Do Lower fNIRS Brain Signals Predict Better Focus?

Two supplements lowered the same brain-blood-flow signal. Only one improved performance.

A botanical blend and a green tea extract produced the same acute change on functional near-infrared spectroscopy (fNIRS): lower oxygenated hemoglobin (oxyHb) in the prefrontal cortex during cognitive tasks. If lower oxyHb reliably meant a sharper brain, that convergence would be persuasive. It does not. One small randomized trial found better working-memory performance alongside the lower signal; another found no measurable cognitive benefit despite the same directional change.

The useful conclusion is narrower. fNIRS can detect that these formulations alter prefrontal hemodynamics during a task. It cannot yet tell us whether that alteration improves cognition, reflects vascular effects, or does both. Blood flow is biology. It is not a report card.

The positive trial: a three-ingredient blend improved N-back performance

In a double-blind, placebo-controlled, between-group randomized trial, Best et al. (2022) gave 40 healthy adults a single dose of a blend containing Bacopa monnieri, American ginseng (Panax quinquefolius), and whole coffee fruit extract, or placebo. Researchers assessed cognition and prefrontal hemodynamics 45 minutes later.

Compared with placebo, the active group had higher N-back accuracy (p<.01), faster response time (p<.05), and lower prefrontal oxyHb (p<.001). The reported p values indicate statistically significant between-group differences, but the available report does not provide effect sizes or confidence intervals here. That makes the practical size of the cognitive improvement difficult to judge.

The authors proposed neural efficiency: participants may have recruited fewer prefrontal resources while performing as well or better. That is plausible, but it remains an interpretation rather than a demonstrated mechanism. The trial did not include an independent test of processing efficiency, such as response-time variability, error patterns, or performance across escalating task loads.

It also cannot identify the active ingredient. Three botanicals arrived together, and whole coffee fruit creates a caffeine-related attribution problem. The paper does not report ingredient doses with enough detail to compare the blend cleanly with prior Bacopa or ginseng trials. A 45-minute window also fits poorly with the expected absorption timeline for Bacopa bacosides, which require gut hydrolysis before absorption. Any acute effect could instead involve faster-acting components, including ginsenosides or caffeine-related effects.

The green tea trial produced the same signal without better cognition

Cha et al. (2024) tested a single 900 mg dose of a GCG-rich green tea extract against placebo in 35 healthy adults, also using a double-blind, between-group randomized design. Each dose contained about 50.8 mg of gallocatechin gallate (GCG), 215.6 mg total catechins, 34.7 mg caffeine, and 10.1 mg L-theanine.

The active group showed lower prefrontal oxyHb during mental arithmetic, Corsi block-tapping, and verbal-fluency tasks. Yet the study reported no significant behavioral cognitive improvements on any task. The authors noted a similar earlier pattern with EGCG: reduced cerebral blood flow with stable cognitive performance.

This result does not prove that GCG impairs cognition. A small trial can miss modest effects. But it does directly weaken the claim that lower task-related prefrontal oxyHb is a validated proxy for better performance. The same biomarker direction accompanied a behavioral gain in one study and a behavioral null result in another.

GCG has an in-vitro rationale for central nervous system access: the paper cites work finding roughly 2.4-times higher blood-brain-barrier permeability than EGCG. Cell-model permeability, however, does not establish meaningful concentrations in the human brain 45 minutes after a dose. Human central pharmacokinetic data for GCG are still lacking.

Why a lower fNIRS signal has several possible meanings

fNIRS measures changes in oxygenated hemoglobin near the cortical surface. During cognitive tasks, the signal reflects neural activity indirectly through neurovascular coupling. It can also respond to vascular and autonomic state. A lower average prefrontal oxyHb value therefore cannot, by itself, separate reduced neural demand from altered cerebral blood flow.

That distinction matters for both botanical trials because each formulation included a plausible vasoactive ingredient. The green tea extract contained 34.7 mg caffeine, while the other blend contained whole coffee fruit. Wei et al. (2025) found that changes in cardiac vagal activity moderated the relationship between cognitive performance and prefrontal hemodynamics. Caffeine and other compounds could affect the measured signal through autonomic or vascular pathways without proving a change in neural efficiency.

Yao et al. (2024) offers a related warning. Older adults with hypertension and healthy older adults had comparable gait performance during dual-task walking, but their cortical fNIRS activity differed with vascular status. Hemodynamic differences can therefore exist without matching differences in task output.

The analysis choices also matter. Both botanical studies focused on the direction of oxyHb change. Ji et al. (2024) linked cognitive compensation in high performers to low-frequency oscillation patterns in prefrontal fNIRS signals. Mean oxyHb amplitude does not capture that spectral information, spatial patterning, or lateralization.

More activation can accompany better performance

Neural efficiency is real in some settings, but it is not a universal rule that less prefrontal activation means better cognition. In a study of right dorsolateral prefrontal transcranial direct current stimulation, Figeys et al. (2023) found load-dependent increases in oxyhemoglobin that tracked improved cognitive performance. The direction was opposite to the botanical findings.

Other intervention studies reach the same broader conclusion: the relationship depends on the task and population. Kjellenberg et al. (2024) found that short, frequent physical-activity breaks improved working memory in adolescents while preserving cerebral blood flow. Zhong et al. (2024) reported correlated working-memory and fNIRS changes after 12 weeks of volleyball training in an 80-child randomized trial. These studies support fNIRS as a sensitive research tool, not as a single-direction cognitive-benefit meter.

What the evidence supports—and what it does not

Both primary trials were small, between-group experiments: 40 participants for the botanical blend and 35 for GCG-rich green tea extract. Randomization helps, but at those sample sizes, baseline cognitive differences can still influence between-group results. Small studies also provide limited precision for modest nootropic effects, making both positive and null findings fragile until independent teams replicate them.

The GCG study adds further reasons for caution. It was retrospectively registered because of proprietary issues, was conducted and funded by the manufacturer Amorepacific Inc., and recruited Amorepacific employees. Those features do not invalidate its hemodynamic result, but they raise the risk of bias and limit confidence in any broad interpretation.

Reduced prefrontal oxyHb after a supplement is an interesting pharmacodynamic observation, not evidence that the supplement improves cognition.

The Best trial is worth following because it measured a statistically significant working-memory signal alongside fNIRS change. It is still one small, multi-ingredient, acute study without a caffeine-only control. The GCG trial supplies the necessary counterexample: a near-identical hemodynamic direction with no significant behavioral gain.

Do not buy an acute botanical nootropic because a small fNIRS trial showed lower prefrontal oxyHb. For Bacopa, the more replicated cognitive literature concerns supplementation over weeks to months, not a single dose assessed 45 minutes later. For GCG-rich green tea extract, one small acute human trial found a hemodynamic effect but no significant cognitive benefit. The next informative study would use an independent, adequately powered crossover design with caffeine-controlled arms and prespecified cognitive and fNIRS analyses.