Biomarkers vs Outcomes
A personal essay, not an evidence assessment. Our graded assessments of individual ingredients — written to a published standard, from full texts — are in the evidence section. Nothing here is a statement about what any product does.
A biomarker can move dramatically while a person feels exactly the same.
This is one of the most important ideas I have learned while reading biomedical research.
We measure:
hormones, inflammatory markers, neurotransmitter metabolites, blood lipids, BDNF, oxidative-stress markers, brain signals, gene expression.
These measurements can be scientifically valuable.
But the moment a biomarker moves, there is a temptation to translate that change into a human benefit.
That translation is not automatic.
The biomarker may sit somewhere along a pathway
Suppose an intervention increases a molecule associated with neuronal plasticity.
Interesting.
But does memory improve?
Learning?
Attention?
Daily function?
Perhaps.
Perhaps not.
The biomarker can tell us that biology changed.
It cannot always tell us whether the change mattered to the person.
Surrogates are attractive because they are convenient
Real outcomes can be difficult.
Long-term cognition takes time to study.
Disease events may take years.
Performance can be noisy.
Patient-important outcomes can require huge samples.
Biomarkers can often be measured quickly and objectively.
That makes them useful.
It also creates temptation.
A shorter, cheaper study can produce a biologically impressive result without demonstrating the consumer outcome everyone actually cares about.
Some biomarkers are exceptionally useful
I do not want to imply that all biomarkers are weak.
In some medical contexts, particular markers are strongly validated and clinically actionable.
The problem is not the existence of a biomarker.
The problem is assuming that every measurable intermediate variable is a validated substitute for a meaningful outcome.
Association does not complete the chain
Even if biomarker Y is associated with better cognition, increasing Y artificially does not automatically mean cognition will improve.
This is the same logical problem that appears with mechanisms.
The direction of causality may differ.
Y may be a consequence rather than a cause.
It may be one component of a much larger system.
Or intervention-induced changes may not behave like naturally occurring differences.
What matters for MindHeaven?
If a study shows:
biomarker changed I want us to write:
biomarker changed.
Not:
brain performance improved unless brain performance was actually measured and improved.
That sounds almost absurdly simple.
Yet this small discipline would remove a remarkable amount of exaggeration from supplement marketing.
I want the reader to see the evidence chain
Ideally:
Intervention
↓
Biological change
↓
Functional outcome
↓
Meaningful human benefit A study may demonstrate only the second step.
That is useful information.
But it should remain on the second step.
One of the easiest ways to look scientific is to measure something sophisticated.
One of the hardest disciplines in science is admitting that the sophisticated measurement may not yet tell us what we want to know.