Objective:

To examine whether frontal high-definition tDCS (HD-tDCS) paired with modified constraint-induced language therapy (mCILT) differentially benefits participants with the nonfluent/agrammatic variant (naPPA) versus logopenic variant (lvPPA).

Background:

Primary Progressive Aphasia (PPA) comprises neurodegenerative syndromes with variant-specific language network damage. naPPA affects frontal regions, while lvPPA involves posterior temporoparietal areas. While behavioral therapies are standard care, transcranial direct current stimulation (tDCS) has emerged as a promising potential intervention. Two frameworks exist for optimizing tDCS stimulation location: the restitution hypothesis proposes stimulating damaged regions to rescue function, while the compensation hypothesis suggests stimulating preserved regions to enhance compensatory mechanisms.

Design/Methods:

We analyzed two participants from an ongoing NIH-funded R01 crossover study. Participants received 10 sessions of active or sham HD-tDCS over the left frontotemporal region paired with mCILT across two separate 12-week treatment arms. We examined noun and verb accuracy during a 96-item picture naming task using within-subject paired t-tests comparing baseline to 6-week performance.

Results:

The naPPA participant demonstrated significant noun improvement during active stimulation (change score = +0.219, p = 0.017) and trending verb improvement (CS = +0.219, p = 0.07), with minimal sham response for nouns (CS = +0.031, p = 0.745) and verbs (CS = +0.063, p = 0.423). The lvPPA participant showed significant improvement in nouns during sham stimulation (CS = +0.313, p = 0.005) but not verbs (CS = +0.063, p = 0.325) and no significant improvement during active stimulation for nouns (CS = -0.063, p = 0.325) and verbs (CS = +0.063, p = 0.325).

Conclusions:

Findings support the restitution hypothesis: naPPA showed selective improvement with active stimulation while lvPPA showed no significant benefit. However, individual-level factors like low baselines and testing fatigue could introduce confounds, while near-significant statistics demonstrate insufficient power. Group-level analyses of the full R01 cohort are essential to address limitations and establish evidence-based treatment protocols.