Re: β-Carotene: a Miss for Epidemiology

Extract

In his editorial, Marshall (1) discusses most of the arguments that explain the lack of protective effect of β-carotene supplementation and does not agree with the interpretation of the results from nutritional epidemiology as the best explanation.

All nutritional epidemiologists agree that dietary assessment methods are rather imprecise if not erroneous, rendering the nutrient exposure estimation exceedingly complex.

This argument is weaker when prospective epidemiologic studies rely on plasma β-carotene concentrations. Although intraindividual variations in plasma levels exist (2), the large numbers of measurements in prospective studies (3,4) are likely to alleviate this problem. But the other argument put forth by Marshall (1), which is that β-carotene is confounded by other nutrients present in the same food as β-carotene, is valid in these studies, and plasma levels of β-carotene may also be the marker of fruit and vegetable intake.

We propose that nutritional epidemiologists also take into consideration factors other than food intake that might influence the plasma β-carotene levels. We have some human and experimental evidence that plasma β-carotene levels are influenced by other types of exposure (e.g., exposure to oxidative stress and xenobiotics).