In this study we examined the relationships between levels of several components of plasma lipoproteins and severity of coronary artery disease in 65 men and 42 women who underwent coronary arteriography for suspected coronary disease. Severity of coronary atherosclerosis was scored as the extent of disease seen at arteriography. Univariate analyses of the relationships between the plasma lipoprotein parameters and score for severity of atherosclerosis revealed a marked difference between men and women. In men, the score for severity of atherosclerosis was strongly related to the low-density lipoprotein (LDL) cholesterol and apolipoprotein B concentrations, whereas in women it was related to the triglyceride concentrations in plasma intermediate-density lipoprotein (IDL) and LDL and to the cholesterol and apolipoprotein B concentrations in IDL. The significance of these correlations was not negated by possible confounding factors such as alcohol intake, diabetes, and treatment with thiazides and ,-adrenergic blockers. Stepwise regression analyses of data adjusted for weight and age indicated that 22% of the variation in the score for severity of atherosclerosis could be accounted for by levels of LDL cholesterol in men. No other lipoprotein parameter could account for any further variation. In contrast, cholesterol did not account for any variation in the score for severity of atherosclerosis in women, whereas plasma triglyceride accounted for 16% of the observed variation in this group. No relationships were found between score for severity of atherosclerosis and highdensity lipoprotein cholesterol or plasma apolipoprotein A-I concentrations in either group. This study indicates the importance of the triglyceride-rich lipoproteins in the development of coronary atherosclerosis in women. Circulation 71, No. 5, 881-888, 1985. EPIDEMIOLOGIC STUDIES have clearly established that a high plasma cholesterol level is associated with an increased risk of coronary artery disease (CAD). This risk is associated primarily with the cholesterol level of the low-density lipoprotein (LDL) fraction." 2 In -contrast, the cholesterol level in the high-density lipoprotein (HDL) fraction is an important determinant of reduced risk.3 Of the lipid-related indexes, the ratio of plasma cholesterol to HDL cholesterol has been suggested to be the most powerful predictor of premature development of CAD.4 In 1978, a report from this laboratory was among the first to relate lipoprotein lipid concentrations to the severity of CAD, as determined from the extent of disease in eight major segments of the coronary circu-

Objective-To examine relationships between blood pressure during childhood and both placental weight and body size at birth, in an Australian population. Design-A follow up study of a birth cohort, undertaken when cohort members were aged 8 years. Setting-Adelaide, South Australia. Subjects-830 children born in the Queen Victoria Hospital in Adelaide, South Australia, during 1975-6. Main outcome measures-Systolic and diastolic blood pressure measured when the children were aged 8 years.Results-Blood pressure at 8 years was positively related to placental weight and inversely related to birth weight, after adjusting for the child's current weight.For diastolic pressure there was a decrease of 1.0 mm Hg for each 1 kg increase in birth weight (95% confidence interval (CI) = -0.4 to 2.4) and an increase of 0.7 mm Hg for each 100 g increase in placental weight (95% CI = 0.1 to 1.3). Diastolic pressure was also inversely related to chest circumference at birth, independently ofplacental weight, with a decrease of 0.3 mm Hg for each 1 cm increase in chest circumference (95% CI = 0.2 to 0.5). Conclusions-These findings are further evidence that birth characteristics, indicative of fetal growth patterns, are related to blood pressure in later life. (Arch Dis Child 1996;74:538-541)

During 1975-1977 twenty-nine males surviving acute myocardial infarction at an age between 40-44 years were registered in Gothenburg, Sweden. Twenty-five of these were studied and compared with two control groups. One group, the reference group (RG, n = 76), was randomly selected from the male population from which the acute myocardial infarction (AMI) group was derived. A second group, the matched control group (MC, n = 47), consisted of men with no history of coronary heart disease, matched with patients for age, serum cholesterol and body weight index. Serum triglyceride levels were higher and alphalipoprotein cholesterol lower in the AMI group than in RG. Prior to infarction, patients had a higher degree of physical activity at work and a higher tobacco consumption than RG. When AMI cases were compared with MC subjects lower alphalipoprotein cholesterol levels were found in AMI, and they also had a higher tobacco consumption prior to infarction. There was a negative correlation between alphalipoprotein cholesterol levels and tobacco consumption in the RG. The differences in alphalipoprotein cholesterol levels between AMI cases and controls could not attributed to smoking habits, but smoking may at least to some extent exert its effect as a risk factor through influence on alphalipoprotein cholesterol levels.

A prospective follow-up study of infants selected by cord blood total cholesterol (TC) and low-density-lipoprotein cholesterol (LDL-C) levels from 2000 consecutive live births was undertaken to reassess the role of cord blood screening in the diagnosis of familial hypercholesterolaemia (FH). Mean values for serum cholesterol were (mmol/l +/- S.D.): TC, 1.83 +/- 0.56; LDL-C, 0.90 +/- 0.49; HDL-C, 0.70 +/- 0.33; TG, 0.38 +/- 0.16. Seventy-three of 117 infants who had had a cord TC and/or LDL-C greater than 95th percentile, and 373 control group children (cord TC and/or LDL-C less than 95th percentile) were followed up at age 3--12 months. Six of the 117 were hypercholesterolaemic (HC), and one child had an HC parent: positive detection rate greater than or equal to 0.05%; false positive rate greater than or equal to 3.7%. Four control-group children were HC and had an HC parent; false negative rate greater than or equal to 1.1%. With the possible exception of detecting FH in a child with a known affected parent, cord blood screening appears to be unreliable for the diagnosis of FH.