In response to “The Relationship of Sugar to Population-Level Diabetes Prevalence: An Econometric Analysis of Repeated Cross-Sectional Data,” a correlation study appearing in the February issue of the journal PLoS ONE, the American Beverage Association issued the following statement:


“This study does not show – or even attempt to show – that consuming sugar causes diabetes. In fact, its most robust finding confirms the well-established relationship between obesity and diabetes risk.  Importantly, however, the study’s conclusions on sugar and diabetes should be viewed cautiously given that the underlying model failed to consider the potential impact of solid fats – such as butter, cheese and lard – or factor for family history.”

Additional Background Information:

 On the Article:

This is an epidemiological (or correlation) study, which cannot and does not prove causation – as noted by co-author Dr. Robert Lustig in a press release issued by Stanford University. Lead author Dr. Sanja Basu noted that “the findings do not prove sugar causes diabetes.”  The release also notes that Basu “was cautious about possible policy implications of his work, stating that more evidence is needed before enacting widespread policies to lower sugar consumption.” The authors note there are several limitations to their analysis:  the potential exists for ecological fallacies; the authors track caloric availability, rather than consumption; the analysis does not distinguish between any specific added sugars (such as sucrose or HFCS) or any specific source of sugar (such as soda or processed food); the statistical power is limited based on the ecological approach; and the database used does not distinguish between Type 1 and Type 2 diabetes. The primary finding of the study confirms the well-established connection of obesity prevalence and diabetes.  Beyond that relationship the model used needs to be viewed with caution as the estimate of the relationship of sugar availability and diabetes risk did not account for solid fat intake, dairy intake or any factors of heritability which would augment the findings concerning sugar. The study also failed to differentiate between Type 1 (10 percent of cases) and Tpe 2 (90 percent of cases) diabetes.  Importantly, Type 1 diabetes is not thought to be affected by diet.

On Diabetes:

According to the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) of the National Institutes of Health (NIH), , those at greater risk for Type 2 diabetes include certain racial and ethnic groups (such as Hispanic/Latino, Asian Americans and Pacific Islanders, and American Indians and Alaska Natives), as well as those who: are over age 45; have a family history of diabetes; are overweight; do not exercise regularly; have low HDL cholesterol or high triglycerides; have high blood pressure; have impaired glucose tolerance (IGT) and/or impaired fasting glucose (IFG); have a history of cardiovascular disease; have polycystic ovary syndrome; have other clinical conditions associated with insulin resistance. Many of these factors were omitted in the Basu et al study. In addition, women who had gestational diabetes, or who have had a baby weighing 9 pounds or more at birth, are at greater risk. Again, this type of diabetes was not accounted for in the Basu et al study. Additionally, a study published late last year in the Journal of Nutrition which looked at eight European countries found no association between digestible carbohydrate, including sugar, and diabetes risk.


# # #

The American Beverage Association is the trade association representing the broad spectrum of companies that manufacture and distribute non-alcoholic beverages in the United States.  For more information on ABA, please visit the association’s Web site at or call the ABA communications team at (202) 463-6770.