The perspective by Krude and Blankenstein1 rightly highlights the importance of learning more about the natural history of newborns with mildly elevated thyroid stimulating hormone (TSH) values. Unfortunately, the studies they cite do not reassure us that TSH values or ‘numbers’ that are mildly elevated are of no consequence. The “normal intelligent quotient (IQ)” in the Swedish study of babies with untreated mild to moderate TSH elevation was lower than that seen in untreated patients with a raised TSH that subsequently normalised.2 The study by Oken et al3 did not show a relationship between thyroid hormone levels and IQ, but TSH will be a more telling reflection of what is happening at the neuronal level. Small fluctuations in TSH ‘numbers’ have to be taken seriously when the developing brain appears to be so sensitive to …