Anxiety after a stroke: neuromodulation remains experimental

An overview of reviews, not a new trial

Anxiety after a stroke can compromise rehabilitation, participation in activities and quality of life. Researchers examined the available reviews on non-invasive neuromodulation for these symptoms, based on a search of four bibliographic databases up to June 1, 2026.

This overview brings together six pairwise meta-analyses and one Bayesian network meta-analysis. It assesses their methodological quality and the certainty of the evidence, without providing new experimental data from patients.

Improvements in some analyses only

Some estimates suggest an improvement in anxiety symptoms, while others show no clear difference. The available abstract does not provide the effect sizes, sample sizes for each outcome or follow-up durations needed to assess the practical importance of these changes.

Crucially, the reviews could include studies primarily focused on pain, sleep disorders or depression after a stroke. A reduction in an anxiety score in these populations does not demonstrate effectiveness in patients recruited for clinically defined post-stroke anxiety.

Seven reviews do not mean seven independent confirmations

All included reviews have low or very low methodological quality. Trials overlap across the reviews: their inclusion in several reviews therefore does not amount to multiple independent confirmations.

Small sample sizes for some outcomes, heterogeneous interventions and inconsistencies between the source publications also limit interpretation. The authors call for particular scrutiny of estimates that pool different neurological populations.

No protocol can be recommended on this basis

The authors’ conclusion remains cautious: non-invasive neuromodulation remains experimental for anxiety after a stroke. Current evidence supports neither the routine use of any particular protocol nor the replacement of usual multidisciplinary care.

Future trials will need to define anxiety from the outset, standardise stimulation and sham procedures, and then measure the persistence of benefits and safety. A favourable signal on a score does not yet meet these requirements.

What we can take away

Non-invasive neuromodulation shows promising signals, but its effectiveness for clinically defined post-stroke anxiety remains uncertain. This overview does not provide a basis for choosing a protocol for routine practice.