Growth hormone treatment in the United States: demographic and diagnostic features of 2331 children.

Demographic, diagnostic, and baseline clinical data were collected for a large cohort (N = 2331) of children who started treatment with biosynthetic human growth hormone (GH) between October 1985 and October 1987. Eighty-one percent met classic criteria for GH deficiency and were classified as having idiopathic GH deficiency (59%), organic GH deficiency (18%), or septo-optic dysplasia (4%). The remaining 19.8% had short stature of varied causes. Height standard deviation score at diagnosis, maximum GH response to stimulation, and heights of parents were examined according to gender, race, age at diagnosis, and previous treatment history. The predominance of boys in all subgroups except septooptic dysplasia, and the observation that girls with idiopathic GH deficiency were comparatively shorter than boys at diagnosis, suggest ascertainment bias. Black children with idiopathic GH deficiency were shorter than white children at diagnosis, and their low overall representation (6.0%) compared with their percentage in the at-risk population (12.9%) also suggest ascertainment bias among races. These data provide a profile of GH deficiency as it is currently defined and expose possible inherent biases in the diagnostic process. Now that GH supply is no longer limited, criteria for its use should be formulated to avoid apparent underascertainment or late diagnosis of GH deficiency in girls and black children.