Cut off value of TSH for early prediction of poor neurodevelopment of 5 to 9 year old children in Sri Lanka

Abstract

INTRODUCTION AND AIMS: To assess the relationship between TSH and neuro-development of children between 5- 9 years of age and to identify the cutoff value of TSH for early prediction of poor neuro developmental outcome. METHODS: A descriptive cross sectional study was carried out in 20 randomly selected estates in the Ratnapura district of Sri Lanka from August to December 2010. A validated neurodevelopment assessment tool was used to assess the neuro-development of 1683 randomly selected children. Neuro¬development was expressed as a General Quotient (GQ) score and a score <100 was identified as evidence of poor neuro-development Of 1683 children screened, 519 were randomly selected with consent from the parents; blood was assayed for TSH using the 3rd generation TSH Chemiluminescent immunometric assay. ROC curve analysis was used to find out the cut off value of TSH that best predicts poor neuro¬development in children. RESULTS: Of 519 children whose TSH levels were assayed, 494 had TSH levels within the normal range between 0.60 to 5.40 ulU/ml, of whom 16% had a GQ score <100. There was a significant association between TSH levels and neuro-development [p=0.0001). The best cutoff value of TSH to predict a GQ score <100 was 1.5 p.IU/ml. [Area under cure=73.4%, sensitivity= 97.0% and specificity =70.6%). CONCLUSIONS: The best cutoff value of TSH for prediction of poor neuro-development in children 5 to 9 years of age is 1.50 p.IU/ml.