Skip down to Figure 1 to see the chart. Note that the response to placebo is normally extremely strong in depression studies, so it’s normal to see both groups improve drastically like this.
The only thing that jumps out to me is that the probiotic group started with HAM-D depression inventory scores that were notably higher. When measuring absolute reduction of HAM-D scores, this actually leaves the more depressed patients with more room to improve and can therefore inflate the perceived effectiveness of their response, placebo or otherwise. And indeed, both groups had basically identical scores at the end of the trial. They only really diverged at follow-up where the probiotic group had improved slightly more than placebo.
I’ve seen enough probiotic stories that failed to pan out later that I’d take this one with a huge grain of salt. Trying probiotics might be worthwhile if you can afford them, but the probiotic-depression link never seems to actually play out from online anecdotes of people taking them. Note that we’ve been seeing small studies about this exact topic for a long, long time.