Objectives. To describe the psychoeducational characteristics of children with congenital hypothyroidism (CH) identified through newborn screening and to study changes over time.
Method. Examined were 83 children with early-treated CH, who were long-time participants in a prospective study of outcome after newborn screening, and 120 control children who were classmates (n = 80) or siblings (n = 42). Children were tested during the third (53 children with CH and 46 control children) or the sixth (51 children with CH and 76 control children) grades at school with 21 children with CH being seen in both grades. Test instruments included multiple measures of achievement and cognitive abilities as well as behavior rating scales completed by parents and teachers.
Results. CH was associated with a slightly increased risk of learning disabilities in grade 3 but not grade 6. Third grade CH children scored lower than control children on tests of reading comprehension and arithmetic but did not differ on word recognition, writing, or spelling. Sixth grade CH children performed similar to controls on basic achievement tests but were reported to be doing poorer in several subject areas. For children with CH in grade 3, delayed skeletal maturity at diagnosis was associated with poorer word recognition ability and a longer period for normalizing thyroid hormone in infancy was correlated with weaker skill in learning sound–symbol correspondences.
Conclusion. Early-treated CH is associated with mild delays in several basic achievement areas (reading comprehension and arithmetic) at the third grade level, with catch up by the sixth grade. However, as other findings indicate cognitive problems do persist into adolescence in memory, attention, and visuospatial processing areas, the implications of these deficits for other educational accomplishments needs additional follow-up. congenital hypothyroidism, thyroid hormone, newborn screening, achievement, behavior, attention.
- CH =
- congenital hypothyroidism •
- SD =
- standard deviation •
- HSC =
- Hospital for Sick Children •
- L-T4 =
- levothyroxine •
- GORT =
- Gilmore Oral Reading Test •
- WJRMT =
- Woodcock Johnson Reading Mastery Task •
- WRAT-R =
- Wide Range Achievement Test-Revised •
- TOWL =
- Test of Written Language •
- GFW =
- Goldman Fristoe Woodcock
Early identification and treatment of congenital hypothyroidism (CH) with newborn screening is associated with the virtual elimination of mental retardation and significantly improved intellectual functioning.1 ,2 However, IQs are significantly and persistently reduced,3–5 while subtle impairments may occur in language,6visuospatial,7 ,8 neuromotor,9 ,10memory,11 attention,12 ,13hearing,14 and auditory discrimination abilities.7 The degree and type of impairment reflect cause,7 age at onset,15severity,5 ,16 and duration of neonatal hypothyroidism,7 as well as adequacy of treatment dosage8 ,11 ,17 ,18 and compliance.19
Several groups of researchers have examined educational achievement in early-treated CH children. Although outcome is usually reported to be quite favorable, arithmetic10 ,20 and reading comprehension19 may be weaker than normal. Because these studies may have lacked proper controls, were based results on nonvalidated instruments given once, and did not control for educational experience, it is still not known how psychoeducational functioning is exactly affected and if abilities change over time.
We describe presently the findings of a longitudinal cohort study, in which the psychoeducational characteristics of school-aged children with CH were examined in detail relative to classmate and sibling controls. All children were evaluated in the same 2 grades at school, which represent major transition points in elementary education. To allow for uniform teacher effects and educational experiences, assessments were conducted only during the latter part of the school year. Our goals were to describe the psychoeducational profile and specific learning disabilities of children with CH, the effects of early hypothyroid and treatment factors, and changes over time.
The CH group consisted of 50 girls and 33 boys, who were born between 1976 and 1985. Fifty-three children were evaluated in the third grade and 51 in the sixth, while 21 children were studied in both grades. The mean ± standard deviation (SD) age of the grade 3 sample was 8.8 ± .3 years and of the grade 6 sample was 11.8 ± .4 years.
The oldest children were identified as having CH through an experimental Toronto-based newborn screening program21that evaluated cord blood thyrotropin levels. After November 1978, children were also identified through the province of Ontario newborn-screening program,22 which evaluated for thyrotropin elevations from dried blood spots typically collected on the second or third days of life. Both programs identified several of the same children who were born between 1978 and 1981.
Each child's confirmatory diagnosis of CH was made at The Hospital for Sick Children (HSC), where treatment and subsequent follow-up were provided. HSC is the regional facility of the provincial screening program serving the Toronto and nearby regions catchment area. It is responsible for ∼40% of CH cases annually. All but 1 child (who was studied in her home) also had their CH monitored continuously at HSC until 9 years of age. The mean age of treatment onset was 12.8 ± 5.2 days for grade 3 children and 16.0 ± 9.9 days for grade 6. At both grade levels, 14 children (27% and 28%, respectively) were athyreotic.
All the children were participants in a longitudinal prospective study that began in 1981 and continued past this study. Children received semiannual psychological evaluations until age 2 and annual assessments thereafter. The majority of children born after 1980 were followed since 6 months of age.23 Children from the original cohort who did not participate presently had either dropped out (11%), mostly attributable to family relocation, or now declined (4%), although several of these children did participate in the subsequent study.24 ,25 Also, some children from the original cohort had not yet reached the grade 3 level (3%) or were beyond grade 6 (4%) when this study took place. A comparison of the present sample with those from the original cohort not now participating showed no differences in type or severity of CH or IQ at 5 years of age.
The control group consisted of 42 siblings to control for family background factors and 80 classmates to control for teacher and educational methods. There were 21 siblings and 27 classmates in grade 3 (mean age: 8.6 ± .5 years) and 22 siblings and 54 classmates in grade 6 (mean age: 11.9 ± .4 years), with 1 sibling and 1 classmate participating at both grade levels. In a few cases, more than 1 classmate per CH child participated. The control group did not differ in age or gender composition from children with CH.
Parents of CH children were notified by letter of this study and invited to participate and permit testing of siblings, if applicable. The permission of parents whose children were attending public schools in the immediate Toronto vicinity was also sought for recruiting controls from their child's class.
The process of recruiting controls involved a number of steps. First, we wrote the research offices of 12 major public school boards in Metropolitan Toronto. All but 1 agreed to participate (the latter declined for labor reasons). No private schools were contacted. Next, if the CH child's parents granted us permission to recruit controls from their child's class, the principal of that child's school was contacted in writing and by phone. If the principal agreed (the majority), the child's teacher was then contacted directly. The teacher was asked to distribute home to all the parents of the boys or girls in the class (depending on the sex of the child with CH) a letter describing the project and inviting their child to participate in it. To maintain the anonymity of the child with CH, this child was included in the process (parents knew in advance of the letter) and the teacher was only told an unspecified child from the class was participating in a developmental study at HSC. The child with CH was never identified. Interested control parents were asked to contact us directly to schedule the testing. Teachers of CH children living outside the immediate Toronto vicinity, attending a private school, or from the nonparticipating school board were contacted directly and asked to complete questionnaires, as were teachers of sibling controls.
To allow for a consistent amount of teaching, which could bias achievement test results, all the testing occurred in the spring semester (February to June). Testing was conducted in the school, hospital, or, in a few cases, the child's home.
Tests and Measures
As listed in Table 1, the test battery consisted of 8 of the 11 Wechsler Intelligence Scale for Children-Revised subtests,26 a screening test for learning disabilities,27 and multiple measures of reading, spelling, arithmetic, and writing.28–32 Oral language, phonological processing, and memory tasks known to be associated with learning disabilities were also provided.33–36 Parents and teachers completed standardized questionnaires of school behavior and attention.37–40
For each child with CH, information was available through the medical chart as to etiology, severity, and duration of hypothyroidism. Causes were determined by technetium scans on day of confirmatory diagnosis at HSC. Skeletal maturity based on knee radiographs on the day of diagnosis provided an indicator of onset of CH. These were provided to us by the radiologist as described previously15 and were reported in weeks of gestation or postnatal age using the standards set out by Greulich and Pyle.41 The severity of hypothyroidism was based on levels of thyroid hormone levels at diagnosis. The postnatal duration of CH was arbitrarily defined as the exact age in days when the child's total T4 levels reached 128 nmol/L.7 For each child, an initial severity score was additionally computed based on bone age at diagnosis, initial T4 level, and an etiologic-weighting score derived from the mean T4 level at diagnosis for each etiologic group. Treatment factors included the age at which levothyroxine (L-T4) therapy was initiated and the starting and maintenance dose levels.
Data were analyzed by Student's t tests and analyses of variance. The Bonferroni P correction was applied to tests involving multiple subscales. Whenever justified by previous findings, 1-tailed tests were used; otherwise, 2-tailed. Pearson product–moment correlations were computed between selective biomedical variables and test scores. χ2analyses served to compare paired dichotomous variables (eg, group and learning disabilities).
Comparisons of classmates and siblings on all measures at both grade levels revealed they differed on only 1 variable, Full Scale IQ at the sixth grade level. Classmates scored significantly (P < .05) higher than siblings (103.8 vs 98.6). As a consequence, the results from these 2 subgroups were combined in all subsequent analyses.
Children with CH did not differ from the combined control group in special education placement (4% vs 2%) grade retention (2% vs 0%), or academic problems reported by parents (21% vs 13%). The Einstein,23 a screening test for learning disabilities given to grade 3 students only, showed the CH group was at a slightly increased risk of having a generalized learning disability (23% vs 13% for control children) and more likely to exhibit arithmetic (P < .05), auditory memory (P < .05), and visual memory (P < .01) difficulties.
On intelligence tests, children with CH scored significantly lower than combined control children in Full Scale IQ (105 vs 112;P < .05); Verbal IQ (103 vs 109; P < .01); Performance IQ (107 vs 111; P < .005); and the Vocabulary and Coding subtests (P < .01 for both).
Shown in Table 2 are the achievement test and supplementary cognitive test results of CH and control children. Children with CH did more poorly than control children in oral reading comprehension using the Gilmore Oral Reading Test (GORT;P < .05) but did not differ in silent reading comprehension (using the Woodcock Johnson Reading Mastery Task [WJRMT] Passage Comprehension) or word decoding (Wide Range Achievement Test-Revised [WRAT-R] Reading). Regarding arithmetic, the CH group scored lower than controls on both the WRAT-R and Keymath (data not provided) arithmetic tests (both: P < .05). On the Keymath, the CH group was performing approximately one third of a grade below par, whereas control children were exactly at grade level (P < .05). Children with CH demonstrated more difficulty on subtraction, division, and numerical reasoning items but did not differ from controls on addition or multiplication items. Detailed analyses of children's arithmetic error profiles indicated no unusual problems or specific deficits (K. Petric and J. F. Rovet, unpublished data).
Children were also assigned to different learning disability subtype groups based on their WRAT-R Reading and Arithmetic subtest scores, using a 10th percentile cutoff score as described previously.42 The CH group was significantly more likely than controls to have an arithmetic (13.2% vs 2.2%) or combined arithmetic/reading disability (9.4% vs 4.3%). The groups did not differ in reading disabilities.
The CH group did not differ from control children on tests of phonological processing (WJRMT Word Attack), writing (Test of Written Language [TOWL]), or spelling (WRAT-R Spelling). The results of the supplementary tests also indicated no differences on language and memory tasks.35 ,36
Parents rated children with CH higher than controls in terms of behavior problems (mean t scores: CH = 53.7; control = 48.3; P < .05). The groups differed most on scales of Somatic Complaints (P < .05) and Attention Problems (P < .05). Table 3, which provides the results of the teacher rating scales, shows the teachers (who were blinded to CH status) rated children with CH less favorably in terms of overall and internalizing behavior problems (P < .05), effort at school (P < .05), and school performance (P < .10). The CH group was also rated somewhat lower in attention (P < .10).
In addition, a series of paired comparisons was also conducted between the children with CH and a same-sex and nearest-in-age classmate. The findings revealed that children with CH performed significantly poorer than their matched classmates in oral reading comprehension (P < .01), global arithmetic (P < .01), and fractions and numerical reasoning (P < .05). Parents rated children with CH as more inattentive (P< .01) and having more behavior problems generally (P< .01), whereas teachers rated them as significantly less happy (P < .05) and more impulsive/hyperactive (P < .05).
The CH group did not differ from control children in special education placement (4% vs 7%), grade retention (4% vs 7%), or parent-reported academic problems (22% vs 17%). On the GORT, the CH group read more slowly than children (P < .05) but did not differ from them on other measures of reading, writing, spelling, or arithmetic or supplementary tasks of phonological processing or grammatical sensitivity (Table 2). The groups did not differ in their frequencies of specific learning disability subtypes using the procedures described for grade 3 children. Paired t tests comparisons between children with CH and matched classmate controls also indicated no differences in achievement.
The results of the Denman memory task indicated the CH group achieved significantly lower scores on the Full Scale Memory Quotient (82.1% vs 88.9%; P < .05), which was contributed to primarily by their poorer recall of a story after a 20-minute delay.
There were no meaningful differences between the CH and control groups on the parent rating scales or those completed by teachers, which are shown in Table 3. Although not significant, teachers tended to rate the child in their class with CH lower in school performance, effort, and learning and attentiveness (P < .10).
Across Grade Levels
The CH and control samples were compared using the scores from the later grade at which the child participated if he or she was tested twice. The CH group scored significantly lower than controls in IQ (P < .05), arithmetic (P < .01), reading comprehension (P < .05), and reading rate (P < .01). Parents rated children with CH higher in hyperactivity (P < .05) and lower in social competence (P < .01). Teachers saw the CH children as performing more poorly (P < .05) and not learning as well (P < .05).
For the 21 children with CH studied longitudinally, the results from both grade levels were compared using repeated measures analyses of variance. A significant improvement from the average to high average range was noted in word decoding (P < .05), whereas passage comprehension skills remained relatively weak at the 31st and 39th percentiles respectively. Arithmetic and written language results were also unchanged. There were no significant changes on any of the measures of behavior reported by parents or teachers, with the exception of teachers rating the CH children's school performance less favorably at the sixth than the third grade level (P < .05). For the grade 6 results, a supplementary comparison of the 21 children with those tested only in grade 6 indicated that the children tested twice performed better than those tested once on oral reading (P < .05) and arithmetic (P < .01 for both), signifying perhaps a practice or compliance effect. Children tested twice were also rated higher by their parents higher on scales of social competence (P < .05) and extracurricular activities (P < .05) and lower on indices of behavior problems (P < .05) Teachers also rated these children higher in popularity (P < .05) and lower in depression (P < .05).
Correlations With Biomedical Indices
Children with CH were compared according to cause of hypothyroidism. The results for both grade levels revealed no significant differences between athyreotic, dyshormonogenic, and ectopic groups.
Children were also stratified according to whether their T4 values at diagnosis were above or below 50 nmol/L.16 Children with the lower pretreatment T4 values scored significantly lower than those with higher T4 values in Full Scale IQ (P < .05 for both grades), Performance IQ (P < .05 for grade 6 only), reading (P < 05 for both grades), and arithmetic (P < .05 for grade 3 only). The lower reading scores of children with the lower pretreatment T4 levels reflected difficulties with word decoding, not comprehension, while their arithmetic problems reflected difficulties with division and fractions. Parents of third grade children with lower diagnostic T4 levels reported that their children were more withdrawn (P < .001) and exhibited more somatic complaints (P < .001) as well as more social (P< .05) and attention (P < .05) problems than did parents whose children had higher T4 levels at diagnosis. Teachers also saw the children with lower T4 levels at diagnosis as being less popular (P < .05). There were no effects of initial T4 levels on the behavioral characteristics of sixth grade children.
Children with CH were similarly stratified by their starting L-T4 dose levels (below or above 7.8 μg/kg), as described previously.11 There were no effects of starting dose level on any aspect of achievement at either grade level. However, parents of children in the higher starting dose group reported more internalizing (P < .05) and externalizing (P < .05) behavior problems attributable to elevations on scales of anxiety/depression (P < .05) and thought problems (P < .05), even though these children were said to have better social relations (P < .05). Teachers of third grade students rated the children in the higher dose group as having a higher learning capacity (P < .01), consistent with previous findings.7
Correlations were also computed between behavioral parameters and age at treatment onset, bone age at diagnosis, age at T4 normalization, and initial severity (derivation of composite score described above). There were no effects of age at initiation of L-T4 therapy. In contrast, skeletal maturity at diagnosis was significantly correlated with sixth grade phonological processing skill (WJRMT Word Attack:P < .05), mechanical reading (WRAT-R Reading:P < .10), and grammatical sensitivity (Error Cloze task: P < .05). A longer postnatal duration of hypothyroidism was associated with poorer grammatical sensitivity (P < .05 both grades) and weaker phonological processing ability (using the Goldman Fristoe Woodcock [GFW] Sound Analysis test: P < 05 for grade 6 only). Figure 1 shows for grade 6 students the relationships between bone age at diagnosis and WJRMT Word Attack and normalization age and GFW Sound Analysis.
Present results indicate that early-treated CH children do not have an increased risk of severe learning disabilities, although they tend to do more poorly than controls in selective areas, especially at grade 3. These areas include oral language, arithmetic, and reading comprehension, which is consistent with mild, nonverbal learning disability profiles.42 By the sixth grade, we noticed marked improvements in most basic achievement areas. Although this is suggestive of catch up, it may be offset by new emerging problems reflecting the lower ratings sixth grade teachers gave the CH children in overall school performance. This may signify that the nature of the deficit was changing from reading and arithmetic to more complex academic skill areas (eg, science) that make demands on visuospatial abilities, memory, and attention abilities, which also continue to be weaker in these children.24 Indeed, a supplementary analysis of the teacher ratings of the following school subject areas using the Teacher Report Form38 indicated that the CH group (at grade 6) was more likely to obtain below-average ratings in language arts (P < .01), science (P < .05), and social science (P < .05); however, they were at par in French, music/art, writing, spelling, and arithmetic.
For children with CH, correlations were significant certain disease- and treatment-related variables and selective abilities. These abilities were those that were basic to the reading process (eg, phonological processing and grammatical sensitivity), although reading itself was unaffected. In particular, we found that more severe pretreatment hypothyroidism and a greater delay in achieving euthyroidism were both associated with weaker phonological processing and poorer grammatical sensitivity and that these relationships persisted until at least the sixth grade 6 level. However, reading achievement levels were unrelated to severity or duration. One explanation to account for the discrepancy between the skills underlying reading and reading itself is that the poorer phonological and receptive language skills reflected mild to severe sensorineural hearing loss (especially in the high frequency ranges) that is often observed in these children and contributes to weaker phonological awareness as well as indirectly to compromised reading.11The increased incidence of sensorineural hearing loss in CH children is thought to reflect improper cochlear development attributable to its dependence on thyroid hormone in the neonatal period.43
Low T4 levels at diagnosis were also correlated with parent- and teacher-reported behavior concerns including poorer concentration, consistent with previous research.10 A higher starting dose level of L-T4 also contributed to more behavior problems but better learning ability.
Parent and teacher behavior ratings as well as additional studies from our lab13 and others12 indicate that attention may be problematic in children with CH. Although they do not show attention deficit hyperactivity disorder per se, they do indicate greater difficulty concentrating. Their attention difficulties reflect both the severity of their neonatal hypothyroidism,11 as well as their (high) thyroid hormone levels at time of testing.13 ,44 Therefore, although children with CH may not have a reading or arithmetic disability, ultimately, their school performance may be compromised by these attention factors. In light of correlations between current hormone levels and performance on attention tasks,44 it is necessary that these children are properly treated and closely monitored throughout childhood and adolescence. It is also important that their thyroid hormone levels are maintained within the normal range to minimize attention difficulties associated with T4 elevations.
Although the present results are very encouraging, one must be cautioned of biases within our study, which may potentially favor our results or limit interpretations. The present sixth grade classmate control children seemed to be functioning below par (given their lower IQ scores than third grade control children), while the third grade classmates may be a select group from primarily better socioeconomic and nonethnic neighborhoods, thus raising the standards of comparison for the CH group. However, because we did not obtain controls from classes of CH children attending private schools (who were the most privileged) and because the sibling control children (who spanned all socioeconomic classes) did not differ from classmates, the effects of these biases were likely minimal. Bias in our CH sample is also possible given the moderately high dropout rate and the fact that our earliest-identified children (who were treated slightly later and at a lower dose level), were beyond the sixth grade when this research project began. Also, the CH group may have performed very well presently because they were tested repeatedly, whereas control children were not. By the sixth grade, many of the children had received the WRAT-R 5 times and the WJRMT 3 times. Indeed, when we compared sixth grade children according to whether they were assessed at grade 3, those tested at both grade levels performed better on some of the tests. Furthermore, because our children were tested repeatedly (many since infancy) and parents had always received a written report of their children's results, it is possible that their parents were already providing them with additional tutoring as recommended in our reports. The observation that by grade 6, fewer children with CH had repeated a grade than controls, even though they were more likely to be experiencing academic difficulties, supports this interpretation. Additional intervention from tutoring may have offset the problems the children would have otherwise experienced had they not been so closely studied. However, information on tutoring and additional support was not recorded presently.
Although present results may be characteristic of children treated at university teaching hospitals where care is optimal, it is not certain whether these encouraging findings hold for all children with CH, especially those diagnosed and treated in community settings, where treatment is neither as prompt45 nor as thorough.46 Our findings, showing that the selective skills underlying the reading process are sensitive to when euthyroidism was achieved, suggest the importance of early prudent detection, regular follow-up, and close monitoring. Studies comparing children diagnosed and treated in the community versus academic health science centers are also necessary to fully appreciate the true benefits of screening.
In summary, present results show that early-treated CH is associated with mild delays in several fundamental achievement skills, although there seems to be some catch up by the sixth grade level. Children with the most severe hypothyroidism initially or the longest times to achieve euthyroidism had the poorest academic achievement and more behavior concerns. Children receiving a higher starting dose of L-T4 were reported to have better learning ability at the expense of more behavior problems. Improvements in basic skills (eg, reading and arithmetic) at the grade 6 level were offset by teacher-reported difficulties in more complex subject areas (eg, language arts, science, and social science). In light of our findings showing that CH adolescents have persistent visuospatial deficits and memory and attention problems,25 ,47 it is important that additional assistance be provided early, so that later accomplishments are not jeopardized by these deficits. It is also prudent to watch CH children who exhibit attention and memory problems very closely, because their problems may be related to inadequate levels of thyroid hormone and treatment factors at the time of testing. Studies to evaluate the accomplishments of children with CH in late adolescence and early adulthood are necessary to establish the true long-term benefits of newborn screening.
The research was supported by grants from Ontario Mental Health Foundation and Ontario Ministry of Social and Community Services.
We appreciate the contributions of Donna Sorbara, MinNa Hockenberry, Lindsay Ireland, Kathy Petric, Sandra Cole, Wynsome Walker, Jennifer Rovet, and especially Deborah Altman, who coordinated this portion of the study. We thank the principals and teachers from Durham, Durham Separate, East York, Etobicoke, Metropolitan Toronto Separate, Peel, Peel Separate, Scarborough, Toronto, York, York Region, and York Separate Boards of Education. Above all, we thank the children, parents, and siblings who so willingly participated in each phase of the project.
- Received June 25, 1998.
- Accepted June 4, 1999.
Reprint requests to (J.F.R.) Department of Psychology, The Hospital for Sick Children, 555 University Ave, Toronto, Canada M5G1X8. E-mail:
Portions of this study were presented at the International Thyroid Congress; September 10–15, 1995; Toronto, Canada.
- Simons WF,
- Fuggle PW,
- Grant DB,
- Smith I
- ↵Rovet J. Hypothyroidism: intellectual and neuropsychological functioning. In: Holmes C, ed. Psychoneuroendocrinology: Brain, Behavior, and Hormonal Interactions. New York, NY: Springer Verlag; 1990:273–322
- Simons WF,
- Fuggle PW,
- Grant DB,
- Smith I
- Simons WF,
- Fuggle PW,
- Grant DB,
- Smith I
- Tillotson S,
- Fuggle P,
- Smith I,
- Ades A,
- Grant D
- ↵Campos SP, Sandberg DE, Barrick C, Boorhess ML, MacGillivray MH. Outcome of lower L-thyroxine dose for treatment of congenital hypothyroidism. Clin Pediatr. 1994;514–520
- New England Congenital Hypothyroidism Collaborative
- Walfish PG,
- Ginsberg J,
- Rosenberg RA,
- Howard NJ
- ↵Peter F. Screening for congenital hypothyroidism with the TSH test. In: Porter L, ed. Proceedings of the Fourth National Newborn Screening Symposium. Columbus, OH; 1988:147–156
- ↵Rovet J. Behavioral and cognitive abnormalities associated with congenital hypothyroidism. In: Hauser P, Rovet J, eds. Thyroid Diseases of Infancy and Childhood: Effects on Behavior and Intellectual Development. Washington DC: American Psychiatric Press. 1999:85–126
- ↵Wechsler D. The Wechsler Intelligence Scale for Children-Revised. New York, NY: Psychological Corp; 1974
- ↵Gottesman RL, Cerullo FM. Eistein Assessment of School-Related Skills. Cleveland OH: Modern Curriculum Press; 1988
- ↵Jastak S, Wilkinson GS. Wide Range Achevement Test-Revised. Wilmington, DE: Jastak Associates; 1984
- ↵Gilmore JV, Gilmore EC. Gilmore Oral Reading Test. New York, NY: Psychological Corporation; 1968
- ↵Woodcock RN. Woodcock Reading Master Tests-Revised. Circle Pines, MN: American Guidance Service; 1987
- ↵Hammill DD, Larsen SC. Test of Written Language. Austin, TX: Pro-Ed; 1978
- ↵Connelly A, Nachtman W, Pritchett E. Keymath Diagnostic Arithmetic Test. Circle Pines, MN: American Guidance Service; 1981
- ↵Goldman R, Fristoe M, Woodcock RW. GFW Sound-Symbol Tests. Circle Pines, MN: American Guidance Service; 1974
- ↵Kirk S, McCarthy J, Kirk W. Illinois Test of Psycholinguistic Abilities. Urbana, IL: University of Illinois Press; 1968
- ↵Denman SB. Denman Nueropsychology Memory Scale. Charleston, SC; 1984
- ↵Achenbach TM. The Child Behavior Checklist. Burlington, VT: University of Vermont; 1991
- ↵Achenbach TM. Teacher's Report Form and Profile for Ages 5–18. Burlington, VT: University of Vermont; 1991
- ↵Conners KC. Conners' Rating Scales. Toronto, Canada: Multi-Health Systems; 1990
- ↵Ullmann RK, Sleator EK, Sprague RL. ADD-H Comprehensive Teacher's Rating Scale (ACTeRS). Los Angeles, CA: Western Psychological Services; 1991
- ↵Greulich WW, Pyle SI. Radiographic Atlas of Skeletal Development of the Hand and Wrist. 2nd ed. Stanford, CA: Stanford University Press; 1959
- ↵Rovet J. Congenital hypothyroidism. In: Rourke BP, ed. Syndrome of Nonverbal Learning Disabilities: Neurodevelopmental Manifestations. New York, NY: Guilford; 1995:255–281
- Bradley DJ,
- Towle HC,
- Young S
- Rovet JF,
- Ehrlich RM,
- Donner E
- ↵Hollowell JG, Therrell BL, Hannon WH. Newborn Screening for Congenital Hypothyroidism. Oxford, UK: Oxford University Press; 1998
- Ahmed SF,
- Barnes ND,
- Hughes IA
- Copyright © 2000 American Academy of Pediatrics