Gender bias in children receiving growth hormone treatment.

BACKGROUND About twice as many boys than girls are treated with GH. Ascertainment bias is a possible explanation. HYPOTHESES For ascertainment bias, the gender least frequently treated should be relatively shorter, and in an unbiased population sample, equal numbers of boys and girls should be eligible for GH treatment. SUBJECTS AND SETTING In 2007 a total of 1485 Australian children received GH (OZGROW database). Heights were also obtained from two recent unbiased surveys consisting of 3596 and 4794 Australian children. METHODS Numbers of boys and girls treated with GH were determined for each treatment indication. Height sd scores (SDS) at first presentation for GH-treated boys and girls were assessed. Frequency of boys and girls from two unbiased populations with height SDS less than -2.326 were recorded. OUTCOMES OUTCOMES included gender frequencies and height SDSs. HYPOTHESES were formed before interrogation of preexisting databases. RESULTS More boys than girls received GH (P = 3.68 x 10(-20)). By indication: biochemical GH deficiency (P = 0.001), cranial irradiation (P = 0.002), slow growing (P = 2.09 x 10(-16)), and chronic renal failure (P = 0.061). Approximately equal numbers of girls and boys were treated for hypoglycemia (P = 0.543). Slow-growing girls were relatively shorter than boys for ages spanning 4.50-8.49 yr (P = 3.80 x 10(-4)), but boys were relatively shorter in the 6.00- to 17.99-month age group (P = 0.011). Biochemical boys were relatively shorter than girls (P = 0.023). In the two unbiased surveys, boys outnumbered girls 11 to six and 16 to eight for height SDS less than -2.326. CONCLUSIONS There is a gender bias in this GH-treated population. Ascertainment bias does not appear to be the major cause.