© 2005 Oxford University Press
CORRESPONDENCE |
Re: Fruit and Vegetable Intake and Risk of Major Chronic Disease
Affiliation of authors: Department of Agrotechnology and Food Sciences, Product Design and Quality Management Group, Wageningen University, Wageningen, The Netherlands
Correspondence to: Matthijs Dekker, PhD, Department of Agrotechnology and Food Sciences, Product Design and Quality Management Group, Wageningen University, P.O. Box 8129, 6700 EV Wageningen, The Netherlands (e-mail: matthijs.dekker{at}wur.nl).
We read the results on the associations between fruit and vegetable intake and risk of chronic diseases in the Journal (1) with great interest. The results were astonishing: Increased intake had no statistically significant association with cancer risk and had a small, statistically significant association with reduced risk of cardiovascular disease. An editorial in the same issue (2) questioned the accuracy of the intake data as well as the analysis of the multivariate measurement errors and especially the correlations between measurement errors of different variables. Many previous epidemiologic studies on fruit and vegetable intake have reported on their associations with cancer (3) and cardiovascular disease. Many of these studies have shown statistically significant reductions in disease risk, others have shown reductions in disease risk that were not statistically significant, and only a few have shown increased disease risk.
The fact that the effects are not randomly scattered around the no-effect level suggests that fruit and vegetable intake is indeed associated with reduced risk of chronic disease but that its statistical significance cannot be reliably assessed with epidemiologic studies as they are currently undertaken. The reason for this phenomenon may well be that it is not fruits and vegetables themselves that provide protection against chronic diseases but rather certain components within them. One can expect a clear answer from studies that assess associations between food intake and disease only if the levels of these components are relatively constant or if their variability is leveled out by studying a large enough cohort and/or a studying the cohort for a long enough time.
To test whether these conditions are fulfilled in current epidemiological studies, one needs quantitative information on variability in the amounts of active components. Candidate components showing the biggest potential as anticancer compounds are the secondary plant metabolites. The levels of these components in different cultivars show considerable variation. For example, levels of glucosinolates vary by more than 100-fold among different vegetables of the Brassica genus. This variation is even further expanded by taking into account the many ways and conditions of processing, storage, and preparation (4). So by analyzing food intake without any knowledge of the cultivars or the conditions in the food production chain, including the consumer, one introduces enormous uncertainty in the actual intake of protective components from these foods. The consequences of this uncertainty on the outcome of epidemiologic cohort studies have been simulated by Monte Carlo simulations (5). Even if one assumes a strong protective effect of a certain component in fruit and vegetables against cancer, these simulations showed that this effect would not be statistically significantly assessed in an epidemiologic cohort study with food intake as the main input.
We therefore recommend that food intake assessments include, at least, information on the method of preparation and type of (pre)processed products that the consumer is using. Predictive models to correct the level of certain components for the processing or preparation method may help to enhance the sensitivity of future nutritional epidemiologic studies. This additional information may lead to a more consistent outcome of this type of study, ultimately giving better clues to the question of which components in which fruits and vegetables may protect against chronic diseases.
REFERENCES
(1) Hung HC, Joshipura KJ, Jiang R, Hu FB, Hunter D, Smith-Warner SA, et al. Fruit and vegetable intake and the risk of major chronic disease. J Natl Cancer Inst 2004;96:1577–84.
(2) Schatzkin A, Kipnis V. Could exposure assessment problems give us wrong answers to nutrition and cancer questions? J Natl Cancer Inst 2004;96:1564–5.
(3) Steinmetz KA, Potter JD. Vegetables, fruit and cancer prevention: a review. J Am Diet Assoc 1996;96:1027–39.[CrossRef][Web of Science][Medline]
(4) Dekker M, Verkerk R, Jongen WM. Predictive modelling of health aspects in the food production chain, a case study on glucosinolates in cabbage. Trends Food Sci 2000;11:174–81.[CrossRef]
(5) Dekker M, Verkerk R. Dealing with variability in food production chains: a tool to enhance the sensitivity of epidemiological studies on phytochemicals. Eur J Nutr 2003;42:67–72.[CrossRef][Web of Science][Medline]
Related Correspondence
Response to this Correspondence
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2005 97: 608.
J Natl Cancer Inst 2005 97: 608-609.
This article has been cited by other articles:
|
|
 |
|
  V. Benetou, P. Orfanos, P. Lagiou, D. Trichopoulos, P. Boffetta, and A. Trichopoulou Vegetables and Fruits in Relation to Cancer Risk: Evidence from the Greek EPIC Cohort Study Cancer Epidemiol. Biomarkers Prev., February 1, 2008; 17(2): 387 - 392. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||