Children and teens who drink a lot of soda pop are twice as likely to steal, beat someone up or bring a weapon to school compared with peers who don't drink it, according to a new review of Minnesota student survey data.

The data, compiled by state officials after an inquiry from the Star Tribune, is far from suggesting cause and effect. It is likely, researchers say, that other social or biological factors could make teens prone to both violence and drinking large amounts of pop.

But when coupled with a study released Tuesday, which evaluated similar survey results in Boston, the data provide some of the first published evidence that pop consumption might have any relationship to youth aggression.

"If we want to understand youth violence and we want to reduce it, then we want to look at everything that can impact it," said Sara Solnick, chairwoman of the University of Vermont's economics department and co-author of the Boston study. "This was something that was not on the radar."

The Boston study was based on a 2008 survey by 1,878 ninth- to 12th-graders. Researchers compared students' self-reported violent behaviors with the number of cans of non-diet carbonated beverages they said they consumed in the most recent seven-day period. (A 20-ounce bottle counted as two cans.)

The study, published in Injury Prevention, found the probability of violence was 9 to 15 percentage points higher among Boston teens who drank five or more cans in a one-week period than among teens who drank four or fewer of the beverages. (Violence was defined as whether students carried weapons in the previous year or attacked classmates, young relatives or people they were dating.)

The American Council on Science and Health, which often provides pro-business perspectives, responded quickly, calling Solnick's results flawed, partly because they were based on self-reports from students who can exaggerate. "It's a shame this poor excuse for science got so much attention," said the council's Dr. Gilbert Ross.

But a similar pattern shows up in figures from Minnesota youth surveys.

After the announcement of the Boston study, the Star Tribune requested similar data from the Minnesota Department of Health, which oversees a survey of students every three years in public schools statewide.

Among the Minnesota findings:

• Among students who drank no pop during the day before they completed the 2010 Minnesota Student Survey, 17 percent admitted "beating up" someone in the prior year. That number jumped to 37 percent among students who drank three or more sodas during the previous day.

• 5.5 percent of sixth-graders who drank no pop during the previous day said they had run away from home at least once in the prior year. The number jumped to 17.3 percent among those who drank at least three glasses of pop.

• 13.3 percent of the no-pop students reported they had stolen something in the previous year; that number was 30.7 percent for heavy soda drinkers.

The Minnesota and Boston surveys are not identical. Students in Boston were asked about pop consumption in the previous week, Minnesota students in the previous day. Boston students were asked about cans of non-diet drinks. The Minnesota survey used "glasses" as the serving size, and wasn't specific about the type of pop consumed. Most of the responding students in Boston were inner-city minorities. The Minnesota survey was statewide.

Poverty? Diet?

The Minnesota student data were provided to the Star Tribune on Monday in raw form, meaning the information had not been weighted to account for such demographic factors as race or economic status. That leaves open the possibility, for example, that pop consumption is simply more common among low-income students, who are also more prone, statistically, to violence.

National data have already shown higher rates of consumption of sugar-sweetened beverages by certain ethnic, racial and low-income groups, said Mary Story, a University of Minnesota expert in youth dietary habits who was not involved with the Boston study.

"The association does not surprise me at all," Story said. "I think this is all about poverty. Poor children have a lot working against them. They are more likely to have a poor diet and drink more soda and sugary drinks."

The Boston study did weight the data to factor out race and gender, but not economic status. Heavy pop drinkers were also more likely to smoke and drink alcohol, factors that are strongly associated with violence. But Solnick said the relationship between pop consumption and aggression held up after accounting for those other factors.

Solnick said there is no reason to think pop consumption causes students to be aggressive. Anything from low blood sugar to poor parenting could be at fault. It's possible, she said, that pop consumption is a "red flag" of an overall poor diet, and that the absence of key nutrients makes students prone to aggression.

"Maybe," Solnick said, "the soda is just telling us what's not there."

Jeremy Olson • 612-673-7744