• Research Review
• Journal Review

Title: Association of smoking with serum and dietary levels of antioxidants in adults: NHANES III, 1988-1994.

Authors: Wei W, Kim Y, Boudreau N.

Reference: American Journal of Public Health 2001;91:258-164.

Reviewer: Joaquin Barnoya

Background: Smoking has been recognized as a risk factor for many chronic diseases. Oxidants and free radicals, contained in tobacco smoke, are known to cause damage to lipids, proteins, DNA, and carbohydrates. In defending the body against such insults, vitamin E, vitamin C, carotenoids, and selenium play an important role.

Carotenoids found in plant foods, function as vitamin A precursor. The 5 major serum carotenoids are lycopene, lutein/zeaxanthin, beta-carotene, beta-cryptoxanthin, and alpha-carotene. Association of lung cancer with low levels of serum and dietary intake of beta-carotene has been well established. Studies with other carotenoids are limited and inconsistent.

In this study the authors investigated the associations between smoking and serum concentrations and dietary intake of vitamin C, vitamin E, 5 major carotenoids, and selenium.

Methods: The National Health and Nutrition Examination Survey III (NHANES III) was conducted from 1988 to 1994 to examine the health and nutritional status of the civilian non-institutionalized US population. This study includes apparently healthy adults aged 17 to 50 years of age whose serum cotinine levels were available. Pregnant and lactating women were excluded as well as any adult suffering from a chronic disease. Subjects were classified as smokers (serum cotinine levels >14 ng/ml) or nonsmokers (serum cotinine levels <14 ng/ml). Intake data on vitamin C, vitamin E, beta-carotene, and selenium were available in the database. NHANES III collected and analyzed blood samples in designated laboratories.Means and medians of serum levels and dietary intakes of antioxidants were used in the analysis. Crude and adjusted values were tested with 2-tailed t test or analysis of variance.

Results: Included in the analysis were 3972 men and 3901 women; 37.4% of males and 25.8% of females were smokers. Smokers had lower education level, annual incomes, and leisure time physical activity, but consumed more alcohol. When compared to nonsmokers, a smaller percentage of smokers took vitamin or mineral supplements.

In both genders, significantly lower serum concentrations of vitamin C, vitamin E, alpha-carotene, beta-carotene, beta-cryptoxanthin, and lutein/zeaxanthin were observed in smokers when compared to nonsmokers. When adjusted for social factors, the reduction in serum levels of vitamin E and lycopene between smokers and nonsmokers was no longer seen. Regarding selenium, although the reduction seen was slight, it remained statistically significant after adjustment. Some gender

differences were observed for certain measure. Males had significant lower mean serum levels of vitamin C (p=.01), alpha-carotene (p=.003), and beta-carotene (p<.001). For vitamin E, beta-cryptoxanthin, lutein/zeaxanthin, lycopene, and selenium no gender differences were observed.

Smokers, compared to nonsmokers, had significantly lower dietary intakes of vitamin C and beta-carotene. The difference observed in vitamin E intake, was no longer statistically significant after adjusting for social factors.

Smokers were classified as nonsmokers (cotinine level <14 ng/ml), light smokers (14-100 ng/ml), moderate smokers (100-200 ng/ml), or heavy smokers (>200 ng/ml). Inverse and generally linear relations were found between serum levels of cotinine levels for vitamin C and E and beta-carotene in both genders. Vitamin C and beta-carotene showed differential and significant declines (p<.001) while the decline in vitamin E was not significant. For beta-carotene the biggest decrease

in serum levels was noted between smokers and light smokers, further increase in smoking intensity yielded non-significant decreases. The inverse association observed persisted after adjustment for dietary intake of a corresponding nutrient.

Dietary levels were divided into four groups according to percentiles. Serum levels of antioxidant and dietary intake were positively associated and generally linear for both sexes, even after adjustment for cotinine levels. Linearity was clearer for vitamin C (P<.001) than for beta-carotene (p<.001) or for vitamin E (not significant).

This analysis of the NHANES III data is consistent with previous studies that have shown that smokers had significantly (p<.001) lower serum levels and dietary intakes of vitamin C than did nonsmokers of both sexes. The decrease in serum levels of vitamin C appears to be the result of a combination between low dietary intakes and the effects of smoking. Previous studies have shown that lower levels of vitamin C are caused by impaired absorption or increased breakdown. These results add up to such evidence, as vitamin C was inversely associated with serum cotinine levels, even after adjustment for dietary intake.Regarding vitamin E, so far published literature is inconclusive. One advantage of this study is that by including only apparently healthy individuals the authors prevented the confounding of such factors such as disease and aging.

Smoking significantly reduced the serum levels of beta-carotene, alpha-carotene, beta-cryptoxanthin, and lutein/zeaxanthin. Lycopene was the only exception. The findings of beta-carotene are consistent with previous published data. The effect of smoking on the levels of serum beta-carotene, after adjustment for dietary intake, has been attributed to the enhanced turnover rate resulting from oxidative stress. Results for alpha-carotene and beta-cryptoxanthin yielded by this analysis are also consistent with previously published results. For lutein/zeaxanthin previous results are inconsistent. Results from this paper show a clear and significant reduce levels in smokers. Lycopene levels association with smoking has been reported elsewhere to be negative but insignificant. Although this data show slight reduction in serum levels, this difference is no longer significant after adjustment for social factors.

In conclusion, this study found that smoking negatively influenced serum concentrations of alpha-carotene, beta-carotene, lutein/zeaxanthin, beta-cryptoxanthin, and vitamin C.