Speech-in-noise perception is a marker of preclinical Alzheimers disease


Central auditory function is affected early in the course of Alzheimer’s disease, and may be useful as a prodromal disease marker. Speech-in-noise perception is a simple test of central auditory function.


We used Speech Reception Threshold (SRT) as a measure of speech-in-noise perception in 160,905 healthy subjects in UK Biobank. We assessed odds ratios for incident Alzheimer’s disease by SRT quintile at a mean follow up of 10 years. In a subgroup of 37,082 healthy subjects with contemporaneous cognitive testing and brain imaging we assessed the relationships of SRT with cognitive performance and structural neuroanatomy (using voxel based morphometry (VBM)).


Poorer SRT was associated with greater odds of incident Alzheimer’s disease (odds ratio for quintile 5 relative to quintile 1: 2.91 (95%CI 1.92–4.57) p=0.000001). Poorer SRT was associated with poorer cognitive performance in multiple domains. VBM revealed symmetrical grey matter correlates of SRT in a distributed network of auditory and speech perception areas, as well as medial temporal lobe structures.


Speech-in-noise perception is a simple non-invasive assay of cognition, brain structure and risk of incident Alzheimer’s disease, and as such may be readily translatable as a cost-effective tool for dementia risk stratification at scale.