The effect of simulated visual impairment on speech‐reading ability

Abstract

Speech-reading is the ability to recognise certain sounds visually, using both the movements of the speaker's mouth, plus other sources of visual information, such as gestures and body language, facial expressions and situational clues. This study aimed to determine the effect of mild degrees of visual impairment on speech-reading ability. A group of 15 normally-sighted subjects was assessed with different levels of simulated visual loss using occlusion foils, in the presence of constant background noise to simulate hearing loss. The visual impairments created ranged from 20 to 13 dB log contrast sensitivity (contrast thresholds 1-5%) (measured using the backlit Melbourne Edge Test) with acuity dropping from 6/6 to 6/24 (logMAR 0.0-0.6). The speech-reading ability with simulated impairment and noise was quantified as the number of words recognised correctly when watching a video of an actor speaking meaningful sentences. The introduction of the mildest simulated visual impairment (VA 6/6 logMAR 0.0; contrast sensitivity 20 dB, contrast threshold 1%) was sufficient to cause a statistically significant reduction in speech-reading ability (t-test, p < 0.001), and as the level of visual impairment was increased, the speech-reading ability became progressively worse throughout the range of impairments tested. This study shows that if hearing is compromised, then an observer's speech-reading ability is sensitive to changes in vision. It is likely that visual loss will have consequences for speech-reading ability. This will be particularly relevant to the elderly population who are often found to have a dual sensory loss that causes compromise to both visual and auditory sensory reception. Some of the difficulties in communication reported by the elderly with reduced hearing could be visual as opposed to auditory in origin.