtime.  The initial dose of GH was 0.13 mg/d dose (0.51 vs 0.65 mg/d in the GH vs. placebo group).  Dose was reduced by half in the event of side effects.¹⁰⁴ The results showed that GH treatment in postmenopausal women with abdominal obesity reduced CRP and IL-6 levels.¹⁰⁴ These were associated with a reduction in abdominal obesity and hepatic fat content.  This suggests an overall reduction in the risk of CVD. Circulating markers of endothelial dysfunction (vascular adhesion molecule-1) were unaffected by treatment.¹⁰⁴  This study did not support a direct GH treatment on vascular adhesion molecule (VAM).  Recent studies by Franco et al¹⁰⁴ suggested that important risk factors for atherosclerosis and its progress such as body fat (BF) distribution, serum lipid pattern, and subclinical inflammation, are however significantly improved by GH treatment.

Growth Hormone and Obesity

Obesity is costing millions of dollars in medical care a year.  Morbid obesity is defined as body mass index (BMI) more than 40 kg/m2 and is a great concern due to both health risk concerns and resistance of therapy.¹¹² Bariatric  surgery is recommended for adults with BMI more than 35-40 kg/m with a co-morbidity or more than 40 kg/m2 (class III obesity). IGF-I, mediator of the GH effects, is used as a measure of GH bioactivity because it almost accurately reflects GH secretory status.  Obesity per se is almost always associated with a normal IGF-I level.  Therefore, to make the diagnosis of idiopathic GHD, an IGD-I level below the age-corrected lower limit of normal should also be present.¹⁴⁷ Age, gender, body composition, nutritional-driven compoments, and glucose homeostasis, have affected IGF-I metabolism.¹⁰⁵ 

     Twenty-four women (body mass index: 44.4 +-7.6 kgm2, aged 36.8 +- 11.7 yr) undergoing laparoscopic-adjustable silicone gastric banding (LASGB) and with GH deficiency after LASGB was included in the study. Group A (n=12) included a standardized diet regimen (55%CHO, 25% fat, 20% protein, 30 g fiber) exercise program plus recombinant human GH (0.5 +- 0.13 mg every day), and group B (n=12) included a standarized diet regimen and exercise program.  The exercise program consisted of 60-90 min/d moderate-intensity active (e.g. brisk walking). ¹⁵⁹ Follow-up was in 6 months.  The results showed the excess body weight loss did not differ between groups A and B after 3 and 6 months.  At 3 months, LBM loss was lower (p=0.001) and fat mass (FM) loss was higher (p=0.000) in Group A than Group B.  At 3 (p=0.0003 and 0.0005, respectively) and 6 months (p=0.0001 and 0.0002 respectively), the percent changes of FM and lean body mass were significantly higher in Group A than Group B. 

  In both groups fasting and post glucose are significantly reduced.  The homeostasis model assessment of insulin and insulin sensitivity indexes and total to high-density lipoprotein cholesterol ratio improved only in Group A.¹⁵⁹ 

     Other simple measures, besides IGF-I, to reflect efficacy of GH replacements are changes in extracellular water (ECW) ¹²⁰ measured by BIA and normalization of other compartments of the body, such as body fat and body cell mass.  The effect of GH to increase ECW through its antinatriuretic action occurs within 3 to 5 days of starting GH replacement.  Increased ECW is one of the measurements that changes most consistently during GH treatment, it may be a more useful endpoint for monitoring GH replacement than other aspects of  body composition. The dose that normalized tissue hydration was similar to the dose that normalizes serum IGF in most of the study subjects.¹²⁰ 

     A 32-week study was conducted by Hartman et al ¹¹¹ to determine whether improvements in aerobic exercise capacity with GH treatment in adults with GHD are related to changes in physical activity or affected by the GH dosing regimen.  There were 29 patients in the study. ¹¹¹ The study compared an individualized GH dosing regimen (ID) with a fixed body weight-based dosing regimen. ^{111 146} In summary, GH replacement therapy in GH-deficient adults improved VO2 max( maximal oxygen consumption) similarly with both dosing regimens, without any influence of physical activity.  There was no effect on submaximal exercise performance.  These results areclinically relevant because they demonstrate dosing regimens now recommended by consensus clinical guidelines will improve aerobic exercise capacity. ^{111 146} 

GH enhances whole body protein synthesis, increasing plasma glucose level and lipolysis via the direct effect on adipocytes as well as lipid oxidation by increasing substrate availability.^{117 118}  A study of 16 males and 7 females showed that GH treatment was able to decrease 11beta-HSD1 mRNA and increase 11beta-HSD2 and, accordingly may be able to reduce the amount of locally produced cortisol in adipose tissue.¹⁵⁵ This results in improved insulin sensitivity and glucose tolerance as well as an improved adipokine profile and reduced fat mass, especially visceral fat mass, because the density of glucocorticoid receptor is higher in visceral adipose tissue.  Abdominal obesity is associated with reduced spontaneous GH secretion and elevated biomarkers of systemic inflammation.   

C-reactive protein (CRP) is an important predictor of CV disease (CVD) events because it modulates the expression of cellular adhesion molecules and reduces nitric oxide synthesis.¹⁷¹ Increased Nox, nitrate and nitrite, bioavailability in response to GH is IGFI dependent.¹⁶⁸ Circulating vascular adhesion molecule (VAM) levels reflect endothelial dysfunction and atherosclerosis.⁸³  Circulating VAMS  correlate positively with body  mass index (BMI), whereas there has been no clear relationship between these markers and adipose tissue (AT)  distribution.¹⁰²         

Hypopituitary patients with GH deficiency accumulate abdominal fat and exhibit the same metabolic risk factors as those present in patients with the metabolic syndrome.  GH-deficient patients not receiving GH replacement have  premature atherosclerosis, increased CVD mortality, and increase levels of CRP and Il-6 that decreased in response to GH replacement.

One-year GH treatment in abdominal obese women reduced visceral adipose tissue (VAT) and low-density lipoprotein cholesterol, whereas body weight was unaffected.¹⁰²  Insulin Insensitivity improved 12 months within the GH-treated group, but there was no between-group differences.  Obesity is related to several metabolic disturbances such as insulin resistance, impaired insulin secretion, non-insulin-dependent diabetes mellitus (NIDD), hypertension, dyslipidemia and CV disease. This complex is often called Metabolic Syndrome (MetS). The complications of obesity have been attributed to increases in visceral adipose tissue (VAT) with an associated rise in portal vein free fatty acid levels.¹⁵⁴

GH is a lipolytic hormone that decreases VAT, increases energy expenditure, and increases lean mass in individuals with GH deficiency.  GH therapy in AGHD also improves total and low-density lipoprotein cholesterol.  Obesity is associated with a decrease in IGF-I and deficits in GH secretion.  Free IGF-I in obesity is increased in obesity, whereas total IGF-I is consistently lower in obese men and women.  Free IGF-I might serve to inhibit GH secretion through a negative feedback loop. Visceral obese men who receive GH treatment had a striking reduction in VAT along with increase in lean mass.  GH decreases low-density lipoprotein cholesterol.

The objective of Pasarica et al¹⁵⁴ was to determine the effect of growth hormone on body composition and visceral adiposity in middle-age men with visceral obesity.  Fasting insulin, glucose and the quantitative insulin sensitivity check index for insulin increased during GH therapy.  The effects of GH on fatness and visceral adiposity disappeared shortly after GH withdrawal, but weight remained increased over baseline when compared with the placebo group. Waist circumference did not change during therapy.  The final 1-year observation, waist circumference was greater in the GH-treated subjects.  Fasting glucose demonstrated a decrease in insulin.  In viscerally obese subjects, supraphysiological GH administration is not an effective treatment.¹⁵³ 

Optimum Growth Hormone Levels

In the decision making process with regard to when to institute growth hormone therapy in adult growth hormone deficiencies, there has been an evolution over time as to what IGF-1 levels should be a potential “threshold” of therapy.   

In the 1980s, data collected from North American and British Commonwealth deficiency literature revealed an average IGF-1 of an adult growth hormone deficient patient to be approximately 100 ng/ml.  Subsequently, the value was adopted as a “cutoff value” when trying to determine if a patient’s IGF-1 levels were considered deficient or not.  Recent research, however, shows the hGH levels in the lower range of normal can lead to morbidity and actually predict poor health. 

In 1997, a two-year study performed by Verhelst assessed the impact of growth hormone therapy on patients who had an initial IGF-1 level of 120 ng/ml or lower. His treatment group demonstrated a 50% decline in risk for all-cause hospitalization.  This included hospital admission for medical diagnoses, as well as psychiatric etiologies or trauma as reason for admission. Growth hormone replacement was associated with a global improvement in patients’ outcomes. The treatment group also demonstrated a 45% decline in days missed from work. ⁴⁴

As the cost of treating patients with growth hormone goes down over time, there will eventually be a medical-economic point at which evaluation and treatment of suboptimal growth hormone levels will become a potentially cost-effective intervention when considering the economic burden of the costs of hospitalization and absenteeism. 

Also in 1997, Johannson reported beneficial responses to therapy in a patient group using IGF-1 values of 160 ng/ml or below as inclusion criteria.  The treatment group showed significant reductions in abdominal fat mass and waist-to-hip ration, in addition to improved measures of glucose metabolism and reduced diastolic blood pressure. ⁴⁵

In 1998, the working group on adult growth hormone deficiency and replacement published a position paper by Carroll, et al in the Journal of Clinical Endocrinology and Metabolism. The landmark study placed growth hormone and IGF-1 measurement in the context of a clinical syndrome with proper diagnosis and treatment with consideration for accompanying signs and symptoms. ⁹

They recommended using a constellation of symptoms in determining when to institute growth hormone therapy and recommended treatment to values at, or above, the mean for a normal age-related population.   

In 2003, Mukherjee reviewed statistical confidence levels with regard to growth hormone therapy and found an IGF-1 value of 194 ng/ml was the level associated with the 95^th percentile confidence value to know a patient had been adequately supplemented and that a deficiency state had been “avoided.” ¹²

The accompanying graph, slide07 from Mukherjee, demonstrates a well-preserved IGF-1 value curve between age 40 and 60, regarding confidence interval for avoiding deficiency symptoms. Over the 20-year period between ages 41 and 60, this confidence interval varied by only 0.14 standard deviations. To age 60, IGF-1 levels maintained in the 190-194 ng/ml and above range were associated with the lowest levels of health risk. As the slide demonstrates, this value range was found to be applicable out to age 70, as well. 

In one of the largest and longest trials to date, a study by Svensson (published in 2004) reported beneficial outcomes based on supplementing IGF-1 levels averaging 210 ng/ml for women and 289 ng/ml for men. ⁴⁶

Given the historic evolution of IGF-1 values and related clinical responses in previous studies, the treatment goal for IGF-1 levels should be set at approximately 250 ng/ml. That treatment level approximates the average IGF-1 levels associated with both sexes at age 40 and is not associated with supraphysiologic values at any age. The upper limit of normal for IGF-1 values at age 75 is still 290 ng/ml. This treatment target also approximates the levels associated with best outcomes in the literature.

Diagnosis of Growth Hormone Deficiency in Adults

Diagnostic Tests

Five provocative tests to stimulate GH secretion to determine a peak response have been used for diagnostic purposes in hypopituitarism and GHD.  The five tests are insulin tolerance test (ITT), arginine-GH releasing hormone (Arg-GHRH), GHRH-GH releasing peptide (CHRP)-6, and GHRH-Pyridostigmine, clonidine and L-Dopa.^{145 152}  Clonidine and L-Dopa are considered inadequate.^{82 173}   ITT (95.1 micro/liter) has been found to have a high sensitivity and specifity in all age groups, ⁷⁴ but may carry increased risk in patients with seizure disorders or CV disease and requires constant monitoring even in healthy adults   Insulin-induced hypoglycemia stimulates GH secretion by somatostatin suppression and alpha-adrenergic activity.  The average dose of insulin is 0.05-0.1 IU/lg iv, with the exception of patients with a high risk of hypoglycemia, in whom the dose of insulin should be reduced by 50%.  The positive result of the test is when GH level increases at least 5-fold after insulin. The peak, however, is reduced In obesity. The peak is lower in women than men.  ITT is the recommended test of choice in adults who have suspected GHD. The test is potentially hazardous especially in patients who have heart disease, seizure disorders, and mental retardation.¹²⁰

The Arg-GHRH stimulation test (4.1 microg/liter) benefits from the potentiating of both substances and reproducible stimulation of the GH secretion is achieved.¹¹⁸ The peak  GH response by Arg-GHRH seems to be independent of age and sex but is reduced by obesity ¹⁴⁷  BMI-related references ranges have been  established.⁹⁵ ARG-GHRH is safe and has less discomfort than the ITT; therefore a good alternative for the ITT except for patients who have primary hypothalamic lesion.  A promising test is the GHRH-GHRP-6 stimulating test which is not influenced by age or body composition. Testing should be conducted after discontinuation of GH treatment to avoid possible suppression of endogenous responses.  The interval between the reevaluation and the discontinuation of GH should not be less than 1 month.  (This interval is based on personal practice) ^{147 96}

The diagnosis of severe GHD in adults has been considered to be a peak GH response to ITT of less than 3- 5 microg/L depending on BMI. The specific cutoff values for Arg-GHRH are between 16.5 and 20.3 mico/L.  For GHRH-GHRP-6 stimulation test is less than 10 microg/L and a patient with more limited degree of hypopituitarism and two or fewer pituitary hormone deficiences. Two independent stimulation tests are recommended to diagnose severe GHD in adults. In patients with low GH levels, other hormones should be evaluated with special emphasis on TSH, cortisol and gonadotropines ^{97 152} and replaced as needed.

The alternative GH deficiency diagnostic method is IGF-I and IGFBP-3 (Insulin like growth factor binding protein).^{145 152}These markers decrease during adult life.  IGF-I is regarded as the best marker of GH therapy monitoring.  IGF-I  levels in adults may be affected by many various disorders, like liver diseases, malnutrition, diabetes, renal failure, and thyroid gland dysfunction, and, therefore, its measurements are less useful in diagnosis than in therapy.¹⁴⁵

The utility of IGF-1 measurement versus measurement of induced growth hormone secretion has been the subject of much discussion and debate.  In the past, the insulin tolerance test was regarded as a requirement to adequately assess a subject’s growth hormone secretion abilities, but this has been called into question in recent years.  The insulin tolerance test requires the medically supervised intravenous administration of insulin to achieve a blood glucose level of 40mg/Dl or less, with average blood glucose levels during this test at approximately 25mg/DL.  The assumption with regard to this test is that under extreme hypoglycemic conditions, growth hormone (a counter-regulatory hormone) release will occur. Growth hormone levels are measured at baseline 30, 60 and, often, 90 minutes after insulin infusion.  If, in the presence of this degree of hypoglycemia, growth hormone secretion occurs to levels of above 3 to 5ng/ml, a patient is judged “normal.”  As noted in the slide, there is no connection between a patient’s state during an insulin infusion challenge versus a normal glucose state in the individual’s day-to-day life.  The average subject will never have glucose levels as low as those seen in an insulin tolerance test—and this intervention may not adequately reflect a patient’s true physiologic state.  The test has remained in the medical lexicon despite its lack of correlation to normal physiology.   

When the insulin tolerance test was developed, growth hormone was still harvested from cadavers and was made as an extract of human tissue.  As such, it was quite limited in quantity and carried risk of disease transmission, especially prion diseases.  Given its difficulty in acquisition and danger with regard to exposure, there was a strong rationale for rationing. The use of the insulin tolerance test was meant to guarantee that only patients with the least ability to secrete growth hormone would be treated.  In the era before recombinant technology, this was a valid concern.  Since 1986, growth hormone has been commercially produced in the laboratory with recombinant gene technology—a product no longer a human extract, but a pure synthetically produced bioidentical hormone.  The change in production methods negated the previous rationale for rationing and disease transmission concerns. Despite this change in availability, there has been a persistence of the old diagnostic paradigm.   

Despite readily available recombinant forms of growth hormone, a slow accommodation in some circles exists, regarding diagnostic investigation—especially in the IGF-1 versus the insulin tolerance test (ITT) debate of which is the most useful in assessing growth hormone status.  

Data will be shown, demonstrating that IGF-1 is already accepted as a diagnostic and therapeutic marker in its own right. We will examine the independent and statistically significant use of IGF-1 as a marker of disease risk and will review some growth hormone intervention studies, which have been undertaken in the absence of “deficiency states” and their demonstrated clinical utility.   

The ITT is not a true indicator of a subject’s underlying hormonal state.  As mentioned, it is not representative of a normal physiologic condition and does not reflect growth hormone secretion patterns in the absence of hypoglycemia.  It is also far less reproducible than it is portrayed to be.   

Pandian, et al, concluded in 2004 that the ITT, like all growth hormone stimulation tests, suffered from statistical lack of reproducibility for a given individual and was deemed not an adequate test as a “gold standard” for the diagnosis of suboptimal growth hormone levels. ¹ slide08  

Vestergaard, in 1997, found a coefficient of variation in patients undergoing repeat ITT of 41% in men and 104% in women.  Of note, approximately one-third of the patients had normal growth hormone release patterns on one test, but their second evaluation actually satisfied criteria for deficiency. ²  slide09   

Hoeck, in 1995, examined normal male and female subjects with an average age of 31½.  There were greater than 72 hours between insulin tolerance tests; this study generated glucose levels of less than 25mg/DL during the insulin challenge.  One-fourth of normal female subjects met lab criteria for a deficiency state. If ITT had been used to make a treatment decision in these patients, they would have been inappropriately funneled for therapy, when in fact there was no underlying pathology.  Hoeck noted no correlation between first and second test results, meaning from one test to another, there was no test reproducibility in his subjects.  They also found that results did not correlate with the degree of hypoglycemia achieved.  His conclusion: there was poor reproducibility during repeated testing and no correlation between the results of the two tests. ³ slide10  

Biller, in 2002, demonstrated there was up to a six-fold difference in growth hormone secretion, either increased or decreased, when comparing adults undergoing repeat ITT.  Her conclusion was that with the ITT, physicians would have to choose between high sensitivity versus high specificity when using the ITT, and in this context.

Biller looked at the utility of using IGF-1 values and found that IGF-1 values were stable over time and highly reproducible, showing standard deviation scores and distribution of values, which dependably placed patients in the same diagnostic category with regard to IGF-1 value on repeat testing.  In fact, no subjects were noted to have changed their diagnostic category on subsequent tests.   

Biller also observed an IGF-1 “cut point” for minimizing misclassification for the deficiency state was 127 mcg/L, but the 95% sensitivity value was set at 200mcg/L.  The IGF-1 standard deviation score for 95% sensitivity was -0.12. ⁴ 

What this study shows is that in the clinical setting of evaluating the signs and symptoms of suboptimal growth hormone levels, a standard deviation score of only 0.12 standard deviations below the norm was required to achieve 95% sensitivity. 

Hoeck found a 97% degree of correlation between IGF-1 levels and subsequent result of growth hormone stimulation tests. ⁵  

In 2003, Boquete pointed out further problems with insulin tolerance testing to evaluate growth hormone levels, demonstrating that insulin tolerance test levels varied with patient body mass index (BMI).  He also found that 92% of patients could be properly categorized clinically using IGF-1 values. ⁶ Studies by Bonert and Biller showed similar variability with insulin tolerance testing and body mass index. ^{7 8}

The most important thing to note with these studies is that IGF-1 values are independent of body mass index.   

Pandian endorsed the use of IGF-1 measurement on a routine basis and, most importantly, pointed out the importance of any growth hormone evaluation being placed in a clinical context in association with proper history and physical findings. ¹ slide11

The Growth Hormone Research Society published a position statement that was published in the Journal of Clinical Endocrinology and Metabolism. They concluded that in association with clinical signs and symptoms and in conjunction with proper laboratory evaluation, proper clinical diagnosis can be made in the absence of insulin tolerance testing and may be associated with normal range IGF-1 values. ⁹ slide12  

The Growth Hormone Research Society’s conclusion properly places growth hormone evaluation in a clinical context with a focus on proper diagnosis and the subsequent long-term benefit of proper replacement therapy.  Even among clinicians who feel insulin tolerance testing is required in the diagnostic evaluation of suboptimal growth hormone levels, it is interesting to note within this group there is no controversy regarding the correlation between IGF-1 levels and growth hormone replacement.   

Without a hint of irony, they conclude that the “gold standard” of following growth hormone replacement therapy for adult deficiency involves the use of IGF-1 as a marker of replacement dose and as a titration point for raising or lowering dose.  The illustrated article by Drake is an example of this, pointing out the importance of using IGF-1 to properly titrate growth hormone replacement. ¹⁰

It is interesting that on the one hand, IGF-1 is supposedly inadequate as a marker of growth hormone production for the purpose of diagnosis, but is transmuted from “lead” into “gold” as a standard for judging growth hormone dosing decisions. 

Drake also points out the utility of maintaining IGF-1 levels at between the normal average for a patient’s age and less than two standard deviations above that mean.  This is regarded as typical of replacement therapy. ¹⁰

Further demonstration of the almost 100% correlation between growth hormone dosage and IGF-1 levels is provided by these illustrations taken from Boer, 1996. ¹¹slide13  

They demonstrate a dependable IGF-1 response with regard to dosing in the bar graph on the left. The horizontal axis represents daily growth hormone dose in units per day; the vertical axis represents corresponding IGF-1 levels in response. There is a direct linear relationship between IGF-1 levels and growth hormone dose.  

The illustration on the right points out that IGF-1 levels are dose responsive and well-maintained over time. The horizontal axis in these graphs represents a 12-month period, demonstrating a very stable IGF-1 level in response to a given growth hormone dose. 

Mukherjee looked at the IGF-1/Growth Hormone relationship from a different point of view. Instead of using IGF-1 to diagnose deficiency, she looked at what IGF-1 level would be sufficient to “avoid a deficiency state.”  Her findings reinforced the currently accepted practice of treating IGF-1 levels to the upper half or quartile for normal age-related range and found that her 95% sensitivity measurement for “adequately” addressing a deficiency state with regard to growth hormone was IGF-1 levels, maintained above the age-related mean. ¹²

Looked at from both sides– diagnostic and the levels associated with adequate “avoidance” of deficiency—IGF-1 is a consistently useful (and used) measure. 

Even in articles criticizing growth hormone use without insulin tolerance testing to confirm a deficiency state (as published by Perls, et al in JAMA in 2005), the authors cannot avoid referencing articles, which undermine their insistence on insulin tolerance testing and point out the clinical utility of using IGF-1 assays for the diagnosis of AGHD.¹³

Again, Perls’ referenced article notes the well-accepted connection between growth hormone replacement therapy dose and corresponding utility of monitoring resultant IGF-1 values.¹³

slide14 slide15 – These slides are representative of the current state of the IGF-1 growth hormone discussion among clinicians who deny the utility of IGF-1 values as they relate to underlying growth hormone levels and insist on ITT to satisfy diagnostic criteria for diagnosis, but then abandon this paradigm in their subsequent use of IGF-1 to monitor therapy.  

Next Page

slide01

slide02

slide03

slide04

slide05

slide06

slide07

slide08

slide09

slide10

slide11

slide12

slide13

slide14

slide15