Anemona Hartocollis of the New York Times writes that a Stanford University study has found that customers at New York City Starbucks restaurants bought items with 6 percent fewer calories after the city began requiring chain fast food restaurants to post calorie counts in April 2008.

The study, released Wednesday, also found that people who habitually bought high-calorie foods at Starbucks changed their habits more than the average Starbucks customer, lowering their calories per transaction by 26 percent. Yet, the researchers said, there was no impact on Starbucks’s average profits.

Most of the reduction in calories was related to food purchases, rather than drinks.

The study is among the first to find a direct correlation between people’s food choices and the posting of calorie counts on menu boards, and it could provide encouragement to Congressional efforts to require calorie-posting as a matter of national policy.

Mayor Michael R. Bloomberg, who has been a strong proponent of government-sponsored health campaigns, from banning smoking in bars to banning trans fats in restaurant baked and fried goods and, of course, posting calories, promoted the study Wednesday by sending out a news release hailing it, with a copy of the study attached.

“This study helps confirm what we’ve believed all along,” Mr. Bloomberg said in a written response to the study. “Consumers can make healthier choices when supplied with the right information, and businesses can profit while offering their customers healthier alternatives.”

Starbucks cooperated with the study, giving the researchers information on every transaction in its New York City stores from Jan. 1, 2008, to Feb. 28, 2009. The study also used Starbucks locations in Boston and Philadelphia, where there were no calorie postings, as control groups.

The research supported similar preliminary findings in a study by the New York City health department. But it contradicted a study published in October in the online version of the journal Health Affairs.

That study, conducted by professors at New York University and Yale, tracked customers at McDonald’s, Wendy’s, Burger King and Kentucky Fried Chicken in poor neighborhoods of New York City with high rates of obesity. Using fast food restaurants in Newark, where there is no calorie posting law, as controls, the study found that calorie counts made no difference even for customers who had noticed that the calories were posted.

Unlike the N.Y.U. and Yale study, the Stanford study was not tailored around poor neighborhoods, and it focused on a luxury brand, Starbucks, rather than chains like McDonald’s that put a premium on low prices and whose core business is food. It covered all 222 Starbucks locations in New York City, as well as 94 in Boston and Philadelphia.

At New York Starbucks, the average calories per transaction dropped 6 percent, to 232 calories from 247 calories. The average beverage calories per transaction hardly changed, while average food calories fell by 14 percent, or an average of 14 calories per transaction, the study said. With the cooperation of Starbucks, the researchers were able to look at more than 100 million transactions in the three cities.

Three quarters, or 10 calories, of the drop in calories per transaction came from consumers buying fewer items, and one quarter of the drop, or 4 calories, came from substituting lower calorie items.

Using statistical modeling, the researchers found that calorie reductions in Queens, Brooklyn, Staten Island and the Bronx were comparable to the reduction in Manhattan.

“This confirms our hope that people would use this information in making more healthful decisions and lower calorie choices, and that they’re doing it consistently over time,” Lynn Silver, New York City’s assistant health commissioner for chronic disease control, said.

One of the study’s authors, Phillip Leslie, an associate professor of economics and strategic management at Stanford, said Wednesday that while a 6 percent decrease by itself was unlikely to have an impact on obesity — studies suggest that 25 percent of food is eaten away from home — he believed that calorie posting laws had created an incentive for restaurants to introduce lower-calorie items, which would have a cumulative effect.

“I think the chain restaurants are all over this,” he said. “We’re going to see lots of low calorie items in the future. I think this law actually drives that.”

Professor Leslie said that a breakdown of the study data by ZIP code showed that the effect of calorie posting was greater among more affluent, educated consumers. And he said, people go to Starbucks primarily for coffee, rather than food, which may be the reason that they were willing to be more flexible on their food choices, with a resulting drop in calories.

“Any notion that this is having an identical effect across all chain restaurants is somewhat absurd,” he said.

The study found no significant change to Starbucks’s revenue after the calorie law went into effect, despite complaints from some companies that the law would hurt their bottom line. The study did say that Starbucks stores within 100 meters of a Dunkin’ Donuts had an estimated 3 percent increase in revenue. Professor Leslie interpreted that finding as suggesting that people had been frightened away when they discovered the number of calories in the staple Dunkin’ Donuts products, doughnuts.

The study said that the United States is the most obese country in the world, according to the Organization for Economic Co-Operation and Development, and that between 1995 and 2008, the proportion of the population that was obese rose to 26.6 percent from 15.9 percent.

The study notes that Starbucks-brewed coffee has only 5 calories, while the highest calorie beverage is the 24-ounce Hazelnut Signature Hot Chocolate with whipped cream, at 860 calories. Foods at Starbucks range from about 100 calories for a small cookie to 500 calories for some muffins, the study said.