Individuals were chosen for this study on the basis of having hyperlipidemic risk (T-Chol >= mg/dl, or HDL-C <= mg/dl). How this gene was chosen is unclear. The authors state “In the course of serial investigations of population genetics for hyperlipoproteinemia in a cohort of an area located in east-central Japan, we recognized a correlation between lipoprotein variations and polymorphism of the RP1 locus.” but give no references and so we don’t know how these studies were conducted and how many genes were being examined.
p = 0.0006 for homozygotes vs others having higher triglycerides (40% higher), p=0.04 for lower HDL (12% higher).
Of the 280 individuals homozygous for this variant, approximately half of the individuals (136) presented with hypertriglyceridemia (TG > 150), whereas only a quarter (14) were hypertriglyceridemic among 52 individuals with one or no copies: the authors report this as p = 0.04, but we calculate it as p = 0.004 using two-tailed Fisher’s exact.
Regarding significance: it’s hard to know whether these p-values can be taken at face value, because it’s unclear how the gene was chosen in the first place — whether and how much multiple hypothesis correction needs to occur. To address this, we score this as having three points rather than the four it might merit if no correction need occur.
PMID: 11926512 implies that overall the risk for hypertriglyceridemia (TG > 150) in Japanese is ~5% (9.4% in males, 0.8% in females), so we use 5% to calculate attributable risk of 3.7% (total 8.7%).