Category‐selective semantic deficit for living things in semantic variant primary progressive aphasia

Background Studies of patients with impaired semantic memory have identified cases with a category‐selective semantic deficit (CSSD) for living things—a disproportionate semantic memory loss for animate versus inanimate items. Data are mixed on whether patients with semantic variant primary progressive aphasia (svPPA) exhibit a CSSD, in part because measurement can be confounded by psycholinguistic variables. Furthermore, there exists little consensus regarding the neural substrates of CSSD.Here we examine (1) the effect of animacy on single word comprehension performance in svPPA after accounting for familiarity, frequency, and age of acquisition, and (2) the relationship between cortical atrophy and disproportionate impairment in identifying animate vs. inanimate items. Methods 21 individuals with svPPA completed the Cambridge Semantic Battery‐Word Picture Matching (CSB‐WPM) task, which includes 32 animate and 32 inanimate items. Using a multiple regression model, we assessed the effects on CSB‐WBM accuracy of familiarity, frequency, age of acquisition, and animacy (i.e., animate, inanimate). Cohen’s d was used to measure effect size. We then estimated CSSD per individual (% inanimate correct ‐ % animate correct). For each ROI, we ran a linear regression assessing the relationship between cortical thickness and CSSD, covarying for overall semantic impairment. Results The main effect of animacy remained significant (t = 4.6, p < .001) even after including all psycholinguistic variables in the model [F(4,56) = 13.8, p < .001] (Cohen’s d=1.1). After controlling for overall semantic impairment, greater cortical atrophy was associated with greater CSSD in the right posterior superior temporal gyrus (STG) (β = ‐.41, t = ‐2.38, p = .031). Relationships between CSSD and the right entorhinal/perirhinal cortex (β = ‐.39, t = ‐1.88, p = .081) and right posterior fusiform gyrus (β = ‐.39, t = ‐2.08, p = .056) also trended in this direction. Conclusions Results provide support for an animacy CSSD in svPPA, with individuals showing greater deficits for animate compared with inanimate items even after controlling for familiarity, frequency, and age of acquisition. The right temporal lobe, especially the posterior STG, entorhinal/perirhinal cortex, and the posterior fusiform gyrus, may also play a critical role in the representation of animate items.