Chocolate consumption and incidence of hypertension.

Chocolate Consumption and Incidence of Hypertension To the Editor: Results from the well-conducted randomized trial by Grassi et al1 underline the positive effect of flavanols on blood pressure and endothelial function. However, this study and others2 fail to answer the important questions about the effect of cocoa consumption in a free-living population consuming standard commercial chocolate products. To address this issue, we evaluated the association between chocolate consumption and incidence of hypertension in the Seguimiento Universidad de Navarra Study, a dynamic cohort of university graduates recruited and followed-up biennially through mailed questionnaires.3 The enrollment of participants is permanently open. In this analysis, we have included participants recruited from December 1999 to January 2002. Diet was assessed with a semiquantitative food frequency questionnaire (136 items) validated previously in Spain.4 New cases of medically diagnosed hypertension were reported in the 2-year follow-up questionnaire. Participants with cardiovascular disease, cancer, diabetes or hypertension at baseline, and those with missing values in covariates were excluded. Finally, 5880 participants (mean age: 35.8 years, 61.3% female) from the 6686 eligible answered the 2-year follow-up questionnaire (88% retention in the cohort). After 13 526 person-years of follow-up, we identified 180 new cases of hypertension. Overall, chocolate consumption was not associated with the risk of hypertension (Table 1). Adjustment for other dietary variables presumably related to blood pressure levels did not appreciably change the results. These results can be partly explained by random error in dietary assessment, resulting in a lack of association. Misclassification in the outcome ascertainment and the presence of unmeasured confounding could bias our results as well. Nonetheless, both diet and self-reported hypertension diagnosed have been validated,4,5 and our analyses have been adjusted for potential nondietary and dietary confounders. Differences between the results of short-term trials and our observational study can be explained by several factors. Feeding trials usually use pure dark chocolate, rich in flavanols, whereas the chocolate consumed by the general population has relatively low quantities of cocoa and even lower levels of flavanols. Additionally, individuals do not consume chocolate in isolation, but in a dietary pattern characterized by frequent snacking and high-energy foods that can foster the development of obesity and, consequently, may counteract any beneficial effect of dark chocolate on blood pressure. In fact, chocolate consumption was associated with snacking (as a dichotomous variable) in our study population (P 0.001 for the lineal trend between chocolate consumption and the habit of between-meal snacking, adjusted for age and sex). Therefore, to avoid confusion among the general public, caution is needed when sending messages to the general population about this issue. This is especially true when apparent conflicting information can reach the public, creating confusion and, in the end, distrust in sound nutrition research.