Diagnostic approach to children and adolescents with short stature

INTRODUCTION — Short stature is defined as a height that is 2 standard deviations (SD) or more below the mean height for individuals of the same sex and chronologic age in a given population. This translates to a height that is below the 2.3^rd percentile.

The most common causes of short stature beyond the first year or two of life are familial (genetic) short stature and constitutional short stature (also known as constitutional delay of growth and puberty [CDGP]), which are normal nonpathologic variants of growth. These growth patterns often can be distinguished from one another, but some children have features of both.

This topic will review the diagnostic approach to children with short stature, beginning with a brief review of normal growth and development. The causes of short stature are discussed separately. (See "Causes of short stature".)

NORMAL GROWTH — Serial measurements of growth are an important parameter in monitoring the health of children. A normal pattern of growth usually suggests good general health, while growth that is slower than normal raises the possibility of an underlying subacute or chronic illness, including an endocrinologic cause of growth failure.

Statural growth is a continuous but not linear process. There are three phases of postnatal growth (infantile, childhood, and pubertal), each of which has a distinctive pattern [1]. The phases are similar for boys and girls, but the timing and pace of growth differ, particularly during puberty.

●Intrauterine – Size at birth is determined more by maternal nutrition and intrauterine and placental factors than by genetic makeup. Not all the genes that affect growth may be expressed at birth. As a result, the correlation coefficient between birth length and adult height is only 0.25 [2].

●Infancy – During the first two years of life (infantile phase), linear growth initially is very rapid and gradually decelerates. Overall growth during this period is approximately 30 to 35 cm. The length (and weight) of premature infants must be corrected for gestational age, at least for the first year. However, growth is often accelerated during the second one-half of the first year in otherwise healthy children born early.

An infant's height curve often crosses percentile lines during the first 24 months of life as the growth moves away from the influences of the intrauterine environment and toward the child's genetic potential. As examples, infants born small because of placental or uterine constraint may move upwards across percentile curves ("channeling up"), whereas infants born large because of maternal-fetal overnutrition may cross downward across percentile curves ("channeling down"). After this adjustment period, the correlation between length at two years and adult height is 0.80 [2].

●Childhood – The childhood phase is characterized by linear growth at a relatively constant velocity, with some slowing in later childhood. Most children grow at the following rates (representing the 10^th to 90^th percentiles):

•Age two to four years – 5.5 to 9 cm/year (2.2 to 3.5 inches/year)

•Age four to six years – 5 to 8.5 cm/year (2 to 3.3 inches/year)

•Age six years to puberty:

-4 to 6 cm/year for boys (1.6 to 2.4 inches/year)

-4.5 to 6.5 cm/year for girls (1.8 to 2.6 inches/year)

●Adolescence – The pubertal phase is characterized by a growth spurt of 8 to 14 cm per year due to the synergistic effects of increasing gonadal steroids and growth hormone [3]. In girls, the pubertal growth spurt typically starts around age 10 but may start as early as age 8 for early maturing girls. In boys, the pubertal growth spurt typically starts around age 12 but may start as early as age 10 in early maturing boys [4].

The "rule of fives" incorporates these typical phases of growth and provides an estimate for normal height and height velocity (HV) in a given age group, as shown in the table (figure 1). Actual height and HV in a healthy child can vary substantially around these approximations.

EVALUATION OF GROWTH

Overview

●Goals – The goal of the evaluation of a child with short stature is to identify the subset of children with pathologic causes, such as Turner syndrome, inflammatory bowel disease or other underlying systemic disease, or hormonal abnormality. The evaluation also assesses the severity of the short stature and likely growth trajectory to facilitate decisions about intervention, if appropriate.

●Sequence of the evaluation – Evaluation of a child suspected of having short stature is guided by answering the following questions. Although the causes and clinical presentation of short stature vary by age group, the same questions are relevant for children of any age:

•How short is the child?

•Is the child's height velocity (HV) impaired?

•What is the child's likely (predicted) adult height?

The answers to these questions guide a focused history and physical examination and, in some cases, laboratory evaluation to determine the cause(s) and appropriate management of the child with short stature (algorithm 1).

●Clinical setting – Some components of the evaluation can reasonably be performed in the primary care setting, including initial interpretation of the growth chart and growth potential (based on measured heights of the child's parents), calculation of height velocity (HV), and initial laboratory screening for an underlying systemic or endocrine disease, if suspected based on symptoms. If HV is slow, then bone age determination should be performed, if expert interpretation is available. Other components of the evaluation, including review of the bone age results and the detailed evaluation for causes of short stature, are typically performed by a pediatric endocrinologist, if available.

Referral patterns reveal substantial sex differences in the evaluation and treatment of children with short stature [5-8]. Boys are referred for evaluation more often, at younger ages, and for less severe height deficits compared with girls. As an example, in one retrospective review of 288 children referred to a single center for assessment of short stature, the male:female ratio was 1.9:1 [5]. At the time of referral, the height deficit was significantly greater for girls than boys (median height Z-score -2.4 versus -1.9) and organic disease was more common among girls (40 versus 15 percent). Similarly, studies of growth hormone registries have shown preferential treatment of boys compared with girls with an approximate ratio of 2:1 [6,7,9].

This apparent sex bias may be due to underappreciation of growth problems in girls, leading to fewer evaluations of girls for short stature. Alternatively, it may be due to increased societal pressure for tall stature in boys, leading to increased referral and growth hormone treatment of boys without organic causes of short stature. These findings emphasize the need for accurate growth monitoring during the health care maintenance of all children to ensure appropriate referral and treatment.

Is the child short? — Accurate measurement of length or height is essential; inaccurate measurements or aberrant plotting is one of the more common causes of apparent growth failure. Length is measured lying down and should be used for infants and children up to 24 months of age (figure 2); height is measured standing and should be used for children two years and older (if possible) (figure 3).

●Standard approach using height-for-age growth curves – The child's length or height should be plotted on the growth chart that is appropriate for the child's age and sex (children <24 months (figure 4A-B); children two years and older (figure 5A-B)). Weight should also be measured to provide additional information about the child's nutritional status. (See "Measurement of growth in children".)

Percentiles and Z-scores can be calculated using a calculator for recumbent length (calculator 1) or standing height for boys (calculator 2) or for girls (calculator 3). For the purposes of an endocrine evaluation, short stature is defined as a length or height more than 2 standard deviations (SD) below the mean (ie, a Z-score <-2), which corresponds to a height that is <2.3^rd percentile.

•Height above the 2.3^rd percentile (>-2 SD) – These children generally do not require further specific evaluation unless there are additional reasons for concern, such as progressively decreasing height percentiles (implying growth failure), dysmorphic features, or evidence of underlying systemic disease, or if they are growing well below their genetic potential (eg, they are children of very tall parents).

•Height below the 2.3^rd percentile (≤-2 SD) – These children have short stature and should undergo a more detailed evaluation, starting with evaluation of HV, as outlined in the following sections. The first steps of this evaluation usually can be performed in the primary care setting. Early referral to a specialist is appropriate if the HV is very slow.

•Height less than the 1^st percentile (≤-2.25 SD) – These children have extreme short stature and usually should be referred to an appropriate subspecialist for a detailed evaluation, where available. The first steps in the evaluation will be similar, but a higher index of suspicion for pathologic causes of growth faltering is appropriate.

●Tanner stage adjusted (TSA) growth curves (for selected patients) – For children with late puberty for their age (including those whose puberty is delayed because of medical illness or puberty-suppressing therapy), use of standard height-for-age growth references may exaggerate the child's degree of short stature. For these children, growth potential may be more accurately assessed by using special TSA growth standards, which are captured in this online calculator or by plotting the child's growth on a TSA-adjusted growth curve, selected for their stage of pubertal development (sexual maturity rating) [10].

Is the child's height velocity impaired? — Determination of the child's HV is an essential component of the evaluation of a child for short stature and can be considered a "vital sign" (algorithm 1). This requires serial measurements of height, which should be done along with weight at each well-child visit. Serial measurements help to determine whether the child's HV is within the normal range and whether the child progressively deviates from their previous growth channel (or percentile curve). The HV should be calculated (in cm/year) using accurate measurements of height and an interval between measurements of at least six months. Plotting height and HV immediately after the measurement is taken on the child's growth chart is important because this may help identify a measurement error and permit remeasurement. The data can be easily plotted using the growth charts included in an electronic health record.

For children two years and older, growth failure is likely if:

●The height-for-age curve has deviated downward across two major height percentile curves (eg, from above the 25^th percentile to below the 10^th percentile).

●Or, if the child is growing slower than the following rates:

•Age two to four years – HV less than 5.5 cm/year (<2.2 inches/year)

•Age four to six years – HV less than 5 cm/year (<2 inches/year)

•Age six years to puberty:

-HV less than 4 cm/year for boys (<1.6 inches/year)

-HV less than 4.5 cm/year for girls (<1.8 inches/year)

Short children with HV above these cut points usually have a nonpathologic cause of short stature, such as familial short stature or constitutional delay of growth. For these children, a basic evaluation for short stature usually is sufficient, consisting of a focused history, physical examination, adult height prediction, and bone age determination, as outlined in the following sections [11]. (See 'Prediction of adult height' below and 'Bone age determination' below.)

Short children with HV below these cut points are more likely to have a pathologic cause of short stature and warrant additional attention. In addition to the basic evaluation outlined above, laboratory screening for pathologic causes of growth failure is warranted. The clinician should be particularly alert for subtle symptoms of underlying systemic disease (eg, Crohn disease) and for evidence of Turner syndrome in girls [11,12]. The detailed evaluation may be most appropriately performed by a pediatric endocrinologist, if available. (See 'Laboratory and imaging studies' below.)

Short children with HV substantially below these cut points are most likely to have a pathologic cause of short stature and should undergo a detailed evaluation for pathologic causes of growth failure by a pediatric subspecialist, if available.

Alternatively, for a more precise assessment, the child's HV can be plotted on an HV chart (figure 6A-B) to determine the HV percentile (or SD [Z-score]) for the child's age and sex (note that this is different from the height-for-age percentile). In general, HV between the 10^th and 25^th percentile should raise concern for possible growth failure and an HV below the 10^th percentile warrants a thorough evaluation for growth failure.

Prediction of adult height — Adult height is determined by a combination of genetic potential and many other factors that influence somatic growth and biologic maturation. No method accurately predicts adult height, and there is wide variation in predicted adult height among the different methods. However, an estimate of adult height can be developed using information about heights in the biologic family, combined with information about the child's own growth and level of skeletal maturation. The results help to guide decisions about evaluation and treatment and also provide some information about the possible causes of short stature in a particular patient.

Is the child's growth within the range for the family? — The next step is to determine the height range expected for the child's biologic family and compare it with the child's growth trajectory. In most populations, a predicted adult height that is <63 inches (160 cm) for males and <59 inches (150 cm) for females is considered short stature, corresponding to more than 2 SD below the mean (<2.3^rd percentile) for adults of the same population and sex.

●Midparental height – An estimate of a child's genetic height potential can be obtained by calculation of the midparental height, which is based upon the heights of both parents and adjusted for the sex of the child (algorithm 1). This estimate is also known as the "target height." Whenever possible, the biologic parents' heights should be directly measured rather than self-reported. The calculation can be performed using a calculator (calculator 4) or the following equation:

•For girls, 13 cm (or 5 inches) is subtracted from the father's height and averaged with the mother's height

•For boys, 13 cm (or 5 inches) is added to the mother's height and averaged with the father's height

•For both girls and boys, 8.5 cm (3.3 inches) on either side of this calculated value (target height) represents the 3^rd to 97^th percentiles for anticipated adult height [13]

In this calculation, the 13 cm (or 5 inches) represent the average difference in the average height of adult males and females. Thus, the child grows, on average, to the midparental height percentile.

If there is a large discrepancy between parents' heights, with one very short parent, the possibility of a dominantly inherited disorder should be considered, perhaps within a gene affecting the growth plate [14-17]. (See "Causes of short stature", section on 'Genetic diseases with primary effects on growth'.)

●Projected height – The projected height for a child older than two years is determined by extrapolating the child's growth along the current channel to the 18- to 20-year mark (figure 7). If the child's bone age is delayed or advanced, then the projected height should be plotted based on the bone age rather than the chronologic age. (See 'Bone age determination' below.)

The child's projected height is then compared with the range calculated from the midparental (target) height.

●If the child's projected height is within 8.5 cm (2 SD) of the midparental height, then the child's height is within the expected range for the family. This child probably has familial short stature, which is considered a variant of normal growth. (See "Causes of short stature", section on 'Familial short stature'.)

●If the projected height is more than 8.5 cm (2 SD) below the midparental height, then the child can be considered abnormally short for their biologic family.

Is there evidence of delayed or accelerated growth? — Most children with severe short stature (height <-2.25 SD) or growth failure should undergo a radiographic determination of bone age. This helps to determine whether the child's growth is delayed or accelerated compared with their chronologic age.

Bone age determination — Bone age (also known as skeletal age) is typically determined from a radiograph of the left hand and wrist and requires expert interpretation. The methods used most commonly for determining skeletal age are the Greulich and Pyle Atlas [18] and the Tanner-Whitehouse 3 (TW3) method [19]. In general, TW3 is considered to be more accurate, based on studies in predominantly White European populations [20]. However, bone ages obtained from these methods are less generalizable to other groups of children; Black or African American children tend to have advanced bone age, and Southeast Asian children tend to have delayed bone age, compared with their White peers [21,22].

The bone age determination informs estimates of the child's growth potential and likely adult height, as follows:

●Delayed or advanced bone age is defined as a bone age that is 2 SD or more below or above the mean, respectively. This is approximately 20 percent below or above the chronologic age. This translates to a difference between bone age and chronologic age of approximately 12 months between 2 and 4 years of chronologic age, 18 months between 4 and 12 years, and 24 months after age 12 (figure 8A-B). If the bone age is delayed or advanced near or beyond these parameters, then the projected height should be recalculated based on the bone age rather than the chronologic age. This will provide a more accurate assessment of projected height. As an example, if an eight-year-old boy is 117 cm tall and has a bone age of 6.5 years, this corresponds to the 3^rd percentile for chronologic age but to the 35^th percentile for skeletal age, suggesting that the child may have constitutional delay of growth.

●The bone age can be used to predict the child's adult height. The technique developed by Bayley-Pinneau (BP) is most commonly used for children approximately six years and older [23]. This technique employs a table to translate the child's bone age and chronologic age to a decimal fraction of adult height (eg, 0.75). To predict adult height, the present height is divided by the fraction of adult height.

Other methods use different algorithms to predict adult height from height measurements and bone age data and include the TW3, Roche-Wainer-Thissen (RWT), and Khamis-Roche (KR) methods. RWT and KR incorporate midparental height in addition to bone age, and TW3 uses a skeletal maturity score. Comparison of these methods reveals a wide range of adult height predictions. Overall, the TW3 method (or the previous version, TW2) either underpredicts or overpredicts adult height in approximately 30 percent of individuals [24,25], while the BP and RWT methods tend to overpredict adult height in boys [26]. The BP method is most likely to identify a short child as a candidate for growth hormone therapy [27]. Methods that incorporate bone age are generally more accurate in predicting adult height than the simple midparental height method. However, there is a wide variation in height prediction, and the same method may yield significantly different adult height predictions at different ages [26]. Investigational techniques include automated interpretation of radiographs [28], or use of quantitative ultrasound to assess bone age, which would have the advantage of avoiding ionizing radiation [29].

The results of the bone age determination also provide important information about possible causes of the short stature (algorithm 1):

●A significantly delayed bone age is consistent with constitutional delay of growth and puberty (CDGP), which is considered a variant of normal growth. However, significantly delayed bone age is also seen in many types of pathologic growth failure, including nutritional deficiency, underlying systemic disease (such as inflammatory bowel disease), and growth hormone deficiency. The HV helps to distinguish among these categories: Children with CDGP tend to have normal or low-normal HV until they reach bone age of 11 years in girls or 13 years in boys. By contrast, children with underlying systemic or endocrine disease tend to have progressive decreases in HV. Children with significantly delayed bone age should undergo a focused history and physical examination to evaluate for symptoms and signs of systemic or endocrine disease (see 'Additional evaluation for causes of short stature' below). Their growth should also be carefully monitored.

●A normal bone age is consistent with several diagnostic possibilities: In a child with short parents, a normal bone age supports the diagnosis of familial short stature. However, a normal bone age is also seen in younger girls with Turner syndrome. Moreover, bone age may be only mildly delayed in early or mild forms of some of the systemic diseases that cause growth failure. Therefore, a bone age that is within normal limits suggests that underlying genetic or systemic disease is unlikely but not impossible.

●Advanced bone age is occasionally seen in older children and adolescents with short stature, especially those with precocious puberty and hyperthyroidism. These children usually experience accelerated early growth but are at risk for early epiphyseal closure, resulting in short stature as an adult if not properly diagnosed and treated. Children with autosomal dominant short stature due to aggrecan mutations also have short stature with advanced bone age [30]. (See "Causes of short stature", section on 'Sexual precocity' and "Causes of short stature", section on 'Skeletal dysplasias/growth plate abnormalities'.)

ADDITIONAL EVALUATION FOR CAUSES OF SHORT STATURE — In addition to the evaluation of growth outlined above, a focused history and physical examination contribute information that helps to categorize the cause of short stature (table 1 and algorithm 1). The key information is organized here according to the diagnostic category of the short stature. Further details about each cause of short stature are presented in the linked topic review. (See "Causes of short stature".)

Are there features that suggest that this is a normal variant of short stature? — The two most common causes of short stature are familial (genetic) short stature and constitutional delay of growth and puberty (CDGP; also termed constitutional short stature for prepubertal children). These growth patterns often can be distinguished from one another, but some children have features of both (table 2). Unless there are signs or symptoms of underlying disease, children with these growth patterns usually require only a basic evaluation, including a focused history, physical examination, and bone age determination. (See 'Laboratory and imaging studies' below.)

●In familial short stature, height velocity (HV) and bone age are within the normal range and one or both parents are short. (See "Causes of short stature", section on 'Familial short stature'.)

●In CDGP, the child's growth is appropriate for their bone age, which is delayed compared with chronologic age. If the height is plotted using the bone age rather than chronologic age, the projected height is within the range predicted for the biologic family and often within the normal range for the population (ie, adult height that is ≥63 inches [160 cm] for males and ≥59 inches [150 cm] for females). There is often a family history of delayed growth and/or puberty (a parent who was a "late bloomer"). (See "Causes of short stature", section on 'Constitutional delay of growth and puberty'.)

Are there features suggesting pathologic growth failure? — The history, review of systems, and physical examination should include the following elements to assess for a variety of pathologic causes of short stature (table 1 and algorithm 1). The findings may influence selection of laboratory tests and/or imaging.

Features suggesting underlying systemic disease — A variety of systemic diseases is associated with attenuated growth during childhood, usually with delayed bone age. This is particularly true for inflammatory disease processes (such as Crohn disease or juvenile idiopathic arthritis), diseases that cause malabsorption or malnutrition, or those that increase energy needs (cardiac or immune defects). Therefore, a complete medical history and review of systems are important to the assessment of short stature.

Key elements of the history include (see "Causes of short stature", section on 'Systemic disorders or processes with secondary effects on growth'):

●Gastrointestinal symptoms, including appetite, abdominal pain, diarrhea, and rectal bleeding – Suggests the possibility of Crohn disease or celiac disease.

●Pulmonary symptoms, including severe asthma, recurrent infections – Suggests the possibility of cystic fibrosis or immunodeficiency.

●Recurrent infections – Suggests the possibility of immunodeficiency; recurrent otitis media with the need for myringotomy tubes is associated with Turner syndrome.

●Arthralgia or arthritis – Consistent with inflammatory bowel disease, rheumatic diseases (eg, juvenile idiopathic arthritis), or celiac disease.

●Medications – Prolonged or frequent use of glucocorticoids (including inhaled glucocorticoids) may contribute to growth failure (see "Causes of short stature", section on 'Glucocorticoid therapy'). Use of stimulants for attention deficit hyperactivity disorder also has been associated with mild delay in growth, although this effect usually is transient [31]. (See "Causes of short stature", section on 'Systemic disorders or processes with secondary effects on growth' and "Pharmacology of drugs used to treat attention deficit hyperactivity disorder in children and adolescents", section on 'General adverse effects'.)

Key elements of the physical examination include:

●Weight loss, poor weight gain, underweight for height, and/or delayed puberty – These findings are consistent with many underlying systemic diseases, psychosocial deprivation, or food restriction. By contrast, most endocrine causes of short stature are associated with overweight for height.

●Oral ulcers or large anal skin tags – These findings are common in Crohn disease and may be the presenting symptoms.

Features suggesting genetic or endocrine disease — Endocrine disorders are relatively uncommon causes of short stature but important to diagnose because they respond to specific treatment(s). These endocrine disorders are often characterized by delayed bone age. An exception is precocious puberty, in which accelerated early growth may be accompanied by early epiphyseal maturation and adult short stature.

Key elements of the history relevant to endocrine disease include (see "Causes of short stature", section on 'Endocrine causes of short stature'):

●Sluggishness, lethargy, cold intolerance, constipation – These symptoms suggest hypothyroidism.

●Developmental delay/learning disabilities – Problems with nonverbal learning skills are common in Turner syndrome. Developmental delay is common in Noonan or Russell-Silver syndrome and in pseudohypoparathyroidism. Acquired hypothyroidism is often associated with altered school performance. Many syndromes with developmental delay also include short stature, such as Down, Prader-Willi, and Bloom syndromes.

●Neuropsychological changes – Symptoms of psychiatric disease occur in over one-half of patients with Cushing syndrome of any etiology.

Key elements of the physical examination include:

●Increased weight for height – Obesity is nearly universal in Cushing syndrome (with central fat distribution). Increased weight for height is also consistent with hypothyroidism, growth hormone deficiency, or pseudohypoparathyroidism.

●Facial dysmorphism

•Hypertelorism, downward eye slant, low-set ears – Noonan syndrome

•Prominent forehead, triangular face, downturned corners of the mouth – Russell-Silver syndrome

•Midface hypoplasia, frontal bossing – Achondroplasia

•Midline defects – Associated with hypothalamic-pituitary hormone deficiencies

•Mild dysmorphism also may occur in a heterozygous skeletal dysplasia

●Optic discs – Papilledema suggests a central nervous system mass effect. Optic nerve hypoplasia suggests septo-optic dysplasia, which is associated with pituitary hormone deficiencies. In optic nerve hypoplasia, the optic disc is small, often pale, and surrounded by a yellowish halo, bordered by a ring of pigmentation (double-ring sign) (picture 1).

●Neck and chest

•Goiter – Hypothyroidism

•Webbed neck, shield chest (picture 2) – Characteristic findings in Turner syndrome

•Webbed neck, pectus excavatum – Seen in Noonan syndrome

•Suprascapular fat pad (buffalo hump) and supraclavicular fat pads – Suggests Cushing syndrome. Mild forms of this fat distribution are seen in simple obesity (sometimes termed "pseudo-Cushingoid"), but exogenously obese children are often of normal or slightly increased stature.

●Limbs

•Cubitus valgus (increased carrying angle of the arm), genu valgum – Commonly seen in Turner syndrome or SHOX gene (short stature homeobox) mutations

•Madelung deformity of the forearm (focal dysplasia of the distal radial physis, leading to a prominent ulna and wrist pain) (picture 3 and image 1) – Commonly seen in Turner syndrome or SHOX mutations

•Trident hands (broad, space between middle fingers) – Achondroplasia

•Stocky build – Often seen in Turner syndrome or SHOX mutations

•Long limbs compared with trunk – Spondyloepiphyseal dysplasia

•Short limbs (especially upper arms) compared with trunk – Achondroplasia

•Subtle body disproportion – Either long or short limbs compared with the trunk may suggest a heterozygous skeletal dysplasia; this can be detected by measuring the sitting height-to-standing height ratio [32] (See "Skeletal dysplasias: Approach to evaluation", section on 'Physical examination'.)

●Skin – Atrophied skin and purple-colored striae are characteristic of Cushing syndrome, sometimes with hyperpigmentation.

●Delayed or accelerated puberty or hypogonadism – Most women with Turner syndrome have absent breast development; however, some have delayed or partial pubertal maturation. Hypothyroidism tends to cause pubertal delay, although, on rare occasion, this disorder may present with precocious puberty. Early puberty is occasionally found in children with congenital virilizing adrenal hyperplasia or Cushing disease. Microphallus or cryptorchidism suggest central hypothalamic-pituitary deficiencies. (See "Approach to the patient with delayed puberty".)

●Decreased deep tendon reflexes – Suggests hypothyroidism.

LABORATORY AND IMAGING STUDIES — There is some controversy about the extent of testing that should be performed for children and adolescents with presumed idiopathic short stature (ISS). In otherwise healthy children, the yield of testing is extremely low. One study described the clinical evaluation of 235 children who were referred for short stature and had height <3^rd percentile but normal growth rate and no symptoms [33]. In this group of low-risk short children, nearly 99 percent were ultimately diagnosed with variants of normal growth (23 percent with familial short stature, 41 percent with constitutional delay of growth, and 36 percent with ISS). The cost per new diagnosis identified was more than $100,000, although most patients did not undergo all of the screening tests that have been recommended in consensus guidelines [11]. Because of the low diagnostic yield and high cost, we suggest performing laboratory testing only selectively in otherwise asymptomatic children, as outlined below.

●Basic evaluation – A bone age determination is appropriate for children with short stature, normal growth rate (eg, height velocity [HV] at least 5 cm/year between four and six years of age and at least 4 cm/year between six years and puberty), and no other symptoms. The bone age determination provides a more accurate adult height prediction and helps to clarify the type of growth defect, as described above (see 'Is there evidence of delayed or accelerated growth?' above). Screening for celiac disease is also reasonable for children with gastrointestinal symptoms or first-degree relatives with celiac disease.

●Broader testing – Additional testing may be warranted if the child is severely short (eg, height ≤-2.5 standard deviations [SD], ie, ≤0.6^th percentile [34]), has growth failure (eg, height-for-age curve crossing two major percentile lines, or HV <5 cm/year between four and six years of age and <4 cm/year between six years and puberty), or if the history or physical examination raised suspicion for a specific systemic, endocrine, or genetic disorder. Decisions regarding the extent of testing should be made in conjunction with a pediatric subspecialist but vary with the child's presenting symptoms and the clinical setting. We perform the following battery of tests to screen for several pathologic causes of short stature (table 1), as outlined in a consensus statement [11]:

•Complete blood count (CBC) and erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP).

•Electrolytes, creatinine, bicarbonate, calcium, phosphate, alkaline phosphatase, albumin.

•Celiac serologies (eg, tissue transglutaminase [tTG] immunoglobulin A [IgA] and total IgA) – Several guidelines recommend celiac screening for children with short stature, as well as for other indications. This test is relatively cost-effective and screens for a common disease. Some celiac disease test panels include a measurement of total IgA to exclude the unlikely possibility of a false-negative test in a patient with selective IgA deficiency. (See "Diagnosis of celiac disease in children", section on 'Indications for testing'.)

•Thyroid-stimulating hormone (TSH), free thyroxine (T4), insulin-like growth factor 1 (IGF-1), and insulin-like growth factor binding protein-3 (IGFBP-3). IGFBP-3 has higher sensitivity for predicting the diagnosis of growth hormone deficiency in children <10 years of age as compared with IGF-1.

•Karyotype or chromosomal microarray should be performed in all girls to rule out Turner syndrome and in boys with associated genital abnormalities.

●Additional tests based on clinical features

•Suspected growth hormone deficiency – Patients with reduced HV, low IGF-1 and/or IGFBP-3, and delayed bone age should be evaluated for growth hormone deficiency using provocative tests. (See "Diagnosis of growth hormone deficiency in children".)

•Confirmed pituitary hormone deficiency – Contrast-enhanced magnetic resonance imaging (MRI) of the brain with and without contrast enhancement is appropriate for children after diagnosis of growth hormone and/or other pituitary hormone deficiencies.

●Genetic testing – Genetic testing is increasingly available and is a valuable and cost-effective component of the evaluation for selected patients [35-38]. Indications for genetic testing include:

•Severe short stature (height below -3 SD)

•Multiple pituitary hormone deficiencies

•Severe growth hormone deficiency

•Unequivocal growth hormone insensitivity

•Children born small for gestational age who do not experience catch-up growth

•Additional congenital anomalies or dysmorphic features

•Evidence of a skeletal dysplasia

•Associated intellectual disability or microcephaly

(See 'Are there features suggesting pathologic growth failure?' above and "Causes of short stature", section on 'Pathologic causes of short stature' and "Skeletal dysplasias: Approach to evaluation".)

For these children, an efficient approach is to use a panel designed to detect genetic disorders associated with short stature, selected based on the child's clinical features, such as growth hormone deficiency, skeletal dysplasia, syndromic features (including intellectual disability or autism), or isolated severe short stature. As an example, to evaluate children with severe growth hormone deficiency or multiple pituitary hormone deficiencies, a panel can test for pathogenic variants in genes encoding relevant pituitary transcription factors, hormones, binding proteins, and receptors relevant to the growth hormone-IGF-1 axis. A list of panels and laboratories is available at the Genetic Testing Registry website. This approach tests simultaneously for pathogenic variants in multiple genes. However, it usually detects copy number variants but not specific sequence variants.

●Considerations for ISS – Most children who are evaluated for short stature do not have one of the above indications for genetic testing. If the evaluation excluded familial short stature and constitutional delay of growth and puberty (CDGP) and did not identify a specific cause, they are considered to have ISS (see "Causes of short stature", section on 'Idiopathic short stature'). Although ISS is generally considered a normal variant of growth, a subset of children has mild or subtle forms of a skeletal dysplasia. Therefore, it is reasonable to offer testing with a gene test panel for skeletal dysplasias in children with ISS and especially if the child has any evidence of body disproportion (eg, elevated sitting height-to-height ratio). When selecting a commercial test panel, ensure that it includes ACAN, NPR2, PTPN11, and SHOX. A heterozygous variant in one of these genes is a relatively common cause of subtle skeletal dysplasia, but these genes are not included in all skeletal dysplasia panels.

A decision to perform genetic testing should be made collaboratively with the family. Considerations include the potential benefits of identifying a genetic diagnosis (eg, directing therapy, further evaluation for associated comorbidities, and potential for identifying other affected family members), relatively low yield for children with mild isolated short stature, and potential adverse impact on the family (cost, anxiety, and feeling less normal). Genetic counseling before and/or after testing may be helpful.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Turner syndrome" and "Society guideline links: Growth hormone deficiency and other growth disorders".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: My child is short (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Definition – Short stature is defined as height that is 2 standard deviations (SD) or more below the mean height for children of that sex and chronologic age in a given population. This translates to a height that is below the 2.3^rd percentile. The clinical significance of the short stature depends on many factors, including genetic potential and changes in stature over time (height velocity [HV]). (See 'Is the child short?' above.)

●Initial evaluation

•Height velocity – Determination of the child's growth, or HV, is an essential component of the evaluation for short stature. For children two years and older, growth failure is suggested by a growth pattern that has deviated downward across two major height percentile curves or by growth slower than the following rates:

-Age two to four years – HV less than 5.5 cm/year (<2.2 inches/year)

-Age four to six years – HV less than 5 cm/year (<2 inches/year)

-Age six years to puberty:

HV less than 4 cm/year for boys (<1.6 inches/year)

HV less than 4.5 cm/year for girls (<1.8 inches/year)

(See 'Is the child's height velocity impaired?' above.)

•Height prediction – An estimate of a child's adult height potential can be obtained by calculation of the midparental height (target height), adjusted for the sex of the child (calculator 4). For males and females, 8.5 cm (3.3 inches) on either side of this calculated value represents the 3^rd to 97^th percentiles for anticipated adult height. (See 'Is the child's growth within the range for the family?' above.)

•History and physical examination – The history and physical examination should include (see 'Additional evaluation for causes of short stature' above):

-Family history of growth and pubertal onset

-Review of systems for features suggestive of gastrointestinal, pulmonary, immunologic, or other systemic disease

-Dysmorphic features, especially webbed neck, cubitus valgus, and absent puberty in girls (suggests Turner syndrome), or disproportionate short stature (eg, short limbs compared with trunk)

●Clinical and laboratory testing – Laboratory evaluation for a child with short stature depends on the results of the above evaluation (algorithm 1).

•Basic evaluation – For children with short stature, normal growth rate (eg, HV at least 5 cm/year between four and six years of age and at least 4 cm/year between six years and puberty), and no other symptoms, we suggest a basic evaluation including bone age determination. The bone age result can then be used to refine the estimate for the child's adult height and also informs the evaluation for possible causes of short stature. (See 'Bone age determination' above and 'Laboratory and imaging studies' above.)

•Further evaluation – Children with severe short stature (eg, height ≤-2.5 SD [0.6^th percentile]) or growth failure should be further evaluated with a complete blood count (CBC), erythrocyte sedimentation rate (ESR), tissue transglutaminase (tTG) immunoglobulin A (IgA), creatinine, electrolytes, thyroid-stimulating hormone (TSH), free thyroxine (T4), insulin-like growth factor 1 (IGF-1), and insulin-like growth factor binding protein-3 (IGFBP-3). A karyotype or comparative genomic hybridization should be performed in all girls with unexplained short stature (to evaluate for Turner syndrome) and in short boys with associated genital abnormalities. Abnormal results of these tests and/or symptoms are signs that a disorder causing growth failure warrants further investigation. (See 'Laboratory and imaging studies' above and "Causes of short stature", section on 'Endocrine causes of short stature'.)

•Genetic testing – Genetic testing using gene panels is valuable for selected children, including those with severe short stature, syndromic features, or body disproportion (suggesting a skeletal dysplasia). It may also be appropriate for some other children with short stature who manifest subtle phenotypes consistent with skeletal dysplasias. (See 'Laboratory and imaging studies' above.)

●Causes of short stature

•Normal variants – The two most common causes of short stature are familial (genetic) short stature and constitutional delay of growth and puberty (CDGP), which are variants of normal growth. These growth patterns often can be distinguished from one another, but some children have features of both (table 2). (See 'Are there features that suggest that this is a normal variant of short stature?' above.)

•Pathologic causes – Important pathologic causes of growth failure that may present with short stature and/or delayed puberty include Crohn disease, celiac disease, and Turner syndrome (table 1). (See 'Are there features suggesting pathologic growth failure?' above.)