Plasma levels of total high density lipoprotein cholesterol (HDL) and Its subtractions (HDL 2 and HDL 3 ) were measured In 366 healthy Caucasian males; these values were related to a number of coronary risk factors. On unlvarlate statistical analysis, total HDL was negatively correlated with cigarette consumption, body mass Index, and serum trlglycerldes, and positively associated with level of physical activity and alcohol consumption. HDL 2 showed an Inverse relationship with cigarette consumption, body mass index, trlglycerldes, and systolic blood pressure and a positive relationship with age. HDL 3 was negatively correlated with cigarette smoking, body mass index, and trlglycerldes and positively associated with exercise level and alcohol consumption. Total HDL and HDL 2 were Inversely related to coronary risk rating, but HDL 3 snowed no significant correlation. Many of these relationships became nonsignificant after allowing for the effects of other variables. In particular, none of the HDL measurements correlated significantly with risk score after allowing for the effect of trlglycerldes. There Is Insufficient evidence at present to recommend the Inclusion of HDL subfractions as routine screening tests for heart disease. HDL subfractions and these and other risk factors in a large group of healthy men.
Methods
SubjectsThe study population consisted of Caucasian males attending a hearth screening center in north London. Subjects with a medical history of heart disease or diabetes or with a current complaint of angina, and those with an abnormal fasting glucose level were excluded from the study. Also excluded were subjects with electrocardiographic evidence of ischemic heart disease and those taking antihypertensive or lipid-lowering medication. We were left with a final sample of 366 healthy men.The screening process, which has been described in detail elsewhere, 20 ' 21 included a detailed medical history and physical examination, chest x-ray and 12-lead electrocardiogram, and collection of blood for a number of biochemical and hematological measurements. Blood pressure was taken using a random-zero (Gelman Hawksley) sphygmomanometer, the value recorded being the mean of two measurements. Stage V end-point was used for the diastolic pressure. Weight (kg) and height (m) were recorded for each subject, and body mass index (BMI) was calculated as weight/height 2 . Cigarette smoking, alcohol consumption, and exercise level were assessed by a computerized questionnaire. Ex-smokers and those smoking only pipes or cigars were coded as non-cigarette smokers. Teetotalers (10 subjects) were omitted from the alcohol analysis. All procedures were approved by the ethical committee of BUPA Medical Research, and informed consent was obtained from each subject.The subjects fasted overnight for 14 hours.