Serum concentrations of ß-carotene and -tocopherol are associated with diet, smoking, and general and central adiposity

¹ From the Departments of Medicine and Community Medicine, Lund University, Malmö University Hospital, Malmö, Sweden.

² Supported by the Swedish Cancer Society (2684-B93-05XAA), the Swedish Medical Research Council (B93-39X-09534-03C), the Ernhold Lundström Foundation, and the Swedish Nutrition Foundation.

³ Address reprint requests to P Wallström, The Malmö Diet and Cancer Study, Medical Research Centre, Malmö University Hospital, Malmö SE-205 02, Sweden. E-mail: peter.wallstrom{at}smi.mas.lu.se.

ABSTRACT  
Background: Previous studies of associations between diet, obesity, and blood concentrations of -tocopherol and ß-carotene have been equivocal. Furthermore, most studies used only body mass index (BMI) as an obesity measure.

Objectives: Our objectives were to examine the associations between energy and nutrient intakes, alcohol consumption, tobacco use, and serum cholesterol and serum concentrations of -tocopherol and ß-carotene, and to examine the associations between different measures of general and central adiposity and serum concentrations of -tocopherol and ß-carotene.

Design: This was a cross-sectional, population-based study of 253 men and 276 women aged 46–67 y. Nutrient data were collected by a modified diet history method. Measures of obesity included BMI, percentage of body fat (impedance analysis), waist-to-hip ratio, and waist circumference. The associations between serum nutrient concentrations and the other factors were examined by multiple linear regression.

Results: Twenty-one percent of men and 34% of women used antioxidant supplements. The mean BMI was 26.1 in men and 25.4 in women. Serum ß-carotene concentration was positively associated with serum cholesterol concentration, fiber intake, and ß-carotene intake, and negatively associated with smoking and all measures of obesity. In men, serum ß-carotene concentration was not significantly associated with central adiposity after adjustment for body fat. Serum -tocopherol concentration was positively correlated with serum cholesterol, obesity, and vitamin E intake. In women, serum -tocopherol concentration was also positively associated with intakes of ascorbic acid and selenium. Serum -tocopherol concentration was associated with central adiposity after adjustment for body fat.

Conclusion: Serum ß-carotene and -tocopherol concentrations have different associations with diet, smoking, general adiposity, and central adiposity.

Key Words: Vitamin E • ß-carotene • -tocopherol • diet • obesity • body composition • body mass index • waist-to-hip ratio • smoking • alcohol • cross-sectional studies

INTRODUCTION  
The antioxidants -tocopherol, the dominant component of vitamin E, and ß-carotene contribute to the body's defense against reactive oxygen species (1–5). High blood concentrations of these nutrients were prospectively shown to be associated with low incidences of cardiovascular disease and several cancers (3, 6–8). Similar results, although less conclusive, were obtained in prospective studies of the dietary intake of these nutrients and the risk of cancer and cardiovascular disease (9–11).

It has been argued that low ß-carotene or -tocopherol serum concentrations may not indicate a causal link between these nutrients and disease but rather a relative absence of other protective dietary factors. This was suggested, particularly for ß-carotene, with the results of 3 randomized controlled trials in which relatively long-term ß-carotene supplementation in high-risk (7, 12) and low-risk (13) populations failed to produce any protective effects. Still, low blood concentrations of -tocopherol and ß-carotene are markers of disease risk. Moreover, carotenoids, such as ß-carotene, in plasma or serum may be regarded as markers of fruit and vegetable intake (14, 15) and there is evidence that a high consumption of fruits and vegetables protects against cancer and cardiovascular disease (16).

There are no straightforward relations between the measured average intake of ß-carotene and -tocopherol and ß-carotene and -tocopherol blood concentrations. First, the daily variation in blood concentrations is considerable (17, 18) and may be due in part to a variation in dietary intake. Second, all instruments used for measuring diet in large groups of individuals entail substantial measurement errors (19). Third, there are interindividual differences in biological properties, eg, differences in absorption, metabolism, and tissue distribution of nutrients (5, 20–22). Other factors that may influence these associations include the intake of other nutrients, smoking (23–30), alcohol use (24–29, 31, 32), plasma lipid profile (23, 24, 27), and obesity.

Obesity, as assessed by body mass index (BMI), has been reported to be independently related to -tocopherol and ß-carotene concentrations, even when diet and lipid concentrations are adjusted for (23, 25, 27–30, 32, 33). However, BMI, by definition, is a measure of weight adjusted for height, and thus can be assumed to be a good marker of obesity only when the variation between individuals is mostly because of differences in fat mass, not fat-free mass (34). Direct measures of body composition, such as body fat content or fat distribution [eg, waist-to-hip ratio (WHR)], may be useful in identifying the role of obesity in determining serum concentrations of -tocopherol and ß-carotene. Furthermore, measures of general and central adiposity are known to be associated differently with several chronic conditions, eg, myocardial infarction (35). One could hypothesize that the associations of central adiposity with blood nutrient concentrations may be different from those between general obesity and respective blood markers.

The first objective of this study from the Malmö Diet and Cancer Study (MDCS) cohort was to examine the association between energy and nutrient intakes, alcohol, tobacco use, and serum cholesterol and serum concentrations of -tocopherol and ß-carotene in a population-based sample of middle-aged and older men and women. Our second objective was to examine the associations between serum -tocopherol and ß-carotene and different measures of obesity, reflecting either general adiposity (BMI and percentage of body fat) or central adiposity (waist circumference and WHR).

SUBJECTS AND METHODS  
Study population
The background population for this study consisted of all men and women born between 1926 and 1945 and living in Malmö (n = 53325 in 1991), Sweden's third largest city. The population was identified through the Swedish national population registries. A random sample of 1000 persons was selected from this population. The probands were invited by mail. Several attempts were made to reach each nonresponder by telephone. Five hundred forty-one persons completed the study. We obtained all relevant serum measurements from all but 11 subjects. One person was excluded because of an extremely high vitamin E intake, resulting in a serum -tocopherol value >150 µmol/L, which was 15 SDs beyond the mean of the other participants. Thus, the analytic sample for this study consisted of 529 subjects: 253 men and 276 women. Data were collected during 1991–1994. The design of the MDCS was approved by the Ethics Committee at Lund University.

Data collection
Study subjects visited the MDCS center twice. At the subject's first visit, project staff provided information on the background and aim of the project and detailed instructions about the lifestyle questionnaire and the dietary data collection procedure. Blood samples were drawn and processed. The subjects were not fasting. At the second visit, subjects were randomly assigned to interviewers who worked with each subject to complete the diet history and to check the lifestyle questionnaire for erroneous answers.

Dietary assessment
The MDCS dietary assessment method was described previously (36, 37). Briefly, it is a modified diet history method that combines quantitative and semiquantitative approaches and measures the entire diet, including cooking methods. It consists of 2 parts: a menu book to record descriptions of cooked meals, cold beverages (including juices and alcoholic beverages), and dietary supplements during 7 consecutive days, and a 168-item questionnaire on foods consumed regularly (other than cooked foods) during the past year. Data on the reproducibility (38) and validity (39, 40) of the method were published previously.

The amount of food and nutrients consumed was calculated with KOSTSVAR software (AIVO AB, Stockholm) with the use of the MDCS nutrient database, in which most of the nutrient information comes from the PC-KOST2-93 database (National Food Administration, Uppsala, Sweden). Names and contents of registered dietary supplements were provided by the Medical Products Agency (Uppsala, Sweden). Information on other dietary supplements was gathered from manufacturers, sales representatives, wholesale dealers, and retailers. Nutrient values of dietary supplements that could not be identified were replaced by the median intake from the type of supplement in question (3.8% of cases). If the supplement type could not be identified, the supplement was ignored (0.8% of cases). A person was considered a user of antioxidant supplements if he or she reported consuming a supplement that contained an antioxidant, ß-carotene, or selenium at least once during the registration.

The most common way of consuming vitamin E was in the form of multivitamin tablets. Such tablets typically contained 10 mg all-rac--tocopherol, which, according to current thinking, corresponds to 6.7 mg RRR--tocopherol (6.7 -tocopherol equivalents, -TE). Single-nutrient vitamin E preparations were uncommon but existed in amounts 100 mg all-rac--tocopherol (67 -TE). Use of ß-carotene supplements was uncommon during the time of this study.

Anthropometric examination and impedance analysis
Weight was measured to the nearest 0.1 kg on a balance-beam scale. Standing height was measured with a fixed stadiometer calibrated in centimeters. BMI was calculated as weight in kg divided by height in meters squared (kg/m²). Waist and hip circumferences of each participant were measured by a trained nurse (waist: midway between the lowest rib margin and the iliac crest; hip: horizontally at the level of the greatest lateral extension of the hips) and were used to calculate WHR. Both WHR and waist circumference were used as measures of central adiposity.

Bioelectrical impedance analysis was used to estimate body composition according to procedures provided by the manufacturer (BIA 103 single-frequency analyzer; RJL Systems, Detroit). The algorithm used to estimate body fat from impedance was supplied by the manufacturer and was used to calculate percentage of body fat.

Calculating the underreporting of energy intake
Basal metabolic rate (BMR) was calculated according to the formula of Garby et al (41):

RESULTS  
Descriptive statistics
Participants are described in Table 1. More women than men had never smoked and more men were either ex- or current smokers. More women than men reported low or no alcohol intake, the women were more frequent users of antioxidant supplements, and the men were more likely to be overweight.

View this table:
TABLE 1. . Characteristics of the study population by sex  
Dietary intakes and antioxidant serum concentrations of the participants are shown in Table 2. Women consumed less energy, dietary fiber, vitamin E, and selenium and more ascorbic acid and ß-carotene than did men. Women had 50% higher serum concentrations of ß-carotene than did men, but there were no significant differences in -tocopherol concentrations, neither crude nor when expressed as an -tocopherol–cholesterol ratio. Women had slightly higher serum cholesterol concentrations than did men. There was no significant difference between men and women in underreporting energy intake (50% of men compared with 45% of women), but underreporting was strongly and positively associated with BMI (P <0.001) in both sexes (data not shown). The median intakes of vitamin E supplements were the same in men and women (6.7 -TE/d), although men consumed slightly higher doses overall. The range of intakes was 0.7–182 -TE/d in men and 0.1–108 -TE/d in women. Only 8 persons reported the use of ß-carotene supplements, 7 of whom were women.

View this table:
TABLE 2. . Average daily nutrient intakes and serum nutrient concentrations  
Correlational analysis
The age-adjusted correlations between serum antioxidants, dietary intake, alcohol consumption, tobacco use, physical activity, obesity estimates, and serum cholesterol concentrations are shown in Table 3. The correlations between serum -tocopherol and ß-carotene concentrations were 0.34 in men and 0.25 in women. Serum ß-carotene concentration was positively associated with serum -tocopherol concentrations, serum cholesterol concentration, and the intakes of dietary fiber, ascorbic acid, and ß-carotene in both men and women. There was also a positive association with selenium intake in women. The amount of tobacco smoked was negatively associated with serum ß-carotene concentrations in men. All obesity measures were negatively associated with serum ß-carotene concentrations in men and women, but none, except for percentage of body fat, were statistically significant in men.

View this table:
TABLE 3. . Associations between serum ß-carotene, serum -tocopherol, and other selected factors (age-adjusted correlation coefficients)  
Serum -tocopherol concentrations were strongly correlated with serum cholesterol concentrations in both men and women (Table 3). They were also associated with total intakes of vitamin E and selenium, and, in women, with ascorbic acid. All obesity measures were both significantly and positively associated with serum -tocopherol concentrations in both sexes, except for percentage of body fat in men.

Multivariate analyses
For each combination of sex and serum nutrient, 4 multiple linear regression models were specified, 1 for each obesity measure. The independent associations between serum nutrients and the independent variables were not substantially altered when any of the 4 obesity measures was included. Because BMI was used most commonly in previous studies, we decided to present the models that included BMI as the obesity measure. The description of the results, however, relates to all of the examined models.

The multivariate analyses of serum ß-carotene concentrations, illustrated in Table 4, indicated that cholesterol concentration and general and central adiposity were the strongest markers of serum ß-carotene in both men and women. Cholesterol was positively and obesity was negatively associated with serum ß-carotene concentrations. Intake of ß-carotene was a positive marker of serum ß-carotene concentration in women, but was not significant in men. Current smoking was associated with low ß-carotene concentrations in men, but not in women. In women, there was a negative relation between energy intake and ß-carotene concentration and a positive association between dietary fiber and serum ß-carotene concentration.

View this table:
TABLE 4. . Markers of serum ß-carotene in men and women (multiple linear regression)  
Serum cholesterol concentration was positively associated with serum -tocopherol concentration and was the strongest marker of serum -tocopherol in both men and women, as illustrated in Table 5. Vitamin E intake was associated with serum -tocopherol concentrations, but not as strongly as was serum cholesterol concentrations. In women, serum -tocopherol concentrations were positively associated with the intakes of ascorbic acid and selenium, but negatively associated with the intake of ß-carotene. In contrast with serum ß-carotene, obesity measures were positively associated with serum -tocopherol, although the association was not significant in all analyses (see below).

View this table:
TABLE 5.. Markers of serum -tocopherol in men and women (multiple linear regression)  
The proportion of the variation in serum ß-carotene values explained by the models (adjusted R²) ranged from 22.3% to 25.3% in men and from 22.0% to 23.6% in women. Corresponding ranges for serum -tocopherol were 45.8–48.5% in men and 52.4–53.4% in women. The difference between explanatory values for serum -tocopherol and ß-carotene concentrations was due mainly to the higher correlation of serum cholesterol with serum -tocopherol than with serum ß-carotene concentrations. No significant interactions between intake of ß-carotene or -tocopherol and smoking status were found.

Partial correlation coefficients of BMI, percentage of body fat, waist circumference, and WHR with serum nutrient concentrations, when each of the obesity measures was entered into the separate multivariate models, are shown in Table 6, as was shown previously for BMI in Tables 4 and 5. All obesity measures were negatively associated with serum ß-carotene concentration. Waist circumference and WHR were positively and significantly associated with serum -tocopherol concentration in both men and women, but measures of general obesity, ie, BMI or percentage of body fat, were not significant, with the exception of BMI in men.

View this table:
TABLE 6. . Summary of the independent associations of 4 obesity measures with serum ß-carotene and -tocopherol (partial correlation coefficients)¹  
There was little change in the associations between central adiposity and serum -tocopherol when adjusting for percentage of body fat, indicating an independent effect of central adiposity (Table 6). In contrast, the correlation between central adiposity and serum ß-carotene concentrations became nonsignificant in men, whereas it decreased in women when percentage of body fat was included in the model.

DISCUSSION  
The present study indicated different associations between serum ß-carotene and -tocopherol concentrations and diet, central adiposity, general adiposity, and smoking.

Diet
Fiber intake was an equally consistent marker of elevated serum ß-carotene concentrations as was ß-carotene intake. Vegetables are important sources of fiber and carotenoids other than ß-carotene, and it was shown that carotenoids may interact in vivo (49). Thus, serum ß-carotene concentrations may be increased by the dietary intake of other carotenoids, as reflected by an independent association between high fiber intake and serum ß-carotene concentration.

Serum -tocopherol concentration was positively associated with total vitamin E intake (diet and supplements). In contrast with the relation in men, serum -tocopherol concentrations in women were positively correlated with the intake of ascorbic acid and selenium. These observations appear biologically plausible: selenium is a cofactor of intracellular glutathione peroxidase, which, with vitamin E, aids in the prevention of membrane lipid peroxidation. Thus, a synergistic effect between intracellular vitamin E and glutathione peroxidase is likely (50). Ascorbic acid can convert the oxidized -tocopherol molecule back into -tocopherol, although the importance of this mechanism in vivo is not clear (5). Another possible explanation for the observation is that many women in the MDCS consumed multivitamin supplements, which tend to contain large doses of vitamin E, selenium, and ascorbic acid (51).

Obesity
In the present study, serum ß-carotene concentration was negatively associated with both general and central adiposity, whereas serum -tocopherol concentration was positively associated with central adiposity (and with BMI in men). Some previous studies reported negative associations between -tocopherol concentrations and central adiposity, in both serum (52) and adipose tissue (53). It was suggested that part of the increased incidence of cardiovascular disease associated with central adiposity may be caused by low -tocopherol status. Our findings do not support this hypothesis. We observed a significant, positive association between serum -tocopherol concentration and central adiposity, both unadjusted and adjusted for diet and serum cholesterol concentration. In men, we also observed a positive association between BMI and serum -tocopherol concentration. To test whether these findings were caused by supplement use, we tested the multivariate models with terms for interaction between supplemental vitamin E use and the obesity measures. No significant interactions were found. In addition, we tested the original models without vitamin E supplement users. The results were not substantially altered (data not shown).

The independent negative association between general obesity and serum ß-carotene concentration observed in this study might have at least 2 explanations. The first is that ß-carotene is distributed between plasma and adipose tissue, adipose tissue being the dominant storage tissue in man (22). Accordingly, a person with high fat mass would have a larger proportion of ingested ß-carotene absorbed by fat tissue than would a lean person if all other metabolic factors were equal. A second explanation could be that obese persons consume less food that would increase serum ß-carotene concentrations, and this might not be adequately captured by fiber and ß-carotene intake estimates. Other causes of the negative association between general obesity and serum ß-carotene concentration were suggested by Hebert et al (23), who proposed, for example, a potential influence of differential bias in the reporting of dietary intake. There are few data to support this latter suggestion. Observations that are consistent with overreporting of nutrient-dense foods by obese persons, however, were made in the MDCS cohort (54).

Smoking
Similar to other studies (55), male smokers had lower concentrations of serum ß-carotene than did male nonsmokers. Low ß-carotene concentrations in smokers may be caused either by low ß-carotene intake (55) or by increased metabolism of antioxidants (56). The quantitative contribution of ß-carotene to the serum antioxidant defense in smokers in vivo is not known (5, 56, 57). If this contribution were to be small, it would imply that smoking itself is not a major cause of low ß-carotene concentrations. On the other hand, there were not many heavy smokers in this study—only 35% of male smokers and 15% of female smokers smoked >15 cigarettes/d. This may explain the lack of an association between the amount of tobacco smoked and low ß-carotene concentrations in female smokers.

Limitations
As expected, serum -tocopherol and ß-carotene concentrations, serum -tocopherol in particular, were strongly associated with serum cholesterol concentrations. Another lipid status variable of importance is triacylgycerol, which is usually correlated with serum cholesterol concentrations. Triacylglycerol was not measured in this study. Triacylglycerol was shown to be associated with both ß-carotene (27, 28, 32) and -tocopherol (23, 27, 28, 32, 33), but more strongly associated with serum -tocopherol than with serum ß-carotene concentrations (23, 27). Indeed, the association between blood -tocopherol and triacylglycerol appears to be almost of the same magnitude as that between serum -tocopherol and serum cholesterol concentrations (27). Because serum triacylglycerol is associated more strongly with central adiposity than with general obesity (58, 59), it is possible that the association we observed between serum -tocopherol concentrations and central adiposity may have been due to some extent to confounding by triacylglycerol. We attempted to partly correct for this by adjusting the multivariate models for use of lipid-lowering drugs (mostly fibrates). However, use of these drugs was small (n = 9) and the adjustment did not change the multivariate models (data not shown).

Furthermore, we had access to blood samples from nonfasting subjects only. Because triacylglycerol concentrations vary markedly with the composition and time of the last meal, it is likely that serum nutrient concentrations vary as well. This could result in an increased, seemingly random variation in -tocopherol and ß-carotene concentrations, implying that the true correlations between intakes and serum concentrations might be stronger than what was shown in this study.

Underreporting is a common problem in dietary analyses. Obese adults tend to underreport energy intake (60), which was observed in recent studies from the MDCS (54, 61, 62). Also, in a large, nationwide sample from the United States, Ma et al (63) noted that desserts were the most important source of vitamin E; desserts are likely to be underreported in dietary analyses (60). However, we observed significant associations between serum -tocopherol concentration and central adiposity after adjustment for both energy intake and general obesity. Furthermore, post hoc analyses of the multivariate models, eg, stepwise addition of independent variables in varying order, did not support an important role of underreporting in the present results (data not shown).

The MDCS participants constitute a fairly representative sample of the Malmö population in the age categories studied. Yet, the MDCS participants may be different from the nonparticipants (healthy cohort) (J Manjer, S Carlsson, S Elmståhl, et al, unpublished observations, 2000). A likely result of this is smaller ranges of nutrient, alcohol, and tobacco intakes in the study group than in the background population, which implies that we might have found stronger associations between serum values and the dietary and lifestyle variables had the participation rate been higher. Additionally, the subsample used in this study had a higher participation rate than did the rest of the MDCS (J Manjer et al, unpublished observations, 2000).

Conclusion
We conclude that serum ß-carotene and -tocopherol concentrations in this sample of middle-aged citizens of Malmö were primarily influenced by serum cholesterol concentrations and measures of obesity, but also by smoking and intake of several antioxidant nutrients. Serum -tocopherol concentration was significantly and positively associated with central adiposity in both men and women; however, the results of associations between general obesity and serum -tocopherol concentration were inconclusive. Serum ß-carotene concentration was significantly and negatively associated with both general and central adiposity, but there was no independent association between central adiposity and serum ß-carotene concentrations in men after adjustment for body fat. The differences between measures of central and general obesity in their associations with serum antioxidants may warrant some consideration in future studies.

ACKNOWLEDGMENTS  
We thank Irene Mattisson and Ulla Johansson for helpful suggestions, the staff at the Department of Cariology at Umeå University for the serum nutrient analyses, and the staff and participants of the Malmö Diet and Cancer Study.

REFERENCES  

1. Cross CE, Halliwell B, Borish ET, et al. Oxygen radicals and human disease. Ann Intern Med 1987;107:526–45.
2. Guyton KZ, Kensler TW. Oxidative mechanisms in carcinogenesis. Br Med Bull 1993;49:523–44.
3. Gey KF. Prospects for the prevention of free radical disease, regarding cancer and cardiovascular disease. Br Med Bull 1993;49:679–99.
4. Gutteridge JM. Biological origin of free radicals and mechanisms of antioxidant protection. Chem Biol Interact 1994;91:133–40.
5. Rock CL, Jacob RA, Bowen PE. Update on the biological characteristics of the antioxidant micronutrients: vitamin C, vitamin E and the carotenoids. J Am Diet Assoc 1996;96:693–702.
6. Comstock GW, Bush TL, Helzlsouer K. Serum retinol, beta-carotene, vitamin E, and selenium as related to subsequent cancer of specific sites. Am J Epidemiol 1992;135:115–21.
7. Omenn GS, Goodman GE, Thornquist MD, et al. Risk factors for lung cancer and for intervention effects in CARET, the Beta-Carotene and Retinol Efficacy Trial. J Natl Cancer Inst 1996;88:1550–9.
8. Woodson K, Tangrea JA, Barrett MJ, Virtamo J, Taylor PR, Albanes D. Serum -tocopherol and subsequent risk of lung cancer among male smokers. J Natl Cancer Inst 1999;91:1738–43.
9. World Cancer Research Fund, American Institute for Cancer Research. Food, nutrition and the prevention of cancer: a global perspective. Washington, DC: American Institute for Cancer Research, 1997.
10. Willett WC. Micronutrients and cancer risk. Am J Clin Nutr 1994; 59(suppl):1162S–5S.
11. Key T. Micronutrients and cancer aetiology: the epidemiological evidence. Proc Nutr Soc 1994;53:605–14.
12. Albanes D, Heinonen OP, Taylor PR, et al. Alpha-tocopherol and beta-carotene supplements and lung cancer incidence in the alpha-tocopherol, beta-carotene cancer prevention study: effects of base-line characteristics and study compliance. J Natl Cancer Inst 1996;88:1560–70.
13. Hennekens CH, Buring JE, Manson JE, et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med 1996; 334:1145–9.
14. Martini MC, Campbell DR, Gross MD, Grandits GA, Potter JD, Slavin JL. Plasma carotenoids as biomarkers of vegetable intake: the University of Minnesota Cancer Prevention Research Unit feeding studies. Cancer Epidemiol Biomarkers Prev 1995;5:491–6.
15. Zino SJM, Harman SK, Skeaff CM, Mann JI. Fruit and vegetable consumption and antioxidant status of Seventh-day Adventists. Nutr Metab Cardiovasc Dis 1998;8:297–304.
16. Steinmetz KA, Potter JD. Vegetables, fruit, and cancer prevention: a review. J Am Diet Assoc 1996;96:1027–39.
17. Tangney CC, Shekelle RB, Raynor W, Gale M, Betz EP. Intra- and interindividual variation in measurements of ß-carotene, retinol, and tocopherols in diet and plasma. Am J Clin Nutr 1987;45:764–9.
18. Hunter DJ. Biochemical indicators of dietary intake. In: Willett WC, ed. Nutritional epidemiology. 2nd ed. New York: Oxford University Press, 1998:174–243.
19. Margetts BM, Nelson MA, eds. Design concepts in nutritional epidemiology. 2nd ed. Oxford, United Kingdom: Oxford University Press, 1997.
20. Drevon CA. Absorption, transport and metabolism of vitamin E. Free Radic Res Commun 1991;14:229–46.
21. Traber MG. Regulation of human plasma vitamin E. Adv Pharmacol 1997;38:49–63.
22. van Vliet T. Absorption of beta-carotene and other carotenoids in humans and animal models. Eur J Clin Nutr 1996;50(suppl):S32–7.
23. Hebert JR, Hurley TG, Hsieh J, et al. Determinants of plasma vitamins and lipids: the Working Well Study. Am J Epidemiol 1994; 140:132–47.
24. Hercberg S, Preziosi P, Galan P, et al. Vitamin status of a healthy French population: dietary intakes and biochemical markers. Int J Vitam Nutr Res 1994;64:220–32.
25. Drewnowski A, Rock CL, Henderson SA, et al. Serum ß-carotene and vitamin C as biomarkers of vegetable and fruit intakes in a community-based sample of French adults. Am J Clin Nutr 1997;65: 1796–802.
26. Nierenberg DW, Stukel TA, Baron JA, Dain BJ, Greenberg ER. The skin cancer prevention study group. Determinants of plasma levels of ß-carotene and retinol. Am J Epidemiol 1989;130:511–21.
27. Stryker WS, Kaplan LA, Stein EA, Stampfer MJ, Sober A, Willett WC. The relation of diet, cigarette smoking, and alcohol consumption to plasma beta-carotene and alpha-tocopherol levels. Am J Epidemiol 1988;127:283–96.
28. Bolton-Smith C, Casey CE, Gey KF, Smith WCS, Tunstall-Pedoe H. Antioxidant vitamin intakes assessed using a food-frequency questionnaire: correlation with biochemical status in smokers and non-smokers. Br J Nutr 1991;65:337–46.
29. Goodman GE, Thornquist MD, Kestin M, et al. The association between participant characteristics and serum concentrations of beta-carotene, retinol, retinyl palmitate, and -tocopherol among participants in the carotene and retinol efficacy trial (CARET) for prevention of lung cancer. Cancer Epidemiol Biomarkers Prev 1996;5:815–21.
30. Knekt P, Seppänen R, Aaran R-K. Determinants of serum -tocopherol in Finnish adults. Prev Med 1988;17:725–35.
31. Campbell DR, Gross MD, Martini MC, Grandits GA, Slavin JL, Potter JD. Plasma carotenoids as biomarkers of vegetable and fruit intake. Cancer Epidemiol Biomarkers Prev 1994;3:493–500.
32. Ascherio A, Stampfer MJ, Colditz GA, Rimm EB, Litin L, Willett WC. Correlations of vitamin A and E intakes with the plasma concentrations of carotenoids and tocopherols among American men and women. J Nutr 1992;122:1792–801.
33. Sinha R, Patterson BH, Mangels AR, et al. Determinants of plasma vitamin E in healthy males. Cancer Epidemiol Biomarkers Prev 1993;2:473–9.
34. Willett WC. Anthropometric measures and body composition. In: Willett WC, ed. Nutritional epidemiology. 2nd ed. New York: Oxford University Press, 1998:244–72.
35. World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO consultation on obesity, Geneva, 3–5 June 1997. Geneva: WHO, 1998. [WHO/NUT NCD/98.1.]
36. Callmer E, Riboli E, Saracci R, Åkesson B, Lindgärde F. Dietary assessment methods evaluated in the Malmö Food Study. J Intern Med 1993;233:53–7.
37. Wallström P, Wirfält E, Janzon L, et al. Fruit and vegetable consumption in relation to risk factors for cancer. A report from the Malmö Diet and Cancer Study. Publ Health Nutr (in press).
38. Elmståhl S, Gullberg B, Riboli E, Saracci R, Lindgärde F. The reproducibility of a novel diet history method and an extensive food frequency questionnaire. The Malmö Food Study. Eur J Clin Nutr 1996;50:134–42.
39. Elmståhl S, Riboli E, Lindgärde F, Gullberg B, Saracci R. The Malmö Food Study. The relative validity of a modified diet history method and an extensive food frequency questionnaire for measuring food intake. Eur J Clin Nutr 1996;50:143–51.
40. Riboli E, Elmståhl S, Saracci R, Gullberg B, Lindgärde F. The Malmö Food Study: validity of two dietary assessment methods for measuring nutrient intake. Int J Epidemiol 1997;26(suppl):S161–73.
41. Garby L, Garrow JS, Jorgensen B, et al. Relation between energy expenditure and body composition in man: specific energy expenditure in vivo of fat and fat-free tissue. Eur J Clin Nutr 1988;42:301–5.
42. Black AE, Goldberg GR, Jebb SA, Livingstone MBE, Cole TJ, Prentice AM. Critical evaluation of energy intake data using fundamental principles of energy physiology: 2. Evaluating the results of published surveys. Eur J Clin Nutr 1991;45:583–99.
43. Royal College of Psychiatrists. Alcohol: our favourite drug. London: Tavistock, 1986.
44. Taylor HL, Jacobs DR Jr, Schucker B, Knudsen J, Leon AS, Debacker G. A questionnaire for the assessment of leisure time physical activity. J Chronic Dis 1978;31:741–55.
45. Pero RW, Olsson A, Berglund G, Janzon L, Larsson SA, Elmståhl S. The Malmö biological bank. J Int Med 1993;233:63–7.
46. Pero RW, Olsson A, Bryngelsson C, et al. Quality control program for storage of biologically banked blood specimens in the Malmö Diet and Cancer Study. Cancer Epidemiol Biomarkers Prev 1998;7: 803–8.
47. Thurnham DT, Smith E, Parget Singh F. Concurrent liquid-chromatographic assay of retinol, alpha-tocopherol, beta-carotene, alpha-carotene, lycopene, and beta-cryptoxanthin in plasma, with tocopherol acetate as internal standard. Clin Chem 1988;34:377–81.
48. Wang Y, Ichiba M, Oishi H, Iyadomi M, Shono N, Tomokuni K. Relationship between plasma concentrations of beta-carotene and alpha-tocopherol and life-style factors and levels of DNA adducts in lymphocytes. Nutr Cancer 1997;27:69–73.
49. van Het Hof KH, West CE, Weststrate JA, Hautvast JG. Dietary factors that affect the bioavailability of carotenoids. J Nutr 2000;130: 503–6.
50. Hunt SM, Groff JL. Vitamin E. In: Advanced nutrition and human metabolism. St Paul: West Publishing Company, 1990:246–52.
51. Wallström P, Elmståhl S, Hanson BS, et al. Demographic and psychosocial characteristics of middle-aged women and men who use dietary supplements. Results from the Malmö Diet and Cancer Study. Eur J Public Health 1996;6:188–95.
52. Öhrvall M, Tengblad S, Vessby B. Lower tocopherol serum levels in subjects with abdominal adiposity. J Intern Med 1993;234:53–60.
53. Virtanen SM, van't Veer P, Kok F, Kardinaal AFM, Aro A. Predictors of adipose tissue tocopherol and toenail selenium levels in nine countries: the EURAMIC study. European Multicentre Case-Control Study on Antioxidants, Myocardial Infarction, and Cancer of the Breast. Eur J Clin Nutr 1996;50:599–606.
54. Wirfält E, Mattisson I, Gullberg B, Berglund G. Food patterns defined by cluster analysis and their utility as dietary exposure variables. Public Health Nutr 2000;3:159–73.
55. Rimm EB, Colditz GA. Smoking, alcohol, and plasma levels of carotenes and vitamin E. Ann N Y Acad Sci 1993;686:323–34.
56. Eiserich JP, van der Vliet A, Handelman GJ, Halliwell B, Cross CE. Dietary antioxidants and cigarette smoke-induced biomolecular damage: a complex interaction. Am J Clin Nutr 1995;62(suppl): 1490S–500S.
57. Niki E, Noguchi N, Tsuchihashi H, Gotoh N. Interaction among vitamin C, vitamin E, and ß-carotene. Am J Clin Nutr 1995;62(suppl): 1322S–6S.
58. Lilienthal Heitmann B. The effects of gender and age on associations between blood lipid levels and obesity in Danish men and women aged 35–65 years. J Clin Epidemiol 1992;45:693–702.
59. Seidell JC, Cigolini M, Charzewska J, et al. Fat distribution and gender differences in serum lipids in men and women from four European communities. Atherosclerosis 1991;87:203–10.
60. Macdiarmid J, Blundell J. Assessing dietary intake: who, what and why of under-reporting. Nutr Res Rev 1998;11:231–53.
61. Lahmann PH, Lissner L, Gullberg B, Berglund G. Sociodemographic factors associated with long-term weight gain, current body fatness and central adiposity in Swedish women. Int J Obes Relat Metab Disord 2000;24:685–94.
62. Lindström M, Hanson BS, Brunner E, et al. Socioeconomic differences in fat intake in a middle-aged population: report from the Malmö Diet and Cancer Study. Int J Epidemiol 2000;29:438–48.
63. Ma J, Hampl JS, Betts NM. Antioxidant intakes and smoking status: data from the Continuing Survey of Food Intakes by Individuals 1994–1996. Am J Clin Nutr 2000;71:774–80.