>>sambe+Ad1
From the review, they found 2986 unique papers on the subject, they then assessed 91 papers from the 2986. From these 91, only 1 had a rigorous double blind testing but on only 46 people. The exposures for the light source was 2500 lux, IR light and no special treatment. The results were not significant.
So, lot-researched but maybe not well-enough researched.
>>Loic+cg1
In other words, the statistical power of such studies is so low, that it's like social science at best: a galore of false positive over true positive effects.