Background:
Immigration studies reported that dyslipidemia is more prevalent in people living in a western lifestyle. However, it is not clear what kind of lifestyle and dietary factors contribute to this difference.
Aim:
To investigate whether lifestyle including dietary factors could explain the difference in serum lipids between Japanese in Japan and Japanese-Americans in Hawaii living a western lifestyle.
Methods:
Analyses were conducted in the INTERLIPID study, where four standardized in-depth 24-h dietary recalls were performed. Non-fasting blood was drawn from the participants who included 1,087 Japanese and 176 Japanese-American. Participants’ characteristics were described as means and SD
,
t
test
was used for the comparison of means between Hawaii and Japan. Multiple linear regression models were used to examine relationships between dietary and lifestyle factors and the difference in serum lipids levels (low-density lipoprotein cholesterol [LDL-C] and log transformed triglycerides [log-TG]) between Hawaii and Japan, adjusting for age, sex and site as the basic model. Other confounders were then added separately to the basic model. Percentage reduction of regression coefficient for the difference between Hawaii and Japan was calculated.
Results:
In men, LDL-C levels were significantly higher in Japanese-Americans (138.3 mg/dl) compared to Japanese in Japan (120.3mg/dl, P<0.0001). Similarly, in women LDL-C levels were higher in Japanese-Americans (135.5 mg/dl) than in Japanese in Japan (123.7mg/dl, P<0.001). Median levels of TG were also higher in Japanese-Americans compared to Japanese in Japan (men: 171.0mg/dl and 132.0mg/dl, P<0.001 and women: 132.0mg/dl and 94.5mg/dl, P<0.0001). Mean intakes of most nutrients were significantly different between Japan and Hawaii for both men and women. In multiple linear regression models, for LDL-C, BMI reduced the regression coefficient of site difference by 44.6% in the basic model that included age, sex and site. Added to the basic model separately, alcohol, dietary cholesterol and saturated fatty acid reduced LDL-C site difference by 11%, 15.6% and 13.4%, respectively. For log-TG, BMI reduced the regression coefficient of site difference by 55.8% and other variables tested reduced the site difference by less than 5%.With step by step addition of other variables (BMI, alcohol, dietary cholesterol, fiber and total fat) into the multiple linear regression model, the site LDL-C and log-TG coefficient further reduced from 14.4 to 3.4 (-76.6%) and 0.31 to 0.10 (-68.2%), respectively.
Conclusion:
BMI explained about half of the higher LDL-C and TG levels in Japanese-Americans in Hawaii compared to Japanese in Japan. BMI together with other macro-nutrients contributed to about 70% of the difference. Further investigation is required to explore the remaining difference.
Apolipoprotein A4-1/2 polymorphism and response of serum lipids to dietary cholesterol in humans