Diagnosing familial hypercholesterolaemia (FH) in children

Children with FH have abnormal carotid arteries by the age of 12 so identifying affected individuals is important. Genotyping remains the diagnostic ‘gold standard’ but the test is expensive and not yet widely available and so could total cholesterol (TC) or LDL-C be used instead. We examined the records of 115 children age <16 years, seen in our Lipid Clinic and genotyped. In 69 affected children (G+), mean TC at their initial visit was 7.6 mmol/l (S.E. 0.18, range 4.7–10.9) and LDL-C was 5.64 ± mmol/l (0.18, n = 63, range 3.0–8.94). TC was higher in girls (8.0 ± 0.28 mmol/l, n = 31) than boys (7.26 ± 0.23 mmol/l, p < 0.05) but not LDL-C. There was a trend for both TC and LDL-C to decrease with age. In 46 G− children, sibs of the above, TC was 4.15 ± 0.11 mmol/l (range 2.37–6.05) and LDL-C 2.13 ± 0.12 mmol/l (n = 35, range 0.42–3.70). There was no difference between boys and girls in TC or LDL-C in the G− group. The lowest observed TC in the G+ group and the highest in the G− group defined the range of overlap (4.70–6.05 mmol/l); 21 children were in this area (18% of the total). Similar calculation for LDL-C showed an overlap range of 3.0–3.7 mmol/l, into which 8 children fell (8% of the total). We conclude that (1) TC <4.7 or >6.1 separate G− and G+ children; (2) TC between these values cannot diagnose FH with sufficient certainty; (3) LDL-C <3.0 or >3.7 may be used instead; and (4) the overlap range with LDL-C is much less than for TC, and so LDL-C should be the measurement of choice in the absence of specific genotyping.