Short and Tall Stature

Short or tall stature is primarily a normal variation of height. It is part of the continuum of the normal Gaussian distribution curve which defines the lower and upper limit of normal as the 3rd and 97th percentile. Within this context it is very important to differentiate normal variations in height and growth from pathological conditions. Normal variations in height are familial and idiopathic short or tall stature. Normal variations in growth are diagnosed as constitutional acceleration or constitutional delay of growth and puberty and are observed on all percentiles. Patients with pathological conditions of height usually have a syndromal, skeletal or chromosomal disorders such as Russell-Silver or Marfan syndrome, achondroplasia, Ullrich-Turner or Klinefelter syndrome. Many of the disorders with short stature have intrauterine growth retardation. The (mean) final height does not necessarily follow the target height range, but reaches a height below or above the population 3rd or 97th percentile. Patients with pathological conditions of growth can have short, normal or tall stature. They have a reduced or increased growth rate caused by a wide variety of chronic organic or psychosomatic diseases. At present, treatment with growth hormone (GH) is available for patients with GH deficiency, Ullrich-Turner syndrome, chronic renal insufficiency, small for gestational age and Prader-Willi syndrome. The indication for GH treatment of idiopathic short stature is so far only approved in the USA. Adult height data indicate that patients with documented GH deficiency reach an adult Prof. Dr. Jurgen H. Bramswig Albert Schweitzerstrasse 33 D–48129 Münster (Germany) Tel. +49 251 834 7730, Fax +49 251 834 7735 E-Mail [email protected] © 2008 Nestec Ltd., Vevey/S. Karger AG, Basel 0517–8606/07/0653–0117$23.50/0 Accessible online at: www.karger.com/ane D ow nl oa de d by : 54 .7 0. 40 .1 1 1/ 21 /2 01 8 6: 02 :2 8 A M Bramswig Ann Nestlé [Engl] 2007;65:117–127 118 growth [6–8] . The result of this growth process is a height within the range of the 3rd and 97th percentile or a height just deviating from normal as in short or tall stature. It is important to remember that in most instances normal height, short and tall stature are the result of normal growth. The ethnic variation in height is shown in table 1 for boys and girls. It demonstrates the large variability of height between populations. Normal adult height in the Netherlands is 168.3 cm for women and 182.0 cm for men [9, 10] . This height compares to a mean normal height of 157.9 and 170.0 cm in Japanese women and men, respectively. The difference in the normal mean adult height between these two populations is 10.4 cm for women and 12.0 cm for men. The mean height in the Netherlands is equivalent to tall stature in Japan. Short stature in the Netherlands is just below the mean normal height in Japan. The data in table 1 demonstrate the need to obtain population-specific growth charts to define short and tall stature. In a very height-conscious global society the variation in height is of national and international importance with respect to psychosocial adaptation. The answers to ‘How tall is too tall’ and ‘How short is too short’ depend on the height data obtained in each ethnic group. Extremes of height can be defined and accepted only when the normal variance in height of the reference population is known. For the majority of children, adolescents and adults a height below the 3rd or above the 97th percentile should be seen as part of a continuum of a normal Gaussian distribution curve. Only very few will have a defined abnormality, i.e. a height or growth disorder. Short Stature Short stature is defined as a height below the 3rd percentile. This means that 3% of the population have short stature. Short stature is defined as familial short stature when one or both parents have a height below the 3rd percentile. Children and adolescents with idiopathic short stature have parents with a height in the lower range of normal, but not below the 3rd percentile. Their short stature probably represents the normal variation in height which can be observed in many families in which some children are taller or shorter than their siblings. Familial and idiopathic short stature is best explained by estimating the target height according to Tanner [11] . Target height is the mean height of both parents adding 6.5 cm to obtain the target height for boys and subtracting 6.5 cm to define the target height for girls. The target height range is 8 8.5 cm, corresponding to 2 SD. The adult height of 94% of the patients with familial or idiopathic short stature should fall into this target height range. The majority of patients with familial short stature will have a final height below the 3rd percentile due to the short stature of their parents. The majority of patients with idiopathic short stature will have a final height above the 3rd percentile due to the normal stature of their parents [12] . Puberty should not be delayed but start within the normal age ranges that are known for the different stages of puberty [13, 14] . Familial or idiopathic short stature is diagnosed from previous height measurements which demonstrate growth below, but parallel to the 3rd percentile. Height measurements of both parents should be obtained and Table 1. Adult height (mean/median) in boys and girls of various countries Adult height in boys, cm Adult height in girls, cm mean/median SD +2 SD –2 SD mean/median SD +2 SD –2 SD The Netherlands 1985 182.0 6.70 195.4 175.3 168.3 6.20 180.7 162.1 Germany 1992 179.9 6.40 192.7 173.5 167.0 5.10 177.2 161.9 Sweden 1976 179.8 6.69 193.2 173.1 165.9 6.56 179.0 159.3 Belgium 1986 176.0 6.60 189.2 169.4 163.3 5.71 174.7 157.6 USA 1977 176.8 6.60 190.0 170.2 163.7 5.93 175.6 157.8 England 1995 176.6 6.95 190.5 169.7 163.7 6.02 175.7 157.7 France 1979 174.5 6.00 186.5 168.5 163.2 5.60 174.4 157.6 Spain 1988 175.6 6.04 187.7 169.6 161.3 5.72 172.7 155.6 Turkey 1978 173.5 6.30 186.1 167.2 160.0 5.93 171.9 154.1 Mexico 1975 172.8 7.19 187.2 169.5 160.6 7.39 175.4 153.2 Korea 1979 170.2 5.22 180.6 165.0 157.6 4.76 167.1 152.8 Japan 1990 170.4 5.60 181.6 164.8 157.9 5.00 167.9 152.9 D ow nl oa de d by : 54 .7 0. 40 .1 1 1/ 21 /2 01 8 6: 02 :2 8 A M Short and Tall Stature Ann Nestlé [Engl] 2007;65:117–127 119 documented on the child’s growth chart together with the target height and the target height range. The physical examination and pubertal development should be normal and exclude pathological causes of short stature such as Russell-Silver syndrome, Ullrich-Turner syndrome, hypochondroplasia or other causes of syndromal, chromosomal or skeletal etiology [11, 15] . It is essential that the normal growth curves of the population are available ( table 1 ). For the majority of children, adolescents and adults, a height below the 3rd percentile should be seen as a normal variation in height. Very few children will have a defined clinical abnormality. One of the major differential diagnoses in so-called familial or idiopathic short stature is the variation in growth, i.e. constitutional delay of growth and puberty (CDGP). It is also part of a continuum of the normal Gaussian distribution curve with respect to the age at onset of puberty and the time of the pubertal growth spurt. Statistically, it also occurs with a frequency of 3%. A history of late pubertal development might be obtained from one or both parents or from older siblings. The diagnosis of CDGP is a preliminary diagnosis as long as puberty has not started. The already mentioned history of late puberty in the family and a retarded bone age of –2 or more years can point to the diagnosis. The final diagnosis of CDGP can only be made when the spontaneous onset of puberty occurs outside the +2 SD range of the reference population, i.e. 13.3 years for breast development Tanner stage B2 in girls and 13.6 years for spontaneous testicular growth of 1 3 ml in boys, referring to the data of the Zürich Longitudinal Growth Study [13, 14] . Growth delay and late pubertal development are the major reasons for referral of short patients. The Tanner growth curves for late onset of puberty as well as the data of Buckler and Wild [16] and Rikken and Wit [17] clearly demonstrate that boys and girls with CDGP will not follow their previous height or growth percentiles due to the late onset of their pubertal growth spurt. The already short adolescent will become even ‘shorter’ at a time when girls and boys with a normal onset of puberty have their pubertal growth spurt. A deviation from the growth curve is normal in these patients. It is necessary to explain this normal phenomenon of transient short stature combined with a late pubertal growth spurt in detail to the patient and his/her parents. The adult height will be within the target height range [12] . It is not necessary and usually confusing for the patients and parents, when an extensive workup is performed to confirm or exclude growth hormone (GH) deficiency (GHD) as the cause of this normal variation of growth. A large number of these patients have been tested as ‘transient GHD’ [18–20] . Their correct diagnosis is most likely CDGP. It should be emphasized that CDGP is most often diagnosed in a patient with short stature but is also observed in adolescents with normal or tall stature and occurs, as already stated, with a frequency of 3%. It is one of the major challenges in pediatric endocrinology to differentiate normal variation in height, i.e. familial or idiopathic short stature, and normal variation in growth, i.e. CDGP from pathological causes of short stature and growth. Thus the pathology of short stature includes height and growth disorders. Patients with short stature caused by height disorders have a short stature that follows a defined, disease-specific growth pattern. A typical example is the girl with Ullrich-Turner syndrome [21] . Her growth is clearly abnormal with respect to normal growth charts ( fig. 1 a), but exactly follows the specific and therefore normal growth pattern of Ullrich-Turner patients ( fig. 1 b). The same is true for many other diseases such as Down syndrome [22– 25] , Noonan syndrome [26] , achondroplasia [22, 27, 28] and other skeletal disorders [29] . Disease-specific growth charts have been created for these and other diseases through the tedious efforts of interested and engaged physicians. Thus patients with height disorders have a pathological short stature, but a normal, disease-specific growth pattern. With these disease-specific growth charts adult height can be projected when a patient finds and follows his/her individual growth channel. Deviation from the disease-specific growth curve indicates disease and needs evaluation, perhaps even treatment. The growth pattern is different for patients with pathological short stature caused by growth disorders. The height of these patients initially follows the normal, population-based growth channel indicated by the height of the parents and the target height range. Figure 2 demonstrates the height and growth pattern of twins and the height and target height range of their parents. The twins have an almost identical height and growth pattern for the first 4 years. Then twin B deviates from the previous growth channel whereas twin A continues growing along the 50th percentile. Unfortunately, no workup of this pathological growth pattern was initiated so that the diagnosis of craniopharyngioma was delayed until the age of 8 years. Abnormal growth can present a height in the normal range of the 3rd and 97th percentile or as short or tall stature. Thus, normal height does not exclude a pathological growth process. Growth is the major component of normality and pathology in short, normal or tall stature. D ow nl oa de d by : 54 .7 0. 40 .1 1 1/ 21 /2 01 8 6: 02 :2 8 A M Bramswig Ann Nestlé [Engl] 2007;65:117–127 120 A large variety of mostly chronic diseases changes the previously maintained growth pattern. Height deviates from the normal growth channel when the specific disease becomes growth effective. Height can accelerate, as in patients with precocious puberty or congenital adrenal hyperplasia, or decelerate, as in hypothyroidism, GHD, craniopharyngioma or other mostly chronic diseases of various etiologies [15, 30, 31] . One of the major causes of pathological growth with or without short stature is GHD. The incidence varies from 1: 4,000 [32] to 1: 10,000 [30] . Very few studies have looked into the incidence of GHD in the general population [32] . Most reports are from pediatric endocrine clinics which probably overestimate the true incidence of GHD. Father 97