In 1979, Dan White1 was tried for the assassinations of San Francisco city district Supervisor Harvey Milk and Mayor George Moscone. His lawyers argued that he had diminished capacity and was unable to premeditate his crime. Part of the evidence for his depressed and altered state of mind was that he had recently changed from a health-conscious diet to junk food and Coca-Cola. Although Twinkies, a popular packaged snack cake filled with cream, were mentioned only in passing during the trial, the legal argument became known as the ‘Twinkie Defense’.
The summary is from the peer-reviewed paper “The ‘Twinkie Defense’: the relationship between carbonated non-diet soft drinks and violence perpetration among Boston high school students” by Solnick and Hemenway in the journal Injury Prevention.
Solnick and Hemenway thought White’s defensive attorneys were on to something because, for example, science tells us that there are “several purported pathways linking diet and anti-social behaviour.” Bad guys eat different stuff than good guys. The craven crave sugar more than the cultured.
“One possible explanation for an association between high sugar intake and aggressive behaviour is that that consumption of sugary beverages is a response to abnormally low blood glucose levels, a physiological state that has been linked with irritable and violent behaviour.”
What about micro-nutrients you ask? “Another possibility is that soft drinks replace healthier whole foods in the diet, and that a deficiency of micro-nutrients can lead to violent behaviour.” There’s that sorted.
These facts given, the researchers wondered whether soda pop accounts for violence of the sort often found in teens. To discover whether this were so, the researchers made use of a questionnaire given to Boston high school students.
Besides asking the kids to volunteer from memory how much sugary soda pop they drank in a week, the survey also wondered whether brothers would admit to having punched their sisters, and vice versa. Or whether they’d admit to being “violent” to other kids. Or whether the kids would admit to be “violent” with whom those they were in a “dating relationship.” Or whether they would fess up to carrying a knife or gun “anywhere in the past year.”
Now, if you asked this last question to my high-school mates you would have received a near 100% positive response. To not carry a gun or knife would have been considered unusual. Then again, your author attended a school which celebrated Opening Day of deer season as an official holiday.
The Boston kids were mostly Blacks and Hispanics, with about 8% Asians and 9% Whites. Turns out that, according to the self reports, very few Asian kids said they had five or more pops a week—the cut off selected by the authors as the most indicative. The reader will recall that Asian kids suffer from the stereotype of doing better at school than other groups.
Anyway, our researchers turned up a “dose-response relationship” between sucking soda and violence. The more pop drank, the larger the percentage of admitted violent acts. Very surprisingly, those who drank five or more pops a week also were more likely to “use” tobacco or alcohol.
But raw percentages do not p-values make, and the authors needed p-values else they would not have a paper. So they fed the data into statistical models and found lots of p-values less than the magic number (of 0.05). In particular the p-value for pop consumption was small, thus the conjecture that pop causes rage, and acts of the same, has been proven.
They explored several models, some including booze and cigarettes, others without. Most of the models had a measure called the (pseudo) R-square in the very disreputable neighborhood of 0.07. This crudely means that the model “explains” only 7% of the variance of the data. The remaining 93% remains unexplained. In other words, these models are to statistics what soda pop are to a healthy diet.
Our principal results are that, for Boston high school students, there is a strong, significant association between carbonated non-diet soft drink consumption and the perpetration of violence against siblings, against peers and against dates…We also find a strong association between soft drink consumption and carrying weapons.
The authors worked hard at finding a “causal pathway” to explain these inexplicable results, to say how the caffeine in some pops and the sugar in all of them re-wires the brains of teens and causes them to reach for the knife and to be “violent.”
“It is possible that an underlying organic factor, such as low blood sugar, may lead to both high soft drink consumption and aggressive behaviour.” This suggests that teens with over-active pancreases should be closely monitored lest they biff other students during study hall at unacceptable rates.
Unexplored by our researchers was the hypothesis that kids who eat too much junk food, smoke, and drink were predisposed to act irresponsibly.
1White got off, too.
Thanks to reader Jim Fedako for suggesting this topic.