How Do Language Domains Recover? A Domain-Specific Response Profile to Constraint-Induced Language Therapy in Chronic Post-Stroke Aphasia

Abstract

Background: Global aphasia scores, such as the Western Aphasia Battery (WAB) Aphasia Quotient (AQ), often obscure heterogeneous domain-specific responses to therapy, limiting tailored speech-language therapy (SLT) planning for chronic post-stroke aphasia (PSA). Understanding differential improvements in spontaneous speech, naming and reading, repetition and writing, and comprehension can optimize session design and goal prioritization, particularly in high-burden regions like Pakistan. Objective: To characterize domain-level changes in language function following Constraint-Induced Language Therapy (CILT) in chronic PSA to inform SLT micro-planning. Methods: A prospective pre-post cohort study enrolled 80 adults with chronic PSA from five Sialkot, Pakistan, centers. Participants received CILT (3 sessions/week, 60–90 minutes) for two months, involving graded verbal tasks constraining non-verbal strategies. WAB domain composites (spontaneous speech, naming and reading, repetition and writing, comprehension) and AQ were assessed pre- and post-therapy. Paired t-tests evaluated mean changes and percent-of-scale gains; AQ changes were modeled with paired correlations. Results: Spontaneous speech improved most (+12.60/20, 63%, p<0.001, d=4.12), followed by naming and reading (+10.04/20, 50.2%, p<0.001, d=3.45), repetition and writing (+9.88/20, 49.4%, p<0.001, d=3.38), and comprehension (+4.50/10, 45%, p<0.001, d=2.35). AQ rose from 24.51±5.92 to 77.38±8.48 (p<0.001, d=5.71), shifting all participants from severe/very severe to moderate/mild categories. Conclusion: CILT preferentially enhances expressive domains, particularly spontaneous speech, while yielding robust comprehension gains, guiding SLT to prioritize verbal output tasks for rapid recovery in chronic PSA.