Medical Physiology A Cellular and Molecular Approach, Updated 2nd Ed.


Eugene J. Barrett


Growth from the fertilized ovum to the adult is an exceedingly complex process involving both hyperplasia (an increase in the number of cells) and hypertrophy (an increase in the size of cells) of the cellular elements of body tissues. The timing and capacity for cell division vary among tissues. In the human central nervous system (CNS), neuronal division is essentially complete by the age of 1 year, whereas bone, muscle, and fat cells continue to divide until later in childhood. Other tissues retain the capacity for hyperplasia throughout life; these include the skin, gastrointestinal epithelial cells, and liver.

In humans, the genetic contribution to growth is evident from the observation that midparental height (i.e., the average of paternal and maternal height) is one of the better predictors of a child’s ultimate stature. For domestic animals, breeding based on desired growth characteristics has been a mainstay of animal husbandry for millennia. For humans, in whom such planned breeding does not occur, body height differences among different populations suggest a genetic contribution. Even more striking are some of the differences seen with specific genetic mutations that affect skeletal growth (e.g., achondroplasia). The basis of these population height differences and of those related to genetic syndromes is beyond the scope of this chapter. Rather, I focus on nutritional and hormonal processes in which physiological regulation appears to play an important role across individuals. The impact of environmental and nutritional factors, such as emotional or nutritional deprivation, on growth is most profound when it occurs during periods of tissue hyperplasia, most critically during the first 2 years of life.

The first two major sections of this chapter deal with factors that affect linear growth, whereas the third section deals with factors that regulate body mass. This division is somewhat artificial because changes in linear growth and body mass often occur simultaneously. The control of linear growth in humans depends on multiple hormones, including growth hormone (GH), insulin-like growth factors 1 and 2 (IGF-1 and IGF-2), insulin, thyroid hormones, glucocorticoids, androgens, and estrogens. Among these, GH and IGF-1 have been implicated as the major determinants of growth in normal postuterine life. However, deficiencies (or excesses) of each of the other hormones can seriously affect the normal growth of the musculoskeletal system as well as the growth and maturation of other tissues. The control of body mass depends on many newly discovered humoral factors made in adipose, intestine, hypothalamus, and other tissues that regulate appetite and energy expenditure.


Growth hormone, secreted by somatotrophs in the anterior pituitary, is the principal endocrine regulator of growth

Individuals with excessive GH secretion during childhood develop gigantism, and those with a deficiency of GH develop pituitary dwarfism. It is thus quite clear that GH profoundly affects somatic size. Dramatic examples of the somatotrophic action of GH can be found in the descriptions of Tom Thumb and the Alton giant. GH deficiency resulted in Tom Thumb’s achieving an adult height of approximately 0.9 m. In contrast, the Alton giant had a GH-secreting pituitary tumor present from early childhood and reached an adult height of more than 2.7 m. It is important that, in both cases, the abnormality of GH secretion was present from early life. Children with GH deficiency are of normal size at birth and only subsequently fall behind their peers in stature.

A deficiency of GH beginning in adult life does not result in any major clinical illness. However, it is now appreciated that replacement of GH (clinically available as a recombinant protein) in adults with GH deficiency leads to increased lean body mass, decreased body fat, and perhaps an increased sense of vigor or well-being. An excess of GH after puberty results in the clinical syndrome of acromegaly (from the Greek akron [top] + megas [large]). This condition is characterized by the growth of bone and many other somatic tissues, including skin, muscle, heart, liver, and the gastrointestinal tract. The lengthening of long bones is not part of the syndrome because the epiphyseal growth plates close at the end of puberty. Thus, acromegaly causes progressive thickening of bones and soft tissues of the head, hands, feet, and other parts of the body. If untreated, these somatic changes cause significant morbidity and shorten life as a result of joint deformity, hypertension, pulmonary insufficiency, and heart failure.

GH is made in the somatotrophs throughout the anterior pituitary (see Chapter 47). Like other peptide hormones, GH is synthesized as a larger prehormone (Fig. 48-1). During processing through the endoplasmic reticulum and Golgi system, several small peptides are removed. GH exists in at least three molecular forms. The predominant form is a 22-kDa polypeptide with two intramolecular sulfhydryl bonds. Alternative splicing generates a 20-kDa form of GH. Other GH forms include a 45-kDa protein, which is a dimer of the 22-kDa form, as well as larger forms that are multimers of monomeric GH. There is little information to suggest that the different principal forms of GH (i.e., the 20- and 22-kDa versions) vary in their activity, but the 20-kDa form may exert fewer of the acute metabolic actions of GH. Once synthesized, GH is stored in secretory granules in the cytosol of the somatotrophs until it is secreted.


Figure 48-1 Synthesis of GH. Somatotrophic cells in the anterior pituitary are responsible for the synthesis of GH. The cell transcribes five exons to form GH messenger RNA (mRNA) for either the 22-kDa protein (191 amino acids) or the 20-kDa protein (176 amino acids). Alternative splicing in the third exon, which removes the RNA-encoding amino acids 32 to 46, is responsible for the two isoforms found in the pituitary. Both mRNAs have a signal sequence that causes them to be translated in the rough endoplasmic reticulum (ER) and enter the secretory pathway. Subsequent processing converts the two pre–pro-GHs first to the pro-GHs and then to the mature GHs. The cleavage of the pro-sequence and disulfide bond formation occur during transit through the Golgi bodies. The somatotroph stores mature GH in granules until GHRH stimulates the somatotroph to secrete the hormones. The 22-kDa version is the dominant form of GH.

Growth hormone is one of a family of homologous hormones that exhibit overlap of activity

GH appears to be a single-copy gene, but four other hormones have significant homology to GH. Most striking are three hormones made by the placenta: placental-variant GH (pvGH) and human chorionic somatomammotropins 1 and 2 (hCS1 and hCS2) (Table 48-1). Human genes for these hormones are also located in the GH gene cluster on the long arm of chromosome 17. The multiple genes in this cluster have an identical intron structure and code for proteins of similar size. Substantial amino acid sequence homology occurs among GH and these three other proteins. pvGH, like the 22-kDa GH, is a 191–amino acid peptide. It has a 93% primary sequence homology with GH and virtually the same affinity for the hepatic GH receptor as GH. It appears able to mimic some of the biological actions of GH and may be an important modulator of systemic IGF-1 production during pregnancy. (As discussed later, a major action of GH is to stimulate secretion of IGF-1.) The hCSs are also called human placental lactogens (hPLs). The affinity of the two forms of hCS for the GH receptor is 100- to 1000-fold less than that of either GH or pvGH. As a result, the hCSs are less effective in promoting IGF-1 or IGF-2 production. The somatomammotropins are primarily lactogenic, priming the breast for lactation after birth (see Table 56-7).

Table 48-1 Homology of Growth Hormone to Chorionic Hormones and Prolactin


The pituitary hormone prolactin (PRL; Table 48-1) is the fourth hormone with homology to GH. It has a 16% amino acid homology with pituitary GH. The principal physiological role of PRL involves promoting milk production in lactating women (see Chapter 56). PRL is made by lactotrophs in the anterior pituitary. Its homology to GH suggests that the two hormones, despite their divergent actions, arose from some common precursor by a gene-duplication event. The sequence homology between these proteins is underscored by the observation that GH and PRL have similar affinities for the PRL receptor. The converse is not true—that is, PRL has no significant affinity for the GH receptor and thus has no growth-promoting activity. As discussed later, the PRL and GH receptors are coupled to an intracellular signaling system that involves stimulation of the JAK family of tyrosine kinases (see Chapter 3) as an early postreceptor event.

Men, like women, make PRL throughout their lives. However, no physiological role for PRL in boys or men has been defined. Both men and women with disorders involving hypersecretion of GH or PRL can develop galactorrhea (breast milk secretion). Although GH and PRL are normally secreted by distinct cell populations in the anterior pituitary, some benign, GH-producing pituitary adenomas (i.e., tumors) secrete PRL along with GH.

Somatotrophs secrete growth hormone in pulses

It remains a paradox that whereas growth occurs slowly over months and years, the secretion of GH is highly episodic, varying on a minute-to-minute basis. Most physiologically normal children experience episodes or bursts of GH secretion throughout the day, most prominently within the first several hours of sleep. Underlying each peak in plasma levels of GH, illustrated for an adult in Figure 48-2, are bursts of many hundreds of pulses of GH secretion by the somatotrophs in the anterior pituitary. With the induction of slow-wave sleep, several volleys of GH pulses may occur; it is estimated that more than 70% of total daily GH secretion occurs during these periods. This pulsatile secretion underlines the prominent role of the CNS in the regulation of GH secretion and growth. The circulating GH concentrations may be up to 100-fold higher during the bursts of GH secretion (i.e., the peaks in Fig. 48-2) than during intervening periods. The pattern of bursts depends on sleep-wake patterns, not on light-dark patterns. Exercise, stress, high-protein meals, and fasting also cause a rise in the mean GH level in humans. In circumstances in which GH secretion is stimulated (e.g., fasting or a high-protein diet), the increased GH output results from an increase in the frequency, rather than the amplitude, of pulses of GH that the somatotrophs secrete.


Figure 48-2 Bursts in plasma levels of GH, sampled in the blood plasma of a 23-year-old woman every 5 minutes over a 24-hour period. Each peak in the plasma GH concentration reflects bursts of hundreds of GH secretory pulses by the somatotrophs of the anterior pituitary. These bursts are most common during the first few hours of sleep. The integrated amount of GH secreted each day is higher during pubertal growth than in younger children or in adults. (Data from Hartman ML, Veldhuis JD, Vance ML, et al: J Clin Endocrinol Metab 1990; 70:1375.)

Growth hormone secretion is under hierarchical control from growth hormone–releasing hormone and somatostatin

The coordination of GH secretion by the somatotrophs during a secretory pulse presumably occurs in response to both positive and negative hypothalamic control signals.

Growth Hormone–Releasing Hormone Small-diameter neurons in the arcuate nucleus of the hypothalamus secrete GH-releasing hormone (GHRH), a 43–amino acid peptide that reaches the somatotrophs in the anterior pituitary through the hypophyseal portal blood (Fig. 48-3). As the name implies, this neuropeptide promotes GH secretion by the somatotrophs. GHRH is made principally in the hypothalamus, but it can also be found in neuroectodermal tissue outside the CNS; it was first isolated and purified from a pancreatic islet cell tumor of a patient with acromegaly.


Figure 48-3 Synthesis and release of GHRH and SS and the control of GH release. Small-bodied neurons in the arcuate nucleus of the hypothalamus secrete GHRH, a 43–amino acid peptide that reaches the somatotrophs in the anterior pituitary through the long portal veins. GHRH stimulates the somatotrophs to release GH stored in secretory granules by raising [cAMP]i and [Ca2+]i. Neurons in the periventricular region of the hypothalamus synthesize SS, a 14–amino acid neuropeptide. SS, which also travels to the anterior pituitary through the long portal vessels, is a potent inhibitor of GH secretion. SS acts by inhibiting adenylyl cyclase (AC) and thus lowering [Ca2+]i. PKA, protein kinase A.

Growth Hormone–Releasing Hormone Receptor GHRH binds to a G protein–coupled receptor (GPCR) on the somatotrophs and activates Gαs, which, in turn, stimulates adenylyl cyclase (see Chapter 3). The subsequent rise in [cAMP]i causes increased gene transcription and synthesis of GH. In addition, the rise in [cAMP]i opens Ca2+ channels in the plasma membrane and causes [Ca2+]i to rise. This increase in [Ca2+]i stimulates the release of preformed GH.

Ghrelin A relatively newly discovered hormone, ghrelin consists of 28 amino acids. One of the serine residues is linked to an octanol group, and only this acylated form of the peptide is biologically active. Distinct endocrine cells within the mucosal layer of the stomach release ghrelin in response to fasting. Endocrine cells throughout the gastrointestinal tract also make ghrelin, although the highest ghrelin concentrations are in the fundus of the stomach. The arcuate nucleus of the hypothalamus also makes small amounts of ghrelin. Infusion of ghrelin either into the bloodstream or into the cerebral ventricles markedly increases growth hormone secretion. Indeed, ghrelin appears to be involved in the postmeal stimulation of growth hormone secretion. It has been more difficult to define the extent to which ghrelin—versus GHRH and somatostatin (SS)—contributes to the changes in normal growth hormone secretion in response to fasting, amino acid feeding, and carbohydrate feeding. Ghrelin also appears to stimulate appetite, thereby contributing to body mass regulation as well as linear growth. (See Note: Ghrelin)

Ghrelin Receptor The hormone binds to a GPCR designated the GH secretagogue receptor (GHSR). This receptor was first identified because it binds synthetic peptide ligands that stimulate GH secretion. In this regard, GHSR is like the GHRH receptor (GHRHR); however, GHRH is not a ligand for GHSR.

Somatostatin The hypothalamus also synthesizes SS, a 14–amino acid neuropeptide. SS is made in the periventricular region of the hypothalamus and is secreted into the hypophyseal portal blood supply. It is a potent inhibitor of GH secretion. SS is also made elsewhere in the brain and in selected tissues outside the CNS, such as the pancreatic islet Δ cells (see Chapter 51) and D cells in the gastrointestinal tract (see Chapter 42). Within the CNS, the 14–amino acid form of SS (SS-14) dominates. The gastrointestinal tract predominantly expresses a 28–amino acid splice variant; the N-terminal 14 amino acids of SS-28 are identical to those of SS-14.

It appears that the primary regulation of GH secretion is stimulatory, because sectioning the pituitary stalk, thereby interrupting the portal blood flow from the hypothalamus to the pituitary, leads to a decline in GH secretion. Conversely, sectioning of the stalk leads to a rise in PRL levels, presumably because dopamine made in the hypothalamus inhibits PRL secretion in the anterior pituitary (see Chapter 47). It also appears that the pulses of GH secretion are entrained by the pulsatile secretion of GHRH (as opposed to the periodic loss of SS inhibition). Such a relationship between releasing factors and pulses of pituitary hormone secretion has been directly documented in primates for the secretion of gonadotropin-releasing hormone (GnRH) and luteinizing hormone (LH), but not yet for GHRH.

Somatostatin Receptor SS binds to a GPCR on the somatotrophs and activates Gαi, which inhibits adenylyl cyclase. As a result, [Ca2+]i decreases, thereby diminishing the responsiveness of the somatotroph to GHRH. When somatotrophs are exposed to both GHRH and SS, the inhibitory action of SS prevails.

Both growth hormone and insulin-like growth factor 1, whose secretion is stimulated by growth hormone, negatively feed back on growth hormone secretion by somatotrophs

In addition to being controlled by GHRH and SS released from the hypothalamus, somatotroph secretion of GH is regulated by a negative feedback loop involving IGF-1. As discussed later, GH triggers the secretion of IGF-1 from GH target tissues throughout the body. Indeed, IGF-1 appears to mediate many of the growth-promoting actions of GH. The circulating levels of IGF-1, which produce its endocrineeffects, largely reflect its hepatic synthesis. IGF-1 synthesized in tissues such as muscle, cartilage, and bone may act more in a paracrine or autocrine fashion to promote local tissue growth. An increase in the circulating concentration of IGF-1 suppresses GH secretion through both direct and indirect mechanisms (Fig. 48-4). First, circulating IGF-1 exerts a direct action on the pituitary to suppress GH secretion by the somatotrophs, probably inhibiting GH secretion by a mechanism different from that of SS. In its peripheral target cells, IGF-1 acts through a receptor tyrosine kinase (see Chapter 3), and not by either the Ca2+ or cAMP messenger systems. IGF-1 presumably acts by this same mechanism to inhibit GH secretion in somatotrophs.


Figure 48-4 GH and IGF-1 (also called somatomedin C) negative feedback loops. Both GH and IGF-1 feed back—either directly or indirectly—on the somatotrophs in the anterior pituitary to decrease GH secretion. GH itself inhibits GH secretion (short loop). IGF-1, whose release is stimulated by GH, inhibits GH release by three routes, one of which is direct and two of which are indirect. The direct action is for IGF-1 to inhibit the somatotroph. The first indirect pathway is for IGF-1 to suppress GHRH release in the hypothalamus. The second is for IGF-1 to increase secretion of SS, which, in turn, inhibits the somatotroph.

Second, evidence also points to two indirect feedback pathways by which circulating IGF-1 inhibits GH secretion. IGF-1 appears to suppress GHRH release in the hypothalamus and to increase SS secretion. Yet another feedback system, independent of IGF-1, reduces GH secretion; GH itself appears to inhibit GH secretion in a short-loop feedback system.

Growth hormone has acute anti-insulin metabolic effects and chronic growth-promoting effects mediated by insulin-like growth factor 1

Once secreted, most GH circulates free in the plasma. However, a significant fraction (~40% for the 22-kDa GH) is complexed to a GH-binding protein formed by proteolytic cleavage of the extracellular domain of GH receptors in GH target tissues. This protein fragment binds to GH with high affinity, thereby increasing the half-life of GH and competing with GH target tissues for GH. In the circulation, GH has a half-life of ~25 minutes.

Growth Hormone Receptor GH binds to specific receptors on the surface of multiple target tissues. The monomeric GH receptor is a 620–amino acid protein with a single membrane spanning segment. The molecular weight of the purified receptor (~130 kDa) greatly exceeds that predicted from the amino acid composition (~70 kDa) as a result of extensive glycosylation. The receptor does not resemble any of the GPCRs or receptors with intrinsic tyrosine kinase activity, but rather is a tyrosine kinase–associated receptor that is related to several cytokine receptors (see Chapter 3). The GH receptor forms a dimer when one GH molecule simultaneously binds to sites on two monomers and acts as a bridge. Receptor occupancy increases the activity of a tyrosine kinase (JAK 2 family) that is associated with, but is not an integral part of, the GH receptor. This tyrosine kinase triggers a series of protein phosphorylations that modulate target cell activity.

Acute Effects of Growth Hormone GH has certain acute (minutes to hours) actions on muscle, adipose tissue, and liver that may not necessarily be related to the more long-term growth-promoting actions of GH. These acute metabolic effects (Table 48-2) include stimulation of lipolysis in adipose tissue, inhibition of glucose uptake by muscle, and stimulation of gluconeogenesis by hepatocytes. These actions oppose the normal effects of insulin (see Chapter 56) on these same tissues and have been termed the anti-insulin or diabetogenic actions of GH. Chronic oversecretion of GH, such as occurs in patients with GH-producing tumors in acromegaly, is accompanied by insulin resistance and often by glucose intolerance or frank diabetes.

Table 48-2 Diabetogenic Effects of Growth Hormone




↓ Glucose uptake


↑ Lipolysis


↑ Gluconeogenesis

Muscle, fat, and liver

Insulin resistance

Long-Term Effects of Growth Hormone Through IGF-1 Distinct from these short-term actions of GH is its action to promote tissue growth by stimulating target tissues to produce IGFs. In 1957, Salmon and Daughaday reported that GH itself does not have growth-promoting action on epiphyseal cartilage (the site where longitudinal bone growth occurs). In those experiments, the addition of serum from normal animals, but not from hypophysectomized (GH-deficient) animals, stimulated cartilage growth in vitro (assayed as incorporation of radiolabeled sulfate into cartilage). The addition of GH to GH-deficient serum did not restore the growth-promoting activity seen with normal serum. However, when the GH-deficient animals were treated in vivo with GH, their plasma promoted cartilage growth in vitro. This finding led to the hypothesis that, in animals, GH provokes the secretion of another circulating factor that “mediates” the action of GH. Initially called “sulfation factor” because of how it was assayed, this intermediate was subsequently termed somatomedin because it mediates the somatic effects of GH. The responsible proteins were isolated and purified and were found to be identical to two peptides with a primary structure much like that of proinsulin, and they were termed IGFs (Fig. 48-5). These peptide hormones are made in various tissues, including the liver, kidney, muscle, cartilage, and bone. They are called “insulin-like” growth factors because they exert insulin-like actions in isolated adipocytes and can produce hypoglycemia in animals and humans. The liver produces most of the IGF-1 present in the circulation. IGF-1 production appears to be more closely related to GH secretion than does IGF-2 production.


Figure 48-5 Structure of the IGFs. Insulin, IGF-1, and IGF-2 share three domains (A, B, and C), which have in common a high degree of amino acid sequence homology. The C region is cleaved from insulin (as the C peptide) during processing, but it is not cleaved from either IGF-1 or IGF-2. In addition, IGF-1 and IGF-2 also have a short D domain.


Insulin-like growth factor 1, which interacts with a receptor similar to the insulin receptor, is the principal mediator of the growth-promoting action of growth hormone

The synthesis of IGF-1 and, to a lesser extent, IGF-2 depends on circulating GH. As described earlier, the periodic nature of GH secretion results in a wide range of plasma GH concentrations. In contrast, plasma [IGF-1] does not vary by more than ~2-fold over a 24-hour period. The plasma [IGF-1] in effect integrates the pulsatile, highly fluctuating GH concentration. The reason for the relatively steady plasma levels of IGF-1 is that like GH—but unlike most peptide hormones—IGF-1 circulates bound to several IGF-1–binding proteins. These binding proteins are made principally in the liver, but they are also manufactured by other tissues. More than 90% of IGF-1 measured in the serum is bound to these proteins. At least six distinct IGF-binding proteins have been identified. In addition to providing a buffer pool in plasma of bound IGF, these proteins may also aid the transfer of IGF to the tissue receptors, thereby facilitating the action of these hormones. The local free fraction of IGF-1 is probably the more biologically active component that binds to the receptor and stimulates tissue growth.

Like other peptide hormones, IGF-1 and IGF-2 are synthesized through the secretory pathway (see Chapter 2) and are secreted into the extracellular space, where they may act locally in a paracrine fashion. In the extracellular space, the IGFs encounter binding proteins that may promote local retention of the secreted hormone by increasing the overall molecular size of the complex. This action inhibits the entry of the IGFs into the vascular system. Thus, local concentrations of the IGFs are likely to be much higher than plasma concentrations.

Whether made locally or reaching tissue through the circulation, IGF-1 acts through a specific receptor tyrosine kinase (see Chapter 3), a heterotetramer that is structurally related to the insulin receptor (Fig. 48-6). Like the insulin receptor (see Chapter 51), the IGF-1 receptor has two completely extracellular α chains and two transmembrane β chains. Also like in the insulin receptor, the β chains have the intrinsic tyrosine kinase activity. Binding of IGF-1 to its receptor enhances receptor autophosphorylation as well as phosphorylation of downstream effectors. The structural homology between the insulin and IGF-1 receptors is sufficiently high that insulin can bind to the IGF-1 receptor, although with an affinity that is approximately two orders of magnitude less than that for IGF-1. The same is true for the binding of IGF-1 to insulin receptors. In fact, the homology between the insulin and IGF-1 receptors is so strong that hybrid receptors containing one α-β chain of the insulin receptor and one α-β chain of the IGF-1 receptor are present in many tissues. These hybrid receptors bind both insulin and IGF-1, but their affinity for IGF-1 is greater.


Figure 48-6 Comparison of insulin, IGF-1, and IGF-2 receptors. Both the insulin and IGF-1 receptors are heterotetramers joined by disulfide bonds. For both, the cytoplasmic portions of the β subunits have tyrosine kinase domains as well as autophosphorylation sites. The IGF-2 receptor (also called M6P receptor) is a single polypeptide chain with no kinase domain.

Given the structural similarity between insulin and IGFs and between the insulin receptor and the IGF-1 receptor, it is not surprising that IGFs can exert insulin-like actions in vivo. This effect has been particularly well studied for IGF-1, which, like insulin, induces hypoglycemia when it is injected into animals. This action is largely the result of the increased uptake of glucose into muscle tissue. IGF-1 is less effective in mimicking insulin’s action on adipose and liver tissue; in humans, these tissues have few IGF-1 receptors. In muscle, IGF-1 promotes the uptake of radiolabeled amino acids and stimulates protein synthesis at concentrations that do not stimulate glucose uptake, findings indicating that the growth-promoting actions of IGF-1 are expressed at lower circulating concentrations than those required to produce hypoglycemia.

Insulin-like growth factor 2 has actions similar to those of insulin-like growth factor 1 but is less dependent on growth hormone

The physiology of IGF-2 differs from that of IGF-1 in certain important respects. First, as noted earlier, the synthesis of IGF-2 depends less on circulating GH than does that of IGF-1. In pituitary dwarfism secondary to GH deficiency, the circulating concentration of IGF-1 is decreased, but that of IGF-2 is not. In states of excessive GH secretion, plasma IGF-1 is reliably elevated, whereas plasma IGF-2 is not.

Although IGF-2 also binds to the IGF-1 receptor, it preferentially binds to its own so-called IGF-2 receptor. This IGF-2 receptor consists of a single-chain polypeptide and is structurally very distinct from the IGF-1 receptor (Fig. 48-6). The IGF-2 receptor lacks a tyrosine kinase domain and does not undergo autophosphorylation in response to the binding of either IGF-2 or IGF-1. The IGF-2 receptor also binds mannose-6-phosphate (M6P), but at a site different from that for IGF-2 binding, and the receptor’s physiological role appears to be in processing mannosylated proteins by targeting them for lysosomal degradation. Thus, the term IGF-2 receptor is somewhat of a misnomer; the IGF-2 receptor’s role in the physiological action of IGF-2 is not clear.

Despite these differences, IGF-2 does share with IGF-1 (and also with insulin) the ability to promote tissue growth and to cause acute hypoglycemia. These properties appear to result from IGF-2’s structural similarity to proinsulin and its ability to bind to the IGF-1 receptor.

Although growth rate usually parallels plasma levels of insulin-like growth factor 1, the two diverge both early and late in life

Illustrated in Figure 48-7 is the mean concentration of total IGF-1 (both free and bound) found in human serum as a function of age. Also shown is the normal rate of height increase (cm/yr). During puberty, the greatest growth rates are observed at times when plasma [IGF-1] is highest. A similar comparison can be made using GH, provided care is taken to obtain multiple measurements at each age and thereby account for the pulsatile secretion and marked diurnal changes that occur in plasma [GH].


Figure 48-7 Serum IGF-1 levels and height velocity as a function of age. The red curve shows the mean plasma concentrations of IGF-1 as a function of age in female humans. The curve for male humans is similar, but the peak is shifted to an older age by 3 to 4 years. The brown curve indicates mean female height velocity—the rate at which height increases (cm/yr). The pubertal peak rate of growth corresponds to the peak serum concentrations of IGF-1. (Data from Reiter EO, Rosenfeld RG: Normal and aberrant growth. In Wilson JD, Foster DW, Kronenberg HM, Larsen PR [eds]: Williams Textbook of Endocrinology, 9th ed, pp 1427-1507. Philadelphia: WB Saunders, 1998.)

Plasma Level of IGF-1 as a Measure of Growth Hormone Secretion

The plasma concentration of IGF-1 is a valuable measure of GH secretion. The wide swings in plasma [GH] that result from the pulsatile secretion of this hormone have confounded efforts to use GH measurements to diagnose disorders of GH deficiency or excess. However, an increased circulating concentration of IGF-1 is one of the most useful clinical measures of the excess GH secretion that occurs in acromegaly (i.e., GH excess in adults) and gigantism (i.e., GH excess in children). Measurement of plasma [IGF-1] has also helped to explain the genesis of a particular type of dwarfism known as Laron dwarfism. These patients were initially identified as persons with growth failure mimicking that of typical pituitary dwarfism; however, plasma [GH] is normal or elevated, and treatment with GH is ineffective in reversing the growth failure. It was subsequently demonstrated that these individuals have mutations of their GH receptors that make the receptors nonfunctional. Thus, the mutant GH receptors cannot trigger the production of IGFs. With the availability of recombinant IGF-1, it is possible that effective treatment of these children will restore growth.

Despite the structural similarity of their receptors, IGF-1 and insulin exert different actions on tissues. IGF-1 has a more marked effect on growth, and insulin has a more significant effect on glucose and lipid metabolism. However, the differences in the postreceptor signaling pathways triggered by the two hormones have not been well defined.

Whereas good general concordance exists between growth rate and the plasma [IGF-1] during puberty, the two diverge at both younger and older ages. First, during adulthood, longitudinal growth essentially ceases, yet secretion of GH and of IGF-1 continues to be highly regulated. As both men and women age, the circulating concentrations of both hormones decline. For many years, the continued secretion of these hormones in adults was considered to be largely vestigial. This belief was reinforced by the observation that cessation of GH secretion and the consequent decline of IGF-1 after trauma, tumor, or surgical removal of the pituitary did not result in any clear clinical syndrome. However, with the availability of recombinant human GH, replacement of GH in GH-deficient adults has led to remarkable increases in body muscle mass, decreases in fat mass, and improved nitrogen balance (a measure of protein nutrition). These findings support the conclusion that—even after linear growth ceases after puberty—GH and IGF-1 remain important regulators of body composition and appear to promote anabolic actions in muscle. Indeed, some investigators have suggested that supplementing physiologically normal adults with GH or IGF-1 may reverse some of the effects of aging, including loss of muscle mass, negative nitrogen balance, and osteoporosis.

A second period of life during which divergence between longitudinal growth and IGF-1 occurs is very early childhood (Fig. 48-7). This period is characterized by a very rapid growth rate, but quite low IGF-1 levels. If this time frame is extended back to intrauterine life, then the discordance is even greater. Indeed, children with complete GH deficiency have very low plasma [IGF-1] levels but are of normal length and weight at birth. This observation suggests that during intrauterine life, factors other than GH and IGF-1 are important regulators of growth. One of these additional factors may be insulin, as discussed later.

Another explanation for the divergence between growth rate and IGF-1 levels may be IGF-2, which may be an important mediator of intrauterine growth. The plasma concentration of IGF-2 is greater during fetal life than later, and it peaks just before birth. Plasma [IGF-2] plummets soon after birth, but then it gradually doubles between birth and age 1 year and remains at this level until at least the age of 80 years. Thus, IGF-2 levels are at adult levels during the first several years of life, when IGF-1 levels are low but growth is rapid. However, several other hormones may also contribute to somatic growth during the first several years of life.

By age 3 or 4 years, GH and IGF-1 begin to play major roles in the regulation of growth. The concentrations of these hormones rise throughout childhood and peak during the time of the pubertal growth spurt. The rate of long bone growth in the pubertal growth spurt is exceeded only during intrauterine life and early childhood. The frequency of pituitary GH secretory pulses increases markedly at puberty. The factors responsible for this acceleration are not clear. However, the accompanying sexual maturation likely plays some role, because both estradiol and testosterone appear to promote GH secretion.

In addition to these humoral influences, nutritional factors also modulate both GH secretion and IGF-1 production. In both children and adults, GH secretion is triggered by high dietary protein intake. Teleologically, this is intriguing in that it may provide linkage between the availability of amino acids to serve as substrates for body protein synthesis (growth) and the endocrine stimulus of cells to grow. This relationship is not simple, however, because the rise in GH levels in the setting of protein intake is not sufficient to stimulate IGF-1 production fully. This principle is well illustrated by fasting, which is associated with a decline in IGF-1 even with increased GH. During fasting, insulin levels are low. Increased insulin appears to be required, at least in some tissues, for GH to stimulate IGF-1 effectively.

Thyroid hormones, steroids, and insulin also promote growth

Although the discussion to this point emphasizes the action of GH and the GH-induced growth factors as modulators of somatic growth, we could regard them as necessary but not sufficient agents for normal growth. Certain other hormones, as well as receptive growth-responsive cartilage, are required. Because growth is a difficult phenomenon to study, especially in humans (few scientists want to follow an experiment over 10 to 20 years), much of our understanding of endocrine regulation of normal growth derives from observations of abnormal growth as it occurs in clinical syndromes of endocrine excess or deficiency. Several of the more important of these endocrine influences are illustrated here. The exact mechanism by which growth is regulated by these agents is not always well understood.

Thyroid Hormones Next to GH, perhaps the most prominent among the growth-promoting hormones are the thyroid hormones thyroxine and triiodothyronine, which are discussed in Chapter 49. In many nonhuman species, thyroid hormone plays a major role in tissue growth and remodeling. For example, resorption of the tadpole tail during morphogenesis requires thyroid hormone. In humans, deficiency of thyroid hormones early in life can cause dwarfism or cretinism (see Chapter 49). In children with normal thyroid function at birth, development of hypothyroidism at any time before epiphyseal fusion leads to growth retardation or arrest. Much of the loss in height that occurs can be recovered through thyroid hormone treatment, a phenomenon called catch-up growth. However, because the diagnosis of hypothyroidism may elude detection for many months or years, delays in initiating treatment can lead to some loss of potential growth. A child’s growth curve can provide a particularly sensitive early indicator of hypothyroidism.

Sex Steroids As with thyroid hormones, the importance of sex steroids for growth is most readily understood by considering the effects of deficiency or excess of these hormones. Androgen or estrogen excess occurring before the pubertal growth spurt accelerates bone growth. However, the sex steroids also accelerate the rate at which the skeleton matures and thus shorten the time available for growth before epiphyseal closure occurs. Most of the time, the dominant effect of sex steroids is to narrow the growth window, thereby diminishing ultimate longitudinal bone growth. This effect is well illustrated in settings in which children are exposed to excessive sex steroid at an early age. The sex steroids can come from endogenous sources (e.g., early maturation of the hypothalamic-pituitary-gonadal axis that produces premature puberty, or tumors that secrete estrogen or androgen) or from exogenous sources (e.g., children who take sex steroids prescribed for others). Again, the growth curve is useful in that it typically shows an increase in growth rate, followed by an early leveling off of growth associated with the development of secondary sexual characteristics.

Glucocorticoids An excess of adrenal glucocorticoids inhibits growth. In children who produce too much cortisol, as a result of either adrenal or pituitary tumors (which produce adrenocorticotropic hormone [ACTH] and cause secondary increases in plasma cortisol), growth ceases. The use of synthetic glucocorticoids in treating various serious illnesses (e.g., asthma, organ transplantation, various chronic autoimmune processes) also arrests growth. Restoration of normal growth does not occur until the glucocorticoid levels return toward normal. Neither GH nor IGF-1 concentrations drop significantly during glucocorticoid treatment. The failure of GH administration to correct the growth retardation that occurs in glucocorticoid-treated children further confirms that GH deficiency cannot account for the growth failure associated with glucocorticoid excess. Because linear growth is related to cartilage and bone synthesis at the growth plates, glucocorticoids presumably are acting at least in part at these sites to impair growth. However, the specific biochemical locus at which glucocorticoids act remains unclear. In adults, as in children, glucocorticoid excess impairs tissue anabolism and thus may manifest as wasting in some tissues (e.g., bone, muscle, subcutaneous connective tissue), rather than growth failure. This tissue wasting results in some of the major clinical morbidity of glucocorticoids (i.e., osteoporosis, muscle weakness, and bruising).

In glucocorticoid deficiency, growth is not substantially affected. However, other deleterious effects of cortisol deficiency (e.g., hypoglycemia; see Chapter 51) dominate.

Insulin This hormone is also an important growth factor, particularly in utero. For example, women with diabetes frequently have high blood levels of glucose during pregnancy and deliver babies of high birth weight (fetal macrosomy). The developing fetus, exposed to glucose concentrations that are higher than normal, secretes additional insulin. Hyperinsulinemia results in increased fetal growth. Fetal macrosomy can create significant obstetric difficulties at the time of delivery.

Conversely, infants born with pancreatic agenesis or with one of several forms of severe insulin resistance are very small at birth. One form of this condition, leprechaunism, is the result of a defect in the insulin receptor (see Chapter 51 for the box on this topic). Thus, it appears that insulin, acting through its own receptor, is an intrauterine growth factor.

Severe insulin deficiency produces a marked catabolic effect associated with wasting of lean body mass in both children and adults. The acute adverse effects of such deficiency (dehydration and acidosis) dominate the clinical picture. Milder degrees of insulin deficiency, seen in patients with chronically undertreated diabetes, diminish growth in affected children. Improved diabetes management may allow restoration of normal growth rates and possibly even some transient accelerated or catch-up growth. Thus, with good care, children with diabetes can expect to achieve normal adult height.

The musculoskeletal system responds to growth stimuli of the GHRH-GH–IGF-1 axis

Longitudinal growth involves lengthening of the somatic tissues (including bone, muscle, tendons, and skin) through a combination of tissue hyperplasia and hypertrophy. Each of these tissues remodels its structure throughout life. For bone, longitudinal growth occurs by the hyperplasia of chondrocytes at the growth plates of the long bones, followed by endochondral ossification. The calcified cartilage is remodeled as it moves toward the metaphyses of the bone, where it is eventually replaced by true lamellar and trabecular bone (see Chapter 52). This process continues until epiphyseal closure occurs toward the completion of puberty.

Anabolic-Androgenic Steroids

We are all unfortunately familiar with the potential for abuse of anabolic-androgenic steroids by bodybuilders and competitive athletes. Illicit use of these agents appears to be widespread in sports, where strength is closely linked to overall performance. In addition to naturally occurring androgens such as testosterone, dihydrotestosterone, androstenedione, and dehydroepiandrosterone, many different synthetic androgenic steroids—as well as GH—serve as performance enhancers. In addition to the sought after “beneficial” effects of increasing muscle mass and strength, each of these agents carries with it a plethora of adverse side effects. Some—such as oily skin, acne, and hair growth—are principally cosmetic. Others—including liver function abnormalities, mood changes with aggressive behavior, and hepatocellular carcinoma—are much more serious. Illicit use of these agents by younger athletes, especially teenagers, is also problematic with regard to alterations in growth and sexual maturation.

The process of cartilage formation and longitudinal bone growth begins as the cellular elements capable of forming cartilage divide along the growth plate and then migrate toward the more mature bone. These cells synthesize the extracellular matrix of cartilage, which include type II collagen, hyaluronic acid, and mucopolysaccharides. These cells appear to respond directly to GH by proliferating and increasing production of the extracellular matrix. This response involves local generation of IGF-1 within the cartilage as an early event in the growth process. As the cells more closely approach the already formed cortical and trabecular bone, ossification of the extracellular matrix begins, and eventually the cellular elements become isolated by the calcifying cartilage. However, this calcified cartilage is not structurally the same as normal bone, and soon after formation, it begins to be remodeled by an ingrowth of cellular elements (osteoclasts and osteoblasts) from adjacent bone. Eventually, it is replaced by normal bone and becomes part of the metaphysis of the long bone.

In most children, growth ceases within several years after completion of puberty, when the chondrocytes at the growth plates of the long bones stop dividing and calcify the previously cartilaginous surrounding matrix. After puberty, radial growth occurs as bones increase their diameter through a process of endosteal bone resorption and periosteal bone deposition. This process is not strictly compartmentalized; that is, resorption and deposition of bone occur at both the periosteal and endosteal surfaces. However, during periods of growth, the rate of periosteal deposition exceeds the rate of endosteal resorption, and the bone shafts grow in width and thickness.

As may be expected, numerous disorders disrupt the complex process of endochondral bone growth on a genetic or congenital basis (e.g., defects in collagen or mucopolysaccharide synthesis) and lead to genetic forms of dwarfism. In these settings, the GHRH-GH–IGF-1 axis is entirely intact and appears to regulate normally. No apparent compensation occurs for the short stature by increased GH secretion, a finding suggesting that the axis is not sensitive to the growth process per se, but simply to the intermediate chemical mediator IGF-1.

GH and IGF-1 clearly play important roles in mediating longitudinal bone growth and also modulate growth of other tissues. Thus, proportional growth of muscle occurs as bones elongate, and the visceral organs enlarge as the torso increases in size. The mechanisms by which GH and IGF-1 coordinate this process and the way in which other hormones or growth factors may be involved continue to be investigated. It is clear that, whereas GH and, more recently, IGF-1 have been considered the major hormones responsible for somatic growth, other tissue growth factors play an important, albeit incompletely defined, role. Table 48-3 lists some of these growth factors. In general, the tissue growth factors have more tissue-specific actions on organogenesis and their growth-promoting activity than the IGFs, and they appear to act largely in a paracrine or autocrine fashion.

Table 48-3 Other Growth Factors Affecting Growth

Nerve growth factor (NGF)

Fibroblast growth factor (FGF)

Angiogenesis factor

Vascular endothelial growth factor (VEGF)

Epidermal growth factor (EGF)

Hepatocyte growth factor (HGF)


The multiple hormonal factors that influence longitudinal growth—discussed in the previous two sections—are themselves responsive to the nutritional intake of a growing individual. For example, amino acids and carbohydrates promote insulin secretion, and amino acids stimulate GH secretion. In addition, the availability of an adequate, balanced nutrient supply likely exerts both direct and indirect influences to promote tissue growth. Independent of any hormonal factors, glucose, fatty acids, and amino acids can each influence the transcription of specific genes. Similarly, amino acids can directly activate the signaling pathways involved in regulating mRNA translation.

Beyond the effects of macronutrients, micronutrients can be similarly important in regulating cell growth and, by extension, growth of the organism. An example is iodine, a deficiency of which can produce dwarfism (see Chapter 49). In a more global fashion, the effect of nutrient limitation on height can be appreciated by considering the differences in mean height between men in North Korea (165 cm) and South Korea (171 cm). As mentioned in Chapter 49, nutritional deprivation early in life can markedly limit longitudinal growth. Perhaps equally fascinating, and only recently appreciated, is that nutritional deprivation early in life also appears to predispose affected individuals to obesity when they reach middle age. This phenomenon was first noted in epidemiologic studies from several European countries that revealed a positive correlation between middle-aged obesity and being born during periods of deprivation during and immediately following the Second World War. Such findings suggest that some level of genetic programming occurs early in life that both diminishes longitudinal growth and predisposes persons to body mass accretion.

The balance between energy intake and expenditure determines body mass

At any age or stage of life, the factors that govern body mass accretion relate specifically to the energy balance between intake and expenditure. If energy intake exceeds expenditure over time—positive energy balance (see Chapter 58)—body mass will increase, assuming the diet is not deficient in essential macronutrients or micronutrients. Small positive deviations from a perfect energy balance, over time, contribute to the major increase in body weight—the “obesity epidemic”—that afflicts many middle-aged adults, and increasingly adolescents, in developed societies. For example, if energy intake in the form of feeding exceeds energy expenditure by only 20 kcal (1 tsp of sugar) daily, over 1 year a person would gain ~1 kg of fat and, over 2 decades, ~20 kg.

Indeed, it is remarkable that many adults maintain a consistent body weight for decades essentially in the absence of conscious effort. Thus, a finely tuned regulatory system must in some manner monitor one or more aspect of body mass, direct the complex process of feeding (appetite and satiety) to replete perceived deficiencies, and yet avoid excesses.

Energy expenditure comprises resting metabolic rate, activity-related energy expenditure, and diet-induced thermogenesis

We can group energy expenditure into three components:

1. Resting metabolic rate (RMR). The metabolism of an individual who is doing essentially nothing (e.g., sleeping) is known as the RMR (see Chapter 58), which amounts to ~2100 kcal/day for a 70-kg adult. The RMR supports maintenance of body temperature, the basal functioning of multiple body systems (e.g., heartbeat, gastrointestinal motility, ventilation), and basic cellular processes (e.g., synthesizing and degrading proteins, maintaining ion gradients, metabolizing nutrients).

2. Activity-related energy expenditure. As we wake up in the morning and begin to move about, we do more than resting metabolism. Exercise or physical work can have a major impact on total daily energy expenditure and varies widely across individuals and within an individual on a day-to-day basis. We also expend energy in activities not classically regarded as exercise or heavy work—tapping our foot while sitting in a chair, looking about the room during physiology lecture, typing at a keyboard—activities dubbed non–exercise-associated thermogenesis or NEAT. Such energy expenditures can vary 3- to 10-fold across individuals and can account for 500 kcal or more of daily energy expenditure. NEAT differences, over time, could considerably contribute to differences in weight gain by individuals consuming identical caloric intake.

3. Diet-induced thermogenesis. Eating requires an additional component of energy expenditure for digesting, absorbing, and storing food. Typically, diet-induced thermogenesis accounts for 10% of energy expenditure. Proteins have a higher thermic effect than either carbohydrates or fats.

Each of these three components of energy expenditure can vary considerably from day to day and is subject to regulation. For example, thyroid hormone is a major regulator of thermogenesis (see Chapter 49). Overproduction of thyroid hormone increases both RMR and NEAT, whereas thyroid hormone deficiency has the opposite effect. (See Note: Effect of Hyperthyroidism on Basal Metabolic Rate)

Hypothalamic centers control the sensations of satiety and hunger

Classic studies—in which investigators made lesions in, or electrically stimulated, specific brain regions—identified two areas in the hypothalamus that are important for controlling eating. A satiety center is located in the ventromedial nucleus (VMN; see Fig. 47-3). Electrical stimulation of the satiety center elicits sensations of satiety, even when an animal is in the presence of food. Conversely, a lesion of the satiety center causes continuous food intake (hyperphagia) even in the absence of need. A hunger (or feeding) center is located in the lateral hypothalamic area (LHA; see Fig. 47-3). Electrical stimulation of this center elicits a voracious appetite, even after an animal has ingested adequate amounts of food. A lesion of the hunger center causes complete and lasting cessation of food intake (aphagia).

Leptin tells the brain how much fat is stored

Only in the last decade have we begun to understand regulatory mechanisms that maintain body mass, made possible by the study of mouse models of obesity. One monogenic model is the Ob/Ob strain of hyperphagic mice that develop morbid obesity; affected mice typically weigh > 100% more than unaffected animals of the same strain. In parabiosis experiments, in which an Ob/Ob mouse was surgically connected to a wild-type mouse (Fig. 48-8A), the Ob/Ob mouse lost weight, a finding suggesting that such mice lack a blood-borne factor. Another model of monogenic obesity is the (Db/Db) mouse, named Db because it secondarily develops type 2 diabetes. These mice are hyperphagic, with adult body weights ~100% higher than those of lean littermates. In parabiosis experiments connecting a Db/Db and a wild-type mouse (Fig. 48-8B), the wild-type mouse starved. Finally, in parabiosis experiment connecting an Ob/Ob to a Db/Db mouse (Fig. 48-8C), the Ob mouse lost weight, but the Db mouse remained obese. These results indicate that (1) the Db mouse makes an excess of the blood-borne factor that cures the Ob mouse, (2) the Db mouse lacks the receptor for this factor, and (3) the absence of the receptor in the Db mouse removes the negative feedback, thus leading to high levels of the blood-borne factor.


Figure 48-8 Parabiosis experiments. In parabiotically coupled mice, ~1% of the cardiac output of one mouse goes to the other, and vice versa, so that the animals exchange blood-borne factors.

In 1994, Jeffrey Friedman and his colleagues used positional cloning to identify leptin (from the Greek leptos [thin]), the blood-borne factor lacked by Ob mice. Leptin is a 17-kDa protein made almost exclusively in adipocytes. The replacement of leptin in Ob/Ob mice leads to rapid weight loss. In 1995, Tepper and collaborators cloned the leptin receptor (Ob-r). The deficiency of this receptor in Db mice makes them leptin resistant. Ob-r is a tyrosine kinase–associated receptor (see Fig. 3-11D) that signals through JAK-2 and STAT (see Fig. 4-15).

Although leptin acts on numerous tissues within the body, most importantly it somehow crosses the blood-brain barrier (see Chapter 11) and modulates specific neurons in the arcuate nucleus of the hypothalamus that control feeding behavior. These same neurons also have insulin receptors. Plasma leptin levels in humans appear to rise in proportion to the mass of adipose tissue. Conversely, the absence of leptin produces extreme hyperphagia, as in Ob/Ob mice. Plasma leptin has a half-time of ~75 minutes, and acute changes in food intake or fasting do not appreciably affect leptin levels. In contrast, insulinconcentrations change dramatically throughout the day in response to dietary intake. Thus, it appears that leptin in some fashion acts as a long-term regulator of CNS feeding behavior, whereas insulin (in addition to multiple other factors) is a short-term regulator of the activity of hypothalamic feeding centers.

In addition to its actions in controlling appetite, leptin promotes fuel utilization. Indeed, leptin-deficient humans paradoxically exhibit some characteristics of starvation (e.g., fuel conservation).

Leptin and insulin are anorexigenic (i.e., satiety) signals for the hypothalamus

At least two classes of neurons within the arcuate nucleus contain receptors for both leptin and insulin. These neurons, in turn, express neuropeptides. One class of neurons produces pro-opiomelanocortin (POMC), whereas the other produces neuropeptide Y (NPY) and agouti-related protein (AgRP).

POMC Neurons Both insulin and leptin stimulate the POMC-secreting neurons (Fig. 48-9), which produce POMC (see Fig. 50-4). At their synapses, POMC neurons release a POMC cleavage product, the melanocortin α-melanocyte–stimulating hormone (α-MSH), which, in turn, binds to MC3R and MC4R melanocortin receptors on second-order neurons. Stimulation of these receptors not only produces satiety and decreases food intake—that is, α-MSH is anorexigenic (from the Greek a [not] + orexis [appetite])—but also increases energy expenditure through activation of descending sympathetic pathways. An indication of the importance of this pathway is that ~4% of individuals with severe, early-onset obesity have mutations in MC3R or MC4R. POMC neurons also synthesize another protein—CART or cocaine-amphetamine–related transcript, which, like α-MSH, is anorexigenic.


Figure 48-9 Control of appetite. ARC, arcuate nucleus.

NPY/AgRP Neurons In addition to stimulating POMC neurons, both insulin and leptin also suppress neurons in the arcuate nucleus that release NPY and AgRP at their synapses (Fig. 48-9). NPY activates NPY receptors—predominantly Y1R and Y5R, which are GPCRs—on secondary neurons, thus stimulating eating behavior. AgRP binds to and inhibits MC4R melanocortin receptors on the secondary neurons in the POMC pathway, thereby inhibiting the anorexigenic effect of α-MSH. In other words, both NPY and AgRP are orexigenic. The yellow obese or agouti mouse overexpresses the agouti protein, which inhibits melanocortin receptors. Overinhibition of MC1R on melanocytes inhibits the dispersion of pigment granules (leading to yellow rather than dark fur). Overinhibition of MC3R and MC4R on anorexigenic neurons blocks the action of α-MSH (leading to obesity).

The secondary neurons to which the POMC and NPY/AgRP neurons project are in five major locations (Fig. 48-9 and also see Fig. 47-3):

1. LHA. In this hunger center, NPY/AgRP neurons stimulate—but POMC neurons inhibit—secondary neurons producing the orexigenic peptides melanin-concentrating hormone (MCH) or orexins A and B.

2. VMN. This nucleus is a satiety center.

3. Dorsomedial hypothalamic nucleus (DMN).

4. Paraventricular nucleus (PVN). This nucleus contains neurons that, in turn, project to both cerebral cortex and areas of the brainstem (see Fig. 47-3).

5. Nucleus tractus solitarii (NTS). This nucleus integrates sensory information from the viscera (see Chapter 2) and also receives input from paraventricular neurons.

Ghrelin is an orexigenic signal for the hypothalamus

Signals originating from the periphery can be not only anorexigenic (i.e., promoting satiety)—as in the case of leptin (from adipose tissue) and insulin (from the pancreas)—but also orexigenic (i.e., promoting appetite). One of these is ghrelin, made in response to fasting by specialized endocrine cells in the gastric mucosa. Indeed, systemically administered ghrelin acutely increases food intake when it is given at physiological doses in both animals and humans. Circulating ghrelin concentrations, however, appear to be lower in obese than lean humans, a finding suggesting that ghrelin does not drive the increased caloric intake in the obese. However, gastric bypass procedures in morbidly obese patients cause ghrelin levels to decline dramatically along with decreases in body weight and food consumption.

Human Obesity

One approach for gauging the extent to which human body mass is appropriate for body height is to compute the body mass index (BMI):


BMIs fall into four major categories: (See Note: Body Mass Index)

1. Underweight: less than 18.5

2. Normal weight: 18.5 to 24.9

3. Overweight: 25 to 29.9

4. Obesity: 30 or more

Although a BMI of 30 or more is an indication of obesity, it is not a direct measure of adipose tissue fat mass. Obesity is an area of intense investigation driven in part by the “obesity epidemic” that is adversely affecting the health of a large fraction of the population of developed nations.

The demonstration that replacement of leptin in Ob/Ob mice led to rapid weight loss raised considerable enthusiasm for the potential of leptin as a pharmacological agent in the treatment of human obesity. Indeed, several extremely rare individuals had been identified with autosomal recessive monogenic obesity secondary to leptin deficiency, like the Ob/Ob mouse. As expected, these individuals respond to exogenous leptin administration with a marked reduction in body weight. However, investigators soon appreciated that most obese persons are not leptin deficient. Quite the contrary, human plasma leptin concentrations increase proportionately to BMI, which is a rough estimate of adipose tissue fat mass.

Although obese persons generally are not leptin deficient, approximately one third of obese persons lose weight in response to exogenous leptin. These individuals are leptin resistant, but they eventually respond to sufficiently high levels of the hormone. In the other two thirds of obese persons, the leptin resistance is so severe that they fail to respond even to the exogenous hormone. Lean persons lose weight in response to leptin.

In addition to mutations to the leptin gene, two other extremely rare mutations cause monogenic human obesity. One is mutation of the leptin receptor gene (analogous to the Db mouse) and the other is mutation of the POMC gene (leading to loss of the anorexigenic α-MSH). A more common—although rare—cause of monogenic human obesity is a mutation in the melanocortin MC4 receptor, the target of α-MSH.

Currently, no satisfactory pharmacological approaches are available to treat obesity. Of the two agents currently approved in the United States by the Food and Drug Administration, one is a serotonin re-uptake inhibitor and one blocks fat digestion and therefore absorption within the gastrointestinal tract. Neither agent directly intervenes at targets within the hypothalamic neuroendocrine control system (Fig. 48-9). More importantly, each is limited by side effects, and each is only minimally effective in decreasing weight. Perhaps more promising, but still being tested, are antagonists of the cannabinoid receptors (CB-1 and CB-1), which are GPCRs. These drugs decrease body weight by blocking access of endogenously produced arachidonic acid derivatives known as endocannabinoids, which are ligands of CB-1 and CB-2. These receptors are located in many areas throughout the brain, as well as in peripheral tissues. They are richly represented in the basal hypothalamus as well as within the nucleus accumbens in the limbic system. CB blockers appear to be effective in achieving and maintaining meaningful weight reduction (10 to 20 kg) for more than 1 year. The same agents are also effective in decreasing smoking behavior. Investigators are still unraveling how blocking the cannabinoid receptor affects the output of hypothalamic neurons that regulate appetite.

As discussed previously, ghrelin binds to GHSR, which is present in neurons of the arcuate nucleus as well as vagal afferents. Some hypothalamic neurons themselves contain ghrelin, and injection of ghrelin into the cerebral ventricles stimulates feeding. It is not clear to what extent circulating ghrelin promotes appetite through vagal afferents versus hypothalamic receptors. As noted earlier, ghrelin also promotes the secretion of GH and thus appears to have a role in both longitudinal growth and body mass accretion.

Plasma nutrient levels and enteric hormones are short-term factors that regulate feeding

Investigators have proposed various theories to explain the short-term regulation of food intake, including models focusing on the regulation of levels of blood glucose (glucostatic), amino acid (aminostatic), or lipid (lipostatic). For example, hypoglycemia produces hunger and also increases the firing rate of glucose-sensitive neurons in the hunger center in the LHA, but it decreases the firing rate of glucose-sensitive neurons in the satiety center in the VMN. Hypoglycemia also activates orexin-containing neurons in the LHA.

Feedback from the gastrointestinal tract also controls the short-term desire for food (Fig. 48-9). Gastrointestinal distention triggers vagal afferents that, through the NTS, suppress the hunger center. Peripherally administering any of several gastrointestinal peptide hormones normally released in response to a meal—glucagon; gastrin-releasing peptide (GRP), SS, and peptide YY (PYY) (see Chapter 41); cholecystokinin (CCK, see Chapter 43); and glucagon-like peptide 1 (GLP-1; see Chapter 51)—reduces meal size (i.e., these substances are anorexigenic). The most important is CCK, which is more effective when it is injected directly into the peritoneal cavity; this effect requires an intact vagus nerve. Therefore, CCK—like gastric distention—may act through vagal afferents. Additionally, an oropharyngeal reflex responds to chewing and swallowing; it may meter food intake, thus inhibiting further eating after a threshold.

An important aspect of our increasing understanding of the neuroendocrine control systems that regulate appetite, satiety, and energy expenditure and thereby body mass is the further affirmation that these processes have a genetic and biochemical basis. Two other factors that influence body mass are cortical control (e.g., “will power”) and environment (e.g., the availability of high-calorie foods). Our emerging appreciation of the biological basis of obesity may allow a more scientific and clinical approach to therapeutic interventions—rather than simply blaming affected patients for their obesity.


Books and Reviews

Argetsinger LS, Carter-Su C: Mechanism of signaling by growth hormone receptor. Physiol Rev 1996; 76:1089-1107.

Etherton TD, Bauman DE: Biology of somatotropin in growth and lactation of domestic animals. Physiol Rev 1998; 78:745-761.

Flier JS: Obesity wars: Molecular progress confronts an expanding epidemic. Cell 2004; 116:337-350.

Kojima M, Kangawa K: Ghrelin: Structure and function. Physiol Rev 2005; 85:495-522.

Mayo KE, Godfrey PA, Suhr ST, et al: Growth hormone–releasing hormone: Synthesis and signaling. Recent Prog Horm Res 1995; 50:35-73.

Reiter EO, Rosenfeld RG: Normal and aberrant growth. In Wilson JD, Foster DW, Kronenberg HM, Larsen PR (eds): Williams Textbook of Endocrinology, 9th ed, pp 1427-1507. Philadelphia: WB Saunders, 1998.

Stewart CEH, Rotwein P: Growth, differentiation, and survival: Multiple physiological functions for insulin-like growth factors. Physiol Rev 1996; 76:1005-1026.

Journal Articles

Coleman DL: Effects of parabiosis of obese with diabetes and normal mice. Diabetologia 1973; 9:294-298.

Daughaday WH, Trivedi B: Absence of serum growth hormone binding protein in patients with growth hormone receptor deficiency (Laron dwarfism). Proc Natl Acad Sci USA 1987; 84:4636-4640.

Hartman ML, Clayton PE, Johnson ML, et al: A low-dose euglycemic infusion of recombinant human insulin-like growth factor I rapidly suppresses fasting-enhanced pulsatile growth hormone secretion in humans. J Clin Invest 1993; 91:2453-2462.

Hartman ML, Veldhuis JD, Vance ML, et al: Somatotropin pulse frequency and basal concentrations are increased in acromegaly and are reduced by successful therapy. J Clin Endocrinol Metab 1990; 70:1375.

Tartaglia LA, Dembski M, Weng X, et al: Identification and expression cloning of a leptin receptor, OB-R. Cell 1995; 83:1263-1271.

Zhang Y, Proenca R, Maffie M, et al: Positional cloning of the mouse obese gene and its human homologue. Nature 1994; 372:425-432.