Study Links Teens’ Soda Consumption to Violence

Harvard Magazine

ALREADY IMPLICATED in the obesity and diabetes epidemics, soda may be linked to violence in young people, new research suggests. In a study of 1,878 students at Boston public high schools, heavy soda drinkers were much more prone to violent behavior than other teens.

That finding came about by accident. While seeking to document the incidence of violent behavior among the high-school students, professor of health policy David Hemenway, who directs the Harvard Injury Control Research Center at Harvard School of Public Health, agreed to incorporate unrelated (or so he thought) questions about nutrition at a colleague’s request.

Analyzing the survey, he found surprising correlations. Heavy consumers of nondiet soft drinks—students who had drunk five or more cans in the week preceding the survey—were more likely to have behaved violently toward peers (57 percent, versus 39 percent of respondents who drank less soda); to have behaved violently toward another child in their own families (42 percent, versus 27 percent); to have behaved violently in a dating relationship (26 percent, versus 16 percent); and to have carried a gun or a knife during the past year (40 percent, versus 27 percent). The strength of the effect was on par with the correlation (well known among researchers) between these behaviors and alcohol and tobacco use; in some cases, the correlation with soda was stronger.

Even within the scientific community, people found these results very surprising, Hemenway reports: “When you think about the causes of violence, soft drinks are not on the map of variables that you tend to look at.”

His findings recall the 1979 “Twinkie defense” mounted in the trial that followed the murder of gay-rights activist Harvey Milk; the defense attorney persuaded the jury to render a verdict of voluntary manslaughter in part by arguing that his client’s recent switch from a healthy diet to one high in junk food and soft drinks contributed to mental-health issues that led to the killing. The argument may have been prescient in its recognition that what people put into their mouths influences how they feel and, consequently, behave. But whether this is the case with soda is not yet clear.

The researchers have since tested the correlation, with similar results, in three other datasets: one surveying more than 5,000 adolescents in California, one of nearly 3,000 five-year-olds of low socioeconomic status born in major U.S. cities (the question about guns and knives was omitted in this case), and one of more than 16,000 students in public, private, and parochial high schools across the United States. (Hemenway has not investigated the relationship between soft drinks and violence in adults. Although violent crimes committed by adults tend to make headlines, he says, teenagers behave in physically aggressive ways far more often than adults do.)

Next, Hemenway and his colleague, Sara Solnick ’86, M.P.H. ’90, now of the University of Vermont, plan to perform a similar analysis with objective sources such as police records and school-discipline records. Instead of relying on youths’ self-reporting, such a study could examine whether youths who drink more soda are more likely to be suspended for fighting or arrested.

Other studies have linked soda consumption with depression and suicidal behavior, but Hemenway is not aware of anyone else studying the correlation with violent behavior. One further avenue for research is elucidating the underlying mechanism. It could be that a third variable, such as the quality of parenting, influences both soda consumption and aggressive behavior. (The researchers attempted to control for socioeconomic status and the quality of parenting; when they did, the correlation remained strong.) If there is a cause-effect relationship, the researchers speculate that excess caffeine and sugar (along with the subsequent blood-sugar crash) may leave soda drinkers irritable and prone to aggression; or maybe those who drink soda in place of healthier food miss out on nutrients that promote a calmer demeanor.

One public-policy implication is apparent already: colleges may want to think twice about promoting soft drinks as a safe alternative to alcohol. Although soda doesn’t share alcohol’s acute, motor-skill-impairing effect, it may have emotional effects that build over time—meaning it may be safer just to stick with water.