If information technology and deep-pocketed startups can't make people healthier, what can? One of the most effective proven solutions is shockingly low-tech. In the late 1990s, as diabetes was noticeably on the rise, the National Institutes of Health wanted to see whether there was a better way to intervene, so it set up a test. It took three groups -- hundreds of middle-aged people in each -- from all over the country: One group received metformin, a prediabetes drug; another group, the control, received a placebo; and the third underwent a "lifestyle-modification program," with the goal of at least 7% weight loss.
The study was meant to last five years, but the results were so striking that they were published a year early, in 2002. The lifestyle intervention not only beat out the placebo, but was twice as effective as the drug too: Nearly 60% of the third prediabetes group never got the disease.