The problem of time turns up in multiple kinds of treatments, and I believe it has simply to do with how long the depressive episodes last. Usually by the time people seek treatment that will be at the peak of the episode. Regardless of the treatment, a couple bunch of month will pass and people find themselves getting better. I suspect most depression studies a deeply flawed because of this effect. At some point the treatment will stop working. But it's not that the treatment stopped working, it's that a new depressive episode has begun and the treatment never really did all that much in the first place.