Anders Juul, Sven Kreiborg, and Katharina M. Main
The evaluation of growth charts and pubertal development in children and adolescents is an important tool for any clinician in the assessment of health status. Optimal thriving and height attainment in accordance with family potential can only be achieved in an environment providing optimal socioeconomic conditions, health care, and psychosocial support. Thus, failure to thrive or to grow may be the first indication of an underlying problem that may need attention. In turn, treatment of children may need to consider the specific growth and developmental windows in order not to disturb this delicate balance.
Measurement of growth in different phases of life
The current concept of prenatal and postnatal growth suggests that there are distinct growth phases, which should be considered separately.
Prenatal growth is divided into three trimesters (by convention). The first trimester is characterized by organogenesis and tissue differentiation, whereas the second and third trimesters are characterized by rapid growth and maturation of the fetus. Fetal growth can be assessed by serial ultrasonography in the second and third trimesters. Abdominal circumference, head circumference, and femoral length of the fetus can be determined, and from these parameters fetal weight can be estimated using different algorithms . The fetal weight estimate should be related to normative data. Some reference curves for fetal growth are based on children born prematurely , and hence such curves tend to underestimate normal fetal weights from healthy pregnancies. Alternatively, reference curves based on ultrasound studies of normal healthy infants exist  and should preferably be used. Based on the changes in fetal weight estimates over time, the fetus can be considered as having a normal fetal growth rate, or alternatively as experiencing intrauterine growth restriction (IUGR) . Children born at term (gestational age 37–42 weeks) are considered mature. Children born before 37 weeks of gestation are premature, and children born after 42 weeks of gestation are postmature. At birth, weight and length can be measured and compared to normative data correcting for gestational age at birth. Based on these comparisons, a newborn child can be classified as either appropriate for gestational age (AGA), small for gestational age (SGA), or large for gestational age (LGA).
IUGR fetuses will often end up being SGA at birth, but not necessarily so. Thus, IUGR infants may end up lighter than their genetic potential but remain within normal ranges (i.e., AGA). Therefore, IUGR and SGA are not synonymous entities, although they are often referred to as such in the literature (Figure 2.1). Height velocity in utero is higher than at any time later in life, leading to an average birth length of 50–52 cm and birth weight of 3.5–3.6 kg after 37–42 weeks of gestation. It is therefore not surprising that growth disturbances during this phase may have long‐lasting effects on growth and health later in life. Whereas the first trimester is dominated by tissue differentiation and organ formation, the second and especially third trimesters show a rapid gain first in length and then in weight. Fetal and placental endocrinology is highly complex and hormones such as insulin, leptin, placental growth hormone, insulin‐like growth factor (IGF)‐2, and thyroid hormone are only some of the many growth factors involved in the regulation of fetal growth.
Figure 2.1 Reference ranges for fetal weight according to gestational age during pregnancy denoted by the blue lines (10th, 50th, and 90th percentiles) (8). Panel (a) shows examples of children with normal birth weights at term; a normally growing fetus ending with a birth weight which is appropriate for gestational age (AGA) and (▪) a fetus with third trimester intrauterine growth restriction (IUGR) ending with a birth weight below the genetic potential but within normal limits (AGA). Panel (b) shows examples of fetuses with intrauterine growth retardation (IUGR) ending up AGA (□) or SGA (▪).
Postnatally, height can be determined by measuring length in the supine position in the first 2–3 years of life. After 2–3 years of age standing height can be measured, preferably using a wall‐mounted stadiometer. Height is determined without shoes, shoulders towards the wall, arms hanging down, and the face straight forward (Figure 2.2). The eyes should be horizontally aligned with the external ear opening. The means of three measurements are recorded. The stadiometer should be calibrated on a daily basis.
Figure 2.2 Standing height determined by a wall‐mounted stadiometer (a). Height is recorded as the mean of three measurements. Sitting height is determined by a specifically designed chair (b). Head circumference is determined using a measuring tape (c). Arm span is determined by measuring the distance from fingertips to fingertips (d).
Importantly, the body proportions (such as head circumference, facial appearance, sitting height, and arm span) may be helpful in the differential diagnosis of growth disorders (Figure 2.1). This can simply be done by assessing the sitting height with subsequent calculation of the sitting height to standing height ratio. This enables quantification of whether or not a growth failure is proportional or disproportional (such as in hypochondroplasia). Reference ranges for this ratio exist .
Changes in height can be separated into infant, childhood, and pubertal growth phases according to the infancy–childhood–puberty (ICP) model described by Karlberg . The majority of children will follow the distinct growth patterns of these phases.
After a brief initial weight loss of up to 10% of the birth weight, growth during the first months postnatally follows to a large extent fetal growth rate during the third trimester with 30 g/day and 3.5 cm/month. After that a rapid decline in growth rate occurs, in both weight and height. However, this period still represents a major growth phase during the lifetime with a three‐fold increase in weight over 6 months. Very little is known about the regulatory factors of growth during this period of life, but nutrition and living conditions play a major role. In 2006, the World Health Organization (WHO) published a new growth chart reference for infancy based on breastfed infants from different countries and ethnic origins living under optimal socioeconomic conditions. This chart did not find significant differences in growth patterns between these children, which indicates that genetic differences may first become evident later in life .
In this phase growth is relatively constant, with a gradual decline in growth velocity over time. From 2 to 4 years children grow approximately 7 cm and 2 kg/year. Beyond 5–6 years of age this rate has decreased to approximately 5 cm/year. This growth phase is highly dependent on growth and thyroid hormones.
During the pubertal growth spurt, which typically stretches over 4–5 years, total height gain is on average 20–25 cm for girls and 25–30 cm for boys with large interindividual variations. There is some tendency that early maturers obtain a higher peak height velocity compared to late maturers (Figure 2.3). Sex steroids increase the pulsatile growth hormone secretion, which in turn increases IGF‐1. Weight gain is highly individual and may occur both before and after peak height velocity.
Figure 2.3 (a) Three examples for height curves and (b) height velocity curves from children with early puberty (●), normally timed puberty (□) and delayed puberty (▲). Note that final height is almost the same (a) and that peak height velocity is higher in earlier puberty (b).
Final height has increased over the past century in developed countries due to major improvements of socioeconomic status and health care, a phenomenon which is now predominantly observed in developing countries. However, earlier onset of pubertal development and increased prevalence of childhood obesity has influenced the trajectory of childhood growth within the last one or two generations, and recently, new Danish reference charts for height, weight, and body mass index have been established .
In girls, the onset of the growth spurt is early and may even precede the development of secondary sexual characteristics in some. Typically, breast buds appear before pubic hair at 10–11 years of age, but occasionally this succession may be reversed . Both breast development and pubic hair attainment are graded into five stages (B1–B5 and PH1–PH5) according to Tanner and Whitehouse . The first menstruation, menarche, is a sign of adult‐level estradiol production and follicle maturation and occurs late during the growth spurt at approximately 13 years of age. Height attainment after menarche is small, with 4–8 cm over 1.5–2 years.
In boys, the pubertal growth spurt occurs relatively late during development. Puberty commences with enlargement of testis size from 3 to 4 mL at 11–12 years of age, and this very first sign of pubertal onset is usually not noticed by the boy or even less so by the parents. Pubertal development in boys is graded into five genital stages (G1–G5) according to Tanner and Whitehouse . Testis growth continues and within 6–12 months pubic hair can be seen. Testicular volume can be determined by the use of an orchidometer to which the size of the testes is compared. Maximum height velocity often occurs at a testis size of 10–12 mL at around 14 years of age, at the time when the voice breaks and facial hair appearance occurs. Thus, boys are already relatively virilized at the time of the adolescent growth spurt . In midpuberty, many boys develop physiological gynecomastia, which usually disappears within 6–12 months.
The onset of puberty is approximately a year earlier in girls than boys, which consequently results in earlier growth arrest in girls than boys (14–15 versus 16–17 years of age). The timing of puberty may also differ by 1–2 years according to ethnicity and nationality. Over the past couple of decades a decline in the age of onset of puberty has been observed in many countries [10,11] suggesting that environmental factors and modern lifestyle may affect maturation in addition to hereditary factors.
A significant number of children and adolescents intermittently experience pains, localized to the shins or legs when going to bed after a physically active day. The etiology of this phenomenon is unknown, but local warmth, gentle massage, and mild pain medication, if the child is in real discomfort, can normally ameliorate the problem, which resolves spontaneously.
Evaluation of growth charts
Growth evaluation should be based on observations over time by applying longitudinal measurements of height and weight on an age‐ and gender‐specific growth chart. These charts are available for many populations and also for a variety of growth disorders and syndromes. Due to the secular trend in height, country‐specific reference ranges should be constructed at regular intervals . Repetitive measures of growth will result in a trajectory of growth, which then can be evaluated against family potential (parental stature, growth of siblings). As some children show considerable seasonal variation in growth, follow‐up periods of 6–12 months may be necessary. In children approaching puberty, pubertal staging  will additionally be necessary for adequate assessment.
Growth charts are usually based on cross‐sectional data from children and adolescents, covering 95% of the population (±2 standard deviations). Charts may depict centiles or standard deviation lines. Per definition, 2.5% of the population will be below or above the outer limits (Figure 2.4). In contrast, height velocity curves are based on longitudinal follow‐up studies of healthy children (Figure 2.5).
Figure 2.4 Normal (Gaussian) distribution of heights illustrating the 95% reference interval by percentiles or standard deviations (SDs).
Figure 2.5 Normal height curve (a) based on healthy children. Lines denote mean ±1 standard deviation (SD) and ±2SD. One individual patient is depicted on the curve (●) before and after operation for a pituitary tumor (craniopharyngeoma) resulting in growth hormone deficiency. A typical deceleration is seen prior to diagnosis. Horizontal lines (red arrow) denote bone age. Following operation the child suffers from pituitary insufficiency and is substituted with L‐thyroxine, hydrocortisone growth hormone, (GH) (arrow), and testosterone (Te) (arrow). This results in a final height well within target height. T = target height range, F = father’s, and M = mother’s height expressed as SDs. (b) Normal height velocity curve based on Tanner’s longitudinal study of healthy children. The same child (●) is depicted on this curve illustrating the marked growth acceleration following GH therapy, as well as the acceleration when puberty is initiated.
In the evaluation, both the position within the growth chart in relation to the parental potential and the trend of the individual growth curve are important. Deviations from the expected may represent two separate pathologic conditions. In populations with a significant secular trend in height attainment due to recently improved socioeconomic conditions, the growth of siblings in comparison to the patient may be helpful as well. The simplest method to determine the family growth potential is based on calculation of midparental height (Box 2.1).
Box 2.1 Calculation of family growth potential (equal to target height or genetic height potential)
[Maternal height (cm) + paternal height (cm)]/2 − 6.5 cm
[Maternal height (cm) + paternal height (cm)]/2 + 6.5 cm
To allow for growth variation within a family, the target height range is calculated as midparental height ± 6.5 cm for both genders.
Potential pitfalls of this approach are: (a) the parents differ considerably in height centiles and (b) one of the parents is not of normal stature.
During childhood, most children will follow their trajectory of growth, which ideally should follow the family potential. There are, however, two phases in life where this trajectory may not be followed without necessarily representing pathology: (a) during the first 2 years of life, children may “catch up” or “slow down” depending on their intrauterine growth and size at birth, a phenomenon also called regression towards the mean; and (b) during puberty, early maturation will lead to a growth spurt above average (vice versa for late maturation) and the individual child will therefore almost always deviate upwards (or downwards) compared to the mean on the growth chart (due to the cross‐sectional design of the growth charts). In general, tall children have a tendency to enter puberty early, short children to enter puberty later.
Acute diseases during childhood and adolescence will often only result in a temporary weight loss with rapid catch‐up after recovery. In contrast, height attainment will often get compromised in long‐term or serious illness. These children may show considerable catch‐up growth after recovery, if their bone age allows further growth potential. Thus, growth deceleration is seen commonly in the year(s) prior to diagnosis of severe chronic disease (e.g., brain tumors or malignancies) which is often first noticed in retrospect.
Detailed evaluation of growth includes bone age determination and final height predictions.
Bone age determination
Linear growth continues until the fusion of the ossification centers. Thus, determination of bone maturation may help to assess the growth potential in an individual, as many disorders of growth are associated with either delayed or accelerated bone age. Bone age is mostly measured with a radiograph of the left hand and wrist and a comparison of the epiphyseal growth plates with age‐ and gender‐specific references (Figure 2.6). Two main systems are used clinically: (a) the Greulich–Pyle method  and (b) the Tanner–Whitehouse (TW) method .
Figure 2.6 Two radiographs of the left hands of two healthy children. Note that the mineralization of the small bones has not yet occurred in the younger child (left).
Computer‐based automated bone age determination programs  are increasingly used as they provide rapid and accurate determination of bone age and consequently final height prediction by applying parental height, current height, weight, age at menarche, and secular trend.
Final height prediction
Both methods of bone age determination (Greulich–Pyle and TW methods) can be employed in prediction models (Bayley–Pinneau and TW method, respectively) for final height [14,16], with a broad margin of error. Both methods are based on studies of healthy children who were followed up until final height, in whom bone ages were determined at various ages. Heights at each bone age were assigned a certain percentage of the final height, e.g., a 13‐year‐old boy with a bone age of 14 years is assumed to have reached approximately 90% of his final height according to the Bayley–Pinneau tables, and his current height can then be transformed into a final height estimate. Pitfalls in this approach are (a) the normal biological variation of bone age in comparison to chronologic age which is ±1 year and (b) the fact that prediction models are based on normally growing children and may therefore both underestimate and overestimate final height in pathologic conditions.
Dental age determination
Dental age or dental maturity may be assessed in different ways. The simplest method is to record the teeth erupted and compare to normative data. A more precise method is to judge the development of the teeth from radiographs. Haavikko  has given normative data for individual permanent teeth, while Demirjian  has developed a scoring system based on assessment of all lower left permanent teeth (except the third molar) from an orthopantomogram. Demirjian’s method has gained general recognition as the most precise. In general, the correlation between dental age and bone age is, however, relatively low (Figure 2.7).
Figure 2.7 Illustrative examples showing low correlation between bone age and dental age. (a) A healthy girl, aged 10 years 9 months, with advanced dental maturity (nearly complete permanent dentition: DS4, M1) compared to the skeletal maturity (prepubertal hand–wrist radiograph). (b) A healthy girl, aged 11 years 6 months, with delayed dental maturity (early mixed dentition: DS2, M1) compared to the skeletal maturity (postpubertal hand–wrist radiograph).
Disorders of growth and puberty
Intrauterine growth restriction
Many adverse conditions can lead to impairment of intrauterine growth and development. Infections, medications, environmental chemicals, exposure to tobacco, maternal diseases, and uteroplacental insufficiency may cause early or late growth restriction. A fetus may follow a growth trajectory below normal throughout pregnancy and be born SGA, or growth restriction may have its onset during the third trimester and lead to IUGR.
Over the past few decades research has revealed that antenatal and early postnatal growth patterns may have health consequences in adult life, which may be caused by fetal programming to accommodate adverse conditions. Links have been established to cardiovascular disorders, dyslipidemia, diabetes mellitus, pubertal timing, and reproductive function. The majority of children born SGA or IUGR (80–85%) will show spontaneous catch‐up growth after birth, typically within the first 2–3 years of life. Thus, the remaining 10–15% of IUGR/SGA children do not show catch‐up growth and remain short in childhood and end up as short adults. These children respond well to treatment with biosynthetic growth hormone. Silver–Russell syndrome is associated with prenatal as well as postnatal growth failure, and children typically respond with significant improvement of final height despite the fact that they generally have no evidence of growth hormone deficiency.
Postnatal growth failure
Today, being of short stature is less well accepted by many societies than being tall. Therefore, many children are presented in the clinic (Figure 2.8). In the majority of cases a growth curve evaluation will reveal that the child is within its family potential. A typical growth curve of a child with familial short stature is shown in Figure 2.8(a). These children are typically growing at a normal growth rate and thus following their growth trajectory. These families need reassurance, as there is today no convincing treatment schedule available that will reliably and significantly increase final height. If, however, the child’s position on the growth curve does not correspond to the familial potential or the growth trajectory deviates downwards due to low growth velocity, the child should be investigated further. Many chronic and systemic diseases (e.g., asthma, sleep apnea, malabsorption, and metabolic diseases) and systemic steroid treatment may lead to growth disorders. In rare cases even large doses of inhaled steroids may have a growth inhibitory effect (Figure 2.8f).
Figure 2.8 Illustrative growth curves of children with growth failure. (a) A child with familial short stature who has a sub‐normal predicted adult height in accordance with the short genetic height potential and retarded bone age. (b) A child with growth deceleration due to the development of a benign brain tumor, which was diagnosed and operated upon. Following the operation, growth hormone (GH) therapy was started and a normal final height was obtained. (c) A girl with Turner syndrome diagnosed in late childhood because of growth failure and a height at diagnosis below genetic height potential. Growth hormone treatment results in growth acceleration, and at age 12 years puberty induction was initiated by low‐dose estradiol treatment. (d) Prenatal and postnatal growth failure in a girl diagnosed with Silver–Russell syndrome before and after initiation of growth hormone therapy which results in marked catch‐up growth. (e) A girl with deceleration of growth from 2 to 3 years of age concomitantly with constipation. She was diagnosed with acquired hypothyroidism and substituted with L‐thyroxine which normalized growth. (f) Marked stunting of growth from 5 years of age and delayed bone age in a girl who was erroneously treated with high‐dose inhalation steroids despite the fact that she no longer had asthma. Cessation of therapy accompanied by growth hormone therapy resulted in marked catch‐up growth.
Hormonal insufficiencies such as growth hormone deficiency (Figure 2.8b) or hypothyroidism (Figure 2.8e) and Cushing syndrome typically present with stunting of growth. Likewise, chromosomal aberrations such as Turner syndrome in girls (45,X) (Figure 2.8c), genetic syndromes such as Silver–Russell syndrome (Figure 2.8d) or Noonan syndrome are well‐known etiologies for postnatal growth failure. Also severe neglect or abuse may induce growth retardation (psychosocial dwarfism).
Importantly, skeletal dysplasias can cause severe stunting of growth. In this case, body proportions are skewed as in the typical case of achondroplasia. Achondroplasia is the most frequent form of short‐limb dwarfism. Affected individuals exhibit short stature caused by rhizomelic shortening of the limbs, characteristic facies with frontal bossing and midface hypoplasia, exaggerated lumbar lordosis, limitation of elbow extension, genu varum, and trident hand. Achondroplasia is caused by mutation in the fibroblast growth factor receptor‐3 gene (FGFR‐3). There is some evidence to suggest that a minor fraction of the milder forms of skeletal dysplasias (hypochondroplasia) may also be due to FGFR‐3 mutations. A large number of different skeletal dysplasias can be classified according to clinical and radiologic criteria (Table 2.1).
Table 2.1 Short‐limbed conditions and adult height (cm)
Adult height (cm)
Tall stature and growth acceleration
Tall stature is today socially more accepted, and some studies indicate that tall people may have higher social success and better job prospects. However, for some individuals, extreme tall stature may still present a major disadvantage and give numerous practical problems (Figure 2.9).
Figure 2.9 Illustrative growth curves of children with tall stature and growth acceleration. (a) A boy with marked growth acceleration from early childhood who was diagnosed with Klinefelter syndrome (47,XXY). He will end up above his target height despite advanced bone age. (b) A boy with growth acceleration from early childhood who was diagnosed with double Y syndrome (47,XYY) who will end up with increased final height. (c) A boy with growth acceleration from 10 to 12 years of age who was diagnosed with gigantism and operated on for his growth hormone‐producing pituitary adenoma. (d) A boy with growth acceleration and who was obese (simple obesity) who will end up with a final height within his target range, probably because of his advanced bone age. (e) Increased growth in a girl who presented with precocious puberty (regular menstruation at the age of 9 years), and markedly advanced bone age. She will end up with a final height at the lower end of her target range. (f) A girl with familial tall stature and delayed puberty who was treated with high‐dose estrogen to accelerate epiphyseal fusion. Despite this, she reached a final height above target range.
Most cases are familial (Figure 2.9f), but rare diseases may also be the underlying cause. Supernumerary sex chromosomes, such as Klinefelter syndrome (47,XXY) (Figure 2.9a), the most common sex chromosome abnormality (1:660 newborns), triple X syndrome (47,XXX), and double Y syndrome (47,XYY) (Figure 2.9b) are all characterized by increased growth compared to the reference population, as well as compared to their genetic target. Endocrine disorders, such as gigantism (growth hormone hypersecretion because of pituitary tumor) (Figure 2.9c), are extremely rare, but should be excluded. Other genetic conditions such as Marfan syndrome (long limbs with narrow hands and long slender fingers, and arm span greater than height), Soto syndrome (prominent forehead, large ears and mandibles, and coarse facial features) and homocysteinuria may be found among patients referred because of tall stature.
Tall stature must be distinguished from conditions with temporary growth acceleration that do not lead to increased final height, such as obesity (Figure 2.9d) in childhood, hyperthyroidism, and early sexual maturation (Figure 2.9e). These children will deviate upwards on their growth chart, but their accelerated bone maturation will at the same time lead to premature fusion of the growth plates.
If the estimated final height is unacceptable to the child and family, gender‐specific treatment with sex hormones to accelerate closure of the epiphyseal plates is an available option, which, however, requires careful discussion with the families about potential risks and benefits. This can be done either by induction of early pubertal maturation or by addition of sex hormone during spontaneous puberty to shorten the pubertal growth spurt. In addition, operative epiphysiodesis of growth plates around the knees is an established treatment option for tall stature.
Early pubertal maturation is much more frequently seen in girls than in boys. There are indications that true precocious puberty before the age of 8 years in girls and before the age of 9 years in boys is becoming more frequent in many populations. Foreign adopted children seem to be at greatest risk. There also appears to be a genetic component, as some families present with early puberty over several generations. In girls, early puberty often presents as an idiopathic premature activation of the hypothalamus–pituitary axis and is rarely caused by diseases. Conditions such as intracranial tumors, hydrocephalus, autonomous sex hormone production (gonadal tumors), and disorders of steroid biosynthesis such as congenital adrenal hyperplasia need to be excluded, especially in boys. It is possible to postpone further pubertal development until a more appropriate age by treatment with long‐acting gonadotropin agonists. This treatment can be useful for children who have difficulties in coping with the psychological effects of early maturation or in very short children, in whom predicted final height is extremely low. Paradoxically, these children are usually referred at a time of pubertal growth acceleration, and therefore present with a height in excess of their peers. Parents are usually unaware of the fact that they may end up being very short (Figure 2.9e).
Late puberty is much more frequently seen in boys than in girls as an extreme of the natural gender dimorphism. In most cases, it is a simple delay of maturation without any underlying pathology. Family history may reveal inheritance from one or both parents as a constitutional delay of growth in puberty. These children usually present with short stature compared to their age‐matched friends, lack of secondary sexual characteristics, a growth curve that shows deviation downwards with time, and delayed bone age. In rare cases, delayed puberty is caused by endocrine disorders such as gonadotropin or pituitary insufficiency or developmental disorders of the gonads. Excessive sporting activities and eating disorders can also cause significant delay in physical maturation. In girls, delayed puberty may be caused by a chromosomal disorder, such as Turner syndrome (45,X).
If no pathology is found, simple reassurance may be the only treatment necessary. Delayed puberty itself does not lead to short final height. If puberty is delayed beyond acceptable limits for the child, treatment with low‐dose sex hormones for 6–12 months may help to “kickstart” the process.
Disorders with deviations in dental maturity
Many children with postnatal growth failure also show delay in dental maturity, e.g., growth hormone deficiency. Likewise most patients with Soto syndrome show advanced dental maturity. Thus, the dentist should be aware of the oral manifestations of general diseases and, thereby, contribute to early diagnosis (see Chapter 23).
Prenatal and postnatal growth reflects the general health status of an individual. Growth charts are easy to obtain, noninvasive, and cheap. Many countries offer health services that allow the longitudinal follow‐up of height and weight attainment, together with an evaluation of puberty progression in teenage years. Pathologic growth charts and a bone age that deviates significantly from chronologic age can be the first indicators of a serious underlying condition that needs attention. Thus, knowledge about normal and abnormal growth patterns in children and adolescents is necessary for all medical personnel that are involved in their health care. In addition, the pediatric dentist should be aware of the fact that marked deviations in dental maturity could be part of a general growth problem.
1. 1. Marshall WA, Tanner JM. Variations in the pattern of pubertal changes in girls. Arch Dis Child 1969;44:291–303.
2. 2. Niklasson A, Ericson A, Fryer JG, Karlberg J, Lawrence C, Karlberg P. An update of the Swedish reference standards for weight, length and head circumference at birth for given gestational age (1977–1981). Acta Paediatr Scand 1991;80:756–62.
3. 3. Marsal K. Intrauterine growth restriction. Curr Opin Obstet Gynecol 2002;14:127–35.
4. 4. Hertel T, Scheike T, Juul A, Main KM, Holm K, Bach‐Mortensen N, et al. Body proportions of Danish children. Curves for sitting height ratio, subischial length and arm span. Ugeskr Laeger1995;157:6876–81.
5. 5. Karlberg J. A biologically‐oriented mathematical model, (ICP) for human growth. Acta Paediatr Scand Suppl 1989;350:70–94.
6. 6. WHO Multicentre Growth Reference Study Group. Child growth standards based on length/height, weight and age. Acta Paediatr Scand 2006;450:76–85.
7. 7. Tinggaard J, Aksglæde L, Sørensen K, Mouritsen A, Wohlfahrt Veje C, Hagen C, Mieritz MG, Jørgensen N, Wolthers OD, Heuck C, Petersen JH, Main KM, Juul A. The 2014 Danish references from birth to 20 years for height, weight and body mass index. Acta Paediatr Scand 2014; 103:214–24.
8. 8. Tanner JM, Whitehouse RH. Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Arch Dis Child 1976;51:170–9.
9. 9. Marshall WA, Tanner JM. Variations in the pattern of pubertal changes in boys. Arch Dis Child 1970;45:13–23.
10. 10. Aksglaede L, Sorensen K, Petersen JH, Skakkebæk NE, Juul A. Recent decline in age at breast development. The Copenhagen Puberty Study. Pediatrics 2009;123:e932–9.
11. 11. Sørensen K, Aksglæde L, Petersen JH, Juul A. Recent changes in pubertal timing in healthy Danish boys: Association with body mass index. J Clin Endocrinol Metab 2010;95:263–70.
12. 12. Juul A, Teilmann G, Scheike T, Hertel NT, Holm K, Laursen EM, et al. Pubertal development in Danish children: comparison of recent European and US data. Int J Androl 2006;29:247–55.
13. 13. Greulich WW, Pyle SI, Waterhouse AM. A radiographic standard of reference for the growing hand and wrist. Chicago, IL: Case Western Reserve University, 1971.
14. 14. Tanner JM, Whitehouse RH, Cameron N, Marshall WA, Healhy MJR, Godtstein H. Assessment of skeletal maturity and prediction of adult height, 2nd edn. London: Academic Press, 1983.
15. 15. Thodberg HH, Kreiborg S, Juul A, Damgaard K. The BoneXpert Method for Automated Determination of Skeletal maturity. IEEE Trans Med Imag 2009;28:52–66.
16. 16. Bayley N, Pinneau S. Tables for predicting adult height from skeletal age. J Pediatr 1952;14:423–41.
17. 17. Haavikko K. Tooth formation age estimated on a few selected teeth. A simple method for clinical use. Proc Finn Dent Soc 1974;70:15–19.
18. 18. Demirjian A. Dentition. In: Falkner F, Tanner JM, eds. Human growth. London: Baillière Tindall, 1982.