- Email: [email protected]
Plasma vitamin E, apolipoprotein B and HDL-cholesterol in middle-aged men from Southern Italy
25
Atherosclerosis, 17 (1989) 25-29 Elsevier Scientific Publishers Ireland, Ltd.
ATH 04299
Plasma vitamin E, apolipoprotein B and HDL-cholesterol in middle-aged men from Southern Italy Paolo Rubba ‘, Mario Mancini ‘, Flarninio Fidanza 2, Giovanni Leccia ‘, Rudolph A. Riemersma 3 and K. Fred Gey 4 ’Institute of Internal Medicine and Metabolic Diseases, 2nd Medical School, Universiiy of Naples, Naples (Italy), ’ Institute of Nutrition and Food Science, University of Perugia, Perugia (Italy), 3 Cardiovascular Research Unit, University of Edinburgh, Edinburgh (U.K.), and 4 Vitamin Research Dept., F. Hoffmann-L.a Roche and Co. Ltd., Base1 (Switzerland) (Received 5 May, 1988) (Revised, received 19 December, 1988) (Accepted 22 December, 1988)
Plasma vitamin E, HDL-cholesterol, apolipoprotein B and triglycerides were measured in an apparently healthy, male, random population sample (n = 74) from Southern Italy. Plasma vitamin E concentration was positively correlated to that of serum cholesterol, non-HDL cholesterol, triglycerides and apolipoprotein B (all P < 0.001). The results of partial correlation analysis showed that apo B, the apolipoprotein constituent of LDL, was related to vitamin E independently of serum triglycerides, a fairly accurate marker of VLDL. On the other hand, triglycerides were related to vitamin E independently of apo B. Both correlations were much weaker if an adjustment was performed for non-HDL-cholesterol. No independent relationship was demonstrated between plasma vitamin E and HDL-cholesterol.
Key words: Vitamin E; Apolipoprotein
B; HDL-cholesterol;
Introduction
There is in vitro evidence that oxidative modification of low density lipoprotein (LDL) with generation of lipid peroxide leads to impaired catabolism of modified LDL, along the receptor
Correspondence to: Dr. P. Rubba, Clinica Medica Institute of Internal Medicine and Metabolic Diseases, 2nd Medical School, University of Naples, Via S. Pan&i 5, 80131 Naples, Italy. 0021-9150/89/$03.50
Triglycerides
mediated pathway, while uptake and degradation of these lipoprotein particles by macrophages are enhanced [l]. This mechanism might contribute to atherogenesis in vivo. It has been demonstrated that overnight incubation of LDL with cultured endothelial or smooth muscle cells produces oxidative modifications of LDL. The cell-modified LDL are rapidly taken up by the same receptor that recognizes the chemically modified LDL [2]. Addition of vitamin E to the endothelial cell culture medium completely abolishes the generation of lipid peroxide
0 1989 Elsevier Scientific Publishers Ireland, Ltd.
26 in the LDL [3]. Through this mechanism, antioxidants might inhibit the accumulation of modified LDL by macrophages and thereby foam cell formation [4]. Although in vitro evidence suggesting the operation of this mechanism is strong and consistent, it is not known to what extent oxidative modifications occur in vivo. Natural antioxidants, such as vitamin E, are nevertheless very likely to antagonize the process of oxidative modification of LDL also in vivo. Vitamin E is carried in plasma within the lipoprotein particles [5-S]. However, it is not clear to what extent total plasma vitamin E, in the general population, is related to very low density (VLDL), low density (LDL) or high density (HDL) lipoprotein levels. This question has been addressed by evaluating plasma vitamin E and serum HDL-cholesterol, apolipoprotein B and triglycerides in a population sample from Southern Italy. The present investigation was part of an international collaborative study on fatty acids/antioxidant hypothesis of arteriosclerosis [9].
Duplicate performed ter.
measurements of blood pressure were with a random zero sphygmomanome-
Methods Blood was obtained by venous puncture using 2 Vacutainer tubes, one (for plasma) with lithium heparinate, the second (for serum) without additives. Samples of plasma for analysis of vitamin E were immediately frozen, stored in the laboratory in Naples and flown frozen on dry ice (-56°C) to the analysing laboratory in Basel, Switzerland, within 4 weeks after collection. Plasma vitamin E was determined by an HPLC method [lo], using a fluorimetric detector (excitation 290 nm, emission 330 nm). The coefficient of variation for the estimation of vitamin E was 3.0%. Serum cholesterol (Chol) and triglycerides (TG) were estimated by enzymatic methods [ll]. HDL-cholesterol was determined enzymatically after precipitation of VLDL and LDL by Mg2+dextran [12]. Serum apolipoprotein B concentration was determined by radial immunodiffusion
[131. Population sample A random sample of 120 men in the age range 40-49 years was selected from rural and semiurban communities of Southern Italy (Comunita’ Bussentina, Sapri). Five men had left the country and 115 were invited to attend a clinic: 79 agreed to participate in the study. Five men reported a history of atherosclerotic cardiovascular complications (2 myocardial infarction, 2 angina pectoris, 1 intermittent claudication). They had changed their lifestyles after the development of overt myocardial or lower limb ischaemia and were therefore excluded. The final sample size was 74 (64%). The men attended the clinic in the morning. All the participants were fasting except for 1 or 2 small cups of coffee; no other food had been consumed in the preceding 12 h. A questionnaire was completed on demographic data and social information, past medical history, smoking habits, drug therapy, medically prescribed diet and physical exercise. Anthropometric measurements included height and weight. Body mass index (BMI) was calculated from weight/height 2 (kg/m2).
The distribution of many variables was skewed to the right; therefore median and ranges were also reported and correlations were calculated from log transformed values. The log transformation was used in order to bring the data distributions, which were skewed to the right, closer to normality (as required for the statistical calculations). The statistical analysis included simple linear correlations and partial correlations [14,15]. Results Table 1 illustrates the mean values for anthropometric data and serum lipid values of the population sample under study; apolipoprotein B concentration and calculated values of non-HDLcholesterol were also reported. Table 2 illustrates the direct correlations which exist between plasma vitamin E and total, nonHDL- and HDL-cholesterol in serum. Both total and non-HDL-cholesterol were positively related to vitamin E (P < 0.001). On the other hand, HDL-cholesterol was inversely and weakly related to vitamin E concentration (Table 2). These correlations were obtained by using log transformed
27 TABLE 1 ANTHROPOMETRIC DATA AND LIPIDS, LIPOPROTEINS HEALTHY MEN FROM SAPRI, SOUTHERN ITALY
AND
VITAMIN
E CONCENTRATIONS
IN APPARENTLY
Age range 40-49 yrs; n = 74.
Mean SD Median Min Max
BMI
Blood pressure (mm Hg)
Serum
(kg/m2 )
systolic
Cholesterol (mg/dl)
Apo B
TG
Total
HDL
Non-HDL
(mg/dl)
(mg/dl)
203.1 44.0 202 118 317
45.8 12.8 44 26 76
156.9 47.7 156 66 287
88 19.2 86 45 141
157.1 146.8 126 55 1138
26.2 3.1 26.1 20.3 34.5
128.9 13.8 121 100 168
Diastolic
84.4 8.6 84 68 114
Vit. E (PM/~)
25.1 7.0 24.6 12.3 47.3
TABLE 2 CORRELATIONS (r COEFFICIENTS) BETWEEN SERUM CHOLESTEROL (Chol) VALUES (LOG TRANSFORMED) AND PLASMA VITAMIN E (LOG TRANSFORMED) IN APPARENTLY HEALTHY MEN FROM SAPRI, SOUTHERN ITALY Age range 40-49 yrs; n = 74. Vitamin E
w
0.761 * * * - 0.293 * 0.763 * * *
5
2
Total Chol HDL Chol Non-HDL Chol * PiO.05;
.* .
15
t II
h
.
SERlN
.
47 .
.
.
6%
***P
values. The correlation coefficients were almost the same when non-transformed values were used in the calculations (not shown). Figures 1 and 2 describe the positive associations which exist between vitamin E and apoli-
.
.
.
440
174 TRIGLYCERIDES
I mQ/dl
Fig. 2. Direct correlation between serum triglycerides and plasma vitamin E.
poprotein B or triglycerides (log transformed values). Each point represents one subject. Both correlations were highly significant (P < 0.001). TABLE 3 PARTIAL CORRELATIONS (r COEFFICIENTS) BETWEEN LIPOPROTEIN CONSTITUENTS (LOG TRANSFORMED) AND VITAMIN E (LOG TRANSFORMED) IN APPARENTLY HEALTHY MEN
f
Age range 40-49 years; n = 74.
rz.647 pt. 001
‘46
DO
70 APOLIPOPR~IEIN
B
I
1100
1
140
mg /dl)
Fig. 1. Direct correlation between serum apolipoprotein B and plasma vitamin E.
Variables correlated
Variable kept constant
r
Apo TG, Apo TG,
TG Apo B Non-HDL-Chol Non-HDL-Chol
0.575 *** 0.471 * * * 0.181 0.268 *
B, Vit. E Vit. E B, Vit. E Vit. E
* PcO.05;
*** P
28 Table 3 gives the results of a partial correlation analysis demonstrating that the direct correlations between vitamin E and apolipoprotein B or triglycerides were independent of each other. Both correlations almost completely disappeared, if an adjustment was performed for non-HDL-cholesterol. The weak inverse correlation between vitamin E and HDL was no longer present, when the influence of differences in apolipoprotein B or triglycerides was eliminated (not shown). Discussion This study demonstrates that plasma vitamin E concentration is positively correlated to that of non-HDL-cholesterol, triserum cholesterol, glycerides or apolipoprotein B, in an apparently healthy, male adult population. The inverse relation between vitamin E and HDL-cholesterol disappeared when plasma triglycerides and apolipoprotein B levels were taken into consideration. The correlations are in agreement with the view that vitamin E, cholesterol, triglycerides and apolipoprotein B are normal constituents of the VLDL and LDL particles in vivo. Apolipoprotein B levels have been shown to be strongly correlated to LDL-cholesterol [16]. Both apolipoprotein B and LDL-cholesterol are fairly accurate markers of circulating LDL particles, while fasting serum triglycerides mainly reflect VLDL levels. The results of partial correlation analysis suggest that both VLDL and LDL concentrations influence the vitamin E levels in the plasma of adult men. In fact apo B, the apolipoprotein constituent of LDL, is related to vitamin E independently of serum triglycerides; on the other hand, triglycerides are related to vitamin E independently of apo B. The relations between vitamin E and apolipoprotein B or triglycerides have been demonstrated in a sample of apparently healthy men with normal levels of serum lipids and lipoproteins. It is therefore likely that they reflect physiological conditions occurring in vivo in men. It has been demonstrated that LDL delivers vitamin E to the cells through a pathway which is dependent on the high affinity apo B/E receptor [17]. There is, however, also another mechanism underlying the delivery of vitamin E to peripheral tissues, which involves triglyceride-rich lipoproteins and lipoprotein lipase
[18]. On the basis of in vitro observations it has been estimated that the latter mechanism is ten times more efficient than the former [18]. It should be considered that lipoprotein lipase activity has been demonstrated in the arterial wall [19], where it is bound to the endothelial cells [20]. It is therefore localized in a crucial site for possible lipoprotein/endothelial cell interactions, leading to oxidative modifications of LDL. It has been shown that much of the variability of plasma vitamin E levels is due to differences in the serum cholesterol levels [21]. However, we have demonstrated that triglycerides and VLDL play an important role too. In a small series of hypertriglyceridemic patients (type IIB, IV, V, I) very high amounts of vitamin E were found in VLDL and chylomicrons [7]. On the other hand, in conditions of hypercholesterolemia with normal triglycerides (type IIA) plasma vitamin levels were raised due to a marked increase of vitamin E in the LDL fraction. Despite the fact that HDL carry a significant proportion of vitamin E [5,22], HDL-cholesterol levels do not seem to be much related to the variability of total plasma vitamin E. This notion is in agreement with the observation [7] that in hypertriglyceridemic patients, who have in general low HDL-cholesterol, total serum vitamin E is higher than in normolipidemic controls. In order to express vitamin E levels independently of lipoprotein carriers levels, some form of adjustment must be made. The adjustment is needed in order to account, in each individual, for the different degree of saturation of plasma lipids by vitamin E, which is a major factor for the thermodynamic partitioning of vitamin E between different lipid compartments such as plasma and arterial wall. In overt hyperlipemia only the vitamin E value, adjusted for VLDL, LDL and eventually chylomicron levels, reflects the vitamin status in human erythrocytes or animal tissues [8]. In some cross-cultural comparisons [23] the vitamin E levels have been standardized for serum cholesterol by using a molar ratio vitamin E/total cholesterol. On the basis of our correlation data, we suggest that there is an advantage of standardizing vitamin E levels for apolipoprotein B and triglycerides as well as for non-HDL-cholesterol. This last parameter should probably be preferred,
29 because it reflects much of the variability of both VLDL and LDL levels. In fact partial correlation analysis has shown that both the relations between vitamin E and apolipoprotein B or serum triglycerides almost completely disappear, when non-HDL-cholesterol levels are taken into consideration (Table 3). Non-HDL-cholesterol can be accurately measured with a simple equipment and at a lower cost than apolipoprotein B concentration. The main conclusion of the present study is that plasma levels of vitamin E, a major natural antioxidant, are markedly influenced by both LDL and VLDL concentrations, not only in patients with overt hyperlipidemia but also in the general male population. In all situations where an assessment of vitamin E status is needed, it is important to perform an adjustment for the levels of lipoprotein carriers. A simple way of dealing with this problem is to use the vitamin E/total cholesterol ratio. More accurate statistical approaches, such as analysis of covariance with an adjustment of vitamin E for non-HDL-cholesterol, are to be preferred in population studies.
6 7 8 9
10
11
12
13
14
Acknowledgements We are indebted to Mr. H. Georgi for technical assistance and to Dr. J.P. Vuilleumier for vitamin analysis (Vit. Res. Dept., Hoffmann-La Roche and Co., Ltd., Basel, Switzerland).
15 16
17
References 18 1 Morel, D.W., Di Corleto, P.E. and Chisolm, D.M., Endothelial and smooth muscle cells alter low density lipoprotein in vitro by free radical oxidation, Arteriosclerosis, 4 (1984) 357. 2 Hemiksen, T., Mahoney, E.M. and Steinberg, D., Enhanced macrophage degradation of low density lipoprotein previously incubated with cultured endothelial cells: recognition by the receptor for acetylated low density lipoproteins, Proc. Natl. Acad. Sci. USA, 78 (1981) 6499. 3 Steinbrecher, U.P., Parthasarathy, S., Leake, D.S., Witzum, J.L. and Steinberg, D., Modification of low density lipoprotein by endothelial cells involves lipid peroxidation and degradation of low density lipoprotein phospholipids, Proc. Natl. Acad. Sci. USA, 81 (1984) 3883. 4 Esterbauer, H., Juergens, G., Anchenberger, 0. and Keller, E., Autoxidation of human low density lipoprotein: loss of polyunsaturated fatty acids and vitamin E and generation of aldehydes, J. Lipid Res., 28 (1987) 495. 5 Takahashi, Y., Urono, K. and Kimura, S., Vitamin E
19 20
21
22
23
binding proteins in human serum, J. Nutr. Sci. Vit., 23 (1977) 201. Bieri, J.G., Corash, L. and van Hubbard, S., Medical uses of vitamin E, New Engl. J. Med., 308 (1983) 1063. Lambert, D. and Mourot, J., Vitamin E and lipoproteins in hyperlipoproteinemia, Atherosclerosis, 53 (1984) 327. Gey, K.F., On the antioxidant hypothesis with regard to arteriosclerosis, Bibl. Nutr. Dieta, 37 (1986) 53. Riemersma, R.A., Wood, D.A., Butler, S., Elton, R.A., Oliver, M.F., Salo, M., Nikkari, T., Vartiainen, E., Puska, P., Gey, F., Rubba, P., Mancini, M. and Fidanza, F., Linoleic acid content in adipose tissue and coronary heart disease, Br. Med. J., 292 (1986) 1423-1427. Vuilleumier, J.-P., Keller, H.E., Gysel, D. and Hunziker, F., Clinical chemical methods for the routine assessment of the vitamin status in human populations, Int. J. Vit. Nutr. Res., 53 (1983) 265. Oriente, P., Di Marino, L., Mastranzo, P., Iovine, C. and Patti, L., Simultaneous determination of cholesterol and triglycerides in serum and lipoprotein fractions with enzymatic automated methods. In: A. Burlina and L. Galzigna (Eds.), Clinical Enzymology Symposium 2, Piccin, Padua, 1979, p. 387. Kostner, G.M., Enzymatic determination of cholesterol in high-density lipoprotein fractions prepared by polyanion precipitation, Clin. Chem., 22 (1976) 695. Mancini, Y., Carbonara, A.O. and Heremans, J.F., Immunochemical quantification of antigens by simple radial immunodiffusion, Immunochemistry, 2 (1965) 235. Dixon, W.J. and Massey, F.J., Introduction to Statistical Analysis, McGraw-Hill, New York, 1969. Spiegel, M.R., Statistics, McGraw-Hill, New York, 1961. Olsson, A.G., Holmquist, L., Walldius, G., Hadell, K., Carlson, L.A., Ricciardi, G., Rubba, P., Pauciullo, P. and Mancini, M., Serum apolipoproteins, lipoproteins and fatty acids in relation to ischemic heart disease in Northern and Southern European males, Acta Med. Stand., 223 (1988) 3. Traber, M.G. and Kayden, H.J., Vitamin E is delivered to cells via the high affinity receptor for low density lipoproteins, Am. J. Clin. Nutr., 40 (1984) 747. Traber, M.G., Olivecrona, T. and Kayden, H.J., Bovine milk lipoprotein lipase transfers tocopherol to human fibroblasts during triglyceride hydrolysis in vitro, J. Clin. Invest., 75 (1985) 1729. Zilversmit, D.B., Atherogenesis: a postprandial phenomenon, Circulation, 60 (1979) 473. Olivecrona, T. and Bengtsson, G., Heparin and lipoprotein lipase - how specific is the interaction? In: International Conference on Atherosclerosis, Carlson, L.A., Paoletti, R., Sirtori, C.R. and Weber, G. (Ed), Raven, New York, 1978, p. 153. Stahelin, H.B., Rosel, F., Buess, E. and Brubacher, G., Cancer, vitamins and plasma lipids: Prospective Base1 study, J. Natl. Cancer Inst., 73 (1984) 1463. Bjomson, L.K., Kayden, H.J., Miller, E. and Moshell, A.N., The transport of alpha tocopherol and beta carotene in human blood, J. Lipid Res., 17 (1976) 343. Gey, K.F., Brubacher, G.B. and Stahelin, H.B., Plasma levels of antioxidant vitamins in relation to ischemic heart disease, Am. J. Clin. Nutr., 45 (1987) 1368.
Atherosclerosis, 17 (1989) 25-29 Elsevier Scientific Publishers Ireland, Ltd.
ATH 04299
Plasma vitamin E, apolipoprotein B and HDL-cholesterol in middle-aged men from Southern Italy Paolo Rubba ‘, Mario Mancini ‘, Flarninio Fidanza 2, Giovanni Leccia ‘, Rudolph A. Riemersma 3 and K. Fred Gey 4 ’Institute of Internal Medicine and Metabolic Diseases, 2nd Medical School, Universiiy of Naples, Naples (Italy), ’ Institute of Nutrition and Food Science, University of Perugia, Perugia (Italy), 3 Cardiovascular Research Unit, University of Edinburgh, Edinburgh (U.K.), and 4 Vitamin Research Dept., F. Hoffmann-L.a Roche and Co. Ltd., Base1 (Switzerland) (Received 5 May, 1988) (Revised, received 19 December, 1988) (Accepted 22 December, 1988)
Plasma vitamin E, HDL-cholesterol, apolipoprotein B and triglycerides were measured in an apparently healthy, male, random population sample (n = 74) from Southern Italy. Plasma vitamin E concentration was positively correlated to that of serum cholesterol, non-HDL cholesterol, triglycerides and apolipoprotein B (all P < 0.001). The results of partial correlation analysis showed that apo B, the apolipoprotein constituent of LDL, was related to vitamin E independently of serum triglycerides, a fairly accurate marker of VLDL. On the other hand, triglycerides were related to vitamin E independently of apo B. Both correlations were much weaker if an adjustment was performed for non-HDL-cholesterol. No independent relationship was demonstrated between plasma vitamin E and HDL-cholesterol.
Key words: Vitamin E; Apolipoprotein
B; HDL-cholesterol;
Introduction
There is in vitro evidence that oxidative modification of low density lipoprotein (LDL) with generation of lipid peroxide leads to impaired catabolism of modified LDL, along the receptor
Correspondence to: Dr. P. Rubba, Clinica Medica Institute of Internal Medicine and Metabolic Diseases, 2nd Medical School, University of Naples, Via S. Pan&i 5, 80131 Naples, Italy. 0021-9150/89/$03.50
Triglycerides
mediated pathway, while uptake and degradation of these lipoprotein particles by macrophages are enhanced [l]. This mechanism might contribute to atherogenesis in vivo. It has been demonstrated that overnight incubation of LDL with cultured endothelial or smooth muscle cells produces oxidative modifications of LDL. The cell-modified LDL are rapidly taken up by the same receptor that recognizes the chemically modified LDL [2]. Addition of vitamin E to the endothelial cell culture medium completely abolishes the generation of lipid peroxide
0 1989 Elsevier Scientific Publishers Ireland, Ltd.
26 in the LDL [3]. Through this mechanism, antioxidants might inhibit the accumulation of modified LDL by macrophages and thereby foam cell formation [4]. Although in vitro evidence suggesting the operation of this mechanism is strong and consistent, it is not known to what extent oxidative modifications occur in vivo. Natural antioxidants, such as vitamin E, are nevertheless very likely to antagonize the process of oxidative modification of LDL also in vivo. Vitamin E is carried in plasma within the lipoprotein particles [5-S]. However, it is not clear to what extent total plasma vitamin E, in the general population, is related to very low density (VLDL), low density (LDL) or high density (HDL) lipoprotein levels. This question has been addressed by evaluating plasma vitamin E and serum HDL-cholesterol, apolipoprotein B and triglycerides in a population sample from Southern Italy. The present investigation was part of an international collaborative study on fatty acids/antioxidant hypothesis of arteriosclerosis [9].
Duplicate performed ter.
measurements of blood pressure were with a random zero sphygmomanome-
Methods Blood was obtained by venous puncture using 2 Vacutainer tubes, one (for plasma) with lithium heparinate, the second (for serum) without additives. Samples of plasma for analysis of vitamin E were immediately frozen, stored in the laboratory in Naples and flown frozen on dry ice (-56°C) to the analysing laboratory in Basel, Switzerland, within 4 weeks after collection. Plasma vitamin E was determined by an HPLC method [lo], using a fluorimetric detector (excitation 290 nm, emission 330 nm). The coefficient of variation for the estimation of vitamin E was 3.0%. Serum cholesterol (Chol) and triglycerides (TG) were estimated by enzymatic methods [ll]. HDL-cholesterol was determined enzymatically after precipitation of VLDL and LDL by Mg2+dextran [12]. Serum apolipoprotein B concentration was determined by radial immunodiffusion
[131. Population sample A random sample of 120 men in the age range 40-49 years was selected from rural and semiurban communities of Southern Italy (Comunita’ Bussentina, Sapri). Five men had left the country and 115 were invited to attend a clinic: 79 agreed to participate in the study. Five men reported a history of atherosclerotic cardiovascular complications (2 myocardial infarction, 2 angina pectoris, 1 intermittent claudication). They had changed their lifestyles after the development of overt myocardial or lower limb ischaemia and were therefore excluded. The final sample size was 74 (64%). The men attended the clinic in the morning. All the participants were fasting except for 1 or 2 small cups of coffee; no other food had been consumed in the preceding 12 h. A questionnaire was completed on demographic data and social information, past medical history, smoking habits, drug therapy, medically prescribed diet and physical exercise. Anthropometric measurements included height and weight. Body mass index (BMI) was calculated from weight/height 2 (kg/m2).
The distribution of many variables was skewed to the right; therefore median and ranges were also reported and correlations were calculated from log transformed values. The log transformation was used in order to bring the data distributions, which were skewed to the right, closer to normality (as required for the statistical calculations). The statistical analysis included simple linear correlations and partial correlations [14,15]. Results Table 1 illustrates the mean values for anthropometric data and serum lipid values of the population sample under study; apolipoprotein B concentration and calculated values of non-HDLcholesterol were also reported. Table 2 illustrates the direct correlations which exist between plasma vitamin E and total, nonHDL- and HDL-cholesterol in serum. Both total and non-HDL-cholesterol were positively related to vitamin E (P < 0.001). On the other hand, HDL-cholesterol was inversely and weakly related to vitamin E concentration (Table 2). These correlations were obtained by using log transformed
27 TABLE 1 ANTHROPOMETRIC DATA AND LIPIDS, LIPOPROTEINS HEALTHY MEN FROM SAPRI, SOUTHERN ITALY
AND
VITAMIN
E CONCENTRATIONS
IN APPARENTLY
Age range 40-49 yrs; n = 74.
Mean SD Median Min Max
BMI
Blood pressure (mm Hg)
Serum
(kg/m2 )
systolic
Cholesterol (mg/dl)
Apo B
TG
Total
HDL
Non-HDL
(mg/dl)
(mg/dl)
203.1 44.0 202 118 317
45.8 12.8 44 26 76
156.9 47.7 156 66 287
88 19.2 86 45 141
157.1 146.8 126 55 1138
26.2 3.1 26.1 20.3 34.5
128.9 13.8 121 100 168
Diastolic
84.4 8.6 84 68 114
Vit. E (PM/~)
25.1 7.0 24.6 12.3 47.3
TABLE 2 CORRELATIONS (r COEFFICIENTS) BETWEEN SERUM CHOLESTEROL (Chol) VALUES (LOG TRANSFORMED) AND PLASMA VITAMIN E (LOG TRANSFORMED) IN APPARENTLY HEALTHY MEN FROM SAPRI, SOUTHERN ITALY Age range 40-49 yrs; n = 74. Vitamin E
w
0.761 * * * - 0.293 * 0.763 * * *
5
2
Total Chol HDL Chol Non-HDL Chol * PiO.05;
.* .
15
t II
h
.
SERlN
.
47 .
.
.
6%
***P
values. The correlation coefficients were almost the same when non-transformed values were used in the calculations (not shown). Figures 1 and 2 describe the positive associations which exist between vitamin E and apoli-
.
.
.
440
174 TRIGLYCERIDES
I mQ/dl
Fig. 2. Direct correlation between serum triglycerides and plasma vitamin E.
poprotein B or triglycerides (log transformed values). Each point represents one subject. Both correlations were highly significant (P < 0.001). TABLE 3 PARTIAL CORRELATIONS (r COEFFICIENTS) BETWEEN LIPOPROTEIN CONSTITUENTS (LOG TRANSFORMED) AND VITAMIN E (LOG TRANSFORMED) IN APPARENTLY HEALTHY MEN
f
Age range 40-49 years; n = 74.
rz.647 pt. 001
‘46
DO
70 APOLIPOPR~IEIN
B
I
1100
1
140
mg /dl)
Fig. 1. Direct correlation between serum apolipoprotein B and plasma vitamin E.
Variables correlated
Variable kept constant
r
Apo TG, Apo TG,
TG Apo B Non-HDL-Chol Non-HDL-Chol
0.575 *** 0.471 * * * 0.181 0.268 *
B, Vit. E Vit. E B, Vit. E Vit. E
* PcO.05;
*** P
28 Table 3 gives the results of a partial correlation analysis demonstrating that the direct correlations between vitamin E and apolipoprotein B or triglycerides were independent of each other. Both correlations almost completely disappeared, if an adjustment was performed for non-HDL-cholesterol. The weak inverse correlation between vitamin E and HDL was no longer present, when the influence of differences in apolipoprotein B or triglycerides was eliminated (not shown). Discussion This study demonstrates that plasma vitamin E concentration is positively correlated to that of non-HDL-cholesterol, triserum cholesterol, glycerides or apolipoprotein B, in an apparently healthy, male adult population. The inverse relation between vitamin E and HDL-cholesterol disappeared when plasma triglycerides and apolipoprotein B levels were taken into consideration. The correlations are in agreement with the view that vitamin E, cholesterol, triglycerides and apolipoprotein B are normal constituents of the VLDL and LDL particles in vivo. Apolipoprotein B levels have been shown to be strongly correlated to LDL-cholesterol [16]. Both apolipoprotein B and LDL-cholesterol are fairly accurate markers of circulating LDL particles, while fasting serum triglycerides mainly reflect VLDL levels. The results of partial correlation analysis suggest that both VLDL and LDL concentrations influence the vitamin E levels in the plasma of adult men. In fact apo B, the apolipoprotein constituent of LDL, is related to vitamin E independently of serum triglycerides; on the other hand, triglycerides are related to vitamin E independently of apo B. The relations between vitamin E and apolipoprotein B or triglycerides have been demonstrated in a sample of apparently healthy men with normal levels of serum lipids and lipoproteins. It is therefore likely that they reflect physiological conditions occurring in vivo in men. It has been demonstrated that LDL delivers vitamin E to the cells through a pathway which is dependent on the high affinity apo B/E receptor [17]. There is, however, also another mechanism underlying the delivery of vitamin E to peripheral tissues, which involves triglyceride-rich lipoproteins and lipoprotein lipase
[18]. On the basis of in vitro observations it has been estimated that the latter mechanism is ten times more efficient than the former [18]. It should be considered that lipoprotein lipase activity has been demonstrated in the arterial wall [19], where it is bound to the endothelial cells [20]. It is therefore localized in a crucial site for possible lipoprotein/endothelial cell interactions, leading to oxidative modifications of LDL. It has been shown that much of the variability of plasma vitamin E levels is due to differences in the serum cholesterol levels [21]. However, we have demonstrated that triglycerides and VLDL play an important role too. In a small series of hypertriglyceridemic patients (type IIB, IV, V, I) very high amounts of vitamin E were found in VLDL and chylomicrons [7]. On the other hand, in conditions of hypercholesterolemia with normal triglycerides (type IIA) plasma vitamin levels were raised due to a marked increase of vitamin E in the LDL fraction. Despite the fact that HDL carry a significant proportion of vitamin E [5,22], HDL-cholesterol levels do not seem to be much related to the variability of total plasma vitamin E. This notion is in agreement with the observation [7] that in hypertriglyceridemic patients, who have in general low HDL-cholesterol, total serum vitamin E is higher than in normolipidemic controls. In order to express vitamin E levels independently of lipoprotein carriers levels, some form of adjustment must be made. The adjustment is needed in order to account, in each individual, for the different degree of saturation of plasma lipids by vitamin E, which is a major factor for the thermodynamic partitioning of vitamin E between different lipid compartments such as plasma and arterial wall. In overt hyperlipemia only the vitamin E value, adjusted for VLDL, LDL and eventually chylomicron levels, reflects the vitamin status in human erythrocytes or animal tissues [8]. In some cross-cultural comparisons [23] the vitamin E levels have been standardized for serum cholesterol by using a molar ratio vitamin E/total cholesterol. On the basis of our correlation data, we suggest that there is an advantage of standardizing vitamin E levels for apolipoprotein B and triglycerides as well as for non-HDL-cholesterol. This last parameter should probably be preferred,
29 because it reflects much of the variability of both VLDL and LDL levels. In fact partial correlation analysis has shown that both the relations between vitamin E and apolipoprotein B or serum triglycerides almost completely disappear, when non-HDL-cholesterol levels are taken into consideration (Table 3). Non-HDL-cholesterol can be accurately measured with a simple equipment and at a lower cost than apolipoprotein B concentration. The main conclusion of the present study is that plasma levels of vitamin E, a major natural antioxidant, are markedly influenced by both LDL and VLDL concentrations, not only in patients with overt hyperlipidemia but also in the general male population. In all situations where an assessment of vitamin E status is needed, it is important to perform an adjustment for the levels of lipoprotein carriers. A simple way of dealing with this problem is to use the vitamin E/total cholesterol ratio. More accurate statistical approaches, such as analysis of covariance with an adjustment of vitamin E for non-HDL-cholesterol, are to be preferred in population studies.
6 7 8 9
10
11
12
13
14
Acknowledgements We are indebted to Mr. H. Georgi for technical assistance and to Dr. J.P. Vuilleumier for vitamin analysis (Vit. Res. Dept., Hoffmann-La Roche and Co., Ltd., Basel, Switzerland).
15 16
17
References 18 1 Morel, D.W., Di Corleto, P.E. and Chisolm, D.M., Endothelial and smooth muscle cells alter low density lipoprotein in vitro by free radical oxidation, Arteriosclerosis, 4 (1984) 357. 2 Hemiksen, T., Mahoney, E.M. and Steinberg, D., Enhanced macrophage degradation of low density lipoprotein previously incubated with cultured endothelial cells: recognition by the receptor for acetylated low density lipoproteins, Proc. Natl. Acad. Sci. USA, 78 (1981) 6499. 3 Steinbrecher, U.P., Parthasarathy, S., Leake, D.S., Witzum, J.L. and Steinberg, D., Modification of low density lipoprotein by endothelial cells involves lipid peroxidation and degradation of low density lipoprotein phospholipids, Proc. Natl. Acad. Sci. USA, 81 (1984) 3883. 4 Esterbauer, H., Juergens, G., Anchenberger, 0. and Keller, E., Autoxidation of human low density lipoprotein: loss of polyunsaturated fatty acids and vitamin E and generation of aldehydes, J. Lipid Res., 28 (1987) 495. 5 Takahashi, Y., Urono, K. and Kimura, S., Vitamin E
19 20
21
22
23
binding proteins in human serum, J. Nutr. Sci. Vit., 23 (1977) 201. Bieri, J.G., Corash, L. and van Hubbard, S., Medical uses of vitamin E, New Engl. J. Med., 308 (1983) 1063. Lambert, D. and Mourot, J., Vitamin E and lipoproteins in hyperlipoproteinemia, Atherosclerosis, 53 (1984) 327. Gey, K.F., On the antioxidant hypothesis with regard to arteriosclerosis, Bibl. Nutr. Dieta, 37 (1986) 53. Riemersma, R.A., Wood, D.A., Butler, S., Elton, R.A., Oliver, M.F., Salo, M., Nikkari, T., Vartiainen, E., Puska, P., Gey, F., Rubba, P., Mancini, M. and Fidanza, F., Linoleic acid content in adipose tissue and coronary heart disease, Br. Med. J., 292 (1986) 1423-1427. Vuilleumier, J.-P., Keller, H.E., Gysel, D. and Hunziker, F., Clinical chemical methods for the routine assessment of the vitamin status in human populations, Int. J. Vit. Nutr. Res., 53 (1983) 265. Oriente, P., Di Marino, L., Mastranzo, P., Iovine, C. and Patti, L., Simultaneous determination of cholesterol and triglycerides in serum and lipoprotein fractions with enzymatic automated methods. In: A. Burlina and L. Galzigna (Eds.), Clinical Enzymology Symposium 2, Piccin, Padua, 1979, p. 387. Kostner, G.M., Enzymatic determination of cholesterol in high-density lipoprotein fractions prepared by polyanion precipitation, Clin. Chem., 22 (1976) 695. Mancini, Y., Carbonara, A.O. and Heremans, J.F., Immunochemical quantification of antigens by simple radial immunodiffusion, Immunochemistry, 2 (1965) 235. Dixon, W.J. and Massey, F.J., Introduction to Statistical Analysis, McGraw-Hill, New York, 1969. Spiegel, M.R., Statistics, McGraw-Hill, New York, 1961. Olsson, A.G., Holmquist, L., Walldius, G., Hadell, K., Carlson, L.A., Ricciardi, G., Rubba, P., Pauciullo, P. and Mancini, M., Serum apolipoproteins, lipoproteins and fatty acids in relation to ischemic heart disease in Northern and Southern European males, Acta Med. Stand., 223 (1988) 3. Traber, M.G. and Kayden, H.J., Vitamin E is delivered to cells via the high affinity receptor for low density lipoproteins, Am. J. Clin. Nutr., 40 (1984) 747. Traber, M.G., Olivecrona, T. and Kayden, H.J., Bovine milk lipoprotein lipase transfers tocopherol to human fibroblasts during triglyceride hydrolysis in vitro, J. Clin. Invest., 75 (1985) 1729. Zilversmit, D.B., Atherogenesis: a postprandial phenomenon, Circulation, 60 (1979) 473. Olivecrona, T. and Bengtsson, G., Heparin and lipoprotein lipase - how specific is the interaction? In: International Conference on Atherosclerosis, Carlson, L.A., Paoletti, R., Sirtori, C.R. and Weber, G. (Ed), Raven, New York, 1978, p. 153. Stahelin, H.B., Rosel, F., Buess, E. and Brubacher, G., Cancer, vitamins and plasma lipids: Prospective Base1 study, J. Natl. Cancer Inst., 73 (1984) 1463. Bjomson, L.K., Kayden, H.J., Miller, E. and Moshell, A.N., The transport of alpha tocopherol and beta carotene in human blood, J. Lipid Res., 17 (1976) 343. Gey, K.F., Brubacher, G.B. and Stahelin, H.B., Plasma levels of antioxidant vitamins in relation to ischemic heart disease, Am. J. Clin. Nutr., 45 (1987) 1368.