Start small, think big: Growth monitoring, genetic analysis, treatment and quality of life in children with growth disorders - Chapter 3: Growth failure in adolescents: Etiology, the role of pubertal timing and most useful criteria for diagnostic workup

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Start small, think big: Growth monitoring, genetic analysis, treatment and quality

of life in children with growth disorders

Stalman, S.E.

Publication date

2016

Document Version

Final published version

Link to publication

Citation for published version (APA):

Stalman, S. E. (2016). Start small, think big: Growth monitoring, genetic analysis, treatment

and quality of life in children with growth disorders.

General rights

It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulations

If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.

chapter 3

Growth Failure in Adolescents:

Etiology, the Role of Pubertal

Timing and Most Useful Criteria

for Diagnostic Workup

3

Abstract

Background

The aim of the study was to evaluate the etiology, the role of pubertal timing and most useful criteria for diagnostic workup in adolescents with growth failure.

Methods

Adolescents (n=182) aged 10.0–18.0 years underwent a standardized diagnostic pro-tocol. Constitutional delay of growth and puberty (CDGP) was defined as late pubertal onset or a Tanner stage less than –2 SDS. Dutch and Finnish criteria for growth moni-toring were retrospectively assessed.

Results

In 13 children (7.1%) a specific diagnosis could be established. CDGP was diagnosed in 10% of patients aged ≥13 (girls) or ≥14 years (boys). Sensitivity to detect pathologic causes was 85% and 62% for, respectively Dutch and Finnish criteria for growth moni-toring as used in younger children, but specificity was low (55%–59%).

Conclusions

In adolescents, pathological causes for growth failure and pubertal delay are common, and we recommend a combination of height SDS, distance to THSDS and growth deflection for deciding on further diagnostic testing.

3

Introduction

Growth failure, including short stature, growth retardation, or short in comparison to target height, is generally considered a relatively early sign of various pathological con-ditions. The incidence of detectable causes in children below 10 years varies between 3.0% and 9.5% [1–4]. In two countries evidence-based guidelines have been developed to assist primary health workers in their decision to refer a child for specialist care: in the Netherlands validated up to 10.0 years and in Finland up to 18 years. These consist of a combination of short stature for the population, short for genetic background, and growth deflection [5–7]. The Finnish guideline [6, 7] contains complicated algorithms that can only be used if integrated into an electronic health record system and the change in height SDS (ΔHSDS) can only be calculated up to 12 years of age.

For adolescents with growth failure, the incidence of detectable pathological causes has been reported as markedly lower (1.3%) than in younger children [8]. This can be explained in part by assuming that the majority of congenital disorders have been diagnosed at an earlier age and that acquired disorders causing growth failure in ado-lescence are rare. In this age group, it is also more difficult to develop guidelines for referral for specialist care because of the wide variability of the onset and progression of puberty. For example, in case of delayed pubertal development, growth rate can sub-stantially decrease, leading to a decreasing height standard deviation score (HSDS) and increasing distance to target height SDS (THSDS) [9]. Constitutional delay of growth and puberty (CDGP), the most common cause of delayed puberty [10, 11], is primarily characterized by delayed puberty (pubertal onset >13 in girls and >14 in boys), but also associated with slow growth, delayed bone age and a positive family history for delayed puberty [12–14]. In a recent paper on Danish boys [10], the definition of CDGP for boys was expanded by including a low Tanner score SDS [15, 16].

Here, our first aim is to evaluate the incidence of causes, including the contribution of CDGP, of growth failure in adolescents referred to a general pediatric clinic. Our sec-ond aim is to probe the efficacy of the existing Dutch and Finnish criteria for diagnostic workup in adolescents.

3

Materials and methods

Participants

The pediatric clinic of the general Tergooi Hospitals serves as a regional referral center for children and adolescents with suspected disorders of growth and puberty. During the study period (January 2010–June 2013), 197 patients aged 10.0–18.0 years were referred for a suspicion of growth failure.

In the Netherlands, growth references are available for Dutch, Turkish and Moroccan children. Therefore, referred adolescents of Dutch, Turkish (including similar South Eastern European ethnic backgrounds, e.g. Iranian, Azerbaijani) and Moroccan (in-cluding similar North African ethnic background, e.g. Algerian, Egyptian) ethnicities, were included for evaluation of growth failure. Children with another ethnicity than Dutch, North-African or South-Eastern European, adopted children, and children with missing files were excluded.

Approval for the study was obtained by the Scientific Review Committee of Tergooi Hospitals (letter reference kv/15.04).

Data collection

Upon the first visit the parents completed a questionnaire regarding demographic data, perinatal information, medical history, medication use, growth, and physical/ psychological changes or complaints. The family history included pubertal onset, health status and height of first-degree family members. Specialized nurses completed the data on the growth history of the patient and performed anthropometry (height, weight, head circumference, SH, arm span) and assessed blood pressure and heart rate. Height of both parents was measured, and if the child was disproportionate also pa-rental arm span, SH and head circumference was measured. The following parameters were calculated: birth weight and length standard deviation score (SDS) [17]; height SDS (HSDS, separately for Dutch origin [18] and Moroccan and Turkish origin [19]); target height (in cm and THSDS) [20]; body mass index (BMI) and BMI SDS [21]; head circumference SDS [22]; arm span for height ratio [23]; sitting height (SH SDS) and ratio between SH and height (SH/H SDS) [24]. Pubertal development according to Tanner [25, 26] was expressed as SDS for age and gender using Dutch reference data [15, 27]. Bone age in all children was assessed by a single pediatrician (GK) using the Greulich and Pyle method [28]. Bone age development was defined as bone age minus

3

chronological age. The same strategy for workup was performed in children who did and those who did not fulfill the criteria for growth failure.

In all patients referred for a suspected growth disorder the questionnaire, anthropo-metric measurements, medical history, full physical examination with special attention to dysmorphisms and disproportions, pubertal development and bone age assessment were evaluated by the pediatrician specifically trained in pediatric endocrinology and growth disorders. If insufficient clues for a disturbed growth were found, the patient was discharged from further follow-up or the pediatrician decided on watchful wait-ing. In patients with clues for disturbed growth, additional further investigations were performed. If an immediate clue for a specific diagnosis was present, targeted further investigations for this disease were performed. If no specific clues were found, full laboratory investigations in blood and urine were performed. In case of abnormal IGF-1 levels (less than –1.0 SD), a GH provocation test was performed with clonidine, and if necessary a second test was performed using arginine, after priming with a testoster-one ester (100 mg Sustanon, i.m. 5 days before the test) in boys and estrogen (25 μg ethinylestradiol p.o. for 5 days) in girls. In case of abnormal IGF-1 less than –1.0 SD a peak GH value of >6.7 ng/mL was considered a normal response and in case of an IGF-1 less than –2.0 SDS a peak GH value of >10 ng/mL. In addition, in case of disproportion, defined as SH/H ratio greater than +2.0 SDS [24] and/or a low arm span for height (<96.5%) [23], a radiographic evaluation of the skeleton was performed. Screening for a short stature homeobox containing gene (SHOX) defect was planned in case of a Rap-pold score >4 [23]. In case a girl showed a HSDS >2.0 below THSDS, genetic testing for Turner syndrome was performed except if additional clinical or laboratory findings made such diagnosis unlikely (e.g. low FSH levels, no dysmorphisms). In the case of an abnormal phenotype, the patient was referred to a clinical geneticist for evaluation and if indicated genetic testing was performed. An overview of this clinical assessment is shown in Fig. 1.

3

Definitions

Growth and pathology

We defined growth failure if one or more of the following characteristics would apply: short stature, growth deflection and/or height below target height range. Short stature was defined as a HSDS below –2.0, in line with general consensus [14, 29]. Growth de-flection was defined as a negative change in HSDS of >0.5 SD/1 year, >0.7 SD/2 years or >1.0 SD/undefined time, in line with the definitions used in our recent paper on 3–9.9 year olds [30]. A HSDS >1.6 below THSDS was considered short for target height, in

Assessment by pediatrician Blood count Infection param. Electrolytes FSH, LH Liver/ kidney function param. IgA a.b. celiac d. Total IgA Thyroid function IGF-1 Urine: Glucose, protein, erythro. Targeted further investigations4

Referral: growth disorder? Medical history Physical exam1

Insufficient clues for disturbed

growth n = 133 Clues for disturbed growth

Discharge n = 86

Targeted further investigations

No immediate clue for specific diagnosis

Cytogenetic testing3 1. Skeletal survey

2. Rappold score

Clues for specific

diagnosis Disproportion 1. Girl 2. HSDS >2.0 below THSDS Abnormal phenotype Targeted further investigations2 Referral to clinical geneticist Watchful waiting n = 47

Figure 1. Diagnostic workup in adolescents with growth failure. In all patients clinical characteristics were

evaluated by the pediatrician. If insufficient clues for a disturbed growth were found the patient was dis-charged or the pediatrician decided on watchful waiting. In case of clues for disturbed growth additional investigations were performed. Param., Parameters; FSH, follicle-stimulating hormone; LH, luteinizing hor-mone; IgA a.b. celiac d., immunoglobulin A antibodies for celiac disease; IGF-1, insulin-like growth factor 1; erythro., erythrocytes; GH, growth hormone; SH/H SDS, sitting height/height ratio standard deviation score; SHOX, short stature homeobox containing gene; HSDS, height standard deviation score; THSDS,

tar-get height standard deviation score. 1_{Special attention for dysmorphism, disproportion, anthropometrics.} 2_If

3

line with results of the most recent Dutch nationwide growth study [20]. Patients who were diagnosed with a pathological cause of their growth failure after our diagnostic workup were classified using the European Society for Paediatric Endocrinology (ESPE) Classification into primary or secondary growth disorders, and the remaining patients with short stature as idiopathic short stature (ISS) [14]. Children with ISS were subclas-sified as either familial short stature (FSS) if HSDS was ≥THSDS-1.6, or non-familial short stature (NFSS), if HSDS was <THSDS-1.6 [20, 31]. Children with HSDS≥–2.0 and no pathology were classified as non-short.

Pubertal onset

CDGP was defined in a similar fashion as reported by Lawaetz et al. [10]. The first (clas-sical) type, CDGP type 1, consisted of adolescents with Tanner stage G1 at age ≥14 years in boys or B1 ≥13 years in girls [14], usually in combination with slow growth, delayed bone age (>1.0 years) and a positive family history for delayed puberty. Comorbidity had to be excluded. The second type, CDGP type 2, consisted of adolescents who were pubertal but at a stage expected for a much younger individual, defined by a Tanner genital stage less than –2.0 SDS [10, 15, 16].

Although, obviously, in younger prepubertal adolescents (boys <14 years and girls <13 years) CDGP cannot be diagnosed, we tried to obtain an impression on the likelihood of future development of CDGP by collecting information about pubertal stage SDS, bone age delay, short stature for TH and the family history of pubertal delay (maternal menarche ≥15.0; late paternal growth spurt and shaving).

Assessment of criteria for diagnostic workup

In patients with known causes for growth failure we retrospectively assessed both Dutch and Finnish growth criteria. The Dutch criteria that have been suggested for 3–9.9-year-old children [5, 32] were applied to the teenagers with pathology and included: 1) HSDS less than –2.5; 2) HSDS less than –2.0 and dysmorphic features or disproportion; 3) HSDS less than –2.0 and HSDS>1.6 SD below cTHSDS; 4) HSDS less than –2.0 and height deflection (>1.0 over an undefined time; 5) HSDS >2.5 below cTHSDS and 6) growth deflection (>0.5 SD/1 year, >0.7 SD/2 years or >1.0 SD/undefined time) irrespective of height SDS. We also assessed the predictive value of the Finnish algorithm [6, 7]. The Finnish guideline includes HSDS, distance between HSDS and THSDS (THSDSDEV_{) and HSDS change over time (ΔHSDS). This results in the following}

3

apply the Finnish criteria to our population, Dutch reference data for calculating HSDS and THSDS were used [18, 20].

Statistical analyses

All data were collected from case records and analyzed in SPSS, version 21 (IBM Corp., Armonk, NY, USA). Descriptive statistics were used to quantify the incidence of patho-logic causes for growth failure and to classify patients for CDGP. Independent t-tests for continuous variables and χ2-tests for categorical variables were used to compare characteristics between short and non-short adolescents. Sensitivity, specificity and likelihood ratios were calculated for all criteria, using MedCalc for Windows, version 12.7.8 (MedCalc Software, Ostend, Belgium).

Results

Participants

After excluding 15 cases, 182 children (99 boys, 83 girls) were available for analysis (Fig. 2). At the time of their visit to the growth clinic, 123 children met the characteristics of growth failure as defined at the time of analysis. Patient characteristics categorized for short and non-short adolescents are shown in Table 1.

Diagnoses

Known causes of growth failure were detected in 13 adolescents (7.1%) (Table 2), which were subclassified into three with a primary growth disorder (cases 1, 2 and 7) and ten with a secondary growth disorder. Seven of them were referred only because of growth failure (cases 1, 2, 8–12) without any other physical complaints.

The remaining 70 children with a HSDS less than –2.0 were classified as ISS. The num-ber of NFSS (n=54) was considerably greater than that of FSS (n=16) (Fig. 2). Most of the 99 adolescents without pathology and a height SDS ≥–2.0 had no previous height measurement below –2.0 SDS, nor a height SDS more than 1.6 SD below TH.

3

n = 15 excluded

Adopted children (n = 5), children with another ethnicity than Dutch, North-African or Sounth-Eastern European (n = 3) and missing data (n = 7)

Patients referred for suspicion of growth failure ≥10 years

n = 197

Patients included for analyses

n = 182 Boys n = 99 Girls n = 83 Non-short

n = 44 Idiopathicn = 49 Primaryn = 1 Secondaryn = 5 Non-shortn = 55 Idiopathicn = 21 Primaryn = 2 Secondaryn = 5

Previous HSDS <-2.0 n= 2 No previous HSDS <-2.0 n= 42 HSDS >1.6 below THSDS n=12 HSDS ≤1.6 below THSDS n=32 Previous HSDS <-2.0 n= 3 No previous HSDS <-2.0 n= 52 HSDS >1.6 below THSDS n=16 HSDS ≤1.6 below THSDS n=39 Non-familial short stature n= 37 Familial short stature n= 12 Non-familial short stature n= 17 Familial short stature n= 4

Figure 2. Overview of the participants. One hundred and eighty-two adolescents were included. The majority

were non-short (HSDS ≥–2.0) or were diagnosed as ISS, of whom the majority was non-familial short with normal pubertal development. Six boys and seven girls were diagnosed with a pathological cause for their growth failure. ISS, Idiopathic short stature; HSDS, height SDS; CDGP, constitutional delay of growth and puberty; THSDS, target height SDS.

Table 1. Characteristics of the study group: short vs. non-short

HSDS <-2.0 (n=79) HSDS ≥-2.0 (n=130) Sign Age at presentation, mean (SD), years 13.3 (1.9) 13.0 (1.7) p=0.198

Age range, years 10.2–18.0 10.3–17.8 –

Gestational age in weeks, mean (SD) 39.2 (2.3) 39.5 (1.9) p=0.334

Birth weight SDS, mean (SD) –0.4 (1.2) –0.1 (1.0) p=0.053

Birth length SDS, mean (SD) –0.5 (1.4) 0.1 (1.4) p=0.013

THSDS, mean (SD) –0.6 (0.5) –0.2 (0.6) p<0.001

HSDS, mean (SD) –2.5 (0.4) –1.4 (0.6) p<0.001

HSDS less than –2.5, no. (%) 32 (40.5) – –

Any of the six criteria positive, no. (%) 67 (84.8) 20 (19.4) p<0.001

BMI SDS at presentation, mean (SD) –0.5 (1.0) –0.1 (1.0) p=0.012

Tanner stage, mean (SD)

G 2.2 (1.0) 2.5 (1.2) p=0.208 B 1.7 (1.1) 2.6 (1.3) p=0.004

Tanner stage SDS, mean (SD)

G –1.1 (0.7) –0.5 (0.7) p<0.001 B –0.7 (0.5) –0.4 (1.0) p=0.119

SD, standard deviation; HSDS, height standard deviation score; THSDS, target height standard deviation score; BMI, body mass index; G, genitals (male); B, breast (female).

3

Table 2.

Growth characteristics in children with growth failure with known cause (n = 13)

Case Pathologic diagnosis a Age, M/F HSDS <-2.0 Dutch guideline b Finnish guideline b Indication

for diagnostic workup – Dutch guideline

Indication

for diagnostic workup – Finnish guideline

1) 2) 3) 4) 5) 6) 1) 2) 3) 1 SGA 11.8, F + + -+ -+ - (-0.40/5.8yr) + + -+ + 2 SGA 12.8, M + + -+ -- (--0.20/10.0yr) + + a + + 3

Known cardiac disorder (TGA)

15.8, M -+ (-0.72/2.0yr) -a + -4

Known pulmonary disorder (severe asthma with steroid treatment)

12.6, M -- (--0.95/9.5yr) -a -5 Methylphenidate c 12.9, M + -- (--0.50/6.1yr) -a -6 Methylphenidate c 12.3, F -+ (-1.03/3.2yr) -a + -7 46 XX gonadal dysgenesis 13.6, F -+ (-1.73/8.4yr) -a + -8 Celiac disease 12.2, F + + -+ -- (--0.80/4.5yr) + -a + + 9 Celiac disease 11.8, F + -+ + -- (--1.26/8.8yr) + -+ + 10 Isolated GH deficiency 12.4, F + + + -+ -- (--2.60/6.8yr) + -a + + 11 IGF insensitivity ( IGF1R defect) 15.3, M + + + + + -- (--1.14/4.8yr) + + a + + 12 Hypothyroidism (Hashimoto) 15.7, M + + -+ + + - (-1.40/7.0yr) + + a + + 13 Psychosocial deprivation 10.5, F + + + + -- (--0.10/4.3yr) + + -+ +

Number fulfilling criterion

9 7 3 7 4 2 3 8 5 0 11 8 aAge > 12.0 years, so Δ HSDS/SD could not be calculated. bClassified by ESPE [14]. cFor criteria and numbering, see Table 4. dGrowth rate decreased after the st art of therapy and increased markedly after lowering of the dosage (case 6) and cessation of medication (case 5). M, Male; F, female; HSDS, height SDS; THSDS, target height SDS; +, does comply; –, does not comply; SGA, small for gest ational age; TGA, transposition of the great arteries; GH, growth hormone; IGF

, insulin-like growth factor

3

Pubertal onset

Of the 70 adolescents with an age above the classical cut-off limit for delayed puberty, five children (7.1%) were still prepubertal, and thus were labeled as delayed pubertal development. One of them was diagnosed with gonadal dysgenesis explaining her delayed puberty. She was the only patient out of the 13 children with a known cause for growth failure who showed delayed puberty (case 7). The remaining four adolescents were thus diagnosed as CDGP (type 1), and all were short for TH, had a delayed bone age (–1.8 to –2.7), and three had a positive family history for delayed puberty (Table 3). Three of them were short (HSDS –2.7 to –3.2). Three other pubertal adolescents with an age above the cut-off limit showed a Tanner SDS less than –2.0 and were classified as CDGP type 2. All had a bone age delay (2.6 – 3.3 years) and two out of these three were short for TH and had positive family history for delayed puberty. Thus, out of 70 adolescents with an age ≥13.0 (F) or ≥14.0 (M), seven complied with the diagnosis CDGP (10.0%).

Out of 112 (including nine patients with known causes for their growth failure) prepu-bertal adolescents with an age below the cut-off limit, many showed at least one clinical feature compatible with future CDGP. Table 3 shows the numerical data when relatively mild (and arbitrary) cut-off limits were used.

Table 3. Subcategorization of adolescents referred for growth failure, according to the likelihood of

constitu-tional delay of growth and puberty (CDGP)]

CDGP type 1 a _{CDGP type 2} b _{Possible future CDGP} c

Boys 4 2 5

Short for TH 4 2 1

Delayed bone age 4 2 3

Positive family history delayed puberty

3 2 3

Girls - 1 15

Short for TH - - 8

Delayed bone age - 1 8

Positive family history delayed puberty

- - 7

Total 4 3 20

a_{Adolescents with Tanner stage G1 at age > 14 years in boys and/or B1 > 13 years in girls.} b_Pubertal

adoles-cents with pubertal stage SDS less than –2.0. c_{Younger prepubertal adolescents (boys < 14 years and girls <}

13 years) if three out of the following four conditions were met: 1) a pubertal stage SDS for age and sex of less than –1.0 (as a continuous variable); 2) bone age delay > 1.0 year; 3) short for TH; 4) positive family history of pubertal delay (maternal menarche ≥ 15.0 years, late paternal growth spurt and shaving).

3

Assessment of criteria for diagnostic workup

Assuming that out of the 182 subjects 13 were “diseased” and 169 “non-diseased”, we calculated the sensitivity, specificity, and positive likelihood ratios (LR+) of suggested selection criteria for 3–10-year-old children and adolescents [5, 32] in the Netherlands for referral to specialist care for diagnostic workup. We also investigated these param-eters for the Finnish algorithm [6, 7]. The results are shown in Tables 2 and 4. The highest sensitivity (85%), as is desirable in referred patients, is found for a combi-nation of the six Dutch criteria.

Table 4. Sensitivity, specificity and positive likelihood ratio

Sensitivity Specificity Positive Likelihood ratio (LR+) Dutch algorithm

Criterion 1

Current criterion adolescents, HSDS <-2.5

54 (25-81) 85 (79-90) 3.64 (1.96-6.77) Criterion 2 HSDS <-2.0 + dysmorphisms/ disproportions 23 (5-54) 92 (87-96) 3.00 (0.98-9.21) Criterion 3 HSDS <-2.0 + HSDS >1.6 below THSDS 54 (25-81) 68 (60-75) 1.69 (0.97-2.92) Criterion 4 HSDS <-2.0 + height deflection >1.0 SDS 31 (9-61) 86 (80-91) 2.17 (0.88-5.31) Criterion 5 HSDS >2.5 below THSDS 15 (2-45) 93 (88-96) 2.17 (0.54-8.67) Criterion 6

Recent growth deflection, defined as >0.5 SDS/1 year, >0.7 SDS/2year or >1SD/undefined time

23 (5-54) 89 (83-93) 2.05 (0.70-6.04)

Criterion 1 and 2 54 (25-81) 81 (74-87) 2.84 (1.57-5.14) Criterion 1 and 5 54 (25-81) 82 (76-88) 3.03 (1.67-5.52) Criterion 1, 2 and 5 54 (25-81) 79 (72-85) 2.53 (1.41-4.52) Criterion 1, 2, 4 and 5 62 (32-86) 74 (67-80) 2.36 (1.43-3.89) Any of the 6 growth criteria positive 85 (55-98) 55 (47-63) 1.88 (1.41-2.50)

Finnish algorithm Criterion 1 HSDS ≤ -2.2414 62 (32-86) 72 (65-79) 2.21 (1.35-3.63) Criterion 2 THSDSDEV _{≥ 2.2414 * SD} 38 (14-68) 76 (69-82) 1.59 (0.76-3.31) Criterion 3 ΔHSDS ≤ -2.2414 0 (0-71) 84 (70-93) 0.00 Any of the 3 growth criteria positive 62 (32-86) 59 (51-67) 1.51 (0.95-2.40)

3

Of the single criteria a height SDS less than –2.5 and a height SDS less than –2.0 in combination with a HSDS >1.6 below TH had the highest sensitivity (both 54%), at specificities between 68% and 85%. The complete Finnish algorithm gave a lower outcome (sensitivity 62%). The specificities of both approaches were similar (55% and 59%).

Discussion

The present study evaluated etiology and criteria for diagnostic workup in adolescents with growth failure in clinical practice. First, our results show that in 13 cases (7%) a specific diagnosis could be established for their growth failure. Second, CDGP was found in 10% of the adolescents who were old enough to allow for this diagnosis. Third, the Dutch evidence-based auxological criteria (for 3.0–9.9-year-old children) for specialized diagnostic workup showed a high sensitivity (85%), which was higher than the sensitivity obtained by using the Finnish algorithm (62%). However, the overall specificities for both guidelines are too low for population screening.

The prevalence of pathologic causes for growth failure in adolescents in our study is similar to those in previous observations in children up to 10 years, varying between 3.0% and 9.5% [1–4]. This contrasts to a study in an academic setting on 235 children and adolescents (mean age 10.1 years), which reported an incidence of 1.3 % [8]. This discrepancy may be explained by exclusion in the latter study of children with low height velocity and/or abnormal symptoms, and a high percentage of missing data. Thus, although pathological causes for growth failure are usually uncovered at a younger age, significant pathology in adolescents can still be found.

Growth disorders which one may expect in adolescents are acquired disorders or congenital disorders with a relatively mild phenotype. We did not detect individuals with Turner syndrome or a SHOX defect. Likely, most children with these conditions have been diagnosed at an earlier age. We detected celiac disease in two adolescents, illustrating that celiac disease should be ruled out in any child or adolescent with growth failure [8, 33]. Furthermore, we demonstrate that not all adolescents with a pathological cause for their growth failure were in fact short (HSDS less than –2.0). Two cases (case 4 and 7) presented with a HSDS just above –2.0 in combination with growth deflection. The other two (case 3 and 6) showed growth deflection but their height was still normal.

3

The term “CDGP” has been used for a condition characterized by delayed pubertal on-set (≥13 in females or ≥14 years in males), relatively short stature, delayed bone age and usually a positive family history for delayed puberty. With this classical definition, it has been difficult to establish the diagnosis in older adolescents who may have entered late into puberty, but in whom the exact age at pubertal onset is uncertain. The development of stage line diagrams for Tanner stages (puberty nomograms) [10, 15, 16] has now made it possible to express Tanner stage as SDS for age, so that a Tanner stage less than –2.0 SDS can be used as an additional criterion for CDGP (type 2). In our study, 10% of the patients presenting at an age above the cut-off limits, could be diagnosed as CDGP, and one patient was diagnosed as gonadal dysgenesis. In younger adolescents at least a similar percentage would be expected with time.

Our second aim was to probe the efficacy of various referral criteria for diagnostic workup in adolescents. A retrospective investigation of various criteria for specialized diagnostic workup showed a high sensitivity (85%) when applying the evidence-based Dutch criteria as used in younger children, consisting of a combination of HSDS, height distance to THSDS and growth rate. So, these criteria appear suitable for pa-tients referred to secondary or tertiary care clinics. However, for population screening the specificities of the Dutch and Finnish guidelines are too low. Similarly to our find-ings in 0–3 and 3–9.9 years old children [30], two essential growth criteria are most important for adolescents: for acquired growth disorders height deflection is the major criterion, while in congenital growth disorders the distance to target height is most important [5]. Therefore, physicians are advised to collect data on previous growth data and parental height for proper analysis of growth at the first visit to the clinic, Obviously, this should always be combined with taking a proper medical history and physical examination [34].

The establishment of auxological criteria in adolescents for population screening is complicated, mainly because growth rate may substantially decrease in adolescents with a late or relatively late puberty, in contrast with usually stable growth in childhood. This implies that parameters for growth monitoring used in younger children can be expected to be less specific in adolescents than in childhood. Indeed, our study shows that the evidence-based criteria for referral and diagnostic workup in 3.0–9.9 years old children show low specificities and positive likelihood ratios, rendering these criteria unsuitable for routine growth monitoring in the general population. However, the sensitivity for detecting pathology in our cohort of children referred for growth failure in a secondary hospital setting is 85%. Therefore, we feel that these criteria may still

3

serve as practical guidelines in the diagnostic workup of children referred to secondary and tertiary care for suspected disorders of growth.

We acknowledge several limitations of this study. First, this is a cohort studied in a re-gional, general hospital in a well-developed health care system (encompassing primary, secondary and tertiary health care), and the diagnostic yield and the nature of diagnoses may be different in academic centers and in different health care systems. Second, the total number of studied cases is limited, so analyses had to be limited to descriptive statistics. For similar reasons, the 95% confidence intervals of likelihood ratios for the elements of growth aberration are wide. Third, we have not been able to investigate the specificity of the various criteria in a general population, although the low specificity in the referred cases make it likely that population specificity may be even lower. Fourth, the lack of routine assessment by a clinical geneticist in the majority of cases might have led to undiagnosed growth disorders that are associated with only slightly unusual phenotypes. However, this mirrors the general policy of most general hospitals, in which patients usually are only referred if the phenotype is clearly abnormal.

In conclusion, pathological causes for growth failure in adolescents were found in 7% and CDGP in 10%. A combination of height SDS, distance to THSDS and growth deflec-tion is recommended in adolescents to guide the decision to perform further diagnostic testing.

Acknowledgments

We wish to thank Antti Saari, Ulla Sankilampi and Leo Dunkel for providing us with the algorithms of the Finnish referral criteria; and Bea Jansen van ‘t Land, Desirée Paap, Lidi Schilperoort, Irene Bergsma and Ingrid van de Woude (specialized nurses of the Tergooi growth clinic, Tergooi Hospitals, Hilversum, The Netherlands) for their good care of the patients at the growth clinic and for performing accurate auxological measurements.

Disclosure Statement

3

References

1. Ahmed ML, Allen AD, Sharma A, Macfarlane JA, Dunger DB. Evaluation of a district growth screening programme: the Oxford Growth Study. Arch Dis Child 1993;69:361–5

2. Grote FK, Oostdijk W, SM dMK-S, van Dommelen P, van Buuren S, et al. The diagnostic work up of growth failure in secondary health care; an evaluation of consensus guidelines. BMC Pediatr 2008;8:21. 3. Lindsay R, Feldkamp M, Harris D, Robertson J,

Rallison M. Utah Growth Study: growth standards and the prevalence of growth hormone deficiency. J Pediatr 1994;125:29–35.

4. Voss LD, Mulligan J, Betts PR, Wilkin TJ. Poor growth in school entrants as an index of organic disease: the Wessex growth study. BMJ 1992;305:1400–2. 5. Grote FK, van Dommelen P, Oostdijk W, de Muinck

Keizer-Schrama SM, Verkerk PH, et al. Developing evidence-based guidelines for referral for short stature. Arch Dis Child 2008;93:212–7 6. Saari A, Harju S, Makitie O, Saha MT, Dunkel L,

et al. Systematic growth monitoring for the early detection of celiac disease in children. JAMA Pediatr 2015;169:e1525.

7. Saari A, Sankilampi U, Hannila ML, Saha MT, Makitie O, et al. Screening of turner syndrome with novel auxological criteria facilitates early diagnosis. J Clin Endocrinol Metab 2012;97:E2125–32.

8. Sisley S, Trujillo MV, Khoury J, Backeljauw P. Low incidence of pathology detection and high cost of screening in the evaluation of asymptomatic short children. J Pediatr 2013;163:1045–51.

9. Palmert MR, Dunkel L. Clinical practice. Delayed puberty. N Engl J Med 2012;366:443–53.

10. Lawaetz JG, Hagen CP, Mieritz MG, Blomberg Jensen M, Petersen JH, et al. Evaluation of 451 Danish boys with delayed puberty: diagnostic use of a new puberty nomogram and effects of oral testosterone therapy. J Clin Endocrinol Metab 2015;100:1376–85. 11. Sedlmeyer IL, Palmert MR. Delayed puberty: analysis

of a large case series from an academic center. J Clin Endocrinol Metab 2002;87:1613–20.

12. Prader A. Delayed adolescence. Clin Endocrinol Metab 1975;4:143–55.

13. Rogol AD, Hayden GF. Etiologies and early diagnosis of short stature and growth failure in children and adolescents. J Pediatr 2014;164(5 Suppl):S1–14. 14. Wit JM, Ranke MB, Kelnar CJ. ESPE classification of

pediatric endocrine diagnoses. Horm Res Paediatr 2007;68:1–120.

15. van Buuren S. Puberty Plot Pro Web Application, TNO Quality of Life. 2008. Available at: http://vps. stefvanbuuren.nl/puberty/.

16. van Buuren S, Ooms JC. Stage line diagram: an age-conditional reference diagram for tracking development. Stat Med 2009;28:1569–79. 17. Niklasson A, Ericson A, Fryer JG, Karlberg J,

Lawrence C, et al. 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.

18. Schönbeck Y, Talma H, van Dommelen P, Bakker B, Buitendijk SE, et al. The world’s tallest nation has stopped growing taller: the height of Dutch children from 1955 to 2009. Pediatr Res 2013;73:371–7. 19. van Buuren S. TNO Growth Calculator. 2013.

Available at: http://groeiweb.pgdata.nl/general.asp/. 20. van Dommelen P, Schönbeck Y, van Buuren S. A

simple calculation of the target height. Arch Dis Child 2012;97:182.

21. Cole TJ, Roede MJ. Centiles of body mass index for Dutch children aged 0–20 years in 1980 – a baseline to assess recent trends in obesity. Ann Hum Biol 1999;26:303–8.

22. Fredriks AM, van Buuren S, Burgmeijer RJ, Meulmeester JF, Beuker RJ, et al. Continuing positive secular growth change in The Netherlands 1955–1997. Pediatr Res 2000;47: 316–23. 23. Rappold G, Blum WF, Shavrikova EP, Crowe BJ,

Roeth R, et al. Genotypes and phenotypes in children with short stature: clinical indicators of SHOX haploinsufficiency. J Med Genet 2007;44:306–13. 24. Fredriks AM, van Buuren S, van Heel WJ,

Dijkman-Neerincx RH, Verloove-Vanhorick SP, et al. Nationwide age references for sitting height, leg length, and sitting height/height ratio, and their diagnostic value for disproportionate growth disorders. Arch Dis Child 2005;90:807–12. 25. Marshall WA, Tanner JM. Variations in pattern

of pubertal changes in girls. Arch Dis Child 1969;44:291–303.

26. Marshall WA, Tanner JM. Variations in the pattern of pubertal changes in boys. Arch Dis Child 1970;45:13–23.

27. Mul D, Fredriks AM, van Buuren S, Oostdijk W, Verloove-Vanhorick SP, et al. Pubertal development

3

in The Netherlands 1965–1997. Pediatr Res 2001;50:479–86.

28. Greulich WW, Pyle SI. Radiograph Atlas of Skeletal Development of the Hand and Wrist, 2nd ed. Stanford: Stanford University Press, 1959. 29. Cohen P, Rogol AD, Deal CL, Saenger P, Reiter EO,

et al. Consensus statement on the diagnosis and treatment of children with idiopathic short stature: a summary of the Growth Hormone Research Society, the Lawson Wilkins Pediatric Endocrine Society, and the European Society for Paediatric Endocrinology Workshop. J Clin Endocrinol Metab 2008;93:4210–7. 30. Stalman SE, Hellinga I, van Dommelen P, Hennekam RC, Saari A, et al. Application of the Dutch, Finnish and British screening guidelines in a cohort of children with growth failure. Horm Res Pediatr 2015;84:376–82.

31. Hermanussen M, Cole J. The calculation of target height reconsidered. Horm Res 2003;59:180–3. 32. Oostdijk W, Grote F, Wit JM, de Munick

Keizer-Schrama SM. [NVK Guideline Short Stature]. Available at: http://www.nvk.nl/Portals/0/richtlijnen/ kleine%20lengte/kleinelengte.pdf.

33. van Rijn JC, Grote FK, Oostdijk W, Wit JM. Short stature and the probability of coeliac disease, in the absence of gastrointestinal symptoms. Arch Dis Child 2004;89:882–3.

34. Oostdijk W, Grote FK, de Munick Keizer-Schrama SM, Wit JM. Diagnostic approach in children with short stature. Horm Res 2009;72:206–17.