The paper is published in the open access journal PLoS One, and linked from this piece: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.00307

The paper is published in the open access journal PLoS One, and linked from this piece: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0030719 http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjourna... Those speculating that the causation might lie in the other direction should at least bear in mind that this is a prospective cohort study (following a particular group of people over a long period of time). It found that – after controlling for income – people who worked longer hours were more likely to experience a subsequent major depressive episode. Of course you might like to speculate that there is a hidden variable: perhaps some psychological factor that both predisposes people to work long hours and also increases their risk of depression. This research doesn’t rule out that possibility, as far as I can tell. To do that you would need to randomly assign people to the long-hours or short-hours condition and follow them for a period of time. I imagine it would be rather difficult to carry out a study of that sort.