- Email: [email protected]
High density lipoprotein cholesterol in male relatives of patients with coronary heart disease
269
32 (1979) 269-276 @ Elsevier/North-Holland Scientific Publishers, Ltd.
Atherosclerosis,
HIGH DENSITY LIPOPROTEIN CHOLESTEROL IN MALE RELATIVES OF PATIENTS WITH CORONARY HEART DISEASE
HORACE MICHELI, DANIEL POMETTA, CONSTANTIN JORNOT and JEAN-RAOUL SCHERRER Divisions de Diabktologie et d ‘Znformatique, Geneva, Geneva (Switzerland)
Department
of Medicine
of the Uniuersity
of
(Received 3 August, 1978) (Revised, received 25 September, 1978) (Accepted 26 September, 1978)
Summary To study factors that play a role in the familial occurrence of coronary heart disease, very low density lipoprotein (VLDL) triglycerides, low density lipoprotein (LDL) cholesterol and high density lipoprotein (HDL) cholesterol were measured after preparative ultracentrifugation in first degree male relatives of coronary patients and in control subjects. The HDL cholesterol concentration was significantly lower in relatives of 20-71 years old than in controls. No increase of serum and LDL cholesterol was found. A low level of HDL cholesterol was observed even in the younger relatives who are less likely to have cardiovascular disease. In older relatives low HDL cholesterol was found in the presence or absence of clinical evidence of coronary artery disease. The HDL-cholesterol concentration was inversely related to the VLDL triglycerides both in relatives and controls, but the regression lines were different (P < 0.001) for the relative (y = -0.166x + 0.43) and for the controls (y = -0.191x + 0.49). A low HDL cholesterol level appears to be a marker of relatives of coronary patients. Key words:
Coronary cholesterol
heart disease - Familial - VLDL triglycerides
occurrence
-
HDL
cholesterol
Supported by a grant (6.13000.74) of the Swiss National Science Foundation. Reprint request to Dr. H. Micheli. Division de Diab&Aogie, HBpital Cantonal. CH-1211 Switzerland.
-
LDL
Geneva 4,
270
Introduction Chance alone may produce impressive familial clusters of a common disease. However, in coronary heart disease familial clustering is too strong to be ascribed wholly to chance [ 11. A positive family history of coronary heart disease (CHD) should be recognized as a major risk factor. The influence on the occurrence of coronary heart disease of elevated serum cholesterol and of low density lipoprotein cholesterol (LDL cholesterol) as well as elevated serum triglycerides and very low density lipoprotein triglycerides (VLDL triglycerides) has been extensively studied. More recently interest has focused on high density lipoprotein (HDL) in CHD patients as they have been found to have low HDL cholesterol in comparison to normal subjects [Z-8]. A few studies have shown that a low HDL cholesterol content may have a predictive value for the further development of coronary disease [9-l 11. It was therefore of interest to determine if the lipoproteins of the close relatives of CHD patients differ from the lipoproteins of a healthy population and if alterations similar to those observed in CHD patients could be found. The purpose of our work was to determine lipoprotein cholesterol and triglyceride content, and particularly HDL cholesterol, in relatives of patients hospitalized for acute myocardial infarction, and to compare the results to those obtained in an apparently healthy population. This study will report the results obtained in males aged 20-71 years, who are first degree relatives of coronary patients. Material and Methods Ninety-three consecutive patients hospitalized for acute myocardial infarction in the Department of Medicine of the University of Geneva between January 1974 and December 1976 were admitted to this study. The diagnosis of acute myocardial infarction was assessed by at least two of the following criteria: clinical history, acute ECG changes, and creatine phosphokinase elevation. Patients were excluded from the study if they were 55 years old and over for the men and 70 years and over for the women, or if it was evident that fewer than two male or female first degree relatives might be examined. The difference in age selection for the men and the women was chosen because of the scarcity of myocardial infarction among women below the age of 55 years. First degree relatives (brothers, sons, fathers) were submitted during the following weeks to a medical examination. Blood samples were taken without venous stasis after an overnight fast. The control group was made up of 370 subjects randomly selected from 640 apparently healthy white- and blue-collar male civil servants of the City of Geneva. Subjects with clinical and electrocardiographic evidence of ischaemic heart disease, peripheral vascular disease or other chronic or acute disease were excluded [ 121. White collars represented 39% and blue collars 61% of the population. Of the 225 blue collars 55 were gardners, 50 firemen and 120 manual workers (jani-
271
tors, carpenters, repairmen, plumbers, etc.). Serum and lipoprotein cholesterol was determined by a modification of Abel1 and Kendall’s method [ 131. Saponification with NaOH was followed by nonane extraction. Of the nonane phase 0.3 ml was evaporated in less than 15 min with aeration at 80°C before the Liebermann-Burchard reaction. Triglycerides were measured according to the method described by Soloni [ 141. Automatic dilutors, samplers and pipettes were used. VLDL, LDL and HDL were separated after preparative ultracentrifugation in two steps at densities 1.006 and 1.063, based on the method of Havel, Eder and Bragdon [15]. The recovery was 93% for cholesterol and 92% for triglycerides. The day-today variations were up to 3.7% for cholesterol and 5.5% for triglycerides. During a day the variations were up to 2.6% for cholesterol and 2.5% for triglycerides. Chemical and ultracentrifugation procedures were standardized by sample exchanges between 4 European laboratories. The results of the quality controls have been published elsewhere [ 121. Statistical analysis of the differences about the means between two samples was performed using the Student t-test, the Mann-Whitney U-test and comparison of regression lines. Before calculation serum and VLDL triglyceride concentrations were transformed to their logarithms. Results Out of 93 families who were eligible for this study, 4 refused to participate because of lack of interest. For one additional family the death of the proband disallowed contact with the relatives. Thus 88 families entered the study (Table 1). Fifty-seven families were related to a male proband and 31 to a female proband. Altogether 159 male first degree relatives 20 years old and over were eligible. One hundred and three (65%) were examined (Table 2). Fifty-six (35%) were not. Out of the non-examined relatives 33 (21% of the eligible relatives) did not attend the medical examination because they lived outside the Geneva area, 4 (2.5%) had a disabling disease, 15 (9%) refused to participate because of lack of interest. For an unknown reason 4 (2.5%) relatives who had previously accepted did not come. The number of dead relatives differs in the 3 generations under consideration (Table 1). All the sons of the patients were alive; in contrast, only one of the 31 fathers of female probands and 15 of the 57 fathers of male probands were still alive. This high mortality is in relation to the age range of the probands. The mean serum cholesterol was not elevated in the first degree relatives of patients with myocardial infarction when compared to the healthy controls. In the age group 40-59 years, the mean serum cholesterol was even lower (P < 0.05) among the relatives of CHD patients than in the control group (Fig. 1). As far as the LDL cholesterol is concerned there was no difference in any age group between the relatives of CHD patients and the controls (Fig. 1). No difference in serum and VLDL triglycerides was observed between relatives and controls at age 40 years and over. In contrast serum and VLDL triglycerides were elevated in the 20-39-year-old relatives of CHD patients (P < 0.05) (Fig. 2).
272 TABLE
1
NUMBER
(AGE
OF MALE
20 AND
OVER)
FIRST
DEGREE
RELATIVES
OF MALE
AND FEMALE
PROBANDS Male probands
Relatives
(n = 5’7)
Female probands
(n = 31)
Age range 41-69 Mean age 61 yrs SD t 7
Yrs
AlI probands (n = 88)
Age range 31-54 Mean age 46 yrs SD f 6 SOtU examined not examined alive dead Bra them examined not examined alive dead
YIS
20
25
45
-
7 -
1 -
31
15
52
24 13
15 24
39 37
Fathers
examined not examined alive
TABLE
6
6 1 30
9 42
10 72
2
AGE DISTRIBUTION EXAMINED Number
OF THE
FIRST
DEGREE
MALE
RELATIVES
Age (YW
Relatives Controls
20-39
Fig. 1. Serum dial infarction
HEALTHY
CONTROLS
Total
20-39
40-59
60-71
48 158
33 163
22 49
LO-59
AND
EGO-71 years
103 370
20-39
, ,
and LDL cholesterol concentrations in male, first degree relatives and in control subjects. Means + SEM.
Fig. 2. Serum and VLDL triglyceride concentrations in male, cardial infarction and in control subjects. Means f SEM.
60- 71years
40-5s
of patients
first degree relatives
with myocar-
of patients
with myo-
273
cholesterol g/l 0.6
N+t
1 5 mmolll
I
I 25
l p
I
I
LO-59
20-39
I
60-71 years
Fig. 3. HDL cholesterol concentrations in male. tion and in control subjects. Means 2 SEM.
first degree
relatives
of patients
with myocardial
infarc-
HDL cholesterol mmol/l
0.1
0.3
1 g/l
3
I
VLDL triglycerides
IO
mmol/l
Fig. 4. Regression lines with their 95% confidence limits for ordinate triglycerides in first degree relatives and in controls. For the VLDL used.
of HDL cholesterol versus VLDL triglycerides a logarithmic scale is
The HDL cholesterol was significantly lower among the relatives of CHD patients than among the controls in any age group considered (Fig. 3). The regression lines of HDL cholesterol versus VLDL triglycerides are y = (-0.16605 f 0.02812)x + (0.43023 f 0.0107) for the relatives, and y = (-9.19126 + 0.01786)x + (0.49184 f 0.00723) for the controls (P < 0.001) (Fig. 4). Discussion The results show that the male first degree relatives of CHD patients differ from the healthy controls by a low HDL cholesterol concentration. In the younger age group (20-39 years) an increase in VLDL and serum triglycerides has also been observed. In contrast, the LDL cholesterol, which is thought to be particularly atherogenic [ 161, was not increased in the relatives of the CHD
274
patients and no increase in the serum cholesterol was observed. Even in the age group 40-59 years, the low HDL cholesterol added to LDL and VLDL cholesterol in normal ranges resulted in a lower total serum cholesterol (P< 0.05) among the CHD patients’ relatives than among the controls. HDL has been thought to play a role in cholesterol esterification [ 171 and to favour the removal of cholesterol from the tissues [l&19]. The low HDL cholesterol measured in families of CHD patients is therefore of great interest and could be the expression of a familial defect of cholesterol exchange. HDL cholesterol concentrations measured in our control population, as well as in 3 other apparently healthy European populations studied simultaneously 1121, have shown that the HDL cholesterol did not differ between these 4 different populations nor within the populations in the different age groups, in contrast to LDL cholesterol and VLDL triglycerides. As standardized methods were used, the HDL cholesterol concentration appears to be a more constant parameter in different populations than the VLDL triglycerides or LDL cholesterol. Therefore the low HDL cholesterol observed in the adult male relatives of CHD patients appears to be an important observation, independent of age. It characterizes the family of CHD patients even at a lower age when cardiovascular disease is usually not apparent. and VLDL triglyAs a high inverse correlation between HDL cholesterol cerides has been shown [20,21] the low HDL cholesterol among the young relatives (20-39 years) could be at least partly related to the increased VLDL triglyceride concentration. However, this explanation is not valid for relatives 40 years old and over who had a low HDL cholesterol concentration while their VLDL triglyceride concentrations did not differ from the controls. In order to analyse further the relationship between the VLDL triglycerides and the HDL cholesterol, regression analysis of the logarithms of VLDL triglycerides versus the HDL cholesterol was performed for the controls and for the relatives, irrespective of their age. The regression analysis confirms the negative correlation between VLDL triglycerides and HDL cholesterol with a correlation coefficient of -0.51 (P< 0.001) for the relatives and of -0.49 for the controls (P< 0.001). As shown in Fig. 4 the slope is -6.166 for the relatives of CHD patients and -0.199 for the controls. The intercept has been calculated for VLDL triglycerides at 1.0 mmole/l. As shown in the figure the intercept is at 0.49 g/l HDL cholesterol for the controls and at 0.43 g/l for the relatives of patients with myocardial infarction; a significant difference (P< 0.001). Therefore the regression lines and their envelopes are different for the controls and the relatives of CHD patients. As shown in the figure, for VLDL triglycerides below 3.1 mmoles/l the relatives of CHD patients have lower HDL cholesterol than the controls: the two populations are different. It is only when the VLDL triglycerides are at 3.1 mmoles/l or above that the envelopes of the regression lines overlap and that HDL cholesterol is indistinctly low both in relatives and controls. Changes in lipoprotein lipids can be induced by diet. Carlson et al. [22] have shown in both type IV and type V hyperlipoproteinemia that when the VLDL is drastically reduced within 5 days by diet and drugs there is no simultaneous
215
increase in HDL cholesterol. None of the relatives of the CHD patients had been on lipid-lowering drugs and we have no evidence of changes in dietary habits among the relatives of CHD patients during the days prior to the medical examination. Therefore the low HDL cholesterol of the CHD relatives cannot be explained by recent dietary changes. Because of clustering of coronary disease in families one could postulate that the low HDL cholesterol concentrations observed were not a familial preexisting trait but rather were associated to the presence of coronary atherosclerotic lesions, as coronary patients have been shown to have low HDL cholesterol. Among the 103 relatives, 10 had a clinical history of angina pectoris or evidence of old myocardial infarction. All of them were 40-66 years old. No difference could be found between their HDL cholesterol (0.48 + 0.04 SEM g/l) and the HDL cholesterol of the 29 relatives of the same age range who showed no evidence of cardiovascular disease (0.47 + 0.04 g/l). Therefore low HDL cholesterol does not appear to be present only in the subjects with clinical manifestations of coronary atherosclerosis, but to be characteristic of all groups of relatives. Furthermore, when one considers the 29 younger relatives aged 20-29, who were less likely to have extensive coronary disease, their LDL cholesterol and VLDL triglycerides were not different from the controls. In contrast, their HDL cholesterol was already significantly lower than the HDL cholesterol of the controls. Low HDL cholesterol appears to be an early characteristic of relatives of CHD patients which may precede for a long period the occurrence of coronary disease. Whether the factor responsible for low HDL cholesterol is related to environment or to a genetic condition needs to be clarified. Very little is known yet about factors influencing HDL cholesterol. References 1 Rissanen, A.M. and NikkiM. E.C., Coronary artery disease and its risk factors in families of young men with angina pectoris and in controls, Brit. Heart J., 39 (1977) 875-883. 2 Barr, D.P., Russ, E.M. and Eder, H.A.. Protein-lipid relationship in human Plasma, Part 2 (In atherosclerosis and related conditions), Amer. J. Med., 11 (1951) 480493. 3 NikkiM, E., Studies in the lipid--protein relationship in normal and pathologic sera and the effect of heparin on serum lipoproteins, Stand. J. Clin. Lab. Invest., 5 (Suppl. 8) (1953) l-101. 4 Lewis, B., Chait. A., Oakley, C.M.O. et al., Serum lipoprotein abnormalities in patients with ischaemic heart disease - Comparisons with a control population, Brit. Med. J.. 3 (1974) 489493. 5 Micheli. H.. Pometta, D. and Bloch. A., Cholesterol et triglycerides isolees Par ultracentrifugation -Relation avec l’atherosclerose coronarienne, Schweiz. Med. Wschr., 104 (1974) 1794-1796. 6 Carlson, L.A. and Ericsson, M.. Quantitative and qualitative serum lipoprotein analysis. Part 2 (Studies in male survivors of myocardial infarction), Atherosclerosis, 21 (1975) 417433. 7 Wiklund. 0.. Gustafson. A. and Wilhelmsen, L., Alpha-lipoprotein cholesterol in men after myocardial infarction compared with a population sample, Artery, 1 (1975) 399-405. 8 Rhoads. G.G., Gulbransen. C.L. and Kagan. A., Serum lipoproteins and coronary heart disease in a population study of Hawaii Japanese men, New Engl. J. Med., 294 (1976) 293-298. 9 Medalie. J.H., Kahn, H.A.. Neufeld. H.N. et al., Five-year myocardial infarction incidence, Part 2 (Association of single variable to age and birth place), J. Chron. Dis., 26 (1973) 329-349. 10 Gordon, T., Castelli. W.P., Hjortland, M.C. et al., High density lipoprotein as a protective factor against coronary heart disease -The Framingham Study, Amer. J. Med., 62 (1977) 707-713. 11 Miller, N.E., Forde. Q.H.. Thelle. S.S. et al., The TromsS heart study - High-density lipoprotein and coronary heart-disease. A prospective case-control study, Lancet. 1 (1977) 965-968. 12 Lewis, B.. Chait, A., Sigurdson, G. et al., Serum lipoproteins in four European communities - A quantitative comparison, Europ. J. Clin. Invest., 8 (1978) 165-173.
276 13 Abell, L.L., Levy, B.B., Brodie, B.B. et al., Simplified method for the estimation of total cholesterol ln serum and demonstration of its specificity, J. Biol. Chem.. 195 (1952) 367-366. 14 Solon& F.G.. Simplified manual micro-method for determination of serum triglycerides, Clin. Chem.. 17 (1971) 529-534. 15 Carlson, K.. Lipoprotein fractionation, J. Clin. Path., 26 (Suppl. 5) (1973) 32-37. 16 Fredrickson, D.S. and Levy, RI., Familial hypercholesterolemia. In: J.B. Stanbuay, J.B. Wyngaarden and D.S. Fredrickson (Eds.), The Metabolic Basis of Inherited Disease, 3rd edition, McGraw-Hill, New York, NY, 1972.545-614. 17 Norum, K.R., Glomset. J.A.. Nichols, A.V. et al.. Plasma lipoproteins in familial ledtbln-cholesterol acyltranferase deficiency -Effects of incubation with 1ecitbln:cholesterol acyltransferase in vitro, Stand. J. Clln. Lab. Invest., 36 (1975) 31-55. 18 Stem, Y., Glangeaud. M.C.. Fainary, M. et al.. The removal of cholesterol from aortic smooth muscle cells in culture and Landschtlta ascites cells by fractions of human high density lipoproteins, Biochim. Biophys. Acta. 380 (1975) 106-118. 19 Bondjers, G., Gustafson, A., Kral, J. et al,, Cholesterol content in arterial tissue in relation to serum lipoproteins in man, Artery, 2 (1976) 200-207. 20 Wilson, D.E. and Lees, R.J., Metabolic relationships among the plasma lipoproteins -Reciprocal changes in the concentrations of very low and low density lipoproteins in men, J. Clin. Invest., 51 (1972) 1051. 21 Miller, G.J. and Miller, N.E., Plasma high-density lipoprotein concentration and development of ischaemic heart-disease, Lancet, l(l975) 16-19. 22 Carlson. L.A., Olsson, A.G. and Ballantyne. D.. On the rise in low density and high density llpoproteins ln response to the treatment of hypertriglyceridemia in type IV and type V hyperlipoproteinemias, Atherosclerosis, 26 (1977) 603-609.
32 (1979) 269-276 @ Elsevier/North-Holland Scientific Publishers, Ltd.
Atherosclerosis,
HIGH DENSITY LIPOPROTEIN CHOLESTEROL IN MALE RELATIVES OF PATIENTS WITH CORONARY HEART DISEASE
HORACE MICHELI, DANIEL POMETTA, CONSTANTIN JORNOT and JEAN-RAOUL SCHERRER Divisions de Diabktologie et d ‘Znformatique, Geneva, Geneva (Switzerland)
Department
of Medicine
of the Uniuersity
of
(Received 3 August, 1978) (Revised, received 25 September, 1978) (Accepted 26 September, 1978)
Summary To study factors that play a role in the familial occurrence of coronary heart disease, very low density lipoprotein (VLDL) triglycerides, low density lipoprotein (LDL) cholesterol and high density lipoprotein (HDL) cholesterol were measured after preparative ultracentrifugation in first degree male relatives of coronary patients and in control subjects. The HDL cholesterol concentration was significantly lower in relatives of 20-71 years old than in controls. No increase of serum and LDL cholesterol was found. A low level of HDL cholesterol was observed even in the younger relatives who are less likely to have cardiovascular disease. In older relatives low HDL cholesterol was found in the presence or absence of clinical evidence of coronary artery disease. The HDL-cholesterol concentration was inversely related to the VLDL triglycerides both in relatives and controls, but the regression lines were different (P < 0.001) for the relative (y = -0.166x + 0.43) and for the controls (y = -0.191x + 0.49). A low HDL cholesterol level appears to be a marker of relatives of coronary patients. Key words:
Coronary cholesterol
heart disease - Familial - VLDL triglycerides
occurrence
-
HDL
cholesterol
Supported by a grant (6.13000.74) of the Swiss National Science Foundation. Reprint request to Dr. H. Micheli. Division de Diab&Aogie, HBpital Cantonal. CH-1211 Switzerland.
-
LDL
Geneva 4,
270
Introduction Chance alone may produce impressive familial clusters of a common disease. However, in coronary heart disease familial clustering is too strong to be ascribed wholly to chance [ 11. A positive family history of coronary heart disease (CHD) should be recognized as a major risk factor. The influence on the occurrence of coronary heart disease of elevated serum cholesterol and of low density lipoprotein cholesterol (LDL cholesterol) as well as elevated serum triglycerides and very low density lipoprotein triglycerides (VLDL triglycerides) has been extensively studied. More recently interest has focused on high density lipoprotein (HDL) in CHD patients as they have been found to have low HDL cholesterol in comparison to normal subjects [Z-8]. A few studies have shown that a low HDL cholesterol content may have a predictive value for the further development of coronary disease [9-l 11. It was therefore of interest to determine if the lipoproteins of the close relatives of CHD patients differ from the lipoproteins of a healthy population and if alterations similar to those observed in CHD patients could be found. The purpose of our work was to determine lipoprotein cholesterol and triglyceride content, and particularly HDL cholesterol, in relatives of patients hospitalized for acute myocardial infarction, and to compare the results to those obtained in an apparently healthy population. This study will report the results obtained in males aged 20-71 years, who are first degree relatives of coronary patients. Material and Methods Ninety-three consecutive patients hospitalized for acute myocardial infarction in the Department of Medicine of the University of Geneva between January 1974 and December 1976 were admitted to this study. The diagnosis of acute myocardial infarction was assessed by at least two of the following criteria: clinical history, acute ECG changes, and creatine phosphokinase elevation. Patients were excluded from the study if they were 55 years old and over for the men and 70 years and over for the women, or if it was evident that fewer than two male or female first degree relatives might be examined. The difference in age selection for the men and the women was chosen because of the scarcity of myocardial infarction among women below the age of 55 years. First degree relatives (brothers, sons, fathers) were submitted during the following weeks to a medical examination. Blood samples were taken without venous stasis after an overnight fast. The control group was made up of 370 subjects randomly selected from 640 apparently healthy white- and blue-collar male civil servants of the City of Geneva. Subjects with clinical and electrocardiographic evidence of ischaemic heart disease, peripheral vascular disease or other chronic or acute disease were excluded [ 121. White collars represented 39% and blue collars 61% of the population. Of the 225 blue collars 55 were gardners, 50 firemen and 120 manual workers (jani-
271
tors, carpenters, repairmen, plumbers, etc.). Serum and lipoprotein cholesterol was determined by a modification of Abel1 and Kendall’s method [ 131. Saponification with NaOH was followed by nonane extraction. Of the nonane phase 0.3 ml was evaporated in less than 15 min with aeration at 80°C before the Liebermann-Burchard reaction. Triglycerides were measured according to the method described by Soloni [ 141. Automatic dilutors, samplers and pipettes were used. VLDL, LDL and HDL were separated after preparative ultracentrifugation in two steps at densities 1.006 and 1.063, based on the method of Havel, Eder and Bragdon [15]. The recovery was 93% for cholesterol and 92% for triglycerides. The day-today variations were up to 3.7% for cholesterol and 5.5% for triglycerides. During a day the variations were up to 2.6% for cholesterol and 2.5% for triglycerides. Chemical and ultracentrifugation procedures were standardized by sample exchanges between 4 European laboratories. The results of the quality controls have been published elsewhere [ 121. Statistical analysis of the differences about the means between two samples was performed using the Student t-test, the Mann-Whitney U-test and comparison of regression lines. Before calculation serum and VLDL triglyceride concentrations were transformed to their logarithms. Results Out of 93 families who were eligible for this study, 4 refused to participate because of lack of interest. For one additional family the death of the proband disallowed contact with the relatives. Thus 88 families entered the study (Table 1). Fifty-seven families were related to a male proband and 31 to a female proband. Altogether 159 male first degree relatives 20 years old and over were eligible. One hundred and three (65%) were examined (Table 2). Fifty-six (35%) were not. Out of the non-examined relatives 33 (21% of the eligible relatives) did not attend the medical examination because they lived outside the Geneva area, 4 (2.5%) had a disabling disease, 15 (9%) refused to participate because of lack of interest. For an unknown reason 4 (2.5%) relatives who had previously accepted did not come. The number of dead relatives differs in the 3 generations under consideration (Table 1). All the sons of the patients were alive; in contrast, only one of the 31 fathers of female probands and 15 of the 57 fathers of male probands were still alive. This high mortality is in relation to the age range of the probands. The mean serum cholesterol was not elevated in the first degree relatives of patients with myocardial infarction when compared to the healthy controls. In the age group 40-59 years, the mean serum cholesterol was even lower (P < 0.05) among the relatives of CHD patients than in the control group (Fig. 1). As far as the LDL cholesterol is concerned there was no difference in any age group between the relatives of CHD patients and the controls (Fig. 1). No difference in serum and VLDL triglycerides was observed between relatives and controls at age 40 years and over. In contrast serum and VLDL triglycerides were elevated in the 20-39-year-old relatives of CHD patients (P < 0.05) (Fig. 2).
272 TABLE
1
NUMBER
(AGE
OF MALE
20 AND
OVER)
FIRST
DEGREE
RELATIVES
OF MALE
AND FEMALE
PROBANDS Male probands
Relatives
(n = 5’7)
Female probands
(n = 31)
Age range 41-69 Mean age 61 yrs SD t 7
Yrs
AlI probands (n = 88)
Age range 31-54 Mean age 46 yrs SD f 6 SOtU examined not examined alive dead Bra them examined not examined alive dead
YIS
20
25
45
-
7 -
1 -
31
15
52
24 13
15 24
39 37
Fathers
examined not examined alive
TABLE
6
6 1 30
9 42
10 72
2
AGE DISTRIBUTION EXAMINED Number
OF THE
FIRST
DEGREE
MALE
RELATIVES
Age (YW
Relatives Controls
20-39
Fig. 1. Serum dial infarction
HEALTHY
CONTROLS
Total
20-39
40-59
60-71
48 158
33 163
22 49
LO-59
AND
EGO-71 years
103 370
20-39
, ,
and LDL cholesterol concentrations in male, first degree relatives and in control subjects. Means + SEM.
Fig. 2. Serum and VLDL triglyceride concentrations in male, cardial infarction and in control subjects. Means f SEM.
60- 71years
40-5s
of patients
first degree relatives
with myocar-
of patients
with myo-
273
cholesterol g/l 0.6
N+t
1 5 mmolll
I
I 25
l p
I
I
LO-59
20-39
I
60-71 years
Fig. 3. HDL cholesterol concentrations in male. tion and in control subjects. Means 2 SEM.
first degree
relatives
of patients
with myocardial
infarc-
HDL cholesterol mmol/l
0.1
0.3
1 g/l
3
I
VLDL triglycerides
IO
mmol/l
Fig. 4. Regression lines with their 95% confidence limits for ordinate triglycerides in first degree relatives and in controls. For the VLDL used.
of HDL cholesterol versus VLDL triglycerides a logarithmic scale is
The HDL cholesterol was significantly lower among the relatives of CHD patients than among the controls in any age group considered (Fig. 3). The regression lines of HDL cholesterol versus VLDL triglycerides are y = (-0.16605 f 0.02812)x + (0.43023 f 0.0107) for the relatives, and y = (-9.19126 + 0.01786)x + (0.49184 f 0.00723) for the controls (P < 0.001) (Fig. 4). Discussion The results show that the male first degree relatives of CHD patients differ from the healthy controls by a low HDL cholesterol concentration. In the younger age group (20-39 years) an increase in VLDL and serum triglycerides has also been observed. In contrast, the LDL cholesterol, which is thought to be particularly atherogenic [ 161, was not increased in the relatives of the CHD
274
patients and no increase in the serum cholesterol was observed. Even in the age group 40-59 years, the low HDL cholesterol added to LDL and VLDL cholesterol in normal ranges resulted in a lower total serum cholesterol (P< 0.05) among the CHD patients’ relatives than among the controls. HDL has been thought to play a role in cholesterol esterification [ 171 and to favour the removal of cholesterol from the tissues [l&19]. The low HDL cholesterol measured in families of CHD patients is therefore of great interest and could be the expression of a familial defect of cholesterol exchange. HDL cholesterol concentrations measured in our control population, as well as in 3 other apparently healthy European populations studied simultaneously 1121, have shown that the HDL cholesterol did not differ between these 4 different populations nor within the populations in the different age groups, in contrast to LDL cholesterol and VLDL triglycerides. As standardized methods were used, the HDL cholesterol concentration appears to be a more constant parameter in different populations than the VLDL triglycerides or LDL cholesterol. Therefore the low HDL cholesterol observed in the adult male relatives of CHD patients appears to be an important observation, independent of age. It characterizes the family of CHD patients even at a lower age when cardiovascular disease is usually not apparent. and VLDL triglyAs a high inverse correlation between HDL cholesterol cerides has been shown [20,21] the low HDL cholesterol among the young relatives (20-39 years) could be at least partly related to the increased VLDL triglyceride concentration. However, this explanation is not valid for relatives 40 years old and over who had a low HDL cholesterol concentration while their VLDL triglyceride concentrations did not differ from the controls. In order to analyse further the relationship between the VLDL triglycerides and the HDL cholesterol, regression analysis of the logarithms of VLDL triglycerides versus the HDL cholesterol was performed for the controls and for the relatives, irrespective of their age. The regression analysis confirms the negative correlation between VLDL triglycerides and HDL cholesterol with a correlation coefficient of -0.51 (P< 0.001) for the relatives and of -0.49 for the controls (P< 0.001). As shown in Fig. 4 the slope is -6.166 for the relatives of CHD patients and -0.199 for the controls. The intercept has been calculated for VLDL triglycerides at 1.0 mmole/l. As shown in the figure the intercept is at 0.49 g/l HDL cholesterol for the controls and at 0.43 g/l for the relatives of patients with myocardial infarction; a significant difference (P< 0.001). Therefore the regression lines and their envelopes are different for the controls and the relatives of CHD patients. As shown in the figure, for VLDL triglycerides below 3.1 mmoles/l the relatives of CHD patients have lower HDL cholesterol than the controls: the two populations are different. It is only when the VLDL triglycerides are at 3.1 mmoles/l or above that the envelopes of the regression lines overlap and that HDL cholesterol is indistinctly low both in relatives and controls. Changes in lipoprotein lipids can be induced by diet. Carlson et al. [22] have shown in both type IV and type V hyperlipoproteinemia that when the VLDL is drastically reduced within 5 days by diet and drugs there is no simultaneous
215
increase in HDL cholesterol. None of the relatives of the CHD patients had been on lipid-lowering drugs and we have no evidence of changes in dietary habits among the relatives of CHD patients during the days prior to the medical examination. Therefore the low HDL cholesterol of the CHD relatives cannot be explained by recent dietary changes. Because of clustering of coronary disease in families one could postulate that the low HDL cholesterol concentrations observed were not a familial preexisting trait but rather were associated to the presence of coronary atherosclerotic lesions, as coronary patients have been shown to have low HDL cholesterol. Among the 103 relatives, 10 had a clinical history of angina pectoris or evidence of old myocardial infarction. All of them were 40-66 years old. No difference could be found between their HDL cholesterol (0.48 + 0.04 SEM g/l) and the HDL cholesterol of the 29 relatives of the same age range who showed no evidence of cardiovascular disease (0.47 + 0.04 g/l). Therefore low HDL cholesterol does not appear to be present only in the subjects with clinical manifestations of coronary atherosclerosis, but to be characteristic of all groups of relatives. Furthermore, when one considers the 29 younger relatives aged 20-29, who were less likely to have extensive coronary disease, their LDL cholesterol and VLDL triglycerides were not different from the controls. In contrast, their HDL cholesterol was already significantly lower than the HDL cholesterol of the controls. Low HDL cholesterol appears to be an early characteristic of relatives of CHD patients which may precede for a long period the occurrence of coronary disease. Whether the factor responsible for low HDL cholesterol is related to environment or to a genetic condition needs to be clarified. Very little is known yet about factors influencing HDL cholesterol. References 1 Rissanen, A.M. and NikkiM. E.C., Coronary artery disease and its risk factors in families of young men with angina pectoris and in controls, Brit. Heart J., 39 (1977) 875-883. 2 Barr, D.P., Russ, E.M. and Eder, H.A.. Protein-lipid relationship in human Plasma, Part 2 (In atherosclerosis and related conditions), Amer. J. Med., 11 (1951) 480493. 3 NikkiM, E., Studies in the lipid--protein relationship in normal and pathologic sera and the effect of heparin on serum lipoproteins, Stand. J. Clin. Lab. Invest., 5 (Suppl. 8) (1953) l-101. 4 Lewis, B., Chait. A., Oakley, C.M.O. et al., Serum lipoprotein abnormalities in patients with ischaemic heart disease - Comparisons with a control population, Brit. Med. J.. 3 (1974) 489493. 5 Micheli. H.. Pometta, D. and Bloch. A., Cholesterol et triglycerides isolees Par ultracentrifugation -Relation avec l’atherosclerose coronarienne, Schweiz. Med. Wschr., 104 (1974) 1794-1796. 6 Carlson, L.A. and Ericsson, M.. Quantitative and qualitative serum lipoprotein analysis. Part 2 (Studies in male survivors of myocardial infarction), Atherosclerosis, 21 (1975) 417433. 7 Wiklund. 0.. Gustafson. A. and Wilhelmsen, L., Alpha-lipoprotein cholesterol in men after myocardial infarction compared with a population sample, Artery, 1 (1975) 399-405. 8 Rhoads. G.G., Gulbransen. C.L. and Kagan. A., Serum lipoproteins and coronary heart disease in a population study of Hawaii Japanese men, New Engl. J. Med., 294 (1976) 293-298. 9 Medalie. J.H., Kahn, H.A.. Neufeld. H.N. et al., Five-year myocardial infarction incidence, Part 2 (Association of single variable to age and birth place), J. Chron. Dis., 26 (1973) 329-349. 10 Gordon, T., Castelli. W.P., Hjortland, M.C. et al., High density lipoprotein as a protective factor against coronary heart disease -The Framingham Study, Amer. J. Med., 62 (1977) 707-713. 11 Miller, N.E., Forde. Q.H.. Thelle. S.S. et al., The TromsS heart study - High-density lipoprotein and coronary heart-disease. A prospective case-control study, Lancet. 1 (1977) 965-968. 12 Lewis, B.. Chait, A., Sigurdson, G. et al., Serum lipoproteins in four European communities - A quantitative comparison, Europ. J. Clin. Invest., 8 (1978) 165-173.
276 13 Abell, L.L., Levy, B.B., Brodie, B.B. et al., Simplified method for the estimation of total cholesterol ln serum and demonstration of its specificity, J. Biol. Chem.. 195 (1952) 367-366. 14 Solon& F.G.. Simplified manual micro-method for determination of serum triglycerides, Clin. Chem.. 17 (1971) 529-534. 15 Carlson, K.. Lipoprotein fractionation, J. Clin. Path., 26 (Suppl. 5) (1973) 32-37. 16 Fredrickson, D.S. and Levy, RI., Familial hypercholesterolemia. In: J.B. Stanbuay, J.B. Wyngaarden and D.S. Fredrickson (Eds.), The Metabolic Basis of Inherited Disease, 3rd edition, McGraw-Hill, New York, NY, 1972.545-614. 17 Norum, K.R., Glomset. J.A.. Nichols, A.V. et al.. Plasma lipoproteins in familial ledtbln-cholesterol acyltranferase deficiency -Effects of incubation with 1ecitbln:cholesterol acyltransferase in vitro, Stand. J. Clln. Lab. Invest., 36 (1975) 31-55. 18 Stem, Y., Glangeaud. M.C.. Fainary, M. et al.. The removal of cholesterol from aortic smooth muscle cells in culture and Landschtlta ascites cells by fractions of human high density lipoproteins, Biochim. Biophys. Acta. 380 (1975) 106-118. 19 Bondjers, G., Gustafson, A., Kral, J. et al,, Cholesterol content in arterial tissue in relation to serum lipoproteins in man, Artery, 2 (1976) 200-207. 20 Wilson, D.E. and Lees, R.J., Metabolic relationships among the plasma lipoproteins -Reciprocal changes in the concentrations of very low and low density lipoproteins in men, J. Clin. Invest., 51 (1972) 1051. 21 Miller, G.J. and Miller, N.E., Plasma high-density lipoprotein concentration and development of ischaemic heart-disease, Lancet, l(l975) 16-19. 22 Carlson. L.A., Olsson, A.G. and Ballantyne. D.. On the rise in low density and high density llpoproteins ln response to the treatment of hypertriglyceridemia in type IV and type V hyperlipoproteinemias, Atherosclerosis, 26 (1977) 603-609.