Congenital hypothyroidism: treat children but don't forget their parents.

Abstract

Congenital hypothyroidism (CH), detected in 1:3500 to 1:4000 newborns world-wide, is one of the most common preventable causes of mental retardation if treated early (1). In the 1970s neonatal screening programmes were started to guarantee detection and treatment from the first weeks of life. Prior to the onset of screening, when clinical diagnosis was commonly delayed, the neurodevelopmental prognosis in CH children was poor. A clear inverse relationship was evident between age at start of treatment and intelligence quotient (IQ), the worst outcome (75% with IQs < 80) being observed in those infants diagnosed at the ages of 6 and 9 months (2). Screening programmes have been extremely successful in the detection and early treatment of infants with CH, and in preventing the serious neuropsychological sequelae of this condition. Although neuropsychological follow-up studies in general have been favourable, slightly lower IQs compared with unaffected controls have been reported, particularly in the most severely affected infants (1, 2). Some follow-up programmes report good psychometric outcomes, with IQs similar to controls (3–5) and no apparent impediment in school performances (3). Other reports indicate a mild decrease in IQs of CH infants in spite of early substitution treatment (6–9). The latter studies also report poorer motor skills (6, 10, 11), defective language abilities (12) and learning problems in school (13). A 1996 metanalysis of literature data, which included 675 patients and 570 controls in North America and Europe, showed a trend toward lower IQs and slightly poorer motor skills in CH patients compared with controls (14). Pooling of data showed a significant deficit of the mean IQ of 6.3 points (95% confidence interval: 4.7–7.8). A common problem of these studies is the difficulty in controlling for appropriate modulating influences of neuropsychological development such as genetic background, socio-economic status, cultural level of the family and education burden. While false positive cases, siblings, matched or unselected normal children have been used as controls (14), unaffected identical twins raised in the same family would be the best control when complex functions such as intelligence are investigated. Two genetically identical twins, one of whom had thyroid agenesis and the other who was unaffected were described in this Journal in 1997 (15). In spite of early (4th week of life) and adequate Lthyroxine (L-T4) therapy (9 mg/kg per day), the CH twin showed IQ scores that were in the normal range but lower compared with the unaffected twin. At the age 7 of years the deficit in total IQ scores was 7 points (108 vs 115), a figure similar to that estimated in the metanalysis of literature data. Several risk factors for the eventual outcome of the neuropsychological development in early-treated CH have been evaluated. These include: (i) pretreatment serum T4 concentration; (ii) neonatal skeletal maturation; (iii) aetiology of CH; (iv) age at start of treatment; (v) starting L-T4 dose; (vi) adequacy of substitution treatment in the first 2 years of life and even afterwards; (vii) socio-economic class of the family. Most follow-up studies found a significant correlation between cognitive development and pretreatment serum T4 (4, 6, 8–10, 16). In a multicentre survey in the UK (16), the relationship between IQ and serum T4 level at diagnosis was discontinuous with a threshold of 3.3 mg/dl (42.8 nmol/l). Below this level, a 10-point IQ deficit was observed as compared with control subjects of similar social class. Above this threshold, children were unlikely to suffer more than a 5-point deficit. A skeletal development at the time of diagnosis <36 weeks of gestation (12) or an area of the knee epiphysis < 0.05 cm (9) were both found to be associated with significantly lower IQs compared with control groups. Overall, the severity of foetal-neonatal hypothyroidism, defined by skeletal maturation at diagnosis and/or pretreatment serum T4, is the most important independent risk factor for the eventual neuropsychological outcome (14), partly due to an effect on brain development determined prenatally. Small amounts of maternal T4 pass the placenta (17), but it is difficult to believe that this limited maternal supply of T4 is sufficient in most cases of severe CH. In keeping with these observations, several studies found that children with severe CH due to thyroid agenesis achieve significantly lower IQs compared with hypothyroid children with thyroid ectopy who display some degree of residual thyroid function (6, 10, 18). European Journal of Endocrinology (1999) 141 101–104 ISSN 0804-4643