Family history evaluation as a predictive screen for childhood hypercholesterolemia. Pediatric Practice Research Group

A study was conducted to evaluate the efficacy of family history factors as screening criteria for childhood hypercholesterolemia. When they were seen for routine care at one of eight office practices, 1005 prepubertal children underwent random serum cholesterol determinations. Parental and grandparental histories of cardiovascular risk factors and atherosclerotic complications prior to 55 years of age were also obtained. Of the initial group, 274 children had total cholesterol levels greater than or equal to 175 mg/dL, and 175 of these children returned for retesting after an overnight fast. A total of 88 children were found to have low-density lipoprotein-cholesterol (LDL-C) values greater than or equal to 90th percentile for age and sex. Maternal and paternal histories of hypercholesterolemia were significantly associated with elevated LDL-C (odds ratio = 7.3 and 2.9, respectively), but had extremely low sensitivities (0.09, 0.15) despite modest positive predictive values (0.42, 0.22). Grandparental histories of sudden death, peripheral vascular disease, and gout were associated with elevated LDL-C, but sensitivities and positive predictive values for all of these factors were less than 0.22. Family history factors most commonly recommended as criteria for cholesterol screening in children did not identify half of all the children with elevated LDL-C and did not selectively identify the most severely affected children. Adding information concerning the presence of childhood obesity did not result in appreciable improvement in LDL-C detection beyond that achieved by family history factors alone. It was concluded that if thorough identification of young children with elevated LDL-C is desired, inclusive population screening rather than a family history-based strategy would be the most effective approach.