BIOLOGICAL AND LIFE-STYLE FACTORS, AND LIPID AND LIPOPROTEIN LEVELS AMONG JAPANESE-AMERICANS IN SEATTLE AND JAPANESE MEN IN JAPAN

Background. It has been previously shown that Japanese Americans in Se attle have significantly higher cholesterol levels than native Japanes e. The present study examines the association of biological and lifest yle factors with plasma lipid and lipoprotein levels among Japanese Am ericans (JA) and native Japanese (NJ) to determine if these associatio ns are consistent between these high and low cholesterol populations. Methods. Study samples consisted of 710 JA male and 728 JA female volu nteers living in the Seattle area and a random sample of 3833 NJ male urban workers who participated in parallel cardiovascular disease scre ening and lifestyle surveys for 1989-1994. Multiple regression analysi s was conducted to examine the association of lifestyle and biological factors with lipid and lipoprotein levels. Results. Alcohol consumpti on was positively and linearly associated with high density lipoprotei n cholesterol (HDL-C) levels and negatively associated with both low d ensity lipoprotein cholesterol (LDL-C) levels and the ratio of total c holesterol (TC)/HDL-C (P < 0.05 to P <0.001) among JA males and JA fem ales and NJ males. Current smoking habit was observed to be negatively associated with HDL-C levels and positively with TC/HDL-C ratio and l og TG levels (logarithmic transformation of triglyceride values) (P < 0.05 to P < 0.001) among all three groups. Body mass index (BMI) was n egatively associated with HDL-C levels and positively associated with log TG and TC/HDL-C ratio among all three groups (P < 0.05 to P < 0.00 1). Moderate alcohol consumption was negatively associated with log TG levels among JA males and females (P < 0.05), whereas heavy alcohol c onsumption was positively associated with log TG levels in NJ males (P < 0.001). Smoking was positively associated with TC and LDL-C levels (P < 0.05) among JA males, whereas a negative association (P < 0.05) w as observed in NJ males. Conclusion. Overall, the fitted models were c onsistent between JA males and females and NJ males with the exception of smoking on TC and LDL-C. The results suggest that moderate alcohol consumption favourably influences lipid profiles in both high and low cholesterol populations. The results also indicate that light alcohol consumption is associated with decreased triglyceride levels, whereas heavy alcohol consumption is associated with increased triglyceride l evels.