The six domains
- Processing speed. How fast you take in and respond to information. Declines earliest and most reliably.
- Attention. Holding focus, switching between tasks, filtering noise. Highly sensitive to sleep, stress and hearing.
- Memory. Which subdivides further: working memory (holding something in mind now), episodic memory (events), semantic memory (facts), and procedural memory (how to do things).
- Executive function. Planning, sequencing, inhibiting an impulse, switching strategy.
- Language. Word-finding, comprehension, fluency.
- Visuospatial ability. Judging distance, navigating, mental rotation.
They are related but separable. A person can be slow and sharp, or fast and forgetful, and the combinations are informative.
What age does to each
Declines: processing speed first and most consistently. Working memory. Episodic memory for new information. Divided attention. Some aspects of executive function.
Holds or improves: vocabulary, which continues rising into the seventies. General knowledge. Procedural memory — skills, once learned, are remarkably durable. Pattern recognition within a familiar domain. Emotional regulation, on average.
The net effect in real life is usually far less dramatic than the speed curve suggests, because experience compensates for a great deal.
How to read a trial claim
Studies measure one domain at a time, with a specific task. This is where supplement claims either hold up or evaporate.
The ginkgo trial that found something measured recall of agreed appointments — episodic memory, in one specific form. It found no effect on recognition memory, which is a different task in the same domain.
The theanine and caffeine studies measured attention-switching. Not memory, not executive function. Attention.
The lutein trial measured attention too, over twelve months.
So “supports cognitive function” on a box could rest on any of those, or on none. The question that makes it readable is: which domain, measured how, for how long, in how many people.
What “supports healthy cognitive function” is allowed to mean
It is structure-function language — the regulated phrasing a dietary supplement may use. It cannot claim to treat, cure or prevent a disease, because that would make the product a drug.
The phrase is therefore doing less work than it appears to. It is not meaningless, but it is not a clinical claim either, and knowing the difference is most of what protects you in this category.
Applying it to a label
Take the six ingredients here and ask the question of each: four have randomised human trials, and those trials measured episodic recall in one case and attention in the other two. Two have no cognitive trial at all.
That is a more informative description than “supports cognitive function”, and it is available for any product whose label prints its amounts.