It is becoming increasingly apparent that pyridoxine responsive homocy
stinuria patients are being missed by newborn screening programs. The
possibility that screening for homocystine, rather than the methionine
, might be more effective was investigated by comparing homocystine an
d methionine levels of non-responsive patients at diagnosis. The plasm
a methionine levels of 11 infants were much more abnormal than the hom
ocystine levels. Urine homocystine was low or not detectable and alway
s less than methionine. Therefore, methionine determination is much mo
re effective than homocystine determination for newborn screening for
homocystinuria. It seems that a second blood specimen at a later age m
ay be required to find the pyridoxine responsive infants. (C) 1997 Els
evier Science Ireland Ltd.