Esther Sternberg makes an interesting point about how urban planning can affect not only social interactions, but also health outcomes. Her data point concerns the proximity of neighbors to one another, and the positive effects of increased social interaction. I've always been struck by how much healthier my lifestyle is when I am living in New York, where I walk far more than anywhere else -- Clive Thompson once wrote a great piece on why New Yorkers live longer -- or my parents' house in suburban California, where perfect weather, easy access to the Santa Ana River Trail and the beachfront bike paths of Orange County encourage me to cycle (especially if my dad permits me to borrow his ultra lightweight road bike).
As a journalist in California's Inland Empire, I did a lot of reporting on planned communities being built by Lewis Homes, one of the more progressive builders in the region. (That isn't saying much.) One community I studied, Terra Vista, included walking and bike paths running through most of its tracts, easily connecting folks to the nearby shopping centers, where they did grocery shopping, patronized a Starbucks, etc. Despite the nod toward a walkable community, however, almost no one used the paths. This is no doubt partly because people who care very much about walking places tend not to move into suburban housing tracts in commuter towns, but I think it's also due to a mistaken notion of what makes walking places pleasant.
Were I given the choice between walking to work on a tranquil suburban trail that winds through a few parks... or else across town on Canal Street, up through the Lower East Side, and into my office... I'd definitely choose the chaotic, noisy, smelly New York City commute. It's not boring, due to the street-scape and the varied people who inhabit it. Jane Jacobs described it a an urban ballet -- which is a lot more diverting everyday than watching grass grow.