Article Text


Genetic and environmental variation in serum lipoproteins in relation to coronary heart disease
  1. Forbes W. Robertson,
  2. Alastair M. Cumming
  1. Department of Genetics, University of Aberdeen, Aberdeen


    Variation in the serum concentration of VLDL, IDL, LDL, and HDL was studied in (a) 192 survivors of myocardial infarction under the age of 50 living in the north-east of Scotland; (b) 250 relatives, mostly first degree; and (c) 259 unrelated individuals, comprising mostly spouses and their relatives.

    The biochemical characterisation of the lipids, which were separated by preparative ultracentrifugation, included (a) determination of total serum cholesterol and triglyceride; and (b) determination of the content of cholesterol in the four fractions, and also of triglyceride in the VLDL fraction.

    In male survivors of infarction there is a significant negative regression of VLDL and a significant positive regression of HDL on peak aspartate aminotransferase activity in blood samples taken within 24 hours of the incident. In later samples correlations were not evident.

    Serum lipoprotein concentrations have been compared in samples taken at different times after the incident. The average value of samples taken within 24 hours is similar to the mean scores for LDL and VLDL of repeat samples taken, on average, 9 months later.

    The incidence of hyperlipidaemia in samples taken within 24 hours of the infarction is about the same for total cholesterol and triglyceride and LDL or VLDL. Respectively, 18·3%, 16·1%, and 7·5% of the survivors exceed the 90th centile value for either LDL, VLDL, or both fractions, while, for the 95th centile limit, the corresponding figures are 8·6%, 9·7%, and 2·2%.

    In samples taken within 24 hours of the incident, the LDL concentration is, on average, 20 to 30% higher, and the VLDL concentration 50 to 60% higher, than the controls. The corresponding HDL difference in males, though negative, is trivial and statistically insignificant, but in females the difference is greater and significantly so.

    VLDL levels are substantially increased in regular cigarette smokers in both sexes, especially in males. HDL levels tend to be lower, though not significantly so. The frequency of persons who are or have been regular cigarette smokers is higher in index cases than in controls.

    Correlation analysis of individual variation of controls and index case relatives indicates a substantial level of independence between the major lipoprotein fractions. There is no correlation between LDL and HDL. There are positive correlations between VLDL, IDL, and LDL, though when IDL is held constant by multiple regression, the correlation between VLDL and LDL is removed. There is a low but consistent negative correlation between VLDL and HDL cholesterol.

    About a quarter of the variance of VLDL in males is accounted for by multiple regression on measures of body fatness, that is, relative weight and subscapular skinfold thickness. In females only about 10% of the variance is thus accounted for, and only in males do LDL levels show correlated changes with fatness. There are no significant differences for either relative body weight or subscapular skinfold thickness between the means of first degree relatives of survivors of infarction and controls.

    The concentrations of VLDL and LDL in first degree relatives of survivors of infarction are significantly higher than the control means. HDL shows no significant difference.

    The evidence for genetic variation in serum lipoprotein is based on the polygenic model and the analysis of the correlation between parents and offspring and the correlation between sibs. From the pooled regression on single parents, heritabilities for HDL, LDL, and VLDL work out at 0·67 ± 0·21, 0·36 ± 0·18, and 0·23 ± 0.20, respectively. The lower heritability of VLDL is consistent with the variance of repeat measurements on the same control individuals, which is much higher for VLDL than for LDL or HDL. The high correlation between VLDL and measures of fatness, for which heritability estimates are statistically insignificant, as well as the association between raised VLDL concentration and smoking, also provide confirmatory evidence of the major importance of non-genetic causes in the variation of VLDL.

    The results are discussed in relation to the origin of the effects of smoking, variation in proneness to coronary disease, and the biological significance of differences in HDL concentration.

    Statistics from

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.