Religion and Well-being

Tom Rees has highlighted an interesting bit of social science research on his blog, Epiphenom:

The well-being of religious adherents follows a clear U-shape, with the least happy being those people with moderate faith.

Also included on the plot was the well-being of atheists, agnostics, and “no religion” (nihilists?) as measured in the study. It seems to me that the biggest take-home from this figure is that the individuals most certain in their beliefs were the happiest — either those with strong religious belief or atheism. Maybe it’s too simplistic to assume that atheists tend to be more certain or definitive in their worldview than others who follow no religion?