Here`s a simple study, though one of the first of its type, showing a beneficial result from adding vitamin D supplementation to an antidepressant:
42 depressed subjects received fluoxetine 20 mg daily plus either 1500 IU of vitamin D or placebo, for 8 weeks.
The vitamin D group had significantly better mood improvement!
I think I would want to see this study replicated, but in the meantime I think it is reasonable to suggest vitamin D supplementation for most cases of depression, particularly in a region where there is a long, dark winter season.
Of note, in this study, the authors found that vitamin D levels were inversely correlated with depression severity ratings before the treatment trial began. So it may be true that correcting vitamin D deficiency is the therapeutic mechanism, as opposed to supplementing vitamin D beyond the usual healthy range.
The doses I suggest are 3000 IU per day.
On a contrary note, there was a large, important meta-analysis published in Lancet Diabetes & Endocrinology by Autier et al. on December 6, 2013, which closely reviewed existing evidence about Vitamin D status and supplementation.
The authors conclude that low vitamin D status is a result of illness, not a cause of illness, in most clinical situations. And they conclude that vitamin D supplementation had no effect on disease occurrence, except for a small effect in elderly women.
The full text is not yet available to me, so I am not sure what the authors looked at with respect to mood disorders and vitamin D.
But the findings are another example of the need to contain our excitement when we see strong correlations between variables. In this case, low vitamin D has been correlated with higher incidences of various diseases, leading to the speculation that supplementation could be an effective treatment for the disease. But this paper shows that much of this correlation is due to non-causal association.
The area of vitamin D supplementation remains important to research further, because this type of supplementation is quite safe, the prevalence of mild deficiency in the population is significant, and there are particular conditions which may indeed improve with higher-dose supplementation.
Furthermore, there have not been enough very simple augmentation studies, such as the first study mentioned above, which could show whether a simple supplement such as vitamin D could play a role in treating depression.
So, for now, I stand by the recommendation to try Vitamin D supplementation as a part of mood disorder treatment, unless there is some contraindication to this (such as hypercalcemia), but with the acknowledgment that the evidence for this is much weaker than I would like.