Atherosclerosis,
21 (1975) 417433
0 Elsevier Scientific Publishing Company, Amsterdam
QUANTITATIVE PART
AND
1. STUDIES
QUALITATIVE
IN HEALTHY
LARS A. CARLSON King Gustaf V Research Uppsala (Sweden)
AND
MARTIN
Institute,
417 - Printed in The Netherlands
SERUM
LIPOPROTEIN
ANALYSIS
MEN AND WOMEN
ERICSSON
Stockholm,
and Department
of Geriatrics,
University
of Uppsala,
(Received October 21st, 1974) (Accepted November 28th, 1974)
SUMMARY
Preparative ultracentrifugal and electrophoretic was performed in 30-70-years-old healthy, fasting
analysis of serum lipoproteins males (N = 80) and females
(N = 77), randomly selected from the Uppsala region, Sweden. The concentrations of cholesterol and triglycerides in total serum and in VLDL, LDL and HDL lipoproteinclassesare reported. Total serum, VLDL and LDL triglycerides and cholesterol concentrations increased with age, while HDL cholesterol and triglyceride concentrations did not vary with age. Overweight persons had higher total serum triglyceride, higher
VLDL
cholesterol
and triglyceride
and lower
HDL
cholesterol
levels. The
upper 90 % population limit values for non-overweight males/females were: total triglycerides (mmol/l) 2.5/2.0, total cholesterol (mg/lOO ml) 298/300, VLDL triglyceride 1.80/1.05, VLDL-cholesterol 32/33, LDL triglyceride 0.69/0.69, LDL cholesterol 210/218, HDL triglyceride 0.32/0.34 and HDL-cholesterol 69/93. The 2 major differences between males and females were that females had lower VLDL but higher HDL concentrations. For VLDL there was a very strong and for LDL a moderately strong positive correlation between cholesterol and triglyceride contents. In HDL however, the measured amounts of cholesterol and triglycerides did not correlate at all. Sinking pre-b lipoprotein was found in about 25% of cases and a second pre-/I band floating at d 1.006, late pre-/I, was found in 35% of male and 25% of female subjects. Subjects with sinking pre-j3 lipoprotein did not differ from other subjects with regard to the concentration of cholesterol and triglycerides in the 3 lipoprotein classes. Males, but not females, with the late pre-/3 (LPB), had an increased amount of cholesterol in VLDL and a raised cholesterol-triglyceride ratio
Supported
by Grants from the Swedish Medical Research Council.
418
L. A. CARLSON,
in this lipoprotein
class. Also the LDL triglyceride
level was increased
M. ERICSSON
in males with
the late pre-/3 lipoprotein.
Key words:
Agarose gel electrophoresis - Healthy men and women - Late pre-P lipoprotein - Preparative ultracentrifugation - Serum lipoproteins Sinking pre-/3 lipoprotein
INTRODUCTION
There is a close association between ischaemic cardiovascular diseases and hyperlipidaemia. Various forms of hyperlipidaemia such as pure hypertriglyceridaemia, pure hypercholesterolaemia and combined hyperlipidaemia (elevation of both triglyceride and cholesterol concentrations) are frequently found in patients with ischaemic heart disease (IHD) 1,s. The serum lipids do not, however, circulate freely in the blood but are bound to the (VLDL), the low (LDL) and the contain cholesterol and triglycerides Thus hypercholesterolaemia as well
serum lipoproteins (LP), high (HDL) density LP and their concentrations as hypertriglyceridaemia
mainly to the very low classes. These LP’s all may vary independently. may occur as a result of
elevation of levels in any of the LP classes, either alone or in combination. To detect the LP abnormality underlying hyperlipidaemia it is necessary to measure the concentration of cholesterol and triglycerides in the 3 lipoprotein classes. However, to define the abnormalities in the 3 major LP classes we must first be certain of the “normal” concentrations. The purpose of this study is to establish the fasting concentrations of cholesterol and triglycerides
in the 3 main ultracentrifugal
LP classes in healthy
male and female
subjects from the Uppsala region, Sweden. These values will then serve as a reference base for studies in patients with IHD. In addition to the quantitative LP analysis, lipoprotein electrophoresis was performed on whole serum as well as on ultracentrifugal fractions in order to identify the presence of particular LP classes such as floating-b LP, sinking pre-/3 LP and late pre-@ LP3. METHODS
Lipoprotein separation After an overnight fast, blood was drawn the following morning by venipuncture without stasis, with the subjects in the sitting position. The blood was allowed to clot at room temperature for 2 hr; serum was recovered after centrifugation at 5700 x g min and 5 % EDTA was added in the proportion 1 :lOO (v/v). The electrophoretic and ultracentrifugal lipoprotein analyses were then carried out within 2 hr of this in most cases. In some instances serum was kept at +4”C for 1 or 2 days before analysis.
QUANTITATIVE
AND
Pl
QUALITATIVE
P3
SERUM LIPOPROTEIN
P4
ANALYSIS,
PART
1
419
95,
Fig. 1. Agarose gel lipoprotein electrophoresis of 4 different sera. P 1 = normal, P 2 = Type III, P 3 = normal with sinking pre-,8, and P 4 = normal with late pre-/J. At the top of the figure the slit where the samples are applied is visible. S =- Whole serum showing from top to bottom the 3 major lipoprotein classes in P 1 (normal) : /3 (LDL), pre-/I (VLDL) and a (HDL). A is the top fraction after ultracentrifugation at d 1.006 and contains normal pre-/3 (VLDL) in P 1, whilst in P 2 VLDL with near ,+mobility, the so-called floating or broad-/? characteristic of Type III hyperlipoproteinemia and also some normal pre-/I can be seen. P 3 contains a normal and P 4 a late and a normal pre-fi. R is the bottom fraction after ultracentrifugation at d 1.006 and shows a normal /I and CIin P 1, P 2 and P 4. P 3 contains in addition a sinking pre-fl.
The ultracentrifugal lipoprotein separation was carried out as described in detail elsewhere3a4. In principle, serum was centrifuged at d 1.006 for 16 hr at 40,000 rpm in a rotor 40.3 (Spinco Model L preparative ultracentrifuge, Beckman Instrument Corp., Calif., U.S.A.). The harvested top fraction was the VLDL. The bottom fraction was then centrifuged at d 1.063, at 40,000 rpm for 20 hr. After centrifugation, the top and bottom fraction, the LDL and HDL, were obtained. Electrophoresis was performed according to the procedure of Noble5 in 1% Agarose gel containing 0.25 % bovine serum albumin (Miles Laboratories Inc., Kankakee, Ill., U.S.A.) in barbital buffer pH 8.6, ionic strength 0.05 for 1 hr at 17 volts/cm. Staining was done with Sudan Black. Whole serum as well as top and bottom fractions after ultracentrifugation at d 1.006 were simultaneously subjected to electrophoresis (Fig. 1). The presence of floating-b, sinking pre-/3 and late pre-/3 LP was ascertained by inspection of the electrophoresis of the top and bottom fractions after ultracentrifugation at d 1.006. An example of these 3 LP groups is shown in Fig. 1. The late pre-#? had slightly greater mobility than the floating-/3 LP. No floating-p LP was observed in this material. Chemical analysis
Whole serum as well as the isolated lipoprotein classes were extracted manually with isopropanol. Cholesterol and triglycerides were then determined by automated techniques with an Auto-Analyzer Model 2 (Technicon Corp.)6. Standards used were triolein and cholesterol, obtained from Sigma Chemical Co., St. Louis, MO. The cholesterol values obtained with this method agreed well with those obtained by the
420
L. A. CARLSON.
Sperry-Webb
method.
The triglyceride
M. ERICSSON
values for whole serum were about 0.2 mmol/l
higher than those obtained by the Carlson method7. The recovery in the lipoprotein analysis, i.e., the sum of cholesterol and triglycerides in the VLDL, LDL and HDL fraction,
was always within 100 * 10 % of the total serum cholesterol
and triglycerides.
Statistical methods Statistical calculator.
calculations
were performed
according
to Snedecor
on a Mae1 4000
MATERIAL
The subjects were randomly selected from the Uppsala region and twice (first early and then late in the 1960) offered a free health examination; about 85 % of 8. The investigation comprised medical history, those selected accepted the offer physical examination, routine blood and urine analysis and resting ECG. On these 2 occasions, from those declared healthy or having only minor ailments we invited a random selection of about 20-25 subjects in each decade to undergo another examination in 1971, an analysis of fasting blood lipids. The subjects presented themselves at the outpatient clinic of the Geriatrics Department after an overnight questioned on medical history since the last examination, present current therapy. Weight (in light underclothes) and height (without
fast. They were health and any shoes) were re-
corded and a fasting venous blood sample was taken. Subjects reporting occurrence of any significant disease (e.g., endocrine disorders, myocardial infarction, malignancy) since the previous examination, subjects on permanent therapy (e.g., B-blocking agents for hypertension, oral contraceptives, etc.), and those reporting recent infection or weight reduction
were excluded.
Fifteen females and 20 males were excluded
on these grounds. All other persons were included in the total material, which comprised 80 males and 77 females. In some of the results overweight subjects have been excluded. (Overweight was defined as weight/height index above 1.10, weight in kg divided
by height in cm minus
100).
RESULTS
The mean concentration of cholesterol and triglycerides in total serum and in the 3 LP classes is given in Table 1 for each group of males and females. Because of the skew distribution of plasma triglycerides, VLDL triglycerides and cholesterol, and LDL triglycerides, the mean for these has also been calculated by logarithmic values (Table 2). The lipid values in relation to age are shown in Figs. 2, 3, 4 and 5.
Age and lipoproteins In both sexes there was an increase with age in triglyceride and cholesterol concentrations in total serum, VLDL and LDL but not in HDL. However, upon linear regression analysis it was found that only the increases in the female subjects
1
(lllmOl/l)
AND CHOLESTEROL
(ITlg/lm
IIll)
CONCENTRATION
FOR MALES
AND FEMALES
IN TOTAL
0.97 *to.15
19 260 *13
It9
1.63 ho.16
1.82 1tO.18
286 It9
1.36 i-o.1 1
1.58 hO.13
1.54 *0.11
1.57 +0.21
1.75 +0.13
1.83 kO.14 287 *13
253 It14
1.17 ho.08
1.26 50.10
219 zt8
251 It11
1.31 10.16
245
251
1.8 7 Ito.
282 *11
0.76 +0.08 0.78 10.09
1.00 zlIo.11
0.69 ho.06
0.52 AO.06
0.85 ho.08
0.82 10.15
250 117 289 &lo
0.59 10.07
213 17
0.78 &to.12
1.03 *to.16
0.99 50.10
1.09 kO.12
2.20 &IO.19
257 i-15
259 *13
1.80 h-o.14
0.99 kO.13
-0
1.01 10.12
1.89 *0.14
242 It11
1.74 Iko.17
1.78 hO.17 244 It11
T
T
T
-0
TG
Chol
TG
-0
VLDL
Total
TG = triglyceride, Chol = cholesterol.
Females 30-39 N=15T N= 12-O 40-49 N=l5T N= 12-O 50-59 N=24T N= 18-O 60-69 N=23T N= 13-O
Males 30-39 N=lST N=17-0 4049 N=l7T N= 13-O 50-59 N=25T N==17-0 60-69 N=20T N= 14-O
Groups
21 *3
15 12
15 *2
14 *2
16 +3 18 *2
10 %2
16 1-3
12 It2
18 It3
20 13
*3
26 It3
21
It3
It3
22
20
*3
-0
21
T
Chol
0.59 Yko.05
0.58 Ito.
0.51 &IO.04
0.41 +0.03
0.54 zko.03
0.55 kO.03
0.52 xto.04
0.50 kO.04
T
TG
LDL
0.53 50.06
0.55 zko.04
0.47 *to.03
0.39 *0.02
0.54 *0.04
0.53 kO.03
0.53 &-0.04
0.50 *0.04
-0.
Mean value & S.E.M. is given for total material (T) and after elimination of overweight persons (-0) above 1.10.
TRIGLYCERIDE
TABLE AND IN THE
3 MAJOR LIPOPROTEIN
CLASSES
ill
190
184 I!I~
159 1-12
133 +7
168 ik8
160 1t6
175 It10
162 +9
T
Chol
185 &-I1
0.25 &to.01
0.24 AO.02
0.29 *0.02
0.30 *0.01
187 +9
0.26 AO.02
0.25 zko.02
0.24 hO.01
0.24 +0.02
0.23 zto.01
-0
0.26 hO.02
0.26 zto.02
0.25 +0.02
0.26 &O.Ol
0.23 *to.01
0.23 Ito.
T
157 0.28 I!K15 kO.02
127 +6
173 &lo
158 +7
173 +12
162 19
-0
TG
HDL
69 &4
69 zt3
76 *5
75 13
66 *4
*3 65 -1-4
64 +3
60 +4
53 zt3
51 AI4
54 A3
-0
63
56 *4
51 ItI2
50 $_3
53 *3
T
Chol
defined as a weight/height index (kg/(cm -100))
SERUM
[lolog(10
VALUES
Females 3&39 N=15T N=12-0 40-49 N=15T N=12-0 5C59 N=24T N=18-0 60-69 N=23T N=13-0
Males 30-39 N= 18T N= 17-o 40-49 N=l7T N-13-O 50-59 N=25T N=17-0 60-69 N=20T N= 14-O
Groups
1.0722 10.1102
1.1395 *to.0854
0.8401 kO.0577
0.9077 zto.0570
1.1868 10.0396
1.2258 &IO.0387
0.8494 i 0.0598
1.1177 *0.0379 1.1745 +0.0365 1.1758 kO.0305
1.1572 +0.0468
1.2144 *0.0330
1.2367 ho.0308
0.9423 10.0485
0.8899 * 0.0470
0.7260 hO.0514
1.0594 *0.0273
1.0853 &0.0311
0.8582 io.0513
0.8341 *0.0496
0.8178 +0.0440
0.7239 +0.0290
0.7251 & 0.0248 1.0622 &0.0831 1.1476 kO.0513
1.1373 10.0707 1.2409 *0.0551
0.7341 Iko.0339
0.6619 ho.0321 0.6880 10.0331
1.1272 It 0.0493 1.1119 f0.0821
0.5828 *to.0214
0.7143 hO.0375
0.7106 kO.0235
0.7086 & 0.0302
0.6762 *to.0335
IN VLDL
0.6046 ztO.0241
0.7138 10.0282
0.7269 iO.0239
0.6970 &0.0313
0.6779 -CO.0316
T
(mg/lOOml)]
0.9420 10.0953 0.9836 hO.0917
1.2225 ho.0629
1.3302 &to.0563
0.9484 *0.0570
1.0405 10.0530
1.2474 10.0350
1.3058 10.0349
0.6880 &0.0506
1.2786 *0.0652
1.2937 50.0572
0.9645 10.0521
0.9975 kO.0488
1.2406 +0.0332
1.2573 zto.0319
1.2429 + 0.0648
1.2110 10.0391
1.2206 *0.0381
0.9414 =‘cOo.O528
1.2237 kO.0657
-0
Plog
LDL TG
CONCENTRATION
0.9531 10.0511
T
-0
T
T
-0
VLDL
VLDL TG
Total TG
Chol
VLDL AND LDL AND CHOLESTEROL
and symbols as in Table 1.
x mmol/l)]rN TOTALSERUM,
Triglyceride values multiplied by 10. Other conditions
ON LOGARITHMIC
CALCULATED
2
TRIGLYCERIDE CONCENTRATION
TABLE
r ?
QUANTITATIVE
-MALES
AND QUALITATIVE
(n.
SERUM LIPOPROTEIN
-FEMALES
60)
(n=77)
ANALYSIS,
PART
1
423
MEANtSEM
TOTAL CHOLESTEROL
TOTAL TRIGLYCERIDES
mg/lOOml 320 -I
-
70 AGE
Fig. 2. Mean concentration and females.
e-a
30
40
YEARS
of triglycerides and cholesterol in total serum in relation to age in males
MALES (n.60)
-
FEMALES (n=77)
VLDL TRIGLYCERIDES
VLDL
lolog (lO~mmol/l)
lolog mg/lOO ml
AGE
AGE
YEARS
of triglycerides and cholesterol in VLDL lipoprotein
MALES (n-60)
M
FEMALES (n-77)
LDL TRIGLYCERIDES
LDL CHOLESTEROL
lolog ( 10 ammol/ I )
mg/lOOml
0.8
200 &@=O
0.7
MEAN t SEM
CHOLESTEROL
YEARS
Fig. 3. Mean concentration males and females.
-
50 60 70 AGE YEARS
in relation to age in
MEAN* SEM
160
0.6 i -f I
30
I
40
I
50
I
60
1
70
AGE YEARS
Fig. 4. Mean concentration males and females.
I
30
I
40
I
50
60
1
70
AGE YEARS
of triglycerides and cholesterol in LDL lipoprotein
in relation to age in
424
L. A. CARLSON, M. ERICSSON .-.
HOL
MALES
(n-80)
TRIGLYCERIDES
mmol/l
o-o
FEMALES
HOL
CHOLESTEROL
(n=77)
+r
SEM
mg/lpOmI
30 .25 .20
AGE
AGE
YEARS
Fig. 5. Mean concentration males and females.
YEARS
of triglycerides and cholesterol in HDL lipoprotein
in relation to age in
were statistically significant. In whole serum triglycerides increased by 0.005 (P > 0.05) and 0.02 (P < 0.01) mmol/l/year in males and females respectively. Corresponding figures for cholesterol were 0.05 (P > 0.05) and 2.4 (P < 0.001) mg/lOO ml/year. In females, VLDL triglycerides and LDL cholesterol increased by 0.014 (P < 0.01) mmol/l and 2.0 (P < 0.001) mg/lOO ml/year. not vary with age.
HDL triglycerides
and cholesterol
did
Lipoprotein composition The relationship between cholesterol and triglycerides in males in the 3 LP classes are illustrated in Figs. 6, 7 and 8, and the linear regression analysis for males and females is given in Table 3. For VLDL there was a strong linear relation between the cholesterol and triglycerides content of this LP class. The relation contained the origin. There was also a significant positive linear relation for LDL. This, however, was not as strong as for VLDL and did not contain the origin. There was no relation between the cholesterol and the triglyceride content in HDL. A very similar relationship existed between cholesterol and triglycerides in the LP classes for female and for male subjects (Table 3).
. . . .**.* ..* /......’ VLDL
COMPOSITION.
??
??
0”
??
+:I . .
0.
4
.
.
:
Normal men .
??
r zO.90
a.
Fig. 6. Relationship
between the content of VLDL cholesterol
and triglycerides
in male subjects.
QUANTITATIVE
300
1
AND QUALITATIVE
LDL COMPOSITION.
Fig. 7. Relationship
SERUM LIPOPROTEIN
ANALYSIS,
PART
425
i
Normal men
.
between the content of LDL cholesterol and triglycerides in male subjects.
TABLE 3 LINEARRE~RESSIONANDCORRELATIONANALYSISF~RTHE
(Y)
AND TRIGLYCERIDES, InItIOl/l (X) IN THE
Females
b
VLDL LDL HDL
mg/lOOml,
CLASSES
Males
LP class
CHOLESTEROL,
RELATIONBETWEEN
LP
a
r
P
b
a
Y
P
19
1
0.90
< 0.001
25
-3
0.92
< 0.001
169
75
0.73
< 0.001
180
75
0.72
< 0.001
-16
56
-0.08
> 0.05
-10
70
-0.04
on Y-axis; r = linear regression
b = linear regression
coefficient; a = intercept statistical significance of b and r.
1 0.05
coefficient:
P =
Sex and lipoprotein concentration There were several differences between male and female subjects. The 2 most constant were : (l), Females had lower VLDL-triglyceride and cholesterol levels than
HDL COMPOSITION.
Normal
men
.* . 84.
??
.
-
..
r-0.08
O-5 TRIGLYCERIDES
Fig. 8. Relationship
m m.fi
between the content
of HDL cholesterol
and triglycerides
in male subjects.
426
L. A. CARLSON, M. ERICSSON
males except for those in the oldest age group (Fig. 3). This was reflected in the lower triglyceride cholesterol
levels in total serum in females (Fig. 4). (2), Females had higher HDL (all age groups) and triglyceride levels (with the exception of the oldest age
group) than males (Fig. 5). The most characteristic difference between males and females regarding LDL cholesterol levels was the rise with age in females but not in males, (Fig. 4).
which caused
the male and female curves
to intersect
at age 45-50 years
Influence of overweight Elimination of overweight glyceride as well as the VLDL
persons from the study lowered the total serum tritriglyceride and cholesterol levels in all male and
female age groups (Tables 1 and 2). The effect of this procedure was highly significant. There was no effect on total serum, LDL cholesterol or LDL triglyceride levels. HDL-cholesterol but not triglyceride levels were significantly increased when overweight persons were eliminated from the male and female groups. It can be seen that overweight healthy persons differed from the rest of the
-----TOTAL TOTAL
,,
MATERIAL
(1)
TRIGLYCERIDES
10
I5
2.0
2.5
3.0
-
OVERWEIGHT
TOTAL
CHOLESTEROL
EXCLUDED
C-0)
3.5”
mmolll
60.
T _-0
nz77 I-l:55
40_
-F1.0
1.5
2.0
2.5
3.0
3.5 mmolll
400 mg/lOOml
Fig. 9. Cumulative frequency distribution curves for triglycerides and cholesterol in total serum. Male values not corrected for age, female values corrected by linear regression equation to age 50 yr. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated.
QUANTITATIVE
__- _ VLDL
AND
QUALITATIVE
TOTAL MATERIAL
SERUM LIPOPROTEIN
(T)
-OVERWEIGHT
TRIGLYCERIDES
05
1.0
VLDL
1.5
2.0
ANALYSIS,
2.5
3.0mmollk
PART EXCLUDED
10
20
30
40
50
0.5
1.0
I I
I,
I
1.5
2.0
25
3.0mmolll
10
20
30
1 1,
40
60
70mg1100
ml
___--
I I 1 I ’ ,
’
t-0)
CHOLESTEROL
%------l?/
h
427
1
T “~77
-0 n- 55
,
,
50
60
(
70mgllOO
ml
Fig. 10. Cumulative frequency distribution curves for VLDL triglycerides and cholesterol. Male values not corrected for age, female values corrected by linear regression equation to age 50 yr. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated.
population cholesterol
by having higher total triglycerides, and lower HDL cholesterol levels.
Frequency distribution and population limits The cumulative frequency distribution
higher
VLDL
curves for cholesterol
triglycerides
and
and triglycerides
in total serum and in the LP classes are shown in Figs. 9, 10, 11 and 12. Since total, VLDL and LDL lipids increased with age in females, all the female values have been adjusted to age 50 yr by using the linear regression coefficients obtained for each lipid LP in relation to age. In general the curves for total triglycerides, VLDL and LDL have the typical appearance of a frequency distribution that is skewed to the right, i.e., the mean is greater than the median. An example of this can be seen in the median value (= 50 % cumulative %) for total triglycerides in males of I .5 mmol/l (Fig. 9), while the mean value was 1.9 (Table 1). The shape of the other curves indicates a more normal distribution. The effect of elimination of overweight persons from the group under study is easily seen. The upper part of the frequency curve for total triglycerides and for VLDL was shifted to the left. For HDL-cholesterol the lower part was shifted to the right, as expected. The LDL curves were not influenced by this elimination.
L.
428 ------
TOTAL
-
MATERIAL (T)
0.4
0.6
OVERWEIGHT
EXCLUDED c-01
_LDL CHOLESTEROL
LOL TRIGLYCERIDES
0.2
A. CARLSON, M. ERICSSON
0.8
1.0
1.2
1.4
mmol/l
a L
60
0.2
0.4
0.6
a8
1.0
12 1.4 mmol/l
100
150
200
250
300
350
mgllO0 ml
Fig. 11. Cumulative frequency distribution curves for LDL triglycerides and cholesterol. Male values not corrected for age, female values corrected by linear regression equation to age 50 yr. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated.
We have indicated in the figures the limits covering 95, 90 and 85 % of the population. In the non-overweight population the upper 90 % population limit was for males/females (aged 50 yr) : total triglycerides 2.5/2.0, total cholesterol 298/300, VLDL triglyceride 1.80/1.05, VLDL cholesterol 32/23, LDL triglyceride 0.69/0.69, LDL cholesterol 210/218, HDL triglyceride 0.32/0.34 and HDL-cholesterol 69/93. Floating-p, late pre-/3 and sinking pre-/I LP The number of these electrophoretic LP’s is given in Table 4. No floating-8 was observed. A late pre-p (see Fig. 1) was found in 35 ‘A of males and 25 ‘A of females. The sinking pre-/I was detected in 25 and 22 % of males and females. Overweight did not appear to be associated with these LP’s. Subjects with sinking pre-/I LP’s did not differ from other subjects with regard to lipid concentration in the 3 LP classes. However, males having the late pre-/3 LP differed from those without this LP (Table 4). The VLDL cholesterol content was raised, leading to an increased cholesteroltriglycerides ratio (P < 0.001). Men with late pre-p LP also had an increased amount of LDL triglycerides.
QUANTITATIVE
AND QUALITATIVE
-------
TOTAL
HDL
TRIGLYCER~S
MATERIAL
SERUM LIPOPROTEIN
(1)
-
ANALYSIS,
OVERWEIGHT
PART EXCLUOEO
429
1 t-0)
HOL CHOLESTEROL
6. LO
P
20
OA
05
# LA!2 I
1 ns77 -0 n=55
/
,
a-
’
30
0.6
nmmlll
o i
.*
40
/
, .*’
-0’
50
60
70
60 90 100 mg/lOO ml
Fig. 12. Cumulative frequency distribution curves for HDL-triglycerides and cholesterol. No age corrections. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated. TABLE 4 NUMBER
OF SUBJECTS
WITH
LATE
PRE-B
AND SINKING
PRE-B
LP
Subjects
N”
Late pre-/3
Sinking pre-B
Males Males (-obese)
55 42
19 16
14 IO
Females Females (-obese)
50 38
13 9
11 8
8 Number of subjects with sufficiently complete analysis to permit visual assessment of the presence of late pre-/3 and sinking pre-/? LP. For explanation of procedure see Methods and Fig. 1.
DISCUSSION
Age Females (but not males) showed statistically significant increases in serum, VLDL and LDL cholesterol and triglyceride concentrations with age. The finding that males did not show any increase might at a first sight appear contradictory to earlier studies on serum lipids. However, the increase in males is most pronounced in the
430 TABLE
L. A. CARLSON,
hf. ERICSSON
5
LIPOPROTEIN CHOLESTEROL MEAN VALUE
COMPOSITION (mg/lm &
Groups
Males No late pre-8 N= 36 Late pre-/I N= 19 Females No late pre-b N = 34 Late pre-/I N= 13
ml)
1N SUBJECTS
WITH
CONCENTRATION
AND
WITHOUT
IN THE
3
LP
LATE
CLASSES
PRE-j!i. AND
TRIGLYCERIDE RATIO
(mIIlOl/l)
OF CHOL/TG
IN
AND
VLDL.
S.E.M. VLDL
LDL
HDL
TG
Chol
Chol/TG
TG
Chol
1.15*0.11
21+2
18&l
0.51 + 0.03
161 f
1.23*0.18
28&4
2441
0.60 A 0.04
0.83 i
16h
0.09
0.70*0.10
14*3
2
19+ 19*2
1
TG
Chol
0.25 zt 0.01
52 f
171 4 7
0.27 f
0.02
51 * 2
0.50 & 0.03
160 + 7
0.28 zt 0.01
67 + 3
0.55 & 0.04
176 i_ 13
0.27 5 0.02
64 + 4
6
2
TG = triglycerides, Chol = cholesterol.
younger age groups. Carlson thus found that in 26-39-yr-old males, cholesterol and triglyceride levels increased about 10 times more per year than in males aged 40-72 yrs. In the Stockholm Prospective Study, cholesterol and triglyceride levels increased only until about age 50 yr in males while they continued to rise in femaleslO. This greater increase with age in women is illustrated by the values for the mean cholesterol content in age groups 25-30/6&65 yr, which were 2291331 in females and 2461278 mg/lOO ml in maleslo. Corresponding triglyceride values were 1.00/1.22 and 1.39/1.32 mmol/l. The present material is limited and the study begins with subjects of age 30 yr. This explains why statistically significant increases with age were obtained here for females but not for males. Sex The observed higher serum triglyceride levels in males and the similar cholesterol levels in the 2 sexes is in agreement with earlier and more extensive studieslOJ1. The sex difference in total serum triglycerides can now be explained at the lipoprotein level; it is entirely due to higher VLDL levels in men than in women, LDL triglycerides being similar in the 2 sexes. As a result of the lower VLDL levels, the VLDL cholesterol content is lower in females than in males. This difference is, however, not observable in total serum cholesterol (which is the sum of VLDL, LDL and HDL cholesterol), because of the higher HDL cholesterol levels in females. A higher amount of cholesterol in females than in males in the Cohn fraction IV + V + VI, corresponding to HDL, was reported in 1951 by Russ et al. 12. The increased amount of HDLin females has been ascribed to an increase of the HDLa fractionls.
QUANTITATIVE
AND QUALITATIVE
SERUM LIPOPROTEIN
ANALYSIS,
PART
1
431
Overweight It is known
from
several
studies
that there is a positive
correlation
between
fasting serum triglyceride concentrations and various measures or indices of adipose tissue mass10J4J5. Either a weaker10 or no correlation16 has been found between serum cholesterol and adiposity. In the present study, overweight influenced serum triglyceride but not cholesterol levels. The cumulative frequency distribution showed that the influence of overweight only operated at triglyceride concentrations above the median value (Fig. 6). Approximately 90% of overweight subjects had serum triglyceride levels above the median value. The findings for serum lipids are explained by the fact that overweight influenced VLDL but not LDL concentrations. Elimination of overweight subjects shifted the lower part of the HDL cholesterol distribution curve to the right but had no effect on HDL triglyceride levels (Fig. 9). This effect may be explained by the usual negative correlation between VLDL and HDL cholesterol. Exclusion of overweight subjects from the present study was mainly an exclusion of persons with high VLDL- and thus probably also low HDL cholesterol concentrations. Normal values for LP classes One object of this study was to establish “normal” limits for the concentration and composition of the different lipoprotein classes. The limits can easily be obtained from the cumulative diagrams and depending on the way in which normality is defined, one can select any desired upper - or lower - cutting point. There are 2 major advantages in using cumulative frequency curves rather than the mean value and multiples of the standard deviation for finding upper normal limits of the population. Firstly, it does not matter what kind of distribution (normal, skewed to the left or right) we are dealing with; one just enters the curve at the required level. Secondly, single very extreme values which may influence the mean and particularly the standard deviation do not influence the value for e.g., the 95th or lower percentiles. Sinking pre-/3 LP Subjects with sinking pre-p LP had the same amount of cholesterol and triglycerides in the 3 LP classes as other subjects, suggesting that neither LP metabolism nor LP composition were greatly altered in the former group. The sinking pre-/? LP appears to be the same lipoprotein as the so-called Lp(a) discovered in Norway by BergI’. Using an immunologic technique, the frequency of Lp(a) was found to be 35 % in Norway’s. Our finding of 25 and 22 ‘A of sinking pre-@ LP in males and females (Table 3) is lower, but as mentioned previously, immunologic techniques were used in the Norwegian study which may have been more sensitive than our ultracentrifugal-electrophoretic visual inspection method. In a study with a visual inspection technique made in the U.S.A., the frequency of sinking pre-p LP was 25 %lg, a figure similar to that obtained in the present study. However, a recent study from Seattle using sensitive immunologic technique reported a frequency of Lp(a) of
432 approx.
L. A. CARLSON, bf. ERICSSON 80°A20. In northern
Sweden
an extra pre-p
lipoprotein
has been found in approx. 15-20 % of healthy male subject@i. band with sinking pre-p LP has, however, not been established. Late pre-p LP Males with late pre-p
LP had an increase
in VLDL
known
as pre-11,
The identity
of this
cholesterol-triglycerides
ratio. This suggests that the late pre-/3 has a high cholesterol-triglycerides males this LP, being visible in about one-third of the population, would
ratio. In then have
the following characteristics : (1) d < 1.006, (2), Electrophoretic mobility in agarose close to p-LP, and (3) High cholesterol-triglycerides ratio. The late pre-@ LP thus resembles the floating or broad Type III B-LP. The additional finding that subjects with a late pre-@ LP had increased LDL triglyceride content but normal cholesterol levels suggests that they had increase in the Sf 12-20 LP or LDLi LP. This increase, as well as the presence of late pre-/3 LP, may be due to an increased amount of socalled remnants or intermediary particles in fasting blood. Remnants are believed to represent degradation products of normal chylomicra and VLDL that accumulate in Type III due to a catabolic defect 22. The presence of such LP’s in about one-third of the male population in the presence of normal VLDL triglycerides is an interesting possibility. The reproducibility and genetics of the occurrence of the late pre-p LP in a subject, as well as its physiologic and clinical importance, has to await further studies. In his original description of the electrophoretic separation of lipoproteins in Agarose gel, Noble5 noted that “plasma samples from different subjects show pre-,!l lipoproteins of different mobilities”. No further characterization of these pre-B LP’s has so far appeared. The fact that these pre-j3 LP’s5 were easily seen in electrophoresis of whole serum makes it likely that they correspond to the sinking pre-/I LP, as this pre-p-abnormality is easily seen in whole serum (Fig. l), while the late pre-/3 LP is more clearly seen in electrophoresis of the VLDL fraction (Fig. 1). Lipoprotein composition The positive relationship between VLDL cholesterol and triglyceride content, containing the origin, indicates that this LP class was composed of homogenous particles. When the VLDL concentration increased, this relation suggested that the increase was due to an increased number of VLDL particles of identical composition (at least with regard to cholesterol/triglycerides). There was also a significant positive relation between LDL cholesterol and triglyceride content. However, this was not as strong nor did it contain the origin. This was probably because the LDL fraction, defined by density 1 JO61.063, was composed of at least 2 rather different LP classes, LDLr (d1.006-1.019) andLDLs(d1.019-1.063). The former is more triglyceride-rich and the latter more cholesterol-rich, and their concentrations may vary independently. Interestingly there was no relation between HDL cholesterol and triglyceride content, indicating that these 2 lipid components vary independently of each other. Since few studies have been made on the triglyceride content of HDL, it is difficult
QUANTITATIVE
AND
QUALITATIVE
SERUM LIPOPROTEIN
ANALYSIS,
PART
433
1
to evaluate this. Also since this fraction contains all material with d > 1.063 in plasma, caution should be exercised as the small amount of triglyceride found by the present methods of determination has not been chemically identified as triglyceride.
REFERENCES CARLSON, L. A., Serum lipids in men
with myocardial
infarction,
Actu Med. Stand., 167 (1960)
399. GOLDSTEIN, J. L.,
HAZZARD, W. R., SCHROTT, H. G., BIERMAN, E. L. AND MOTULSKY, A. G., Hyperlipidemia in coronary heart disease, J. C/in. Invest., 52 (1973) 1533-1577. CARLSON, K., Lipoprotein fractionation, J. C/in. Path., Suppl. 5, 26 (1973) 32. CARLSON, L. A., The effect of nicotinic acid treatment on the chemical composition of plasma lipoprotein classes in man - Drugs affecting lipid metabolism, Advunc. Exptl. Med. Biol., 4 (1969)
327. separation of plasma lipoproteins in agarose gel, J. Lipid Res., 9 5 NOBLE, R. P., Electrophoretic (1968) 693. 6 RUSH, R. L., LEON, L. AND TURRELL, J., Automated simultaneous cholesterol and triglyceride determination on the Auto Analyzer@ II instrument. In: Advances in Automated Analysis, Vol. Z, Thurman,
1971, p. 503.
of serum triglycerides, J. Atheroscler. Res., 3 (1963) 334. 7 CARLSON, L. A., Determination 8 ERICSSON, P., Lung Function and Physical Work Capacity in Healthy Elrler1.vPeople (Thesis 102), Acta Univ. Upsal., 1971. 9 CARLSON, L. A., Serum lipids in normal men, Acta Med. Stand., 167 (1960) 377.
No.
10 CARLSON, L. A., AND LINDSTEDT, S., The Stockholm Prospective Study, Part 1 (The initial values for plasma lipids), Actu Med. Scund., Suppl. No 493, 1969. 11 WOOD, P. D. S., STERN, M. P., SILVERS, A., REAVEN, G. M. AND GROEBEN, J., Prevalence of plasma lipoprotein abnormalities in a free-living population of the Central Valley, California,
Circulation, 45 (1972) 114. 12 Russ, E. M., EDER, H. A. AND BARR, D. P., Protein-lipid relationship in human plasma, (In normal individuals), Amer. J. Med., 11 (1951) 468. 13 BARCLAY, M., BARCLAY, R. K. AND SKIPSKI, V. P., High density lipoprotein concentration
Part
I
in men
and women, Nafure (Lond.), 200 (1963) 362. 14 ALBRINK, M. J. AND MEIGS, J. W., Interrelationship between skinfold thickness, serum lipids and blood sugar in normal men, Amer. J. Clin. Nutr., 15 (1964) 255. 15 BIERMAN, E. L., PORTE, D. AND BAGDADE, J. P., Hyperglyceridemia and glucose intolerance in man, In : B. JEANRENAUD AND D. HEPP (Eds.), Adipose Tissue, Regulation and Metabolic Functions, Academic Press, New York, N.Y., 1970, p. 209. 16 KEYS, A. et al., Coronary heart disease among Minnesota business and professional men followed 15 years, Circulation, 28 (1963) 381. 17 BERG, K., The Lp system, Ser. Haematol., 1 (1968) I1 I. 18 MONN, E., BERG, K., REINSKON, T. AND TEISBERG, P., Serum protein polymorphisms among Norwegian Lapps, Human Hered., 21 (1971) 134. 19 ELLEFSON, R. D., JIMENEZ, B. S. AND SMITH, R. C., Pre-P lipoprotein of high density in human blood, Mayo C/in. Proc., 46 (1971) 328. 20 ALBERS, J. AND HAZZARD, W., Quantitative genetic analysis of the human plasma Lp(a) lipoprotein. In G. SCHETTLER AND A. WEIZEL (Eds.), Atherosclerosis ZZZ,Springer, Berlin, 1974, p. 469. 21 DAHL~N, G., ERICSON, C., FORSBERG, C., LUNDKVIST, L. AND SVKRDSUDD, K., Studies on an extra pre-/? lipoprotein fraction, Actu Med. &and., Suppl. 53 1, 1972. 22 HAZZARD, W. R., PORTE, D. AND BIERMAN, E. L., Abnormal lipid composition of chylomicrons in broad-p disease (Type III hyperlipoproteinemia), J. C/in. Invest., 49 (1970) 1853.
21 (1975) 417433
0 Elsevier Scientific Publishing Company, Amsterdam
QUANTITATIVE PART
AND
1. STUDIES
QUALITATIVE
IN HEALTHY
LARS A. CARLSON King Gustaf V Research Uppsala (Sweden)
AND
MARTIN
Institute,
417 - Printed in The Netherlands
SERUM
LIPOPROTEIN
ANALYSIS
MEN AND WOMEN
ERICSSON
Stockholm,
and Department
of Geriatrics,
University
of Uppsala,
(Received October 21st, 1974) (Accepted November 28th, 1974)
SUMMARY
Preparative ultracentrifugal and electrophoretic was performed in 30-70-years-old healthy, fasting
analysis of serum lipoproteins males (N = 80) and females
(N = 77), randomly selected from the Uppsala region, Sweden. The concentrations of cholesterol and triglycerides in total serum and in VLDL, LDL and HDL lipoproteinclassesare reported. Total serum, VLDL and LDL triglycerides and cholesterol concentrations increased with age, while HDL cholesterol and triglyceride concentrations did not vary with age. Overweight persons had higher total serum triglyceride, higher
VLDL
cholesterol
and triglyceride
and lower
HDL
cholesterol
levels. The
upper 90 % population limit values for non-overweight males/females were: total triglycerides (mmol/l) 2.5/2.0, total cholesterol (mg/lOO ml) 298/300, VLDL triglyceride 1.80/1.05, VLDL-cholesterol 32/33, LDL triglyceride 0.69/0.69, LDL cholesterol 210/218, HDL triglyceride 0.32/0.34 and HDL-cholesterol 69/93. The 2 major differences between males and females were that females had lower VLDL but higher HDL concentrations. For VLDL there was a very strong and for LDL a moderately strong positive correlation between cholesterol and triglyceride contents. In HDL however, the measured amounts of cholesterol and triglycerides did not correlate at all. Sinking pre-b lipoprotein was found in about 25% of cases and a second pre-/I band floating at d 1.006, late pre-/I, was found in 35% of male and 25% of female subjects. Subjects with sinking pre-j3 lipoprotein did not differ from other subjects with regard to the concentration of cholesterol and triglycerides in the 3 lipoprotein classes. Males, but not females, with the late pre-/3 (LPB), had an increased amount of cholesterol in VLDL and a raised cholesterol-triglyceride ratio
Supported
by Grants from the Swedish Medical Research Council.
418
L. A. CARLSON,
in this lipoprotein
class. Also the LDL triglyceride
level was increased
M. ERICSSON
in males with
the late pre-/3 lipoprotein.
Key words:
Agarose gel electrophoresis - Healthy men and women - Late pre-P lipoprotein - Preparative ultracentrifugation - Serum lipoproteins Sinking pre-/3 lipoprotein
INTRODUCTION
There is a close association between ischaemic cardiovascular diseases and hyperlipidaemia. Various forms of hyperlipidaemia such as pure hypertriglyceridaemia, pure hypercholesterolaemia and combined hyperlipidaemia (elevation of both triglyceride and cholesterol concentrations) are frequently found in patients with ischaemic heart disease (IHD) 1,s. The serum lipids do not, however, circulate freely in the blood but are bound to the (VLDL), the low (LDL) and the contain cholesterol and triglycerides Thus hypercholesterolaemia as well
serum lipoproteins (LP), high (HDL) density LP and their concentrations as hypertriglyceridaemia
mainly to the very low classes. These LP’s all may vary independently. may occur as a result of
elevation of levels in any of the LP classes, either alone or in combination. To detect the LP abnormality underlying hyperlipidaemia it is necessary to measure the concentration of cholesterol and triglycerides in the 3 lipoprotein classes. However, to define the abnormalities in the 3 major LP classes we must first be certain of the “normal” concentrations. The purpose of this study is to establish the fasting concentrations of cholesterol and triglycerides
in the 3 main ultracentrifugal
LP classes in healthy
male and female
subjects from the Uppsala region, Sweden. These values will then serve as a reference base for studies in patients with IHD. In addition to the quantitative LP analysis, lipoprotein electrophoresis was performed on whole serum as well as on ultracentrifugal fractions in order to identify the presence of particular LP classes such as floating-b LP, sinking pre-/3 LP and late pre-@ LP3. METHODS
Lipoprotein separation After an overnight fast, blood was drawn the following morning by venipuncture without stasis, with the subjects in the sitting position. The blood was allowed to clot at room temperature for 2 hr; serum was recovered after centrifugation at 5700 x g min and 5 % EDTA was added in the proportion 1 :lOO (v/v). The electrophoretic and ultracentrifugal lipoprotein analyses were then carried out within 2 hr of this in most cases. In some instances serum was kept at +4”C for 1 or 2 days before analysis.
QUANTITATIVE
AND
Pl
QUALITATIVE
P3
SERUM LIPOPROTEIN
P4
ANALYSIS,
PART
1
419
95,
Fig. 1. Agarose gel lipoprotein electrophoresis of 4 different sera. P 1 = normal, P 2 = Type III, P 3 = normal with sinking pre-,8, and P 4 = normal with late pre-/J. At the top of the figure the slit where the samples are applied is visible. S =- Whole serum showing from top to bottom the 3 major lipoprotein classes in P 1 (normal) : /3 (LDL), pre-/I (VLDL) and a (HDL). A is the top fraction after ultracentrifugation at d 1.006 and contains normal pre-/3 (VLDL) in P 1, whilst in P 2 VLDL with near ,+mobility, the so-called floating or broad-/? characteristic of Type III hyperlipoproteinemia and also some normal pre-/I can be seen. P 3 contains a normal and P 4 a late and a normal pre-fi. R is the bottom fraction after ultracentrifugation at d 1.006 and shows a normal /I and CIin P 1, P 2 and P 4. P 3 contains in addition a sinking pre-fl.
The ultracentrifugal lipoprotein separation was carried out as described in detail elsewhere3a4. In principle, serum was centrifuged at d 1.006 for 16 hr at 40,000 rpm in a rotor 40.3 (Spinco Model L preparative ultracentrifuge, Beckman Instrument Corp., Calif., U.S.A.). The harvested top fraction was the VLDL. The bottom fraction was then centrifuged at d 1.063, at 40,000 rpm for 20 hr. After centrifugation, the top and bottom fraction, the LDL and HDL, were obtained. Electrophoresis was performed according to the procedure of Noble5 in 1% Agarose gel containing 0.25 % bovine serum albumin (Miles Laboratories Inc., Kankakee, Ill., U.S.A.) in barbital buffer pH 8.6, ionic strength 0.05 for 1 hr at 17 volts/cm. Staining was done with Sudan Black. Whole serum as well as top and bottom fractions after ultracentrifugation at d 1.006 were simultaneously subjected to electrophoresis (Fig. 1). The presence of floating-b, sinking pre-/3 and late pre-/3 LP was ascertained by inspection of the electrophoresis of the top and bottom fractions after ultracentrifugation at d 1.006. An example of these 3 LP groups is shown in Fig. 1. The late pre-#? had slightly greater mobility than the floating-/3 LP. No floating-p LP was observed in this material. Chemical analysis
Whole serum as well as the isolated lipoprotein classes were extracted manually with isopropanol. Cholesterol and triglycerides were then determined by automated techniques with an Auto-Analyzer Model 2 (Technicon Corp.)6. Standards used were triolein and cholesterol, obtained from Sigma Chemical Co., St. Louis, MO. The cholesterol values obtained with this method agreed well with those obtained by the
420
L. A. CARLSON.
Sperry-Webb
method.
The triglyceride
M. ERICSSON
values for whole serum were about 0.2 mmol/l
higher than those obtained by the Carlson method7. The recovery in the lipoprotein analysis, i.e., the sum of cholesterol and triglycerides in the VLDL, LDL and HDL fraction,
was always within 100 * 10 % of the total serum cholesterol
and triglycerides.
Statistical methods Statistical calculator.
calculations
were performed
according
to Snedecor
on a Mae1 4000
MATERIAL
The subjects were randomly selected from the Uppsala region and twice (first early and then late in the 1960) offered a free health examination; about 85 % of 8. The investigation comprised medical history, those selected accepted the offer physical examination, routine blood and urine analysis and resting ECG. On these 2 occasions, from those declared healthy or having only minor ailments we invited a random selection of about 20-25 subjects in each decade to undergo another examination in 1971, an analysis of fasting blood lipids. The subjects presented themselves at the outpatient clinic of the Geriatrics Department after an overnight questioned on medical history since the last examination, present current therapy. Weight (in light underclothes) and height (without
fast. They were health and any shoes) were re-
corded and a fasting venous blood sample was taken. Subjects reporting occurrence of any significant disease (e.g., endocrine disorders, myocardial infarction, malignancy) since the previous examination, subjects on permanent therapy (e.g., B-blocking agents for hypertension, oral contraceptives, etc.), and those reporting recent infection or weight reduction
were excluded.
Fifteen females and 20 males were excluded
on these grounds. All other persons were included in the total material, which comprised 80 males and 77 females. In some of the results overweight subjects have been excluded. (Overweight was defined as weight/height index above 1.10, weight in kg divided
by height in cm minus
100).
RESULTS
The mean concentration of cholesterol and triglycerides in total serum and in the 3 LP classes is given in Table 1 for each group of males and females. Because of the skew distribution of plasma triglycerides, VLDL triglycerides and cholesterol, and LDL triglycerides, the mean for these has also been calculated by logarithmic values (Table 2). The lipid values in relation to age are shown in Figs. 2, 3, 4 and 5.
Age and lipoproteins In both sexes there was an increase with age in triglyceride and cholesterol concentrations in total serum, VLDL and LDL but not in HDL. However, upon linear regression analysis it was found that only the increases in the female subjects
1
(lllmOl/l)
AND CHOLESTEROL
(ITlg/lm
IIll)
CONCENTRATION
FOR MALES
AND FEMALES
IN TOTAL
0.97 *to.15
19 260 *13
It9
1.63 ho.16
1.82 1tO.18
286 It9
1.36 i-o.1 1
1.58 hO.13
1.54 *0.11
1.57 +0.21
1.75 +0.13
1.83 kO.14 287 *13
253 It14
1.17 ho.08
1.26 50.10
219 zt8
251 It11
1.31 10.16
245
251
1.8 7 Ito.
282 *11
0.76 +0.08 0.78 10.09
1.00 zlIo.11
0.69 ho.06
0.52 AO.06
0.85 ho.08
0.82 10.15
250 117 289 &lo
0.59 10.07
213 17
0.78 &to.12
1.03 *to.16
0.99 50.10
1.09 kO.12
2.20 &IO.19
257 i-15
259 *13
1.80 h-o.14
0.99 kO.13
-0
1.01 10.12
1.89 *0.14
242 It11
1.74 Iko.17
1.78 hO.17 244 It11
T
T
T
-0
TG
Chol
TG
-0
VLDL
Total
TG = triglyceride, Chol = cholesterol.
Females 30-39 N=15T N= 12-O 40-49 N=l5T N= 12-O 50-59 N=24T N= 18-O 60-69 N=23T N= 13-O
Males 30-39 N=lST N=17-0 4049 N=l7T N= 13-O 50-59 N=25T N==17-0 60-69 N=20T N= 14-O
Groups
21 *3
15 12
15 *2
14 *2
16 +3 18 *2
10 %2
16 1-3
12 It2
18 It3
20 13
*3
26 It3
21
It3
It3
22
20
*3
-0
21
T
Chol
0.59 Yko.05
0.58 Ito.
0.51 &IO.04
0.41 +0.03
0.54 zko.03
0.55 kO.03
0.52 xto.04
0.50 kO.04
T
TG
LDL
0.53 50.06
0.55 zko.04
0.47 *to.03
0.39 *0.02
0.54 *0.04
0.53 kO.03
0.53 &-0.04
0.50 *0.04
-0.
Mean value & S.E.M. is given for total material (T) and after elimination of overweight persons (-0) above 1.10.
TRIGLYCERIDE
TABLE AND IN THE
3 MAJOR LIPOPROTEIN
CLASSES
ill
190
184 I!I~
159 1-12
133 +7
168 ik8
160 1t6
175 It10
162 +9
T
Chol
185 &-I1
0.25 &to.01
0.24 AO.02
0.29 *0.02
0.30 *0.01
187 +9
0.26 AO.02
0.25 zko.02
0.24 hO.01
0.24 +0.02
0.23 zto.01
-0
0.26 hO.02
0.26 zto.02
0.25 +0.02
0.26 &O.Ol
0.23 *to.01
0.23 Ito.
T
157 0.28 I!K15 kO.02
127 +6
173 &lo
158 +7
173 +12
162 19
-0
TG
HDL
69 &4
69 zt3
76 *5
75 13
66 *4
*3 65 -1-4
64 +3
60 +4
53 zt3
51 AI4
54 A3
-0
63
56 *4
51 ItI2
50 $_3
53 *3
T
Chol
defined as a weight/height index (kg/(cm -100))
SERUM
[lolog(10
VALUES
Females 3&39 N=15T N=12-0 40-49 N=15T N=12-0 5C59 N=24T N=18-0 60-69 N=23T N=13-0
Males 30-39 N= 18T N= 17-o 40-49 N=l7T N-13-O 50-59 N=25T N=17-0 60-69 N=20T N= 14-O
Groups
1.0722 10.1102
1.1395 *to.0854
0.8401 kO.0577
0.9077 zto.0570
1.1868 10.0396
1.2258 &IO.0387
0.8494 i 0.0598
1.1177 *0.0379 1.1745 +0.0365 1.1758 kO.0305
1.1572 +0.0468
1.2144 *0.0330
1.2367 ho.0308
0.9423 10.0485
0.8899 * 0.0470
0.7260 hO.0514
1.0594 *0.0273
1.0853 &0.0311
0.8582 io.0513
0.8341 *0.0496
0.8178 +0.0440
0.7239 +0.0290
0.7251 & 0.0248 1.0622 &0.0831 1.1476 kO.0513
1.1373 10.0707 1.2409 *0.0551
0.7341 Iko.0339
0.6619 ho.0321 0.6880 10.0331
1.1272 It 0.0493 1.1119 f0.0821
0.5828 *to.0214
0.7143 hO.0375
0.7106 kO.0235
0.7086 & 0.0302
0.6762 *to.0335
IN VLDL
0.6046 ztO.0241
0.7138 10.0282
0.7269 iO.0239
0.6970 &0.0313
0.6779 -CO.0316
T
(mg/lOOml)]
0.9420 10.0953 0.9836 hO.0917
1.2225 ho.0629
1.3302 &to.0563
0.9484 *0.0570
1.0405 10.0530
1.2474 10.0350
1.3058 10.0349
0.6880 &0.0506
1.2786 *0.0652
1.2937 50.0572
0.9645 10.0521
0.9975 kO.0488
1.2406 +0.0332
1.2573 zto.0319
1.2429 + 0.0648
1.2110 10.0391
1.2206 *0.0381
0.9414 =‘cOo.O528
1.2237 kO.0657
-0
Plog
LDL TG
CONCENTRATION
0.9531 10.0511
T
-0
T
T
-0
VLDL
VLDL TG
Total TG
Chol
VLDL AND LDL AND CHOLESTEROL
and symbols as in Table 1.
x mmol/l)]rN TOTALSERUM,
Triglyceride values multiplied by 10. Other conditions
ON LOGARITHMIC
CALCULATED
2
TRIGLYCERIDE CONCENTRATION
TABLE
r ?
QUANTITATIVE
-MALES
AND QUALITATIVE
(n.
SERUM LIPOPROTEIN
-FEMALES
60)
(n=77)
ANALYSIS,
PART
1
423
MEANtSEM
TOTAL CHOLESTEROL
TOTAL TRIGLYCERIDES
mg/lOOml 320 -I
-
70 AGE
Fig. 2. Mean concentration and females.
e-a
30
40
YEARS
of triglycerides and cholesterol in total serum in relation to age in males
MALES (n.60)
-
FEMALES (n=77)
VLDL TRIGLYCERIDES
VLDL
lolog (lO~mmol/l)
lolog mg/lOO ml
AGE
AGE
YEARS
of triglycerides and cholesterol in VLDL lipoprotein
MALES (n-60)
M
FEMALES (n-77)
LDL TRIGLYCERIDES
LDL CHOLESTEROL
lolog ( 10 ammol/ I )
mg/lOOml
0.8
200 &@=O
0.7
MEAN t SEM
CHOLESTEROL
YEARS
Fig. 3. Mean concentration males and females.
-
50 60 70 AGE YEARS
in relation to age in
MEAN* SEM
160
0.6 i -f I
30
I
40
I
50
I
60
1
70
AGE YEARS
Fig. 4. Mean concentration males and females.
I
30
I
40
I
50
60
1
70
AGE YEARS
of triglycerides and cholesterol in LDL lipoprotein
in relation to age in
424
L. A. CARLSON, M. ERICSSON .-.
HOL
MALES
(n-80)
TRIGLYCERIDES
mmol/l
o-o
FEMALES
HOL
CHOLESTEROL
(n=77)
+r
SEM
mg/lpOmI
30 .25 .20
AGE
AGE
YEARS
Fig. 5. Mean concentration males and females.
YEARS
of triglycerides and cholesterol in HDL lipoprotein
in relation to age in
were statistically significant. In whole serum triglycerides increased by 0.005 (P > 0.05) and 0.02 (P < 0.01) mmol/l/year in males and females respectively. Corresponding figures for cholesterol were 0.05 (P > 0.05) and 2.4 (P < 0.001) mg/lOO ml/year. In females, VLDL triglycerides and LDL cholesterol increased by 0.014 (P < 0.01) mmol/l and 2.0 (P < 0.001) mg/lOO ml/year. not vary with age.
HDL triglycerides
and cholesterol
did
Lipoprotein composition The relationship between cholesterol and triglycerides in males in the 3 LP classes are illustrated in Figs. 6, 7 and 8, and the linear regression analysis for males and females is given in Table 3. For VLDL there was a strong linear relation between the cholesterol and triglycerides content of this LP class. The relation contained the origin. There was also a significant positive linear relation for LDL. This, however, was not as strong as for VLDL and did not contain the origin. There was no relation between the cholesterol and the triglyceride content in HDL. A very similar relationship existed between cholesterol and triglycerides in the LP classes for female and for male subjects (Table 3).
. . . .**.* ..* /......’ VLDL
COMPOSITION.
??
??
0”
??
+:I . .
0.
4
.
.
:
Normal men .
??
r zO.90
a.
Fig. 6. Relationship
between the content of VLDL cholesterol
and triglycerides
in male subjects.
QUANTITATIVE
300
1
AND QUALITATIVE
LDL COMPOSITION.
Fig. 7. Relationship
SERUM LIPOPROTEIN
ANALYSIS,
PART
425
i
Normal men
.
between the content of LDL cholesterol and triglycerides in male subjects.
TABLE 3 LINEARRE~RESSIONANDCORRELATIONANALYSISF~RTHE
(Y)
AND TRIGLYCERIDES, InItIOl/l (X) IN THE
Females
b
VLDL LDL HDL
mg/lOOml,
CLASSES
Males
LP class
CHOLESTEROL,
RELATIONBETWEEN
LP
a
r
P
b
a
Y
P
19
1
0.90
< 0.001
25
-3
0.92
< 0.001
169
75
0.73
< 0.001
180
75
0.72
< 0.001
-16
56
-0.08
> 0.05
-10
70
-0.04
on Y-axis; r = linear regression
b = linear regression
coefficient; a = intercept statistical significance of b and r.
1 0.05
coefficient:
P =
Sex and lipoprotein concentration There were several differences between male and female subjects. The 2 most constant were : (l), Females had lower VLDL-triglyceride and cholesterol levels than
HDL COMPOSITION.
Normal
men
.* . 84.
??
.
-
..
r-0.08
O-5 TRIGLYCERIDES
Fig. 8. Relationship
m m.fi
between the content
of HDL cholesterol
and triglycerides
in male subjects.
426
L. A. CARLSON, M. ERICSSON
males except for those in the oldest age group (Fig. 3). This was reflected in the lower triglyceride cholesterol
levels in total serum in females (Fig. 4). (2), Females had higher HDL (all age groups) and triglyceride levels (with the exception of the oldest age
group) than males (Fig. 5). The most characteristic difference between males and females regarding LDL cholesterol levels was the rise with age in females but not in males, (Fig. 4).
which caused
the male and female curves
to intersect
at age 45-50 years
Influence of overweight Elimination of overweight glyceride as well as the VLDL
persons from the study lowered the total serum tritriglyceride and cholesterol levels in all male and
female age groups (Tables 1 and 2). The effect of this procedure was highly significant. There was no effect on total serum, LDL cholesterol or LDL triglyceride levels. HDL-cholesterol but not triglyceride levels were significantly increased when overweight persons were eliminated from the male and female groups. It can be seen that overweight healthy persons differed from the rest of the
-----TOTAL TOTAL
,,
MATERIAL
(1)
TRIGLYCERIDES
10
I5
2.0
2.5
3.0
-
OVERWEIGHT
TOTAL
CHOLESTEROL
EXCLUDED
C-0)
3.5”
mmolll
60.
T _-0
nz77 I-l:55
40_
-F1.0
1.5
2.0
2.5
3.0
3.5 mmolll
400 mg/lOOml
Fig. 9. Cumulative frequency distribution curves for triglycerides and cholesterol in total serum. Male values not corrected for age, female values corrected by linear regression equation to age 50 yr. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated.
QUANTITATIVE
__- _ VLDL
AND
QUALITATIVE
TOTAL MATERIAL
SERUM LIPOPROTEIN
(T)
-OVERWEIGHT
TRIGLYCERIDES
05
1.0
VLDL
1.5
2.0
ANALYSIS,
2.5
3.0mmollk
PART EXCLUDED
10
20
30
40
50
0.5
1.0
I I
I,
I
1.5
2.0
25
3.0mmolll
10
20
30
1 1,
40
60
70mg1100
ml
___--
I I 1 I ’ ,
’
t-0)
CHOLESTEROL
%------l?/
h
427
1
T “~77
-0 n- 55
,
,
50
60
(
70mgllOO
ml
Fig. 10. Cumulative frequency distribution curves for VLDL triglycerides and cholesterol. Male values not corrected for age, female values corrected by linear regression equation to age 50 yr. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated.
population cholesterol
by having higher total triglycerides, and lower HDL cholesterol levels.
Frequency distribution and population limits The cumulative frequency distribution
higher
VLDL
curves for cholesterol
triglycerides
and
and triglycerides
in total serum and in the LP classes are shown in Figs. 9, 10, 11 and 12. Since total, VLDL and LDL lipids increased with age in females, all the female values have been adjusted to age 50 yr by using the linear regression coefficients obtained for each lipid LP in relation to age. In general the curves for total triglycerides, VLDL and LDL have the typical appearance of a frequency distribution that is skewed to the right, i.e., the mean is greater than the median. An example of this can be seen in the median value (= 50 % cumulative %) for total triglycerides in males of I .5 mmol/l (Fig. 9), while the mean value was 1.9 (Table 1). The shape of the other curves indicates a more normal distribution. The effect of elimination of overweight persons from the group under study is easily seen. The upper part of the frequency curve for total triglycerides and for VLDL was shifted to the left. For HDL-cholesterol the lower part was shifted to the right, as expected. The LDL curves were not influenced by this elimination.
L.
428 ------
TOTAL
-
MATERIAL (T)
0.4
0.6
OVERWEIGHT
EXCLUDED c-01
_LDL CHOLESTEROL
LOL TRIGLYCERIDES
0.2
A. CARLSON, M. ERICSSON
0.8
1.0
1.2
1.4
mmol/l
a L
60
0.2
0.4
0.6
a8
1.0
12 1.4 mmol/l
100
150
200
250
300
350
mgllO0 ml
Fig. 11. Cumulative frequency distribution curves for LDL triglycerides and cholesterol. Male values not corrected for age, female values corrected by linear regression equation to age 50 yr. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated.
We have indicated in the figures the limits covering 95, 90 and 85 % of the population. In the non-overweight population the upper 90 % population limit was for males/females (aged 50 yr) : total triglycerides 2.5/2.0, total cholesterol 298/300, VLDL triglyceride 1.80/1.05, VLDL cholesterol 32/23, LDL triglyceride 0.69/0.69, LDL cholesterol 210/218, HDL triglyceride 0.32/0.34 and HDL-cholesterol 69/93. Floating-p, late pre-/3 and sinking pre-/I LP The number of these electrophoretic LP’s is given in Table 4. No floating-8 was observed. A late pre-p (see Fig. 1) was found in 35 ‘A of males and 25 ‘A of females. The sinking pre-/I was detected in 25 and 22 % of males and females. Overweight did not appear to be associated with these LP’s. Subjects with sinking pre-/I LP’s did not differ from other subjects with regard to lipid concentration in the 3 LP classes. However, males having the late pre-/3 LP differed from those without this LP (Table 4). The VLDL cholesterol content was raised, leading to an increased cholesteroltriglycerides ratio (P < 0.001). Men with late pre-p LP also had an increased amount of LDL triglycerides.
QUANTITATIVE
AND QUALITATIVE
-------
TOTAL
HDL
TRIGLYCER~S
MATERIAL
SERUM LIPOPROTEIN
(1)
-
ANALYSIS,
OVERWEIGHT
PART EXCLUOEO
429
1 t-0)
HOL CHOLESTEROL
6. LO
P
20
OA
05
# LA!2 I
1 ns77 -0 n=55
/
,
a-
’
30
0.6
nmmlll
o i
.*
40
/
, .*’
-0’
50
60
70
60 90 100 mg/lOO ml
Fig. 12. Cumulative frequency distribution curves for HDL-triglycerides and cholesterol. No age corrections. Curves are given for total material, and after exclusion of overweight subjects. The upper 85, 90 and 95th percentiles are indicated. TABLE 4 NUMBER
OF SUBJECTS
WITH
LATE
PRE-B
AND SINKING
PRE-B
LP
Subjects
N”
Late pre-/3
Sinking pre-B
Males Males (-obese)
55 42
19 16
14 IO
Females Females (-obese)
50 38
13 9
11 8
8 Number of subjects with sufficiently complete analysis to permit visual assessment of the presence of late pre-/3 and sinking pre-/? LP. For explanation of procedure see Methods and Fig. 1.
DISCUSSION
Age Females (but not males) showed statistically significant increases in serum, VLDL and LDL cholesterol and triglyceride concentrations with age. The finding that males did not show any increase might at a first sight appear contradictory to earlier studies on serum lipids. However, the increase in males is most pronounced in the
430 TABLE
L. A. CARLSON,
hf. ERICSSON
5
LIPOPROTEIN CHOLESTEROL MEAN VALUE
COMPOSITION (mg/lm &
Groups
Males No late pre-8 N= 36 Late pre-/I N= 19 Females No late pre-b N = 34 Late pre-/I N= 13
ml)
1N SUBJECTS
WITH
CONCENTRATION
AND
WITHOUT
IN THE
3
LP
LATE
CLASSES
PRE-j!i. AND
TRIGLYCERIDE RATIO
(mIIlOl/l)
OF CHOL/TG
IN
AND
VLDL.
S.E.M. VLDL
LDL
HDL
TG
Chol
Chol/TG
TG
Chol
1.15*0.11
21+2
18&l
0.51 + 0.03
161 f
1.23*0.18
28&4
2441
0.60 A 0.04
0.83 i
16h
0.09
0.70*0.10
14*3
2
19+ 19*2
1
TG
Chol
0.25 zt 0.01
52 f
171 4 7
0.27 f
0.02
51 * 2
0.50 & 0.03
160 + 7
0.28 zt 0.01
67 + 3
0.55 & 0.04
176 i_ 13
0.27 5 0.02
64 + 4
6
2
TG = triglycerides, Chol = cholesterol.
younger age groups. Carlson thus found that in 26-39-yr-old males, cholesterol and triglyceride levels increased about 10 times more per year than in males aged 40-72 yrs. In the Stockholm Prospective Study, cholesterol and triglyceride levels increased only until about age 50 yr in males while they continued to rise in femaleslO. This greater increase with age in women is illustrated by the values for the mean cholesterol content in age groups 25-30/6&65 yr, which were 2291331 in females and 2461278 mg/lOO ml in maleslo. Corresponding triglyceride values were 1.00/1.22 and 1.39/1.32 mmol/l. The present material is limited and the study begins with subjects of age 30 yr. This explains why statistically significant increases with age were obtained here for females but not for males. Sex The observed higher serum triglyceride levels in males and the similar cholesterol levels in the 2 sexes is in agreement with earlier and more extensive studieslOJ1. The sex difference in total serum triglycerides can now be explained at the lipoprotein level; it is entirely due to higher VLDL levels in men than in women, LDL triglycerides being similar in the 2 sexes. As a result of the lower VLDL levels, the VLDL cholesterol content is lower in females than in males. This difference is, however, not observable in total serum cholesterol (which is the sum of VLDL, LDL and HDL cholesterol), because of the higher HDL cholesterol levels in females. A higher amount of cholesterol in females than in males in the Cohn fraction IV + V + VI, corresponding to HDL, was reported in 1951 by Russ et al. 12. The increased amount of HDLin females has been ascribed to an increase of the HDLa fractionls.
QUANTITATIVE
AND QUALITATIVE
SERUM LIPOPROTEIN
ANALYSIS,
PART
1
431
Overweight It is known
from
several
studies
that there is a positive
correlation
between
fasting serum triglyceride concentrations and various measures or indices of adipose tissue mass10J4J5. Either a weaker10 or no correlation16 has been found between serum cholesterol and adiposity. In the present study, overweight influenced serum triglyceride but not cholesterol levels. The cumulative frequency distribution showed that the influence of overweight only operated at triglyceride concentrations above the median value (Fig. 6). Approximately 90% of overweight subjects had serum triglyceride levels above the median value. The findings for serum lipids are explained by the fact that overweight influenced VLDL but not LDL concentrations. Elimination of overweight subjects shifted the lower part of the HDL cholesterol distribution curve to the right but had no effect on HDL triglyceride levels (Fig. 9). This effect may be explained by the usual negative correlation between VLDL and HDL cholesterol. Exclusion of overweight subjects from the present study was mainly an exclusion of persons with high VLDL- and thus probably also low HDL cholesterol concentrations. Normal values for LP classes One object of this study was to establish “normal” limits for the concentration and composition of the different lipoprotein classes. The limits can easily be obtained from the cumulative diagrams and depending on the way in which normality is defined, one can select any desired upper - or lower - cutting point. There are 2 major advantages in using cumulative frequency curves rather than the mean value and multiples of the standard deviation for finding upper normal limits of the population. Firstly, it does not matter what kind of distribution (normal, skewed to the left or right) we are dealing with; one just enters the curve at the required level. Secondly, single very extreme values which may influence the mean and particularly the standard deviation do not influence the value for e.g., the 95th or lower percentiles. Sinking pre-/3 LP Subjects with sinking pre-p LP had the same amount of cholesterol and triglycerides in the 3 LP classes as other subjects, suggesting that neither LP metabolism nor LP composition were greatly altered in the former group. The sinking pre-/? LP appears to be the same lipoprotein as the so-called Lp(a) discovered in Norway by BergI’. Using an immunologic technique, the frequency of Lp(a) was found to be 35 % in Norway’s. Our finding of 25 and 22 ‘A of sinking pre-@ LP in males and females (Table 3) is lower, but as mentioned previously, immunologic techniques were used in the Norwegian study which may have been more sensitive than our ultracentrifugal-electrophoretic visual inspection method. In a study with a visual inspection technique made in the U.S.A., the frequency of sinking pre-p LP was 25 %lg, a figure similar to that obtained in the present study. However, a recent study from Seattle using sensitive immunologic technique reported a frequency of Lp(a) of
432 approx.
L. A. CARLSON, bf. ERICSSON 80°A20. In northern
Sweden
an extra pre-p
lipoprotein
has been found in approx. 15-20 % of healthy male subject@i. band with sinking pre-p LP has, however, not been established. Late pre-p LP Males with late pre-p
LP had an increase
in VLDL
known
as pre-11,
The identity
of this
cholesterol-triglycerides
ratio. This suggests that the late pre-/3 has a high cholesterol-triglycerides males this LP, being visible in about one-third of the population, would
ratio. In then have
the following characteristics : (1) d < 1.006, (2), Electrophoretic mobility in agarose close to p-LP, and (3) High cholesterol-triglycerides ratio. The late pre-@ LP thus resembles the floating or broad Type III B-LP. The additional finding that subjects with a late pre-@ LP had increased LDL triglyceride content but normal cholesterol levels suggests that they had increase in the Sf 12-20 LP or LDLi LP. This increase, as well as the presence of late pre-/3 LP, may be due to an increased amount of socalled remnants or intermediary particles in fasting blood. Remnants are believed to represent degradation products of normal chylomicra and VLDL that accumulate in Type III due to a catabolic defect 22. The presence of such LP’s in about one-third of the male population in the presence of normal VLDL triglycerides is an interesting possibility. The reproducibility and genetics of the occurrence of the late pre-p LP in a subject, as well as its physiologic and clinical importance, has to await further studies. In his original description of the electrophoretic separation of lipoproteins in Agarose gel, Noble5 noted that “plasma samples from different subjects show pre-,!l lipoproteins of different mobilities”. No further characterization of these pre-B LP’s has so far appeared. The fact that these pre-j3 LP’s5 were easily seen in electrophoresis of whole serum makes it likely that they correspond to the sinking pre-/I LP, as this pre-p-abnormality is easily seen in whole serum (Fig. l), while the late pre-/3 LP is more clearly seen in electrophoresis of the VLDL fraction (Fig. 1). Lipoprotein composition The positive relationship between VLDL cholesterol and triglyceride content, containing the origin, indicates that this LP class was composed of homogenous particles. When the VLDL concentration increased, this relation suggested that the increase was due to an increased number of VLDL particles of identical composition (at least with regard to cholesterol/triglycerides). There was also a significant positive relation between LDL cholesterol and triglyceride content. However, this was not as strong nor did it contain the origin. This was probably because the LDL fraction, defined by density 1 JO61.063, was composed of at least 2 rather different LP classes, LDLr (d1.006-1.019) andLDLs(d1.019-1.063). The former is more triglyceride-rich and the latter more cholesterol-rich, and their concentrations may vary independently. Interestingly there was no relation between HDL cholesterol and triglyceride content, indicating that these 2 lipid components vary independently of each other. Since few studies have been made on the triglyceride content of HDL, it is difficult
QUANTITATIVE
AND
QUALITATIVE
SERUM LIPOPROTEIN
ANALYSIS,
PART
433
1
to evaluate this. Also since this fraction contains all material with d > 1.063 in plasma, caution should be exercised as the small amount of triglyceride found by the present methods of determination has not been chemically identified as triglyceride.
REFERENCES CARLSON, L. A., Serum lipids in men
with myocardial
infarction,
Actu Med. Stand., 167 (1960)
399. GOLDSTEIN, J. L.,
HAZZARD, W. R., SCHROTT, H. G., BIERMAN, E. L. AND MOTULSKY, A. G., Hyperlipidemia in coronary heart disease, J. C/in. Invest., 52 (1973) 1533-1577. CARLSON, K., Lipoprotein fractionation, J. C/in. Path., Suppl. 5, 26 (1973) 32. CARLSON, L. A., The effect of nicotinic acid treatment on the chemical composition of plasma lipoprotein classes in man - Drugs affecting lipid metabolism, Advunc. Exptl. Med. Biol., 4 (1969)
327. separation of plasma lipoproteins in agarose gel, J. Lipid Res., 9 5 NOBLE, R. P., Electrophoretic (1968) 693. 6 RUSH, R. L., LEON, L. AND TURRELL, J., Automated simultaneous cholesterol and triglyceride determination on the Auto Analyzer@ II instrument. In: Advances in Automated Analysis, Vol. Z, Thurman,
1971, p. 503.
of serum triglycerides, J. Atheroscler. Res., 3 (1963) 334. 7 CARLSON, L. A., Determination 8 ERICSSON, P., Lung Function and Physical Work Capacity in Healthy Elrler1.vPeople (Thesis 102), Acta Univ. Upsal., 1971. 9 CARLSON, L. A., Serum lipids in normal men, Acta Med. Stand., 167 (1960) 377.
No.
10 CARLSON, L. A., AND LINDSTEDT, S., The Stockholm Prospective Study, Part 1 (The initial values for plasma lipids), Actu Med. Scund., Suppl. No 493, 1969. 11 WOOD, P. D. S., STERN, M. P., SILVERS, A., REAVEN, G. M. AND GROEBEN, J., Prevalence of plasma lipoprotein abnormalities in a free-living population of the Central Valley, California,
Circulation, 45 (1972) 114. 12 Russ, E. M., EDER, H. A. AND BARR, D. P., Protein-lipid relationship in human plasma, (In normal individuals), Amer. J. Med., 11 (1951) 468. 13 BARCLAY, M., BARCLAY, R. K. AND SKIPSKI, V. P., High density lipoprotein concentration
Part
I
in men
and women, Nafure (Lond.), 200 (1963) 362. 14 ALBRINK, M. J. AND MEIGS, J. W., Interrelationship between skinfold thickness, serum lipids and blood sugar in normal men, Amer. J. Clin. Nutr., 15 (1964) 255. 15 BIERMAN, E. L., PORTE, D. AND BAGDADE, J. P., Hyperglyceridemia and glucose intolerance in man, In : B. JEANRENAUD AND D. HEPP (Eds.), Adipose Tissue, Regulation and Metabolic Functions, Academic Press, New York, N.Y., 1970, p. 209. 16 KEYS, A. et al., Coronary heart disease among Minnesota business and professional men followed 15 years, Circulation, 28 (1963) 381. 17 BERG, K., The Lp system, Ser. Haematol., 1 (1968) I1 I. 18 MONN, E., BERG, K., REINSKON, T. AND TEISBERG, P., Serum protein polymorphisms among Norwegian Lapps, Human Hered., 21 (1971) 134. 19 ELLEFSON, R. D., JIMENEZ, B. S. AND SMITH, R. C., Pre-P lipoprotein of high density in human blood, Mayo C/in. Proc., 46 (1971) 328. 20 ALBERS, J. AND HAZZARD, W., Quantitative genetic analysis of the human plasma Lp(a) lipoprotein. In G. SCHETTLER AND A. WEIZEL (Eds.), Atherosclerosis ZZZ,Springer, Berlin, 1974, p. 469. 21 DAHL~N, G., ERICSON, C., FORSBERG, C., LUNDKVIST, L. AND SVKRDSUDD, K., Studies on an extra pre-/? lipoprotein fraction, Actu Med. &and., Suppl. 53 1, 1972. 22 HAZZARD, W. R., PORTE, D. AND BIERMAN, E. L., Abnormal lipid composition of chylomicrons in broad-p disease (Type III hyperlipoproteinemia), J. C/in. Invest., 49 (1970) 1853.
