When University of Pennsylvania researcher Christina Roberto set out to study Philadelphia’s soda tax, she wasn’t convinced she would find much health impact.

“We know that it is really hard to shift a person’s weight and keep that weight off, and it’s a really tall order to ask a policy like this to produce a health effect,” she said.

Indeed, Roberto’s study, published this month in The Lancet , found that average BMI, a measure of a person’s body fat based on their height and weight, is still increasing in Philadelphia.

But compared to control groups outside the city, BMIs in Philadelphia are rising at a slower rate. The study found “limited evidence” that BMIs and the prevalence of obesity decreased in the city three years after the implementation of the 2017 tax, its authors wrote.

Frequently drinking sugary drinks is associated with weight gain, obesity and other conditions like diabetes . Philadelphia’s soda tax, the first in a major American city, aims to curb sales by adding 1.5 cents per ounce to distributors’ costs for sodas, energy drinks, mixers, and other sugary beverages. The increased costs are mostly passed on to consumers who are charged higher prices. The tax, which is sent to the city’s general fund, partially funds pre-kindergarten education and improvements to parks and community centers.

BMI alone is often not a good predictor of an individual person’s health, and people’s body shapes and compositions can vary depending on their race, gender, and age, according to the American Medical Association . The AMA recommends that clinicians working with individual patients use BMI alongside other measures to look at weight and body fat and determine a course of care.

But to research health across large populations, the BMI measurement is generally regarded as a more accurate data point, according to the AMA and other scientists. BMI is “probably the best we can do,” Iliya Gutin, then a program officer at the National Academies of Sciences, Engineering and Medicine, told the New York Times last year , when the AMA recommended doctors avoid using BMI alone to assess a patient’s health.

While the soda tax’s impact on average BMI in Philadelphia is small, it’s significant for researchers. “We’re not seeing giant reductions. It’s the fact that we’re seeing reductions at all” that’s surprising, Roberto said.

Increasing research on the soda tax’s impact

Philadelphia’s soda tax was controversial when instituted in 2017, and the American Beverage Association spent millions lobbying against it. Detractors said that the tax would hurt local businesses and customers and would not achieve its public health goals.

But the uproar has died down over the years , with the beverage industry focusing instead on preventing the implementation of similar taxes elsewhere, and scientific research is now beginning to examine the tax’s impact.

Another study published earlier this year in JAMA Health Forum found that Philadelphia had seen some of the biggest decreases in sugary beverage sales among cities that enacted similar taxes. Sales here dropped by nearly half in the years after the tax was implemented .

Roberto’s study looked at electronic health records from Penn Medicine, comparing Penn patients from Philadelphia and the city’s collar counties in Pennsylvania and New Jersey, where beverage taxes were not implemented.

More research is needed on how beverage taxes affect health, Roberto said, but she’s been encouraged by research showing similar benefits in other cities and countries that have implemented taxes.

“When you conduct one study, it’s hard to be confident about having the definitive answer,” she said. “But I’m starting to get more confident that the taxes may actually be improving health.”

‘A clear no-brainer’

Even small improvements in BMI — or just a decrease in soda sales and consumption — points to the efficacy of the tax, Roberto said.

“To me, this policy is a clear no-brainer. We know from lots of data that sugary drinks aren’t good for our health, and we tax other things, like tobacco, that we know are not good for our health,” she said. “Even if we didn’t see a BMI effect, getting people to drink fewer sugary drinks is a good thing.”

And, because the tax funds pre-K programs — and good education is another good predictor of overall health — it might also be impacting public health in other ways, Roberto noted.

Roberto said she’s hopeful that the findings in Philadelphia will influence other cities considering implementing a beverage tax.

“Philly can serve as an example with the more data we accumulate. You might get more cities and states to do this, and that might improve the health of Americans across the whole country,” she said.