Also, I forgot to mention that bacopa has been shown to possess acetylcholinesterase inhibiting properties. This AChE-I effect is probably more well-known from things like huperzine, galantamine, and sage, but may also play a role in bacopa’s ability to improve cognition.
In healthy elderly volunteers, 300mg/day (10% yield, 5% total saponin content determined using HPLC) decreased AChE by slightly more than 10% at 4 weeks, approximately 35% at 8 weeks, and approximately 30% at 12 weeks, while 600mg/day decreased AChE by roughly 5-10% at weeks 4, 8, and 12. Additionally, AChE levels remained decreased even 4 weeks after cessation of bacopa after the 12 weeks of daily use.
In this study, both doses benefited working memory (reached significance at 8 and 12 weeks, but not 4 weeks), with 300mg seeming to be superior, and both studies still noted benefits 4 weeks after cessation of use.
Edit: brief and imperfect comparision of the AChE-I activity of bacopa to huperzine and galantamine. I say “imperfect” because (1) I’m comparing the effects of bacopa from one study to the results of huperzine and galantamine in another study, which is itself a bit flawed/imperfect, and (2) the huperzine/galantamine study was in young subjects, while the bacopa study was in elderly subjects, and we can’t inherently assume the magnitude of effects would be the same for these different populations. That said:
we saw that the 300mg/day bacopa decreased AChE activity to ~64% of pre-treatment levels (after 8 weeks of daily use), while 100 and 200 mcg huperzine-A decreased AChE to 83% and 72% of pre-treatment levels, and 4 and 8 mg galantamine decreased it to 85% and 75% of pre-treatment levels, but these were with a single dose (link to the info on AChE-Is below).