Objectives: We have previously reported that brachial-ankle PWV (baPWV) is more strongly related to lipoprotein subclasses than carotid-femoral PWV (cfPWV) among middle-aged white men. The current study examined the associations of baPWV, cfPWV and femoral ankle PWV (faPWV) with lipoprotein subclasses in Japanese Americans, Koreans, and Whites. Methods: The ERA JUMP Study is an international population-based study of subclinical atherosclerosis in men aged 40-49 without clinical CVD in the US, Korea, Hawaii and Japan. Participants for the current study were 310 whites, 303 Japanese Americans and 302 Koreans. PWV was assessed using an automated waveform analyzer (VP2000, Omron, Japan). Lipoprotein subclasses were assessed by NMR lipoproteins. Multiple linear regressions were performed to analyze the association of each PWV with NMR lipoprotein distributions after adjusting for age, body mass index, systolic blood pressure, and other potential confounders. Results: Both baPWV and faPWV had significant association with total, large, and small LDL-P after adjusting for race and other confounders, and across three racial groups except for faPWV in whites ( Table ). cfPWV had also significant associations with total and small LDL-P after adjusting for race and other confounders but these associations were weaker than those with baPWV and faPWV. Similarly, both baPWV and faPWV had significant associations with VLDL-P (total, large and small VLDL-P) and HDL-P (total, large, medium, and small HDL-P) but the associations of cfPWV with these lipoproteins were weaker and mostly non-significant. Conclusions: As compared to cfPWV, both baPWV and faPWV had stronger associations with lipoprotein distributions across three racial groups with very different lifestyle. Our results suggest that baPWV and faPWV are more related to atherosclerosis than cfPWV whereas cfPWV may represent arteriosclerosis. Table 1. Association between LDL subclasses(LDL-P) and carotid-femoral, brachial-ankle and femoral-ankle pulse wave velocities among young men in Korea, Hawaii and North America: Multiple linear regression derived beta (β) values (SE) @ Large LDL Small LDL Total LDL Particles cfPWV All participants $ −0.074 (.050) 0.405 (.204) * 0.152 (.072) * Whites −0.045 (.081) 0.442 (.309) 0.197 (.094) * Koreans −0.179 (.084) * 1.143 (.375) ** 0.337 (.140) * Japanese Americans 0.060 (.98) 0.064 (.361) 0.016 (.151) faPWV All participants $ −0.252 (.067) *** 1.201 (.264) *** 0.331 (.099) *** Whites 0.007 (.100) 0.368 (.389) 0.165 (.122) Koreans −0.252 (.113) * 1.609 (.502) ** 0.472 (.189) * Japanese Americans −0.501 (.152) ** 2.266 (.552) *** 0.698 (.225) ** baPWV All participants $ −0.310 (.053) *** 1.334 (.210) *** 0.404 (.075) *** Whites −0.212 (.108) * 1.220 (.415) ** 0.425 (.131) ** Koreans −0.245 (.079) ** 1.398 (.351) *** 0.493 (.126) *** Japanese Americans −0.194 (.098) * 1.132 (.357) ** 0.392 (.143) ** * p<0.05, ** p<0.01, *** p<0.001 # After excluding participants with diabetes mellitus, and those taking medicines for diabetes mellitus, hypertension and hyperlipidemia, number of participants were 248 Whites, 262 Koreans and 192 Japanese Americans. @ Adjusted for age, systolic blood pressure, body mass index, fasting glucose, pack-years smoking, alcohol intake and LDL cholesterol. $ Additionally adjusting for race.