Evidence Tables

Wheeler P, Balk E, Cole C, et al. Criteria for Determining Disability in Infants and Children: Short Stature. Rockville (MD): Agency for Healthcare Research and Quality (US); 2003 Mar. (Evidence Reports/Technology Assessments, No. 73.) Evidence Tables.

Evidence Table 1. Studies Evaluating Short Stature Secondary to Medical Impairment Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Disease/Condition Type (N)	Study Design (Duration)
Abbott 1982 82277019	Location: US	Short stature due to growth hormone deficiency and/or other pituitary hormone deficiency	None	Isolated short stature	Prospective cross-sectional cohort
	Setting: In-patient
	Mean age: 11±3 y
	Age range: 4–18 y
	Mean height: ND
	Height range: 81–132 cm
	Male: 64%
	Race: White 82%, Black 18%
	Enrolled: ND
	Evaluated: 11
	Number of sites: 1
Angehrn 1979 80093987	Location: Switzerland	Russell-Silver syndrome:	None	RSS	Prospective longitudinal cohort
	Setting: ND	Height < 2 SDS
	Mean age: 4±2.2 y	Low normal growth velocity
	Age range: ND	Low birth weight
	Mean height, SDS: -4.4±1.0	Moderate bone age retardation
	Male: 80%	Small triangular face and skull of large appearance
	Race: ND
	Enrolled: ND
	Evaluated: 20
	Number of sites: 1
Frisch 1990 90350511	Location: Austria	Isolated GH deficiency, or Multi-pituitary hormone deficiency	None	GH deficiency	Prospective cross-sectional
	Setting: Clinic
	Mean age: 12±3 y
	Age range: 6.8–15.7 y
	Mean height, SDS: -2.5±0.9
	Height range, SDS: -4.5 to -0.9
	Male: 67%
	Race: ND
	Enrolled: ND
	Evaluated: 23
	Number of sites: 1
Gordon 1984 84214634 Gordon 1982 82268405	Location: US	Cases - Constitutional short stature <5^th percentile height	Cases - Abnormal GH studies	Isolated short stature	Prospective cross-sectional cohort
	Setting: Psychiatric history clinic	Skeletal maturation delayed > 13 months	Psychiatric history
	Mean age:	Height velocity < 4 cm/y	Controls - Psychiatric history
	Cases 10±2.2 y	Controls - Normal height
	Controls 10±21. y	Matched for age, IQ, SES and sex
	Age range:
	Case 6–12 y
	Control 6–12 y
	Mean height, SDS: <5^th percentile
	Male: 83%
	Race: ND
	Mean SES: 3.5 (Hollinghead)
	Enrolled: ND (23 controls)
	Evaluated: 24 (23)
	Number of sites: 1
Holmes 1986 Holmes 1985 86060101	Location: US	Height < -2 SDS	Additional medical diagnoses (e.g., IDDM)	Isolated short stature:	Prospective longitudinal cohort
	Setting: Pediatric endocrinology clinic	Patients at pediatric endocrinology clinic		GH deficiency (17)	(Mean 3.1 y Range 2.2–4.0 y)
	Mean age, initial: 12 y			Constitutional delay (21)
	Mean age, follow-up: 15 y			Turner syndrome (9)
	Age range: ND
	Mean height, SDS: < -2
	Male: ND
	Race: ND
	SES: Middle class 21%
	Enrolled: 76
	Evaluated: 47
	Number of sites: 1
Kranzler 2000 20102780	Location: US	Referred Cases - Height < 5^th percentile	IQ < 75	Isolated short stature	Prospective cross-sectional cohort
	Setting: Region	Pediatric endocrinology clinics (n=27)	Serious psychiatric or medical problem
	Mean age: 10±1.4 y	Non-referred - Identified in local public schools as shortest 10% of class (n=34)	English not primary language
	Age range: 6–12 y	Controls - Height in middle 50%
	Mean height, SDS:	At local public schools
	Referred -2.7±1.0
	Non-referred -1.7±0.5
	Controls +0.1±0.4
	Height range, SDS:
	Referred -4.5 to -1.3
	Non-referred -3.2 to -1.3
	Controls -0.7 to +0.7
	Male: 51%
	Race: ND
	Enrolled: ND
	Evaluated: 61 (29 controls)
	Number of sites: 2
Lai 1994 95243689	Location: UK	Russell-Silver Syndrome (met 3 of 5 criteria)	None	RSS	Prospective cross-sectional
	Setting: Clinic	Age 6–12 y
	Mean age: 9±2 y
	Age range: ND
	Mean height, SDS: -2.2±1.5
	Male: 80%
	Race: White 95%; West Indian 5%
	SES: 80% Middle-class or higher
	Enrolled: 30
	Evaluated: 25
	Number of sites: 3
McCauley 1987 87161055 McCauley 1986 86141341	Location: US	Cases - Turner syndrome	Verbal IQ < 79–80	Turner syndrome (17)	Prospective cross-sectional cohort
	Setting: Pediatric endocrinology clinic	Followed at Endocrine and Genetics Clinics		Isolated short stature (16)
	Mean age:	Controls - Girls with height < 5^th percentile from cause other than Turner syndrome
	Turner 13±2.4 y	Followed at Endocrine and Genetics Clinics
	SS 13±2.8 y
	Age range: ND
	Mean height:
	Turner 52±5.0 in
	SS 53±4.5 in
	Male: 0%
	Race: White 94%, Black 6%
	SES:41±16 (Hollingshead revised)
	Enrolled: 23 (22 controls with SS)
	Evaluated: 33 total
	Number of sites: 1
Meyer-Bahlburg 1978 78184456	Location: US	Isolated growth hormone deficiency, or Multiple hormone deficiency	None	Isolated growth hormone deficiency (9 total)	Prospective cross-sectional cohort
	Setting: Pediatric endocrine clinic			Multiple hormone deficiency (13 total)
	Mean age:
	GH def 12 y
	Multi-hormone def 13 y
	Age range:
	GH def 4–18 y
	Multi-hormone def 6–18 y
	Mean height, SDS:
	GH def -3.6
	Multi-hormone def -3.9
	Height range, SDS:
	GH def -4.8 to -2.4
	Multi-hormone def -6.0 to -1.9
	Male: 73%
	Race: ND
	Enrolled: 29
	Evaluated: 22 (age ≤ 18 y; 29 total)
	Number of sites: 1
Pollitt 1964	Location: US	Significant short stature thought to be due to growth hormone deficiency - either pan-hypopituitary or isolated	None	Isolated short stature	Prospective cross-sectional cohort
	Setting: Endocrine clinic	Pediatricians thought were well-adjusted psychologically and had cooperative parents
	Mean age: ND
	Age range: 3–16 y
	Mean height: ND
	Height range: 70–126 cm
	Male: 40%
	Race: ND
	Enrolled: ND
	Evaluated: 15
	Number of sites: 1
Robinson 1983 84052395	Location: US	Monosomy X karyotype (45, X) and partial monosomy X	None	Turner Syndrome	Prospective cross-sectional cohort
	Setting: Region			Monosomy (6)
	Mean age: ND			Partial monosomy (3)
	Age range: 7–15 y
	Mean height: ND
	Male: 0%
	Race: ND
	Enrolled: 9
	Evaluated: 9
	Number of sites: ND
Ross 1997 97320537	Location: US	Turner syndrome, confirmed by karyotyping	Estrogen treatment	Turner syndrome	Prospective cross-sectional cohort (derived from RCT of GH treatment)
	Setting: ND	Age ≥ and <12 y	Androgen treatment
	Mean age: 9.5 y
	Age range: ND
	Mean height: ND
	Male: 0%
	Race: 87% White
	SES: 47 (Hollingshead)
	Enrolled: ND
	Evaluated: 40
	Number of sites: 2
Rovet 1985	Location: Canada	Height more than 2 SD below mean	History of familial short stature	Isolated short stature	Prospective cross-sectional
	Setting: Clinic
	Mean age: 14±1 y
	Age range: ND
	Mean height: 139±9 cm
	Male: 100%
	Race: ND
	Enrolled: 27
	Evaluated: 25
	Number of sites: 1
Siegel 1986	Location: US	Idiopathic hypopituitarism	Deprivation dwarfism	Isolated short stature	Prospective cross-sectional cohort
	Setting: Endocrinology clinic		Craniopharyngioma	GH def (28)
	Mean age: 12±2.3 y		Age < 6 y	Multiple hormone def (14)
	Age range: 6–16 y		IQ < 30
	Mean height, SDS: < -2
	Male: 71%
	Race: ND
	SES, Social position: High 5%, High average 21%, Average 31%, Low average 33%(14), Low 10%
	Enrolled: 53
	Evaluated: 42
	Number of sites: 2
Siegel 1994	Location: US	Cases - Height < 3^rd percentile	Known mental impairment	Isolated short stature	Prospective cross-sectional cohort
	Setting: Endocrinology clinic	Controls - Normal height	Severe psychiatric	GH Def (87)
	Mean age:		Chronic medical illness	Idiopathic SS (90)
	GHD 11±2.7 y		Non-English speaking
	ISS 11±2.6 y
	Control 11±2.8 y+		Short stature secondary to systemic disease, tumor, or genetic abnormalities
	Age range: ND
	Mean height, SDS:
	GHD -2.7±0.6
	ISS -2.8±0.6
	Control +0.1±0.9
	Male: 75%
	Race: White 91%
	Enrolled: 177 (90 controls)
	Evaluated: 177 (90)
	Number of sites: 27
Siegel 1998 98356098	Location: US	Cases - Girls with Turner syndrome, GH deficiency, or idiopathic short stature	Did not follow through with 3 y of GH therapy	Turner syndrome (22)	Prospective longitudinal cohort
	Setting: Clinic	Controls - Matched for age and socioeconomic status		GH deficiency (13)	(3 y)
	Initial mean age: 10 y			Idiopathic short stature (12)
	Control 11 y
	Initial mean height:
	TS 120±13 cm
	GHD/ISS 125±12 cm
	Control 145±17 cm
	Male: 0%
	Race: ND
	Enrolled: 88 (Controls 67)
	Evaluated: 47 (25)
	Number of sites: ~27
Skuse 1997 Gilmour 1996 96408881 Skuse 1994 95252652	Location: UK	Cases - Height < 5th percentile (< -2 SDS)	Cases - Bone age delayed > 2 y	Isolated short stature (15)	Prospective cross-sectional cohort
	Setting: Growth clinics	Height velocity over the last year above -1.5 SDS	Medical treatment for growth retardation	Russell-Silver syndrome (7)
	Mean age:	Age 6–11 y	Russell-Silver children who had IQ < 72
	Case 8±1.5 y	Prepubertal
	Control 9±1.4 y	Controls - From each case's school class, matched for age, sex, and SES
	Age range:
	Case 6–11 y
	Control 2–11 y
	Mean height, SDS:
	Case -2.5±0.3
	Control -0.3±1.1
	Male: 72%
	Race: ND
	SES (Parent profession): 59%
	advantaged/professional ; 14%
	clerical/craft; 14% manual
	Enrolled: 31 (25 controls)
	Evaluated: 22 (22)
	Number of sites: 2
Stathis 1999 20099119	Location: Australia	Cases - Height < 3^rd percentile	PPVT-R score < 50	Isolated short stature	Prospective cross-sectional cohort
	Setting: Region	Age 4–6 y	Cerebral palsy
	Mean age: 5.5 y	Controls - Height between 10^th and 90^th percentile	Other serious neurological disorder
	Age range: 4–6 y	Age 4–6 y
	Mean height: <3^rd percentile
	Male: 53%
	Race: ND
	Enrolled: ND
	Evaluated: 113 (3,178 controls)
	Number of sites: 1
Steinhausen 1976 77086274	Location: Germany	Cases - Height < 3^rd percentile	None	Isolated short stature	Prospective cross-sectional cohort
	Setting: Inpatient	Controls - Age and sex matched from local schools with “normal” stature		GH deficiency (16)
	Mean age:			Constitutional delay (16)
	SS 14±2.7 y
	Control 14±2.6 y
	Age range:
	SS 9–18 y
	Control 9–18 y
	Mean height, SDS: <3^rd percentile
	Male: 81%
	Race: ND
	SES: Lower class 59%, Middle class 41%
	Enrolled: ND
	Evaluated: 32 (32 controls)
	Number of sites: 1
Tanner 1975 75217260	Location: UK	Russell-Silver syndrome	None	RSS	Unclear
	Setting: Growth disorder clinic	Height < -2 SDS
	Mean age: ND	Adjusted birth weight ≥ 2 SD below the mean of the Tanner-Thomson standard
	Age range: 1.4–10 y
	Mean height, SDS: -3.6
	Height range, SDS: -6.3 to -2.2
	Male: 56%
	Race: ND
	Enrolled: 39
	Evaluated: 39
	Number of sites: 1
Voss 1994 Voss 1991 92151680 Downie 1997 97159125 Downie 1996 96408882	Location: UK	Cases - Height <3^rd percentile, starting school between 1985 and 1986	Cases - Organic disease (e.g., Turner syndrome or celiac disease); Immigrant parents	Isolated short stature	Prospective longitudinal cohort (up to 9 y)
	Setting: Region	Controls - From same school class	Controls - None
	Mean age: ND	Height between 10^th and 90^th percentile
	Age range: Data at ages 5–6 y, 7–9 y, 11–13 y, mean 13 y (see comments)	Matched for age, sex and school class
	Mean height (age 5–6 y), SDS:
	Case -2.3±0.3
	Control +0.1±0.6
	Mean height (mean age 13, No GH): Case 139 cm
	Control 155 cm
	Male: 54%
	Race: ND
	SES: Class I/II 20%; Class III 49%; Class IV/V 31%; 25% unemployed fathers
	Enrolled: 140 (140 controls)
	Evaluated: 140 (140) (See comments)
	Number of sites: 1
Wilson 1986	Location: US	National Health Examination Survey (NHES):	None	Isolated short stature	Prospective cross-sectional cohort
	Setting: ND	1963-1965: 7,119 children 6–11 y
	Mean age: ND	1966-1970: 6,768 adolescents 12–17 y
	Age range: 6–17 y
	Mean height, SDS: ND
	Male: ND
	Race: ND
	Enrolled: ND
	Evaluated: 6,768–7,119 (all heights)
	Number of sites: ND
Young-Hyman 1986	Location: US	Children with SS who presented to endocrine clinics for evaluation	Chronic health problem	Isolated short stature	Prospective cross-sectional cohort
	Setting: Endocrine clinic		Mental disability
	Mean age: 12 y
	Age range: ND
	Mean height, SDS: ND
	Male: 78%
	Race: ND
	SES: Upper middle class
	Enrolled: 27
	Evaluated: 27
	Number of sites: 1

Evidence Table 1. Studies Evaluating Short Stature Secondary to Medical Impairment Part II

Author, Year	Predictors	Predictor Measures	Outcomes	Outcome Measures
Abbott 1982 82277019	Isolated short stature	ND	Cognitive/Academic	IQ (WISC)
				Academic achievement (WRAT)
				Visual-motor skills (Developmental Test of Visual-Motor Integration)
Angehrn 1979 80093987	Russell-Silver syndrome	Height < -2 SDS, low normal growth velocity, low birth weight for gestational age, moderate bone age retardation, typical small triangular face, and large skull	Cognitive/Academic	IQ (ND)
			Psychomotor	Psychomotor development (Denver developmental screening test)
Frisch 1990 90350511	Isolated short stature/GH deficiency	GH peak < 5 ng/ml in 2 stimulation tests	Cognitive/Academic	IQ (HAWIK & HAWIE, German version of Wechsler IQ test)
	Hypopituitary GH deficiency			Academic performance (Repetition of grade)
			Behavioral/Psychological	Personality structure tests (Freiburger Personality Questionnaire, Giessen Test)
				Sociability, Social activity, Neuroticism (Hamburger Neuroticism-Extroversion Scales)
Gordon 1984 84214634 Gordon 1982 82268405	Isolated short stature	Height < 5^th percentile	Cognitive	IQ (WISC-R)
				Math, reading, comprehension (PIAT)
			Behavioral	School adjustment, Internalization (CBCL - Teacher)
Holmes 1986	Isolated short stature: CGD (21)	Retarded bone age, growth velocity within 2 SD of normal for age, projected adult height normal	Cognitive/Academic	IQ (WISC-R)
Holmes 1985 86060101	GHD (17)	Maximum GH ≤ 7 ng/ml on 9 samples
	TS (9)	ND
Kranzler 2000 20102780	Isolated short stature	Referred: Height < 5^th percentile	Cognitive/Academic	Verbal and non-verbal intelligence (K-BIT)
		Non-referred: Height < 10^th percentile of class		Educational achievement (KTEA brief form)
			Behavioral	Adaptive and problem behaviors (BASC - Teacher)
Lai 1994 95243689	Isolated short stature	Russell-Silver syndrome	Cognitive/Academic	IQ (WISC)
				Reading comprehension (Neale Analysis of Reading Ability)
				Arithmetic comprehension (ND)
				Cognitive processing (Matching Familiar Figure Test)
				Special educational needs
McCauley 1987 87161055 McCauley 1986 86141341	Turner syndrome	Karyotype showing absence of all or part of an X chromosome	Cognitive	IQ (WISC)
			Behavioral	Behavior factor (Piers-Harris Scale - Children self-report)
	Isolated short stature	Height < 5^th percentile	Psychological	Depression (Children's Depression Inventory)
Meyer-Bahlburg 1978 78184456	Isolated growth hormone deficiency, and multiple hormone deficiency	GH deficient or panhypopituitary	Cognitive	IQ (WISC)
Pollitt 1964	Hypopituitary short stature	Panhypopituitary	Cognitive	IQ (WISC Scale for Children - age > 5 y)
		Isolated GH deficiency
		Turner syndrome
		Primordial
		Undiagnosed
Robinson 1983 84052395	Turner Syndrome	45, X monosomy or partial monosomy on karyotype from chromosomal exam of amniotic membrane	Cognitive	IQ (Wechsler Preschool and Primary Scale of Intelligence)
				Language development (“A comprehensive speech and language evaluation”)
			Physical Development	Neuromuscular maturation (Bruininks-Oseretsky Test of Motor Proficiency)
Ross 1997 97320537	Turner syndrome	Karyotype 45,X	Cognitive/Academic	IQ (WISC-R)
				Academic achievement (WART - reading)
Rovet 1985	Isolated short stature	SD from mean	Cognitive	IQ (WISC)
Siegel 1986	Hypopituitary isolated short stature	GH secretion < 7 ng/dL, delayed bone age, growth < 4 cm in past year, height < -2 SDS or below the 3^rd percentile	Cognitive/Academic	IQ (WISC-R)
				Academic achievement (WRAT)
				Visual-motor integration skills (Bender Visual-Motor Gestalt Test: a significant visual-motor integration deficit was defined as if the developmental score was < 16^th percentile, and four or more errors were significant indicators of brain injury)
Siegel 1994	Isolated short stature GH deficient Idiopathic short stature	Height < 3^rd percentile	Cognitive/Academic	IQ (Slosson Intelligence Test)
				Academic achievement (WRAT - revised)
Siegel 1998 98356098	Isolated short stature	Turner syndrome GHD/ISS	Cognitive/Academic	IQ (Slosson Intelligence Test)
				Academic achievement (WRAT - revised)
Skuse 1997 Gilmour 1996 96408881 Skuse 1994 95252652	Isolated short stature and Russell-Silver syndrome	Height < -2 SDS	Cognitive	IQ (WISC for children - version III)
			Behavioral	Competence and self-concept (Harter Self Perception Profile for Children, Harter Pictorial Scale of Perceived Competence and Social Acceptance for Young Children, CHIPS)
Stathis 1999 20099119	Isolated short stature	Height < 3^rd percentile	Cognitive	Receptive vocabulary (PPVT-R, which correlates closely with intellectual ability)
Steinhausen 1976 77086274	Isolated short stature	Height < 3^rd percentile GH deficiency or no endocrine disease	Cognitive/Academic	Primary mental abilities (LPS)
				IQ (WISC)
				Academic performance (Grade level for age)
				School achievement (ND)
			Behavior	Extraversion and Neuroticism (Hamburg Neuroticism Extraversion Scale)
Tanner 1975 75217260	Russell-Silver syndrome	Height < -2 SDS, adjusted birth weight< -2 SDS, no other obvious cause of short stature	Cognitive/Academic	IQ (ND)
				Academic achievement (Grade repetition)
Voss 1994 Voss 1991 92151680 Downie 1997 97159125 Downie 1996 96408882	Isolated short stature	Height < 3^rd percentile	Cognitive/Academic	IQ (Short form IQ scale of the British Ability Scales)
				Attainment (Word reading and basic number skills scales of British Ability Scales)
			Behavioral	Behavioral problems (Rutter's Behavior Questionnaire - Teacher)
				Locus of control (Nowicki and Strickland scale)
Wilson 1986	Isolated short stature	Height < 5^th percentile	Cognitive/Academic	IQ (WISC)
				Academic achievement (WRAT)
Young-Hyman 1986	Isolated short stature	ND	Cognitive	IQ (WISC-R)
				Visual-motor integration skills (Bender Visual-Motor Gestalt Test)

Evidence Table 1. Studies Evaluating Short Stature Secondary to Medical Impairment Part III

Author, Year	Associations found	Potential Biases	Comments
Abbott 1982 82277019	Mean (SD) IQ of children with SS due to pituitary hormone deficiency (n=11) was Full Scale = 88 (17), Verbal = 89 (20), Performance = 89 (16), Reading = 88 (17), Spelling = 85 (14), and Arithmetic = 83 (15)	Compared to population norms	Study was government and hospital funded
	Mean (SD) WRAT scores for children with SS (n=11) were Reading = 88 (17), Spelling = 85 (14), and Math = 83 (15)
	Visual-Motor Integration skills were significantly below average among children with SS
Angehrn 1979 80093987	Mean (range) IQ for children with SS due to RSS (n=9) was 91 (38 to 110); 3 subjects had low IQs of 85, 83, and 38; an additional 5 subjects were “probably of normal intelligence”	Incomplete testing	No data on funding source
	Of subjects with RSS, 3 of these children had microcephaly; 3 children had sub-normal IQ, 2 of whom had microcephaly	IQ test used not reported
	More than half of the psychomotor activities on the Denver developmental screening test occurred at an age > 90^thpercentile in subjects with RSS (n=14)	Incomplete and unclear reporting
		Small sample size for IQ data
Frisch 1990 90350511	The median (range) IQ of children with GHD (n=23) were Full Scale = 115 (68–131), Performance = 114 (61–136), Verbal = 104 (72–128); full score and performance IQ were significantly higher than population norms; verbal IQ was similar to population norms	Compared to population norms	No data on funding source
	Performance Behavior (zeal and school aversion) was similar in children with GHD (n=20) compared to population norms	Reporting incomplete
	Sociability in children with GHD (n=20) than population norms; Social activity in children with GHD was similar to population norms (No further data)
	Emotional Stability (including mood, control, depression) in children with GHD was similar to population norms (No further data)
	Delayed Schooling was more common in children with multiple pituitary hormone deficiencies and SS (33%) than normal population (2.1%)
Gordon 1984 84214634 Gordon 1982 82268405	Mean (SD) IQ for children with SS due to constitutional delay (n=24) were total 108 (14), verbal 107 (16), and performance 108 (14).	Comparisons of cognitive function between short children and controls of questionable value since controls were matched for overall IQ	Study was government and private funded
	Mean (SD) scores on PIAT for children with SS (n=24) were math 103 (16), reading recognition 102 (11), and comprehension 102 (11); these were, similar to scores among children with normal stature who were matched by IQ (n=23): 103 (9.6), 102 (7.5), and 103 (10), respectively	No outline of the statistical analyses used in the paper; therefore, it is hard to evaluate the significance	There is no evidence that these behavior problems constitute a disability
	Mean (SD) teacher CBCL scores for children with SS (n=24) were total 31.2 (26.5) and for controls (n=23) 27.5 (27.6); conduct problem scores were 22.6 (4.3) and 24.1 (6.8), respectively; inattentive/passive 12.7 (4.8) and 12.1 (4.8), respectively; hyperactive 13.7 (5.9) and 11.9 (4.3), respectively; and socialization 14.7 (2.5) and 14.3 (3.0), respectively		The subjects and controls were matched for IQ and had equal school performance
Holmes 1986 Holmes 1985 86060101	Mean (SD) Verbal IQ in children with isolated SS (n=47) was 102 (13); among these children, those with GH deficiency (n=17) verbal IQ was 97 (13), those with constitutional delay (n=21), 104 (14) and those with TS (n=9), 103 (14); these were all similar to population norms	40% non-participation rate	No data on funding source
		Those declining participation more likely to have constitutional delay and to have received optimal treatment benefits
		Study presented only partial data, focusing on significant results
		IQ compared only to population norms
Kranzler 2000 20102780	Mean (SD) IQ scores of referred children with SS (n=27) were Verbal = 101 (16), Matrices = 106 (14), and Composite = 103 (14); of non-referred children with SS (n=34) were Verbal = 104 (15), Matrices = 105 (14), and Composite = 105 (14); and of normal controls (n=29) were verbal = 110 (12), matrices = 110 (17), and composite = 112 (13);	Study of cognitive skills excluded children with low IQ	Study was government, privately and pharmaceutical funded
	Post hoc analysis found that referred short children had significantly lower Verbal IQ than normal controls; other comparisons were statistically similar	Definition of non-referred short children sample vague
	Mean (SD) composite and spelling Educational Achievement (KTEA) scores were similar for referred short children, 106 (18) and 105 (15), non-referred short children, 107 (19) and 104 (19); and, controls, 114 (17) and 102 (17).	Predominantly post hoc analyses
	Mean (SD) mathematics and reading KTEA scores were lower for referred short children, 107 (21) and 104 (15) and for non-referred short children, 105 (19) and 106 (18) than for controls, 121 (20) and 115 (18).
	Teacher ratings on BASC for Externalizing behavior, Internalizing behavior, School Problems composite, Behavior symptoms index and Adaptive Skills composite were each similar among referred short children, non-referred short children and normal controls
Lai 1994 95243689	The mean (SD) IQ of children with RSS (n=25) were Full Scale = 86 (24), range 46–130, Verbal = 89 (23), range 52–141, and Performance = 84 (24), range 47–127; these were significantly lower than population norm, 100 (15)	Compared to population norms and controls from prior study	No data on funding source
	In children with RSS, larger Head Circumference was significantly associated with higher IQ (r=0.5–0.6)
	Reading Comprehension in children with RSS was delayed by a mean of 15.4 months; Arithmetic Competence was delayed by a mean of 7 months
	Cognitive Processing, measured by reaction time and error scores in children with RSS (n=22) was similar to normal controls from an earlier study (n=90)
	Special Education needs were necessary in 9 (36%) children with RSS, primarily in reading and language skills; 12 (48%) children received speech therapy
McCauley 1987 87161055 McCauley 1986 86141341	Mean verbal and performance IQ (SD) for girls with TS (n=17) were 95 (11) and 91 (10), respectively; those IQ (SD) scores for girls with SS not due to TS (n=16) were 100 (12) and 108 (14), respectively.	Excluded children with low IQ	Study was hospital funded
	The difference in performance IQ between the groups was significant; however, both groups had verbal and performance IQ scores within 1 SD of the mean
	Girls with TS had significantly lower performance on arithmetic, digit span, picture completion, object assembly, and coding subscales than girls with SS; performance on other subscales were similar for the two groups
	Self-reported Piers-Harris behavior factor and Children's Depression Inventory scores were similar in girls with TS and girls with SS;
	“Although the TS children indicated low self-concept they did not endorse features of clinical depression.”
Meyer-Bahlburg 1978 78184456	Mean IQ of children with isolated GHD (n=13) was 101, and of children with multiple hormone deficiency (n=9) was 102	Compared to population norms	No data on funding source
	Patients with multiple hormone deficiencies have somewhat lower IQs than patients with isolated GHD, after controlling for SES
Pollitt 1964	Mean (SD) IQ for children over age 5 with SS (n=13) were total 103 (16), verbal 103 (15), and performance 102 (16);	Compared to population norms	No data on funding source
	These IQ scores were not significantly different than normal IQ		Descriptive study
Robinson 1983 84052395	Mean (SD) IQ of girls with TS (n=9) were Full Scale = 87 (20), Verbal = 93 (25), and Performance = 83 (12). Siblings (n=17) had significantly higher Full Scale IQ = 111 (15) and Performance IQ = 107 (14), but similar Verbal IQ =113 (15).	Small sample	No data on funding source
	The TS girls were at non-significantly greater risk of developing Language Disorders than their unaffected siblings	Handwriting problems not defined
		Methods and results not fully reported
		Most analyses were done for all subjects (girls and boys with various sex chromosomal anomalies) together
Ross 1997 97320537	Mean (SD) Full scale IQ in girls with TS with GH treatment (n=20) was 98 (14) and without GH (n=20) was 99 (15)	Indirectly compared to population norms	Study was government funded
	Mean (SD) Academic Achievement (WART) reading score in girls with TS with GH treatment (n=20) was 100 (14) and without GH treatment (n=20) was 97 (17), similar to population norms
Rovet 1985	The mean IQ of boys with SS (n=25) were Verbal = 98, Performance = 102, and Full Scale = 101	Compared to population norms	Study was hospital funded
Siegel 1986	Mean (SD) IQ for children with hypopituitary SS (n=42) were verbal 94 (18), range 52 to 127, and performance 94 (16), range 49–123; Both the verbal and performance IQ were significantly below normal IQ	Compared to population norms	Study was hospital funded
	Mean (SD) WRAT scores for children with hypopituitary SS (n=42) were reading 96 (16.9), range 52 to 131, and math 85 (13.9), range 52–126;	Excluded children with severe mental retardation
	The WRAT reading score was in normal range, but the WRAT math score was significantly below the normal score (100)
	Significant visual-motor integration deficits occurred in 11 (26%) of children with hypopituitary SS (n=42), indicating “a significant number of hypopituitary youngsters have visual-motor integration problems.”
Siegel 1994	Mean (SD) IQ for children with GHD (n=87) was 110 (16) for children with ISS (n=90) was 107 (19); compared to 116 (15) for control children (n=90);	Excluded children with “known mental impairment”	No data on funding source
	Significantly more GHD children (9%) and ISS children (18%) had IQs < 90, than control children (3%)
	Mean (SD) WRAT scores for children with GHD (n=87) were reading 103 (18), spelling 98 (17), and math 99 (17); for children with ISS (n=90) were reading 102 (18), spelling 98 (17), and math 96 (21); and for control children (n=90) were reading 105 (14), spelling 103 (14), and math 105 (16)
	Significantly more ISS children (22%) had a skill deficit in math (score < 80) compared to GHD (10%) and normal children (7%)
Siegel 1998 98356098	The mean IQ of girls with TS (n=22) was 102 (11), and with GHD/ISS (n=25) was 105 (15); these were not significantly different than controls (n=25) with 109 (17).	CBCL completed by parent, typically mother	Study was pharmaceutical company funded
	Mean (SD) WRAT scores for children with TS (n=22) were reading 98 (15), spelling 95 (19), and math 90 (16); for children with GHD/ISS (n=25) were reading 105 (18), spelling 99 (21), and math 102 (22); and for control children (n=25) were reading 98 (16), spelling 100 (15), and math 99 (20). No significant differences among groups.
Skuse 1997 Gilmour 1996 96408881 Skuse 1994 95252652	Mean (SD) IQ of children with SS (n=22) were Full Scale = 96 (19), Verbal = 95 (20), and Performance = 97 (17); for normal children (n=22), Full Scale = 107 (17), Verbal = 105 (17), and Performance = 108 (17);	Excluded Russell-Silver Syndrome children who had low IQ	Study was pharmaceutical funded
	Children with SS had statistically lower Full Scale IQ and Performance IQ than controls, but statistically similar Verbal IQ	Incomplete reporting of results.
	Conduct, Global, Physical, Athletic, Social, and Scholastic competencies measured by Teacher CBCL were similar among SS children (n=17) and controls	Teacher CBCL completed only on subset of subjects
Stathis 1999 20099119	Mean receptive vocabulary score on PPVT-R test for boys with SS (n=60) was 92, which was significantly lower for normal height boys, 98 (n=1,660)	Study of correlation of PPVT-R score to height excluded those with score < 50	No data on funding source
	Mean receptive vocabulary score on PPVT-R test for girls with SS (n=53) was 92, which was significantly lower for normal height girls, 95 (n=1,508)		PPVT-R score was adjusted for maternal age, education, non-English speaking background, age at evaluation, maternal depression at birth and maternal height
Steinhausen 1976 77086274	Mean IQ for children with SS (n=32) were total 103, verbal 102, and performance 103.	Not all of the psychological tests are published and validated; the one well-known test (WISC) was used without data in control group	No data on funding source
	Primary mental abilities on LPS intelligence test were significantly different among children with SS (n=32) and controls (n=32) for only 2 of 14 measures of intelligence; Spatial orientation was 4.7 for SS children and 6.2 for controls; Speed of closure was 4.4 for SS children and 5.3 for controls	Many comparisons with some statistical significance but questionable clinical significance; the intelligence (LPS) and personality (CPQ) variables are not well defined
	Grade level was appropriate in 59% of children with SS (n=17); Academic achievement was adequate in 48% of SS children	No statistical comparison was done for social and coping behavior; and, school performance; Average scores only
	Among children with short stature (n=32) mean extraversion and neuroticism scores on HANES were 11.4 and 12.9, respectively. These were similar to controls (11.6 and 15.6, respectively).	No outline of how the subset 17 kids were selected
		School achievement not defined
Tanner 1975 75217260	Mean (SD) IQ for children with SS due to RSS (n=11) was 103 (21), range 70–130	IQ test not described and performed in only a sub-sample	No data on funding source
	Of subjects with RSS and head circumference < 3^rdpercentile (n=18), 83% were at “normal” schools, 17% were at schools for educationally subnormal children (“usually IQ < 80”);
	Of subjects with head circumference > 3^rdpercentile (n=19), all were at normal schools
Voss 1994 Voss 1991 92151680 Downie 1997 97159125 Downie 1996 96408882	Mean (SD) IQ in children aged 7–9 y with SS (n=140)was 102 (15.2) and for controls with normal height (n=140), IQ was 104 (14.1); there was no significant difference between the two groups; there was also no significant difference in Verbal scales, Non-verbal scales, Speed of Information Processing, or Reading or Number Attainment, after controlling for SES	None noted	Studies were hospital and pharmaceutical funded
	Behavior Disturbance (total Teacher RBQ score > 9) occurred in 29% of children aged 7–9 y with SS (n=132), and in 21% of control children (n=132); the difference was not significant;		Downie studies included subsets of subjects in Voss studies; one examined subjects treated with GH; one examined untreated subjects
	Hyperactivity (Activity subscale score ≥ 3) occurred in 16% of SS children compared to 8% of control children (p=0.003);		Different sized subsets of subjects, at different ages, analyzed in different articles
	Mean (SD) Total Behavior scale score for children with SS was 6.2 (6.2), compared to 5.1 (6.5) for control children (NS)
	Conduct subscale score was 1.2 (2.3) for children with SS, compared to 1.0 (2.1) for control children (NS)
	Emotional subscale score was 1.2 (1.4) for children with SS, compared to 1.2 (1.7) for control children (NS)
	Activity subscale score was 1.7 (2.1)for children with SS, compared to 1.1 (1.6) for control children (NS after controlling for SES)
	Mean (SD) IQ in children aged 13 y with SS, not treated with GH, (n=84–106 for different tests) and controls (n=93–119), were 103 (17) and 109 (16), respectively (p=0.005);
	Reading Attainment of 44.3 (8.8) and 47.9 (8.5), respectively (p=0.002);
	Math Attainment of 40.2 (7.2) and 43.5 (9.2), respectively (p=0.003);
	Locus of Control score of 16.6 (5.0) and 14.3 (5.1), respectively (p=0.001), Behavior score of 6.8 (7.5) and 5.3 (6.8), respectively (NS);
	The statistical significance of these differences were reduced, but remained significant, after controlling for SES
Wilson 1986	Mean IQ (WISC) for children with SS, aged 12–17 (N~350) was 92	Prevalence and correlation data only	Study was government funded
	Mean IQ (WRAT) for children with SS, aged 12–17 (N~350) was 91	Incomplete reporting
	Among children of all heights, aged 6–11 (n=7,119), both WISC and WRAT scores showed a significant but small correlation with height (r=0.18, WISC; r=0.17, WRAT)
	Among children of all heights, aged 12–17 (n=6,768), both WISC and WRAT scores showed a significant but small correlation with height (r=0.20, WISC; r=0.19, WRAT)
	“The association between height and IQ scores is less among the subjects with a high family income”
Young-Hyman 1986	Mean (SD) IQ for children with SS (n=27) was 110 (16), range was 85 to 142	How significance was defined for perceptual-motor skill test was not outlined in the paper	No data on funding source
	Mean (SD) perceptual-motor skill for children with SS (n=27) was significantly delayed at 9.2 y (1.2 y), compared to chronological age 11.8 y

Evidence Table 2. Studies Evaluating Short Stature Secondary to Skeletal Dysplasia Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Disease / Condition Type	Study Design Mean Duration (Range)
Alston 1983	Location: UK	Cases - Osteogenesis imperfecta	None reported	Osteogenesis imperfecta	Prospective cross-sectional cohort
	Setting: Brittle Bone Society, Health Authorities	Controls - Same school class or neighborhood as cases, matched for age, sex and social class
	Mean age: ND
	Age range: 5–16 y
	Mean height: ND
	Male: ND
	Race: ND
	Enrolled: 40 (40 controls)
	Evaluated: 40 (40)
	Number of sites: ND
Apajasalo 1998 98163312	Location: Finland	Cases - Diagnosis of skeletal dysplasia	Cases - Incomplete questionnaires	Skeletal dysplasia	Prospective cross-sectional
	Setting: Clinic	Age 12–15 y		Achondroplasia
	Mean age: ND	Controls - Age-matched Helsinki elementary school pupils who completed the HRQOL questionnaire		Cartilage hair hypoplasia
	Age range: 12–15 y	Age 12–15 y		Diastrophic dysplasia
	Mean height: ND
	Male: 42%
	Race: ND
	Enrolled: 21 (239 controls)
	Evaluated: 19 (239)
	Number of sites: 1
Bailey 1971 72080259	Location: US	Diagnosis of achondroplasia	None reported	Achondroplasia	Retrospective cross-sectional
	Setting: Clinic
	Mean age: ND
	Age range: 3 days – 72 y
	Mean height: ND
	Male: 51%
	Race: ND
	Enrolled: ND
	Evaluated: 41 (no data on number < 21 y)
	Number of sites: 1
Benson 1978 79027354	Location: US	Diagnosis of osteogenesis imperfecta from 1924 to 1974	Spine X-ray films not available	Osteogenesis imperfecta	Retrospective longitudinal case series (Mean [Range] 5 [0–17] y)
	Setting: Shriner's Hospital
	Mean age: ND
	Initial age range: 1 mo – 16 y
	Mean height: ND
	Male: 44%
	Race: White 94%; Black 6%
	Enrolled: 143
	Evaluated: 126
	Number of sites: 1
Bleck 1981 82026195	Location: US	Diagnosis of osteogenesis imperfecta: congenita or tarda	Death	Osteogenesis imperfecta	Prospective longitudinal cohort (Mean [Range] 4 [1–15] y)
	Setting: ND		Too young to enable a satisfactory follow-up
	Initial mean age: 9y		Moved out of the state
	Follow-up mean age: 13 y
	Initial age range: 3 mo – 21 y
	Follow-up age range: 4–26 y
	Mean height: ND
	Male: ND
	Race: ND
	Enrolled: 30
	Evaluated: 24
	Number of sites: ND
Brinkmann 1993 94091382	Location: Germany	Cases - Clinical and radiographic diagnosis of achondroplasia	None reported	Achondroplasia	Prospective cross-sectional
	Setting: Clinic	Controls - Closest sibling;
	Mean age: 9 y	Other causes of short stature
	Age range: ND	Matched for age, sex and SES
	Mean height: ND	Normal height children, matched for age, sex and SES
	Male: 47%%
	Race: ND
	SES: Primarily middle-class
	Enrolled: 30 (90 controls)
	Evaluated: 30 (90)
	Number of sites: 1
Cox 1982 83016754	Location: US	Medically diagnosed osteogenesis imperfecta, types I, III or IV	History of head injury	Osteogenesis imperfecta	Prospective cross-sectional cohort
	Setting: Dept of Communicative Disorders		Prior surgery
	Mean age: 10 y		Treatment for ear disease Noise exposure (For current review, excluded those age ≥ 18 y)
	Age range: 4–17 y
	Mean height: ND
	Male: 53%
	Race: ND
	Enrolled: 30
	Evaluated: 30 (15 met criteria for current review)
	Number of sites: 1
Daly 1996 96219608	Location: UK	Patients registered at OI clinic who had been recently seen	No reply to the questionnaire	Osteogenesis imperfecta	Prospective cross-sectional cohort
	Setting: Osteogenesis imperfecta clinic
	Mean age: 7.4 y
	Age range: 3.5–16.3 y
	Mean height: ND
	Male: 49%
	Race: ND
	Enrolled: 59
	Evaluated: 51
	Number of sites: 1
Engelbert 1997	Location: Netherlands	Diagnosis of osteogenesis imperfecta	Parents declined to participate	Osteogenesis imperfecta	Prospective cross-sectional cohort
Possible overlap with Engelbert 1998, Engelbert 1999 Engelbert 2000 Engelbert 2001	Setting: Children's hospital	Visit source hospital = 1 per year
	Mean age:
	Type I OI: 7.3±3.5 y
	Type III OI: 6.5±2.9 y
	Type IV OI: 9.1±5.0 y
	Age range:
	Type I OI: 2–16 y
	Type III OI: 2–12 y
	Type IV OI: 1–17 y
	Mean height: ND
	Male: 54%
	Race: ND
	Enrolled: 68
	Evaluated: 61
	Number of sites: 1
Engelbert 1998 98366964	Location: Netherlands	Children with osteogenesis imperfecta who regularly attended a referral center	None reported	Osteogenesis imperfecta	Prospective cross-sectional cohort
Possible overlap with Engelbert 1997, Engelbert 1999 Engelbert 2000 Engelbert 2001	Setting: Children's hospital
	Mean age: 8 y
	Age range: 1–15 y
	Mean height, SDS:
	Type I OI -1.6±1.3
	Type III OI -8.0±2.0
	Type IV OI -2.8±1.7
	Male: 47%
	Race: ND
	Enrolled: 47
	Evaluated: 47
	Number of sites: 1
Engelbert 1999 99381568	Location: Netherlands	Diagnosis of osteogenesis imperfecta	Intramedullary fixation surgery within 6 months	Osteogenesis imperfecta	Prospective longitudinal cohort (Mean [Range] 1.2 [0.7–1.8] y)
Possible overlap with Engelbert 1997, Engelbert 1998 Engelbert 2000 Engelbert 2001	Setting: Children's hospital		Any other disability or impairment
	Mean age:		Parents' refusal in follow-up
	Initial: 7.3±3.8 y		Unable to be reached
	Follow-up: 9.4±3.6 y		Surgical interventions in the last half year of the study
	Age range:
	Initial: 1–15 y
	Follow-up: 3–17 y
	Mean height: ND
	Male: 50%
	Race: ND
	Enrolled: 54
	Evaluated: 44
	Number of sites: 1
Engelbert 2000 20429012	Location: Netherlands	Diagnosis of osteogenesis imperfecta	Too young to expect walking	Osteogenesis imperfecta	Retrospective cross-sectional cohort
Possible overlap with Engelbert 1997, Engelbert 1998 Engelbert 1999 Engelbert 2001	Setting: Children's hospital	Completed questionnaire
	Mean age: 11.2±5.2 y
	Type I OI: 10.8±5.8 y
	Type III OI: 9.4±2.9 y
	Type IV OI: 13±4.6 y
	Age range: 1–28 y
	Type I OI: 1–28 y
	Type III OI: 5–15 y
	Type IV OI: 4–20 y
	Mean height: ND
	Male: 45%
	Race: ND
	Enrolled: 98
	Evaluated: 70
	Number of sites: 1
Engelbert 2001 21334060	Location: Netherlands	Diagnosis of osteogenesis imperfecta	None reported	Osteogenesis imperfecta	Prospective cross-sectional cohort
Possible overlap with Engelbert 1997, Engelbert 1998 Engelbert 1999 Engelbert 2000	Setting: Hospital
	Mean age: 12.6±3.2 y
	Age range: 8–16 y
	Mean height: ND
	Male: 50%
	Race: ND
	Enrolled: 40
	Evaluated: 40
	Number of sites: 1
Falvo 1973 74013834	Location: US	Diagnosis of osteogenesis imperfecta	None reported	Osteogenesis imperfecta	Unclear
	Setting: ND
	Mean age: 13.5 y
	Age range: 4–34 y
	Mean height: ND
	Male: 27%
	Race: ND
	Enrolled: ND
	Evaluated: 11
	Number of sites: 1
Fowler 1997 97356792	Location: US	Diagnosis of achondroplasia	None reported	Achondroplasia	Prospective cross-sectional
Possible overlap with Pauli 1995	Setting: Bone dysplasia clinic
	Mean age: ND
	Age range: 1–60 months
	Mean height: ND
	Male: 48%
	Race: White 88%, Black 2%, Hispanic 3%, Korean/Asian/Indian 5%
	Enrolled: 93
	Evaluated: 37
	Number of sites: 1
Pauli 1995 95193803	Location: US	Clinical diagnosis of achondroplasia	None reported	Achondroplasia	Prospective longitudinal (N = 53) and cross-sectional (N = 22) cohorts
Possible overlap with Fowler 1997	Setting: Bone clinic				(To age > 1 y)
	Mean age: ND
	Age range: Infants, young children
	Mean height: ND
	Male: 52%
	Race: White 85%; Black 3%; Hispanic 4%; Korean 7%; Asian Indian 1%
	Enrolled: 75
	Evaluated: 75
	Number of sites: 1
Hecht 1991 92152267	Location: US	Infants with clinical diagnosis of achondroplasia	None reported	Achondroplasia	Prospective cross-sectional cohort
	Setting: Clinic
	Mean age: 15±5 mo
	Age range: 8.5–26 mo
	Mean height: ND
	Male: 46%
	Race: ND
	Enrolled: 13
	Evaluated: 13
	Number of sites: ND
Hunter 1998 98299528	Location: Multinational	Cases - Diagnosis of primary skeletal dysplasia	None reported	Skeletal dysplasia	Prospective cross-sectional
	Setting: Clinic	Age < 16–18 y old (by site)
	Mean age: ND	Controls - Unaffected siblings
	Age range: ND
	Mean height: ND
	Male: 49%
	Race: White 100%
	Enrolled: 55 (37 controls)
	Evaluated: 55 (37)
	Number of sites: 6
Kuurila 2000 20379515	Location: Finland	Diagnosis of osteogenesis imperfecta	Insufficient cooperation	Osteogenesis imperfecta	Prospective cross-sectional cohort
	Setting: Dept of clinical genetics	Age 4–16 y	No audiometry performed
	Mean age: 10 y		Moderate mental retardation.
	Age range: 4–16 y
	Mean height: ND
	Male: 42%
	Race: White
	Enrolled: 60
	Evaluated: 45
	Number of sites: 1
Norimatsu 1982 82163323	Location: Japan	Osteogenesis imperfecta	None reported	Osteogenesis imperfecta (congenita and tarda)	Retrospective longitudinal case series
	Setting: Children's Hospital	Treated from 1959 to 1980			(Mean [Range] 7.6 [1–20] y)
	Initial mean age: ND	Followed by spinal x-ray
	Follow-up mean age: 15.5 y
	Initial age range: 12 d – 14 y
	Follow-up age range: 4–28 y
	Mean height: ND
	Male: 68%
	Race: (Japanese)
	Enrolled: 30
	Evaluated: 22
	Number of sites: 1
Poussa 1991 92054945	Location: Finland	Diagnoses of diastrophic dysplasia	None reported	Diastrophic dysplasia	Prospective longitudinal cohort 13.5 y
	Setting: Genetics and bone clinics	Age ≤ 20 y
	Mean age: ND
	Age range: 6–20 y
	Mean height: ND
	Male: ND
	Race: White
	Enrolled: 101
	Evaluated: 38
	Number of sites: 2
Reid 1988 89193370	Location: US	Achondroplasia	Seen only in orthopedics clinic (not in genetics clinic)	Achondroplasia	Prospective longitudinal cohort 20–45 mo
Possible overlap with Stokes 1988	Setting: Genetics clinic	Age < 7 y
	Mean age: ND	Seen in genetics clinic
	Age range: < 7 y
	Mean height: ND
	Male: ND
	Race: ND
	Enrolled: 62
	Evaluated: 26
	Number of sites: 1
Stokes 1988 88110722	Location: US	Achondroplasia	Acute medical or surgical problems	Achondroplasia	Prospective cross-sectional cohort
Possible overlap with Reid 1988	Setting: Hospital, genetics clinic, LPA convention		Hypochondroplasia
	Mean age: ND		Technically unsatisfactory spirometry
	Age range: ND		Too young to cooperate with spirometry
	Mean height: ND
	Male: 50%
	Race: ND
	Enrolled: 102 (including adults)
	Evaluated: 24 children
	Number of sites: 1
Reite 1972 72182015	Location: US	Children with “severe” osteogenesis imperfecta	None reported	Osteogenesis imperfecta	Retrospective cross-sectional
	Setting: Clinic
	Mean age: ND
	Age range: 6–17 y
	Mean height: ND
	Male: ND
	Race: ND
	Enrolled: ND Evaluated: 12
	Number of sites: 2
Rogers 1979 79200737	Location: US	Diagnosis of skeletal dysplasia leading to short stature	Syndromes of low-birth-weight dwarfism	Skeletal dysplasia and achondroplasia	Prospective cross-sectional cohort
	Setting: Pediatric clinic research unit		Multiple malformations
	Mean age: ND		Mucolipidosis or mucopolysaccharidosis
	Age range: 5 mo – 15 y		Cardiac arrest
	Mean height: ND
	Male: ND
	Race: ND
	Enrolled: 68 Evaluated: 68
	Number of sites: 1
Ruiz-Garcia 1997 97346439	Location: Mexico	“Classical” achondroplasia diagnosed in Genetics Department	History of neurosurgical procedures	Achondroplasia	Prospective cross-sectional cohort
	Setting: Genetics clinic
	Mean age: 4.5 y
	Age range: 3 mo – 17 y
	Mean height: ND
	Male: 49%
	Race: ND
	Enrolled: 39
	Evaluated: 39
	Number of sites: 1
Shurka 1976	Location: Israel	Dwarfism due to bone disease	(For current review, age ≥ 18 y)	Bone disease (External chondromatosis, Achondroplasia, Undefined)	Prospective cross-sectional cohort
	Setting: Pediatric endocrinology clinic
	Mean age: 15 y
	Age range: 9–17 y
	Mean height: 134 cm
	Height range: 104–157 cm
	Male: 46%
	Race: ND
	SES: Primarily average or high Enrolled: 12
	Evaluated: 11 (7 met criteria for current review)
	Number of sites: 1
Skeletal Dysplasia Group 1989 89192477	Location: UK	Diagnosis of skeletal dysplasia	Osteogenesis imperfecta	Morquio disease	Retrospective cross-sectional case series
	Setting: Workgroup's registers	Adequate cervical spine films (Morquio disease: other disease sub-groups did not meet criteria for current review)	Sclerosing bone dysplasias or tumor-like disorders
	Mean age: ND		Malformation syndromes
	Age range: 1–15 y
	Mean height: ND
	Male: 49%
	Race: ND
	Enrolled: 182
	Evaluated: 15 (children with Morquio disease)
	Number of sites: ND
Stewart 1989 90125184	Location: Scotland	Diagnosed based on their history of spontaneous fractures.	None reported	Osteogenesis imperfecta	Prospective cross-sectional cohort
	Setting: Dep. Of medical genetics	Contacted via Brittle Bon Society.
	Mean age: ND	For current review, only children aged 10–19 were included
	Age range: 10–19 y
	Mean height: ND
	Male: 43% for the larger group
	Race: ND
	Enrolled: 56
	Evaluated: 56 (13 met criteria for current review)
	Number of sites: ND
Thompson 1999 99221075	Location: US	Cases - clinical and radiographic diagnosis of ACH	Severe psychiatric disorder	Achondroplasia	Prospective cross-sectional cohort
	Setting: Genetics clinic	Premature controls- premature infants with hydrocephalus	Uncontrolled seizure
	Mean age:	Normal controls - normally developing children of average stature	Other neurological disorder
	Cases: 89 mo	All - IQ>69	Child abuse
	Premature: 107 mo		Sensory deficit
	Controls: 111 mo		Unable to speak English
	Age range: ND
	Mean height: ND
	Male: 52%
	Race: White 74%, Black 6%, Hispanic 13%, Other 7%
	SES: Low 31%, Middle 19%, High 50%
	Enrolled: 16 (21 premature, 17 controls)
	Evaluated: 16 (21, 17)
	Number of sites: 1
Waters 1993 94029092	Location: Australia	Patients with dysplasia (implied)	Refusal to participate	Achondroplasia	Prospective cross-sectional cohort
	Setting: Sleep lab and clinic
	Mean age: ND
	Age range: 1–31 y
	Mean height: ND
	Male: 50%
	Race: ND
	Enrolled: 20
	Evaluated: 20
	Number of sites: 1

Evidence Table 2. Studies Evaluating Short Stature Secondary to Skeletal Dysplasia Part II

Author, Year	Predictors	Predictor Measures / Definitions	Outcomes	Outcome Measures
Alston 1983	Osteogenesis imperfecta	ND	Cognitive/Academic	Non-verbal intelligence (Raven's Matrices) Reading (Primary Reading Test, Wide Span Reading Test, or Gapadol Reading Test)
				Spelling (Graded Word Spelling Test) Writing speed (Number of words written in 5 minutes)
			Motor function	Mobility (Walk vs wheelchair)
Apajasalo 1998 98163312	Achondroplasia	Disproportionate short limbed short stature, characteristic facial features and joint and spine disorders	Physical development	Overall quality of life (Health related quality of life, 16 dimensions)
	Cartilage-hair hypoplasia	Congenital disproportionate short-limbed short stature, thin/sparse hair, defective immunity		Mobility dimension
	Diastrophic Dysplasia	Severe physical handicap due to short-limbed short stature, progressive joint limitations, kyphoscoliosis, foot deformities		Vision dimension
				Hearing dimension
Bailey 197172080259	Achondroplasia	ND	Motor function	Elbow flexion (Angle, Self-reported disability)
Benson 1978 79027354	Osteogenesis imperfecta, mild	Thicker cortical bone with visible trabeculae, not severe osteoporosis of vertebrae	Motor function	Walking ability (Independent walkers, Walked with appliance, Sitter)
	Osteogenesis imperfecta, moderate	ND		Scoliosis (spinal curvature on spine radiography)
	Osteogenesis imperfecta, severe	Thin osteoporotic long bones and severe osteoporosis of vertebrae
Bleck 1981 82026195	Osteogenesis imperfecta congenita	Severe type	Physical development/Motor function	Independence (0 = No achievement possible
	Osteogenesis imperfecta tarda	ND		1 = Complete dependence
				2 = Partial dependence
				3 = Independence through technical help
				4 = Completely independent, but not normal in timing, effort expenditure, or external appearance
				5 = Normal activity)
				Mobility efficiency (0 = None
				1 = Mobility not functional enough to be used
				2 = Moves about immediate surroundings
				3 = Travels in the community
				4 = Travel beyond the community Activities of daily living (ADL: 0 = No ADL possible
				1 = Meets own essential needs
				2 = Completes personal care
				3 = Household activity possible
				4 = Normal ADL)
				Ambulation (0 = None
				1 = Walking possible but not functional enough to be useful
				2 = Household walker
				3 = Walk outside home but limited to neighborhood
				4 = Community walker)
Brinkmann 1993 94091382	Achondroplasia	Short limbs and characteristic skull changes with macrocephaly	Cognitive	Cognitive Abilities Test, German version
				Verbal component
				Arithmetic component
				Reasoning component
			Physical development	Motor development (Months to sit, stand, walk alone)
			Physical development	Delayed speech (parental report)
			Sensory	Hearing deficit (parental report)
			Pain	Limb, joint, back pain
Cox 1982 83016754	Osteogenesis imperfecta	Idiopathic bone fragility, blue sclerae, hearing impairment	Sensory	Hearing loss (Pure tone thresholds > 30 dB hearing loss at 250–1,000 Hz and/or > 25 db hearing loss at 2,000–6,000 Hz)
Daly 1996 96219608	Osteogenesis imperfecta Shapiro classification:	Fractures at birth, bones radiographically abnormal	Physical development/Motor function	Mobility (Wheelchair, Aided walker, Independent)
	Congenita A			Patterns of development (Abnormal arrested, Delayed arrested, Normal arrested, Delayed, Normal)
	Congenita B	Fractures at birth, bones radiographically normal
	Tarda A	First fracture by walking stage, bones narrow, osteopenic
	Tarda B	First fracture after walking, bones radiographically normal
	Osteogenesis imperfecta Sillence classification:	Autosomal dominant, normal stature, blue sclerae
	I A
	I B	IA plus dentinogenesis imperfecta
	III	Autosomal recessive, multiple fractures at birth, progressively deforming
	IV A	Autosomal dominant, bone fragility, normal sclerae
	IV B	IV A plus dentinogenesis imperfecta
Engelbert 1997	Osteogenesis imperfecta	Classified according to Sillence method (see Daly 1996 , above)	Motor function	Self-care (Pediatric Evaluation of Disability Inventory, PEDI, range 0–100, median 50. Score scaled for children older than 7.5 years such that normal score=100)
	I, III and IV			Mobility (PEDI)
Engelbert 1998 98366964	Osteogenesis imperfecta	Classified according to Sillence method (see Daly 1996, above)	Motor function	Scoliosis (Cobb angle > 40°)
	I, III and IV			Kyphosis (Cobb angle < 10° or > 40°)
Engelbert 1999 99381568	Osteogenesis imperfecta	Classified according to Sillence method (see Daly 1996, above)	Motor function	Level of ambulation (Bleck classification)
	I, III and IV
Engelbert 2000 20429012	Osteogenesis imperfecta	Classified according to Sillence method (see Daly 1996, above)	Motor function	Level of ambulation (Bleck classification)
	I, III and IV
Engelbert 2001 21334060	Osteogenesis imperfecta	Classified according to Sillence method (see Daly 1996, above)	Motor function	Self-care (Pediatric Evaluation of Disability Inventory, PEDI, range 0–100. Score scaled for children older than 7.5 years such that normal score=100)
	I, III and IV			Mobility (PEDI)
				Level of ambulation (Bleck classification)
			Motor function	Joint range of motion: arms, legs (0=normal, 4=maximally decreased)
				Muscle strength: arms, legs (0=no muscle activity, 5=normal muscle strength)
Falvo 1973 74013834	Osteogenesis imperfecta	Generalized osteoporosis, blue sclerae, dentinogenesis imperfecta, history of fracture	Pulmonary	Pulmonary function tests (Vital capacity (VC), Residual volume (RV), 1-second forced expiratory volume divided by forced vital capacity (FEV₁/FVC))
	Mild	No noticeable deformity of long bones		Arterial blood analysis (Oxygen partial pressure, carbon dioxide partial pressure)
	Moderate	Bowing of lower extremity long bones only
	Severe	Bowing of all long bones
Fowler 1997 97356792	Achondroplasia	ND	Physical development	Motor development (Denver Developmental Screening Test: Months to sit, stand, walk alone)
Pauli 1995 95193803	Achondroplasia	ND	Neurological	Neurological examination (Truncal tone, Limb tone, Arm strength, Leg strength)
				Neurological history (Developmental delays, Asymmetry, Weakness, Sensory abnormality, Arching / opisthotonus, Clonus, Seizure, Apnea)
			Pulmonary	Polysomnography (Central apnea, Central hypopnea, Obstructive apnea)
Hecht 1991 92152267	Achondroplasia	ND	Cognitive	Mental development index (Bayley Scales of Infant Development)
				Psychomotor development index (Bayley)
			Pulmonary	Sleep apnea
Hunter 199898299528	Chondrodysplasia	ND	Psychological	Depression (Beck Depression Inventories)
				Anxiety (Speilberger State-Trait Anxiety Inventories)
Kuurila 2000 20379515	Osteogenesis imperfecta	Classified according to Sillence method (see Daly 1996, above)	Sensory	Hearing loss (Pure tone < 20 dB, Conductive < 15 dB)
	I, III and IV
Norimatsu 1982 82163323	Osteogenesis imperfecta	Per Falvo 1973 (see above)	Motor function	Walking ability (Wheelchair bound, ambulates with assistance)
	Mild, Moderate, Severe			Scoliosis (Mild <30°; Moderate 30°-50°; Severe > 50°, by Cobb's method)
Poussa 1991 92054945	Diastrophic dysplasia	Short-limbed short stature, joint limitations, spinal and foot deformities, dysmorphic earlobes, cleft palate	Motor function	Scoliosis (>10° by Cobb's method)
Reid 1988 89193370	Achondroplasia	ND	Pulmonary	Obstructive sleep apnea (>5 obstructive or >2 central apneas per hour)
			Neurological	Paresis (not defined)
Stokes 1988 88110722	Achondroplasia	Rhizomelia, macrocephaly with midfacial hypoplasia, spinal canal narrowing	Pulmonary	Forced vital capacity
Reite 1972 72182015	Osteogenesis imperfecta	ND	Cognitive	IQ (WISC)
Rogers 1979 79200737	Achondroplasia	ND	Cognitive	IQ (Age< 2 y: Bayley Scales of Infant Development; Age 2–6 y: Stanford-Binet Intelligence Scale; Age 6–15 y: WISC)
	Other skeletal dysplasia	Multiple diagnoses
Ruiz-Garcia 1997 97346439	Achondroplasia	Autosomal dominant, abnormally short stature, macrocephaly, rhizomelia	Neurological	Neurological examination (Quadriparesis, Weakness, Sensory deficit)
				Compressive neural syndrome (craniocervical compression, lumbar compression)
			Hearing Loss	Brainstem auditory evoked response
Shurka 1976	Skeletal dysplasia	External chondromatosis, achondroplasia, or undefined	Cognitive	IQ (WISC)
Skeletal Dysplasia Group 1989 89192477	Morquio disease	ND	Neurological	Neurological complications (ND)
Stewart 1989 90125184	Osteogenesis imperfecta	ND	Sensory	Hearing loss (Pure tone < 30 dB, conductive < 15 dB)
Thompson 1999 99221075	Achondroplasia	ND	Cognitive	IQ (WISC-R)
				Academic achievement (WRAT-R)
			Motor function	Gross motor arm and leg coordination (McCarthy Scales)
			Visual-motor skills	Motor-based design copying task (Beery Test of Visual Motor Integration)
				Motor-free spatial matching task (Judgment of Line Orientation)
				Fine motor skills
Waters 1993 94029092	Achondroplasia	ND	Pulmonary	Sleep apnea (> 5 apneic episodes per hour)

Evidence Table 2. Studies Evaluating Short Stature Secondary to Skeletal Dysplasia Part III

Author, Year	Associations found	Potential Biases	Comments
Alston 1983	In children with osteogenesis imperfecta (n=40) mean non-verbal intelligence (104), reading ability (100) and spelling ability (102) were similar to control children (n=40, 105, 102, 104, respectively). Writing speed was significantly slower (51 words in 5 minutes) compared to controls (61 words).	None noted	Study was privately funded
	Of children with osteogenesis imperfecta (n=40), 38% required wheelchair use.
Apajasalo 1998 98163312	Among adolescents with skeletal dysplasia (n=19), overall quality of life (HRQOL=92) was significantly lower than in controls (n=239, 95) in univariate and multivariate analyses.	Few definitions of outcome measures	No data on funding source
	Among adolescents with skeletal dysplasias (n=19), the quality of life dimension mobility was significantly lower than for normal children. Children with diastrophic dysplasia and achondroplasia had mean scores of ~81 (from graph); those with cartilage-hair hypoplasia had a mean score of 100.	Reporting of analyses incomplete
	Those with skeletal dysplasia had similar scores to normal children for vision and hearing. Children with diastrophic dysplasia had mean scores of 94 for vision and 93 for hearing; those with achondroplasia had mean scores of 96 and 95, respectively; those with cartilage-hair hypoplasia had mean scores of 96 and 85, respectively.
	In the analysis of 16 dimensions of quality of life, there were no statistically significant differences among the different groups of children with skeletal dysplasia.
Bailey 1971 72080259	Infants with achondroplasia, aged 0–2 years, had mean elbow flexion deformity of 11°, ranging from 5°–25°; Children, aged 3–12 years, had mean elbow flexion of 19°, ranging from 10°–40°; adolescents, aged 13–20 years, had mean elbow flexion of 23.5°, ranging from 0°–40°.	No data on actual disability due to limited joint movement specifically in children.	No data on funding source
		Unknown number of subjects in relevant age ranges.
Benson 1978 79027354	Among children with osteogenesis imperfecta (n=126) 48% were wheelchair bound (sitters), 25% walked with appliances, and 28% were independent walkers.	Incomplete data available.	No data on funding source
	Among those children with osteogenesis imperfecta who had spine radiography performed (n=103), 38% had straight spines, 7% had scoliosis between 0°–9°, 19% had scoliosis between 10°–19°, 14% had scoliosis between 20°–49°, 14% had scoliosis between 50°–79°, and 9% ≥ 80°.	Incomplete reporting
Bleck 1981 82026195	Among children with osteogenesis imperfecta congenita, severe type (n=12), mean independence, mobility, activities of daily living, and ambulation scores after orthotic and mobility management were 2.6 (partial dependence-independence through technical help), 3.0 (travels in community), 2.5 (completes personal care-household activity possible), and 0.7 (no ambulation-walking possible but not functional), respectively.	No statistical analysis.	No data on funding source.
	Among children with osteogenesis imperfecta tarda (n=12), mean independence, mobility, activities of daily living, and ambulation scores after orthotic and mobility management were 3.4 (independence through technical help-completely independent but not normal), 3.8 (almost travel beyond community), 3.7 (almost normal activities of daily living), and 2.8 (almost walk outside home but limited to neighborhood), respectively.		No statistical comparison.
Brinkmann 1993 94091382	Among children with achondroplasia (n=30), mean Cognitive Abilities Scores were Total 48, Verbal 46, Arithmetic 49, and Reasoning 51. Compared to normal controls (n=30), Total (56), Verbal (54) and Arithmetic (55), children with achondroplasia had significantly lower scores; Reasoning scores (control 55) were not significantly different. Overall comparisons to siblings (n=30) and other short children (n=30) were similar.	Limited statistical analyses.	Study was government funded
	Children with achondroplasia (n=30) were able to sit alone at a median of 12 months (compared to 7 months for normal controls, n=30), according to parental reports, to stand alone at 14 months (10 months), and to walk alone at 18 months (12 months). Siblings (n=30) and short controls (n=30) had motor development similar to other controls. No statistical analysis reported.	No data on sibling controls for factors that were reported by parents.
	Among children with achondroplasia (n=27) 63% were reported by their parents to have hearing deficits, compared to 18% of short controls (n=28) and 19% of normal controls (n=27).
	Among children with achondroplasia (n=27) 41% were reported by their parents to have delayed speech development, compared to 14% of short controls (n=28) and 7% of normal controls (n=27).
	Among children with achondroplasia (n=30) limb, joint or back pain was present by parental report in 80%, compared to 43% of short controls (n=30) and 10% of normal controls (n=30).
Cox 1982 83016754	Of children with osteogenesis imperfecta (n=15) 33% had hearing loss	All subjects from 5 families with history of osteogenesis imperfecta.	Study was university funded
		Limited data on hearing function
Daly 1996 96219608	Among children with osteogenesis imperfecta classified by Shapiro as congenita B (n=31) 19% had abnormal arrested development, 19% had delayed arrested development, 29% had normal arrested development, 29% had delayed development, and 3% had normal development. Among these children, 87% were wheelchair bound and 13% were independent walkers.	Only a parental questionnaire	No data on funding source
	Among children with osteogenesis imperfecta tarda A (n=15) none had abnormal arrested development, 27% had delayed arrested development, 7% had normal arrested development, 20% had delayed development, and 47% had normal development. Among these children, 27% were wheelchair bound, 13% were aided walkers, and 60% were independent walkers.	No statistical analysis.
	Too few children had osteogenesis congenita A (n=1) or tarda B (n=4) for analysis.	Definitions of development unclear
	Among (the same sample of) children with osteogenesis imperfecta classified by Sillence as Type I (n=15) none had abnormal arrested or delayed arrested development, 7% had normal arrested development, 27% had delayed development, and 67% had normal development. Among these children, 7% were aided walkers and 93% were independent walkers.
	Among children with osteogenesis imperfecta Type III (n=29) 21% had abnormal arrested development, 34% had delayed arrested development, 28% had normal arrested development, 17% had delayed development, and none had normal development. Among these children, 97% were wheelchair bound and 3% were independent walkers.
	Among children with osteogenesis imperfecta Type IV (n=7) none had abnormal arrested development, 14% had delayed arrested development, 14% had normal arrested development, 43% had delayed development, and 29 had normal development. Among these children, 57% were wheelchair bound, 14% were aided walkers and 29% were independent walkers.
Engelbert 1997	Among children aged ≤ 7.5 years with osteogenesis imperfecta Type I (n=16) mean Self-care Functional Skills score on PEDI (normal median 50) was 39 and mean Mobility score was 31. Among children with Type III (n=8) mean Self-care score was 31 and Mobility score was 5. Among children with Type IV (n=6) mean Self-care score was 37 and Mobility was 48.	Interpretation of scales difficult due to different norms used and unclear description of scales	No data on funding source
	Among children aged > 7.5 years with osteogenesis imperfecta Type I (n=16) mean Self-care Functional Skills score on PEDI (normal = 100) was 100 and mean Mobility score was 100. Among children with Type III (n=6) mean Self-care score was 70 and mean Mobility was 41. Among children with Type IV (n=9) mean Self-care score was 93 and Mobility was 62.	No explicit comparison to normal children
		Unclear how much overlap in sample with other Engelbert papers.
Engelbert 1998 98366964	Among children with osteogenesis imperfecta scoliosis > 10° was present in 12% of those with Type I (n=17), 63% with Type III (n=16), and 71% with Type IV (n=14).	No comparison with normal controls	No data on funding source
	Among children with osteogenesis imperfecta kyphosis < 10° was present in 6% of those with Type I (n=17), 38% with Type III (n=16), and 14% with Type IV (n=14).	Unclear how much overlap in sample with other Engelbert papers.
	Among children with osteogenesis imperfecta kyphosis > 40° was present in 6% of those with Type I (n=17), 38% with Type III (n=16), and 14% with Type IV (n=14).
Engelbert 1999 99381568	At follow-up, of children with osteogenesis imperfecta Type I (n=19) 53% were community walkers, 26% were neighborhood walkers, 16% were household walkers, and 5% were exercise walkers.	No children with recent intramedullary fixation surgery or other disability or impairments.	No data on funding source
	Of children with Type III (n=13) 31% were household walkers, 8% were exercise walkers, 38% were bottom shufflers, 15% could sit unsupported, and 8% could only sit supported.	PEDI scale scores reported for small to very small samples of children. Appears to be earlier subset of data reported earlier in Engelbert 1997. Not analyzed here.
	Of children with Type IV (n=10) 10% were community walkers, 10% were neighborhood walkers, 10% were household walkers, 40% were exercise walkers, 20% could weight bear on feet, and 10% were bottom shufflers.	Unclear how much overlap in sample with other Engelbert papers.
Engelbert 2000 20429012	At follow-up, of children with osteogenesis imperfecta Type I (n=41) 59% were community walkers without aides, 5% were community walkers with aides, 10% were neighborhood walkers without aides, 12% were household walkers without aides, 7% were household walkers with aides, and 7% were therapy walkers with aides.	Unclear how much overlap in sample with other Engelbert papers.	No data on funding source.
	Of children with Type III (n=11) 9% were neighborhood walkers with aides, 27% were household walkers with aides, 19% were therapy walkers with aides, and 45% were non-walkers.	No comparison with normal children
	Of children with Type IV (n=18) 28% were community walkers without aides, 11% were community walkers with aides, 6% were neighborhood walkers with aides, 11% were household walkers with aides, 22% were therapy walkers with aides, and 22% were non-walkers.
Engelbert 2001 21334060	Among children aged > 8 years with osteogenesis imperfecta Type I (n=17) mean Self-care Functional Skills score on PEDI (normal = 100) was 100 and mean Mobility score was 92. Among children with Type III (n=11) mean Self-care score was 81 and mean Mobility was 55. Among children with Type IV (n=12) mean Self-care score was 100 and Mobility was 67.	Unclear how much overlap in sample with other Engelbert papers, especially those with PEDI analyses (1997 and 1999)	No data on funding source.
	Among children with osteogenesis imperfecta Type I (n=17) the average child was a community walker without aides. Among children with Type III (n=11) the average child was a therapy walker with aides. Among children with Type IV (n=12) the average child was a household walker without aides.	No comparison to normal children.
	Among children with osteogenesis imperfecta Type I (n=17), the average joint range of motion on a scale of 0 (normal) to 4 (maximally decreased) was 0.5 for arms and 0 for legs. Among children with Type III (n=11) the average arm range of motion was 1.8 and leg range of motion was 3.3. Among children with Type IV (n=12) the average arm range of motion was 1.3 and leg range of motion was 1.7.	Meaning of scales unclear.
	Among children with osteogenesis imperfecta Type I (n=17), the average muscle strength on a scale of 0 (no muscle activity) to 5 (normal muscle strength) was 4.5 for arms and 4.8 for legs. Among children with Type III (n=11) the average arm strength was 3.5 and leg strength was 3.6. Among children with Type IV (n=12) the average arm strength was 4.5 and leg strength was 3.8.
Falvo 1973 74013834	Among children with osteogenesis imperfecta mean pulmonary vital capacity was 85% of predicted normal (n=10), mean residual volume was 119% of predicted normal (n=10), mean FEV₁/FVC was 99% of predicted normal (n=10), mean arterial oxygen partial pressure was 88 mm Hg (n=8), and mean arterial carbon dioxide partial pressure was 35 mm Hg (n=8).	Limited analyses	Study was government and privately funded.
	“No patient had severe hypoxemia or hypercapnea. Reduction of VC and increase in RV were found only in patients with kyphoscoliosis. Other parameters of pulmonary function were within normal limits.”
Fowler 1997 97356792	Among children with achondroplasia, by parental report, the mean age of sitting unsupported was 12 months (n=37), the mean age of standing alone was 19 months (n=21), and the mean age of walking alone was 18 months (n=24). “Virtually all fine motor skills were delayed when compared with children without achondroplasia.”	Includes subjects in Pauli 1995	No data on funding source
		Data on motor development from few subjects enrolled in study
Pauli 1995 95193803	Among children with achondroplasia, 72% had decreased truncal tone (n=39), 70% had decreased limb tone (n=44), 32% had abnormal arm strength (n=40), and 50% had abnormal leg strength (n=40). No child had marked abnormalities in neurological examination.	Earlier subset of subjects in Fowler 1997	Study was hospital funded
	Among children with achondroplasia (n=52) 42% had an abnormality by neurological history. These included disproportionate developmental delays (10%), asymmetries (21%), weakness (14%), sensory abnormalities (4%), arching/opisthotonos (10%), clonus (8%), seizures (8%), and apnea (27%).
	Among children with achondroplasia (n=35) by polysomnography 40% had mild to severe central apnea, 66% had mild to severe central hypopnea, and 23% had mild obstructive apnea
Hecht 1991 92152267	Among infants with achondroplasia (n=13) mean BSID Mental Development Index score was 97 (1 or 2 infants had scores <70, text and table disagree) and mean Psychomotor Development Index score was 63 (62% scored <50).	5 of 13 patients were initially seen (and thus included) for apnea, motor delay or neurological complication	Study was government funded
	Among infants with achondroplasia (n=13) 77% had normal polysomnography. Two infants (15%) had obstructive apnea, 1 had hypoxemia and respiratory acidosis.	Reporting errors
Hunter 1998 98299528	Among children with chondroplasias (n=55) 7% had moderate or severe depression per the Beck Depression Inventories, which was not significantly different than among normal siblings (11%, n=35).	None noted	No data on funding source
	Among children with chondroplasias (n=55) the mean Speilberger Child State Anxiety score was 28.4 and Trait Anxiety score was 32.5, which were significantly lower (implying less anxiety) than general population means for boys (31.0 and 36.7, respectively) and girls (30.7 and 38.0, respectively).
Kuurila 2000 20379515	Among children with osteogenesis imperfecta (n=45), 7% had hearing loss: 2 children had conductive hearing loss and 1 had sensorineural deafness.	Descriptive study only	No data on funding source.
Norimatsu 1982 82163323	Among children with osteogenesis imperfecta congenita (n=8) 38% were wheelchair bound; the remainder walked with assistance.	Multiple subjects were closely related.	No data on funding source
	Among children with osteogenesis imperfecta tarda (n=14) 14% were wheelchair bound; the remainder walked with assistance. Among children with osteogenesis imperfecta congenita (n=8) 88% had severe scoliosis (59°–139°; one had mild scoliosis (17°). Among children with osteogenesis imperfecta tarda (n=14) 21% had severe scoliosis (55°–125°), 14% had moderate scoliosis (32° & 33°), 64% had mild scoliosis (5°–28°).	Descriptive study only
Poussa 1991 92054945	Among children with diastrophic dysplasia (n=38) 29% had scoliosis. Of those ≤ 10 years (n=17) 6% had scoliosis; of those between 11 and 20 years (n=21) 48% had scoliosis. Those with scoliosis (n=11) had a mean magnitude of curvature of 39° (range 15°–80°).	Descriptive analysis only	No data on funding source.
			Subjects are part of the study. Descriptive data only.
Reid 1988 89193370	Among children with achondroplasia (n=26) 85% had a history of respiratory abnormalities including pneumonia, loud snoring, cyanotic spells or apnea in the preceding 6 months; 35% had obstructive sleep apnea.	Descriptive study only	No data on funding source
	Among children with achondroplasia (n=26) 42% had paresis.	No specific definitions of conditions reported
		Appears to be subgroup of subjects included in Stokes 1988
Stokes 1988 88110722	Among children with achondroplasia (n=24) mean forced vital capacity was 72% of predicted, significantly lower than population norms.	Limited details of pediatric sub-sample.	Study funded by government, hospital and private sources
		Appears to include subjects in Reid 1988
Reite 1972 72182015	Among children with osteogenesis imperfecta (n=12) the mean IQ was 107, ranging from 78–133.	Only children with “severe” osteogenesis imperfecta	Study was government and privately funded.
		Incomplete data reported
Rogers 1979 79200737	Among school aged children with achondroplasia (n=19) mean Full scale IQ was 96, Verbal IQ was 95, and Performance IQ was 100; 5% had IQ<70.	Descriptive study only	Study was government funded
	Among preschool aged children with achondroplasia (n=15) mean Full scale IQ was 97; 13% had IQ<70.
	Among school aged children with other skeletal dysplasias (n=22) mean Full scale IQ was 104, Verbal IQ was 104, and Performance IQ was 103; none had IQ<70.
	Among preschool aged children with other skeletal dysplasias (n=12) mean Full scale IQ was 100; none had IQ<70.
Ruiz-Garcia 1997 97346439	According to summary table, among children with achondroplasia (n=39) 59% had hypotonia, 15% had quadriparesis, 31% had weakness, and 10% had a sensory deficit.	No definitions of outcomes	No data on funding source
	According to text, 49% had hypotonia, 13% had quadriparesis, 26% had weakness, and 5% had sensory deficit. In additon, 31% had compressive neural syndromes (18% craniocervical, 13% lumbar).	Results reported in text differ from results reported in table
	According to the text, among children with achondroplasia (n=32), 50% had abnormal hearing as tested by brainstem auditory evoked response.	Descriptive study only.
Shurka 1976	Among children with skeletal dysplasias (n=7) mean Full scale IQ was 102, Verbal IQ was 103, and Performance IQ was 99.	Very small sample size.	No data on funding source
		Multiple, poorly defined causes of dwarfism.
Skeletal Dysplasia Group 1989 89192477	Among children with Morquio disease (n=15) 33% had “known neurological complications.”	Limited data reported	No data on funding source
		Conditions and outcomes not defined
		Only subset met criteria for inclusion in SSA report
Stewart 1989 90125184	Among children with osteogenesis imperfecta (n=13), 15% had hearing loss.	Incomplete reporting	Study was private funded
		Descriptive only
Thompson 1999 99221075	Among children with achondroplasia mean WISC-R Verbal IQ was 94 (n=16), Performance IQ was 101 (n=16), WRAT-R Spelling score was 88 (n=14), Arithmetic was 89 (n=15), and Sentence writing was 80 (n=12). These scores were all statistically similar to normal controls (IQ ND, Spelling 97, Arithmetic 96, Sentence writing 95, n=17).	Biased sample as some children dropped who were unable to complete the testing for many measures and excluded children with IQ<69.	Study was government funded.
	Among children with achondroplasia (n=13) Gross Motor Arm Coordination score on McCarthy Scales was 73 and Leg Coordination was 79, significantly lower than normal controls (93, 104, respectively; n=13).		Subjects were recruited as part of larger previous study of hydrocephalus
	Among children with achondroplasia (n=13) visual motor skills were significantly lower (Beery 82, JLO 85) than among normal controls (Beery 92, JLO 100, n=12). Various measures of fine motor skills were similar between the two groups.
Waters 1993 94029092	Among children with achondroplasia (n=20) 75% had > 5 apneic episodes per hour of sleep.	Descriptive study only	No data on funding source

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Asthma - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Balfour-Lynn 1986	Location: UK	Chronic perennial asthma	None reported	Asthma	Prospective longitudinal cohort
87074993	Setting: Pediatric asthma clinic				(Mean 8.9 y)
	Initial mean age: 7.5 y
	Initial age range: 2–12 y
	Follow-up mean age: 21 y
	Follow -up age range: 13–26 y
	Mean height: ND
	Male: 68%
	Race: ND
	SES: ND
	Enrolled: 66
	Evaluated: 66
	Number of sites: 1
Cernelc 1975	Location: Yugoslavia	Asthmatic children with normal or pathological lung function tests	None reported	Asthma	Prospective cross-sectional
76016195	Setting: Clinic
[Also in cardiac section]	Mean age: ND
	Mean height: ND
	Male: ND
	Race: ND
	SES: ND
	Enrolled: ND
	Evaluated: 337
	Number of sites: ND
Ferguson 1982	Location: Canada	Children with a diagnosis of asthma and short stature (height < 3^rdpercentile)	None reported	Asthma	Preliminary study of a prospective longitudinal study
82190765	Setting: Regional
	Mean age: ND
	Age range: ~3–17 y
	Mean height: ND
	Male: 89%
	Race: ND
	SES: ND
	Enrolled: 36
	Evaluated: 36
	Number of sites: 1
Hauspie 1979	Location: Belgium and Hungary	Hungary : asthmatic boys residing in the National Sanatorium for Children	None reported	Asthma /Respiratory disease	Retrospective cross-sectional cohort
80114229	Setting: Sanatorium; Asthma clinic	Belgium: boys with chronic aspecific respiratory affection, asthma or asthma and eczema.
	Mean age: ND
	Age range: 3–16 y
	Mean height: ND
	Male: 100%
	Race: ND
	SES: ND
	Enrolled: 1107
	Evaluated: 1107
	Number of sites: 2
Klein 1991	Location: US	Children with moderately severe, non-steroid-dependent asthma	> 6 short courses of steroids during previous year	Asthma	Prospective cross-sectional
91314975	Setting: Clinic
	Mean age: 8.5 y
	Age range: 2–15 y
	Mean height: ND
	Male: ~ 69%
	Race: ND
	SES: ND
	Enrolled: 176
	Evaluated: 176
	Number of sites: 4
Martin 1981	Location: Australia	Asthmatic children	None reported	Asthma	Prospective longitudinal cohort
82110081	Setting: Region	Aged 7 y in 1964			(14 y follow-up)
	Initial age: 7 y
	Follow-up ages: 10 y, 14 y, 21 y
	Mean height: ND
	Male: 63%
	Race: ND
	SES: ND
	Enrolled: 375
	Evaluated: 315
	Number of sites: 1
McNicol 1970	Location: Australia	Asthmatic children	None reported	Asthma	Prospective longitudinal cohort
71076839	Setting: Unclear				(Ages 7 and 10)
	Ages: 7 and 10 y
	Mean heights: 123 cm (age 7 y); 138 cm (age 10)
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 226
	Evaluated: 226
	Number of sites: ND
Neville 1996	Location: UK	Tayside Childhood Asthma Project (1990-92)	Loss to follow-up	Asthma	Secondary analysis of prospective longitudinal cohort
96427603	Setting: Pediatric practices	Possible asthma, based on medication prescriptions, from chart review
	Mean age: ND	Age 3.5–13 y
	Age range: 3.5–13 y	Height and weight measurements available
	Mean height: ND
	Male: 57%
	Race: ND
	SES: “Broad range”
	Enrolled: 2915
	Evaluated: 699
	Number of sites: 12
Rona 1980	Location: UK	Parental reporting of asthma in last 12 months	None reported	Asthma	Retrospective cross-sectional cohort
81025712	Setting: Schools
	Mean age: ND
	Age range: 5–11 y
	Mean height: ND
	Male: 68%
	Race: ND
	SES: Proportionally more lower SES areas
	Enrolled: 102
	Evaluated: 102
	Number of sites: 28
Sant'Anna 1996	Location: Brazil	Children with asthma followed in Division of Allergy, Clinical Immunology and Rheumatology	Phenotypical deviations	Asthma	Retrospective cross-sectional case series
97295750	Setting: Pediatric immunology clinic		Clear changes in body proportions
	Mean age: ND		Other chronic diseases
	Age range: 6 mo – 16 y		Rash and/or atopic dermatitis
	Mean height: ND
	Male: 58%
	Race: ND
	SES: ND
	Enrolled: 514
	Evaluated: 514
	Number of sites: 1
Spock 1965	Location: US	Children with bronchial asthma	Concomitant chronic illness	Asthma	Retrospective longitudinal cohort
	Setting: Pediatrics department	Age > 4 y at the first visit	Cyproheptadine use		(19 y)
	Mean age: 7 y	Followed for minimum of 3 y
	Age range: 4–12 y
	Mean height: ND
	Male: 65%
	Race: White
	SES: ND
	Enrolled: 200
	Evaluated: 200
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Asthma - Part II

Author, Year	Severity of Illness Groups	Outcomes
Balfour-Lynn 1986	Treatment required to control asthma	Initial growth retardation (not defined)
87074993	• Bronchodilators alone or children whose asthma resolved before puberty	Growth along centile lines
	• Sodium cromoglycate and bronchodilators in children whose asthma did not improve at puberty	Final adult height
	• Steroids and bronchodilators	Percentile, based on Tanner and Whitehouse growth chart
Cernelc 1975	Pulmonary function tests	Height by age (linear regression)
76016195	• Normal	Compared to normal values for Slovenian children
	• Pathological (not defined)
Ferguson 1982	Asthma symptom score, based on frequency of wheezing in past year, duration of wheezing, level of therapy required, and bronchodilator medication days in past year (16 point scale)	Height
82190765	• Mild: score 1–7	Percent average height for age
	• Moderate: score 8–12
	• Severe: score 13–16
Hauspie 1979	Hungarian sample: Ovsáth's system, based on based on frequency of attacks, seriousness of dyspnea, duration of dyspnea, and number of symptom-days per year	Height Z-scores
80114229	• Class II: Less severe
	• Class III: More severe
	• Class IV: Most severe
	Belgian sample: Disease diagnosis
	• Chronic aspecific respiratory affection
	• Asthma
	• Asthma and eczema
	Belgian sample: Treatment required to control asthma:
	• No corticotherapy
	• Interrupted corticotherapy
	• Continuous corticotherapy for at least 6 months
Klein 1991	Asthma severity score (by event), based on hospitalization and medication requirements in previous year (score: 0–60)	Height
91314975	• Mild: score < 25	Height percentile
	• Moderate/severe ≥ 25	Based on National Center for Health Statistics Growth Chart
Martin 1981	Asthma grade, based on frequency and recency of wheezing	Percentage of subjects <10^thpercentile
82110081	• Grade A: ≤ 5 episodes of wheezing	Based on Australian Department of Health
	• Grade B: > 5 episodes of wheezing but no wheezing within 12 months of exam	Anthropometric tables
	• Grade C: Episodic asthma over a number of years and wheezing within 12 mo of exam	At ages 10, 14, and 21 y
	• Grade D: Very frequent or chronic unremitting asthma
McNicol 1970	Asthma grade, based on frequency and recentness of asthma episodes	Height
71076839	• Grade I: ≤ 5 episodes of asthma
	• Grade II: ≤ 20 episodes of asthma or no asthma within 12 mo of exam
	• Grade III: >20 episodes and still having asthma within 12 mo of exam
Neville 1996	Asthma severity (British Thoracic Society treatment steps), based on medication requirements	Height
96427603	• Step 1: Beta-agonist only in past year	SDS, based on control group within study
	• Step 2: Beta-agonist and cromoglycate
	• Step 3: Beta-agonist and low dose inhaled steroid
	• Step 4: Beta-agonist and > 400 μg/day inhaled steroid
Rona 1980	Frequency of asthma episodes	Height SDS, based on Scottish and English norms
81025712	• ≤ 2 episodes in the past year	• Unadjusted
	• ≥ 3 episodes in the past year	• Adjusted for parents' height, social class, number of siblings
	Hospitalization for asthma	Height gain, cm/year
		• Unadjusted
		• Adjusted
Sant'Anna 1996	Asthma severity, based on the International Consensus Report on Diagnosis and Treatment of Asthma	Percentage with short stature
97295750	• Mild	< 3^rdpercentile, based on National Center of Health Statistics, US, 1977
	• Moderate
	• Severe
Spock 1965	Clinical status (not defined)	Linear growth rate
	• Poor	Percentage of normal, based on Stuart and Wetzel's anthropometic charts
	• Fair
	• Good
	• Excellent

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Asthma - Part III

Author, Year	Associations found	Potential Biases	Comments
Balfour-Lynn 1986	Comparing children whose asthma was controlled with bronchodilators alone or had resolved at puberty (n=18, Group I) with those whose asthma required sodium cromoglycate and had not resolved at puberty (n=19, Group II), and with those whose asthma required treatment with steroids (n=29, Group III), there was no statistical difference in growth retardation at the start of the study. However, “growth retardation” was not defined. Of children in Group I, 2 of 18 (11%) had initial height < 10^th percentile. Of children in Group II, 1 of 19 (5%) had initial height < 10^th percentile. Of children in Group III, 3 or 29 (10%) had initial height < 10^th percentile. No child had height < 3^rd percentile.	Severity of asthma based on treatment used, not objective measures	No data on funding source
87074993	All children “grew along expected centile lines without appreciable deviation, from initial age (mean 7.5 y) until age 10 y, regardless of treatment required. 30 of 66 (45%) children had “physiological decelerating growth velocity pattern of delayed puberty.” No data are reported regarding association with asthma severity.	Reported data incomplete
	Comparing the three groups of asthmatic children, based on treatment requirements, no statistical difference was found in final adult height. Possible minor trend toward taller adult height from Group I to II to III seen in graph.	Study sample poorly defined
		6 subjects lost to follow up prior to obtaining final adult height but were followed for mean 8.3 years (6.3–14.6 y)
Cernelc 1975	Asthmatic children with normal pulmonary function tests (n=225) were not significantly different in height for age compared to asthmatic children pathological pulmonary function tests (n=112).	Limited data and analysis	No data on funding source
76016195	No data on growth velocity.	Study sample poorly defined	Data on asthmatic children with bronchiectasis (n=22) not included because children reported to be both significantly and not significantly more likely to have growth retardation
Ferguson 1982	The percentages of average normal height for age (50^th percentile) of short children whose asthma was mild (n=19, 90.7%), moderate (n=13, 90.2%) and severe (n=4, 88.5%) were not significantly different among the three groups.	Study limited to children with height < 3^rd percentile, however results reported as percentage of average normal height	Study was government and private funded.
82190765	No significant correlation was found between symptom score and percentage of average normal height for age of individual subjects (r=0.20).
	No data on growth velocity
Hauspie 1979	Mean height of Class II (relatively mild) Hungarian asthmatic boys (n=222) is normal (~ 0 SDS, from graph). The mean height of Class III asthmatic boys (n=253) was significantly lower (~ -0.3 SDS) and that of Class IV asthmatic boys (n=25) was even lower (~ -0.9 SDS). “A gradually increasing degree of growth retardation in relation to the seriousness of the disease can be observed.”	Chronic aspecific respiratory affection includes illnesses other than asthma.	No data on funding source.
80114229	In a separate sample of Belgian asthmatic boys, those with chronic aspecific respiratory affection (n=204) had normal mean height (~ -0.1 SDS, from graph). Those with asthma (n=323) were statistically significantly but minimally shorter (~ -0.2 SDS) and those with asthma and eczema (n=80) were even shorter (~ -0.4 SDS).	Data reported mostly in graphical format only
	Belgian asthmatic boys requiring no corticotherapy (n=245) and requiring interrupted corticotherapy (n=130) had normal mean height (~ -0.1 SDS, from graph). Those who required continuous corticotherapy (n=16) were statistically shorter (~ -0.5 SDS).	Study limited to boys
	No data on growth velocity.
Klein 1991	Children (n=176 total) with mild asthma (ND on n) had mean ± SD height of 145±10 cm; those with moderate asthma were on average 148±10 cm. The height difference was not significantly different. No data reported on children with severe asthma.	Reported data very incomplete.	No data on funding source
91314975	“There was no relationship between the severity of asthma (as assessed by a clinical and medications score) and the height percentile.” Reported data are incomplete.	Mean heights reported as 44.8 cm and 47.8 cm. We understood this to mean 144.8 cm and 147.8 cm, however, reported numbers may refer to percentiles.
	No data on growth velocity	Children requiring > 6 short courses of steroids within previous year excluded.
Martin 1981	The prevalence rates of height < 10^th percentile in 10 years old children whose asthma was rated Grade A (mild, n=64), Grade B (moderate, n=88), Grade C (severe, n=106), and Grade D (very severe, n=57) were 9%, 8%, 8%, and 15% respectively. Only children with very severe asthma had significant higher prevalence of height < 10^th percentile at age 10 y compared to non-asthmatic children (n=62, 4%).	None noted.	No data on funding source
82110081	The prevalence of short stature in the asthmatic children when they were 14 years old was similar to when they were 10 years old. Height < 10^th percentile in Grade A asthmatics was 6%, Grade B 5%, Grade C 5%, and Grade D 17%. Only children with very severe asthma had significant higher prevalence of height < 10^th percentile at 14 years of age compared to non-asthmatic children (1%).
	At age 21 years, the prevalence of short stature was similar among all asthma grades. Height < 10^th percentile Grade A asthmatics was 15%, Grade B 15%, Grade C 9%, and Grade D 19%. There was no difference in the prevalence of height < 10^th percentile at 21 years of age, between the 4 groups and non-asthmatic children (12%).
	No direct data on growth velocity
McNicol 1970	The mean heights of children at both ages 7 and 10 years were the same (123–124 cm at age 7 years; 137–139 cm at age 10 years) regardless of frequency of asthmatic episodes among 276 asthmatic children and 94 non-asthmatic controls. No significant difference was found between controls and any degree of asthma severity.	None noted	No data on funding source
71076839	No data on growth velocity
Neville 1996	Asthmatic children who used only beta agonists (n=537) had a mean height of +0.04 SDS. Those who used beta agonists and cromoglycate (n=64) had a mean height of +0.04 SDS. Those who used beta agonists and low dose inhaled steroids (n=39) had a mean height of +0.08. Those who used beta agonists and > 400 μg/day inhaled steroids (n=59) had a mean height of -0.38. Only those on high dose inhaled steroids were significantly shorter than normal.	Large number of potential subjects excluded due to missing height data	Study was government, privately, and pharmaceutical funded
96427603	No data on growth velocity.	Statistical analysis implied only
		Asthma severity based on medication use
Rona 1980	Among children with asthma episodes, those with ≤ 2 episodes in previous year (n=33) were not significantly shorter (-0.1 SDS, - 0.2 SDS, respectively) than controls (n=4334) in both unadjusted model and model adjusted for parental height, social class, and number of siblings. Those with ≥ 3 asthma episodes in previous year (n=69) were significantly shorter (-0.4 SDS) than controls in both models. Overall, “there was a highly significant negative relationship between unadjusted [and adjusted] height and number of [asthma] episodes.	Retrospective survey of parents (for medical history) and school records	Study was government funded.
81025712	“Hospitalization for asthma was more closely associated with shorter stature for boys than for girls, but the number of children hospitalized was too small for useful analysis.”	Logic of adjusting for number of siblings unclear.
	In adjusted model, height velocity was non-significantly lower in boys with ≤ 2 asthma episodes per year (n=22, 5.34 cm/y) and with ≥ 3 episodes per year (n=41, 5.57 cm/y) than non-asthmatic boys (n=1873, 5.72 cm/y), and in girls with ≤ 2 episodes per year (n=9, 5.67 cm/y) and with ≥ 3 episodes per year (n=16, 5.50 cm/y) than non-asthmatic girls (n=1552, 5.93 cm/y).
Sant'Anna 1996	The prevalence of short stature (< 3^rd %tile) in children with mild asthma (n=211, 6%) was non-significantly lower than those with moderate asthma (n=154, 11%) and severe asthma (n=149, 11%).	Multiple height-related exclusion criteria	No data on funding source
97295750	No data on growth velocity
Spock 1965	The mean linear growth rate in percentage of normal was 109% for asthmatic patients with excellent clinical status (n=68), 107% for asthmatic patients with good clinical status (n=86), 102% for asthmatic patients with fair clinical status (n=30), and 97% for asthmatic patients with poor clinical status (n=16). The correlation between severity of clinical status and the linear growth rate in percentage of normal was not statistically significant.	Degree of clinical status not defined	No data on funding source.
	“No statistically significant correlation was found between the duration and/or severity of asthmatic symptoms and growth impairment for height or weight.”	The percentage of children requiring steroids was unusually low at 12%

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Cardiac Disease - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Baum 1980	Location: US	Diagnosis of acyanotic congenital heart disease	None reported	Congenital heart disease:	Prospective cross-sectional cohort
81065058	Setting: Clinic	Age 2 to 6 y		Patent ductus arteriosus
	Mean age: 3.8±1.0 y			Ventricular septal defect
	Age range: ND			Atrial septal defect
	Mean height: 99 cm			Pulmonic stenosis
	Male: 77%			Aortic stenosis
	Race: ND			Atrioventricular canal
	SES: ND			Pulmonary artery band
	Enrolled: 26			Right ventricular band
	Evaluated: 26			Mitral stenosis
	Number of sites: 1			Mitral insufficiency
Cernelc 1975	Location: Yugoslavia	Children with congenital heart disease with or without cyanosis	None reported	Congenital heart disease:	Prospective cross-sectional
76016195	Setting: Clinic			Not defined
[Also in asthma section]	Mean age: ND
	Mean height: ND
	Male: ND
	Race: ND
	SES: ND
	Enrolled: ND
	Evaluated: 22
	Number of sites: ND
Feldt 1969	Location: US	Diagnosis of congenital heart disease	Trisomy 21	Congenital heart disease:	Prospective longitudinal cohort
69176079	Setting: Clinic		Gonadal dysgenesis	Ventricular septal defect	(5 mo)
	Mean age: ND		Clinically recognized maternal rubella syndrome	Tetralogy of Fallot
	Age range: 1 wk – 17 y			Aortic stenosis
	Mean height: ND			Pulmonary stenosis
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 463
	Evaluated: 311
	Number of sites: 1
Levy 1977	Location: US	Diagnosis of ventricular septal defect	None reported	Ventricular septal defect	Retrospective longitudinal cohort
78125814	Setting: Cardiology clinic				(mean[range] 6.4 [4–8] y)
	Mean age: ND
	Age range: ND
	Mean height: ND
	Male: 50%
	Race: ND
	SES: ND
	Enrolled: 1,210
	Evaluated: 777
	Number of sites: ND
Strangway 1976	Location: Canada	Diagnosis of congenital heart disease	Additional major congenital lesion	Congenital heart disease:	Retrospective longitudinal cohort
76101945	Setting: Cardiology clinic		Mental retardation	Tetralogy of Fallot	(5 y)
	Mean age: ND		Undergone corrective surgery	Ventricular septal defect
	Age range: < 2–11 y		Acute or chronic intercurrent illness	Atrial septal defect
	Mean height: ND			Pulmonary stenosis
	Male: ND			Aortic stenosis
	Race: ND			Coarctation of aorta
	SES: ND			Patent ductus arteriosus
	Enrolled: 568			Arterioventricular canal
	Evaluated: 568			MIscellaneous
	Number of sites: 1
White 1970	Location: US	Diagnosis of congenital heart disease	Tetralogy of Fallot with total corrective surgery	Congenital heart disease:	Prospective cross-sectional
72024496	Setting: Adolescent cardiac clinic	Age < 19 y		Ventricular septal defect
	Mean age: ND			Atrial septal defect
	Age range: 12–18 y			Patent ductus arteriosus
	Mean height: ND			Pulmonary stenosis
	Male: ND			Aortic stenosis
	Race: White 96%			Coarctation of aorta
	SES: Generally middle class			Aortic insufficiency
	Enrolled: 80			Undiagnosed acyanotic
	Evaluated: 80
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Cardiac Disease - Part II

Author, Year	Predictors (N)	Outcomes
Baum 1980	Heart Failure (not defined)	Height by age (linear regression)
81065058		Height percentile
		Based on National Center for Health Statistics growth charts
Cernelc 1975	Cyanosis	Height by age (linear regression)
76016195		Compared to normal values for Slovenian children
[Also in asthma section]
Feldt 1969	Pulmonic valve pressure gradient	Growth failure preoperative height (and weight) ≤ -2 SDS
69176079	Aortic valve pressure gradient	Height SDS based on Reed and Stuart, 1959
	Tetralogy of Fallot lesion severity
	• Hematocrit
	• Oxygen saturation (cyanosis)
	• Magnitude of right to left shunt
	• History of hypoxemia
	Ventricular septal defect shunt severity
	• Pulmonary vascular disease (pressure)
Levy 1977	Ventricular septal defect severity	Height SDS
78125814	• I: Trivial to mild, pulmonary artery mean pressure < 20 mm Hg
	• II: pulmonary artery mean pressure ≥ 20 mm Hg and pulmonary to systemic resistance ratio < 0.20
	• III-IV: Severe: pulmonary artery mean pressure ≥ 20 mm Hg and pulmonary to systemic resistance ratio ≥ 0.20
Strangway 1976	Cardiac problems	Height velocity fraction of the normal mean for the same age and sex
76101945	• Cyanosis
	• Cardiac enlargement (cardiothoracic ratio > 59% age < 2 y, > 55% age >2 y)
	• Congestive heart failure
White 1970	Cyanosis	Short stature, height < 2^ndpercentile
72024496

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Cardiac Disease - Part III

Author, Year	Associations found	Potential Biases	Comments
Baum 1980	Children with congenital heart disease with early heart failure (n=12) were significantly shorter for age than asymptomatic children (n=14). On average, children age 2–6 years with heart failure were 5–9 cm shorter than those without. The differences in height were greater with older age.	None noted	No data on funding source.
81065058	Six (50%) of the 12 children with early heart failure had height below the 5^th percentile, while none of the 14 asymptomatic children had height below the 10^th percentile.
	No data on growth velocity.
Cernelc 1975	Children with congenital heart disease and cyanosis (n=11) were significantly more likely to be short for age than normal population, as opposed to children with congenital heart disease without cyanosis (n=11) who were not significantly different in height for age than normal population.	Limited data and analysis	No data on funding source
76016195		Study sample poorly defined
[Also in asthma section]
Feldt 1969	Among children with aortic or pulmonary stenosis (n=73), those “with mild pressure gradients (< 50 mm Hg) across either the pulmonic or the aortic valve usually were of normal weight and height. In contrast, patients with severe valvular gradients were more likely to have severe growth failure.”	No statistical analyses	Study was government funded.
69176079	The severity of growth failure of children with Tetralogy of Fallot (n=83) “was not related to any criteria for seveity of the lesion such as hemaglobin value, oxygen saturation…, magnitude of right to left shunt, or history of hypoxemic spells.” Specifically height SDS was not associated with hemoglobin level.	Growth retardation included both short stature and low weight
	Among children with ventricular septal defect (n=155) “there was no correlation between the extent of growth failure and the degree of pulmonary vascular disease assessed either at cardiac catheterization of from pressure measurements taken at the time of operation.”
	No data correlated height velocity to severity of disease.
Levy 1977	For children with ventricular septal defect, the mean initial height SDS for those with mild disease (n=492) was -0.23 SDS, for those with moderate disease (n=170) -0.63 SDS, and for those with severe disease (n=115) -0.86. The height of children in each category was statistically significantly lower than normal children. The statistical significance of the association between height and severity of disease is implied only.	Reporting of statistical analyses unclear and incomplete	Study was government funded.
78125814	No data on growth velocity
Strangway 1976	Among children aged 0 to 2 years with congenital heart disease (n=181), height velocity percentile was significantly lower in children with cyanosis (-14%) than those without cyanosis (+7%), was similar in children with cardiac enlargement (+2%) as without cardiac enlargement (+1%), and was non-significantly lower in children with congestive heart failure (-20%) than without congestive heart failure (+2%).	Because of the calculation of height velocity, some subjects missing follow-up measures of height were excluded from analyses.	Study was government funded.
76101945	Among children aged 2 to 11 years with congenital heart disease (n=387), height velocity percentile was similar in children with cyanosis (+10%) as those without cyanosis (+7%), was similar in children with cardiac enlargement (+15%) as without cardiac enlargement (+6%), and was non-significantly lower in children with congestive heart failure (-13%) than without congestive heart failure (+7%).
White 1970	Among children with congenital heart disease, height < 2^nd percentile was more common among those with cyanosis (n=27, 19%) than those without cyanosis (n=53, 2%)	No statistical analyses performed	Study was partial government funded.
72024496		Reporting of data unclear

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Insulin Dependent Diabetes Mellitus - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Arreola 1991 93191566	Location: Mexico	Patients with type I diabetes seen at outpatient diabetes clinic from 1984 to 1998	None reported	Insulin dependent diabetes	Retrospective longitudinal cohort (4–5 y)
	Setting: Outpatient diabetes clinic	Age 1–16 y
	Mean age: ND
	Age range: 1–16 y
	Mean height: ND
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 198±8 each year (1984-1988)
	Evaluated: 198±8 each year
	Number of sites:1
Court 1982 83021090	Location: UK	Children with diabetes attending clinics in the Newcastle region	None reported	Diabetes (implied insulin dependent)	Retrospective cross-sectional cohort
	Setting: Region
	Mean age: 13±3.7 y
	Age range: ND
	Mean height, SDS: -0.69
	Male: 54%
	Race: ND
	SES: ND
	Enrolled: 121
	Evaluated: 111
	Number of sites: ND
Herber 1988 88180314	Location: UK	Patients with Type I diabetes who have attended the pediatric diabetic clinic for at least 3 years	None reported	Insulin dependent diabetes	Retrospective longitudinal cohort (3 y)
	Setting: Pediatric diabetic clinic
	Initial mean age: 10 y
	Follow-up mean age: 13 y
	Age range: ND
	Initial height, SDS: -0.24±1.16
	Follow-up height, SDS: -0.20±1.08
	Male: 64%
	Race: ND
	SES: ND
	Enrolled: 67
	Evaluated: 67
	Number of sites: 1
Izumi 1995 96105591	Location: Japan	Children participated in Summer camp program	None reported	Insulin dependent diabetes mellitus	Retrospective longitudinal cohort (18 y)
	Setting: Summer camp	More than 3 determinations of Hgb_A1c
	Initial mean age: ND	Age < 18 y at final examination.
	Follow-up mean age: 14 y
	Age range: ND
	Mean height: ND
	Male: 40%
	Race: ND
	SES: ND
	Enrolled: 385
	Evaluated: 107
	Number of sites: ND
Jivani 1973 73134278	Location: UK	Children with diabetes > 3 y follow-up	Other condition which might have affected growth (12 subjects with fair control were not analyzed)	Diabetes (implied insulin dependent)	Retrospective longitudinal cohort (mean [range] 7 [3–13] y)
	Setting: Children's hospital
	Mean age: ND
	Age range: 9 mo - 13 y
	Mean height: ND
	Male: 46%
	Race: ND
	SES: ND
	Enrolled: 116
	Evaluated: 104
	Number of sites: 1
Pitukcheewanont 1995 96136504	Location: US	Type I diabetes	< 3 visits during study	Insulin dependent diabetes	Retrospective longitudinal cohort (6 y)
	Setting: Clinic		Other problems that might affect growth (thyroid disease, cerebral palsy, foster care placement, methylphenidate therapy)
	Mean age: 11.5±3.8 y		Precocious or delayed puberty
	Age range: ND
	Mean height: ND
	Male: 49%
	Race: White 100%
	SES: ND
	Enrolled: 82
	Evaluated: 82
	Number of sites: 1
Rosenbloom 1982 83032889	Location: US	Children with Type I diabetes	None reported	Insulin dependent diabetes mellitus	Prospective longitudinal cohort (ND)
	Setting: Clinic	Pre-puberty
	Mean age: ND	Duration of DM >3 y
	Age range: ND
	Mean height: ND
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 309
	Evaluated: 142
	Number of sites:8
Salardi 1987 87127000	Location: Italy	Children with Type I diabetes	None reported	Insulin dependent diabetes mellitus	Prospective longitudinal cohort (Mean [Range] 4.5 [1–10.7] y)
	Setting: Pediatric clinic
	Mean age: 7.4±3.6 y
	Age range: 0.7–15 y
	Mean height: ND
	Male: 44%
	Race: ND
	SES: ND
	Enrolled: 79
	Evaluated: 79
	Number of sites: 1
Soliman 1996 96412723	Location: Oman	Prepubertal children with diagnosis of Type I diabetes mellitus	Intra-uterine growth retardation	Insulin dependent diabetes mellitus	Prospective longitudinal cohort (1 y)
	Setting: Diabetes clinic		Systemic or endocrine disease
	Mean age: 9 y		Dysmorphic trait
	Age range: 2–12 y		Central nervous system irradiation
	Mean height: ND		Hypothyroid
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 47
	Evaluated: 45
	Number of sites: 1
Vanelli 1992 93052062	Location: Italy	Diabetic children referred to diabetes clinic at diagnosis	Endocrinological disease affecting linear growth	Insulin dependent diabetes mellitus	Prospective longitudinal cohort (ND)
	Setting: Diabetes unit	Treated with daily insulin
	Mean age: ND	Subset analysis of girls who became diabetic at onset of puberty
	Age range: 1.5–21 y
	Mean height: ND
	Male: 0% (analyzed)
	Race: ND
	SES: ND
	Enrolled: 204
	Evaluated: 42
	Number of sites: 1
Wise 1992 92387067	Location: US	Children and adolescents with diabetes	Other problems that might affect growth (thyroid disease, cerebral palsy, foster care placement, methylphenidate therapy, or precocious or delayed onset of puberty)	Insulin dependent diabetes mellitus	Prospective longitudinal cohort
	Setting: Pediatric diabetes clinic
	Mean age: ND
	Age range: ND
	Mean height: ND
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 122
	Evaluated: 122
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Insulin Dependent Diabetes Mellitus - Part II

Author, Year	Severity of Illness Groups	Outcomes
Arreola 1991 93191566	Diabetes control	Height velocity
	• HgbA1c	cm/month
	• Poor control: HgbA1c > 11%
	• Good control: HgbA1c < 11%
Court 1982 83021090	Diabetes control (text unclear)	Height SDS, based on Tanner 1976
	• Poor control: Glucosuria (>2%) frequency >75% or 24 hour urine glucose > 50 g	Growth velocity SDS
	Good control: Glucosuria frequency < 10% or 24 urine glucose < 20 g
Herber 1988 88180314	HgbA1c	Change in height SDS
Izumi 1995 96105591	HgbA1c category	Height SDS, based on Japanese Ministry of Education statistics
	• <10.0%
	• 10.0–12.4%
	• ≥12.0% (unclear why overlap)
Jivani 1973 73134278	Diabetes control	Height percentile, based on Tanner, 1966
	• Poor: Ketonuria, glycosuria up to 5% at clinic, BS > 250 mg/dL, severe or frequent hypoglycemic reactions
	• Good: No ketonuia, ≤ 1% glycosuria at clinic, blood sugar < 150 mg/dL, rare mild hypoglycemic reactions
Pitukcheewanont 1995 96136504	HgbA1c	Height SDS, based on National Center for Health Statistics
		Height velocity, SDS
Rosenbloom 1982 83032889	Limited joint mobility (“associat[ed with] early microvascular complications)	Height percentile, based on National Center for Health Statistics
	• None	< 25%
	• Mild: Involving 1 or 2 interphalangeal joints, 1 large joint, or only the metacarpophalangeal joints bilaterally
	• Moderate/Severe: Involving 3 or more interphalangeal joints or 1 finger joint and 1 large joint bilaterally or cervical spine involvement or obvious hand deformity at rest
Salardi 1987 87127000	HgbA1c	Height SDS, based on Tanner's centiles
		Height velocity SDS, based on Tanner's centiles
Soliman 1996 96412723	Diabetes control	Height velocity SDS, based on age-matched population
	• HgbA1c < 10%
	• HgbA1c > 10%
Vanelli 1992 93052062	Diabetes severity (apparently post hoc grouping)	Peak height velocity (during pubertal growth) cm/year
	• Lower insulin requirement: Mean ± SD insulin 37±1.5 U/m²/day; HgbA1c 10±1.8%
	• Higher insulin requirement: Insulin 42±0.5 U/m²/day; HgbA1c 9±2.4%
Wise 1992 92387067	HgbA1c level categories	Height velocity
	• 6.0–7.9	Change in SDS, based on National Center for Health Statistics
	• 8.0–9.9
	• 10.0–11.9
	• 12.0–13.9
	• 14.0–15.9
	• 16.0–21.9

Evidence Table 3. Association of Decrease Growth Velocity with Severity of Disease Insulin Dependent Diabetes Mellitus - Part III

Author, Year	Associations found	Potential Biases	Comments
Arreola 1991 93191566	Children with diabetes with HgbA1c > 11% (no data on number of subjects) had significantly lower height velocity (0.54 cm/month, from graph) compared to those with HgbA1c < 11% (0.22 cm/month). There was a significant correlation between HbA1c level and height velocity.	True number of subjects included not reported.	No data on funding source
Court 1982 83021090	For children with diabetes that was under poor control (n=57) mean height (-0.96 SDS) was significantly lower than those under good control (n=54, -0.22 SDS).	Determination of diabetes control based on glucosuria frequency and 24 hour urine glucose instead of HgbA1c	No data on funding source.
	Mean height velocity was significantly lower for children under poor control (-1.22 SDS) than those under good control (-0.61 SDS).
Herber 1988 88180314	In children with diabetes (n=67), “no significant effect on [change in height] was apparent with differing levels of HgbA1c.”	Specific data on height SDS in relation to HbA1c levels was not given.	No data on funding source.
		Overall good control (mean HbA1c levels at commencement 10.2%; mean HbA1c levels at end 9.9%.
Izumi 1995 96105591	Among children with diabetes, mean height of those with HgbA1c < 10% (n=28) was -0.54 SDS, of those with HgbA1c of 10–12.4% (n=53) was -0.76, and of those with HgbA1c ≥ 12% (n=26) was -0.75. Height was not significantly associated to HgbA1c level.	Categories of severity of diabetes overlapping.	No data on funding source.
	No data on high velocity.	Unclear if statistical analysis based on groups or on individual HgbA1c levels
Jivani 1973 73134278	The mean height percentile among girls with diabetes with poor control (n=22) was 38% and among those with good control (n=34) was 40%. There was no significant difference between the two groups.	Eligibility criteria unclear.	No data on funding source.
	The mean height percentile among boys with diabetes with poor control (n=19) was 35% and among those with good control (n=29) was 29%. There was no significant difference between the two groups.	Age of children at time of follow-up analysis unclear.	The text and table of the criteria for assessment of diabetic control were not concordant.
	Among both boys and girls with diabetes, height velocity was not associated with level of diabetes control	Statistical analysis unclear.
		Determination of diabetes control based on factors other than HgbA1c
Pitukcheewanont 1995 96136504	Among children with type I diabetes (n=82), there was no association between mean HgbA1c and mean height SDS in univariate analysis. In multivariate analysis, controlling for age of diabetes onset and initial height, there remained no association between HgbA1c and height.	Retrospective analysis	No data on funding source.
	There was no association between mean HgbA1c and mean height velocity at any Tanner stage in univariate analysis.
Rosenbloom 1982 83032889	Diabetic children with mild (n=31), moderate (n=19) or severe (n=24) limited joint mobility were significantly more likely to be below the 25^th percentile for height (~75%, from graph) than those without limited joint mobility (n=68, ~37%)). However, the correlation between HgbA1c and joint mobility was poor.	Joint mobility is a poor proxy for severity of diabetes.	Study hospital funded
	No data on height velocity	Only looked at quartiles of height.
Salardi 1987 87127000	In children with insulin dependent diabetes (n=79), height SDS and height velocity SDS were not significantly correlated with HgbA1c.	Only summary of results reported.	Study was government funded
Soliman 1996 96412723	Children with diabetes whose HbA1c < 10% (n=30) had significant higher growth velocity (+0.75 SDS) than DM subject HbA1c > 10% (n=15, -1.6 SDS).	None noted	No data on funding source.
Vanelli 1992 93052062	In a subset of girls who became diabetic at onset of puberty (n=42), the 23 treated with greater amounts of insulin (mean 41.5 U/m²/d) and who had lower mean HgbA1c (8.9%) had significantly higher mean peak growth velocity (8.5 cm/y) than the 19 treated with less insulin (37.3 U/m²/d and higher HgbA1c (10.1%), at 6.9 cm/y.	Study poorly defined.	No data on funding source
		Unclear a priori hypotheses
		Definitions of sample groups poorly defined
		Single analysis of subgroup only
Wise 1992 92387067	Among children with diabetes (n=122) a significant linear relationship was seen between HgbA1c and growth velocity. HgbA1c < 8% was associated with growth acceleration (+0.10 SDS); the most severe growth retardation occurred when HgbA1c levels were > 16% (-0.07 SDS).	None noted	No data on funding source

Evidence Table 3. Association of Decreased Growth Velocity with severity of Disease β-thalassemia - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Constantoulakis 1975 75129833	Location: Greece	Diagnosis of homozygous β- thalassemia	None reported	β thalassemia major	Retrospective longitudinal cohort
	Setting: Outpatients				(Range 4–7 y)
	Mean age: ND
	Age range: 7 months - 28 y
	Mean height: ND
	Male: 58%
	Race: ND
	SES: ND
	Enrolled: 229
	Evaluated: 171
	Number of sites: 1
Kattamis 1970 71003038	Location: Greece	Diagnosis of homozygous β-thalassemia	Clinical picture of thalassemia intermedia and hemoglobin between 7–10 g/dL	β thalassemia major	Prospective cross-sectional cohort
	Setting: Clinic		Age > 11 y
	Mean age: ND
	Age range: 1–11 y
	Mean height: ND
	Male: 53%
	Race: ND
	SES: ND
	Enrolled: 74
	Evaluated: 74
	Number of sites: 1
Madeddu 1978 78167848	Location: Italy	Homozygous β thalassemia	None reported	β thalassemia major	Prospective cross-sectional
	Setting: Clinic
	Mean age: ND
	Age range: 2–13 y
	Mean height: ND
	Male: 46%
	Race: ND
	SES: ND
	Enrolled: 50
	Evaluated: 50
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease β-thalassemia - Part II

Author, Year	Severity of Illness Groups	Outcomes
Constantoulakis 1975 75129833	Multiple linear regression including	Height percentiles, based on various Greek sources
	• Mean hemoglobin level over previous 2–4 years
	• Total blood transfused since birth
	• “Index of severity of disease” based on amount of blood transfused during previous 2 years
	• (Age)
Kattamis 1970 71003038	β-thalassaemia severity	Height percentile, based on Children's Medical Center, Boston growth chart
	• Group I: Pre-transfusion hemoglobin > 8 g/dL, frequent transfusions
	• Group II: Unable to follow regularly, pre-transfusion hemoglobin 6–8 g/dL, either very low SES or had to travel “very long distance” for transfusion
	• Group III: Transfused only when hemoglobin < 6 g/dL, usually < 5 g/dL
Madeddu 1978 78167848	Hemoglobin concentration between two transfusions	Height percentile, based on Tanner, 1962

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease ß-thalassemia-Part III

Author, Year	Associations found	Potential Biases	Comments
Constantoulakis 1975 75129833	Among children will ß-thalassemia major (n=171) there was a trend toward lower height percentiles being associated with low hemoglobin levels and severity of disease; however, neither was statistically significant in a linear regression controlling for age. Total blood transfused since birth was not associated with height.	No data on hemoglobin levels, total blood transfused since birth, and index of severity of disease. Provides only the regression results.	No data on funding source
	No data on height velocity.	Definition of the severity of the disease unclear.
		Large numbers of subjects were not included in analyses. No explanation.
		Model adjusted for non-independent variables.
Kattamis 1970 71003038	Children with ß-thalassemia major (n=38) had significantly higher mean height percentile (56%) than children with moderate (n=14, 22%) and severe (n=22, 7%) disease.	Thalassemia severity based on how low hemoglobin allowed to fall before transfusion	No data on funding source
	No data on height velocity.
Madeddu 1978 78167848	Children with ß thalassemia who were <3^rd percentile for height (n=18) had mean hemoglobin of 6.6 g/dL, which was not significantly lower than those whose height was 3^rd –25^th percentile (n=12, 7.2 g/dL) and whose height was 25^th–75^th percentile (n=19, 7.0 g/dL) (One child had height above 75^th percentile).	Cross-sectional study cannot evaluate previous degree of anemia	No data on funding source
	No data on height velocity.

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Inflammatory Bowel Disease - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases/Conditions	Study Design (Duration)
Farmer 1979 80023814	Location: US	Crohn's disease established by clinical, radiographic and/or histological means	Unable to obtain follow-up data	Crohn's disease	Retrospective longitudinal case series
	Setting: Clinic	Age < 20 y			(Mean 7.7 y)
	Mean age: ND (41% < 16 y)	Diagnosis made form 1955-1974
	Age range: < 20 y
	Mean height: ND
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 522
	Evaluated: 522
	Number of sites:1
Griffiths 1993 93345883	Location: Canada	Diagnosis of Crohn's disease	Transferred after initial diagnosis and referred for consultation only	Crohn's disease	Retrospective longitudinal case series
	Setting: Clinic	Tanner stage 1 or 2			(≥ 2 y)
	Mean age: 11±2.3 y
	Age range: 5–17 y
	Mean height, SDS: -1.1±1.3
	Male: 66%
	Race: ND
	SES: ND
	Enrolled: 100
	Evaluated: 100
	Number of sites: 1
Saha 1998 98182041	Location: Finland	Diagnosis of inflammatory bowel disease	None reported	Inflammatory bowel disease: Ulcerative colitis Crohn's disease	Retrospective longitudinal cohort
	Setting: Pediatrics department				(13 y)
	Mean age: 7 y
	Age range: ND
	Mean height: ND
	Male: 43%
	Race: ND
	SES: ND
	Enrolled: 47
	Evaluated: 47
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Inflammatory Bowel Disease - Part II

Author, Year	Severity of Illness Groups	Outcomes
Farmer 1979 80023814	Location of Crohn's disease	Short stature
	•Ileocolic pattern	Height < 3^rd percentile, based on growth charts
	•Colon pattern
	•Small intestine pattern
	•Anorectal pattern
Griffiths 1993 93345883	Severity of gastrointestinal symptoms over given year	Height velocity cm/year
	•Quiescent: No symptoms, normal complete blood count and sedimentation rate	Percent with growth retardation < 4 cm/year
	•Mild: Some symptoms, no prednisone treatment needed	Height SDS, based on National Center for Health Statistics
	•Moderate: Intermittent exacerbations requiring prednisone or nutritional support	Change in height SDS
	•Severe: Chronic unremitting symptoms requiring prednisone or alternate treatment
	Location of Crohn's disease
	•Small bowel
	•Ileocecal
	•Ileocolic
	•Colon
Saha 1998 98182041	Severity of Crohn's disease/ulcerative colitis	Height SDS, based on Finnish population mean
	12 point scale based on number of relapses, exacerbations, medications and need for surgery (correct scoring mechanism not clear)	Height velocity SDS
	•Mild: 1–3 points
	•Moderate: 4–6 points
	•Severe: ≥ 7 points

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Inflammatory Bowel Disease - Part III

Author, Year	Associations found	Potential Biases	Comments
Farmer 1979 80023814	Among children with Crohn's disease likelihood of short (< 3^rd height percentile) was not related to location of disease. Of those with ileocolic disease (n=175) 7% were short; of those with colonic disease (n=162) 8% were short; of those with small intestine disease (n=151) 6% were short; of those with anorectal disease (n=34) 9% were short.	No statistical analyses.	No data on funding source
	No data on height velocity	“Severity” based on location of disease
Griffiths 1993 93345883	Among children with Crohn's disease, “severity of gastrointestinal symptoms was the major factor influencing linear growth during the first two years [after diagnosis]. Height velocity significantly decreased with increasing gastrointestinal symptoms during both year and year two, after adjusting for corticosteroid use. During year one, height velocity of children with quiescent disease (n=14) was ~5.7 cm/year (from graph), with mild disease (n=54) was ~4.6 cm/year, with moderate disease (n=23) was ~3 cm/year, and with severe disease (n=9) was ~2.7 cm/year. During year two, height velocity of children with quiescent disease (n=23) was ~7.3 cm/year, with mild disease (n=44) was ~4.5 cm/year, with moderate disease (n=24) was ~4 cm/year, and with severe disease (n=8) was ~4.1 cm/year.	Retrospective	No data on funding source
	Among children with Crohn's disease who were followed to maturity, initial and ultimate height were not associated with disease severity. Initial and ultimate height severity were -0.88 and -0.43 SDS for those with mild disease (n=26), -1.17 and -1.15 SDS for those with moderate disease (n=15), -1.48 and -0.70 SDS for those with moderate or severe disease with sustained remission (n=14), and -1.12 and -1.37 SDS for those with chronically severe disease (n=12).
	Among children with Crohn's disease who were followed to maturity, change in height was significantly associated with disease severity. Change in height was 0.58 SDS for those with mild disease (n=26), 0.03 SDS for those with moderate disease (n=15), 0.68 SDS for those with moderate or severe disease with sustained remission (n=14), and -0.11 SDS for those with chronic severe disease (n=12).
	Among children with Crohn's disease, at year two (n=100), disease location was not significantly associated with height velocity (cm/year) or percentage with height velocity < 4 cm/year. Among those with long-term follow-up (n=65), disease location was not associated with initial height SDS, ultimate height SDS, or change in height SDS.
Saha 1998 98182041	Among children with ulcerative colitis (n=29), mean overall change in height over 4 years was -0.16 SDS for those with mild disease, -1.09 SDS for those with moderate disease, and -1.74 SDS for those with severe disease. This trend did not reach statistical significance. The overall change in height velocity was significantly greater in those with mild disease (+1.14 SDS) than those with moderate disease (+0.39 SDS) and those with severe disease (+0.12 SDS).	Height outcomes are difficult to interpret.	Study was hospital and privately funded.
	Among children with Crohn's disease (n=18), mean overall change in height over 4 years was +0.51 SDS for those with mild disease, -0.27 SDS for those with moderate disease, and -0.11 SDS for those with severe disease. This trend did not reach statistical significance. The overall change in height velocity was significantly greater in those with mild disease (+0.42 SDS) than those with moderate disease (+0.09 SDS) and those with severe disease (-1.09 SDS).	Numbers of subjects in each category not reported.

Evidence Table 3. Association of Decreased Growth Velocity with Severity of disease Juvenile Rheumatoid Arthritis - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Bernstein 1977 87026009	Location: US	Children with diagnosis of juvenile rheumatoid arthritis	None reported	Juvenile rheumatoid arthritis	Retrospective longitudinal cohort (mean [range] 6.6[2.3–11] y, up to age of 15)
	Setting: Clinic
	Mean age: ND
	Initial age range: 1–12 y
	Mean height: ND
	Male: 24%
	Race: ND
	SES: ND
	Enrolled: 31
	Evaluated: 31
	Number of sites: 1
Polito 1997 97283189	Location: Italy	Children with active juvenile rheumatoid arthritis for ≥ 1 y	Steroid use	Juvenile rheumatoid arthritis:	Retrospective longitudinal case series
	Setting: Pediatric clinic	Follow-up for ≥ 1 y	Growth hormone therapy	Pauciarticular onset	(≥ 1 y)
	Initial mean age: 7 y			Systemic onset
	Follow-up mean age: 12 y			Polyarticular onset
	Age range: ND
	Initial mean height, SDS: +0.34
	Follow-up mean height, SDS: -0.05
	Male: 26%
	Race: ND
	SES: ND
	Enrolled: 58
	Evaluated: 58
	Number of sites:1
Saha 1999 99374796	Location: Finland	Children with mild to moderate juvenile onset chronic arthritis	Ullrich-Turner syndrome	Juvenile rheumatoid arthritis:	Retrospective longitudinal case series
	Setting: Hospital based clinic	Prepuberty		Pauciarticular onset	(1–11 y, until age 12 y)
	Mean age: 4 y	Followed for ≥ 1 y		Systemic onset
	Age range: 0.9–11 y			Polyarticular onset
	Mean height: ND
	Male: 36%
	Race: ND
	SES: ND
	Enrolled: 64
	Evaluated: 64
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Juvenile Rheumatoid Arthritis - Part II

Author, Year	Severity of Illness Groups	Outcomes
Bernstein 1977 87026009	Mode of onset of juvenile rheumatoid arthritis	Mean percent change in height per year
	• Systemic
	• Polyarticular
	• Pauciarticular
Polito 1997 97283189	Type of juvenile rheumatoid arthritis	Change in height SDS, based on Tanner 1976
	• Systemic	Change in height SDS per year
	• Polyarticular
	• Pauciarticular
	Number of joints affected (systemic or polyarticular only)
	Sum of periods of time with disease flares (systemic or polyarticular only) joint swelling, tenderness, functional impairment, ESR > 25 mm/hr)
	Functional class, not defined (systemic or polyarticular only)
Saha 1999 99374796	Severity of juvenile chronic arthritis score	Height velocity SDS, based on Tanner 1966
	6 point scale based on local anti-inflammatory administrations and number of systemic antirheumatics	Height SDS
	• Mild: Score 1–3
	• Moderate: Score ≥ 4
	Type of disease
	• Pauciarticular
	• Polyarticular
	• Systemic
	• Iridocyclitis

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Juvenile Rheumatoid Arthritis - Part III

Author, Year	Associations found	Potential Biases	Comments
Bernstein 1977 87026009	Among children with juvenile rheumatoid arthritis, those with systemic disease (n=13) had significantly greater mean change in height per year (-0.85%) than those with polyarticular disease (n=9, +0.04%) and those with pauciarticular disease (n=9, -0.01%).	Definition of outcome measure (percent change in height per year) not well defined	Study was privately funded.
Polito 1997 97283189	Among children with juvenile rheumatoid arthritis, those with pauciarticular disease (n=21) were less likely to lose > 1 height SDS between first and last visits (0%) than those with systemic disease (n=19, 26%) or those with polyarticular disease (n=18, 28%). No statistical analysis was performed.	Retrospective analysis	No data on funding source
	Among children with either systemic or polyarticular disease (n=37), change in height SDS per year was not associated with number of affected joints but was associated with the sum of the periods of disease flares. Children in functional class I (not defined) had significantly smaller change in height SDS per year (-0.01 SDS) than those in functional class II (-0.06 SDS)	Inclusion criteria was no steroids, thus children with most severe disease, were excluded.
		Functional class not defined.
Saha 1999 99374796	Among children with juvenile rheumatoid arthritis (n=64), by multivariate analysis of covariance “the most obvious decrease in [height] velocity was seen in children with polyarthritis and growth velocity was affected slightly more in children with moderate than mild disease,” adjusting for sex and duration of disease.	Exclusion criteria unclear.	Study hospital and privately funded
	“Heights were more affected in children with polyarthritis and the severity of the disease had an additive effect,” adjusting for glucocorticoid use and duration of disease.	No children with severe arthritis.
		Limited reporting of data.
		Statistical significance of relevant findings not reported.

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Chronic Kidney Disease / Kidney Failure - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Claris-Appiani 1989 91002131	Location: Italy	Children with chronic kidney disease	Acute obstruction	Chronic kidney disease	Prospective longitudinal cohort
	Setting: ND	Pre-dialysis	Pyelonephritis	Stage 3
	Mean age: 1.6–9 y	Stable	Hypertension	Stage 4
	Age range: ND		Hyposthenuria
	Mean height: ND		Salt wasting
	Male: 65%		Steroid treatment
	Race: ND
	SES: ND
	Enrolled: 17
	Evaluated: 17
	Number of sites: ND
Ismaili 2001 21219717	Location: Belgium	Infants with kidney dysplasia and chronic kidney disease	Obstructive renal dysplasia	Chronic kidney disease	Retrospective longitudinal cohort
	Setting: Pediatric nephrology			Kidney dysplasia	(4 y)
	Initial mean age: 6 mo			Stage 4
	Follow-up mean age: 4 y			Stage 5
	Age range: ND
	Mean height: ND
	Male: 55%
	Race: ND
	SES: ND
	Enrolled: 11
	Evaluated: 11
	Number of sites: 1
Karlberg 1996 96384503	Location: Europe	Onset of chronic kidney disease age < 6 mo	Primary glomerulonephropathy	Chronic kidney disease	Prospective longitudinal case series
	Setting: Pediatric nephrology clinics	GFR < 3 SD below mean for age	Systemic metabolic disorder	Stage 4	(Mean [Range] 7.3 [3–16] y)
	Mean age: ND	≥ 2 height measurements before age 1 y and followed until at least age 3 y		Stage 5
	Age range: 0–5 y
	Mean heights (various ages), SDS: -1.05 to -3.09
	Male: 70%
	Race: ND
	SES: ND
	Enrolled: 73
	Evaluated: 61
	Number of sites: 22
Konrad 1995 96162281	Location: Europe	Autosomal recessive polycystic kidney disease	Single growth measurement only	Autosomal recessive polycystic kidney disease	Prospective longitudinal case series
	Setting: Pediatric nephrology clinics	Prepuberty		Stage 1	(≥ 1 y)
	Initial mean age: 1.7 y	Girls ≤ 9 y old		Stage 2
	Initial age range: 1–8 y	Boys ≤ 10 y old		Stage 3
	Follow-up man age: 5.6 y	Age ≥ 1 y at first assessment		Stage 4
	Follow-up age range: 2–9 y	≥ 1 y follow-up
	Mean height: ND	Subset of Karlberg, 1996 [#226] and Schaefer, 1996 [#3001]
	Male: 59%
	Race: ND
	SES: ND
	Enrolled: 67
	Evaluated: 58
	Number of sites: 22
Schaefer 1996 96384504	Location: Europe	Born between 1970 and 1992	Primary glomerulonephropathy	Chronic kidney disease	Prospective longitudinal case series
	Setting: Pediatric nephrology clinics	Marked reduction in functional kidney mass	Systemic metabolic disorder	Stage 4	(Median [Range] 3.3 [0.5–10] y)
	Initial median age: 2.6 y			Stage 5
	Initial age range: 0–8 y
	Follow-up age: ~ 10 y
	Height range, SDS: -4.2 to +4.1
	Male: ND
	Race: White 96%, Arab or Asian 4%
	SES: ND
	Enrolled: 500
	Evaluated:321
	Number of sites: 22
Norman 2000 21024968	Location: UK	Children scheduled for GFR measurements	Congenital growth abnormalities	Chronic kidney disease	Prospective cross-sectional cohort
	Setting: Clinic		Receiving dialysis	Stage 1
	Mean age: ND			Stage 2
	Age range: 2–17 y			Stage 3
	Mean height: ND			Stage 4
	Male: 63%
	Race: ND
	SES: ND
	Enrolled: 95
	Evaluated: 95
	Number of sites: 1
Rizzoni 1984 85010647	Location: Italy	Children with GFR < 70 ml/min/1.73 m²	Proteinuria	Chronic kidney disease	Retrospective longitudinal case series
	Setting: Pediatric clinic	≥ 1 bone age determined	Nephrotic syndrome	Stage 2	(Mean [Range] 4.3 [1–12] y)
	Mean age: 6 y	Follow-up for ≥ 1 y	Steroid treatment	Stage 3
	Age range: < 6 mo – 15 y			Stage 4
	Initial mean height, SDS: -1.6			Stage 5
	Follow-up mean height, SDS: -1.5
	Male: 69%
	Race: ND
	SES: ND
	Enrolled: 47
	Evaluated: 47
	Number of sites: 1
Schärer 1999 20068988	Location: US	Prepubertal children with idiopathic nephrotic syndrome	Chronic kidney failure	Idiopathic nephrotic syndrome	Prospective longitudinal cohort
	Setting: Clinic				(Mean [Range] 9.0 [2–19] y)
	Mean age: 4.5 y
	Age range: 1.2–10 y
	Mean height: ND
	Male: 67%
	Race: ND
	SES: ND
	Enrolled: 45
	Evaluated: 38
	Number of sites: ND
Tejani 1983 84068349	Location: US	Nephrotic syndrome	Diffuse increase in mesangial cells or with mesangial deposition of immunoglobulin	Focal segmental sclerosis	Retrospective longitudinal case series
	Setting: Pediatric nephrology clinic	Biopsy-proven focal segmental sclerosis			(Steroid resistant: 7±5 y; Steroid sensitive: 13±8 y))
	Mean age at onset:	Biopsy performed because of lack of response to steroids or relapse after treatment with steroids
	Steroid resistant 7.7±3.7 y; Steroid sensitive, 3.5±2.5 y
	Age range at onset: 10 mo – 15 y
	Mean height: ND
	Male: 42%
	Race: White 16%, Black 50%, Hispanic 33%
	SES: ND
	Enrolled: 25
	Evaluated: 24
	Number of sites: 1
Tsau 1989 90155156	Location: Taiwan	Nephrotic syndrome beginning before age 15 y	Development of chronic kidney insufficiency	Nephrotic syndrome	Retrospective longitudinal cohort
	Setting: Hospital	Followed up for at least 2 y			(5.5±2.8 y)
	Mean age of onset: 5 y
	Age range: ND
	Mean height: ND
	Male: 83%
	Race: Chinese
	SES: ND
	Enrolled: 52
	Evaluated: 52
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Chronic Kidney Disease / Kidney Failure - Part II

Author, Year	Severity of Illness Groups	Outcomes
Claris-Appiani 1989 91002131	GFR, Schwartz formula estimate	Height SDS, based on Tanner 1976
		Growth velocity per year SDS
Ismaili 2001 21219717	Chronic kidney disease severity, Schwartz formula estimate and direct measurement	Height SDS, based on Tanner 1966
	• GFR < 15 mL/min/1.73 m²
	• GFR > 15 mL/min/1.73 m²
Karlberg 1996 96384503	GFR, age-specific Schwartz formulae estimates	Height SDS, based on normal Swedish children
Konrad 1995 96162281	GFR, Schwartz formula estimate	Height SDS, based on First Zurich Longitudinal Growth Study
	Chronic kidney disease severity
	• GFR < 60 mL/min/1.73 m²
	• GFR > 260 mL/min/1.73 m²
Schaefer 1996 96384504	GFR, age-specific Schwartz formulae estimates	Height SDS
	Chronic kidney disease severity	Height velocity SDS
	• GFR < 25 mL/min/1.73 m²
	• GFR > 25 mL/min/1.73 m²
Norman 2000 21024968	Chronic kidney disease severity, based on GFR direct measurements	Height SDS, based on normal UK population
	• Normal: GFR > 75 mL/min/1.73 m²
	• Mild: GFR 50–75 mL/min/1.73 m²
	• Moderate: GFR 25–50 mL/min/1.73 m²
	• Severe: GFR < 25 mL/min/1.73 m²
Rizzoni 1984 85010647	GFR, estimate or direct measurement	Height SDS, based on Tanner 1976
		Height velocity, cm/year
Schärer 1999 20068988	Final kidney status	Height SDS
	• Reached final height in absence of kidney function deterioration or ESRD	Change in height SDS
	• Reached final height with serum creatinine > 1.2 mg/dL after age 13.6 years
	• (Lost to follow-up before kidney function deterioration -- not evaluated in this report)
	• Developed ESRD before reaching adult height
	Idiopathic nephrotic syndrome severity (mean serum albumin level and serum protein level)
Tejani 1983 84068349	Responsiveness (improvement of proteinuria) of focal segmental sclerosis to steroids	Growth retardation
	• Steroid responsive: Complete remission within 8 wk of prednisone, persisting for 2 mo after cessation of treatment	Height for age < 5^th percentile at follow-up
	• Steroid resistant: No remission after 8 wk of treatment
Tsau 1989 90155156	Nephrotic syndrome severity	Change in Height SDS, based on Department of Health, Taiwan
	• Less favorable clinical courses, including frequent relapsing, steroid-dependent, and steroid-resistant nephrotic syndromes	Growth velocity index
	• Had responded to steroid therapy initially with occasional or no relapse	Ratio of actual growth velocity to normal growth velocity for age (% of expected growth)

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Chronic Kidney Disease/Kidney Failure - Part III

Author, Year	Associations found	Potential Biases	Comments
Claris-Appiani 1989 91002131	Among children with pre-dialysis chronic kidney disease (n=17) there was a significant positive correlation between height SDS and creatinine clearance.	Univariate analysis only was performed for height.	No data on funding source
	In multivariate analysis, height velocity was not associated with creatinine clearance, controlling for bone age, BUN, change in BUN, PTH, calcitriol, caloric and protein intake.	No univariate analysis for height velocity and creatinine clearance. Creatinine clearance not independent of other variables.
		Far too few subjects for multivariate analysis
Ismaili 2001 21219717	Among infants with chronic kidney disease with GFR < 15 mL/min/1.73 m² at age 6 months (n=5) mean height did not improve between age 6 months (-2.6 SDS) and 4 years (-2.3 SDS). Among those with GFR > 15 mL/min/1.73 m² (n=6) mean height increased from age 6 months (-2.6 SDS) to 4 years (-1.4 SDS).	Small sample size	No data on funding source.
Karlberg 1996 96384503	Among children with chronic kidney disease (n=47) change in height SDS between 9 months and 2 years was not associated with GFR in a multiple linear regression.	No data on variables in model.	Study was pharmaceutical funded.
	Among infants with chronic kidney disease (n=14) change in height SDS during the “first postnatal months of life” were not associated with GFR.	Many children not included in all analyses	Same dataset as Schaefer 1996
Konrad 1995 96162281	Among girls with autosomal recessive polycystic kidney disease, those with GFR < 60 mL/min/1.73 m² (n=9) had mean height of -2.1 SDS; those with higher GFR (n=15) had mean height of -0.3 SDS. Boys with GFR < 60 mL/min/1.73 m² (n=10) had mean height of -1.5 SDS; those with higher GFR (n=24) had mean height of -0.5 SDS. There was a significant correlation between GFR and height SDS among girls but not among boys.	None noted	No data on funding source.
	No data on height velocity.		Subset of Karlberg, 1996 [#226] and Schaefer, 1996 [#3001]
Schaefer 1996 96384504	Among children with chronic kidney disease (n=321) height velocity was positively correlated with GFR among 6 year olds and among children of 4 other age groups (of 9 age groups total).	Numbers of children analyzed with GFR > 25 mL/min/1.73 m² and with lower GFR not reported Results for specific age groups not reported	Study was pharmaceutical funded.
	Children with GFR > 25 mL/min/1.73 m² had consistently higher annual height velocities than children with lower GFR.	Results for specific age groups not reported	Same dataset as Karlberg 1996
	Children with GFR > 25 mL/min/1.73 m² were taller (mean height -1.65 SDS) than those with lower GFR (-2.79 SDS).
Norman 2000 21024968	Among children with chronic kidney disease, those with GFR > 75 mL/min/1.73 m² (n=35) had mean height of +0.40 SDS, those with GFR 50–75 mL/min/1.73 m² (n=23) had mean height of -0.34 SDS, those with GFR 25–50 mL/min/1.73 m² (n=19) had mean height of -0.58 SDS, those with GFR < 25 mL/min/1.73 m² (n=18) had mean height of -1.52. Those with moderate and severe chronic kidney disease had significantly lower height than those with normal GFR.	None noted	No data on funding source.
	No data on height velocity
Rizzoni 1984 85010647	Among children with chronic kidney disease (n=47) there was no association between GFR and growth velocity. or height SDS.	Limited data reported	No data on funding source
Schärer 1999 20068988	Children with nephrotic syndrome who reached final height without deterioration of kidney function (n=16) had a mean change in height SDS from onset of kidney disease to final height of +0.27 SDS; those who developed serum creatinine > 1.2 mg/dL before reaching final height (n=8) had a mean change in height SDS of -0.08 SDS; those who developed ESRD before reaching final height (n=9) had a mean change in height of -1.35 SDS. Children who developed ESRD had a significant decrease in their height SDS compared to other groups.	Analyses performed on only subgroups of subjects	No data on funding source.
	Among children with steroid resistant nephrotic syndrome in whom steroid therapy was stopped before puberty (n=16) change in prepubertal height SDS and height velocity SDS correlated positively and significantly with total protein and albumin levels.	Categorization of children by severity of disease is retrospective. Categorized only after reaching final height.
		Serum albumin and protein are poor markers of disability-related severity of disease
Tejani 1983 84068349	In children with focal segmental sclerosis, those with steroid responsive disease (n=10) were significantly less likely to have short stature (< 5^th percentile) than those with steroid resistant disease (n=14) despite greater steroid doses among the former.	Detailed results not reported.	No data on funding source
	No data on growth velocity
Tsau 1989 90155156	Among children with nephrotic syndrome, those with relapsing disease (n=29) had significantly lower height velocity (80% of normal) and change in height SDS (-0.63 SDS) than those without relapsing disease (n=23, 104%, +0.19 SDS). This association was true for up to 7 years of follow-up	Cannot distinguish the effect of steroid use from severity of disease. Those with relapsing disease required significantly longer duration of steroid use (8 months) than those without relapsing disease (4 months).	No data on funding source

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Human Immunodeficiency Virus - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Brettler 1990 91011887	Location: US	Hemophilia	None reported	HIV (with hemophilia)	Prospective and retrospective longitudinal cohort
	Setting: Hemophilia clinic	HIV infection			(ND)
	Median age: 8 y	2 y of growth data before HIV diagnosis
	Age range: 2.5–15 y	Age ≤ 12.5 y in 1980
	Mean height: ND
	Male: 100%
	Race: ND
	SES: ND
	Enrolled: 36
	Evaluated: 36
	Number of sites: 1
Matarazzo 1994 95143841	Location: Italy	Perinatal HIV infection	None reported	HIV	Prospective longitudinal cohort
	Setting: Pediatric clinic				(24 mo)
	Median age: 2.6 y
	Age range: 0–8 y
	Mean height: ND
	Male: 49%
	Race: ND
	SES: ND
	Enrolled: 24
	Evaluated: 24
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Human Immunodeficiency Virus - Part II

Author, Year	Severity of Illness Groups	Outcomes
Brettler 1990 91011887	Progression to AIDS or ARC (not defined)	Growth failure
		Decrease in > 15 percentile points in height or weight for age for 2 consecutive years
Matarazzo 1994 95143841	Clinical condition	Height SDS, based on Tanner 1966
	• Stable: not defined	Height velocity SDS
	• Clinical deterioration: not defined

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Human Immunodeficiency Virus - Part III

Author, Year	Associations found	Potential Biases	Comments
Brettler 1990 91011887	Among boys with hemophilia and HIV infection “progression to AIDS or [AIDS related complex]” was more common in those with “growth failure” over 2 years (n=9, 55%) than those with normal growth (n=27, 7%). In multivariate analysis, “growth failure was the strongest prognostic variable” for progression to AIDS or AIDS related complex, adjusted for CD4, p24 antigenemia, and possibly other variables.	Methods, predictors and outcomes poorly defined	No data on funding source
		Variables included in multivariate analysis not reported
		Definition of growth failure includes weight
Matarazzo 1994 95143841	Among children with perinatal HIV infection (n=24), those with stable disease courses (n=14) had mean height at 1 year follow-up of -0.7 SDS, mean height at 2 year follow-up of -0.7 SDS, mean height velocity at 1 year follow-up of +1.2 SDS, and mean height at 2 year follow-up of +0.4 SDS. Children with unstable disease courses (n=10) had lower heights and height velocities. Mean height at 1-year follow-up was -2.0 SDS, mean height at 2 year follow-up of -2.6 SDS, mean height velocity at 1 year follow-up of -2.6 SDS, and mean height at 2 year follow-up of -3.0 SDS.	Direct comparison of children with stable and unstable disease had to be calculated from reported raw data.	Study was government and privately funded
	“At follow-up [after 2 years], most children with normal growth maintained a stable clinical condition, whereas those with growth reduction showed progression in the severity of the disease…Growth failure was a simple, sensitive predictor of clinical course [stable condition vs clinical deterioration].”	Clinical condition not defined.
		No statistical analysis.

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Atopic Dermatitis / Eczema - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Massarano 1993 93312068	Location: UK	Diagnosis of eczema	Chronic non-atopic disease which would limit growth	Atopic eczema	Prospective cross-sectional cohort
	Setting: Pediatric clinic	Prepubertal
	Mean age: 6 y
	Age range: 2.3–12 y
	Mean height: ND
	Male: 68%
	Race: ND
	SES: ND
	Enrolled: 68
	Evaluated: 68
	Number of sites: 1
Patel 1988 99014336	Location: UK	Atopic dermatitis that began before age 1 y	Systemic steroids	Atopic dermatitis	Prospective longitudinal cohort
	Setting: Pediatric clinic	Prepubertal, age 2–11 y			(2 y)
	Mean age: 5.5±0.3 (SEM)	Referred by pediatrician or dermatologist because of severity of disease
	Age range: 2–10 y
	Mean height, initial, SDS: +0.1±0.1 (SEM)
	Mean height, follow-up, SDS: -0.1±0.2 (SEM)
	Male: 55%
	Race: ND
	SES: ND
	Enrolled: 80
	Evaluated: 80
	Number of sites:1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Atopic Dermatitis / Eczema - Part II

Author, Year	Severity of Illness Groups	Outcomes
Massarano 1993 93312068	Maximum percentage of skin surface area affected by atopic eczema (erythema, vesicle, and crusts)	Height SDS, based on Tanner 1966
	Subgroups by percentage of skin surface area affected
	• < 50%
	• > 50%
Patel 1988 99014336	Percentage body surface area affected by atopic dermatitis	Height SDS, based on Tanner 1966
	• < 50%	Height velocity SDS
	• > 50%
	Coexistence of disease
	• With asthma
	• Without asthma

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Atopic Dermatitis / Eczema - Part III

Author, Year	Associations found	Potential Biases	Comments
Massarano 1993 93312068	Among children with atopic eczema (n=68), in a multivariate regression analysis, height SDS was correlated with surface area of eczema, controlling for parental height, diet, duration of eczema, treatment and asthma.	None noted	No data on funding source
	Those children with < 50% of skin affected (n=41) had mean height of +0.11 SDS; those with 50% of skin affected (n=27) had significantly shorter mean height of -0.83 SDS.
	No height velocity data.
Patel 1988 99014336	Among children with atopic dermatitis height SDS and height velocity SDS “did not differ between patients with < 50% body surface area (n=43) involvement and those with ≥ 50% body surface area involvement (n=37)…and between patients without asthma (n=36) and those with asthma (n=44).”	Complete results not reported.	No data on funding
		No statistical analyses reported.

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Miscellaneous Diseases - Part I

Author, Year	Demographics	Inclusion Criteria	Exclusion Criteria	Types of Diseases / Conditions	Study Design (Duration)
Evliyaoglu 1996 97097381	Location: Turkey	Homozygote sickle cell anemia	None reported	Homozygous sickle cell anemia	Prospective cross-sectional cohort
	Setting: Pediatric hematology clinic
	Mean age: 10 y
	Age range: 5.5–15 y
	Mean height, percentile: 20±22%
	Height range, percentile: 3–75%
	Male: ND
	Race: ND
	SES: ND
	Enrolled: 24
	Evaluated: 24
	Number of sites: 1
Rasat 1995 95372015	Location: New Zealand	Congenital adrenal hyperplasia diagnosed and treated at one hospital between 1962 and 1993	None reported	Congenital adrenal hyperplasia	Retrospective longitudinal case series
	Setting: ND				(Median [Range] 14 [2–31] y)
	Median age: 15 y
	Age range: 2–31 y
	Mean height: ND
	Male: 50%
	Race: ND
	SES: ND
	Enrolled: 16
	Evaluated: 9
	Number of sites: 1
Samson-Fang 1998 99066563	Location: US	Children with cerebral palsy	Genetic disorders	Cerebral Palsy	Retrospective cross-sectional cohort
	Setting: Rehabilitation center	Age < 10 years old at the mid-point of assessing growth interval	Complicating medical issues
	Mean age: 4.5 y	Minimum time interval of measurements of 0.8 years
	Age range: 1–12 y
	Mean height, SDS: -1.2±1.5
	Male: 53%
	Race: White 80%, Black 20%
	SES: ND
	Enrolled: 175
	Evaluated: 81
	Number of sites: 1

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Asthma Miscellaneous Diseases - Part II

Author, Year	Severity of Illness Groups	Outcomes
Evliyaoglu 1996 97097381	Severity of sickle cell anemia	Height percentile (not defined)
	• Mild: No blood transfusion or crises
	• Severe: Blood transfusion needed early in life, frequent crises
Rasat 1995 95372015	Number of episodes of biochemical escape from congenital adrenal hyperplasia suppression (morning 17-OH Progesterone > 60 nmol/L)	Height SDS. based on Tanner 1966
	• < 3 episodes over at least 2 years
	• > 3 episodes
Samson-Fang 1998 99066563	Cerebral palsy cognitive impairment	Growth velocity SDS, based on Rikken 1992
	• Cognitive impairment: Learning disability, or delay in problem solving or receptive language skills
	• No cognitive impairment
	Type of cerebral palsy
	• Spastic
	• Extrapyramidal
	• Mixed
	Ambulatory status
	•Ambulatory: Functionally ambulated with or without aids at home
	•Non-ambulatory

Evidence Table 3. Association of Decreased Growth Velocity with Severity of Disease Abbreviations

Author, Year	Associations found	Potential Biases	Comments
Evliyaoglu 1996 97097381	Among children with sickle cell anemia, those with mild disease (n=12) had significantly greater mean height percentile (22.5%) than those with severe disease (n=12, 17.3%).	Only mild and severe sickle cell.	No data on funding source
	No data on height velocity	Height measurement (percentile) not clear.
		Unclear why all children had low height percentile
Rasat 1995 95372015	Among children with congenital adrenal hyperplasia there was no significant difference in final height SDS in those with < 3 biochemical escape episodes (n=4) and those with > 3 episodes (n=5).	Very small numbers	No data on funding source
	No data on height velocity.	Results shown only graphically.
Samson-Fang 1998 99066563	Among children with cerebral palsy those with cognitive impairment (n=62) had significantly lower mean growth velocity (-1.25 SDS) than those with no cognitive impairment (n=19, -1.25 SDS))	Exclusion criteria limit most severe impairments due to cerebral palsy	Study was privately funded.
	Those age ≥ 2 years who were non-ambulatory (n=35) had significantly lower mean growth velocity (-1.20 SDS) than those who were ambulatory (n=37, -0.35 SDS).
	There was no significant difference in mean height velocity among those with spastic disease (n=22, mean height velocity = -0.69 SDS), those with extrapyramidal disease (n=36, -0.86 SDS), and those with mixed disease (n=22, -1.50 SDS)