Natural Solutions Radio header image

More Proof Soft Drinks Tied to Diabetes

One can -- or one more can -- of soda daily significantly increased the risk of diabetes in a European population, researchers found.

Consuming an additional 12 ounces of sugar-sweetened beverages -- a standard size can of soda -- daily was associated with a 22% (95% CI 1.09 to 1.38) increased risk for diabetes, and an extra can of artificially-sweetened beverage put the risk 52% higher (95% CI 1.26 to 1.83), according to Dora Romaguera-Bosch, PhD, MSc, of Imperial College London in England, and colleagues at the InterAct consortium.

However, only the association between sugar-sweetened beverages and diabetes risk remained significant after adjustment for body mass index and energy intake (HR 1.18, 95% CI 1.06 to 1.32), they wrote online in the journal Diabetologia.

Soda and other artificially- and sugar-sweetened beverages may contribute to diabetes due to their effect on body weight and blood glucose, the researchers wrote. A link between sugary drink consumption and diabetes has also been seen in prior research in American populations.

The authors studied associations between consumption of juices, nectars, sugar-sweetened soft drinks, and artificially-sweetened soft drinks with diabetes incidence in a case-cohort study of 15,374 participants from eight European countries, and including 11,684 incident cases of diabetes.

Cohorts from eight countries were involved in the EPIC (European Prospective Investigation into Cancer and Nutrition)-InterAct study. InterAct specifically looked at incidence of diabetes.

Participants answered dietary questionnaires that measured:

Food intake over the year prior to the survey Individual energy intake Consumption of juices, nectars, and soft drinks Sugary drink consumption was measured in tiers that included less than one glass monthly, one to four glasses monthly, more than one to six glasses weekly, and one or more glasses daily.

Covariates included smoking status, alcohol intake, education, physical activity, BMI, and, "in most participating centers," baseline chronic conditions, including hypertension, hyperlipidemia, and cardiovascular disease.

Raw scores were also calculated with adjustment for age, sex, BMI category, and physical activity levels.

Compared with those who consumed lower levels of sugary soft drinks, high-level consumers were more likely to be male, physically active, less educated, smokers, and have a higher waist circumference. Juice and nectar high consumers were mostly younger, female, physically active, former smokers, and better educated than those with lower juice and nectar consumption.

Although soft drink consumption was linked with diabetes incidence, there was no association between diabetes and consumption of juices and nectars. One or more glasses daily of soft drink was associated with a 58% increased risk of diabetes compared with those who consumed the lowest levels of soft drinks (P<0.0001).

When this comparison was adjusted for BMI, this trend remained significant, though the association was not as strong (P=0.0005).

Artificially-sweetened soft drink consumption was associated with diabetes risk prior to adjustment for BMI (P<0.0001), but this association lost significance after adjustment (P=0.24).

Consumption of sugar-sweetened soft drinks remained significantly associated with diabetes both before (P<0.0001) and after (P=0.013) adjustment for BMI.

"The observed association between sugar-sweetened soft drinks and diabetes in the present analysis is of similar magnitude as the association reported in a meta-analysis of eight prospective studies, which was based on 15,043 diabetes cases mostly from the U.S.A.," they concluded.

They also noted that future studies should examine whether BMI acts as a mediator or confounder with the association.

The authors said the study was limited by the clinical definition of type 2 diabetes, lack of follow-up dietary data, a definition of juices and nectars that included those with and without added sugar, and possible measurement errors. The study was also limited by center-specific factors, possible residual confounding, and possible reverse causality.