Human growth hormone in currently recommended dosage is effective in m
any short children, irrespective of their endogenous growth-hormone st
atus. This suggests that present dosing is pharmacological rather than
physiological. As for any drug, issues of safety should be of paramou
nt concern. Reassuring short-term data with pharmacological dosing or
long-term data with replacement dosing cannot guarantee the ultimate s
afety of this form of therapy. The risk of future malignancy should be
of particular concern. Poorly growing children without classic (sever
e) growth-hormone deficiency constitute an increasing proportion of ch
ildren treated with human growth hormone. There are no satisfactory cr
iteria for the diagnosis of neurosecretory growth-hormone dysfunction.
The closer to puberty these children are treated, the less likely it
is that there will be benefits in terms of increased final height. Rec
ommendations as to a 'safety first' approach to growth-hormone treatme
nt are given. A multicentre controlled trial is urgently needed to est
ablish the benefits of treating children with neurosecretory growth ho
rmone dysfunction.