In a study of Japanese men, a nonlinear association of alcohol consumption
and blood pressure was not observed. This cross-sectional study was conduct
ed to clarify the effect of infrequent or light alcohol consumption on bloo
d pressure in normotensive Japanese men. The subjects were 2,179 male worke
rs ranging in age from 40 to 54 years. Drinking habits were represented by
three indices: episodic alcohol consumption, frequency of drinking, and mon
thly alcohol consumption. Seven items were analyzed as covariates: age, bod
y mass index, salt intake, physical activity, and levels of gamma-glutamyl
transpeptidase, uric acid, and plasma glucose. The association between drin
king habits and blood pressure was evaluated by using analysis of covarianc
e. Both episodic alcohol consumption and frequency of drinking were signifi
cantly associated with blood pressure in subjects aged 45 to 54 years. In t
hose aged 40 to 44 years, there was a significant association between frequ
ency of drinking and diastolic blood pressure. Significant associations of
monthly alcohol consumption with blood pressure in every age group also wer
e found. In subjects aged 45 to 54 years, linear associations between episo
dic alcohol consumption, frequency of drinking, monthly alcohol consumption
, and adjusted blood pressure were observed. In those aged 40 to 44 years,
nonlinear associations of monthly alcohol consumption and frequency of drin
king with adjusted blood pressure were observed, and threshold effects of 5
40 ml of alcohol per month and a frequency of 1 to 12 days per month were f
ound. In conclusion, habitual alcohol consumption increases blood pressure
linearly in normotensive workers aged 45 to 54 years, although light consum
ption does not affect blood pressure in those aged 40 to 44 years. (C) 2001
Elsevier Science Inc. All rights reserved.