Nonlinear Models for the Relation between Cardiovascular Risk Factors and Beverage-type Effects: Is the Effect Due to Wine, Beer, or Spirits?

• Activity Number: 200
• Type: Contributed
• Date/Time: Tuesday, August 10, 2004 : 9:00 AM to 10:50 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Nav Randhawa*+
• Companies: CHEO Research Institute
• Keywords: nonlinear models ; Box-Tidwell power-linear model ; cardiovascular disease ; differential beverage-type effects

Abstract

An inverse association between moderate alcohol consumption and coronary heart disease has been shown in many epidemiological studies. However, it is not clear whether any specific type of alcoholic drink has a particular benefit. Nonlinear models are proposed for investigating the association between differential beverage type effects and risk factors for coronary heart disease (CHD). The parametric nonlinear models discussed are similar to those considered in Ambler et al. (2003) based on Box and Tidwell's power-linear model. The nonparametric generalized additive models provide a contrast to these models. The risk factors high-density lipoprotein cholesterol (HDL), systolic blood pressure (SBP), and fibrinogen are considered using data from the Whitehall II study of British civil servants. The results indicate that a substantial benefit is derived from the overall consumption of alcohol rather than specific beverage types, although increasing wine consumption is associated with slightly more favorable levels of all three cardiac risk factors. Findings also suggest that alcohol consumption relates positively with HDL and negatively with SBP and fibrinogen.