[Relations of serum apolipoprotein A-I, A-II and B levels to smoking, drinking and body mass].

To examine whether the serum apolipoprotein A-I level (Apo A-I), serum apolipoprotein A-II level (Apo A-II) and serum apolipoprotein B level (Apo B) are related to cigarette-smoking habits, drinking frequency and body mass index (BMI) independent of serum lipoprotein cholesterol levels (HDLC, LDLC or VLDLC), we statistically analyzed the data on Apo A-I, Apo A-II, Apo B, HDLC, LDLC and VLDLC and the life style data obtained from health examinations of 256 male residents aged 40 to 49 mostly randomly selected from two Japanese areas, Ninohe, Iwate and Ishikawa, Okinawa. The results were as follows: (1) HDLC was strongly positively correlated with Apo A-I and Apo A-II, while Apo B was strongly correlated with LDLC and VLDLC. (2) According to univariate analyses, Apo A-I, Apo A-II, Apo B, HDLC, LDLC and VLDLC were not associated with smoking. On the other hand, drinking frequency was positively associated with Apo A-I, Apo A-II and HDLC. Apo A-I and HDLC were negatively correlated with BMI, whereas Apo B, LDLC and VLDLC were positively correlated with BMI. (3) According to the results of multi-dimensional analyses of covariance, Apo A-II was positively correlated with drinking frequency independent of Apo A-I and HDLC, especially among the individuals with increased HDLC. The same multivariate analysis showed that Apo B was positively associated with smoking independent of LDLC and VLDLC among the individuals without increased VLDLC. From these results, we conclude that Apo A-II may be effective as a biological marker for alcohol drinking independent of Apo A-I and HDLC, while cigarette smoking may affect Apo B through a certain direct mechanical effect. (4) Increased HDLC in obese individuals (BMI greater than or equal to 27) or non-drinkers was associated with remarkably increased Apo A-I, while decreased HDLC in thin individuals (BMI less than 21) was associated with remarkably decreased Apo A-II.