Picture Description and Functional Communication Rating Correlates in Variants of Primary Progressive Aphasia

Generalization of speech-language pathology treatment is an important goal in clinical practice and research. Functional communication rating scales are often used to investigate potential treatment effects in daily life. Connected speech samples more closely reflect meaningful changes in communication as a result of treatment than test/re-test outcomes. Little attention has been directed to this relationship in aphasia due to neurodegenerative etiologies.

We investigated whether correlations between ratings on the Georgetown University Center for Aphasia Research and Rehabilitation Functional Communication Scale (CARR-FCS) and total content units (CUs) from spoken descriptions of the Cookie Theft Picture from the Boston Diagnostic Aphasia Examination (CTP-BDAE), versus correlations with scores on the Boston Naming Test (BNT), differed among variants of primary progressive aphasia (PPA) at baseline testing and 6-12 month follow-up (ClinicalTrials.gov Identifier: NCT02675270). At baseline and follow-up, for semantic variant PPA (svPPA), we hypothesized that performance on picture description will correlate more strongly with the CARR-FCS, compared to confrontation naming, as those with svPPA tend to have especially poor performance on confrontation naming [compared to logopenic variant PPA (lvPPA) and nonfluent agrammatic PPA (nfaPPA)], a task that may not be reflective of functional communication ability because communication via other means (e.g., circumlocutions) is not captured. We did not expect to find this effect in lvPPA or nfaPPA because, although anomia is present in all PPA variants, it is less severely compromised in these variants. At follow-up, for nfaPPA, we hypothesized that performance on picture description will correlate more strongly with the CARR-FCS, compared to confrontation naming, because those with nfaPPA can have a marked decline in confrontation naming due to worsening apraxia of speech.

At two time points, we calculated correlation coefficients between care partners' ratings on the CARR-FCS and total CUs from spoken descriptions of the CPT-BDAE and BNT scores of individuals with PPA. We compared the size of the correlations.

Correlations were significantly stronger between total CUs and ratings on the CARR-FCS, compared to BNT scores and CARR-FCS ratings, for svPPA at both time points, but were not significantly different for lvPPA and nfaPPA.

These findings suggest that, while confrontation naming performance may be a proxy for functional communication in lvPPA and nfaPPA, a measure of connected speech is more representative of functional communication in svPPA because confrontation naming is typically more impaired in svPPA than in either of the other two variants.