They don't appear to have run a control with caffeine alone. Caffeine is a well-established athletic performance-enhancing drug in its own right. Isn't this a serious flaw with the swim-test portion of this study?
Figure 1 shows that 5,10,20,50 mg/kg caffeine has an effect on the forced-swim test, but not as strong as imipramine. Table 1 shows that the same doses of caffeine don’t make the mice more active in their cage, which might help address your concern.
"Table 1 shows that the same doses of caffeine don’t make the mice more active in their cage"
As a former rat owner, and a former locked-up felon, if a rat (or human) is in a cage, it tends to know and act like it's in one. Caffeine isn't going to make you suddenly more energetic in your cage unless you've got some other underlying issues, so that part of the study is instantly suspect.
Mice will run for miles (literally!) if there’s a treadmill or wheel in their cage. I could certainly imagine there’s some sort of multiplicative effect, where a substances causes (say) a 1.2x increase in activity, but that’s also not really how caffeine hits me: too much and I’m nonspecifically jittery.
Regardless, that could certainly explain the caffeine-only part, but it’s harder to square with the caffeine*anti-depressant effect.
I had thousands of other data points in the GP cell right there with me. Plenty of scientific basis for my conclusion, as I had a good 50 weeks to observe.