Subjects were 21,572 Japanese men (age range: 35–60 years) who were divided into non-, light (<22 alcohol="alcohol" analysis.="analysis." analysis="analysis" and="and" between="between" but="but" by="by" covariance="covariance" day="day" drinkers.="drinkers." ethanol="ethanol" font="font" g="g" heavy="heavy" intake="intake" investigated="investigated" logistic="logistic" nbsp="nbsp" of="of" ratio="ratio" regression="regression" relationship="relationship" tg="tg" the="the" using="using" very="very" was="was">
Log-transformed TG/HDL-C ratio was significantly lower in light, heavy, and very heavy drinkers than in nondrinkers and was lowest in light drinkers. Odds ratios for high TG/HDL-C ratios in light and heavy drinkers versus nondrinkers were significantly lower than a reference level of 1.00 (light drinkers: 0.63, 95% CI [0.57, 0.71], p < .01); heavy drinkers: 0.75, 95% CI [0.69, 0.81], p < .01]). Odds ratios for high waist-to-height ratio of subjects with versus subjects without high TG/HDL-C ratios were significantly higher than the reference level in non-, light, heavy, and very heavy drinkers and were significantly lower in heavy and very heavy drinkers than in nondrinkers (nondrinkers: 3.84 [3.42,4.31]; light drinkers: 3.65 [2.97,4.48]; heavy drinkers: 3.17 [2.84, 3.54], p < .05 compared with nondrinkers; very heavy drinkers: 2.61 [2.29, 2.97], p < .01 compared with nondrinkers).
Alcohol drinking is inversely associated with TG/ HDL-C ratio and confounds the relationship between TG/HDL-C ratio and obesity.
Read Full Abstract
Request Reprint E-Mail: wakabaya@hyo-med.ac.jp
