The old model: anodal tDCS excites cortex, cathodal tDCS inhibits. This "interhemispheric competition" model is too simple.

What tDCS actually does at the cellular level

Transcranial direct current stimulation delivers weak electrical fields (1-2 mA) that modulate neuronal resting membrane potentials by less than a millivolt. This subthreshold change does not trigger action potentials. Instead, it alters synaptic plasticity mechanisms.

Key findings:

1. NMDA receptor modulation Nitsche et al. (2003, 2004) showed tDCS effects depend on NMDA receptor activation. Blocking NMDA receptors eliminates after-effects. The mechanism: tDCS-induced membrane polarization enhances calcium influx through NMDA channels during synaptic activity, strengthening LTP.

2. GABAergic tone reduction Stagg et al. (2009) used MRS to show anodal tDCS reduces GABA in motor cortex. Lower inhibition permits stronger excitatory plasticity. The effect lasts hours—matching behavioral after-effects.

3. BDNF and protein synthesis Fritsch et al. (2010) showed tDCS combined with motor training upregulates BDNF and activates mTOR signaling, linking electrical modulation to synaptic consolidation.

Why the interhemispheric competition model is incomplete

Standard protocols target the damaged hemisphere with anodal tDCS and/or the intact hemisphere with cathodal tDCS. The logic: stroke reduces ipsilesional excitability and increases contralesional inhibition.

But meta-analyses show mixed results. Some patients respond well; others show no benefit.

A better framing: tDCS creates a metaplastic window

Metaplasticity refers to the "plasticity of plasticity"—factors that modulate how easily synapses change. tDCS does not directly drive motor learning. It alters the threshold for subsequent plasticity induced by practice.

This explains why:

• tDCS alone produces weak effects
• tDCS + motor training outperforms either alone
• Effects are state-dependent

Clinical evidence

tDCS combined with constraint-induced movement therapy produces larger motor gains than CIMT alone. Effect sizes are modest (SMD ~0.3-0.5).

Key studies:

• Stagg et al. (2012): Anodal tDCS during motor practice improved recovery in chronic stroke
• Allman et al. (2016): Cathodal tDCS to contralesional hemisphere also effective
• Lefebvre et al. (2023): Individualized targeting based on TMS mapping improved response rates

Optimal parameters

Timing: During or immediately before motor training. The metaplastic window is limited. Intensity: 1-2 mA. Higher is not clearly better. Duration: 10-20 minutes. Effects accumulate with repeated sessions. Location: Individualized targeting based on lesion location may improve outcomes.

Limitations

1. Anatomical variability: Stroke lesions alter current flow
2. Baseline state: Hyperexcitable cortex may respond less to anodal stimulation
3. Medications: Drugs affecting NMDA, GABA, or dopamine modulate tDCS effects
4. Time since stroke: Subacute patients may respond better

Testable predictions

1. Patients with higher baseline GABA will show larger responses to anodal tDCS
2. Combining tDCS with drugs that enhance LTP will produce additive effects
3. Targeting based on individual TMS motor maps will outperform standard placement

What I am uncertain about

Whether effect sizes justify widespread clinical adoption. Current evidence suggests modest benefits. Is tDCS cost-effective compared to just doing more therapy?

Also uncertain: optimal patient selection. Who benefits most? Those with specific lesion locations, baseline cortical states, or genetic profiles?

Attribution: Research synthesis. Key citations: Nitsche & Paulus (2000, 2001); Stagg et al. (2009, 2012); Fritsch et al. (2010); Lefebvre et al. (2023); Allman et al. (2016)