Basic and Clinical Endocrinology 7th International student edition Edition



Dennis Styne MD

Assessment of growth in stature is an essential part of the pediatric examination. Growth is an important index of physical and mental health and of the quality of the child's psychosocial environment; chronic problems in any of these areas may be reflected in a decreased growth rate. We shall consider influences on normal growth, the normal growth pattern, the measurement of growth, and conditions that lead to disorders of growth.



The growth of a fetus begins with a single fertilized cell and ends with differentiation into more than 200 cell types, length increasing by 5000-fold, surface area by 6 × 106-fold, and weight by 6 × 1012-fold. Overall, the growth of the fetus is dependent upon the availability of adequate oxygen and nutrition. It is orchestrated by a group of growth factors, all operating according to a basic genetic plan. This genetic plan is especially important early in gestation, whereas the maternal environment is of more importance late in gestation.

The classic definition of intrauterine growth retardation or intrauterine growth restriction (IUGR) is a birth weight below the fifth percentile or below 2500 g for a term baby in the United States, though some suggest that the designation small-for-gestational-age (SGA) be used for infants born weighing less than the tenth percentile and IUGR be reserved for those lower than the third percentile and/or those with demonstrated decrease of intrauterine growth rate. The term symmetric IUGR indicates that the head as well as the body is small, while asymmetric IUGR indicates that the head is relatively spared from the growth failure;


asymmetric IUGR is more common in the developed world, and symmetric IUGR is more common in the developing world. The former group shows “catch-up growth” more frequently than the latter, but 10–30% of IUGR infants remain short as children and adults, in contrast to appropriate-for-gestational-age premature infants, who are smaller at birth but generally experience catch-up growth in the first 2 years.


The placenta acts as an endocrine organ that influences most aspects of fetal growth, including the supply of adequate nutrition and oxygen and regulation of hormones and growth factors. Aberrant delivery or control of any of these factors will affect fetal growth; placental weight is usually directly related to birth weight.


The hormones that mediate postnatal growth do not play the same roles in fetal growth. Growth hormone (GH) is present in very high concentrations in the fetus, in contrast to the limited presence of GH receptors. While this discrepancy suggests limited activity of GH in the fetus, GH does play a role in fetal growth that is reflected in their average birth weight 1 SD below the mean that is found in growth hormone-deficient infants. Infants with Laron's syndrome (GH resistance due to reduced or absent GH receptors) have elevated GH and low serum IGF-I levels; they also have decreased birth length and weight. Thyroid hormone deficiency does not directly affect human birth weight, but prolonged gestation is a feature of congenital hypothyroidism, and this factor will itself increase weight. Placental lactogen exerts no effect on birth size in human beings. However, the concentration of placental growth hormone (from the GHV gene) is significantly decreased in the serum of a pregnant woman bearing a fetus with intrauterine growth retardation.


Oncogenes may be responsible for neoplastic growth in postnatal life, but expression of these genes is important in the normal development of many fetal organs. Remarkably, the same oncogenes that cause postnatal neoplasia are prevented from causing tumors in the normally differentiating fetus. For example, a mutation in the von Hippel-Lindau gene predisposes to retinal, cerebellar, and spinal hemangioblastomas, renal cell carcinomas, and pheochromocytomas, but the normal VHL gene is expressed in all three germ cell layers of the embryo and in the central nervous system, kidneys, testis, and lung of the fetus, suggesting a role in normal fetal development for this gene.


IGF-I in the fetus is regulated by metabolic factors rather than undergoing significant influence by GH, as is the case in postnatal life. One explanation is that there are fewer GH receptors in the fetus than after birth. In the human fetus, serum GH falls during later gestation owing to maturation of central nervous system negative control, while serum IGF-I and IGFBP-3 rise during gestation, demonstrating their independence from GH stimulation. Artificial elevation of maternal IGF-I can increase growth in mouse and rat fetuses and overcome the restraint imposed by uterine limitations of litter size.

Studies of knockout mice, which lack various growth factors or binding proteins, indicate a role for IGF-II in growth during early gestation and one for IGF-I during later gestation. Knockout of type 1 IGF receptors leads to a more profound growth failure than is found in IGF-I knockout mice alone, suggesting that factors other than IGF-I (eg, IGF-II) exert effects upon fetal growth through the type 1 receptor.

Study of transgenic mice overexpressing IGF-binding proteins supports the concepts that excess IGFBP-1 stunts fetal growth while excess IGFBP-3 leads to selective organomegaly. For example, overexpression of IGFBP-3 in mice led to organomegaly of the spleen, liver, and heart, though birth weight was not different from that of wild-type mice.

While controversy remains over some of the data regarding IGFs and fetal growth, a summary of the complex IGF system in the fetus, based upon the evidence from various species, appears to apply to the human being: (1) IGFs are detectable in many fetal tissues from the first trimester onward; (2) levels of IGFs in the fetal circulation increase during pregnancy, and at term the levels of IGF-I are directly related to birth weight; (3) in mice, disruption of the IGF gene leads to severe growth retardation; (4) at the end of the first trimester, there is a striking increase in IGFBP-1 and IGFBP-2 levels in amniotic fluid; (5) the major binding proteins in the human fetus are IGFBP-1 and IGFBP-2; (6) from as early as 16 weeks, there is an inverse correlation between fetal levels of IGFBP-1 and birth weight; (7) in the mother, circulating levels of IGF-I and IGFBP-1 increase during pregnancy; (8) maternal levels of IGFBP-1 are elevated in severe preeclampsia and intrauterine growth retardation; and (9) fetal levels of IGFBP-1 are elevated in cases of intrauterine growth retardation, especially those associated with specific evidence


of reduced uteroplacental blood flow. Levels of IGFBP-1 appear to be a sensitive indicator of the short- or long-term response to reduced fetal nutrition.


While insulin is a major regulatory factor for carbohydrate metabolism, many lines of evidence demonstrate its importance in fetal growth as well. Macrosomia is a well known effect of fetal hyperinsulinism such as found in the infant of the diabetic mother. Errors in the normal pattern of IGF-II gene expression from the paternal chromosome and type 2 IGF receptor for IGF-II from the maternally derived gene may underlie the pathogenesis of Beckwith-Wiedemann syndrome. Affected infants are large and have elevated insulin concentrations. At the other end of the spectrum is the small-for-gestational-age (SGA) infant born to a diabetic mother with vascular disease, or under extremely tight control, or who has eclampsia or preeclampsia; this demonstrates that limited nutrient delivery compromises the growth of the infant.

Just as increased insulin stimulates fetal growth, syndromes of fetal insulin deficiency such as are found in congenital diabetes mellitus, pancreatic dysgenesis, or fetal insulin resistance (eg, leprechaunism) are characterized by intrauterine growth retardation.


Epidermal growth factor (EGF) is involved with fetal growth, and expression varies with disordered fetal growth. Microvilli purified from the placentas of infants with intrauterine growth retardation (IUGR) have decreased or absent placental epidermal growth factor receptor (EGF-R) phosphorylation and tyrosine kinase activity. Maternal smoking decreases birth weight by an average of 200 g, with the major effect occurring late in pregnancy; the placenta responds to smoking by significant changes in its vascularity, which leads to fetal hypoxia. There are decreased numbers of EGF-Rs and a reduced affinity of these receptors for EGF in the placentas of women who are smokers. Hypertensive patients also have decreased numbers of placental EGF-Rs, which may result in IUGR.

EGF levels in amniotic fluid are normally increased near term but decreased in pregnancies complicated by IUGR—though not, conversely, increased in infants who are large for gestational age. EGF levels in the first urines to be voided by IUGR and macrosomic infants are lower than in control infants.

EGF administered to fetal monkeys results in histologic and biochemical maturation of their lungs, leading to improved air exchange and a diminished requirement for respiratory support. Surfactant apoprotein A concentration and the lecithin:sphingomyelin ratio are both significantly higher in the amniotic fluid of the EGF-treated fetuses. Whereas birth weight is not affected by EGF, adrenal and gut weights, standardized for body weight, are increased significantly. Furthermore, EGF stimulates gut muscle, enzyme maturation, and gut size and content, improving the ability of the infant to absorb nutrients. Lastly, EGF advances the maturation of the fetal adrenal cortex, increasing the expression of 3β-hydroxysteroid dehydrogenase. Since EGF can be absorbed orally, this raises a question about whether EGF could be a useful treatment for premature infants or, postnatally, for causing more rapid maturation of the neonate and improving survival in premature infants.

Fibroblast Growth Factor

Genetically engineered fibroblast growth factor receptor (FGF-R)-deficient mice are severely growth-retarded and die before gastrulation. Aberrant FGF signaling during limb and skeletal development in the human being can lead to dysmorphic syndromes. For example, achondroplasia is due to mutations in the transmembrane domain of the type 3 fibroblast growth factor receptor.

Genetic, Maternal, & Uterine Factors

Maternal factors, often expressed through the uterine environment, exert more influence on birth size than paternal factors. The height of the mother correlates better with fetal size than the height of the father. However, there is a genetic component to length at birth that is not sex specific. Firstborn infants are on the average 100 g heavier than subsequent infants; maternal age over 38 years leads to decreased birth weight; and male infants are heavier than female infants by an average of 150–200 g. Poor maternal nutrition is the most important condition leading to low birth weight and length on a worldwide basis. Chronic maternal disease and eclampsia can also lead to poor fetal growth. Maternal alcohol ingestion has severe adverse effects on fetal length and mental development and predisposes to other physical abnormalities seen in the fetal alcohol syndrome such as microcephaly, mental retardation, midfacial hypoplasia, short palpebral fissures, wide-bridged nose, long philtrum, and narrow vermilion border of the lips; affected infants never recover from this loss of length but attain normal growth rate in the postnatal period. Abuse of other substances and chronic use of some medications (eg, phenytoin) can cause intrauterine growth retardation. Cigarette smoking causes not only retarded intrauterine growth but also decreased postnatal growth for as long as 5 years after parturition. Maternal infection—most commonly toxoplasmosis, rubella, cytomegalovirus infection, herpes


simplex infection, and HIV infection—leads to many developmental abnormalities as well as short birth length. In multiple births, the weight of each fetus is usually less than that of the average singleton. Uterine tumors or malformations may decrease fetal growth.

Chromosomal Abnormalities Malformation Syndromes

Many chromosomal abnormalities that lead to malformation syndromes also cause poor fetal growth. Other malformation syndromes associated with a normal karyotype are characterized by intrauterine growth retardation. In most cases, endocrine abnormalities have not been noted. For further discussion of this extensive subject, the reader is referred to other sources listed in the references at the end of this chapter.


Many lines of evidence demonstrate long-lasting effects of abnormalities in fetal growth. Inanition during the last two trimesters of pregnancy, which occurred during the famine in Holland during World War II, led to an 8–9% decrease in birth weight; however, female infants born under these conditions later gave birth to normal-sized infants. On the other hand, in the Dutch and Leningrad famines, babies born after early gestational starvation of their mothers but with improved maternal nutrition in late gestation were of normal size at birth. However, the female infants born of normal size after this early gestational maternal starvation themselves gave birth to small babies (IUGR of 300–500 g decrease). In other populations, women with a history of IUGR tend to have IUGR babies themselves, and some studies show that generations of malnutrition must be followed by generations of normal nutrition before there is correction of the birth weight of subsequent babies.

Numerous studies from around the world indicate a relationship between low birth weight or low weight at 1 year of age and chronic disease in adulthood. Birth weight—not prematurity—is inversely related to cardiovascular mortality and the insulin resistance syndrome (syndrome X), which consists of (1) elevation of systolic and diastolic blood pressure, (2) impaired glucose tolerance, and (3) elevated triglycerides among other features. The individuals most affected were those with the largest placentas but smaller birth weights. This is attributed to fetal and neonatal “metabolic programming,” in which early adjustments to enhance survival in difficult intrauterine circumstances set the stage for later disorders.

Insulin resistance, which might be the basis for most or all of these complications, may spare nutrients from utilization in muscle, thus leaving them available for the brain, a mechanism which would serve to minimize central nervous system damage in the fetus during periods of malnutrition. Studies of otherwise normal thin children who had a history of IUGR demonstrated insulin resistance before the teenage years, suggesting early metabolic programming. Present-day adults who were born in the Netherlands during the Dutch famine and had the lowest birth weights and the lowest maternal weights (those subjects whose mothers experienced malnutrition during the last two trimesters) show a degree of insulin resistance which is directly related to their degree of IUGR, further documenting the relationship between fetal undernutrition and adult insulin resistance.

On the other hand, large babies born to mothers with diabetes mellitus develop childhood obesity after a period of normal weight between 1 year and 5 years of age. Remarkably, studies of the offspring of Dutch mothers exposed to famine during World War II in the first two trimesters (the time in which birth weight is least affected by maternal starvation) demonstrated a twofold increase in the incidence of obesity at 18 years of age compared with a 40% decrease of the incidence of obesity if the individual was exposed to famine in the last trimester (the time in which birth weight is most negatively impacted by maternal starvation).


Postnatal growth in stature and weight follows a characteristic pattern in normal children (Figures 6-1 and 6-2). The highest overall growth rate occurs in the fetus, the highest postnatal growth rate just after birth, and a slower growth rate follows in mid childhood (Figures 6-3and 6-4). There are two periods characterized by brief growth spurts in childhood: the infant-childhood growth spurt between 1˝ years and 3 years and the mid childhood growth spurt between 6˝ years and 7 years. In addition, there is an “adiposity rebound” of accelerating weight gain and rising BMI in mid childhood after a previous period of relative stability of weight gain. An early adiposity rebound is a risk factor for the development of obesity later in childhood and thereafter.

After another plateau, the striking increase in stature known as the pubertal growth spurt follows, causing a second peak of growth velocity. The final decrease in growth rate then ensues, until the epiphyses of the long bones fuse and growth ceases.

Endocrine Factors


As discussed in Chapter 5, somatotropin, or growth hormone (GH), is suppressed by hypothalamic growth hormone release-inhibiting factor (or somatostatin) and










stimulated by growth hormone-releasing hormone (GHRH). The gene for GH is located on the long arm of chromosome 17 in a cluster of five genes: GHN codes for human GH (a single 191-amino-acid polypeptide chain with a molecular weight of 22 kDa); GHV codes for a variant GH produced in the placenta; CSH1 and CSH2 code for prolactin; and CSHP1 codes for a variant prolactin molecule. A 20-kDa variant of pituitary growth hormone accounts for 5–10% of circulating GH.


Figure 6-1. A. Chart appropriate for boys in most clinical situations showing the third to the ninth percentiles. (Redrawn from the original developed by the National Center for Health Statistics in collaboration with the National Center for Chronic Disease Prevention and Health Promotion, 2000.) B. Growth chart appropriate for boys in the extremes of growth showing standard deviations from the mean. (Redrawn and reprinted with permission of Genentech, Inc. Sources of data: 1976 study of the National Center for Health Statistics [NCHS; Hyattsville, MD]; Hamill PVV et al: Physical growth: National Center for Health Statistics percentiles. Am J Clin Nutr 1979;32:607.)


Figure 6-2. A. Growth chart appropriate for girls in most clinical situations showing the third to ninth percentiles. (Redrawn from the original developed by the National Center for Health Statistics in collaboration with the National Center for Chronic Disease Prevention and Health Promotion, 2000.) B. Growth chart appropriate for girls in the extremes of growth, showing standard deviations from the mean. (Redrawn and reprinted with permission of Genentech, Inc. Sources of data: 1976 study of the National Center for Health Statistics [NCHS; Hyattsville, MD]; Hamill PVV et al: Physical growth: National Center for Health Statistics percentiles. Am J Clin Nutr 1979;32:607.)


Figures 6-3 and 6-4. Incremental growth charts for boys (Figure 6-3) and girls (Figure 6-4). Height velocity can be compared with the percentiles on the right axis of the charts. (Redrawn with permission from Genetech, Inc.)

The effects of growth hormone are mainly mediated by the insulin-like growth factors, but GH also has direct effects such as lipolysis, increased amino acid transport into tissues, and increased protein synthesis in liver. Growth hormone produces insulin resistance and is a diabetogenic substance, increasing blood sugar. GH is secreted in a pulsatile manner, so that serum concentrations are low much of the day but peak during short intervals. Values are higher in the immediate neonatal period, decrease through childhood, and rise again as a result of increased pulse amplitude (but not frequency) during puberty. GH secretion falls again during aging. GH secretion is decreased significantly in obesity but rises in states of starvation.

Growth hormone circulates in plasma bound to a protein, the growth hormone-binding protein (GHBP), with a sequence equivalent to that of the extracellular membrane domain of the growth hormone receptor. The physiology of the GH-binding protein appears to be of great importance in growth. For example, obese patients have lower plasma GH concentrations but higher GHBP levels, while starvation raises GH concentrations and lowers GHBP levels. Patients with abnormalities of the GH receptor (eg, Laron dwarfism) also have the defect reflected in the serum GHBP concentrations; those with decreased numbers of GH receptors have decreased serum GHBP concentrations.

GH exerts its effects on growth mainly through the insulin-like growth factors (IGFs) and their binding proteins. IGF-I and IGF-II have structures similar to that of the proinsulin molecule but differ from insulin in regulation, receptors, and biologic effects. The structure of the insulin-like factors (originally called sulfation factor and then somatomedin), the genes responsible for their production, and information about their physiology have been elucidated. Recombinant IGFs are available for clinical studies.

The single copy gene for prepro-IGF-I is located on the long arm of chromosome 12. Posttranslational processing produces the 70-amino-acid mature form, and alternative splicing mechanisms produce variants of the molecule in different tissues and developmental stages. The IGF-I cell membrane receptor (the type I receptor) resembles the insulin receptor in its structure of two α and two β chains. Binding of IGF-I to type I receptors stimulates tyrosine kinase activity and autophosphorylation of tyrosine residues in the receptor. This leads to cell differentiation or division (or both). IGF-I receptors are down-regulated by increased IGF-I concentrations, while decreased IGF-I concentrations increase IGF-I receptors.

IGF molecules in the circulation are mostly bound to a variety of IGF-binding proteins (IGFBPs); at present, information is available about the molecular weights and other properties of six IGFBPs. IGFBP-1 and IGFBP-3 have been extensively studied. IGFBP-1 is a 25-kDa protein. The serum levels of IGFBP-1 are inversely proportionate to insulin levels; this protein does not appear to be regulated by GH. IFGBP-1 is mainly inhibitory of IGF action. It is present in high concentrations in fetal serum and amnionic fluid. Serum values in blood are inversely proportionate to birth weight.

IGF-I circulates bound to IGFBP-3 and an acid labile subunit in a 150-kDa complex. Serum IGFBP-3 concentrations are directly proportionate to GH concentrations but also to nutritional status—in malnutrition, IGFBP-3 and IGF-I levels fall while GH rises. IGF-I directly regulates IGFBP-3 as well. IGFBP-3 rises with advancing age through childhood, with highest values achieved during puberty; however, the pattern of change of IGF-I at puberty is different from that of IGFBP-3. The molar ratio of IGF-I to IGFBP-3 rises at puberty, suggesting that more IGF-I is free to influence growth during this period.

IGF-I is produced in most tissues and appears to be exported to neighboring cells to act upon them in a paracrine manner or upon the cell of origin in an autocrine manner. Thus, serum IGF-I concentrations may not reflect the most significant actions of this growth factor. The liver is a major site of IGF-I synthesis, and much of the circulating IGF-I probably originates in the liver; serum IGF-I concentrations vary in liver disease with the extent of liver destruction. IGF-I is a progression factor, so that a cell which has been exposed to a competence factor such as platelet-derived growth factor (PDGF) in stage G0 of the cell cycle and has progressed to G1 can, with IGF-I exposure in G1, undergo division in the S phase of the cell cycle. Aside from the stimulatory effects of IGF-I on cartilage growth, IGF-I has stimulatory effects upon hematopoiesis, ovarian steroidogenesis, myoblast proliferation and differentiation, and differentiation of the lens.

IGF-I was in short supply until production by recombinant DNA technology became possible. IGF-I administration in clinical trials increases nitrogen retention and decreases BUN—and, in GH-resistant patients (Laron dwarfs), IGF-I stimulates growth without the presence of GH. Thus, IGF-I may prove useful in


treatment of various clinical conditions from pathologic short stature to catabolic states, including the postoperative period and burns.

IGF-II is a 67-amino-acid peptide. The gene for prepro-IGF-II is located on the short arm of chromosome 11, close to the gene for preproinsulin. The type II IGF receptor preferentially binds IGF-II and is identical to the mannose 6-phosphate receptor, a single-chain transmembrane protein. While most of the effects of IGF-II appear mediated by its interaction with the type I receptor, independent actions of IGF-II via the type II receptor are described.

Plasma concentrations of the IGFs vary with age and physiologic condition. IGF-I concentrations are low at term in neonates and remain relatively low in childhood until a peak is reached during puberty, with values rising higher than at any other time in life. Serum IGF-I then decreases to adult levels, values higher than in childhood but lower than in puberty. With advancing age, serum GH and IGF-I decrease. IGF-I concentrations are more highly correlated in monozygotic twins than in same-sex dizygotic twins, indicating a genetic effect upon IGF-I regulation.

GH deficiency leads to lower serum IGF-I and IGF-II concentrations, while GH excess leads to elevated IGF-I but no rise in IGF-II above normal. Because serum IGF-I is lower during states of nutritional deficiency, IGF-I is not a perfect tool in the differential diagnosis of conditions of poor growth, which often include impaired nutritional state. IGF-I suppresses GH secretion, so that patients who lack GH receptors (Laron dwarfs) and are unable to produce IGF-I have elevated GH concentrations but negligible IGF-I concentrations.


As noted above, congenital hypothyroid newborns are of normal length, but if untreated they manifest exceedingly poor growth soon after birth. Infants with untreated congenital hypothyroidism will suffer permanent mental retardation. Acquired hypothyroidism leads to a markedly decreased growth rate but no permanent intellectual defects. Bone age advancement is severely delayed in hypothyroidism, usually more so than in GH deficiency, and epiphysial dysgenesis is seen when calcification of the epiphyses progresses. The normal decrease in the upper to lower segment ratio with age (Figure 6-5) is delayed and therefore elevated, owing to poor limb growth in hypothyroidism.


Gonadal sex steroids exert an important influence on the pubertal growth spurt, while absence of these factors is not of major importance in prepubertal growth. Gonadal and adrenal sex steroids in excess can cause a sharp increase in growth rate as well as the premature appearance and progression of secondary sexual features. If unabated, increased sex steroids will cause advancement of skeletal age, premature epiphysial fusion, and short adult stature.


Figure 6-5. Normal upper to lower segment (US:LS) ratios, based on findings in 1015 white children. Values are slightly lower for black children. (Reproduced, with permission, from McKusick V: Hereditable Disorders of Connective Tissue, 4th ed. Mosby, 1972.)

The pubertal rise in gonadal steroids exerts direct and indirect effects upon IGF-I production. Sex steroids directly stimulate the production of IGF-I from cartilage. They also increase GH secretion, which stimulates IGF-I production indirectly. Both actions appear important in the pubertal growth spurt (see Chapter 15).


Endogenous or exogenous glucocorticoids in excess will quickly stop growth; this effect occurs more quickly than weight gain. The absence of glucocorticoids has little effect on growth if the individual is clinically well in other respects (ie, if hypotension and hypoglycemia are absent).



Other Factors


Genetic factors influence final height. Correlation is found between midparental height and the child's height; appropriate methods to utilize this phenomenon and determine the target height for a child are presented in Figure 6-6. There is a heritable pattern to birth length, postnatal increase in length, and the intrinsic rate of change in growth. These effects are sex-specific.


Worldwide, the most common cause of short stature is poverty and its effects. Thus, poor nutrition, poor hygiene, and poor health influence growth both before and after birth. In people of the same ethnic group and in the same geographic location, variations in stature are often attributable to these factors. For example, Japanese individuals born and reared in North America after World War II are generally taller than Japanese-born immigrants to North America. Conversely, when socioeconomic factors are equal, the differences in average height between various ethnic groups are mainly genetic. The new growth charts for children in the United States from the CDC are not ethnicity-specific because it is believed that the major differences between growth in ethnic groups are due to socioeconomic status and nutrition rather than genetic endowments.


The influence of malnutrition accounts for much of the socioeconomic discrepancy in height noted above, but malnutrition may occur in the midst of plenty and must always be suspected in disorders of growth. Other factors may be blamed for poor growth when nutritional deficiencies are actually responsible. For example, Sherpas were thought to have short stature mainly because of genetic factors or the effects of great altitude on the slopes of Mount Everest; nutritional supplementation increased stature in this group, however, demonstrating the effects of adequate nutrition. The developed world places a premium on appearance, and women portrayed as beautiful in the media are characteristically thin. Significant numbers of children, chiefly teenagers, voluntarily decrease their caloric intake even if they are not obese; this accounts for some cases of poor growth. Chronic disease, which hampers adequate nutrition, often leads to short stature. For example, bronchopulmonary dysplasia decreases growth to some degree because it increases metabolic demands, shifting nutrient usage from growth; improved nutrition will increase growth in these patients. Feeding problems in infants, resulting from inexperience of parents or poor child-parent interactions (maternal deprivation), may account for poor growth. Fad diets such as poorly constructed vegan diets that put children at risk for vitamin B12 or iron deficiency as well as dietary manipulation for expected benefit, such a low-fat diet, may place children at risk for deficiency of fat-soluble vitamins. Deliberate starvation of children by caregivers is an extreme form of child abuse that may be first discovered because of poor growth.

There are significant endocrine changes associated with malnutrition. Decreased GH receptors or postreceptor defects in GH action, leading to decreased production of IGF-I and decreased concentration of serum IGF-I are notable. The characteristic results of malnutrition are elevation of serum GH and decrease in IGF-I. Remarkably, obesity increases IGF-I concentrations by increasing GH receptors even though GH secretion is suppressed to levels suggesting GH deficiency. IGFBP-1, a suppressor of IGF-I effects, is elevated in obesity.


Aberrant intrafamilial dynamics, psychologic stress, or psychiatric disease can inhibit growth either by altering endocrine function or by secondary effects on nutrition (psychosocial dwarfism or maternal deprivation). It is essential to differentiate these situations from true disease states.


Even aside from the effects of poor nutrition, many chronic systemic diseases interfere with growth. For example, congestive heart failure and asthma, if uncontrolled, are associated with decreased stature; in some cases, final height is in the normal range because growth continues over a longer period of time. Children of mothers with HIV infection are often small at birth and have an increased incidence of poor postnatal growth, delayed bone age development, and reduced IGF-I concentrations; in addition, thyroid dysfunction may develop, further complicating the growth pattern. Infants born of HIV-infected mothers but themselves not infected may exhibit catch-up growth.

Catch-Up Growth

Correction of growth-retarding disorders may be temporarily followed by an abnormally high growth rate as the child approaches normal height for age. This catch-up growth will occur after initiation of therapy for hypothyroidism and GH deficiency, after correction of glucocorticoid excess, and after appropriate treatment of many chronic diseases such as celiac disease. Catch-up growth is usually short-lived and is followed by a more average growth rate.




Accurate measurement of height is an essential part of the physical examination of children and adolescents. The onset of a chronic disease may often be determined by an inflection point in the growth chart. In other cases, a detailed growth chart will indicate a normal constant growth rate in a child observed to be short for age. If careful growth records are kept, a diagnosis of constitutional delay in growth and adolescence or genetic short stature may be made in such a patient; without previous measurements, the child might be subjected to unnecessary diagnostic testing or months of delay may occur as the child's growth is finally carefully monitored.


The National Center for Health Statistics revised the growth charts for children in the United States (Figures 6-1 and 6-2). The new charts display the third and 97th percentiles rather than the fifth and 95th percentiles, and standard deviations are also available. Charts displaying BMI by age contain data appropriate for the evaluation of obesity and underweight.

But growth charts still leave 6 out of 100 healthy children outside of their boundaries, with 3 out of 100 below the more worrisome (to parents) “lower limits of normal.” It is both unnecessary and impractical to evaluate 3% of the population. Instead, the examining physician should determine which short children warrant further evaluation and which ones (and their parents) require only reassurance that the child is healthy. When parents see that their child is below the third percentile and in a section of the chart colored differently from the “normal area,” they assume that there is a serious problem. Thus, the format of the chart can dictate parental reaction to height, since all parents want their children to be in the “normal range.” Figures 6-1 and 6-2 furnish data necessary to evaluate the height of children at various ages using percentiles or the standard deviation method (SD) used by the WHO. Standard deviation determination is more useful in extremely short children below the second or first percentile.

Pathologic short stature is usually more than 3.5 SD below the mean, whereas the third percentile is only at 2 SD below the mean. However, a diagnosis of pathologic short stature is best not based on a single measurement. Serial measurements are required because they allow determination of growth velocity, which is a more sensitive index of the growth process than a single determination. A very tall child who develops a postnatal growth problem will not fall 3.5 SD below the mean in height for some time but will fall below the mean in growth velocity soon after the onset of the disorder. As Figures 6-3 and 6-4 demonstrate, growth velocity varies at different ages, but as a rough guide, a growth rate of less than 5 cm per year between age 4 years and the onset of puberty is abnormal. In children under 4 years of age, normal growth velocity changes more strikingly with age. Healthy term newborns tend to be clustered in length measurements around 21 inches (mostly owing to difficulties in obtaining accurate measurements). In the ensuing 24 months or so, the healthy child's height will enter a channel on the growth chart and remain there throughout childhood. Thus, a child with constitutional delay in growth or genetic short stature whose height is at the mean at birth and gradually falls to the tenth percentile at 1 year of age and to the fifth percentile by 2 years of age may in fact be healthy in spite of crossing percentile lines in the journey to a growth channel at the fifth percentile. Although the growth rate may decrease during these years, it should not be less than the fifth percentile for age. A steeper decrease in growth rate may be a sign of disease. When a question of abnormal growth arises, previous measurements are always helpful; every physician treating children should record supine length (under 2 years of age) or standing height (after 2 years of age) as well as weight at every office visit. As the child leaves infancy, height and growth velocity should be determined in relation to the standards for the child's age on a graphic chart with clear indication of the child's position (supine or standing) at measurement.

Patients who cannot be measured in the standing position (eg, because of cerebral palsy) require other approaches: The use of arm span is a possible surrogate for the measurement of height, and there are formulas available for the calculation of height based upon the measurement of upper arm length, tibial length, and knee length (see below).

This discussion presupposes accuracy of measurements, which should be available for children followed by a single physician or group. However, it is reported that screening examinations in the real world fall short of that ideal. Forty-one percent of a presumably normal population screened at a school in England met the criteria for evaluation of abnormal growth (approximately two-thirds grew faster than the normal growth category and one-third were in the slower than normal category), leading to an unreasonable size of a referral population, all due to simple measuring error.

Relation to Midparental Height: The Target Height

There is a positive correlation between midparental height (the average of the heights of both parents) and the stature of a child. One way to use this relationship




of parents' heights to the expected heights of children within a given family is the calculation of target adult height range using parents' heights and correcting these heights for the sex of the child. For boys, add 5 inches to the mother's height, add the result to the father's height, and divide by 2. This is the target height, and it is expected that sons of these parents will reach a height within 2 SD of this target—or, for simplicity, within 4 inches above and 4 inches below the target height. (Two inches approximates 1 SD for adults.) For girls, subtract 5 inches from the father's height and add the result to the mother's height and divide by 2, leading to the target height for the girl. The range for girls will also be within 4 inches above and below this target. In effect, this corrects the North American growth charts for the particular family being considered. The calculated target height corresponds to the 50th percentile for the family, and the limits of the ą 2 SD approximate the fifth to 95th percentile for the family (see Figures 6-1 and 6-2). This method is useful only in the absence of disease affecting growth, and the prediction is more valid when the parents are of similar rather than of widely different heights. Figures 6-6 and 6-7 demonstrate the calculation of target height and the ranges. When there is a large discrepancy between the heights of the mother and the father, prediction of final height becomes difficult. A child may follow the growth pattern of the shorter parent more closely than the midparental height. Furthermore, a boy may, for example, follow the growth of a short mother rather than a taller father.


Figure 6-6. Determination of target height in a shorter family. This 10-year-old boy is 124 cm tall (we measure patients in cm), his mother is 61 inches tall (adults recall their heights in inches and feet but, if they are available, their height should be actually measured in centimeters), and the father is 63 inches tall. Five inches are added to the mother's height to convert her height percentile to the equivalent percentile on a boy's chart. (If we were considering a daughter whose height is plotted on a girl's chart, the mother's height would be directly plotted and 5 inches would be subtracted from the father's height to correct his height percentile to the equivalent for an adult

A parent who spent the growing years in poverty, with chronic disease, or in an area of political unrest might have a falsely lowered adult height. Of course, the height of an adopted child will have no relationship to the adoptive parents' heights.

Technique of Measurement

Length and height must be measured accurately. Hasty measurements derived from marks made on paper at an infant's foot and head while the infant is squirming on the paper on the examining table are useless. Infants must be measured on a firm horizontal surface with a permanently attached rule, a stationary plate perpendicular to the rule for the head, and a movable perpendicular plate for the feet. One person should hold the head stable while another makes sure the knees are straight and the feet are firm against the movable plate. Children over age 2 are measured standing up. These measurements cannot be accurately performed with the measuring rod that projects above the common weight scale; the rod is too flexible, and the scale footplate will in fact drop lower when the patient stands on it. Instead, height should be measured with the child standing back to the wall with heels at the wall, ankles together, and knees and spine straight against a vertical metal rule permanently attached to the wall or to a wide upright board. Height is measured at the top of the head by a sliding perpendicular plate (or square wooden block). A Harpenden stadiometer is a mechanical measuring device capable of such accurate measurement. Standing height is on the average 1.25 cm less than supine length, and it is essential to record the position of measurement each time; shifting from supine height at 2 years to standing height at 2˝ years can falsely suggest an inadequate growth rate over that 6-month period. It is preferable to measure in the metric system, since the smaller gradations make measurements more accurate by eliminating the tendency to round off numbers. Growth is not constant but is characterized by short spurts and periods of slowed growth. The interval between growth measurements should be adequate to allow an accurate evaluation of growth velocity. Appropriate sampling intervals vary with age but should not be less than 3 months in childhood, with a 6-month interval being optimal.

The problem of measuring the growth rate of children with orthopedic deformities or contractures is significant, as these patients may have nutritional or endocrine disorders as well. The measurement of knee height, tibial length, or upper arm length correlates well with standing height (r = .97); thus, these measurements may be translated, using special linear regression




equations, into total height, which is then plotted on standard growth charts. Specialized laser-calibrated devices to measure tibial length (“kneeometry”) are reported to be accurate for assessment of short-term growth down to weekly intervals.


Figure 6-7. Determination of target height in a taller family. This 10-year-old boy is 124 cm tall (we measure patients in centimeters), his mother is 65 inches tall (adults recall their heights in inches and feet but, if they are available, their height should be actually measured in centimeters), and the father is 73.5 inches tall. Five inches are added to the mother's height to convert her height percentile to the equivalent percentile on a boy's chart. (If we were considering a daughter whose height is plotted on a girl's chart, the mother's height would be directly plotted and 5 inches should be subtracted from the father's height to correct his height percentile to the equivalent height percentile for an adult woman.) Her corrected height and that of the father are plotted at the far right of the chart where adult heights are displayed. The midparental height is calculated by adding the father's height to the corrected mother's height and the sum is divided by 2; the result is the target height. The limits of 2 SD above and below the target height are displayed by plotting 2 SD (which equals approximately 4 inches) above and below the target height. This is equivalent to moving the 50th percentile for the United States population to a conceptual 50th percentile for the family under consideration. It is evident that the height of the child, which is below the third percentile for the United States, is even farther outside the bounds of the percentiles described by ą SD from the target height, and, thus, the child appears to fall far outside the genetic pattern of the family. The growth velocity and the degree of skeletal maturation are some of the other factors necessary to evaluate this analysis in more detail.

In addition to height or length, other significant measurements include (1) the frontal-occipital head circumference; (2) horizontal arm span (between the outspread middle fingertips with the patient standing against a flat backboard); and (3) the upper segment (US) to lower segment (LS) ratio. For the latter, the LS is measured from the top of the symphysis pubis vertically to the floor with the patient standing straight, and the US is determined by subtracting the LS from the standing height measurement noted above. (Normal standard US:LS ratios are shown in Figure 6-5.) Sitting height is used in some clinical studies of growth, but the sitting stadiometer is rarely available.

Height & Growth Rate Summary

In summary, we may consider three criteria for pathologic short stature: (1) height more than 3.5 SD below the mean for chronologic age; (2) growth rate more than 2 SD below the mean for chronologic age; and (3) height more than 2 SD below the target height when corrected for midparental height.


The measured weight should be plotted for age on standard graphs developed by the National Center for Health Statistics (NCHS), which are available from various companies producing baby food or growth hormone. Variation in the weights of children in the USA due to differing diets or activity regimens makes it difficult to exactly compare percentiles of height with percentiles of weight. Body mass index (BMI) charts displaying percentiles of BMI (weight in kilograms divided by height in meters squared) are widely available and provide an excellent way to assess nutritional status.


Skeletal development is a reflection of physiologic maturation. For example, menarche is better correlated with a bone age of 13 years than with a given chronologic age. Estrogen plays the major role in advancing skeletal maturation. Patients with aromatase deficiency, who cannot convert estrogen from testosterone, and patients with estrogen receptor defects, who cannot respond to estrogen, grow taller well into their twenties without having epiphysial fusion. Bone age indicates remaining growth available to a child and can be used to predict adult height. However, bone age is not a definitive diagnostic test of any disease; it can assist in diagnosis only when considered along with other factors.

Bone age is determined by comparing the appearance and stage of fusion of epiphyses or shapes of bones on the patient's radiograph with an atlas demonstrating normal skeletal maturation for various ages. The Greulich and Pyle atlas of radiographs of the left hand and wrist is most commonly used in the USA, but other methods of skeletal age determination such as Tanner and Whitehouse maturity scoring are preferred in Europe. Any bone age more than 2 SD above or below the mean for chronologic age is out of the normal range. For newborn infants, knee and foot radiographs are compared with an appropriate bone age atlas; for late pubertal children, just before epiphysial fusion, the knee atlas will reveal whether any further growth can be expected or whether the epiphyses are fused.

Height is predicted by determining bone age and height at the time the radiograph was taken and consulting the Bayley-Pinneau tables in the Greulich and Pyle skeletal atlas of the hand. The Roche, Wainer, and Thissen (RWT) method of height prediction uses patient weight and midparental height—in addition to the variables noted above—to calculate predicted height (Table 6-1). Recently, the method was simplified by eliminating the necessity for a bone age assessment in the Khamis-Roche method; results are almost as accurate as the RWT method in white American children. Height prediction by any method becomes more accurate as the child approaches the time of epiphysial fusion.

Table 6-1. The RWT method for predicting adult stature.1

Tables of Multipliers for Boys


Tables of Multipliers for Girls


Recumbent length


Midparental stature

Skeletal age

Adjustment factor


Recumbent length


Midparental stature

Skeletal age

Adjustment factor



















































































































































































































































































































































































































































































































































































































































































































































































































































The RWT method predicts the height of an individual at 18 years of age; after this age the average total increase in statuture is 0.6 cm for girls and 0.8 cm for boys.


























































The RWT method predicts the height of an individual at 18 years of age; after thisage the average age total increase in stature is 0.6 cm for girls and 0.8 cm for boys.
 Recumbent length is measured in cm (add 1.25 cm to the standing height, without shoes, if that is available). Weight is measured in kg. The midparental height is calculated by adding the standing height of each parent in cm (without shoes) and dividing by two; if the parents' heights are unknown, in the USA a height of 174.5 cm can be substituted for the father's height or 162 cm for the mother's height. The skeletal age is determined from an x-ray of the left wrist and hand comparing it to the Greulich and Pyle atlas.
   A prediction is made by:

1. Recording the child's data as noted below.

2. Finding the multipliers from the charts on these pages, making sure the positive and negative signs are retained for the calculations.

3. Multiplying the data by the multipliers, taking note of the positive or negative sign.

4. Adding the products to the adjustment factor, taking note of the sign of the factor; the result is a prediction of the height at 18 years of age.

1Modified and reproduced, with permission, from Roche AF, Wainer H, Thissen D: The RWT method for the prediction of adult stature. Pediatrics 1975;56:1026, as modified in Styne DM: Growth Disorders. Page 99 in: Handbook of Clinical Endocrinology. Fitzgerald PA (editor). Jones Medical Publications, 1986.










There are many causes of decreased childhood growth and short adult height (Table 6-2). The following discussion covers only the more common conditions, emphasizing those that might be included in an endocrine differential diagnosis. Shorter than average stature need not be considered a disease, since variation in stature is a normal feature of human beings, and a normal child should not be burdened with a misdiagnosis. While the classifications described below may apply to most patients, some will still be resistant to definitive diagnosis.

  1. Constitutional Short Stature

Constitutional short stature (constitutional delay in growth and adolescence) is not a disease but rather a variation from normal for the population and is considered a slowing of the pace of development. There is usually a delay in pubertal development as well as a decrease in growth (see Constitutional Delay in Adolescence in Chapter 15). It is characterized by moderate short stature (usually not far below the third percentile), thin habitus, and retardation of bone age. The family history often includes similarly affected members (eg, mother with delayed menarche or father who shaved late and kept growing past his teen years).

All other causes of decreased growth must be considered and ruled out before this diagnosis can be made with confidence. The patient may be considered physiologically (but not mentally) delayed in development. Characteristic growth patterns include normal birth length and height, with a gradual decrease in percentiles of height for age by 2 years; on the contrary, a rapid decrease in percentiles is an ominous sign of pathology. Onset of puberty is usually delayed for chronologic age but normal for skeletal age. Adult height is in the normal range but varies according to parental heights. The final height is often less than the predicted height, because growth is less than expected during puberty.

  1. Genetic Short Stature

Short stature may also occur in a familial pattern without retarded bone age or delay in puberty; this is considered “genetic”short stature. Affected children are closer to the mean on the normal population growth charts after correction for midparental height by calculation of the target height (Figures 6-6 and 6-7). Adult height depends on the mother's and father's heights. Patients with the combination of constitutional short stature and genetic short stature are quite noticeably short due to both factors and are the patients most likely to seek evaluation. Boys are brought to consultation more often than girls.

  1. Prematurity & Intrauterine Growth Retardation

While the majority show catch-up growth, 30% of IUGR infants may follow a lifelong pattern of short stature. Symmetric IUGR infants are most likely to demonstrate this finding. In comparison, appropriate-for-gestational-age premature infants will usually catch up to the normal range of height and weight by 1–2 years of age. Severe premature infants with birth weights less than 800 g (that are appropriate for gestational age), however, may maintain their growth retardation at least through their third year; only follow-up studies will determine whether this classification of premature infants reaches reduced adult heights. Bone age, age at onset of puberty, and yearly growth rate are normal in IUGR patients, and the patients are characteristically thin. Within this grouping are many distinctive genetic or sporadically occurring syndromes. The most common example is Russell-Silver dwarfism, characterized by small size at birth, triangular facies, a variable degree of asymmetry of extremities, and clinodactyly of the fifth finger. Intrauterine infections with Toxoplasma gondii, rubella virus, cytomegalovirus, herpesvirus, and human immunodeficiency virus are noted to cause IUGR. Furthermore, maternal drug usage, either illicit (eg, cocaine), legal but ill-advised (eg, alcohol during pregnancy), or legally prescribed medication (eg, phenytoin) may cause IUGR. Reports of other syndromes in small-for-gestational-age infants can be found in sources listed in the bibliography.

While IUGR is not an endocrine cause of short stature, several studies report increased growth velocity when GH is administered. GH is presently approved for use in intrauterine growth retardation. The effects of final height are known in only a few patients, but the early results suggest a beneficial action.

  1. Syndromes of Short Stature

Many syndromes include short stature as a characteristic feature. Some include intrauterine growth retardation and some do not. Common ones are described briefly below. Laurence-Moon, Biedl-Bardet, or Prader-Willi syndrome combine obesity with short stature, as do the endocrine conditions, hypothyroidism, glucocorticoid excess, pseudohypoparathyroidism, and GH deficiency. Moderately obese but otherwise normal children without these conditions tend to have slightly advanced bone age and advanced physiologic maturation




with increased stature during childhood and early onset of puberty. Thus, short stature in an overweight child must be considered to have an organic cause until proved otherwise.

Table 6-2. Causes of abnormalities of growth.


Nonendocrine causes

Endocrine disorders

   Constitutional short stature

 GH deficiency and variants

   Genetic short stature

      Congenital GH deficiency

   Intrauterine growth retardation

  With midline defects

   Syndromes of short stature

  With other pituitary hormone deficiencies

      Turner's syndrome and its variants

  Isolated GH deficiency

      Noonan's syndrome (pseudo-Turner's syndrome)

  Pituitary agenesis

      Prader-Willi syndrome

      Acquired GH deficiency

      Laurence-Moon and Bardet-Biedl syndromes

  Hypothalamic-pituitary tumors

      Other autosomal abnormalities and dysmorphic syndromes

  Histiocytosis X


  Central nervous system infections

   Chronic disease

  Head injuries

      Cardiac disorders

   GH deficiency following cranial irradiation

  Left-to-right shunt

  Central nervous system vascular accidents

  Congestive heart failure


      Pulmonary disorders

  Empty sella syndrome

  Cystic fibrosis

      Abnormalities of GH action


  Laron's dwarfism

      Gastrointestinal disorders


  Malabsorption (eg, celiac disease)

   Psychosocial dwarfism

  Disorders of swallowing


      Hepatic disorders

   Glucocorticoid excess (Cushing's syndrome)

      Hematologic disorders


  Sickle cell anemia




      Renal disorders

   Disorders of vitamin D metabolism

  Renal tubular acidosis

   Diabetes mellitus

  Chronic uremia

   Diabetes insipidus, untreated

      Immunologic disorders


  Connective tissue disease


  Juvenile rheumatoid arthritis


  Chronic infection


      Central nervous system disorders




  Decreased availability of nutrients


  Fad diets


  Voluntary dieting


  Anorexia nervosa


  Anorexia of cancer chemotherapy



Nonendocrine causes

Endocrine disorders

   Constitutional tall stature

   Pituitary gigantism

   Genetic tall stature

   Sexual precocity

   Syndromes of tall stature


      Cerebral gigantism

   Infants of diabetic mothers

      Marfan's syndrome




      Beckwith-Wiedemann syndrome


      XYY and XYYY syndromes


      Klinefelter's syndrome


Turner's Syndrome & Its Variants

While classic Turner's syndrome of 45,XO gonadal dysgenesis (see Chapter 14) is often correctly diagnosed, it is not always appreciated that any phenotypic female with short stature may have a variant of Turner's syndrome. Thus, a karyotype determination should be done for every short girl if no other cause for short stature is found, especially if puberty is delayed (see Chapters 13, 14, and 15).

Noonan's Syndrome (Pseudo-Turner Syndrome)

This syndrome shares several phenotypic characteristics of Turner's syndrome, including short stature, webbed neck, low posterior hairline, and facial resemblance to Turner's syndrome, but the karyotype is 46,XX in the female or 46,XY in the male with Noonan's syndrome, and other features clearly differentiate it from Turner's syndrome—eg, in Turner's syndrome there is characteristically left-sided heart disease and in Noonan's syndrome right-sided heart disease. Noonan's syndrome is an autosomal dominant disorder at gene locus 12q24 (see Chapters 14 and 15).

Prader-Willi Syndrome

This condition is characterized by fetal (poor intrauterine movement) and infantile hypotonia, acromicria (small hands and feet), developmental delay, almond-shaped eyes, and extreme obesity. Glucose intolerance and delayed puberty are characteristic. This syndrome is due to deletion of the small nuclear riboprotein polypeptide N (SNRPN) on paternal chromosome 15 (q11-13), uniparental disomy of maternal chromosome 15, or methylation of this region of chromosome 15 of paternal origin. If a mutation of the same locus is derived from the mother, Angelman's syndrome results (see Chapter 15).

Laurence-Moon Syndrome & Biedl-Bardet Syndrome

Biedl-Bardet syndrome, associated with mutations on chromosome 16 (q21), is characterized by developmental delay, retinitis pigmentosa, polydactyly, and obesity. Laurence-Moon syndrome is characterized by developmental delay, retinitis pigmentosa, delayed puberty, and spastic paraplegia. Both syndromes are associated with poor growth and obesity. They are inherited as autosomal recessive disorders (seeChapter 15).

Autosomal Chromosomal Disorders & Syndromes

Numerous other autosomal chromosomal disorders and syndromes of dysmorphic children with or without mental retardation are characterized by short stature. Often the key to diagnosis is the presence of several major or minor physical abnormalities that indicate the need for karyotype determination. Other abnormalities may include unusual body proportions, such as short extremities, leading to aberrant US:LS ratios, and arm spans quite discrepant from stature. Some conditions, such as trisomy 21 (Down's syndrome), are quite common, while others are rare. Details of these syndromes can be found in the references listed at the end of the chapter.

Skeletal Dysplasias

There are more than 100 known types of genetic skeletal dysplasias (osteochondrodysplasias). Often they are noted at birth because of the presence of short limbs or trunk, but some are only diagnosed after a period of postnatal growth. The most common condition is autosomal dominant achondroplasia. This condition is characterized by short extremities in the proximal regions, a relatively large head with a prominent forehead due to frontal bossing and a depressed nasal bridge, and lumbar lordosis in later life. Intelligence is normal. Mutations of the tyrosine kinase domain of the fibroblast growth factor receptor gene (FGFR3 gene locus 4p16.3) have been described in this condition. Adult height is decreased, with a mean of 132 cm for males and 123 cm for females. Limb lengthening operations are used to increase stature in a few centers. Children with achondroplasia who have received GH have in some instances demonstrated improved growth; however, one child experienced atlanto-occipital dislocation. The potential for abnormal brain growth and its relationship to aberrant skull shape mandates the caution that GH is not considered established therapy for this condition. Hypochondroplasia is manifested on a continuum from severe short-limbed dwarfism to apparent normal development until puberty, when there is an attenuated or absent pubertal growth spurt, leading to short adult stature. This disorder may be caused by an abnormal allele in the achondroplasia gene.

  1. Chronic Disease

Severe chronic disease involving any organ system can cause poor growth in childhood and adolescence. In many cases, there will be adequate physical findings by the time of consultation to permit diagnosis; in some cases, however—most notably celiac disease and regional enteritis—short stature and decreased growth


may precede obvious signs of malnutrition or gastrointestinal disease. In some cases, growth is only delayed and may spontaneously improve. In others, growth can be increased by improved nutrition; patients with gastrointestinal disease, kidney disease, or cancer may benefit from nocturnal parenteral nutritional infusions. Cystic fibrosis combines several causes of growth failure: lung disease impairs oxygenation and predisposes to chronic infections, gastrointestinal disease decreases nutrient availability, and late-developing abnormalities of the endocrine pancreas cause diabetes mellitus. Children with cystic fibrosis experience decreased growth rates after one year of age following a normal birth size. The pubertal growth spurt is often decreased in magnitude and delayed in its timing: secondary sexual development may be delayed especially in those with impaired pulmonary function. Study of growth in these patients allowed development of a cystic fibrosis specific growth chart and indicates that a reasonable outcome is an adult height in the 25th percentile. Children with congestive heart failure due to a variety of congenital heart diseases or acquired myocarditis grow poorly unless successfully treated with medications or surgery; patients with cyanotic heart disease experience less deficit in growth.

Celiac disease may present initially with growth failure. Early diagnosis can be made by determination of antigliadin, antiendomysial, or antireticulin antibodies while on a normal wheat-containing diet. These studies may be falsely positive, and a biopsy may still be required for diagnosis. On a gluten-free diet, patients experience catch-up growth which is strongest in the first year of therapy but continues for several more. Adult height may still be impaired, depending upon the duration of the period without treatment. Untreated patients with celiac disease have decreased serum IGF-I concentrations, presumably due to malnutrition while IGF-I concentrations rise with dietary therapy; thus serum IGF-I in this condition, as in many with nutritional deficiencies, is not a reliable indicator of GH secretory status.

Crohn's disease is associated with poor growth and decreased serum IGF-I concentrations. On an elemental diet growth rate increases and with glucocorticoid therapy (moderate doses), growth rate improves even though serum IGF-I decreases.

Patients with chronic hematologic diseases, such as sickle cell anemia or thalassemia, often have poor growth, delayed puberty, and short adult stature. Juvenile rheumatoid arthritis may compromise growth before or after therapy with glucocorticoids. GH treatment is reported to increase the growth rate of these children, but it is too early to draw conclusions about the efficacy and safety of such therapy in juvenile rheumatoid arthritis.

Chronic renal disease is known to interfere with growth. Hypophosphatemic vitamin D-resistant rickets will usually lead to short adult stature, but treatment with 1,25-hydroxyvitamin D3 (cholecalciferol) and oral phosphate in most cases will lead to increased—if not normal—adult stature. Children with chronic renal failure are reported to have increased growth rate with improved nutrition and GH therapy.

Proximal and distal renal tubular acidosis may both cause short stature. Proximal renal tubular acidosis demonstrates bicarbonate wasting at normal or low plasma bicarbonate concentrations; patients have hypokalemia, alkaline urine pH, severe bicarbonaturia, and, later, acidemia. The condition may be inherited, sporadic, or secondary to many metabolic or medication-induced disorders. Distal renal tubular acidosis is caused by inability to acidify the urine; it may occur in sporadic or familial patterns or be acquired as a result of metabolic disorders or medication therapy. Distal renal tubular acidosis is characterized by hypokalemia, hypercalciuria, and occasional hypocalcemia. The administration of bicarbonate is the primary therapy for proximal renal tubular acidosis, and proper treatment can substantially improve growth rate.

Obstructive sleep apnea is associated with poor growth. The amount of energy expended during sleep in children with sleep apnea appears to limit weight and length gain, a pattern which reverses with the resolution of the obstruction.

Hemoglobin, white blood cell count, erythrocyte sedimentation rate, serum carotene and folate levels, antigliadin, antiendomysial, antireticulin or tissue transglutaminase antibodies, plasma bicarbonate levels, and liver and kidney function should be assessed in short but otherwise apparently healthy children before endocrine screening tests are done. Urinalysis should be performed, with attention to specific gravity (to rule out diabetes insipidus) and ability to acidify urine (to evaluate possible renal tubular acidosis). All short girls without a diagnosis should undergo karyotype analysis to rule out Turner's syndrome. A list of chronic diseases causing short stature is presented inTable 6-2.

  1. Malnutrition

Malnutrition (other than that associated with chronic disease) is the most common cause of short stature worldwide. Diagnosis in the developed world is based on historical and physical findings, particularly the dietary history. Food faddism and anorexia nervosa—as well as voluntary dieting can cause poor growth. Infection with parasites such as Ascaris lumbricoides or Giardia lamblia can decrease growth. Specific nutritional deficiencies can have particular effects upon growth. For example, severe iron deficiency can cause a thin


habitus as well as growth retardation. Zinc deficiency can cause anorexia, decreased growth, and delayed puberty, usually in the presence of chronic systemic disease or infection. Children with nutritional deficiencies will characteristically demonstrate failure of weight gain before growth rate decreases, and weight for height will decrease. This is in contrast to many endocrine causes of poor growth, where weight for height remains in the normal or high range. There are no simple laboratory tests for diagnosis of malnutrition, though serum IGF-I concentrations are low in malnutrition, as they are in GH deficiency.

  1. Medications

Children with hyperactivity disorders (or those incorrectly diagnosed as such) are frequently managed with chronic dextroamphetamine or methylphenidate administration. In larger doses, these agents can decrease weight gain—probably because of their effects on appetite—and they have been reported to lower growth rate, albeit inconsistently. These drugs must be used in moderation and only in children who definitely respond to them.

Exogenous glucocorticoids are a potent cause of poor growth (see below).


  1. Growth Hormone Deficiency & Its Variants (Table 6-3)

The incidence of GH deficiency is estimated to be between 1:4000 and 1:3500 in Utah and in Scotland, so the disorder should not be considered rare. Using the conservative criteria of height less than the third percentile and growth velocity less than 5 cm per year, the incidence of endocrine disease in 114,881 Utah children was 5%, with a higher incidence in boys than girls by a ratio of over 2.5:1. In this population, 48% of the children with Turner's syndrome or growth hormone deficiency were not diagnosed prior to the careful evaluation afforded by this study.

There may be abnormalities at various levels of the hypothalamic-pituitary GH-IGF-I axis. Most patients with idiopathic GH deficiency apparently lack GHRH. One autopsied GH-deficient patient had an adequate number of pituitary somatotrophs that contained considerable GH stores; the pituitary gland produced growth hormone, but it could not be released. Long-term treatment of such patients with GHRH can cause GH release and improve growth. Patients with pituitary tumors or those rare patients with congenital absence of the pituitary gland lack somatotrophs. Several kindreds have been described that lack various regions of the GH gene responsible for producing GH. Alternatively, gene defects responsible for the embryogenesis of the pituitary gland may cause multiple pituitary deficiencies. Absence of thePIT1 gene (a pituitary-specific transcription factor) causes deficient GH, TSH, and prolactin synthesis and secretion. Mutations of the PROP1gene cause deficiencies of GH, TSH, FSH, LH production, and possibly ACTH.

Table 6-3. Postulated disorders of hGH release and action.

Site of Defect

Clinical Condition


Idiopathic GH deficiency due to decreased GHRH secretion; hypothalamic tumors

Pituitary gland

Dysplasia, trauma, surgery, or tumor of the pituitary gland; gene defect with impaired GH biosynthesis

Sites of IGFproduction

Laron's dwarfism with high GH and low IGF concentrations (GH receptor defect)
Pygmies with normal GH, low IGF-1, and normal IGF-2 concentrations


Glucocorticoid-induced growth failure. Resistance to IGF-1

Congenital Growth Hormone Deficiency

Congenital GH deficiency presents with slightly decreased birth length (-1 SD) but decreased growth rate soon after birth. The disorder is identified by careful measurement in the first year and becomes more obvious by 1–2 years of age. Patients with classic GH deficiency have short stature, obesity with immature facial appearance, immature high-pitched voice, and delay in skeletal maturation. Less severe forms of partial GH deficiency are described with few abnormal characteristics apart from short stature. Growth hormone deficient patients lack the lipolytic effects of growth hormone partially accounting for the pudgy appearance. There is a higher incidence of hyperlipidemia with elevated total cholesterol and LDL in GH deficiency, and longitudinal studies demonstrate elevation of HDL with GH treatment. Males with GH deficiency may have microphallus (penis less than 2 cm in length at birth), especially if the condition is accompanied by gonadotropin-releasing hormone (GnRH) deficiency (Figure 6-8). GH deficiency in the neonate or child can also lead to symptomatic hypoglycemia and seizures; if ACTH deficiency is also present, hypoglycemia is usually more severe. The differential diagnosis


of neonatal hypoglycemia in a full-term infant who has not sustained birth trauma must include neonatal hypopituitarism. If microphallus (in a male subject), optic hypoplasia, or some other midline facial or central nervous system defect is noted, the diagnosis of congenital GH deficiency is more likely (see below). Congenital GH deficiency is also statistically correlated with breech delivery. Intelligence is normal in GH deficiency unless repeated or severe hypoglycemia or a significant anatomic defect has compromised brain development. When thyrotropin-releasing hormone (TRH) deficiency is also present, there may be additional signs of hypothyroidism. Secondary or tertiary congenital hypothyroidism is not usually associated with mental retardation as is congenital primary hypothyroidism, but a few cases of isolated TRH deficiency and severe mental retardation have been reported.


Figure 6-8. A 12-month-old boy with congenital hypopituitarism. He had hypoglycemic seizures at 12 hours of age. At 1 year, he had another hypoglycemic seizure (plasma glucose, 25 mg/dL) associated with an episode of otitis media, and it was noted that his penis was quite small. At 12 months, length was 66.5 cm (-2 SD) and weight was 8.5 kg (-3 SD). The penis was less than 1.5 cm long, and both testes were descended (each 1 cm in diameter). Plasma GH did not rise above 1 ng/mL after arginine and levodopa testing. (No insulin was given because of the history of hypoglycemia.) LH rose very little after administration of GnRH (gonadorelin), 100 ľg. Serum thyroxine was low (T4, 6.6 ľg/dL; T4 index, 1.5), and after administration of 200 ľg of protirelin (TRH), serum TSH rose with a delayed peak characteristic of tertiary hypothyroidism. Plasma ACTH rose only to 53 pg/mL after metyrapone. Thus, the patient had multiple defects in the hypothalamic-pituitary axis. He was given six doses of 2000 units each of chorionic gonadotropin (hCG) intramuscularly over 2 weeks, and plasma testosterone rose to 62 ng/dL, indicating normal testicular function. He was then treated with 25 mg of testosterone enanthate every month for 3 months, and his phallus enlarged to 3.5 × 1.2 cm without significant advancement of bone age. With hGH therapy (0.05 mg/kg intramuscularly every other day), he grew at a greater than normal rate for 12 months (catch-up growth), and growth then continued at a normal rate.

Congenital GH deficiency may present with midline anatomic defects. Optic hypoplasia with visual defects ranging from nystagmus to blindness is found with variable hypothalamic deficiency, including diabetes insipidus; about half of patients with optic hypoplasia have absence of the septum pellucidum on CT scan or MRI, leading to the diagnosis of septo-optic dysplasia. Septo-optic dysplasia is most often sporadic in occurrence, but families are reported with mutations of the HESX1 locus (3p21.2-p21.1), a homeobox gene. Cleft palate or other forms of oral dysraphism are associated with GH deficiency in about 7% of cases; thus, such children may need more than nutritional support to improve their growth. An unusual midline defect associated with GH deficiency is described in children with a single maxillary incisor.

Congenital absence of the pituitary, which occurs in an autosomal recessive pattern, leads to severe hypopituitarism, including hypoglycemia and hypopituitarism; affected patients have shallow development or absence of the sella turcica. This defect is quite rare but clinically devastating if treatment is delayed.

Hereditary GH deficiency is described in several kindreds. Recent biochemical techniques have defined various genetic defects of the GHNgene (17q22-24) in affected families. Type IA GH deficiency is inherited in an autosomal recessive pattern, and patients have deletions, frameshifts, and nonsense mutations in the GH genome; unlike those with classic sporadic GH deficiency, some of these children are reported with short birth lengths. Patients with absent or abnormal GH genes do initially respond to exogenous hGH administration, but some soon develop high antibody titers that eliminate the effect of therapy; one reported kindred had a heterogeneous response as one sibling continued to grow and did not develop blocking antibodies while the opposite effect occurred in two other siblings in the same family. Patients with high titers of blocking antibodies


are reported to benefit from IGF-I therapy in place of GH therapy. Type IB patients have autosomal recessive splice site mutations and incomplete GH deficiency; they are less severely affected. Type II patients have autosomal dominant GH deficiency due to splice site or missense mutations; and type III patients have X-linked GH deficiency often associated with hypogammaglobulinemia. A few patients are described with abnormalities of the GRF gene, while others are described with mutant GH molecules.

Acquired Growth Hormone Deficiency

Onset of GH deficiency in late childhood or adolescence, particularly if accompanied by other pituitary hormone deficiencies, is ominous and may be due to a hypothalamic-pituitary tumor. The development of posterior pituitary deficiency in addition to anterior pituitary deficiency makes a tumor even more likely. The empty sella syndrome is more frequently associated with hypothalamic-pituitary abnormalities in childhood than in adulthood; thus, GH deficiency may be found in affected patients.

Some patients, chiefly boys with constitutional delay in growth and adolescence, may have transient GH deficiency on testing before the actual onset of puberty; when serum testosterone concentrations begin to increase in these patients, GH secretion and growth rate also increase. This transient state may incorrectly suggest bona fide GH deficiency but does not require therapy. Conditions that cause acquired GH deficiency—craniopharyngiomas, germinomas, gliomas, histiocytosis X, etc—are described in Chapter 5 and 15. It is remarkable that after craniopharyngioma removal, some patients, mainly obese subjects, continue to grow quite well in spite of the absence of GH secretion. This appears to be caused by hyperinsulinemia.

Cranial irradiation of the hypothalamic-pituitary region to treat head tumors or acute lymphoblastic leukemia may result in GH deficiency approximately 12–18 months later, owing to radiation-induced hypothalamic (or perhaps pituitary) damage. Higher doses of irradiation such as the 24 Gy previously routinely used for the treatment of central nervous system leukemia have greater effect (final height may be as much as 1.7 SD below the mean) than the 18 Gy used more routinely now. Girls treated at an early age with this newer regimen still appear to be at risk for growth failure. All children must be carefully observed for growth failure after irradiation. If these patients receive spinal irradiation, upper body growth may also be impaired, causing a decreased US:LS ratio. Abdominal irradiation for Wilms' tumor may also lead to decreased spinal growth (estimated loss of 10 cm height from megavoltage therapy for treatment at 1 year of age, and 7 cm from treatment at 5 years of age). Others receiving gonadal irradiation (or chemotherapy) have impaired gonadal function and lack onset or progression of puberty and have diminished or absent pubertal growth spurt.

Other Types of Growth Hormone Dysfunction or Deficiency

Other disorders of GH production or action are not manifested in the classic manner of GH deficiency.

Laron's syndrome (primary GH resistance or insensitivity or primary IGF-I deficiency) is due to GH receptor or postreceptor defects in an autosomal recessive pattern. Patients with decreased or absent GH receptors have decreased serum GHBP levels, while those with postreceptor defects have normal GHBP concentrations. Affected children are found throughout the world, including Israel, where the syndrome was first reported, and Ecuador, where several generations of a large kindred were studied in great detail; defects in various kindreds include nonsense mutations, deletions, RNA processing defects, and translational stop codons. Serum GH is elevated, with decreased or absent IGF-I. The growth deficiency does not respond to GH treatment. Patients are short at birth, confirming the importance of IGF-I in fetal growth that was apparent in IGF-I gene knockout experiments in mice. The head circumference and jaw are small, and there is some intellectual impairment. About one-third have hypoglycemia, and half of boys have microphallus. Patients treated with recombinant DNA-derived IGF-I grew at an improved rate but did not respond as well to IGF-I as GH-deficient children do to GH treatment, indicating the direct role of GH in fostering growth above the effect of IGF-I itself.

Other forms of GH resistance are described, but the majority of patients with disorders of the GH axis have abnormalities of GH secretion, not action. Very short, poorly growing children with delayed skeletal maturation, normal GH and IGF-I values, and no signs of organic disease have responded to GH therapy with increased growth rates equal to those of patients with bona fide GH deficiency. These patients may have a variation of constitutional delay in growth or genetic short stature, but a subtle abnormality of GH secretion or action is possible.

A few patients are reported with defects of the IGF-I gene or with deficiency of the IGF-I receptor.

Why do certain normal children, perhaps within a short family, have stature significantly lower than the mean? There is no definite answer to this persistent question, but some patients have decreased serum GHBP concentrations, which suggests a decrease in GH receptors in these children. A minority of short, poorly growing children have definable genetic abnormalities


of their GH receptors. It is likely that short stature is the final common pathway of numerous biochemical abnormalities.

With plentiful GH supplies, there is increasing pressure to treat more children—usually boys—who are not severely short and are not growing extremely slowly and do not have greatly delayed bone ages. Some studies suggest that increased final height can be achieved with such treatment, but others do not, and the treatment is not standard and has not proved to be effective.

Pygmies have normal serum GH, low IGF-I, and normal IGF-II concentrations. They will not respond to exogenous GH with improved growth rate or a rise in IGF-I. Thus, they have a congenital inability to produce IGF-I, which has greater importance in stimulating growth than IGF-II. Pygmy children are reported to lack a pubertal growth spurt, suggesting that IGF-I is essential to attain a normal peak growth velocity. Efe pygmies, the shortest of the pygmies, are significantly smaller at birth than neighboring Africans, and their growth is slower throughout childhood, leading to statures displaced progressively below the mean. Presumably, IGF-I therapy would increase growth rates in this population during childhood and puberty, but such therapy has not yet been reported.

Adults who had growth hormone deficiency in childhood or adolescence have decreased bone mass compared with normals even when bone mass is corrected for their smaller size. Many of these patients were treated with GH for various periods, but it appears that GH therapy may not completely reverse the effects of GH deficiency on skeletal density.

Diagnosis of Growth Hormone Deficiency

Because basal values of serum GH are low in normal children and GH-deficient patients alike, the diagnosis of GH deficiency rests upon demonstration of an inadequate rise of serum GH after provocative stimuli or upon some other measure of GH secretion. This process is complicated because different radioimmunoassay systems vary widely in their measurements of GH in the same blood sample (eg, a result on a single sample may be above 10 ng/mL in one assay but only 6 ng/mL in another). The physician must be familiar with the standards of the laboratory being used. Most insurance companies and state agencies will accept inability of GH to rise above 10 ng/mL with stimulation as inadequate and diagnostic of GH deficiency.

Another complicating factor is the state of pubertal development. Prepubertal children secrete less GH than pubertal subjects and, especially as they approach the onset of puberty, may have sufficiently reduced GH secretion to falsely suggest bona fide GH deficiency. This factor is sometimes addressed by administering a dose of estrogen to such subjects before testing. The very concept of GH testing provides a further complication. GH is released in episodic pulses. While a patient who does not secrete GH in response to standard challenges is generally considered to be GH-deficient, a normal GH response to these tests may not rule out GH deficiency. Testing should occur after an overnight fast; carbohydrate or fat ingestion will suppress GH response. Obesity suppresses GH secretion, and a chubby child may falsely appear to have GH deficiency. Because 10% or more of healthy children will not have an adequate rise in GH with one test of GH reserve, at least two methods of assessing GH reserve are necessary before the diagnosis of classic GH deficiency is assigned. Of course, if GH rises above 10 ng/mL in a single test, classic GH deficiency is eliminated. Serum GH values should rise after 10 minutes of vigorous exercise; this is used as a screening test. After an overnight fast, GH levels should rise in response to arginine infusion (0.5 g/kg body weight [up to 20 g] over 30 minutes), oral levodopa (125 mg for up to 15 kg body weight, 250 mg for up to 35 kg, or 500 mg for over 35 kg), or clonidine (0.1-0.15 mg/m2 orally). Side effects of levodopa include nausea; those of clonidine include some drop in blood pressure and drowsiness.

GH levels also rise after acute hypoglycemia due to insulin administration; however, this test carries a risk of seizure if the blood glucose level drops excessively. An insulin tolerance test may be performed if a 10-25% dextrose infusion is available for emergency administration in the face of hypoglycemic coma or seizure, if the patient can be continuously observed by a physician, and if the patient has no history of hypoglycemic seizures. The patient must have a normal glucose concentration at the beginning of the test in the morning after an overnight fast (water intake is acceptable). Regular insulin, 0.075–0.1 unit/kg in saline, may be given as an intravenous bolus. In 20–40 minutes, a 50% drop in blood glucose will occur, and a rise in serum GH and cortisol and ACTH should follow. Serum glucose should be monitored, and an intravenous line must be maintained for emergency dextrose infusion in case the patient becomes unconscious or has a hypoglycemic seizure. If dextrose infusion is necessary, it is imperative that blood glucose not be raised far above the normal range, since hyperosmolality has been reported from overzealous glucose replacement; undiluted 50% dextrose should not be used (see Chapter 5).

A family of penta- and hexapeptides called growth hormone-releasing peptides (GHRPs) stimulate GH secretion in normals and in growth hormone-deficient subjects. GHRPs act via ghrelin receptors that are different from the GRF receptors, and their effects are additive to that of GRF. These agents are used in diagnosis and therapy.



Patients who respond to pharmacologic stimuli (eg, levodopa, clonidine, or insulin) but not to physiologic stimuli such as exercise or sleep were said to have neurosecretory dysfunction; these patients have decreased 24-hour secretion of hGH (or integrated concentrations of hGH) compared with healthy subjects, patterns similar to those observed in GH-deficient patients. It is not clear how frequently this condition is encountered.

This long discussion of the interpretation of GH after secretagogue testing brings into question the very standard for the diagnosis of GH deficiency. It is clear that pharmacologic testing cannot always determine which patients truly need GH therapy, and some authorities suggest we abandon such dynamic testing.

Serum IGF-I and IGFBP-3 measurements are alternative methods for evaluating GH adequacy. Serum IGF-I values will be low in most GH-deficient subjects, but, as noted above, some short patients with normal serum IGF-I concentrations may require GH treatment to improve growth rate. In addition, starvation will lower IGF-I values in healthy children and incorrectly suggest GH deficiency. Children with psychosocial dwarfism—who need family therapy or foster home placement rather than GH therapy—have low GH and IGF-I concentrations and may falsely appear to have growth hormone deficiency. Likewise, patients with constitutional delay in adolescence will have low IGF-I values for chronologic age but normal values for skeletal age and may have temporarily decreased GH response to secretagogues. Thus, IGF-I determinations are not infallible in the diagnosis of GH deficiency. They must be interpreted with regard to nutrition, psychosocial status, and skeletal ages. IGFBP-3 is GH-dependent, and if its concentration is low, it is more indicative of GH deficiency than IGF-I determination.

Although pharmacologic tests of GH secretion and serum IGF-I and IGFBP-3 values will usually identify those individuals who have classic GH deficiency, the diagnosis will remain in doubt in some cases. This should not lead to the conclusion that all short children should receive GH therapy. Only about half of very short children (height well below the fifth percentile or > 2.5–3.5 SD below the mean) who grow very slowly (growth velocity below the fifth percentile for age) with delayed bone ages, meticulously studied in research protocols, were shown to benefit from hGH therapy in the absence of classic GH deficiency. However, children meeting less stringent criteria have been treated with GH in controlled trials, producing some increase in growth rate in some but not all studies; it is not yet clear whether this treatment increases adult height. Thus, in the absence of classic GH deficiency, no clearly recognized measurement can predict which short child is likely to respond to GH therapy before it is instituted. The Growth Hormone Research Society produced criteria that attempt to deal with the diagnosis of growth hormone deficiency in childhood in spite of the uncertainty of the methods. These criteria use clinical findings of various conditions associated with growth hormone deficiency, the severity of short stature, and the degree and duration of decreased growth velocity to identify individuals that may have GH deficiency. The guidelines and diagnostic considerations in this chapter include most of the GH Society criteria. (See “Consensus guidelines” reference in the Short Stature section at the end of this chapter for details of the GH Society statement.) A 3- to 6-month therapeutic trial of GH therapy may be necessary. In patients who were diagnosed late, have entered puberty, and appear to have limited time to respond to GH before epiphysial fusion causes the cessation of growth, a GnRH agonist has been used to delay epiphysial fusion in clinical trials.

GH therapy will increase adult height in Turner's syndrome if started early enough; the addition of low-dose oxandrolone will further increase growth rate. Estrogen must be used during the adolescent years only in low doses and only after the normal age of onset of puberty is reached to preserve adult height.

Treatment of Growth Hormone Deficiency


Before 1986, the only available method of treatment for GH deficiency was replacement therapy with human GH (hGH) derived from cadaver donors. In 1985 and thereafter, Creutzfeldt-Jakob disease, a degenerative neurologic disease rare in patients so young, was diagnosed in some patients who had received natural hGH 10–15 years before. Because of the possibility that prions contaminating donor pituitary glands were transmitted to the GH-deficient patients, causing their deaths, natural growth hormone from all sources was removed from distribution. Recombinant growth hormone now accounts for the world's current supply.

Commercial growth hormone is currently available in the 191-amino-acid natural-sequence form (somatropin) and, less commonly, in the 192-amino-acid methionyl form (somatrem). There is no convincing evidence that either form is clinically superior to the other, but somatropin is prescribed more often. GH is now available in virtually unlimited amounts, allowing innovative treatment regimens not previously possible owing to scarce supplies; however, the potential for abuse of GH in athletes or in children of normal size whose parents wish them to be taller than average must now be addressed.

GHRH was isolated, sequenced, and synthesized and is now available for use in diagnosis and treatment. GH-deficient patients demonstrate lower or absent GH secretion after administration of GHRH. However,


episodic doses of GHRH can restore GH secretion, IGF-I production, and growth in children with idiopathic GH deficiency. The ability of GHRH administration to cause pituitary GH secretion further supports the concept that idiopathic GH deficiency is primarily a disease of the hypothalamus, not of the pituitary gland.

IGF-I is now produced by recombinant DNA technology. Although no long-term human treatment program has yet been reported, initial studies suggest that IGF-I may be a useful treatment for short stature, particularly in Laron dwarfism (and perhaps for African pygmies, should treatment be desired) where no other treatment is possible. Growth disturbances due to disorders of GH release or action are shown in Table 6-3. GH-deficient children require biosynthetic somatropin (natural sequence GH) or somatrem (methionyl GH) at a dose of 0.3 mg/kg/wk administered in one dose per day six or seven times per week during the period of active growth before epiphysial fusion. A depot preparation that may be given every 2 weeks or every 4 weeks is now available; a larger volume and sometimes multiple injections are required. The increase in growth rate (Figures 6-9, and 6-11) is most marked during the first year of therapy. Older children do not respond as well and may require larger doses. Higher doses are approved by the FDA for use in puberty. GH will not increase growth rate without adequate nutrition and euthyroid status. During the roughly 50 years since the first use of GH in children, long-term effects are reported in several series. If only the children most recently treated with recombinant GH are considered, the mean final height was 1.4 SD below the mean, a significant improvement over the -2.9 SD mean height at the start of therapy but not a true normalization of height. With earlier diagnosis and treatment, final height might reach genetic potential.

Antibodies to GH may be present in measurable quantities in the serum of children receiving GH. However, a high titer of blocking antibodies with significant binding capacity is rare except in patients with absence or abnormality of GH genes. Only a few patients are reported to have temporarily ceased growing on somatrem therapy because of antibody formation.

GH exerts anti-insulin effects. Although clinical diabetes is not a likely result of GH therapy, the long-term effects of a small rise in glucose in an otherwise healthy child are unknown. Another potential risk is the rare tendency to develop slipped capital femoral epiphyses in children receiving GH therapy. Recent data have weakened the link between slipped capital femoral epiphyses and GH therapy, but the final import of the relationship is not yet clear. Slipped capital femoral epiphyses, if associated with endocrinopathies, are most common in treated hypothyroid patients (50% one series of 80 episodes of slipped capital femoral epiphyses), followed by treated GH-deficient patients (25% of the series). This condition may occur bilaterally, and prophylactic treatment of the nonaffected side is recommended by several authorities. Pseudotumor cerebri may rarely occur with GH therapy and is usually associated with severe headache. It is reported to reverse after cessation of GH therapy, but if allowed to continue it may impair vision and cause severe complications. Organomegaly and skeletal changes like those found in acromegaly are other theoretical side effects of excessive GH therapy but do not occur with standard doses. Furthermore, several cases of prepubertal gynecomastia are reported with growth hormone therapy.

The discovery of leukemia in young adults previously treated with growth hormone was worrisome, but no cause and effect relationship has been established, and GH treatment is not considered a cause of leukemia. GH does not increase the recurrence rate of tumors existing before therapy. Thus, patients with craniopharyngiomas, for example, may receive GH, if indicated, after the disease is clinically stable without significant worry that the GH will precipitate a recurrence. Clinicians usually wait 1 year after completion of tumor therapy before starting patients on GH therapy, but doing so is not a requirement.

There are numerous studies of GH treatment in idiopathic short stature without classic growth hormone deficiency, but they are rarely controlled studies, and some suffer from methodologic flaws that may account for the variable results. The increased growth rate may be accompanied by advancing bone age, according to some studies, leading to a final height of 1.7 SD below the mean, a result indistinguishable from untreated children with the same conditions. The Lawson Wilkins Pediatric Endocrine Society and the American Academy of Pediatrics both recommend a conservative approach to the treatment of idiopathic short stature with growth hormone.

There are other conditions for which the FDA has approved the use of GH. In the most successful series of children with IUGR treated with GH, the agent increased final height between 2.0 SD and 2.7 SD. Girls with Turner's syndrome treated with GH reach a final height of more than 150 cm, an improvement from the average untreated height which is about 144 cm.

Treatment with GH is approved for chronic renal disease in childhood. GH increases growth rate above the untreated state without excessive advancement in bone age. Prader-Willi syndrome may also be treated with GH to improve growth rate and to increase lean tissue mass and bone density.

Monitoring of growth hormone replacement is mainly accomplished by measuring growth rate and annually assessing bone age advancement. Serum IGF-I


and IGFBP-3 will rise with successful therapy while GHBP will not change appreciably, but these factors are not routinely tested after the start of therapy. Clinical studies are under way to evaluate the utility of titrating the dose of GH to restore serum IGF-I to the high-normal range. Serum bone alkaline phosphatase rises with successful therapy. Urinary hydroxyproline, deoxypyridinolone, and galactosyl-hydroxylysine reflect growth rate and are used in clinical studies to reflect increased growth rate with therapy.


Figure 6-9. Examples of abnormal growth charts. Squares (□) represent the growth pattern of a child (such as patient A in Figure 6-11) with precocious sexual development and early excessive growth leading to premature closure of the epiphyses and cessation of growth. Circles (O) represent growth of a boy (such as patient B in Figure 6-11) with GH deficiency who showed progressively poorer growth until 6 years of age, when he was treated with hGH (arrow), after which catch-up growth occurred. The curves describe standard deviations from the mean.

Growth hormone deficiency is associated with an adverse lipid profile with elevated LDL cholesterol and decreased HDL cholesterol in addition to an increased BMI; GH-deficient adolescents treated with GH develop these findings within a few years after discontinuation of GH therapy. Low-dose GH therapy is now approved




for use in adults with childhood-onset growth hormone deficiency and is said to forestall these metabolic changes. One can therefore inform the parents of a child with GH deficiency that the patient may still benefit from GH therapy even after he or she stops growing.


Figure 6-10. Two examples of abnormal growth plotted on a height velocity chart. A: The plot is taken from the data recorded as squares in Figure 6-9, describing a patient with precocious puberty, premature epiphysial closure, and cessation of growth. B: The plot is taken from the data recorded as circles in Figure 6-9, describing a patient with GH deficiency who was treated with hGH (arrow) at age 6. Initial catch-up growth is noted for 2 years, with a lower (but normal) velocity of growth following. These charts display growth rate over growth intervals rather than 12-month intervals as shown on Figures 6-3 and 6-4.


Figure 6-11. Two boys demonstrating extremes of growth. The boy at left in each photograph (A) has precocious puberty due to a central nervous system lesion. At 4˝ years, he was 125.1 cm tall, which is 5 SD above the mean. (The mean height for a 4-year-old is 101.5 cm.) His testes measured 2 × 3.5 cm each, his penis 9.8 × 2.8 cm. He was muscular and had acne and a deep voice. His bone age was 10 years, the testosterone level was 480 ng/dL, and the LH rose after 100 ľg of GnRH (gonadorelin) to a pubertal response. His brain CT scan revealed a hamartoma of the tuber cinereum. The boy at right (B) at 4˝ years was 85 cm tall, which is 4.5 SD below the mean. He had the classic physical and historical characteristics of idiopathic GH deficiency, including early growth failure and a cherubic appearance. His plasma GH values did not rise after provocative testing.


Research into the psychologic outcome of patients with short stature is flawed by lack of consistent methods of investigation and lack of controlled studies, but some results are of interest. Studies vary in concluding whether short stature is harmful to a child's psychologic development or not and whether, by inference, growth hormone is helpful in improving the child's psychologic functioning. Children with growth hormone deficiency are the most extensively studied; earlier investigations suggest that they have more passive personality traits than do healthy children, may have delayed emotional maturity, and suffer from infantilization from parents, teachers, and peers. Many of these children have been held back in school because of their size without regard to their academic abilities. Some patients retain a body image of short stature even after normal height has been achieved with treatment. More recent studies challenge these views and suggest that self-image in children with height below the fifth percentile, who do not have growth hormone deficiency, is closely comparable to a population of children with normal height. These findings may not be representative of the patient population discussed above in that a normal ambulatory population of “short” children may differ from the selected group that seeks medical attention. The data suggest that short stature itself is not cause for grave psychologic concern, and such concerns should not be used to justify growth hormone therapy. Nonetheless, in certain cases depression and suicidal behavior can occur in affected adolescents because of the


psychologic stress associated with short stature and delayed development. We cannot avoid the fact that our “heightist” society values physical stature and equates it with the potential for success, a perception that is not lost on the children with short stature and their parents. A supportive environment in which they are not allowed to act younger than their age nor to occupy a“privileged place” in the family is recommended for children with short stature. Psychologic help is indicated in severe cases of depression or maladjustment.

  1. Psychosocial Dwarfism (Figure 6-12)

Children with psychosocial dwarfism present with poor growth, potbellied immature appearance, and bizarre eating and drinking habits. Parents may report that the affected child begs for food from neighbors, forages in garbage cans, and drinks from toilet bowls. As a rule, this tragic condition occurs in only one of several children in a family. Careful questioning and observation


reveal a disordered family structure in which the child is either ignored or severely disciplined. Caloric deprivation or physical battering may or may not be a feature of the history. These children have functional hypopituitarism. Testing will often reveal GH deficiency at first, but after the child is removed from the home, GH function quickly returns to normal. Diagnosis rests upon improvement in behavior or catch-up growth in the hospital or in a foster home. Separation from the family is therapeutic, but the prognosis is guarded. Family psychotherapy may be beneficial, but long-term follow-up is lacking.


Figure 6-12. Photograph and growth chart of a 9˝-year-old boy with psychosocial dwarfism. He had a long history of poor growth (< 3 cm/yr). The social history revealed that he was given less attention and punished more frequently than his seven sibs. He ate from garbage cans and begged for food, though he was not completely deprived of food at home. When the photograph was taken, he was 99 cm tall (-7 SD) and weighed 14.7 kg (-3 SD). His bone age was 5 years, with growth arrest lines visible. Serum thyroxine was 7.8 ľg/dL. Peak serum GH varied from nondetectable to 8 ng/mL on different provocative tests between age 6 years and 8˝ years. He was placed in a hospital chronic care facility (arrow) for a 6-month period and grew 9 cm, which projects to a yearly growth velocity of 18 cm. On repeat testing, the peak serum GH was 28 ng/mL.

Growth disorder due to abnormal parent-child interaction in a younger infant is called maternal deprivation. Caloric deprivation due to parental neglect may be of greater significance in this younger age group. Even in the absence of nutritional restriction or full-blown psychosocial dwarfism, constant negative interactions within a family may inhibit the growth of a child.

It is essential to consider family dynamics in the evaluation of a child with poor growth. It is not appropriate to recommend GH therapy for emotional disorders.

  1. Hypothyroidism

Thyroid hormone deficiency decreases postnatal growth rate and skeletal development and, if onset is at or before birth, leads to severe developmental delay unless treatment is rapidly provided. Screening programs for the diagnosis of congenital hypothyroidism have been instituted all over the world, and early treatment following diagnosis in the neonatal period markedly reduces growth failure and has virtually eliminated mental retardation caused by this disorder. Indeed, early treatment of congenital hypothyroidism results in normal growth. Acquired hypothyroidism in older children (eg, due to lymphocytic thyroiditis) may lead to growth failure. Characteristics of hypothyroidism are decreased growth rate and short stature, retarded bone age, and an increased US:LS ratio for chronologic age (Figure 7-33). Patients are apathetic and sluggish and have constipation, bradycardia, coarsening of features and hair, hoarseness, and delayed pubertal development. Intelligence is unaffected in late-onset hypothyroidism, but the apathy and lethargy may make it seem otherwise.

The diagnosis of congenital hypothyroidism is usually made on the basis of neonatal screening studies. In this procedure, currently in use in developed countries throughout the world, a sample of blood is taken from the heel or from the umbilical cord at birth and analyzed for total T4 or TSH. A total T4 of under 6 ľg/dL or a TSH over 25 mU/L is usually indicative of congenital hypothyroidism, but values differ by state and laboratory. A low total T4 alone may be associated with low circulating thyroxine-binding globulin (TBG), but the significantly elevated TSH is diagnostic of primary hypothyroidism. The diagnosis may be accompanied by radiologic evidence of retarded bone age in severe congenital hypothyroidism.

In older children, serum TSH is the most reliable diagnostic test. Elevated TSH with decreased free T4 eliminates the potential confusion resulting from the use of total T4, which may vary with the level of TBG or other thyroxine-binding proteins. A positive test for serum thyroglobulin antibodies or thyroperoxidase antibodies would lead to the diagnosis of autoimmune thyroid disease (Hashimoto's thyroiditis) as an explanation for the development of hypothyroidism (see Chapter 7). If both FT4 and TSH are low, the possibility of central hypothyroidism (pituitary or hypothalamic insufficiency) must be considered. This must initiate a search for other hypothalamic-pituitary endocrine deficiencies such as GH deficiency and central nervous system disease (see Chapter 5).

Treatment is by thyroxine replacement. The dose varies from a range of 10–15 ľg/kg in infancy to 2–3 ľg/kg in older children and teenagers. Suppression of TSH to normal values for age is a useful method of assessing the adequacy of replacement in acquired primary hypothyroidism. However, there are additional considerations in treatment of neonates, and consultation with a pediatric endocrinologist is essential in this age group to ensure optimal central nervous system development.

  1. Cushing's Syndrome

Excess glucocorticoids (either exogenous or endogenous) will lead to decreased growth before obesity and other signs of Cushing's syndrome develop. The underlying disease may be bilateral adrenal hyperplasia due to abnormal ACTH-cortisol regulation in Cushing's disease, autonomous adrenal adenomas, or adrenal carcinoma. The appropriate diagnosis may be missed if urinary cortisol and 17-hydroxycorticosteroid determinations are not interpreted on the basis of the child's body size or if inappropriate doses of dexamethasone are used for testing (appropriate doses are 20 ľg/kg/d for the low-dose and 80 ľg/kg/d for the high-dose dexamethasone suppression test) (seeChapter 9). Furthermore, daily variations in cortisol production necessitate several urinary or plasma cortisol determinations before Cushing's disease can be appropriately diagnosed or ruled out. The high-dose dexamethasone test was positive in 68% of a recent series of children with Cushing's disease. The corticotropin-releasing hormone (CRH) test was positive in 80% of affected patients, whereas


MRI of the pituitary was positive in only 52%. Inferior petrosal sampling (see Chapter 9) was 100% accurate in the diagnosis of Cushing's disease. Transsphenoidal microadenomectomy is the treatment of choice.

Exogenous glucocorticoids used to treat asthma or even overzealous use of topical corticosteroid ointments or creams may suppress growth. These iatrogenic causes of Cushing's syndrome, if resolved early, may allow catch-up growth and so may not affect final height. Thus, an accurate history of prior medications is important in diagnosis. Treatment of the underlying disorder (eg, transsphenoidal microadenomectomy for Cushing's disease) will restore growth rate to normal (catch-up growth may occur initially) if epiphysial fusion has not occurred, but final height will depend upon the length of the period of growth suppression.

  1. Pseudohypoparathyroidism

Pseudohypoparathyroidism type 1A is a rare disorder consisting of a characteristic phenotype and chemical signs of hypoparathyroidism (low serum calcium and high serum phosphate), though circulating PTH levels are elevated and target tissues fail to respond to exogenous PTH administration. Children with pseudohypoparathyroidism are short and chubby, with characteristic round facies and short fourth and fifth metacarpals. This constellation of findings is called Albright's hereditary osteodystrophy. Developmental delay is common. The condition is due to a defect in the alpha subunit of the Gs protein transducer (GNAS1 gene). When imprinted paternally, this results in a defect in the guanylyl nucleotide-sensitive regulatory protein that couples PTH-occupied receptors to adenylyl cyclase. Thus, patients with pseudohypoparathyroidism type 1A have a blunted rise of urinary cAMP in response to administration of PTH. Remarkably, this defect occurs in the same regulatory protein system affected in McCune-Albright syndrome, in which hyperactive endocrine events result (see Chapter 15). A rarer variant of this disorder (pseudohypoparathyroidism type 1B), in which administration of PTH produces a rise in nephrogenous cAMP but fails to induce an increase in phosphate excretion, appears to be due to a defect distal to the receptor-adenylyl cyclase complex. Treatment with high-dose vitamin D or physiologic replacement with 1,25-dihydroxyvitamin D3 (calcitriol) and exogenous calcium as well as phosphate-binding agents will correct the biochemical defects and control hypocalcemic seizures in patients with pseudohypoparathyroidism.

Two remarkable patients are reported with pseudohypoparathyroidism and premature Leydig cell maturation, both due to abnormalities in the same G protein. The defective protein was shown to be inactive at normal body temperature, leading to defective PTH activity at the level of the kidney and bone. However, it was hyperactive at the cooler temperatures in the scrotum, leading to ligand-independent activation of Leydig cell function.

Children with the pseudohypoparathyroid phenotype of Albright's hereditary osteodystrophy but with normal circulating levels of calcium, phosphate, and PTH have pseudopseudohypoparathyroidism. They require no calcium or vitamin D therapy. (See Chapter 8.)

  1. Disorders of Vitamin D Metabolism

Short stature and poor growth are features of rickets in its obvious or more subtle forms. The cause may be vitamin D deficiency due to inadequate oral intake, fat malabsorption, inadequate sunlight exposure, anticonvulsant therapy, or renal or hepatic disease. In addition, there are inherited forms of vitamin D-dependent rickets. Classic findings of vitamin D-deficient rickets include bowing of the legs, chest deformities (rachitic rosary), and characteristic radiographic findings of the extremities associated with decreased serum calcium and phosphate levels and elevated serum alkaline phosphatase levels. There are two forms of hereditary vitamin D-dependent rickets. Type I involves a renal 25OHD 1-hydroxylase deficiency, and type II involves an absent or defective vitamin D receptor. However, the most common type of rickets in the United States is X-linked hypophosphatemic rickets, a dominant genetic disorder affecting renal reabsorption of phosphate. It is associated with short stature, severe and progressive bowing of the legs (but no changes in the wrists or chest), normal or slightly elevated serum calcium, very low serum phosphate, and urinary phosphate wasting. Short stature is linked with rickets in other renal disorders associated with renal phosphate wasting. Examples include Fanconi's syndrome (including cystinosis and other inborn errors of metabolism) and renal tubular acidosis.

When treatment is effective in these disorders (eg, vitamin D for vitamin D deficiency or alkali therapy for appropriate types of renal tubular acidosis), growth rate will improve. Replacement of vitamin D and phosphate is appropriate therapy for vitamin D-resistant rickets. It improves the bowing of the legs and leads to improved growth, though there is a risk of nephrocalcinosis. This necessitates annual renal ultrasound examinations when patients are receiving vitamin D therapy (see Chapter 8).

In the Williams syndrome of elfin facies, supravalvular aortic stenosis, and mental retardation with gregarious personality, patients have intrauterine growth retardation and greatly reduced height in childhood and as


adults; this disorder may have infantile hypocalcemia but is no longer considered a disorder of vitamin D since a genetic defect in the elastin gene at 7q11.23 occurs in most affected patients (see Chapter 8 and 15).

  1. Diabetes Mellitus

Growth in type 1 diabetes mellitus depends on the efficacy of therapy; well-controlled diabetes mellitus is compatible with normal growth, while poorly controlled diabetes often causes slow growth. Liver and spleen enlargement in a poorly controlled short diabetic child is known as Mauriac's syndrome, rarely seen now owing to improved diabetic care. Another factor that may decrease growth rate in children with diabetes mellitus is the increased incidence of Hashimoto's thyroiditis; yearly thyroid function screening is advisable, especially as the peripubertal period approaches. Growth hormone concentrations are higher in children with diabetes, and this factor may play a role in the development of complications of diabetes mellitus. IGF-I concentrations tend to be normal or low, depending upon glucose control, but judging from the elevated GH noted above, the stimulation of IGF-I production by GH appears to be partially blocked in these children. (SeeChapter 17.)

  1. Diabetes Insipidus

Polyuria and polydipsia due to inadequate vasopressin (neurogenic diabetes insipidus) or inability of the kidney to respond to vasopressin (nephrogenic diabetes insipidus) leads to poor caloric intake and decreased growth. With appropriate treatment (see 5), the growth rate should return to normal. Acquired neurogenic diabetes insipidus may herald a hypothalamic-pituitary tumor, and growth failure may be due to associated GH deficiency.


An initial decision must determine whether a child is pathologically short or simply distressed because height is not as close to the 50th percentile as desired by the patient or the parents. Performing unnecessary tests is expensive and may be a source of long-term concern to the parents—a concern that could have been avoided by appropriate reassurance. Alternatively, missing a diagnosis of pathologic poor growth may cause the patient to lose inches of final height or may allow progression of disease.

If a patient's stature, growth rate, or height adjusted for midparental height is sufficiently decreased to warrant evaluation, an orderly approach to diagnosis will eliminate unnecessary laboratory testing. The medical history will provide invaluable information regarding intrauterine course and toxin exposure, birth size and the possibility of birth trauma, mental and physical development, symptoms of systemic diseases (Table 6-2), abnormal diet, and family heights and ages at which pubertal maturation occurred. Evaluation of psychosocial factors affecting the family and the relationship of parents and child can be carried out during the history-taking encounter. Often the diagnosis can be made at this point.

On physical examination, present height—measured without shoes on an accurate measuring device—and weight should be plotted and compared with any previous data available. If no past heights are available, a history of lack of change in clothing and shoe sizes or failure to lengthen skirts or pants may reflect poor growth. Questions about how the child's stature compares with that of his or her peers and whether the child's height has always had the same relationship to that of classmates are useful. One of the most important features of the evaluation process is to determine height velocity and compare the child's growth rate with the normal growth rate for age. Adjustment for midparental height is calculated and nutritional status determined. Arm span, head circumference, and US:LS ratio are measured. Physical stigmas of syndromes or systemic diseases are evaluated. Neurologic examination is essential.

If no specific diagnosis emerges from the physical examination, a set of laboratory evaluations may prove useful. Complete blood count, urinalysis, and serum chemistry screening with electrolyte measurements may reveal anemia, abnormalities of hepatic or renal disease (including concentration defects), glucose intolerance, acidosis, calcium disorder, or other electrolyte disturbances. Age-adjusted values must be used, since the normal ranges of serum alkaline phosphatase and phosphorus values are higher in children than in adults. An elevated sedimentation rate, low serum carotene, or positive antigliadin, antiendomysial, antireticulin, or tissue transglutaminase antibody determination may indicate connective tissue disease, Crohn's disease, celiac disease or malabsorption. Serum TSH and free T4 are important measurements to exclude existing thyroid disease. Skeletal age evaluation will not make a diagnosis; however, if the study shows delayed bone age, the possibility of constitutional delay in growth, hypothyroidism, or GH deficiency must be considered. The tests used for the diagnosis of GH deficiency are detailed above. If serum IGF-I is normal for age, classic GH deficiency or malnutrition is unlikely; if serum IGF-I is low, it must be considered in relation to skeletal age, nutritional status, and general health status before interpretation of the value can be made. Serum IGFBP-3 adds to the evaluation of short stature. Serum


gonadotropin and sex steroid determinations are performed if puberty is delayed. Serum prolactin may be elevated in the presence of a hypothalamic disorder. Karyotyping for Turner's syndrome is obtained in any short girl without another diagnosis, especially if puberty is delayed or gonadotropins are elevated. If Turner's syndrome is diagnosed, evaluation of thyroid function and determination of thyroid antibodies is also important. Elevated urinary free cortisol (normal: < 60 ľg/m2/24 h [< 18.7 ľmol/m2/24 h]) signifies Cushing's syndrome. If GH deficiency or impairment is found or if there is another hypothalamic-pituitary defect, an MRI is indicated with particular attention to the hypothalamic-pituitary area to rule out a congenital defect or neoplasm of the area. Ectopic location of the posterior pituitary on MRI is relatively frequent in congenital GH deficiency, as is a decreased pituitary volume.

If no diagnosis is apparent after all of the above have been considered and evaluated, more detailed procedures, such as provocative testing for GH deficiency, are indicated. It must be emphasized that a long and expensive evaluation is not necessary until it is demonstrated that psychologic or nutritional factors are not at fault. Likewise, if a healthy-appearing child presents with borderline short stature, normal growth rate, and short familial stature, a period of observation may be more appropriate than laboratory tests.


  1. Constitutional Tall Stature

A subject who has been taller than his or her peers through most of childhood, is growing at a velocity within the normal range with a moderately advanced bone age, and has no signs of the disorders listed below may be considered to be constitutionally advanced. Predicted final height will usually be in the normal adult range for the family.

Exogenous obesity in an otherwise healthy child will usually lead to moderate advancement of bone age, slightly increased growth rate, and tall stature in childhood. Puberty will begin in the early range of normal, and adult stature will conform to genetic influences. Thus, an obese child without endocrine disease should be tall; short stature and obesity are worrisome.

  1. Genetic Tall Stature

Children with exceptionally tall parents have a genetic tendency to reach a height above the normal range. The child will be tall for age, will grow at a high normal rate, and the bone age will be close to chronologic age, leading to a tall height prediction. Some children with tall stature have been noted to have growth hormone secretory patterns similar to those associated with acromegaly—eg, GH levels increase after TRH administration.

Occasionally, children will be concerned about being too tall as adults. These worries are more common in girls and will often be of greater concern to the parents than to the patient. Final height can be limited by promoting early epiphysial closure with estrogen in girls or testosterone in boys but such therapy should not be undertaken without careful consideration of the risks involved. Testosterone therapy decreases HDL cholesterol levels. Acne fulminans may be caused by testosterone therapy and progression may occur, even after therapy has been withdrawn. Estrogen carries the theoretical risk of thrombosis, ovarian cysts, and galactorrhea, but few of these complications are reported. High-dose estrogen therapy is estimated to decrease predicted final height by as much as 4.5–7 cm but only if started 3–4 years before epiphysial fusion. No therapy to limit stature is warranted until a careful assessment of the parents' and the child's expectations and reasons for seeking therapy is performed. Counseling and reassurance are usually more appropriate than endocrine therapy. Height-limiting therapy is extremely rarely invoked in the present era.

  1. Syndromes of Tall Stature

Cerebral Gigantism

The sporadic syndrome of rapid growth in infancy, prominent forehead, high-arched palate, sharp chin, and hypertelorism (Sotos' syndrome) is not associated with GH excess. Mentation is usually impaired. The growth rate decreases to normal in later childhood, but stature remains tall.

Marfan's Syndrome

Marfan's syndrome is an autosomal dominant abnormality of connective tissue exhibiting variable penetrance. The disorder is due to mutation of the fibrillin-1 gene, 15q21.1. This condition may be diagnosed by characteristic physical manifestations of tall stature, long thin fingers (arachnodactyly), hyperextension of joints, and superior lens subluxation. Pectus excavatum and scoliosis may be noted. Furthermore, aortic or mitral regurgitation or aortic root dilation may be present, and aortic dissection or rupture may ultimately occur. In patients with this syndrome, arm span exceeds height, and the US:LS ratio is quite low owing to long legs. Aortic root ultrasound and slitlamp ophthalmologic examinations are indicated.




Patients with homocystinuria have an autosomal recessive deficiency of cystathionine β-synthase (gene locus 21q22.3) and phenotypes similar to those of patients with Marfan's syndrome. Additional features of homocystinuria include developmental delay, increased incidence of seizures, osteoporosis, inferior lens dislocation, and increased urinary excretion of homocystine with increased plasma homocystine and methionine but low plasma cystine. Thromboembolic phenomena may precipitate a fatal complication. This disease is treated by restricting dietary methionine and, in responsive patients, administering pyridoxine.


Patients with Beckwith-Wiedemann syndrome demonstrate overweight (> 90th percentile birth weight) in 88%, increased postnatal growth, omphalocele in 80%, macroglossia in 97%, and hypoglycemia due to the hyperinsulinism of pancreatic hyperplasia in 63%. Other reported features include fetal adrenocortical cytomegaly, and large kidneys with medullary dysplasia. The majority of patients occur in a sporadic pattern due to a mutation at 11p15.5, but analysis of some pedigrees suggests the possibility of familial patterns.

XYY Syndrome

Patients with one (47,XYY) or more (48,XYYY) extra Y chromosomes achieve greater than average adult heights. They have normal birth lengths but higher than normal growth rates. Excess GH secretion has not been documented (see Chapter 14).

Klinefelter's Syndrome

Patients with Klinefelter's syndrome (see Chapter 15) tend toward tall stature, but this is not a constant feature.


  1. Pituitary Gigantism

Pituitary gigantism is caused by excess GH secretion before the age of epiphysial fusion. The increased growth hormone secretion may be due to somatotroph-secreting tumors, to the constitutive activated GH secretion sometimes found in the McCune-Albright syndrome, or to allelic deletion of the 11q13 locus (a tumor suppressor which is abnormal in tumors of MEN 1 or in spontaneous adenomas); alternatively, it may result from excess secretion of GHRH. Patients—besides growing excessively rapidly—have coarse features, large hands and feet with thick fingers and toes, and often frontal bossing and large jaws. Although this condition is quite rare, the findings appear similar to those observed in the more frequent acromegaly (which occurs with GH excess after epiphysial fusion). Thus, glucose intolerance or frank diabetes mellitus, hypogonadism, and thyromegaly are predicted. Treatment is accomplished by surgery (the transsphenoidal approach is used if the tumor is small enough), radiation therapy, or by therapy with a somatostatin analog.

  1. Sexual Precocity

Early onset of secretion of estrogens or androgens will lead to abnormally increased height velocity. Because bone age is advanced, there will be the paradox of the tall child who, because of early epiphysial closure, is short as an adult. The conditions include complete and incomplete sexual precocity (including virilizing congenital adrenal hyperplasia) (Figures 6-9, 6-10, and 6-11).

  1. Thyrotoxicosis

Excessive thyroid hormone due to endogenous overproduction or overtreatment with exogenous thyroxine will lead to increased growth, advanced bone age, and, if occurring in early life, craniosynostosis. If the condition remains untreated, final height will be reduced.

  1. Infants of Diabetic Mothers

Birth weight and size in infants of moderately diabetic mothers will be quite high, though severely diabetic women who have poor control may have babies with intrauterine growth retardation due to placental vascular insufficiency. Severe hypoglycemia and hypocalcemia will be evident in the babies soon after birth. The appearance and size of such babies is so striking that women have been diagnosed with gestational diabetes as a result of giving birth to affected infants. Infants of diabetic mothers have an increased prevalence of obesity by 10 years of age and thereafter.


Normal Growth

Chard T: Insulin-like growth factors and their binding proteins in normal and abnormal human fetal growth. Growth Regul 1994;4:91.

Himes JH: Minimum time intervals for serial measurements of growth in recumbent length or stature of individual children. Acta Paediatr 1999;88:120.

Juul A et al: Serum levels of insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3) in healthy infants, children, and adolescents: The relation to IGF-I, IGF-II, IGFBP-1, IGFBP-2, age, sex, body mass index, and pubertal maturation. J Clin Endocrinol Metab 1995;80:2534.



Khamis HJ, Roche AF: Predicting adult stature without using skeletal age: The Khamis-Roche method. Pediatrics 1994;94:504.

Le Roith DL, Butler AA: Insulin-like growth factors in pediatric health and disease. J Clin Endocrinol Metab 1999;84:4355.

Pierson M, Deschamps J-P: Growth. In: Pediatric Endocrinology. Job J-C, Pierson M (editors). Wiley, 1981.

Reiter EO, Rosenfeld RG: Normal and aberrant growth. In: Williams Textbook of Endocrinology, 10th ed. Wilson JD et al (editors). Saunders, 2002.

Roche AF, Wainer H, Thissen D: The RWT method for the prediction of adult stature. Pediatrics 1975;56:1027.

Styne DM: Growth. In: Handbook of Pediatric Endocrinology. Williams & Wilkins, 2003. [no pages yet as in press]

Tanner JM, Whitehouse RH: Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Arch Dis Child 1976;51:170.

Thorner MO: The discovery of growth hormone-releasing hormone. J Clin Endocrinol Metab 1999;84:4671.

Short Stature

August GP, Julius JR, Blethen SL: Adult height in children with growth hormone deficiency who are treated with biosynthetic growth hormone: the National Cooperative Growth Study experience. Pediatrics 1998;102:512.

Buchlis JG et al: Comparison of final heights of growth hormone-treated vs. untreated children with idiopathic growth failure. J Clin Endocrinol Metab 1998;83:1075.

Brown P: Human growth hormone therapy and Creutzfeldt-Jakob disease: A drama in three acts. Pediatrics 1988;81:85.

Byard PJ: The adolescent growth spurt in children with cystic fibrosis. Ann Hum Biol 1994;21:229.

Consensus guidelines for the diagnosis and treatment of growth hormone (GH) deficiency in childhood and adolescence: summary statement of the GH Research Society. GH Research Society. J Clin Endocrinol Metab 2000;85:3990.

De Zegher F et al: Early, discontinuous, high dose growth hormone treatment to normalize height and weight of short children born small for gestational age: results over 6 years. J Clin Endocrinol Metab 1999;84:1558.

Fine RN et al: Growth after recombinant human growth hormone treatment in children with chronic renal failure: report of a multicenter randomized double-blind placebo-controlled study. Genentech Cooperative Study Group. J Pediatr 1994; 124:374.

Guyda HJ: Four decades of growth hormone therapy for short children: what have we achieved? J Clin Endocrinol Metab 1999;84:4307.

Harris DA et al: Somatomedin-C in normal puberty and in true precocious puberty before and after treatment with a potent luteinizing hormone-releasing hormone agonist. J Clin Endocrinol Metab 1985;61:152.

Hintz RL: The role of auxologic and growth factor measurements in the diagnosis of growth hormone deficiency. Pediatrics 1998;102:524.

Kaplan SL: Normal growth. In: Rudolph's Pediatrics, 20th ed. Rudolph AM, Hoffman JIE, Rudolph CD (editors). Appleton & Lange, 1996.

Kunwar S, Wilson CB: Pediatric pituitary adenomas. J Clin Endocrinol Metab 1999;84:4385.

Lafferty AR, Chrousos GP: Pituitary tumors in children and adolescents. J Clin Endocrinol Metab 1999;84:4317.

Laron Z: The essential role of IGF-I: lessons from the long-term study and treatment of children and adults with Laron syndrome. J Clin Endocrinol Metab 1999;84:4397.

Levitsky LL: Growth and pubertal pattern in insulin-dependent diabetes mellitus. Semin Adolesc Med 1987;3:233.

Lindsay RM et al: Utah Growth Study: Growth standards and the prevalence of growth hormone deficiency. J Pediatr 1994; 125:29.

Loder RT, Wittenberg B, DeSilva G: Slipped capital femoral epiphysis associated with endocrine disorders. J Pediatr Orthop 1995;15:349.

Magiakou MA et al: Cushing's syndrome in children and adolescents. Presentation, diagnosis, and therapy. N Engl J Med 1994;331:629.

Matarazzo P et al: Growth impairment, IGF I hyposecretion and thyroid dysfunction in children with perinatal HIV-1 infection. Acta Paediatr 1994;83:1029.

Nishi Y et al: Treatment of isolated growth hormone deficiency type IA due to GH-I gene deletion with recombinant human insulin-like growth factor I. Acta Paediatr 1993;82:983.

Oberfield SE, Sklar CA: Endocrine sequelae in survivors of childhood cancer. Adolesc Med 2002;13:161.

Ogilvy-Stuart AL et al: Treatment of radiation-induced growth hormone deficiency with growth hormone-releasing hormone. Clin Endocrinol (Oxf) 1997;46:571.

Parks JS et al: Hereditable disorders of pituitary development. J Clin Endocrinol Metab 1999;84:4362.

Price DA et al: Efficacy and safety of growth hormone treatment in children with prior craniopharyngioma: an analysis of the Pharmacia and Upjohn International Growth Database (KIGS) from 1988–1996. Horm Res 1998;49:91.

Rose SR et al: Diagnosis of central hypothyroidism in survivors of childhood cancer. J Clin Endocrinol Metab 1999;84:4472.

Rosenbloom AL et al: Growth hormone receptor deficiency in Ecuador. J Clin Endocrinol Metab 1999;84:4436.

Rosenfeld RG, Cohen P: Disorders of growth hormone; insulin-like growth factor secretion and action. In: Sperling MA (editor): Pediatric Endocrinology. Saunders, 2002.

Rosenfeld RG et al: Growth hormone therapy of Turner's syndrome: beneficial effect on adult height. J Pediatr 1998; 132:319.

Rosenfeld RG, Rosenbloom AL, Guevara Aguirre J: Growth hormone (GH) insensitivity due to primary GH receptor deficiency. Endocrine Reviews 1994;15:369.

Rosenfeld RL et al: Optimizing estrogen replacement treatment in Turner syndrome. Pediatrics 1998;102:486.

Saenger P: Growth-promoting strategies in Turner's syndrome. J Clin Endocrinol Metab 1999;84:4345.

Saenger P et al: Metabolic consequences of 5-year growth hormone (GH) therapy in children treated with GH for idiopathic short stature. Genentech Collaborative Study Group. J Clin Endocrinol Metab 1998;83:3115.



Smith GC et al: First-trimester growth and the risk of low birth weight. N Engl J Med 1998;339:1817.

Stevenson RD: Use of segmental measures to estimate stature in children with cerebral palsy. Arch Pediatr Adolesc Med 1995; 149:658.

Styne DM et al: Treatment of Cushing's disease in childhood and adolescence by transsphenoidal microadenomectomy. N Engl J Med 1984;310:889.

Tanaka H et al: Effect of growth hormone therapy in children with achondroplasia: growth pattern, hypothalamic-pituitary function, and genotype. Eur J Endocrinol 1998;138:275.

Tanner JM, Davies PS: Clinical longitudinal standards for height and height velocity for North American children. J Pediatr 1985;107:317.

Thomas AG et al: Insulin like growth factor-I, insulin like growth factor binding protein-1, and insulin in childhood Crohn's disease. Gut 1993;34:944.

Thorner M et al: Once daily subcutaneous growth hormone-releasing hormone therapy accelerates growth in growth hormone-deficient children during the first year of therapy. J Clin Endocrinol Metab 1996;81:1189.

Tillmann V et al: The relationship between stature, growth, and short-term changes in height and weight in normal prepubertal children. Pediatr Res 1998;44:882.

Tiulpakov AN et al: Growth in children with craniopharyngioma following surgery. Clin Endocrinol (Oxf) 1998;49:733.

Van Wyk JJ, Smith EP: Insulin-like growth factors and skeletal growth: possibilities for therapeutic intervention. J Clin Endocrinol Metab 1999;84:4349.

Wollmann HA et al: Reference values for height and weight in Prader-Willi syndrome based on 315 patients. Eur J Pediatr 1998;157:634.

Tall Stature

Elliott MR et al: Clinical features and natural history of Beckwith-Wiedemann syndrome: presentation of 74 new cases. Clin Genet 1994;46:168.

Eugster EA, Peskovitz OH: Gigantism. J Clin Endocrinol Metab 1999;84:4379.

Karlberg J, Wit JM: Linear growth in Sotos syndrome. Acta Paediatr Scand 1991;80:956.