Maybe it was the rise of fast-food chains. Or maybe the advent of video games and TV shows that pulled children out of playgrounds and into living rooms. Perhaps the changes were dietary, behavioral, or even learned from parents. But the world's children are increasingly overweight and obese, a problem with global health and thus economic implications. A National Institute of Health study characterized the global rise in child obesity over the last three decades as an "epidemic." According to the Centers for Disease Control and Prevention, the prevalence of childhood obesity in the United States has tripled over that period.

Japan hasn't solved childhood obesity, but it has achieved some of the developed world's greatest successes in keeping children healthy. But the story of how that happened might not be what you think. When the childhood obesity crisis began about 30 years ago, Japan's national rates were already low, but not the lowest in the developed world, as they appear to be today. We think of Japan as such a healthy place that it's easy to mistake their low childhood obesity rates as somehow particular to the country and its people, as something non-transferrable. But Japan was not immune to the rise in childhood obesity, which doubled over 25 years, due in part to an initial uptick.

As other nations saw their children become less and less healthy, Japan pulled off a feat practically unmatched in the developed world: It responded with policies that helped to resist the global trend of childhood obesity, even countering it in the last few years. And one of the policies that has been a big part of that success, according to Japanese health officials, are school lunches.

Japanese school lunches tend to be planned by a nutritionist, include locally grown and fresh ingredients and tend to be dominated by rice, vegetables, soups and fish. The Japanese program is both impressively successful and surprisingly cheap at about $3 per meal. The Washington Post's Chico Harlan reported on the impressively successful and surprisingly cheap lunches (about $3 per meal) in Sunday's paper. The school program actually goes back, he wrote, to the aftermath of World War II, when food shortages were common, and the country had to dramatically rethink how it fed its children:

For a decade after the war, school lunch food was still coming from international donations. Many older Japanese remember postwar school meals of powdered skim milk, bread and daikon radish. But by the 1970s, the school meal came to look much like the modern-day standard. These days, ethnic food (such as Korean or Italian) is mixed in once or twice per week.
Japanese government officials say no other country has copied Japan’s system of made-from-scratch meals eaten in classrooms, or even tried to.

The program's merits aside, it's not really feasible to statistically isolate the role of school lunches in Japan's public health data. Still, common sense dictates that school lunches would play at least a significant role in student health; American parents would not have to look very hard to notice the prevalence of fried food and soft drinks in school lunchrooms. So it's worth looking at the larger trend of child obesity in Japan, in which school officials seem to see signs of the school lunch program's success.

As in every developed country, childhood obesity has indeed increased in Japan since the 1970s. But that increase has been one of the slowest in the developed world. A 2006 study in the International Journal of Pediatric Obesity (yes, the crisis is so bad that it has its own academic journal) found that, from 1976 to 2000, the prevalence of obesity among Japanese schoolchildren increased by only about 0.1 percent per year. The U.S. increase was four times as rapid. The only countries with slower growth than Japan were Finland and the Netherlands, although the data suggests that children in Japan are still the least likely to be obese.

Childhood obesity rates can be difficult to compare across countries, largely because "obese" and "school age" are not defined the same in all countries. Still, the academic literature typically describes Japan as having some of the lowest childhood obesity rates among all developed countries.

According to benchmarks set by the International Obesity Taskforce, Japan's childhood obesity rates between 1976 and 2000 grew from 1.5 percent to 3.8 percent among boys and 1.2 percent to 2.9 percent in girls. Yes, it more than doubled, which both conveys the global nature of the problem – even Japan has been badly effected – and adds some context to the oft-repeated statistic that America's childhood obesity rate has tripled. Still, other states fared far worse during the rise of childhood obesity. In Britain, for example, rates grew from 1.7 to 5.4 percent in boys and 2.6 to 7.8 percent in girls.

Even Iceland, which had even healthier children than Japan as of 1978, when the global epidemic was first beginning, saw its obesity rate among girls more than quadruple, from 0.5 to 4.2 percent. Iceland has all the ingredients of a country ready to resist childhood obesity: an island nation with a steady supply of fish, a tradition of working outdoors, extensive social programs and a prosperous and educated population. So why didn't they match Japan's success?

Academics have been studying that question for years, and will surely continue to. We can get some hints from this chart, below, produced as part of a comprehensive 2010 study by the Japanese pediatric society and translated into English for the journal Pediatrics International. It shows the childhood obesity rates over time for boys (top chart) and girls (bottom chart), with the different dots indicating different age groups:

Something you'll notice in both charts is that the youngest children, the five-year-olds (represented by the little diamonds), showed a remarkable drop in childhood obesity rates. Since 2000, rates have also declined slightly for 8, 11 and 14-year-olds. Those are all school-age years. The rates have only really increased consistently among 17-year-olds, where the rise has been quite pronounced in both genders.

What does this have to do with school lunches? As Harlan reported in his look at school lunches, Japanese schools are unusual in that they offer students no freedom of choice whatsoever. The students, he reported, "get identical meals, and if they leave food untouched, they are out of luck: Their schools have no vending machines. Barring dietary restrictions, children in most districts can’t bring food to school, either, until they reach high school." There are many possible explanations for Japan's unusual and impressive childhood obesity statistics; the International Journal of Pediatric Obesity study suggests that rates may rise at age 15 in part because students spend more time studying and less time outside. But could school lunches play a role?

Tellingly, the International Journal of Pediatric Obesity study finds that, in similar studies in countries such as the United States and Germany, childhood obesity tends to spike early, around age four or five, when students first enter school (although it's not at all clear that entering school is the cause). In Japan, however, the rates seem to stay very low until age 15, when they enter high school and start making their own dietary choices outside of the rigorous – if delicious-sounding – school lunch program.