Beyond Words: Cross-Sectional Analysis of Non-Verbal Auditory Measures Across Cognitive Health, Mild Cognitive Impairment and Alzheimer's Disease Dementia

Abstract Background Speech-in-noise hearing is impaired in early symptomatic Alzheimer’s disease. However, most tests involve the use of verbal stimuli where performance measures may be confounded by linguistic and cultural factors. Non-verbal auditory measures may overcome these issues. Methods 158 cognitively healthy, 26 mild cognitively impaired and 28 participants with Alzheimer’s disease dementia underwent evaluation using the Addenbrooke’s Cognitive Examination (3rd Edition), pure-tone audiometry, speech-in-noise hearing testing with digits and sentences and non-verbal auditory short-term memory for basic sound features. Group-level differences were assessed after adjusting for age, sex and educational attainment. Multinomial logistic regression and receiver operator characteristic metrics were used to test the fit of the model to the diagnosis using verbal and non-verbal auditory variables. Results Non-verbal measures provided a better fit to diagnosis (cognitively normal, mild cognitive impairment or Alzheimer’s disease dementia) (Akaike Information Criteria >10) using logistic regression as compared to verbal measures. There were no statistically significant differences using receiver operator characteristic measures. Conclusions Non-verbal auditory measures are as good as verbal speech-in-noise measures at discriminating between cognitively healthy, mild cognitively impaired and people with Alzheimer’s disease dementia.