Hyposomatotropism (Growth Hormone Deficiency) Workup

Updated: Jan 24, 2019
  • Author: Sunil Kumar Sinha, MD; Chief Editor: Robert P Hoffman, MD  more...
  • Print

Approach Considerations

The diagnosis of growth hormone (GH) deficiency (GHD), or hyposomatotropism, remains controversial. The Growth Hormone Research Society [48, 49, 50] convened an international workshop of acknowledged authorities to address this issue. [48]  The diagnosis of GHD is a multifaceted process requiring comprehensive clinical and auxologic assessment combined with biochemical testing of the GH-insulinlike growth factor (IGF) axis and radiologic evaluation. Biochemical testing of the GH-IGF axis includes radioimmunoassays (RIAs) of GH, IGF, and insulinlike growth factor binding proteins (IGFBPs)

Regarding clinical and auxologic assessment, history taking and physical examination are the most useful diagnostic tools because the diagnosis of GHD rests on clinical judgment. The foundation for the diagnosis of GHD is careful, serial documentation of the patient's height and a determination of height velocity.

In the absence of other evidence of pit-hGH secretory dysfunction, testing for GH secretion is typically unnecessary. 

Diagnostic Criteria for Hyposomatotropism 

Random testing of serum GH concentrations is of no use in establishing the diagnosis of GHD. Provocative GH testing is not the current criteria standard. Current diagnostic criteria include the following:

  • Growth-velocity Z score below -2, evidence of certain genetic mutations (eg, GH1 deletion; IGF1 deletion; mutations involving SHOX, PIT1, PROP1, Turner syndrome, and Prader-Willi syndrome) [51]

  • Predicted adult height (Bayley-Pinneau value more than 1.5 standard deviations below the calculated midparental target height)

  • Serum-free or total IGF-1 Z score below -2 (ie, more than 2 standard deviations below the mean for the patient's age, sex, and Tanner stage)

New Models for Diagnosis

Basal, oscillating, and pulsatile GH inputs and the wide range of intrasubject and intersubject variance in GH pharmacokinetics negate the assumption of a uniform relationship between GH secretion and serum GH concentration. Because of this, peak GH concentration is an oversimplified outcome of GH testing. Bright and colleagues postulated that serum GH concentrations reflect multiple components of GH input and that a composite pharmacokinetic model that accounts for pulsatile (infused), basal, and oscillatory components is required to accurately estimate each individual's pharmacokinetic parameters. [52]  Ongoing research may aid in applying these complex mathematical models to daily practice. [53, 54]


Laboratory Studies

RIA for GH

Many RIAs are available to measure GH levels, and all offer limited accuracy. Repeated measurements may vary by as much as 3-fold, even when the tests are conducted in laboratories with personnel experienced in the procedures. This variation is observed because several molecular forms of GH are identified in the serum and because polyclonal (instead of monoclonal) antibodies are used. To improve standardization, use of a 22-kDa recombinant human growth hormone (rhGH) reference preparation with an assigned potency of 3 IU/mg has been recommended. When assay data are reported, a clear statement of the method should be included. The optimal assay measures the 22-kDa hGH species by using a monoclonal antibody. [14, 15, 21, 55, 23, 56]

Serum GH concentrations remain constitutively elevated from the newborn period to as late as 6 months of age. Therefore, a serum GH level of less than 20 ng/mL in infants younger than 6 months suggests GHD. However, a random hGH level is not diagnostic in patients older than 6 months because hGH is intermittently secreted in brief nocturnal pulses (of 10-15 minutes during deep sleep) beyond early infancy. GHD cannot be diagnosed on the basis of a single random serum GH concentration at any age.


Specific RIAs distinguish IGF-1 and IGF-2. Serum IGF-1 concentrations depend on GH and vary with the patient's age, nutritional status, and sexual maturation. In children younger than 8 years, serum IGF-1 levels may be indistinguishable from levels measured in children with GHD. Concentrations of serum IGF-2 vary less than IGF-1 levels do at a given age; however, serum IGF-2 is less GH dependent than IGF-1.

Rosenfeld and colleagues evaluated the effectiveness of using IGF-1 and IGF-2 RIAs to identify children with GHD. [57] When performed alone, assays for both produced false-positive and false-negative results. However, combined assays helped to correctly identify 96% of children with GHD. Only 0.5% of healthy children had serum concentrations of both IGF-1 and IGF-2 that were below the reference ranges for their age and sex.

Total serum IGF-1 levels represent the combined quantity of unbound IGF-1 (free IGF-1) plus IGF-1 bound to IGFBP-3. Free IGF-1 is postulated to be the bioactive fraction, but it accounts for only a small fraction of the total amount.

Hasegawa and colleagues developed an immunoradiometric assay for free IGF-1 in plasma and reported the relationship of free IGF-1 to GH-secretory status. [58] Low serum levels of free IGF-1 assayed by using this method were highly correlated with complete GHD but not partial GHD. Despite their reduced diagnostic usefulness in patients with partial GHD, free IGF-1 levels may prove useful for assessing compliance with, or the effectiveness of, rhGH therapy.


To diagnose GHD, assaying the serum IGFBP-3 concentration may be superior to measuring the free IGF-1 concentration for at least 2 reasons. First, IGFBP-3 levels vary less with nutritional status than free IGF-1 values. Second, serum IGFBP-3 levels, even in young children, are typically more than 500 mg/mL; therefore, the detection of low levels is feasible.


Imaging Studies


Radiography to assess skeletal maturation, similar to an examination of growth and development, is a useful diagnostic tool to determine a patient's GH secretory status. Anteroposterior radiographs of the left hand and wrist (knee or ankle in children < 1 yr) are used to evaluate the progress of epiphyseal ossification by comparing the results to age-matched and sex-matched reference ranges.

Crude estimates of skeletal maturation can also be obtained by assessing dental eruption. Primary teeth begin to erupt at approximately 6 months of age, and exfoliation starts at 6-12 years.

Height predictions rely on the observation that the greater the delay in bone age relative to chronologic age, the longer the time before epiphyseal fusion occurs and, thus, final height is achieved. The method of height prediction is based on formulas Bayley and Pinneau developed using information from Greulich and Pyle's classic radiographic atlas. [45] Tanner and colleagues and Roche and colleagues subsequently refined these predictions by linking skeletal maturation to a rating of sexual maturity. [59, 46]

Each system is useful for estimating the range of a patient's likely adult height to within 2 inches above or below the predicted value. A major limitation of current methods for predicting height is that the standards Greulich and Pyle established are based on calculations from a few Caucasian children who lived in 2 affluent suburbs in the United States during the 1940s. Normal skeletal maturation varies with ethnicity and is likely to vary with socioeconomic status. Moreover, most industrialized countries are home to a heterogeneous population. A modern reappraisal of these radiographic standards is overdue.


A lateral skull image may provide evidence of enlargement or distortion of the sella turcica, as well as suprasellar calcification, which indicates a craniopharyngioma. As a result of the high false-negative rate of skull findings with plain radiography, MRI is the procedure of choice to exclude intracranial masses or developmental abnormalities arising from pituitary anlagen. Before rhGH therapy is started, patients with GHD should undergo MRI of the brain to exclude the possibility of an organic lesion. [32]

Contemporary MRI techniques can be overly sensitive, with MRIs depicting clinically insignificant signal intensity in the hypothalamus or pituitary gland. Most of these lesions require only clinical evaluation (eg, ophthalmologic examination, growth surveillance). A thickened pituitary stalk or asymmetric elevation of the pituitary contour warrants further evaluation.


Other Tests

Provocative Testing

Provocative GH testing is criticized for several reasons, including the following [60] :

  • None of the tests reproduces the physiologic secretory pattern of GH because they involve the use of pharmacologic stimuli to indirectly assess physiologic GH production.

  • Individual clinicians assign what are essentially arbitrary definitions for subnormal responses (ie, cutoffs for peak serum GH values) to provocation.

  • The reproducibility of provocative tests and GH RIAs is limited. Many pediatric endocrinologists apply other clinical criteria (eg, growth velocity Z score below -2) and do not perform provocative GH tests to diagnose GHD.

Despite limitations, provocative GH tests remain helpful ways to measure GH reserve. Pediatric endocrinologists use physiologic stimuli (eg, strenuous exercise, fasting, deep sleep) and pharmacologic stimuli (eg, clonidine, levodopa-propranolol, glucagon, arginine, insulin) to provoke GH secretion. In euthyroid children, all of the tests must be performed after overnight fasting.

To improve diagnostic sensitivity and specificity, at least 2 provocative tests are performed. Immediately before and during the earliest phases of puberty, GH production is often indistinguishable in unaffected children and in children with GHD. Serum GH concentrations typically rise during puberty. Many investigators suggest that children approaching puberty should be given gonadal steroids to prime the growth hormone-releasing hormone (GHRH)-GH axis before testing.

Most clinicians use a peak serum GH concentration of more than 10 ng/mL (30 IU) to exclude GHD in children. Specific provocative tests are described below.

Insulin tolerance test

Insulin-induced hypoglycemia is the most potent stimulus for GH secretion and the most dangerous tool for provocative GH testing in patients who may have GH deficiency. Insulin tolerance testing takes advantage of the hormonal counterregulatory response to hypoglycemia. In patients without GHD, plasma concentrations of glucagon, epinephrine, norepinephrine, cortisol, corticotropin, and GH are elevated in response to acute hypoglycemia.

To perform the test, patients fast for 8 hours. Then, lispro insulin 0.1 U/kg of body weight is administered rapidly as an intravenous bolus. Serial blood samples are subsequently obtained to measure GH, cortisol, and glucose concentrations at 0, 15, 30, 60, 75, 90, and 120 minutes. With each sample, the blood glucose level is simultaneously determined by using a bedside glucometer to document an appropriate reduction and to ensure safety. Performance of the test is considered adequate when the blood glucose level decreases below 50% of its baseline value.

Adverse effects expected during the procedure include symptoms secondary to hypoglycemia, such as lethargy, shaking, confusion, headache, abdominal pain, nausea, vomiting, syncope, and seizure activity. The test must be performed under the watchful eye of the physician who can begin prompt resuscitation with glucose and/or glucagon as soon as the diagnostic samples have been obtained. To date, the insulin tolerance test is the only provocative test associated with fatalities; therefore, personnel must be trained and conduct the test judiciously.

Clonidine stimulation test

Clonidine acts centrally to stimulate alpha-adrenergic receptors, which are involved in regulating GH release. Serum GH levels are obtained at baseline and at 60 minutes and 90 minutes after the oral administration of clonidine 0.1 mg/kg. Clonidine may induce hypotension during the test. Therefore, warn parents that they may experience lethargy and/or depression for 24 hours after clonidine is administered. [61]

Levodopa-propranolol HCl test

Levodopa is a dopamine receptor agonist. Dopamine is involved in the stimulation of GH secretion. In the converse, beta-adrenergic control negatively regulates GH release. [62]

Propranolol is a beta-blocker used to hinder inhibitory input affecting GH release, while levodopa simultaneously stimulates GH release by means of the dopaminergic pathway. Propranolol 0.75-1 mg/kg is orally administered before levodopa. The dosage of levodopa for levodopa-propranolol HCl testing varies with weight, so that children weighing less than 15 kg receive 125 mg, children weighing 10-30 kg receive 250 mg, and children weighing more than 30 kg receive 500 mg.

Blood samples for GH testing are drawn at 0, 60, and 90 minutes after the administration of levodopa. Adverse effects include nausea and, in rare cases, emesis. In addition, the patient's heart rate may decrease because of the use of propranolol. Closely monitor his or her vital signs, and ensure that appropriate resuscitative measures are available.

Arginine HCl test

Arginine appears to exert a direct depolarizing action on somatropic neurons, increasing GH secretion. After an overnight fast, patients are given 10% arginine HCl in 0.9% NaCl 0.5 g/kg (not to exceed 30 g) as a constant intravenous infusion over 30 minutes. Blood samples for GH testing are obtained at 0, 15, 30, 45, and 60 minutes after the infusion of arginine is begun. Arginine has historically been used as a primer before insulin is administered during insulin tolerance testing. [63, 64]

Glucagon test

Glucagon increases peripheral glucose concentrations by means of glycogenolysis and gluconeogenesis. Because glucagon is rapidly metabolized, an abrupt reduction in serum glucose concentration ensues and triggers the release of counterregulatory hormones. [65]

After fasting overnight, patients receive an intramuscular injection of glucagon 0.03 mg/kg (not to exceed 1 mg). Some clinicians advocate the concomitant use of propranolol to inhibit the catecholaminergic response to hypoglycemia. Serum GH concentrations are determined at 0, 30, 60, 90, 120, 150, and 180 minutes after glucagon administration. Nausea and, occasionally, emesis may occur.