Association or Causation of Sugar-Sweetened Beverages and Coronary Heart Disease

In this issue of Circulation , de Koning and colleagues1 evaluate the association between sugar-sweetened beverage (SSB) consumption, incident coronary heart disease (CHD), and biomarkers associated with cardiovascular risk using data from the Health Professionals' Follow-up Study. This analysis is similar to this research group's previous evaluation using data from the Nurses' Health Study, with similar longitudinal follow-up of >20 years.2 The results mirror those reported previously (relative risk, 1.19; 95% confidence interval, 1.11–1.28) for incident CHD associated with 1 serving per day higher SSB consumption in the Health Professionals' Study and add to the growing body of information that suggests an independent association between SSBs and worse cardiovascular health.Article see p 1735Data from the National Health and Nutrition Examination Survey suggest that ≈2 of every 3 Americans drink SSBs daily, with rates reaching as high as 4 in 5 among young (20–44 years) black men.3 Calorie intake from SSBs among individuals who drink SSBs has also risen from 239 kcal/d in 1988–1994 to 294 kcal/d in 1999–2004, which highlights the relevance of the authors' research question. The analyses appear internally valid, but the question remains as to whether these associations are causal or not, which may be particularly difficult to discern given the risk of residual confounding present in many dietary studies, especially those that rely on self-reported data, as was done in both the Health Professionals' Follow-Up Study and the Nurses' Health Study. A review that uses criteria to assess causation, as outlined by the famed British epidemiologist and biostatistician Sir Austin Bradford Hill (1897–1991) in his 1965 presidential address to the United Kingdom's Royal Society of Medicine, may help place this research into perspective.4 Strength of AssociationBradford Hill's research legacy lay in the association between tobacco and lung cancer, which had a relative …