Which suggests that non-drug interventions are generally better than drug interventions, at least when people actually do them. (Obviously people are much more likely to take a pill than to change their diet, which is part of what makes the research so interesting.)

Not surprisingly, many of the diseases with the lowest adherence are the ones where people would benefit the most. Type 2 diabetes is literally a disease of non-adherence, with 98% of patients not following their doctor's recommendations.

I've learned a ton from reading through the major adherence research findings over the last couple months, and I think there are a ton of ideas that can be applied to entrepreneurship and web startups especially. DiMatteo actually has a brand new book out on this so you don't even really need to read through all the original journal articles anymore to get the main ideas:

http://www.amazon.com/Health-Behavior-Change-Treatment-Adher...