Brain Stimulation Fails Its Most Rigorous Aging Test: Major Pre-Registered Trial Finds No Motor-Learning Boost from tDCS in Older Adults
For nearly two decades, transcranial direct current stimulation—tDCS to its friends—has hovered at the intersection of genuine neuroscience and consumer hype: a battery-grade trickle of current through scalp electrodes, allegedly capable of making the brain learn faster. Now one of the most rigorous tests of that promise in aging humans has come back empty. In a pre-registered, double-blind, randomized, sham-controlled trial published on 29 August 2026 in the journal GeroScience, researchers at KU Leuven in Belgium, led by first author Silke Kerstens and senior author Jean-Jacques Orban de Xivry, report that anodal tDCS, delivered in both its conventional and its more focal high-definition form, produced no measurable improvement in motor sequence learning among 52 older adults. Every group in the study, stimulated or sham, learned a finger-sequencing task at the same rate, and every group retained exactly what it had learned 24 hours later. The current, in other words, added nothing that practice alone could not deliver.
tDCS works on a deceptively simple principle. A weak direct current, in this study one milliampere, is passed between electrodes on the scalp. The current is far too weak to trigger neurons directly; instead, it subtly shifts the resting membrane potential of cortical neurons, nudging them marginally toward greater excitability under the anode, so that when the brain’s own activity arrives, the response is amplified. The approach entered the modern canon in 2000, when researchers showed that weak direct current reliably and reversibly changes the excitability of the human primary motor cortex, the patch of cortex that controls voluntary movement. The classical montage, however, uses large pad electrodes, and much of the current leaks through skin, skull, and cerebrospinal fluid before reaching the cortex, blurring its focus. High-definition tDCS was engineered as the corrective: a small central electrode sits over the individually mapped motor hotspot, surrounded by four return electrodes arranged in a ring that confine the current and sharpen the electric field over the hand-control region.
The stakes are high specifically for older people. Motor sequence learning—the implicit process by which the motor system extracts a hidden repeating pattern from a stream of finger movements—declines measurably with age, and it is central to retraining after stroke and in Parkinson’s disease. If a non-invasive stimulator could reliably amplify it, rehabilitation would gain a cheap, portable multiplier of therapy. Earlier findings fed that hope: stimulation paired with multi-day practice enhanced visuomotor skill acquisition through effects on consolidation; post-training stimulation improved motor memory consolidation in older people; and a randomized controlled trial found that tDCS enhanced motor learning in Parkinson’s disease. Yet the literature has been strikingly uneven, with effect sizes swinging from robust to null across laboratories. A 2023 systematic review and meta-analysis led by members of the same Leuven group concluded that tDCS effects on learning in older adults, with and without Parkinson’s disease, were inconsistent, partly because stimulation parameters and electrode montages vary widely across experimental protocols.
The new study was built to close that gap with unusual methodological discipline. The team pre-registered the trial, enrolled 52 older adults, and ran a double-blind, randomized, sham-controlled, parallel-group experiment with a crossover component embedded inside it. In the parallel arm—the primary contrast—participants were assigned to active or sham stimulation, which guarantees that any between-group difference in learning cannot be contaminated by carryover from a previous stimulation session. Nested within that framework, the crossover comparison allowed each participant’s response to conventional versus high-definition tDCS to be evaluated within-subject, isolating the effect of montage focality from person-to-person variability. The reporting followed CONSORT guidelines for randomized trials, participants and experimenters alike were blind to condition, and sham sessions used identical hardware without sustained current, holding expectation effects constant across groups. The pre-specified question was whether stimulation changed how much was learned and how much survived 24 hours of consolidation.
Learning was quantified with the serial reaction time task, the laboratory standard for implicit motor learning. Participants watched a visual cue appear at one of several positions and pressed the corresponding key with the appropriate finger as quickly as possible. Unknown to them, the cue positions followed a fixed repeating sequence interleaved with random blocks; the gap in reaction time between repeated and random sequences indexes how much of the hidden pattern the motor system has absorbed. Stimulation ran concurrently with practice: one milliampere of anodal current was applied over the individually identified hotspot of the primary motor cortex, either through the conventional large-electrode montage or through the 4×1 high-definition ring, with matched sham hardware in each montage arm. Because active versus sham was tested in parallel groups while montage was tested within participants, the design could answer two separate questions cleanly: does the stimulation beat placebo, and does focality change the answer?
The answer to both was no. All four groups showed robust motor sequence learning: reaction times shortened dramatically as practice progressed and again across sessions, with the main effect of time reaching p < 0.0001—overwhelming statistical evidence that the task engaged the aging motor system and that it adapted. The stimulation itself, however, was inert. The main effect of stimulation group was p = 0.68, and the main effect of montage was p = 0.66, both statistically indistinguishable from chance. Conventional tDCS did not outperform its sham; high-definition tDCS did not outperform its sham; and neither focal nor diffuse current edged out the other. Twenty-four hours later, the pattern repeated with a twist that matters clinically: every group maintained what it had learned overnight, but none enhanced it. The retention comparisons—p = 0.64 for high-definition tDCS, p = 0.76 for conventional tDCS, p = 0.69 for high-definition sham, and p = 0.57 for conventional sham—showed consolidation proceeding identically with or without electricity. Practice built the skill; current changed neither how much was built nor how much was kept.
Why did a manipulation that has succeeded in some laboratories fail here? The evidence points to several converging mechanisms. The aging head is a poor interface between electrode and cortex: as brain tissue shrinks and cerebrospinal fluid spaces enlarge with age, computational models and tissue-thickness measurements show that current is progressively shunted away from the cortex, weakening the field that actually reaches the motor strip. Even in young adults, the electric field induced by a given montage varies substantially between individuals depending on anatomy, so a one-milliampere dose that is effective beneath one scalp may be underdosed beneath another. Plasticity in the aged motor cortex also depends on its neurochemical baseline; studies linking sensorimotor GABA levels to motor-learning-related plasticity in older adults imply that raising cortical excitability cannot rescue learning when the excitatory-inhibitory balance is itself shifted by aging. Finally, dose and timing matter: the field’s most durable positive effects came from stimulation paired with practice across multiple days, or delivered after training to boost consolidation, rather than from a single concurrent session.
The verdict lands in a field already accumulating negative answers. A meta-analysis of cerebellar tDCS found no consistent effect on visuomotor adaptation; a recent experiment showed that complex sequential learning is not facilitated by stimulation over the dorsolateral prefrontal cortex or the motor cortex; and one study of older adults even found that anodal stimulation of the prefrontal cortex slowed sequence learning. Earlier head-to-head comparisons of the two montages had reported differential effects on implicit motor sequence learning, making the within-person montage contrast here especially informative. What separates the Leuven trial is its construction: pre-registration, double blinding, sham control, a parallel-group test of the active-versus-sham question, an embedded crossover to isolate montage focality, and a pre-specified overnight retention endpoint. A null from an underpowered or unblinded study is easy to discount; a null from a design engineered to detect the expected effect is not. The result therefore functions as a calibration point, constraining the plausible effect of single-session anodal tDCS on implicit motor sequence learning in healthy older adults to something very close to zero.
The implications radiate to two audiences. For clinicians and rehabilitation scientists, the message is that in unimpaired older adults the engine of motor learning remains practice itself: the sham groups learned and consolidated as fully as the stimulated groups, meaning therapeutic time is better invested in structured, intensive training than in adjunctive stimulation—at least in this configuration. For the consumer neurotechnology market, where headsets promising cognitive and motor enhancement are increasingly aimed at older buyers, the trial adds a sobering data point: the electrophysiological rationale for anodal tDCS is real, but converting it into faster motor learning in an aged brain is not straightforward, and one session at one milliampere does not do it. The authors are equally careful not to overclaim in reverse. The study tested a single session, one current intensity, one task, and one retention window; it closes those doors specifically, not the entire house.
That restraint shapes the forward agenda. To capture possible long-term or cumulative effects, the team recommends that future studies administer repeated tDCS sessions throughout the motor learning process, pairing stimulation with practice across days as in the multi-day protocols that produced the field’s most durable gains, rather than relying on a single exposure. They also point toward an optimization frontier: individualized modeling of each participant’s head anatomy to predict the delivered electric field, dosing adjusted for age-related tissue changes, and monitoring of neurochemical markers such as cortical GABA that may gate whether excitability shifts can become learning. None of this diminishes the force of the current result. In one of the cleanest tests yet performed, the aging brain learned a new motor sequence, slept on it, kept it—and did so exactly as well with its motor cortex left unstimulated. For a field built on the promise of plugging extra learning into older heads, that is a striking verdict, and it required no current at all to deliver.
Subject of Research: Effects of conventional and high-definition anodal transcranial direct current stimulation on motor sequence learning and 24-hour retention in older adults
Subject of Research: Medicine
Article Title: No effect of anodal high-definition transcranial direct current stimulation during motor sequence learning in older people
Article References: Kerstens, S., Broeder, S., Gilat, M., Nackaerts, E., Vandendoorent, B., Nieuwboer, A., & Orban de Xivry, J.-J. (2026). No effect of anodal high-definition transcranial direct current stimulation during motor sequence learning in older people. GeroScience. https://doi.org/10.1007/s11357-026-02457-8
Image Credits: AI Generated
DOI: 10.1007/s11357-026-02457-8
Keywords: Aging, Neuroplasticity, Consolidation, Motor sequence learning, Transcranial direct current stimulation (tDCS), High-definition transcranial direct current stimulation (HD tDCS), Primary motor cortex, Serial reaction time task, Older adults, Motor memory retention
Cite Scienmag News
APA
MLA
Chicago
Cassandra Pierce. (August 30, 2026). High-definition brain stimulation fails to boost motor learning in older adults. Scienmag. https://scienmag.com/high-definition-brain-stimulation-fails-to-boost-motor-learning-in-older-adults/
Cassandra Pierce. “High-definition brain stimulation fails to boost motor learning in older adults.” Scienmag, 30 August 2026, https://scienmag.com/high-definition-brain-stimulation-fails-to-boost-motor-learning-in-older-adults/. Accessed 30 August 2026.
Cassandra Pierce. “High-definition brain stimulation fails to boost motor learning in older adults.” Scienmag. August 30, 2026. https://scienmag.com/high-definition-brain-stimulation-fails-to-boost-motor-learning-in-older-adults/
Copy citation
Download RIS
Tags: aging and neural plasticityaging and neuroplasticityaging brain and motor sequence learningaging brain plasticitybrain stimulationbrain stimulation for motor skill acquisitionefficacy of tDCS in cognitive enhancementevidence-based neurosciencehigh-definition brain stimulationlimitations of brain stimulation therapieslimitations of tDCS in older populationsmotor learning in older adultsneurostimulation and motor skill acquisitionrandomized controlled trial in aging populationrandomized controlled trial in neurosciencesham-controlled neurostimulation studiestDCStranscranial direct current stimulation
