7.2.8 Tall stature

The term tall stature simply means that the child’s height is above the 97th percentile (corresponding to a standard deviation score (SDS) of +1.88). It says nothing about the underlying cause and is not itself a growth abnormality; indeed, most children with tall stature are normal. Although as many children have heights greater than 2 SD above the mean as have heights greater than 2 SD below the mean, tall stature in childhood usually generates less anxiety than shortness and, therefore, referral for tall stature is less common than it is for short stature.

The initial approach to tall stature is similar to that used for short stature (Box 7.2.8.1). It is important to assess the child’s height against standard growth charts, and to decide whether the child has normal body proportions or not. Where disproportionate growth occurs, it is more likely that there is an underlying genetic syndrome. Furthermore, it is necessary to establish whether or not one or even both parents are tall, and if so whether there is a dominant growth abnormality such as Marfan’s syndrome. Some children clearly have normal genetic tall stature in others pubertal development has been early and some have a combination of both features. Weight needs to be measured as well. Children with simple obesity have heights in the upper half of the target range centiles, so that if parents are moderately tall, the child may well be above the 97th centile for height. Importantly, children whose tallness is out of context with the family patterns and in whom obesity cannot be invoked as the main cause, must be assessed in terms of dysmorphic features, as well as learning difficulties and sexual precocity.

Children with familial or constitutional tall stature have predicted adult height (1) in keeping with their midparental height. Midparental height is caculated as the average of parents' heights, minus 6.6cm for girls, or plus 6.5cm in boys (although +/- 7 cm may be used in the UK). If possible, parental heights should be measured rather than reported. Further, height prediction plays a key role in the management of children with growth disorders and any possible medical treatment is based on the estimated height prognosis. Therefore, accurate height prediction methods are essential. In clinical practice the methods of Bayley and Pinneau (1), and Tanner (2) are most commonly used. The first prediction method is based on bone age (BA) assessment developed by Greulich and Pyle, the second on the method described by Tanner. However, to date, only the Tanner–Whitehouse mark II equation has included samples of tall children (2, 3). Thus, applying standard equations for the determination of adult height to children with growth disorders may not give accurate results. Joss et al. studied the accuracy of height predictions at various ages based on five different prediction methods (4). They concluded that there is no best or most accurate method for predicting adult height in tall children. The method of choice differs with respect to sex and BA (4–7). In addition, correcting factors may improve the accuracy of prediction (4). A similar study performed by De Waal et al. concluded that, in boys and girls, the most reliable prediction method is to extrapolate height SDS for BA determined by Greulich and Pyle (Index of Potential Height; IPH) (6, 7). Furthermore, a new model to predict final height in constitutionally tall children has been reported, but the clinical validity has not been ascertained in larger groups of tall children (8). However, it is important to stress that height prognosis is more accurate when based on growth data derived from tall children (7). In addition, in the third edition of the ‘Assessment of skeletal maturity and prediction of adult height (TW3-method)’ considerable changes have been made taking into account the secular trend in many countries towards rapid maturation (9). In conclusion, although these methods may have small mean errors of prediction, the error in an individual case may be considerable, even so, over recent years no new data set using updated BA and prediction models have been published. Therefore, predicted adult height should be given as height with a confidence limit using the residual SDS of the prediction technique for calculation.

Although most children with tall stature are normal and healthy, a careful clinical evaluation is indicated because tall stature might be part of a disease and/or syndrome. Clinical algorithms may help in making a diagnosis and, in Box 7.2.8.2, a list of the differential diagnoses of tall stature is presented. In this chapter the disorders are considered in three categories:

The assessment of height velocity is crucial in the follow-up of a child presenting with any growth disorder. Height velocity is seldom below the 50th centile in tall stature (10). A child with a height velocity greater than the 97th percentile over 1 year should be investigated immediately, whereas children with velocities between the 75th and 97th percentile should be carefully followed. A height velocity on the 75th percentile over 2 years decreases the chance of being normal to 5%. Thus, children growing fast without any signs of puberty have to be observed closely. Adrenarche regularly presents in tall children with an accentuation of the mid-childhood growth spurt, which may cause concern. However, this accentuation of the mid-childhood growth spurt is self-limiting and is usually combined with an acceleration of BA. Therefore, height for BA remains unchanged, or may even decrease slightly.

The diagnosis is generally made from the family history record of stature and physical examination. Although growth hormone and insulin-like growth factor 1 (IGF-1) and IGF-binding protein-3 (IGFBP-3) status in familial tall stature is often in the upper range of normal, both enhanced secretion of growth hormone and greater efficiency of growth hormone-mediated IGF-1 production might be potential causes of familial tall stature (7, 11).

These overgrowth syndromes are not always sharply defined and comprise a group of disorders with the following common characteristics: increased birthweight and length, overgrowth with advanced BA maturation during the early years of life, and mental retardation.

Accelerated growth is observed in Sotos syndrome only in the first 2 years of life, followed by normal growth at or above the 97th centile throughout childhood, and early adolescence reaching final height within normal range with advanced osseous maturation in childhood. In 50% macrocephaly is of prenatal onset, but it is present in 100% by the age of 1 year. Facial features may resemble acromegaly and mental retardation is variable. This disorder follows an autosomal dominant inheritance pattern, caused by mainly de novo mutations. Mutations in the nuclear receptor SET-domain-containing protein (NSD1) located on chromosome 5q35 are responsible for most of the cases (12).

Accelerated growth and maturation (advanced BA) is common with weight more significantly increased than height. The head circumference may also be increased (reported in 83%), although the forehead is not as prominent as that seen in Sotos syndrome. Features include: camptodactyly, large head and a small but prominent chin. Development may be mildly delayed. It has been suggested that some cases are due to mutations of NSD1, which is the major cause of Sotos syndrome.

Accelerated linear growth and markedly accelerated skeletal maturation, a tendency to be underweight, shallow orbits and broad middle phalanges are the main clinical signs. Developmental delay is the rule and death in the first 2 years of life (from respiratory problems) is reported in most patients, who may have choanal atresia, laryngomalacia, and cerebral atrophy. It is rare, with about 40 cases in literature to date.

Macroglossia, macrosomia, omphalocele as well as neonatal hypoglycaemia are the main abnormalities of this syndrome. Hepatomegaly and cardiovascular defects including isolated cardiomegaly may occur. There is an association with embryonal tumours, mainly Wilms' tumour. Although sporadic, autosomal dominant inheritance with preferential maternal transmission has occurred in about 10–15% of cases. BWS is caused by perturbations of the normal dosage balance of a number of genes clustered on chromosome 11p15, a highly imprinted region in the genome. This region contains paternally expressed insulin-like growth factor 2 (IGF-2), as well as a number of genes and transcripts that control expression of IGF-2 (13). Cytogenetically detectable abnormalities involving 11p15 are found in 1% or less of cases. Clinically available molecular genetic testing can identify several different types of 11p15 abnormalities in individuals with BWS: (1) loss of methylation at DMR2 is observed in 50% of individuals; (2) gain of methylation at DMR1 is observed in 2–7%; (3) paternal uniparental disomy for chromosome 11p15 is observed in 10–20%. Testing reveals mutations in the CDKN1C gene (previously called p57 KIP2) in 40% of familial cases and 5–10% of sporadic cases (individuals with no known family history of BWS).

The diagnosis is based on criteria established in 1996, in which growth pattern is not included (14). The impression of long extremities is exaggerated by poor muscular mass and arachnodactyly. The condition is inherited in an autosomal dominant manner with complete penetrance, but highly variable phenotype even within families. Marfan’s syndrome is caused by abnormalities in the fibrillin (FBN1) gene located on the long arm of chromosome 15. The prevalence is 1: 10 000. Eye (lens subluxation) and heart abnormalities (dilatation with or without dissecting aneurysm) are characteristic of familial Marfan’s syndrome. Height-limiting therapy may be considered because of the excessively tall stature and it may also arrest kyphosis, as well as scoliosis.

In many (∼75%) of these patients there is a marfanoid habitus with tall stature. Additionally, in this syndrome, there is a combination of phaeochromocytoma, medullary thyroid carcinoma (MTC) and mucosal neuroma of the gastrointestinal (GI) tract. Although the association of phaeochromocytoma with neurofibromatosis is well known, the tumours in MEN 2B are true neuromas, consisting mainly of nerve cells. The patients sometimes have cafe-au-lait spots. MEN 2B is caused by alterations of the RET proto-oncogene at chromosome 10q11.2 (15).

This condition, characterized by skeletal abnormalities, including excessive height and length of limb, can be easily diagnosed by the detection of homocysteine in a urine sample. It is an autosomal recessive disorder caused by a deficiency of the enzyme cystathionine beta-synthase (chromosome 21q22.3) producing increased urinary homocysteine and methionine. It is further characterized by mental deficiency, subluxation of the lens, osteoporosis, arterial, and venous thrombosis. Mild hypercystinaemia may be caused by methylenetetrahydrofolate reductase deficiency due to mutations in the 5,10-alpha-methylenetetrahydrofolate reductase gene; MTHFR gene.

Klinefelter's syndrome is characterized by a 47,XXY karyotype (variants; 48,XXXY; 49,XXXXY) and has a prevalence of 1 in 600 males. Tall stature may be seen even before puberty in individuals with karyotypes 47,XXY and 48,XXYY. Disproportionately long limbs may develop during puberty. Furthermore, XYY syndrome is reported with a prevalence of 1/1000 in the normal population, but most boys with 47,XYY karyotype remain undiagnosed.

Gigantism in childhood/adolescence is rare. In most instances, it is caused by a growth hormone-secreting pituitary adenoma. Serum growth hormone is elevated and cannot be suppressed by a glucose load. Circulating IGF-1 levels are also elevated compared with age and puberty-matched normative data.

Growth acceleration is seen in children with hyperthyroidism if it remains untreated. The BA is advanced.

Commonly obese children are tall for age, associated with advanced skeletal maturity and early onset of puberty. Basal growth hormone levels and responses to stimulation tests are attenuated. IGF-1 levels are normal or slightly increased.

Precocious puberty, either central or pseudoprecocious puberty results in tall stature in comparison with normal children of the same chronological age. BA is of course advanced.

As with other disorders of growth, the earlier treatment is introduced, the better the outcome. Two important facts need to be stressed, first, that children gain between 25 and 30 cm in height during puberty and, secondly, the amount of height gained during puberty is relatively resistant to any form of manipulation. Therefore, any treatment, initiated to reduce final height, should start at a height 25–30 cm less than that desired, otherwise little effect on the final height will be achieved. For instance, a girl who is prepubertal at an age of 10 years and has a height of 155 cm will attain a final height in the region of 180–185 cm, assuming that the girl enters puberty within that year.

As a height on the 97th centile varies substantially among various populations (the Scandinavians and the Dutch are among the tallest in the world), the height, which is acceptable in adult life for a given child may vary between countries and is a matter of opinion (7). In general, for boys and their families, heights up to 200 cm are well accepted, whereas for girls, heights greater than 180 cm are often unacceptable. Therefore, in contemplating the question ‘Does the child need treatment’, one needs to determine if the predicted height is acceptable to the child’s family and its peer group. Importantly, the general acceptance of a child within their peer group should never be underestimated. Excessively tall stature may be associated with genuine suffering and psychological problems, including difficulties with self-image and long-term self-esteem. However, there is a paucity of studies focusing on the psychological effects of tallness and its therapy. Tall girls report frequent teasing, and this is probably one of the most compelling reasons for seeking medical therapy. In males, practical issues, such as availability of clothing and driving a car, are factors influencing the decision to treat. Binder et al. reported that overall in their study no major psychosocial difference between treated and untreated subjects could be revealed.

Management of tall children and adolescents remains among the more controversial topics in paediatric endocrinology. Generally, treatment has been aimed at inducing incomplete precocious puberty and accelerating the rate of development of secondary sex characteristics. This concept of treating tall adolescents to reduce final height developed from the work of Albright et al. who used oestrogen in adult patients with acromegaly (16). Thereafter, the use of oestrogen was first introduced and reported by Goldzieher (17). The rational was based on two clinical observations: (1) children with precocious puberty become small adults if premature epiphyseal closure occurs; (2) epiphyseal closure is delayed in the absence of sex steroids. A variety of preparations and dose schedules have been employed over the years. These include injectable oestradiol valerate, implanted oestradiol pellets, oral stilbestrol (3 mg/day), diethylstilbestrol (5 mg/day), ethinyl oestradiol (100–500 µg/day) and conjugated oestrogens (2.5–10 mg/day). In practice, most paediatric endocrinologists use either continuous ethinyl oestradiol (100 µg/day) or conjugated oestrogens (7.5 mg/day), together with progesterone for 7–10days (day (15/18) – 25) at the end of each monthly cycle to promote endometrial shedding (18). Although, there is general agreement that a favourable effect on ultimate height results from such pharmacological therapy, the exact age at which oestrogen treatment should be started and the predicted mature height that would serve as an indication for treatment still remain open questions. The decision to treat has to be judged within the social context, which changes over time (19). Importantly, the eventual loss in stature should not be overestimated. Despite the fact that the earlier treatment starts the better the statural result, it would be unethical and incorrect to induce puberty in a 7- or 8-year-old girl just to reduce her final height (20, 21).

Short-term side effects of oestrogen administration include nausea, weight gain, pigmentation, leg cramps, and transient hypertension. However, it is important to emphasize that very few serious side effects have been reported using even high doses of oestrogens. Thromboembolism is a potential hazard (22). Tall girls on pharmacological doses of oestrogen need close attention paid to the possibility of clotting disorders. Antithrombin III levels might be useful as a screening factor with more sophisticated clotting factor analysis if a positive family history for clotting disorders exists.

In the management of tall girls, auxology should be performed every 3 months. When the height remains unchanged between two successive measurements, radiography of the left hand can be performed to determine epiphyseal fusion, bearing in mind that this may not reflect fusion at other skeletal sites. The overall average decrease in ultimate height varies between 3.6 and 7.6 cm (19). After cessation of therapy a mean (SD) additional height gain of 2.7 (1.1) cm has been observed, which is of the same magnitude in boys (2.4 cm) (6, 7). Generally, menses return promptly after discontinuing oestrogen administration and subsequently fertility has proved to be normal in treated women (7). Later effects of pharmacological oestrogen administration on gonadal function, genital tract and breast neoplasia remain uncertain (23, 24). However, in a recent article focusing on oestrogen ‘induced’ carcinogenesis the weight of evidence indicates that exposure to oestrogen may be an important determinant of the risk of breast cancer (25, 26). Finally, in terms of patient perception, Weimann et al. reported that 84.6% of the previously treated tall women were satisfied with the outcome associated with a mean height reduction of 5.2 cm. Only 15.4% regretted the therapy (27).

Experience with testosterone treatment in boys is limited, because fewer boys complain of tall stature. In the United States, for instance, tall stature is rarely a cause for complaint among adolescent boys. The intramuscular administration of a long-acting depot-form of testosterone (for example, testosterone enanthate) 500 mg every 2 weeks has produced a significant reduction in predicted final height (7). In contrast to the treatment in girls, testosterone may lead to an acceleration of height velocity in the first 3 months, which might cause concern in the treated boys. One major side effect is severe acne, which may need treatment. Finally, despite theoretical concerns, there are reports stating that high doses of long-acting testosterone esters (such as propionate, enanthate, and decanuate) at puberty for tall stature do not impair testicular function on a long-term basis (7, 28). What is the effect in terms of height reduction? As stated above, the ‘uncorrected’ effect of height reductive therapy, i.e. height prediction minus achieved adult height varies with the prediction method applied. Since every single prediction method has its own prediction error, the mean effect may be ‘corrected’ by subtraction of the corresponding mean prediction errors (7). The Bayley–Pinneau prediction showed the greatest mean ‘corrected’ effect of 2.0 cm, while the IPH, being the most accurate method, calculated a mean ‘corrected’ effect of only 0.6 cm (7). Therefore, the ‘corrected’ reductions reported are between 4.7 and 7.5 cm. In addition to the variability in adult height prediction methods, differences in results may be due to study design, comparability of the control group, inclusion criteria (such as age and BA at start of therapy), and therapeutic regimen. It has been clearly shown that height reduction was dependent on the BA at start of therapy: height reduction was more pronounced when treatment was started at a younger BA (7). However, an important issue that caused a significant reduction in the height-limiting effect was the observation of marked additional post-treatment growth after cessation of therapy. This post-treatment growth might partly be explained by late pubertal completion of spinal growth. On the other hand, the additional growth could result from the fact that treatment had been stopped before complete closure of the epiphyses. A significant negative relationship between post-treatment growth and BA at the time of stopping therapy (r² = 0.53; P = 0.001) was observed. The latter contrasts with the opinion of Brämswig and co-workers (29), who advocated short-term therapy and reported significant height reduction (uncorrected: 7.6 cm) with a mean BA (SD) of 15.3(0.8) year at the time of stopping therapy. However, these results have been disputed by Bettendorf et al. (30). Most important, however, is the fact that when therapy was started at a BA of 14 years or older, adult final height significantly exceeded height prognosis at the time of starting the treatment. This suggests that treatment had resulted in induction, rather than reduction of growth.

Finally, there are several reasons for seeking alternative treatment regimens to reduce final height; first the anxiety about side effects of sex steroids, and secondly, the inappropriateness of offering sex steroids to young children of prepubertal age; a treatment that reduces growth during early and mid-childhood is required.

Unlike dopamine agonist drugs, such as bromocriptine (31, 32) intravenous infusion of native somatostatin suppresses growth hormone secretion in both normal individuals and acromegalics, with a rebound of growth hormone secretion after cessation, which is less pronounced when long-acting somatostatin analogues (for example, octreotide) are administered (33). Different therapeutic regimens, for instance 250 μg octreotide twice daily (34); 35–50 µg once or twice daily (35); 50–100 µg infusion for 12 h overnight (36) and 60 µg as a single injection at night (personal data) have been reported. The effect of all these regimens, however, did not differ in the extent that predicted final height (1–7 cm) was reduced (34–36). Octreotide was without effect on fasting blood glucose, insulin, glycated haemoglobin, or serum thyroxine. Tolerability was good, except for the occurrence of gallbladder microlithiasis in one patient (34), whereas sludge in the gallbladder was a common finding. Data on bone maturation were contradictory. A number of important questions regarding the optimal use of octreotide in tall stature, such as timing, duration, and mode of administration remain still unanswered.

Furthermore, if growth hormone secretion is blocked, the likely reduction in growth velocity will be only about 50–60%. Therefore, it is unlikely that octreotide will be able to reduce final height substantially on its own. It is possible that a combination of octreotide to reduce childhood growth, followed by a treatment designed to blunt the pubertal growth spurt through alteration of the timing of pubertal development might be the treatment of choice. In addition, the knowledge of the interaction between growth hormone and its receptor has led to the rational design of growth hormone-receptor antagonists (37). These well-designed drugs may replace simple somatostatin analogue therapy. In adults with acromegaly it has been proven to be the most effective drug to normalize IGF-1 levels (38).

Most patient with gigantism (growth hormone excess) have pituitary macroadenomas with extrasellar extension and some of these subjects may be ‘cured’ following trans-sphenoidal surgery, but persistent growth hormone release from tumour remnants is often observed. Apart from re-operation, resulting in near or total hypophysectomy, alternatives include irradiation and the use of these growth hormone-secretion blocking drugs described.