A Comparison of Item Acquisition and Response Generalization for Semantic Versus Phonological Treatment of Aphasia.

Abstract

Purpose: The purpose of this work is to examine whether therapy-related improvements in trained versus untrained items (acquisition and response generalization, respectively) are differentially affected by phonological versus semantic language treatments and to investigate individual variables associated with treatment response.

Method: Sixty-three participants with chronic poststroke aphasia were included in this retrospective analysis of data from a large, multisite clinical trial with an unblinded cross-over design in which all participants underwent 3 weeks of semantic treatment and 3 weeks of phonological treatment. A linear mixed-effects model was used to examine treatment acquisition and generalization effects for the two treatment types. Multiple regression analyses were also conducted to examine individual participant factors associated with acquisition compared to generalization.

Results: Results showed main effects of outcome type (acquisition vs. response generalization) and treatment type (semantic vs. phonological) on posttreatment changes in naming and an interaction between these factors: For acquisition, phonological treatment resulted in better gains than semantic treatment, whereas for response generalization, semantic treatment resulted in slightly better gains than phonological treatment. There were no significant associates of generalization gains. However, acquisition after phonological treatment was associated with less severe aphasia and higher nonverbal semantic processing abilities at baseline, whereas acquisition after semantic treatment was associated with apraxia of speech.

Conclusions: On average, phonological treatment may be more effective for acquiring trained items, whereas semantic treatment may be more effective for response generalization to untrained items. Moreover, acquisition gains are associated with individual baseline variables. These findings could have clinical implications for treatment planning.

Supplemental material: https://doi.org/10.23641/asha.28410212.