To evaluate the effectiveness of criteria based on child-parent assessment in predicting familial hypercholesterolemia (FH)-causative mutations in unselected children with hypercholesterolemia.

LDLR, APOB, and PCSK9 genes were sequenced in 78 children and adolescents (mean age 8.4 ± 3.7 years) with clinically diagnosed FH. The presence of polygenic hypercholesterolemia was further evaluated by genotyping 6 low-density lipoprotein cholesterol (LDL-C)-raising single-nucleotide polymorphisms.

Thirty-nine children (50.0%) were found to carry LDLR mutant alleles but none with APOB or PCSK9 mutant alleles. Overall, 27 different LDLR mutations were identified, and 2 were novel. Children carrying mutations showed higher LDL-C (215.2 ± 52.7 mg/dL vs 181.0 ± 44.6 mg/dL, P < .001) and apolipoprotein B levels (131.6 ± 38.3 mg/dL vs 100.3 ± 30.0 mg/dL, P < .004), compared with noncarriers. A LDL-C of ~190 mg/dL was the optimal value to discriminate children with and without LDLR mutations. When different diagnostic criteria were compared, those proposed by the European Atherosclerosis Society showed a reasonable balance between sensitivity and specificity in the identification of LDLR mutations. In children without mutation, the FH phenotype was not caused by the aggregation of LDL-C raising single-nucleotide polymorphisms.

In unselected children with hypercholesterolemia, LDL-C levels >190 mg/dL and a positive family history of hypercholesterolemia appeared to be the most reliable criteria for detecting FH. As 50% of children with suspected FH did not carry FH-causing mutations, genetic testing should be considered.