View Photo Gallery: A city grappling with obesity begins stocking corner store shelves with nutritious snack alternatives.

Philadelphia has the highest obesity rate and poorest population of America’s big cities. It also has an ambitious plan — launched out of 632 corner stores — to put healthy food on every table.

The $900,000 investment in better health depends on apples and oranges, chips and candy, $1,200 fridges and green plastic baskets. The results could steer the course of American food policy.

Philadelphia is trying to turn corner stores into greengrocers. For a small shop, it’s a risky business proposition. Vegetables have a limited shelf life, so a store owner must know how much will sell quickly — or watch profits rot away. He also lacks the buying power of large supermarkets and is often unable to meet the minimum orders required by the cheaper wholesalers that grocery stores use.

With shelf space at a premium, shop owners must pick and choose the products they think will sell best. Chips and candy and soda are a sure bet. Eggplant? It’s hard to know.

Access to healthy foods has been a cornerstone of the Obama administration’s food policy, dedicating hundreds of millions of dollars in federal funds to projects like this one. The goal is to eradicate food deserts — low-income areas that lack access to nutritious foods — by 2017.

“More parents will have a fresh food retailer right in their community — a place that sells healthy food, at reasonable prices, so they can feed their families the way they want,” first lady Michelle Obama said when she launched the White House’s $400 million Healthy Food Financing Initiative.

More than just a drain on families, obesity is a huge economic drag: The United States spends $147 billion each year treating the condition.

But even as the White House has scaled up such efforts, a growing body of research has questioned its basic assumption: that people will eat better if given better options. Multiple studies have scoured local, state and national data looking for a causal relationship between weight and access to healthy food. None has found it.

“It’s a theory that makes sense, and it’s intuitive,” says Helen Lee, a policy fellow at the Public Policy Institute of California, whose research focuses on racial disparities in health outcomes. “But my concern would be that we’re investing in a strategy that may not be very promising. If you’re investing government money, you should carefully be evaluating how much you’ve invested and how much you’re getting out of that.”

That’s where Philadelphia comes in. Along with building the country’s largest network of healthy corner stores, the city is conducting the largest study to date of what happens when nutritious options are introduced into neighborhoods that have traditionally gone without. It’s measuring what people bought before, what they’re eating now and whether that improves.

“Availability of these products is definitely changing,” says Giridhar Mallya, director of policy for the Philadelphia Department of Public Health. “Now we’re waiting to see what is actually happening with people’s purchases.”

The Obama administration is watching, too.

“Research hasn’t caught up with all the interventions, because collecting evidence and evaluating it takes time,” Deputy Secretary of Agriculture Kathleen A. Merrigan said. “That’s why we’re excited about efforts like the one that they’re undertaking in Philadelphia.”

Changing habits is a hard sell

The term “food desert” is a relatively new one in public health policy, tracing to a 1995 paper from a government work group in Scotland. Various definitions exist today, and all describe parts of the country, both urban and rural, where there is inadequate access to affordable nutritious foods.

Public health researchers have long known that lower socioeconomic status correlates with worse health, including higher levels of obesity. Numerous studies have also noted connections between access to healthy foods and lower weight. A 2011 article in the Journal Obesity Review found that “greater accessibility to supermarkets or less access to takeaway outlets were associated with a lower prevalence of obesity.”

If governments could improve proximity to healthy foods, the theory went, it could reduce a rapidly rising obesity rate.

“In the U.K., we’d started making policy about this before there was any empirical evidence,” says Neil Wrigley, a professor of geography at Southampton University in England, who works on urban planning research. “Time to time, this happens, where you get policies that outstrip the evidence. Then the evidence needs to catch up.”

Wrigley conducted one of the first studies of a food desert intervention, looking at what happened when a grocery store was brought into an underserved part of Leeds, an industrial city in northern England. Of shoppers surveyed, 45 percent switched to the new store. Their habits, however, barely changed: Consumption of fruits and vegetables increased by one-third of a cup per day — about six grapes or two broccoli florets.

“The results came out quite small, a very modest increase in consumption of nutritious foods,” Wrigley says. “It seemed an almost nonexistent improvement.”

Similar research in the United States shows much the same.

Ohio State University’s Janne Boone-Heineman published a 2011 longitudinal study of food access in Birmingham, Ala., Chicago, Minneapolis and Oakland, Calif. Over 15 years, she traced obesity levels alongside the introduction of healthy food options (grocery stores) and unhealthy venues (fast food restaurants). Her study found no connection between a new grocery store and better health outcomes.

In March, the California institute’s Lee published a paper looking, nationwide, for a connection between proximity to grocery stores, fast food and obesity. RAND Institute’s Roland Sturm published a separate paper this year, one that compared food sold in a neighborhood and children’s diet in California. Neither could find a relationship.

“While some studies find a correlation between food accessibility and BMI and obesity, the causal pathways are not well understood,” the Agriculture Department concluded in a 2009 review of food desert research, noting elsewhere that “interventions aimed at increasing access to healthy foods may not be successful in addressing obesity.”

To date, no study has found a causal relationship between improving access to healthy foods and improving health outcomes. “You have more people starting to poke holes into what’s a simple thesis, that poor people are overweight because they lack access to healthy food,” Lee says. “My concern is that we might be investing in something that might not be very promising, at the cost of not investing in something that works.”

Candy vs. cantaloupe

One pervasive theory of why food access interventions have not worked has to do with what, exactly, corner stores sell. Even when they offer fresh fruit and produce, they also stock chips and candy. The latter are often less costly, more calorie-dense and require little to no preparation — just the sort of thing, in other words, that people will grab on the run.

Others question whether proximity is a good metric for defining access. Adam Drewnowski at the University of Washington recently surveyed Seattle residents on where they bought groceries. He found that most people don’t shop where they live — access is determined as much by price and public transit, for instance, as proximity.

“If you live next to a Mercedes dealership, that doesn’t mean you’ll buy a Mercedes,” he says. “And it’s the same with living next to a grocery store: That doesn’t necessarily mean you’ll start eating salads.”

Others point to a lack of rigorous study of planned interventions: Most have been small in scale, involving a handful of stores. A 2012 review article looked at efforts in 16 cities to improve food access. Only four measured impact on weight; none found any change. And often, after a city’s initial investment, there was no follow-up.

“One big gap in much of the work has been a lack of detailed evaluation,” says Joel Gittlesohn, a public health researcher at Johns Hopkins University who has published extensively on food access issues. “Programs are often implemented by local departments with very little evaluation of what’s going on.”

Philadelphia’s study, distinct in scope and scale, may deliver a breakthrough.

The city has, in many ways, been the epicenter of American efforts to improve food access. Of the country’s 10 largest cities, its population is the lowest-income, and it has higher obesity rates than New York City and Baltimore. It’s home to The Food Trust, a nonprofit that has risen to national prominence as an advocate for increasing food access for low-income Americans.

Working with Food Trust, in the late 2000s Philadelphia began piloting healthy corner stores. In 2010, it ramped up efforts significantly when it received $25.4 million in stimulus funds meant to combat obesity and tobacco use. That initial grant was bolstered with $1.5 million more in funding from the Affordable Care Act’s Prevention and Public Health Fund, a $15 billion commitment to projects that promote preventive health.

“It was a historic investment in public health,” says the public health department’s Mallya, who oversees the initiative for Philadelphia. “I don’t know if we’ll ever get that level of investment again. So for us, it’s very much been transformative.”

‘Just one piece of the puzzle’

The city has recruited 632 corner stores — of 2,500 overall — to its Get Healthy Philly initiative. Of those, 122 have gotten more intensive support, been supplied with new fridges to store produce and connected with wholesalers from whom they can buy at lower prices. It is also working with schools to improve nutrition and helping neighborhoods launch farmers markets, a multifaceted approach officials hope will improve public health.

“Access to healthy food is just one piece of the puzzle, and we are committed to doing the work to help improve public health,” Merrigan, of the USDA, said.

Anecdotal reports from shop owners suggest that sales of fresh produce have indeed increased alongside the surge in supply.

“Almost every day, people grab lettuce or something,” says Catalina Morrell-Hunter, who has owned her corner store in North Philadelphia for 15 years. Apples and oranges go fastest, and cilantro has proved popular in the largely Hispanic neighborhood. “I don’t say I sell like an entire market does. But when people are short a carrot, they can come to the convenience store.”

But whether that will have a health impact remains to be seen. Temple University’s Center for Obesity Research is working with the city to study how shopping habits do, or don’t, change when healthy options are introduced. Last year, before stores added nutritious options, researchers stopped 7,000 shoppers on their way out of the store to look at their purchases. With the new foods now available, researchers are doing another 7,000 stops.

“I don’t think we know much about how well this works,” says Gary Foster, director of the center. “It’s a field in its infancy . . . nobody has really done at such a big scale.”

Foster expects the research on urban corner stores to publish in about a year and, when it does, it will be “the largest study by a long shot.”

When kids come into Guillermo and Denise Rodriguez’s store, they often buy the bananas at her urging. “It’s not a problem getting them to buy the fruit,” Denise says. “It’s a problem trying to get them to keep buying the fruit, and stay off the junk food. You have people who buy what they want to buy.”

Sometimes they’ll hand out fruit for free, to encourage kids to try it. The goal, says Denise, is to familiarize kids with healthier foods.

That’s a hard way to run a business.

“It’s all good but, you know, when the moment this money stops flowing, things go back to normal,” Drewnowski says. “There needs to be a longer-term business model.”

The Rodriguezes say they are committed to the business; aside from the equipment they received from the city, the sales of fruit and vegetables make the new venture sustainable, if not quite profitable right now.

“There’s not too much of a profit,” Denise says. “We’re not really worried too much about the profit right now, not until we see a profit later on. Right now, we’re just selling the fruit and making things healthier for other people. That’s good enough for us.”

Kliff wrote this article with the assistance of the Dennis A. Hunt Fund for Health Journalism, which is administered by the California Endowment Health Journalism Fellowships, a program of the University of Southern California’s Annenberg School for Communication and Journalism.