I have always retained everything, the useless included. Quantity was never the problem. The problem was the time it took to find something I knew, and the path, not always logical, to get there.
With bacopa, 1,000 mg per day, at first, nothing. Nothing at all, for a long time. And then, over time, something changed: I find the drawer faster. I still retain everything, I forget almost nothing, but it is as if my memory were compressing: simplified, without anything being lost.
This is an observation, not a measurement, and I will say what it is worth. But it has one feature I have found with no other molecule in my stack: the delay. And on that precise point, the literature is unusually clear.
after a 300 mg dose: no test moves
Nathan 2001
the length of every trial that finds an effect
Pase 2012
free recall tests improved across the six trials in the review
Pase 2012
Nothing at two hours, everything at twelve weeks
The same Australian team asked the question twice, in the same year, with the same extract at 300 mg. First as a single dose: 38 healthy adults, tests before and two hours after. No change, on any test. The authors conclude that bacopa, at that dose, has no acute effect on cognition.
Then as daily intake, with tests at baseline, five weeks and twelve weeks. This time, visual information processing speed, learning rate and memory consolidation improve compared with placebo, and anxiety drops. With a precision that matters: the maximum effects are only visible after twelve weeks.
Two hours, nothing. Five weeks, a start. Twelve weeks, the effect. That is the curve I lived without knowing it, slower still.
What six twelve-week trials measure
The 2012 systematic review only kept randomised placebo-controlled trials in adults without dementia: six, all run over twelve weeks, with three different extracts dosed at 300 to 450 mg per day. All measured memory. Bacopa improves performance on 9 of the 17 free recall tests, and almost nothing in other domains, partly because they are little studied.
Free recall is retrieving, without a cue, a list of words learned a little earlier. It is not the capacity to store, it is the capacity to bring back out. When I say I find the drawer faster, that is exactly the test I am describing in my own words.
The 2014 meta-analysis, nine trials and 518 participants, all on at least twelve weeks of standardised extracts, points the same way with numbers: the trail-making test B, which measures attention speed and flexibility, is shortened by 17.9 milliseconds, and choice reaction time by 10.6 milliseconds. Small numbers, cleanly measured. The authors conclude that bacopa has the potential to improve cognition, mostly attention speed, and that a large trial against a reference drug will be needed to settle efficacy.
After 55, the effect is clearest
Two trials in older people give the most legible picture. Calabrese, in 2008: 54 participants aged 65 and over, 300 mg per day or placebo for twelve weeks, after six weeks of placebo for everyone. The primary outcome was delayed recall of a word list, and it improves on bacopa. The Stroop test, which measures the ability to ignore distracting information, too. Anxiety and depression scores drop in the bacopa group and rise in the placebo group.
Morgan and Stevens, in 2010: 98 Australians over 55, 300 mg per day of another extract, twelve weeks. Verbal learning, acquisition and delayed recall improve clearly. Complex figure and trail-making tests improve too, without the difference from placebo being significant.
And in 60 older volunteers, at 300 or 600 mg for twelve weeks, working memory improves, the brain's electrical responses to a stimulus arrive earlier, and the activity of the enzyme that breaks down acetylcholine drops in the blood. That is the most often proposed mechanism: a partial brake on that enzyme, like some Alzheimer's drugs, far weaker. It explains a slow effect that settles in.
The price: the gut
Bacopa is not free. In Morgan and Stevens's trial, the bacopa group reports more stools, abdominal cramps and nausea than the placebo group. In Calabrese's, adverse events are as frequent on bacopa as on placebo, nine against ten, and they are mostly stomach upset. Nothing serious in these trials, but enough for some people to stop before twelve weeks, so before they could judge.
What I do not know is the share of everything else. Months of practising my own memory, a job that changed, sleep that changed too. An effect that takes three months to come is an effect impossible to isolate alone. That is the honest limit of every slow molecule, and bacopa is the slowest in my stack.
How I use it
| Question | What the trials say | What I do |
|---|---|---|
| What dose | 300 mg per day of standardised extract, sometimes up to 450; 600 does no better than 300 in the only trial that compares | 1,000 mg per day of standardised extract, more than three times the trials |
| How long before judging | Nothing at two hours, a start at five weeks, the maximum at twelve | I judge nothing before three months, and expect nothing on the day |
| When | One daily dose, no timing in the trials | With a meal, because of the gut |
| What to watch | Nausea, cramps, more frequent stools | If it persists after two weeks, I stop |
My dose deserves a sentence of its own. A thousand milligrams a day is more than three times what the trials use. The only trial comparing 300 and 600 mg sees no clear difference between the two, and no trial has measured 1,000 mg of extract. That is not an argument for this dose, it is a fact I report, and it is the first thing I would cut if my gut protested.
What this article does not say
- That bacopa makes you smarter. The trials measure free recall, attention speed and working memory. Little elsewhere, and the 2012 review says so: the research is in its infancy.
- That my "compressed" memory is a measured effect. It is my way of describing faster retrieval. Recall tests come close, none measures that feeling.
- That the effect lasts after stopping. No included trial measures it. The Thai trial tested four weeks after stopping, without that point being detailed in its abstract.
- That all extracts are equal. Three different extracts across six trials, standardised differently. Raw plant powder is not an extract.
- That three months of intake are without consequence. The trials stop at twelve weeks, and the gut is the first to protest.
- That my dose is the right one. 1,000 mg per day is more than three times the trials, with no evidence that a higher dose does better, and no measurement at that dose.
« Bacopa does not make memory bigger. It makes the path shorter. And it takes three months for the path to appear. »
What Helix does
Bacopa is the simplest case and the one product sheets serve worst: a molecule whose effect is documented, but only after a delay nobody prints on the box. Someone who stops at three weeks because they feel nothing is right about their feeling and wrong about their conclusion.
The engine shows, for each line, the delay before the effect measured in trials, next to the dose. And it only asks your opinion on a slow molecule once the trials say an opinion is possible.
A stack that tells you when to judge each molecule, not just how much to take.
Join the beta