Four pediatricians introduced a portable cholesterol analyzer into their group practice. Their experience is described on the basis of 12 months of screening in 1665 children and adolescents. The overall 50th and 90th percentile values for a subgroup of 1406 routinely screened children were 156 and 197 mg/dL, respectively, but there was marked variation in these values among specific age and sex groups. Cholesterol levels decreased by age group during the early teenage years and increased thereafter, these changes occurring at ages approximately 2 years younger for girls than for boys. Further analysis of screening results for 398 sibling pairs demonstrated significant concordance between paired cholesterol levels when classified by the respective age- and sex-specific 90th percentile values for each member of the pair. Sibling pairs in which both members' cholesterol values exceeded their 90th percentile value were identified 2.4 times as frequently as expected (confidence interval 1.1 to 4.5, P = .029). The observations reported here indicate that office-based cholesterol screening in a pediatric practice may be both practical and useful, although further consideration of screening criteria is needed. Age- and sex-specific reference values for cholesterol levels during childhood could improve screening results. Special emphasis should be directed toward screening siblings of children in whom high cholesterol levels have been detected.
