Segregation analysis was performed on the serum uric acid measurements from
523 randomly ascertained Caucasian families from the NHLBI Family Heart St
udy. Gender-specific standardized residuals were used as the phenotypic var
iable in both familial correlation and segregation analysis. Uric acid resi
duals were adjusted for age, age(2), age(3), body mass index (kg/m(2)), cre
atinine level, aspirin use (yes/no), total drinks (per week), HOMA insulin
resistance index [(glucose * insulin)/22.5], diuretic use (yes/no), and tri
glyceride level. Sibling correlations (r=0.193) and parent-offspring correl
ations (r=0.217) were significantly different from zero, but these two fami
lial correlations were not significantly different from one another. After
adjustment for covariates, the heritability estimate for serum uric acid wa
s 0.399. Segregation analysis rejected the "no major gene" model but was un
able to discriminate between an "environmental" and a "Mendelian major gene
" model. These results support the hypothesis that uric acid is a multifact
orial trait possibly influenced by more than one major gene, modifying gene
s, and environmental factors.