As non-fasting status may affect triglyceride levels, we conducted secondary analyses with the phenotype of race-specific triglyceride residuals adjusted for age, gender, and fasting status and the results were not materially different (data not shown).
