Cognitive enhancement claims, examined

The market for cognitive enhancement is large and the evidence base is weaker than the volume of publication suggests. The pattern of weakness is consistent enough to be worth describing.

Transcranial direct current stimulation

A weak current passed between electrodes on the scalp, intended to shift the excitability of underlying cortex.

The device is trivially simple, which is part of why it produced an enormous research literature.

What the evidence shows. Early studies reported effects across a wide range of cognitive tasks. Larger, better-controlled studies and meta-analyses have found much smaller effects, and several have found none.

Replication has been poor. Effects appear sensitive to electrode montage, current, duration, task, and individual anatomy — which is another way of saying the effect is not robust.

Modelling of current flow indicates that a substantial fraction of applied current is shunted through the scalp, and that reaching a specific cortical region with meaningful current density is harder than the simple picture suggests.

The honest summary. Small effects in some paradigms, high heterogeneity, publication bias likely, and no established basis for consumer devices claiming reliable enhancement.

Clinical applications — particularly in depression, where it has been studied more rigorously — have a somewhat stronger evidence base and are a separate question from enhancement in healthy people.

Transcranial magnetic stimulation

A different proposition, and worth distinguishing clearly.

Magnetic pulses induce currents in cortex strong enough to trigger neuronal firing, unlike direct current stimulation.

Repetitive TMS has regulatory approval for treatment-resistant depression and for some other indications, with a substantial trial base.

It requires clinical equipment, trained operators and repeated sessions. It is not a consumer technology and there is no home version that does the same thing.

Neurofeedback

Real-time display of a brain signal, allowing the user to attempt to modulate it.

The methodological problem. Any training produces improvement. The question is whether improvement is caused by feedback of the specific signal or by the practice, attention and expectation involved.

The test is a sham-controlled trial, where a control group receives feedback from someone else's signal or from a scrambled one.

Studies with proper sham control have generally found much smaller differences between real and sham than uncontrolled studies suggest — and in several cases no difference.

Where the evidence is better. Some clinical applications, particularly in attention deficit disorders, have a more substantial literature, and even there the sham-controlled results are contested.

Consumer devices generally use one or two electrodes, which cannot localise a signal, and report a proprietary composite metric that is not validated against anything.

Nootropics

A category spanning approved medications, dietary supplements and unregulated compounds.

Caffeine has robust evidence for alertness and vigilance, with well-known tolerance and sleep effects. It is the most effective widely available cognitive drug and the least discussed as one.

Prescription stimulants used off-label by people without an attention disorder show, in controlled studies, modest effects on some measures and inconsistent effects on complex cognition — with some evidence that they increase confidence in performance more than performance itself.

They also carry cardiovascular, dependence and sleep risks, and prescription diversion is a genuine concern.

Racetams and most marketed compounds have limited human evidence, and supplement regulation in most jurisdictions does not require efficacy demonstration.

Independent testing has repeatedly found products whose contents do not match their labels.

The pattern

The same features recur across all of these.

Early small studies with positive results. Substantial publication bias. Larger and better-controlled studies finding smaller or absent effects. Meta-analyses with high heterogeneity. Strong placebo responses. And a consumer market that cites the early literature and not the later.

What actually improves cognition

Boring, well-evidenced and free.

Sleep. The single largest effect on cognitive performance available to most people, with a very substantial literature. Sleep deprivation produces impairment comparable to significant alcohol intoxication.

Aerobic exercise, with reasonable evidence for effects on executive function and, over longer periods, on cognitive ageing.

Treating hearing loss, vision problems, depression, anxiety and sleep apnoea, each of which impairs cognition substantially and each of which is treatable.

Practice at the specific task, which reliably improves performance at that task and — this is the well-replicated disappointment of brain training research — transfers poorly to anything else.

None of that is marketable as a device, which is a reasonable explanation for the imbalance in attention.