Neuroanatomical Correlates of Language Processing in Acquired Aphasia: A Lesion– Symptom Mapping Study of 40 Adults in the Kenyan Rift Valley

Author: Mercyrose J. Jesang¹˒ , Benard Kodak², Kenneth Odhiambo³, Beatrice Owiti⁴

Date: 2026

Abstract: Knowledge of how focal brain lesions disrupt language has been built almost entirely on patients from Europe and North America. Stroke profiles, health-system resources and everyday language use differ markedly in sub-Saharan Africa, where many adults routinely use two or more languages. In Kenya, most urban and peri-urban adults use both Kiswahili and English daily, yet no study has related lesion location to language profile in this population. Objective: To examine the relationship between lesion location and specific language deficits among adults with post-stroke aphasia in Nakuru County, Kenya, and to explore whether bilingual language use modulates the pattern of impairment. Methods: A cross-sectional lesion–symptom study was conducted among 40 adults attending speech therapy at Nakuru County Referral Hospital between January and December 2024. All participants had a single, first-ever stroke confirmed on computed tomography (CT) or magnetic resonance imaging (MRI) and aphasia persisting for more than three months. Language was assessed with a shortened, culturally adapted version of the Boston Diagnostic Aphasia Examination covering fluency, auditory comprehension, naming and repetition. Lesion sites were coded from radiology reports by two independent reviewers into frontal, temporal, parietal and subcortical categories. Data were analysed in SPSS version 26 using Pearson (point-biserial) correlation and one-way analysis of variance (ANOVA). Results: Frontal lesions involving Broca’s area were most frequent (17/40, 42.5%), followed by temporal lesions around Wernicke’s area (14/40, 35.0%), parietal lesions (6/40, 15.0%) and subcortical lesions (3/40, 7.5%). Frontal lesions were associated with reduced fluency and impaired repetition (r = 0.71, p < 0.001); temporal lesions with impaired auditory comprehension (r = 0.68, p < 0.001); parietal lesions with anomia (r = 0.59, p = 0.002); and subcortical lesions with mixed, milder deficits (r = 0.42, p = 0.03). Among 23 participants who used Kiswahili and English daily, naming in Kiswahili was better preserved than naming in English following temporal lesions (mean difference 1.2 points, p = 0.04). Conclusion: This study, to our knowledge the first lesion–symptom investigation of aphasia in Kenya, shows that the classical perisylvian organisation of language holds in a Kenyan population, while bilingual experience shapes the profile of impairment. The cohort provides a foundation for larger imaging-based studies and for developing linguistically and culturally appropriate rehabilitation in East Africa.

View Full Text