Executive functions — inhibitory control, cognitive flexibility, monitoring, and attentional allocation — are the cognitive abilities most vulnerable to aging because they depend on prefrontal systems that undergo substantial structural and functional change in later life. But aging is not simply loss. Neuroimaging frameworks such as HAROLD (Hemispheric Asymmetry Reduction in Older adults) and PASA (Posterior-Anterior Shift in Aging) describe how older adults often preserve performance by recruiting additional prefrontal resources during demanding tasks. Whether that extra recruitment is adaptive compensation or neural inefficiency is a central question for any intervention that claims to enhance cognition, because stimulation-induced changes could just as easily reinforce inefficient network dynamics as foster adaptive plasticity.

Cognitive training alone can engage residual plasticity in healthy older adults, but its benefits rarely transfer broadly across cognitive domains. Combining training with non-invasive brain stimulation (NIBS) is thought to strengthen activity-dependent plasticity by modulating the responsivity of networks that are concurrently engaged by the training tasks. tDCS applies a weak electrical current that does not directly fire neurons but instead modifies membrane polarization, shifting the probability that task-engaged neural populations will respond to ongoing input. The dorsolateral prefrontal cortex is considered a prime target for this kind of neuroempowerment because of its role in executive attention, monitoring, relational integration, and flexible control. Yet findings in healthy older adults remain strikingly heterogeneous — some studies report gains after repeated prefrontal stimulation, particularly when paired with cognitive training, while others find little or no advantage over sham conditions.

To cut through that heterogeneity, the Milan team designed a three-arm controlled pre-post study. Thirty-three community-dwelling older adults, screened for cognitive health through the Mini Mental State Examination, Geriatric Depression Scale, and Instrumental Activities of Daily Living scale, were randomly assigned to one of three conditions: active tDCS plus cognitive training, sham tDCS plus the same cognitive training, or a wait-list control. Both training groups completed 24 sessions across eight weeks, three sessions per week, using the same computerized program covering attention, memory, and verbal and non-verbal executive functions. Task difficulty was individualized according to ongoing performance — increasing when participants achieved fewer than 20 percent errors — while the overall training dose and structure remained constant across participants.

In the active condition, a continuous 1.5-milliampere anodal current was delivered for 15 minutes using 35-square-centimeter electrodes, with the anode positioned over F4 to target the right prefrontal cortex and the cathode over the contralateral supraorbital site Fp1 in a cephalic bipolar montage. Cognitive exercises began five minutes after stimulation onset, overlapped with the remaining stimulation period, and continued thereafter — an arrangement intended to promote concurrent engagement and modulation of right-prefrontal networks. Sham stimulation reproduced the device setup and initial cutaneous sensations without sustained neuromodulation. Assessors administering cognitive tests and EEG analysts were blinded to participant condition, and all participants completed every planned session.

Before and after the intervention, participants underwent a comprehensive neuropsychological battery — including the Stroop test, Raven’s Progressive Matrices, phonemic, semantic, and associative verbal-fluency tests, the Trail Making Test, and measures of verbal and visuospatial memory — as well as EEG recording during a 120-trial computerized Stroop-like task. The task presented congruent color-word stimuli on 25 percent of trials and incongruent stimuli on 75 percent, each for 300 milliseconds. EEG was acquired at 1,000 Hz using a 15-channel system, filtered offline between 0.1 and 30 Hz, and corrected for ocular artifacts using a semi-automated Independent Component Analysis procedure. Event-related potentials were averaged separately for congruent and incongruent trials, and the N2 component — a negative deflection peaking around 200 milliseconds, linked to conflict detection, mismatch processing, and response selection — was extracted at frontal sites F3, Fz, and F4.

The behavioral results were selective rather than sweeping. Raven’s Progressive Matrices, which measures non-verbal reasoning through logical rule maintenance, hypothesis generation, and relational integration, showed a significant Group × Time interaction: only the active-tDCS group improved significantly, with scores rising from a mean of 32.8 to 36.2 (Cohen’s d = 0.67), while sham and wait-list groups remained stable. Associative fluency — which requires flexibly generating semantically related responses without relying on a single overlearned retrieval rule — showed an even stronger pattern, improving substantially in the active-tDCS group (d = 0.95) with no comparable change in either control condition. By contrast, other measures such as Stroop interference time, phonemic and category fluency, Trail Making Test performance, and immediate and delayed verbal learning improved over time across the entire sample, without group-specific effects — a pattern consistent with practice effects, procedural learning, or broader training benefits rather than changes specifically attributable to active stimulation.

The electrophysiological evidence was even more striking for its spatial precision. Active tDCS combined with cognitive training produced a significant increase in N2 amplitude at F4 during congruent trials — the waveform became substantially more negative, from a mean of −5.00 to −9.12 microvolts (d = 1.19) — whereas neither the sham-tDCS nor wait-list group showed any comparable change. No significant modulation appeared at F3 or Fz, and N2 latency was unchanged everywhere. The congruent trials were the less frequent stimuli in the task, and the authors interpret the localized N2 enhancement as possibly reflecting increased sensitivity of the stimulated right-prefrontal system to infrequent, salient, or contextually unexpected events rather than modulation of conflict processing alone. The Group × Time interaction for N2 amplitude during incongruent trials approached but did not reach conventional significance, showing a descriptively similar pattern in the active-tDCS group.

The convergence between selective behavioral gains and localized electrophysiological change is what makes this pilot study theoretically relevant. The significant N2 effect appeared precisely at the electrode corresponding to the anodal stimulation site, a topographical correspondence compatible with preferential effects on right-frontal processing — though the authors caution that scalp-recorded activity cannot be uniquely attributed to the underlying right prefrontal cortex because ERP signals reflect volume-conducted activity. Within aging frameworks, greater N2 negativity following active stimulation may reflect enhanced recruitment of right-prefrontal resources during early stimulus evaluation, potentially compensatory upregulation — but increased neural activity cannot in itself distinguish adaptive recruitment from neural inefficiency, and the association between the N2 change and behavioral improvement was not directly tested.

The authors acknowledge several limitations. The sample of 33 participants limits statistical power and precision, and the 15-channel EEG montage and peak-based ERP analysis provided limited spatial resolution. Because stimulus congruency and frequency were not independently manipulated, the contribution of congruency per se cannot be disentangled from that of stimulus rarity. The study also focused on immediate pre-to-post change and therefore does not establish whether the observed effects persist after training ends. Moreover, the design tested the additional contribution of active tDCS when delivered alongside cognitive training rather than isolating a pure stimulation effect, since both intervention groups were exposed to a structured training context involving repeated researcher contact and task engagement. Future factorial designs including cognitive-training-only, tDCS-only, sham, and no-intervention conditions, along with high-density EEG and source reconstruction, would allow the independent and interactive contributions of training and stimulation to be more directly examined — and could help determine whether stimulation-supported neuroempowerment operates by strengthening specific task-relevant prefrontal processes rather than producing broad cognitive enhancement.

Subject of Research: Effects of tDCS combined with cognitive training on executive function and prefrontal ERP markers in older adults

Article Title: Multimodal assessment of NIBS-induced adaptive plasticity in aging prefrontal networks: ERP and cognitive evidence

Article References: Crivelli, D., & Balconi, M. (2026). Multimodal assessment of NIBS-induced adaptive plasticity in aging prefrontal networks: ERP and cognitive evidence. Frontiers in Molecular Neuroscience, 19, Article 1954473. https://doi.org/10.3389/fnmol.2026.1954473

Image Credits: AI Generated

DOI: 10.3389/fnmol.2026.1954473

Keywords: tDCS, cognitive training, executive function, healthy aging, prefrontal cortex, N2 ERP, EEG, neuroplasticity, dorsolateral prefrontal cortex, neuroenhancement, event-related potentials, older adults