RT Journal Article SR Electronic T1 Blood lipids: the relationship with alcohol intake, smoking, and body weight. JF Journal of Epidemiology and Community Health JO J Epidemiol Community Health FD BMJ Publishing Group Ltd SP 197 OP 202 DO 10.1136/jech.46.3.197 VO 46 IS 3 A1 Wannamethee, G A1 Shaper, A G YR 1992 UL http://jech.bmj.com/content/46/3/197.abstract AB STUDY OBJECTIVE--The aim was to assess the interrelationship between alcohol intake, cigarette smoking, body weight, and blood lipid concentrations. DESIGN--This was the cross sectional (screening) phase of a prospective study. The main outcome measure was the blood lipids (serum total cholesterol, HDL cholesterol, and triglycerides). SETTING--General practices in 24 towns (The British Regional Heart Study). SUBJECTS--Subjects were 7735 men aged 40-59 years, selected at random from the age-sex registers of one group practice in each of the 24 towns. RESULTS--Univariate analysis showed little association between alcohol intake and total cholesterol, a strong positive relation with HDL cholesterol, and a significant increase in triglycerides in heavy drinkers. A strong positive association between alcohol intake and body weight was present in non-smokers but not in moderate/heavy smokers. With the exception of HDL cholesterol, the relationships between alcohol intake and serum lipids were significantly different in smokers and non-smokers, apparently due to the opposing effect of smoking on blood lipids and body weight. Total cholesterol and triglycerides were significantly and positively associated with alcohol intake in non-smokers, the cholesterol association being largely mediated by the influence of alcohol on body weight. In smokers, no such association was seen: current smokers who were heavy drinkers or non-drinkers had the lowest mean cholesterol levels. CONCLUSIONS--The association between alcohol intake and body weight and alcohol intake and blood lipids are strongly conditioned by cigarette smoking. Simple standardisation for smoking in multivariate analyses may obscure the independent relationship with alcohol. These findings are of importance in studies seeking to relate alcohol intake, body weight, or cigarette smoking to blood lipid concentrations, or blood lipid concentration to morbidity or mortality.