Determinants of the nutritional status of vitamin E in a non ... - Nature

European Journal of Clinical Nutrition (1997) 51, 723±728 ß 1997 Stockton Press. All rights reserved 0954±3007/97 $12.00

Determinants of the nutritional status of vitamin E in a nonsmoking Mediterranean population. Analysis of the effect of vitamin E intake, alcohol consumption and body mass index on the serum alpha-tocopherol concentration P GascoÂn-Vila1, R Garcia-Closas1, L Serra-Majem1,2, MC Pastor3, L Ribas1, JM Ramon1, A MarineÂ-Font4 and L Salleras5 1

Community Nutrition Research Unit, Institute of Public Health, Campus de Bellvitge, University of Barcelona, Ctra. de la Feixa Llarga, s/n 08907 L'Hospitalet, Spain; 2Department of Preventive Medicine and Public Health, Faculty of Health Sciences, University of Las Palmas, 35080 Las Palmas, de Gran Canaria; 3Laboratory of Biochemistry, Germans Trias i Pujol Hospital, 08910, Badalona; 4 Department of Nutrition and Bromatology, Faculty of Pharmacy, University of Barcelona, 08028 Barcelona; and 5Department of Health and Social Security, Autonomous Government of Catalonia, 08028 Barcelona

Objectives: Study was conducted in order to investigate the association of vitamin E intake and other factors with plasma a-tocopherol concentration in a non-smoking Mediterranean population. Design: A cross-sectional study was conducted in a subsample of a representative sample of the Catalan population. Subjects: Sample size was 143 men and women, aged between 18 and 75 y, and ®nal response rate reached 61.9% of the initial sample. Interventions: Serum alpha-tocopherol concentration standardized by serum total lipids was used as a proxy of the nutritional status of vitamin E. Vitamin E intake and alcohol consumption were estimated by a replicated 24 h recall method. Dietary data were collected in two different periods, winter and summer, in order to account for seasonal variation in nutrient intake, and were corrected for random within-person variability in order to account for day-to-day variation in nutrient intake. Multivariate linear regression models were ®tted in order to estimate the determinants of serum a-tocopherol concentration. Results: In this population study, for each one mg increase in vitamin E intake, serum a-tocopherol concentration increased, on average, 0.66 micromol/L, after adjusting for age, gender, Body Mass Index (BMI), alcohol consumption and energy intake. BMI also in¯uenced signi®cantly serum a-tocopherol concentration, whereas alcohol intake, age and gender did not show signi®cant associations with serum a-tocopherol. Conclusions: The study showed that vitamin E nutritional status was associated to vitamin E intake and BMI in non-smokers. Sponsorship: This study was supported by a research agreement between the Department of Health of the Autonomous Government of Catalonia and the Bosch Gimpera Foundation of the University of Barcelona (Contract 2415/95). Descriptors: vitamin E; BMI; alcohol; intake; a-tocopherol and nutritional status

Introduction Vitamin E is an essential micronutrient which may have protective effects against cardiovascular diseases and several cancers through different biological mechanisms (Grant & De Hoog, 1985; Willett, 1986; Diplock, 1991; Stahelin et al, 1991; Ashwell, 1993; Rimm et al, 1993; Stampfer et al, 1993; Bellizzi et al, 1994; Kushi et al, 1996). It has antioxidant properties, which may prevent cellular damage (Bieri, 1984; Ames, 1983). It may also have antiaggregant effects (Gey, 1989) and inhibit carcinogenesis (Ames, 1983). Knowledge of the determinants of

Correspondence: Prof LluõÂs Serra-Majem, Institut de Salut PuÂblica de Catalunya, Camput de Bellvitge, Universitat de Barcelona, Ctra. de la Feixa Llarga, s/n. 08907 L'Hospitalet, Spain. Received 19 January 1997; revised 12 June 1997; accepted 26 June 1997

the vitamin E nutritional status is, therefore, useful for epidemiologic studies of chronic disease. The concentration of total vitamin E or a-tocopherol in serum is the most frequently used biochemical index of the nutritional status of vitamin E (Gibson, 1990; Hunter, 1990). It does appear that a single measurement of plasma or serum a-tocopherol standardized by lipid concentration can reasonably represent long-term vitamin E intake and could be used in epidemiological studies of diet and disease (Gibson, 1990; Hunter, 1990). Factors such as tobacco and alcohol consumption do appear to be determinants of plasma a-tocopherol levels in some epidemiologic studies (Buiatti et al, 1996), but not in others (Stryker et al, 1988; Comstock et al, 1988). Previous studies have found a relationship between concentration of vitamin E in the adipose tissue and plasma concentration of a-tocopherol (Kayden et al, 1983; Traber & Kayden, 1987; Kaardinal et al, 1995).

Determinants of the nutritional status of vitamin E P GascoÂn-Vila et al

724

The aim of the present investigation is to study vitamin E dietary intake, body mass index (BMI), and alcohol consumption as possible determinants of serum a-tocopherol concentration in a non-smoking Mediterranean population.

et al, 1991). We chose this source of food composition data because of its accuracy, speci®city, and comprehensiveness in foods and nutrients, and because of the similarities in the composition and types of food consumed in Catalonia and in France (Farran et al, 1994).

Methods

Statistical analyses Data analyses were performed using the statistical package STATA 4.0 for Windows (Statistics and Data Analysis, Stata Corporation, College Station, Texas). In order to account for day-to-day variation in nutrient intake, the estimated nutrient intakes were corrected for random within-person error (Beaton et al, 1979). Correction for random within-person error enabled us to estimate usual intake. Usual intake is an estimation of long term intake, and long term intake is correlated with nutritional status biomarker (Willett, 1990). Means, standard deviations and medians were estimated for continuous variables, and frequency distributions were examined for categorical variables. The residual standard multivariate method was used to adjust nutrient intake for total energy intake (Willett & Stampfer, 1986). This method enabled us to isolate the effect of total energy intake on the dependent variable. It overcomes the added variation of nutrient intake related to total energy intake, but unrelated to dietary composition (Willett, 1990; Willett & Stampfer, 1986). Energy-adjusted vitamin E intake, energy-adjusted alcohol consumption and BMI were categorized in tertiles. Serum a-tocopherol levels and other characteristics of the sample were explored by levels of vitamin E intake, alcohol consumption and BMI. The Kruskall±Wallis test was used to assess mean differences between categories. Crude associations between serum alpha-tocopherol concentration and the independent covariates were explored by correlation and linear regression coef®cients. Multivariate linear regression was the method applied to estimate the determinants of serum a-tocopherol concentration. The independent variables considered for the analyses were: gender, age (y), vitamin E intake (mg/d), alcohol consumption (g/ d), BMI (Kg/m2) and total energy intake (Kcal/d). The dependent variable was the serum a-tocopherol concentration standardized by serum total lipid concentration. Assumptions of linear regression models were tested by pair-wise scatter plots, quantile-normal plots of the residuals, residuals versus predicted plots and partial leverage plots. Log transformation of the dependent variable was necessary in order to ful®l the assumption of homoscedasticity of the residuals. Once the ®nal model was ®tted, in¯uential analysis was performed by examination of high leverage, DFITS, DEBeta data points and Cook's distance.

Sample The representativeness of the Catalan population sample was assured by the sampling process. The sample was strati®ed according to household and randomized into subgroupings where municipalities constituted the primary sampling units, and the individuals within these municipalities constituted the ®nal sample units. The sample distribution according to strati®cation by household was as follows: 46 municipalities with less than 10000 inhabitants, 28 municipalities with more than 10 000 and less than 100 000 inhabitants, and 8 municipalities consisting of more than 100 000 inhabitants (Serra-Majem et al, 1996). For biochemical assessment a random subsample was obtained from the representative sample of the Catalan population aged between 18 and 75 y, previously selected for the Nutritional Survey of the Catalan Population (1992±93) carried out by the Department of Health and Social Security of the Government of Catalonia and the University of Barcelona (Serra-Majem et al, 1996). After exclusion of smokers, the study sample was 143 subjects: 45.4% men, 54.6% women. Dietary assessment method Two 24 h recalls per subject collected in two different periods (June±July 1992 and November±December 1992) were used to estimate nutrient intake. Thirty-six previously trained dieticians using standard protocol obtained the information in the participants homes. Each subject was visited by the same dietician in the two interviews. Response rate in the ®rst and second interview were respectively 68.9% and 61.9% of the initial sample. Household measures were used to estimate portion sizes. The dieticians were in charge of conversion of food portion size into grams and milliliters, and of food codi®cation. Anthropometric measures Bathroom scales which were calibrated every day were used to estimate subjects weights their heights were determined using non-extensible wall measures. Laboratory analyses Blood samples were obtained in standardized conditions by vacuum extraction from 12 h fasted individuals. Blood was centrifuged at 3000 rpm and 8 C for 15 min, and serum was frozen at 7 80 C in the health centers of extraction. Serum a-tocopherol was measured by high performance liquid chromatography (HPLC) with a UV detector (280 nm). The mobile phase was methanol: water (95:5), 2 mL/min. (Lehman & Martin, 1982). The range of coef®cients of variation (CV) was 5±5.5% in the sample replicated analysis. In order to avoid misclassi®cation of vitamin E nutritional status, the plasma a-tocopherol concentration was adjusted by serum lipid concentration (total cholesterol and triglycerides) (Willett et al, 1983a). Food composition table The food composition database used to estimate nutrient intake was the ReÂpertoire GeÂneÂral des Aliments (Feinberg

Results Serum concentration of a-tocopherol and characteristics of the sample by tertiles of vitamin E intake are shown in Table 1. No statistical differences were found between categories of vitamin E intake for any of the studied variables from unadjusted analyses. Mean energy intake was 7.9 kJ (1905 kcal), median was 7.6 kJ (1818 kcal), 25th percentile was 6.8 kJ (1639 kcal) and 75th percentile was 8.6 kJ (2062 kcal). Table 2 shows serum concentration of a-tocopherol and characteristics of the sample by tertiles of alcohol consumption. The proportion of men increased with increasing levels of alcohol consumption. Serum concentration of a-tocopherol and


Table 1

725

Serum a-tocopherol concentration and characteristics of the 143 participants by tertiles of vitamin E intake Tertiles of vitamin E intake (mg/d) Low (4.73±8.24)

Medium (8.25±9.28)

Pa

High (9.29±17.89)

Gender (% of males)a

34.8 Mean

s.d.

57.4 Mean

s.d.

44.0 Mean

s.d.

Age (y) Body Mass Index (BMI) Energy-adjusted vitamin E intake (mg/d) Standardized serum a-tocopherol (mmol/L)a

44.3 26.4 9.9

12.8 4.7 8.0

46.9 25.3 9.5

19.9 3.2 13.0

48.5 25.7 11.1

15.3 3.6 9.6

0.5 0.6 0.2

32.4

8.1

31.4

5.8

32.1

8.2

0.7

0.2

a

P Kruskall±Wallis test.

Table 2

Serum a-tocopherol concentration and characteristics of the 143 participants by tertiles of alcohol consumption Tertiles of alcohol consumption (g/d) Low (0.00±3.83)

Gender (% of males)a Age (y) Body Mass Index (BMI) Energy-adjusted vitamin E intake (mg/d) Standardized serum a-tocopherol (mmol/L)a

Medium (3.84±10.20)

Pa

High (10.21±81.37)

29.2 Mean

s.d.

34.0 Mean

s.d.

72.9 Mean

s.d.

43.0 25.4 9.0

17.7 4.2 1.8

48.4 25.9 9.1

15.1 4.3 2.3

48.5 26.0 8.6

15.6 3.0 1.7

30.3

7.7

30.9

8.2

30.6

5.3

0.0003 0.2 0.002 0.8 0.3

a


Table 3

Serum a-tocopherol concentration and characteristics of the 143 participants by tertiles of Body Mass Index (BMI) Tertiles of Body Mass Index (BMI) Low (18.3±24.0)

Medium (24.1±26.8)

Pa

High (26.9±40.2)

Gender (% of males)

32.6 Mean

s.d.

58.3 Mean

s.d.

44.9 Mean

s.d.

Age (y) Energy-adjusted alcohol consumption (g/d)a Energy-adjusted vitamin E intake (mg/d) Standardized serum a-tocopherol (mmol/L)a

36.2 7.0 9.1

13.7 7.0 2.0

49.9 11.9 9.0

15.7 9.9 1.9

53.2 11.5 8.6

14.2 12.7 1.9

30.1

5.7

32.2

5.9

33.4

9.3

0.1 0.0001 0.009 0.5 0.1

a


characteristics of the sample by tertiles of BMI are shown in Table 3. Age increased with increasing tertiles of BMI (P < 0.05). Alcohol consumption was higher in the second and third tertile of BMI than in the ®rst tertile group (P < 0.05). The Pearson correlation coef®cient between the logarithm of vitamin E intake and the logarithm of serum a-tocopherol was 0.15, between the logarithm of BMI and the logarithm of serum a-tocopherol it was 0.17, and for the logarithm of alcohol consumption and the logarithm of serum a-tocopherol it was 0.14 (Table 4). In multivariate analyses, vitamin E intake was signi®cantly associated with serum a-tocopherol concentration, after adjusting for age, gender, BMI, alcohol consumption and energy intake. BMI also in¯uenced serum a-tocopherol concentration, whereas no association was observed between alcohol intake and serum a-tocopherol (Table 5). In our ®nal linear regression model, we used the logarithm of serum a-tocopherol in order to have homoscedasticity of the

residuals. However, for better interpretation of the estimators of the b-parameters, we used the untransformed regression coef®cients, which are unbiased despite heteroscedasticity (Hamilton, 1992). The regression coef®Table 4 Pearson correlation coef®cient between the logarithm of serum a-tocopherol (micromol/L) and the logarithm of the independent variables considered for the analyses: vitamin E intake (mg/d), alcohol consumption (g/day), BMI (Kg/m2), total energy intake (kcal/d), age (y) and gender. Signi®cance level (P) Serum a-tocopherol Pearson correlation coef®cient Vitamin E intake BMI Alcohol consumption Age Energy intake Gender

0.15 0.17 0.14 0.16 0.02 7 0.07

P 0.03 0.04 0.01 0.06 0.5 0.42


726

Table 5 Predictors of serum a-tocopherol concentrationa in multivariate analyses b-coef®cient, standard error, con®dence interval and signi®cance level Independent variable Energy intake Alcohol consumption Age Sex BMI Vitamin E intake Intercept

b-coef®cient (s.e.)

95% CI for b

P

0.0002 (0.0001) 0.002 (0.002)

(0.00002, 0.0003) (ÿ0.0020, 0.0003)

0.03 0.40

0.003 0.094 0.010 0.018 2.387

(ÿ0.0002, 0.0056) (ÿ0.0140, 0.2029) (0.0003, 0.0197) (0.0001, 0.0351) (1.7600, 3.0140)

0.07 0.09 0.04 0.049 < 0.001

(0.001) (0.055) (0.005) (0.009) (0.3169)

R2 0.1056. Adjusted R2 0.0655. a Serum a-tocopherol concentration was transformed to the logarithm

cients for vitamin E intake, BMI and energy intake in the untransformed model were 0.66, 0.34 and 0.004 respectively. Discussion In this cross-sectional study we found that intake of vitamin E was positively associated with serum a-tocopherol concentration. For each one mg increase in vitamin E intake, serum a-tocopherol concentration increased, on average, 0.66 micromol/L, after adjusting for age, gender, BMI, alcohol consumption and energy intake. We used serum a-tocopherol concentration as an indicator of vitamin E nutritional status. a-tocopherol is the most abundant and biologically active form of vitamin E (Tietz, 1986). Serum and plasma a-tocopherol are the most frequently used biomarkers of vitamin E nutritional status because they are technically easiest to obtain and process, and because they are more suitable for large population studies (Gibson, 1990). Concentration of plasma a-tocopherol varies according to method of analysis and serum lipid concentration (Gibson, 1990). High performance liquid chromatography (HPLC) is the method of choice to analyze serum vitamin E concentration (Behrens et al, 1986). This method is relatively simple, and rapid, it requires only small volumes of serum/plasma, and it allows different tocopherols to be differentiated (Bieri, 1984; Bieri et al, 1985; Behrens et al, 1986). The very strong correlation between vitamin E plasma concentration and serum lipid concentration is due to the fact that a-tocopherol is transported in blood as part of a lipoprotein complex (Gibson, 1990). Plasma a-tocopherol concentration standardized by total lipids is established as a biochemical marker of the nutritional status of vitamin E (Willett et al, 1983a). The regression coef®cient for the logarithm of vitamin E intake in the multivariate model was 0.018 (Table 5), which was lower than that observed in previous studies (Willett et al, 1983b; Ascherio et al, 1992). In our study the Pearson correlation coef®cient between the logarithm of energy-adjusted vitamin E intake and the logarithm of the standardized serum a-tocopherol was 0.15 (Table 4). Previous studies have observed correlations from 0.12±0.65 (Willett et al, 1983a; Willett et al, 1983b; Stryker et al, 1988; Knekt et al, 1988; Ascherio et al, 1992). In these studies the highest associations were found in individuals who took vitamin E supplements (Willett et al,

1983b; Ascherio et al, 1992). The lower associations between intake of vitamin E and serum a-tocopherol found in our study could be due to the fact that less than 10% of our study population took vitamin E supplements (Serra-Majem et al, 1996). In this study, vitamin E supplement intake was taken into account in the estimation of total vitamin E intake. A further reduction in the observed association could be caused by differences in the dietary assessment method used to estimate intake. We used a replicated 24 h recall method and the other studies used food frequency questionnaires (FFQ). The 24 h recall method provides high answer rates, it can be used in low literacy population groups, its reproductibility is high, and its cost is relatively low (Block, 1982). However, the high day-to-day ¯uctuation of individual nutrient intake (Ballogh et al, 1971; Beaton et al, 1979) in the diet assessment when using daily methods may lead to misclassi®cation of the subject's long term dietary intake (Witschi, 1990). In order to account for seasonal variation of dietary intake, we collected the data during two different periods: winter and summer. To avoid attenuation of the true associations we further adjusted for within-person variability (Beaton, 1979). We found that for each one unit increase in BMI, serum a-tocopherol increased by 0.34 micromol/l, after adjusting for gender, age, vitamin E intake, alcohol intake and energy intake. The adipose tissue is the most important pool of a-tocopherol in the human body (Bieri, 1972; Hunter, 1990), and Quetelet's Index (weight/height2) is a good estimator of adiposity in epidemiologic studies (Criqui et al, 1982; Willett, 1990). However, the bioavailability of vitamin E in the adipose tissue is not well known, nor are the mechanisms which regulate the release of alpha-tocopherol from the adipocyte into the plasma (Farrel and Bieri, 1975). Adipose tissue concentration of vitamin E in well-nourished populations increases with intake of supplements of vitamin E (Bieri, 1972). Our study population had vitamin E intakes above the recommended dietary allowances (Gutteridge & Haliwell, 1994). The Pearson correlation coef®cient between the logarithm of the BMI and the logarithm of the serum atocopherol concentration was 0.17 (Table 4). Some authors have observed a correlation coef®cient between a-tocopherol concentration in adipose tissue and plasma a-tocopherol levels of 0.34 (Kaardinal et al, 1995). Lifestyle factors other than diet could modify the nutritional status of vitamin E (Stryker et al, 1988). For example, oxidative stress by alcohol consumption could reduce serum concentration of a-tocopherol (Knutse et al, 1993). Alcohol consumption could induce lipid peroxidation (Clot et al, 1994). Intake of vitamin E has an important role in counteracting the oxidative stress caused by alcohol consumption (Nordmann, 1994). However, as in previous studies (Stryker et al, 1988; Comstock et al, 1988), we did not ®nd an association between alcohol intake and serum a-tocopherol concentration in the linear regression model. Low intakes of alcohol have not been related to depletion of antioxidant vitamins in the human body (Clot et al, 1994), but plasma concentration of a-tocopherol is lower in alcoholics than in people with low intakes of alcohol (Lecomte et al, 1994). Epidemiological and experimental studies have shown that the oxidative stress by alcohol consumption depends on the amount and frequency of ethanol intake (Lecomte et al, 1994; Clot et al, 1994; Azzalis et al, 1995). In our study; alcohol intake was low compared to other regions in Spain (Aranceta et al, 1990). Concentration of serum tocopherol


varies according to age (Gibson, 1990; Serra-Majem et al, 1996). The crude association between age and serum vitamin E may be largely due to its association with plasma lipids. In this study, the effect of age as a predictor of serum a-tocopherol disappeared after standardization of serum a-tocopherol by serum cholesterol and triglyceride concentrations. Conclusion Serum a-tocopherol concentration, used as an indicator of the nutritional status of vitamin E, seems to be determined by vitamin E intake and BMI in this non-smoking population group. AcknowledgementsÐThe authors wish to acknowledge Dr. Francesc MaciaÁ for their helpful assistance in the laboratory analysis of vitamin E and to Miss Marta Olmos, Mr. Jordi EspunÄas and Mrs. Neus GalõÂ for their assistance in the ®eld work of the biochemical analysis. This study was supported by a research agreement between the Department of Health of the Autonomous Government of Catalonia and the Bosch Gimpera Foundation of the University of Barcelona (contract 2415/95). Pablo GascoÂn Vila received a fellowship for Research Personel Training from the Spanish Ministry of Education and Culture Ministry.

References Ames BN (1983): Dietary carcinogens and anticarcinogens. Oxygen radicals and degenerative diseases. Science 221, 1256±1264. Aranceta J, PeÂrez C & Eguileor Y (1990): Encuesta Nutricional del Pais Vasco. Vitoria: Gobierno Vasco. Ascherio A, Stampfer MJ, Colditz GA, Rimm EB, Litin L & Willett WC (1992): Correlations of vitamin A and E intakes with the plasma concentrations of carotenoids and tocopherols among American men and women. J. Nutr. 122, 1792±1801. Ashwell M (1993): Diet and Heart Disease (1st edn). A Round Table of factors. British Nutrition Foundation: London. Azzalis LA, Junqueira VB, Simon K, Giavarotti L,Silva MA, Kogake M, Simizu K, Barros SB, Fraga C & Porta EA (1995): Prooxidant and antioxidant hepatic factors in rats chronically fed in ethanol regimen treated with an acute dose of lindane. Free Rad. Biol. Med. 19(2), 147± 159. Balogh M, Kahn HA & Medalie JH (1971): Random repeated 24-hour dietary recalls. Am. J. Clin. Nutr. 24, 304±310. Beaton GH, Milner J, Corey P, McGuire V, Cousins M, Stewart E, de Ramos M, Hewitt D, Gransbsch PV, Kassim N & Little JA (1979): Sources in variance in 24-hour dietary recall data: implications for nutrition study design and interpretation. Am. J. Clin. Nutr. 32, 2546± 2259. Behrens WA, Thompson JN & Medere R (1986): Alpha- and GammaTocopherol concentrations in human serum. J. Am. Coll. Nutr. 5, 91±96. Bellizzi MC, Franklin MF, Duthie GG & James WPT (1994): Vitamin E and coronary heart disease: the European paradox. Eur. J. Clin. Nutr. 48, 822±831. Bieri JG (1972): Kinetics of tissue of alpha-tocopherol depletion and repletion. Ann. N. Y. Acad. Sci. 203, 181±191. Bieri JG (1984): Vitamin E.: In Present Knowledge in Nutrition (5th edn), pp 226±240. The Nutrition Foundation Inc: Washington. Bieri JG, Tolliver TJ & Catignani GL (1985): Simultaneous determination of alpha-tocopherol and retinol in plasma or red cells by high pressure liquid chromatography. J. Liq. Chrom. 8, 473±484. Block GH (1982): A review of validations of dietary assessment methods. Am. J. Epidemiol. 115, 492±505. Buiatti E, MunÄoz N, Kato I, Vivas J, Muggli R, Clummer M, Benz M, Fraceschi S & Oliver W (1996): Determinants of plasma anti-oxidant vitamin levels in a population at high risk for stomach cancer. Int. J. Cancer 65(3), 317±322. Clot P, Tabone M, Arico S & Albano E (1994): Monitoring oxidative damage in patients with liver cirrhosis and different daily alcohol intake. Gut 35(11), 1637±1643. Comstock GW, Menkes MS, Schober SE, Vuilleumier JP & Helsing KJ (1988): Serum levels of retinol, beta-carotene, and alpha-tocopherol in older adults. Am. J. Epidemiol. 127, 114±123. Criqui M, Klauber H, Barret-Connor E, Holdbrook MJ, SuaÂrez L &

Windgard DL (1982): Adjustment for obesity in studies of cardiovascular disease. Am. J. Epidemiol. 116, 685±691. Diplock AT (1991): Antioxidant nutrients and disease prevention: an overview. Am. J. Clin. Nutr. 53, 189S, 193S. Farran A, Boatella J, Serra-Majem L, Ribas L, Rafecas M & Codony R (1994): Criterios generales de elaboracioÂn y utilizacioÂn de tablas y sistemas de datos de composicioÂn de alimentos. Rev. de San. e Hig. Pub. 68, 427±441. Farrell PM & Bieri JG (1975): Megavitamin E supplementation in man. Am. J. Clin. Nutr. 28, 1381±1386. Feinberg M, Favier JC & Ireland Riper J (1991): ReÂpertoire GeÂneÂral Des Aliments. Ciqual-Lavoisier, Paris, ®rst edition. Gey KF (1989): Inverse correlation of vitamin E and ischemic heart disease. Int. J. Vit. Nutr. Res. 30 (suppl), 224±231. Gibson RS (1990): Assessment of the status of vitamins A, D, and E. In: Principles of Nutritional Assessment, (1st edn), ed. RS Gibson, pp 377±412. Oxford University Press, New York. Grant A & DeHoog S (1985): Nutritional Assessment and Support (3rd edn). Grant-Dehoog: Seattle. Gutteridge MC & Haliwell B (1994): Antioxidants in Nutrition, Health and Disease (1st edn), p 67. Oxford University Press: London. Hamilton LC (1992): Regression with Graphics (1st edn). Belmont Duxbury Press: California. Hunter D (1990): Biochemical indicators of dietary intake. In Nutritional Epidemiology (1st edn). ed. WC Willett, pp 143±216. Oxford: Oxford University Press. Kardinaal AF, Van Veer P, Brants HA, Van den Berg H, Schoonhoven J & Hermus RJ (1995): Relations between antioxidant vitamins in adipose tissue, plasma, and diet. An. J. Epidemiol. 141 (5), 440±450. Kayden HJ, Hatam LJ & Traber MG (1983): The measurement of nanograms of tocopherol from needle aspiration biopsies of adipose tissue: normal and abetalipoproteinemic subjects. J. Lipid Res. 24, 652± 656. Knetk P, Maatela J, Aaran RK, Nikkari T, Hakama M, Hakulinen T, Peto R, Saxen E & Teppo L (1988): Serum vitamin E and risk of cancer among Finnish men during a 10-year follow-up. Am. J. Epidemiol. 127, 28±41. Knutse SF, Ringstad J, Nilssen OR & Thomassen Y (1993): A Comparative study of serum selenium and vitamin E levels in a population of male risk drinkers and abstainers. Biol. Trace Elem. Res. 36 (1), 65± 71. Kushi LH, Folsom AR, Prineas RJ, Mink PJ, Wu Y & Bostick RM (1996): Dietary antioxidant vitamins and death from coronary heart disease in postmenopausal women. N. Engl. J. Med. 334, 1156± 1162. Lecomte E, Herberth B, Pirrollet P, Chancerelle Y, Aruand J, Musse N, Paille F, Siest G & Artur Y (1994): Effect of alcohol consumption on blood antioxidant nutrients and oxidative stress indicators. Am. J. Clin. Nutr. 60 (2), 255±261. Lehman J & Martin HL (1982): Improved direct determination of alphaand gamma-tochopherols in plasma and platelets by liquid chromatography, with ¯uorescence detection. Clin. Chem. 28, 1784±1787. Nordmann R (1994): Alcohol and antioxidant systems. Alcohol-alcohol. 29 (5), 513±522. Rimm EB, Stampfer MJ, Ascherio A, Giovannucci E, Colditz GA & Willett W (1993): Vitamin E consumption and the risk of coronary heart disease in men. N. Engl. J. Med. 328, 1450±1456. Serra-Majem L, Ribas L, Garcia-Closas R, Ramon J, Salvador G, Farran A, Serra J, Sabater G, Jover L, GascoÂn P, Tresserras R, SaltoÂ E, ChacoÂn P, Pastor MC, Pvchal A, Lloveras G, Taberner JL & Salleras L (1996): AvaluacioÂ de l'Estat Nutricional de la PoblacioÂ Catalana (1992±93). Department de Sanitat i Seguretat Social. Generalitat de Catalynya: Barcelona. Stahelin HB, Gey KF, Eichholzer M, Ludin E, Bernasconi F, Thurneysen J & Brubacher G (1991): Plasma antioxidant vitamins and subsequent cancer mortality in the 12-year follow-up of the Prospective Basel Study. Am. J.Epidemiol. 133 (8), 766±775. Stampfer MJ, Hennekens CH, Manson JE, Colditz GA, Rosner B & Willett WC (1993): Vitamin E consumption and the risk of coronary heart disease in women. N. Engl. J. Med. 328, 1444±1449. Stryker S, Kaplan L,Stein A, Stampfer MJ, Sober W & Willett WC (1988): The relation of diet, cigarette smoking, and alcohol consumption to plasma beta-carotene and alpha-tocopherol levels. Am. J. Epidemiol. 127, 283±296. Tietz NW (1986): Textbook of Clinical Chemistry. (1st edn) W.B. Saunders Company: Philadelphia. Traber MG & Kayden HJ (1987): Tocopherol distribution and intracellular localization in human adipose tissue. Am. J. Clin. Nutr. 46, 488±495. Willett WC, Stampfer MJ, Underwood BA, Taylor J & Hennekens CH

727


728

(1983a): Vitamins A, E, and Carotene: effects of supplementation on their plasma levels. Am. J. Clin. Nutr. 38, 559±566. Willett WC, Stampfer MJ, Underwood BA, Hennekens CH, Speizer FE & Rosner B (1983b): Validation of dietary questionnaire with plasma carotenoid and alpha-tochopherol levels. Am. J. Clin. Nutr. 38, 631± 639. Willett WC (1986): Selenium, vitamin E, ®ber and the incidence of human cancer: an epidemiological perspective. Avance. Experim. Med. Biol. 206, 27±34.

Willett WC & Stampfer MJ (1986): Total energy intake: Implications for epidemiologic analyses. Am. J. Epidemiol. 124, 17±27. Willett WC (1990): Anthropometric measures and body composition. In: Nutritional Epidemiology (1st edn), ed. WC Willett, pp 143±216, 217± 244. Oxford University Press: Oxford. Witschi JC (1990): Short-term dietary recall and recording methods. In: Nutritional Epidemiology (1st edn) ed. WC Willett, pp 52±68. Oxford University Press: Oxford.