ABO Blood Groups, Lipids, and Coronary CT Imaging in A Japanese Single-Center Cohort

<b>Background and Objectives:</b> Non-O ABO blood groups have been linked to higher coronary risk, plausibly via hemostatic and lipid pathways. However, evidence in Japanese populations and imaging-defined disease is limited. We examined whether ABO status relates to serum lipids and coronary CT imaging findings in Japanese adults. <b>Materials and Methods:</b> We reviewed adults who underwent coronary CT angiography (CCTA) at our institution. After prespecified exclusions, 865 patients comprised the imaging cohort. For lipid analyses, we excluded patients receiving lipid-lowering therapy at the time of blood sampling, leaving 636 patients (lipid subset). ABO blood group was obtained from the medical record as recorded at registration (patient-reported) and was not re-confirmed by laboratory testing for this study. Outcomes were any coronary artery calcium (Agatston score > 0) and ≥50% luminal stenosis on CCTA. <b>Results:</b> In the lipid subset (<i>n</i> = 636), coronary calcium was present in 44-54% of patients across the four ABO groups and did not differ across groups (<i>p</i> = 0.33). Among assessable scans in the imaging cohort, ≥50% stenosis did not differ across the four ABO groups. In multivariable models (<i>n</i> = 636), older age, male sex, hypertension, and diabetes were independently associated with both outcomes (CAC presence and ≥50% stenosis) (all <i>p</i> < 0.05). For ≥50% stenosis, higher High-Density Lipoprotein-cholesterol (HDL-C) was additionally associated with lower odds (<i>p</i> < 0.05). ABO status (O vs. non-O) was not independently associated with either outcome. <b>Conclusions:</b> In Japanese adults undergoing CCTA, type O blood was tied to lower HDL-C and higher diastolic pressure-features that track with cardiometabolic risk-yet ABO type did not independently relate to coronary calcium or CT-defined stenosis once standard risk factors were considered. These data suggest that, in this setting, ABO adds little beyond conventional risk profiling.