Author + information
- Received December 6, 2000
- Revision received August 30, 2001
- Accepted September 4, 2001
- Published online December 1, 2001.
- Brendan M McQuillan, MBBS, FRACP*,†,
- Joseph Hung, MBBS, FRACP, FACC†,
- John P Beilby, PhD, FAACB§,
- Mark Nidorf, MD, FRACP, FACC‡ and
- Peter L Thompson, MD, FRACP, FACC*,†,‡ ()
- ↵*Reprint requests and correspondence: Prof. Peter L. Thompson, Department of Cardiovascular Medicine, Sir Charles Gairdner Hospital, Verdun Street, Nedlands, WA 6009, Australia
This study examined whether dietary intake or plasma levels of antioxidant vitamins were independently associated with common carotid artery intima-media (wall) thickness (IMT) or focal plaque, or both, in a large, randomly selected community population.
Oxidation of low-density lipoprotein (LDL) cholesterol is thought to be important in early atherogenesis. Antioxidant micronutrients may therefore protect against lipid peroxidation and atherosclerotic vascular disease.
We studied 1,111 subjects (558 men and 553 women; age 52 ± 13 years [mean ± SD], range 27 to 77). We measured dietary vitamin intake and fasting plasma levels of vitamins A, C and E, lycopene and alpha- and beta-carotene and performed bilateral carotid artery B-mode ultrasound imaging.
After adjustment for age and conventional risk factors, there was a progressive decrease in mean IMT, with increasing quartiles of dietary vitamin E intake in men (p = 0.02) and a nonsignificant trend in women (p = 0.10). Dietary vitamin E levels accounted for 1% of the variance in measured IMT in men. For plasma antioxidant vitamins, there was an inverse association between carotid artery mean IMT and plasma lycopene in women (p = 0.047), but not in men. None of the other dietary or plasma antioxidant vitamins, nor antioxidant vitamin supplements, were associated with carotid artery IMT or focal carotid artery plaque.
This study provides limited support for the hypothesis that increased dietary intake of vitamin E and increased plasma lycopene may decrease the risk of atherosclerosis. No benefit was demonstrated for supplemental antioxidant vitamin use.
Oxidation of low-density lipoprotein (LDL) is thought to be important in the development of early atherosclerosis (1). Oxidized LDL possesses a number of biologic properties that may promote atherogenesis (2), and agents that protect against lipid peroxidation may therefore reduce the development of atherosclerotic vascular disease. In animal studies, the administration of antioxidants has been shown to inhibit lipid oxidation and reduce the development of atherosclerosis (3,4). However, a large number of epidemi-
ologic studies investigating a relationship between antioxidant vitamins and vascular disease have yielded results that conflict with the results of randomized clinical trials (5). There is limited human data regarding the effect of antioxidant intake on measures of atherosclerosis. High-resolution carotid artery B-mode ultrasound imaging has been used previously to assess the relationship between dietary and plasma antioxidants and early atherosclerosis; however, there were inconsistent correlations in the Atherosclerosis Risk In Communities (ARIC) study in the U.S. (6,7)and the Etude sur le Viellissement Artériel (EVA) in France (8).
To further evaluate the contribution of dietary and plasma antioxidant vitamins to atherosclerotic vascular disease, we used high-resolution carotid artery B-mode ultrasound imaging to determine whether dietary intake or plasma levels of vitamins A, C and E, beta- and alpha-carotene and lycopene were independently associated with carotid artery intima-media thickness (IMT) and focal plaque in a large, randomly selected cross-sectional Australian population with a broad age range and an equal number of men and women.
Subjects were participants in the Perth Carotid Ultrasound Disease Assessment Study (CUDAS). The techniques used in this study have been previously reported (9). This was a random, electoral-roll survey (10)of 1,111 subjects from Perth, Western Australia, with equal numbers of men and women and equal numbers of subjects in each decile of age between 27 and 77 years examined between June 1995 and December 1996 (61% of those eligible agreed to participate). Because of the potential for alterations in reported dietary intake and supplemental vitamin use in subjects with a history of symptomatic vascular disease (n = 87), analyses were restricted to the 1,024 asymptomatic subjects. Written, informed consent was obtained from all study participants. The study protocol was approved by the Institutional Ethics Committee of the University of Western Australia.
In all subjects, a fasting venous blood sample was obtained. The standard procedure was to separate plasma by centrifugation shortly after venipuncture and to transport samples in ice-cooled containers protected from direct light to the laboratory. Plasma levels of antioxidant vitamins (vitamins A, C and E, alpha- and beta-carotene and lycopene) and homocysteine were determined by reverse-phase, high-performance liquid chromatography. The interassay coefficient of variation was 6% for vitamin A, 10% for lycopene and the carotenoids, 6% for vitamin C and 5% for vitamin E in our laboratory. Total cholesterol, high-density lipoprotein (HDL) cholesterol and triglyceride levels were determined enzymatically on a Hitachi 747 autoanalyzer (Tokyo, Japan). The LDL cholesterol was calculated by Friedewald’s method (11).
Risk factor assessment
A self-administered questionnaire similar to that used by the 1994 Australian National Heart Foundation Perth Risk Factor Prevalence Survey was used to record a history of hypertension, hyperlipidemia, diabetes, angina pectoris, myocardial infarction and stroke (12).
All subjects completed a semi-quantitative food-frequency questionnaire prepared by the Australian Commonwealth Scientific and Industrial Research Organization (CSIRO), Division of Human Nutrition, Adelaide, South Australia. This questionnaire was self-administered, but all responses were reviewed by a research nurse on the day of their visit. The average daily intakes of vitamins A, C and E and beta-carotene were derived from the dietary questionnaire, using the QUESTAN program of the CSIRO Dietary Assessment by Computer Program, a previously validated program for estimating vitamin intake from food-frequency data (13). The average daily vitamin intake was derived from dietary estimates and adjusted for supplemental vitamin use.
Carotid artery ultrasound examination
Bilateral carotid artery B-mode ultrasound imaging was performed according to a standardized protocol, as previously described (9). A thorough search of the full length of the distal common carotid artery, carotid artery bulb and internal and external carotid arteries was made to determine the presence of focal plaque. Plaque was defined as a clearly identified area of increased focal thickness (≥1 mm) of the intima-media layer.
The IMT was defined as the distance between the characteristic echoes from the lumen–intima and media–adventitia interfaces (14). Three end-diastolic images were digitized and analyzed from the right and left distal common carotid arteries in each subject, by use of semi-automated edge-detection software (15). The mean IMT measurement for a given subject was the average of these six measurements. Repeat measurement of randomly selected scans revealed no significant variation in the mean IMT measurement obtained during any specific period of the study. The intraobserver coefficient of variability for image acquisition and analysis in our laboratory was 2.9% for sonographer no. 1 and 4.8% for sonographer no. 2. The interobserver coefficient of variability was 5.9%. The mean (±SD) difference in carotid artery IMT between repeat measurements varied from 0.03 ± 0.02 mm for intraobserver variability to 0.05 ± 0.04 mm for interobserver variability.
Dietary vitamin intakes were adjusted for total energy (kilojoule) intake, with the nutrient density and total energy intake included in multivariate models (16). The fat-soluble plasma vitamins were found to be correlated with plasma lipid fractions and were adjusted for plasma total cholesterol as well as triglycerides for all analyses (17). Subjects were classified into gender-specific quartiles according to their dietary intake or plasma antioxidant vitamin concentrations. Logarithmic transformation was used for continuous variables, with skewed distributions where appropriate. Carotid artery mean IMT was treated as a continuous variable, and subjects were also categorized as those with or without focal carotid artery plaque. Analysis of variance was used to relate quartiles of dietary or plasma antioxidant vitamins with mean IMT, and if overall significance was demonstrated, intergroup differences were assessed by multiple comparisons testing and the ttest for linear trends. Logistic regression was used to test the independent relationship between dietary or plasma antioxidant vitamins (independent variables) and focal plaque (dependent variable). The risk of focal plaque for each vitamin quartile, compared with the risk for the lowest quartile, was estimated and expressed as the odds ratio (OR), derived as the anti-logarithm of the logistic regression coefficients. All analyses were adjusted for age, LDL cholesterol, smoking, diabetes, systolic blood pressure, homocysteine and family history. The models were fitted separately for men and women. Statistical significance was taken as a two-sided p value <0.05 (without adjustment for multiple comparisons). Results are expressed as the mean value ± SD or OR with 95% confidence interval (CI), unless otherwise stated. Analysis was performed with SAS statistical software (SAS Institute, Cary, North Carolina).
The clinical characteristics of the study population (n = 1,111; 558 men and 553 women; mean age 52 ± 13 years) are shown in Table 1. There was a similar number of men and women in each age decade from 27 to 77 years. A similar prevalence of conventional vascular risk factors was observed for men and women, except that men were more likely to have higher systolic and diastolic blood pressure, lower HDL cholesterol and higher triglyceride levels, compared with women (Table 1). A previous history of myocardial infarction or stroke was recorded in 8% (n = 87) of the total population.
Dietary and supplement antioxidant intake and plasma antioxidant levels
Table 2demonstrates the average dietary intake of vitamins for subjects initially unadjusted and then adjusted for vitamin supplement use. Women were more likely to consume antioxidant-containing vitamin supplements, compared with men. There were no differences in supplement use across age deciles or in subgroups defined by a history of smoking, diabetes, hypertension or hyperlipidemia. However, subjects with a history of clinical vascular disease had higher reported use of vitamin supplements, compared with asymptomatic subjects (27% vs. 12%; p = 0.001 by the chi-square test). Women reported an average 17% lower dietary energy intake, compared with men (p = 0.0001). Independent of kilojoule intake, women had a lower reported intake of vitamin A, but higher intakes of beta-carotene and vitamin C, with no difference in vitamin E intake. After adjustment for energy intake, women had higher proportional intakes of beta-carotene and vitamins C and E (p < 0.05), with no difference in vitamin A intake between women and men.
The average plasma antioxidant vitamin levels are shown in Table 2. Women had lower plasma vitamin A levels, compared with men, but they had higher alpha- and beta-carotene and vitamin C levels. Plasma vitamin E and lycopene levels were similar between men and women. Plasma antioxidant levels correlated with dietary intake for vitamin C (Spearman rho correlation coefficient [rs] = 0.34, p < 0.001), vitamin E (rs= 0.34, p < 0.001) and beta-carotene (rs= 0.30, p < 0.001), but not for vitamin A (rs= 0.03, p = 0.34). No differences in the relationship between dietary and plasma levels were seen between men and women or in subgroups defined by a history of smoking, diabetes, hypertension, hyperlipidemia or clinical vascular disease.
Dietary and plasma antioxidants and carotid artery atherosclerosis
In the overall population, the average carotid artery mean IMT was 0.71 ± 0.14 mm, and 26% of subjects had focal plaque identified. Men had increased mean IMT values (0.73 ± 0.15 vs. 0.69 ± 0.13 mm; p = 0.001) and a higher prevalence of focal plaque (29% vs. 22%, p = 0.001 by the chi-square test), compared with women. Mean IMT was significantly associated with conventional vascular risk factors, including age, male gender, systolic blood pressure, LDL cholesterol, pack-years of smoking, plasma homocysteine and diabetes.
Table 3shows the relationship between dietaryvitamins and carotid artery IMT, after adjustment for age, gender, total energy intake and conventional risk factors. There was a progressive decrease in mean IMT with increasing quartiles of dietary vitamin E intake in men (test for linear trends, p = 0.02) and a nonsignificant trend in women (test for linear trends, p = 0.10). In men, multiple regression analysis showed that dietary vitamin E was an independent predictor of mean IMT (partial R2= 0.01, p = 0.006 for addition to the model), with the entire model accounting for ∼50% of the variability in mean IMT (model R2= 0.51, p = 0.0001). There was no association between use of any of the antioxidant vitamin supplements and mean IMT.
Table 4shows the relationship between plasmaantioxidant vitamins and carotid artery IMT. There was an inverse association between carotid artery mean IMT and plasma levels of lycopene in women (test for linear trends, p = 0.047), but not in men.
The prevalence of carotid artery plaque decreased with increasing quartiles of dietary vitamin E (p = 0.03) in women, with no significant associations between plaque and dietary antioxidant vitamins in men. The prevalence of carotid artery plaque decreased with increasing plasma levels of alpha-carotene (p = 0.03) and lycopene (p = 0.001), with a nonsignificant trend for vitamin A (p = 0.06) in women and with increasing quartiles of vitamin A (test for trends, p = 0.02) and lycopene (p = 0.001) in men. When adjusted for age, hypertension, LDL cholesterol, smoking, homocysteine, diabetes, waist-to-hip ratio and total energy intake, logistic regression analysis found no residual association between any dietary or plasma antioxidant vitamins or supplemental vitamin use and the risk of focal plaque in men or women.
In this study, supplemental antioxidant vitamin use was not associated with mean IMT or carotid artery plaque. No independent association was seen for any of the dietary or plasma antioxidant vitamins with focal plaque. Dietary intake of vitamin E in men and plasma lycopene levels in women were found to be inversely associated with mean common carotid artery IMT, independent of conventional risk factors; however, it is not clear whether the different associations seen in women versus men were biologically significant or due to statistical chance. The finding of an effect in men may be due to the fact that men have a thicker IMT and therefore a higher chance of showing an effect, but this would not explain the association with plasma levels of lycopene seen in women. The most likely interpretation is that these weak associations are consistent with the generally weak association of antioxidant vitamins with markers of atherosclerosis.
Biologic plausibility versus clinical trial results
There are well-documented biologic mechanisms to anticipate a protective effect of antioxidant vitamins on atherosclerosis. Oxidative modification of LDL cholesterol confers a range of biologic properties that may enhance its role in early atherogenesis (18). Agents that protect LDL from oxidation have been shown in a range of in vitro and animal models to reduce the development and progression of atherosclerosis (19). Among these, the antioxidant micronutrients, including the carotenoids and vitamins C and E, have gained wide interest because of the potential for prevention of atherosclerotic vascular disease in humans. Despite this, randomized, controlled trials of antioxidant vitamins in cardiovascular disease have thus far yielded largely negative results (20). The most recent study of >9,000 participants at high risk of cardiovascular disease followed for 4.5 years showed no benefit of supplemental vitamin E on death or cardiovascular events (21). These studies have not specifically addressed the effects of antioxidant vitamins on the initiation and early progression of atherosclerosis (5).
Antioxidant vitamins and carotid artery IMT
Subclinical atherosclerosis can be accurately assessed in large-population studies by B-mode ultrasound measurement of the carotid artery wall (22–24). Carotid artery IMT correlates strongly with the presence of focal carotid artery plaque and conventional risk factors (23,24). The presence of carotid artery IMT and focal carotid artery plaque has also been found to predict the risk of future vascular events (25–27). Previous information on the relationship between antioxidant vitamins and early carotid artery atherosclerosis has come from only two ultrasound studies, with inconsistent results. In the ARIC study, dietary intake of antioxidant vitamins was assessed in 6,318 women and 4,989 men with asymptomatic coronary heart disease (6). They found that carotid artery IMT was inversely related to dietary vitamin C intake in men and women (test for trends across quintiles of intake, p = 0.035 and p = 0.019, respectively) and inversely related to alpha-tocopherol intake in women (p = 0.033), but only in subjects >55 years old. In a subsequent case-control comparison from the ARIC group, 231 patients (60% men) with increased carotid artery IMT (>90th percentile of IMT for the entire cohort) were compared with 231 matched control subjects (<75th percentile of IMT), with regard to their serum antioxidant vitamin levels, among other factors (7). They observed that patients had lower serum levels of the carotenoids beta-cryptoxanthin (p = 0.02 by the ttest) and lutein plus zeaxanthin (p = 0.05) than control subjects, but, contrary to the hypothesis, patients also had nonsignificantly higher alpha-tocopherol and retinol levels than control subjects. This follow-up study was further confounded by the delay of three years between the carotid artery examination and the collection of the serum antioxidants.
The EVA study examined the associations of plasma carotenoids, selenium, erythrocyte vitamin E and a marker of lipid peroxidation by ultrasound measurement of carotid artery atherosclerosis in an elderly cohort (aged 59 to 71 years) of 476 men and 711 women from France (8). They observed an independent inverse association between erythrocyte vitamin E and common carotid artery IMT in men (test for linear trends across quartiles of vitamin E, p = 0.041) and women (p = 0.045), but no independent association of carotid artery IMT and carotid artery plaque with plasma carotenoids.
In the present study, we also found only weak correlations between antioxidant vitamin intake and carotid artery atherosclerosis. Dietary vitamin E intake was independently and inversely associated with mean IMT in men (test for linear trends, p = 0.02), with a nonsignificant trend in women. The effect of vitamin E intake was modest, adding ∼1% to the variance in measured IMT in men. Among plasma vitamins, only the naturally derived carotenoid lycopene was inversely associated with mean IMT in women. We found no independent association between vitamin supplements and IMT or between any dietary or plasma antioxidant vitamins and the risk of carotid artery plaque. Taken in context with the previous studies, our results suggest that any association between individual antioxidant vitamins and early atherosclerosis is likely to be modest and variable according to gender and the population studied. The modest and variable effects of individual antioxidant vitamins seen in these studies could conceivably be a chance finding, due to the large number of variables studied.
Significance of the findings of this study
This was a cross-sectional study of an atherosclerotic surrogate, but it represents a large number of subjects with an equal gender ratio and a broad age range, with equal numbers in each age decile. The subjects were participants in a random survey of the electoral roll and are likely to reflect the general Australian population. The exclusion of subjects with evident cardiovascular disease should reduce the potential bias of recent changes in diet, because of the occurrence of disease. The negative results are consistent with the findings of the carotid artery substudy of a trial of vitamin E supplementation, which also showed a lack of effect of vitamin E supplementation on the rate of progression of carotid artery atherosclerosis (28).
Although food-frequency questionnaires provide only an approximate measure of usual dietary intake, there was a moderate correlation between plasma vitamin levels and dietary intakes, comparable to correlations found in previous reports (29). Assessment of both measures should allow stratification of subjects within the group, with regard to their vitamin status. The common food sources of several antioxidant vitamins (especially fruits and vegetables) raise the possibility that it is a combination of vitamins that may provide benefit. Some trials (21,30)showing a lack of benefit with vitamin E supplementation may have overlooked an effect of the combination or even pro-oxidant effects of vitamin E given in isolation (31).
This study provides limited support for the hypothesis that higher dietary consumption of vitamin E may reduce the development of early atherosclerosis. We found no protective effect from the use of antioxidant-containing vitamin supplements on carotid artery atherosclerosis. Further studies are required to assess the role of antioxidant vitamin combinations and other potential anti-atherogenic aspects of diet.
☆ This study was supported by a grant from Healthways, the Western Australian Health Promotion Foundation.
Presented in part at the 70th Annual Scientific Session of the American Heart Association, Orlando, Florida, November 9–12, 1997.
- Atherosclerosis Risk In Communities study
- confidence interval
- Commonwealth Scientific and Industrial Research Organization
- Carotid Ultrasound Disease Assessment Study
- Etude sur le Viellissement Artériel
- intima-media thickness
- high-density lipoprotein
- low-density lipoprotein
- odds ratio
- Received December 6, 2000.
- Revision received August 30, 2001.
- Accepted September 4, 2001.
- American College of Cardiology
- Carew T.E,
- Schwenke D.C,
- Steinberg D
- Kritchevsky S.B,
- Shimakawa T,
- Tell G.S,
- et al.
- Iribarren C,
- Folsom A.R,
- Jacobs D.R Jr.,
- Gross M.D,
- Belcher J.D,
- Eckfeldt J.H
- Bonithon-Kopp C,
- Coudray C,
- Berr C,
- et al.
- McQuillan B.M,
- Beilby J.P,
- Nidorf M,
- Thompson P.L,
- Hung J
- ↵Risk Factor Prevalence Study Management Committee. Risk factor prevalence study: survey no. 3, 1989. Canberra: National Heart Foundation of Australia and Australian Institute of Health, 1990:141.
- Friedewald W.T,
- Levy R.I,
- Fredrickson D.S
- Risk Factor Prevalence Study Management Committee
- Pignoli P,
- Tremoli E,
- Poli A,
- Oreste P,
- Paoletti R
- Adams M.R,
- Nakagomi A,
- Keech A,
- et al.
- Willett W.C,
- Howe G.R,
- Kushi L.H
- Thurnham D.I,
- Davies J.A,
- Crump B.J,
- Situnayake R.D,
- Davis M
- Heiss G,
- Sharrett A.R,
- Barnes R,
- Chambless L.E,
- Szklo M,
- Alzola C
- Belcaro G,
- Nicolaides A.N,
- Laurora G,
- et al.
- Bots M.L,
- Hoes A.W,
- Koudstaal P.J,
- Hofman A,
- Grobbee D.E
- SECURE Investigators,
- Lonn E,
- Yusuf S,
- Dzavik V,
- et al.
- Hunter D
- Upston J,
- Terentis A,
- Stocker R