From October 2002 to June 2004
Project Leader(s): Dr Nick Miller
Staff: D Burn, T McGee, U Brechany
Contact: Dr Nick Miller
Sponsors: Rhoda Lockhart Research Fund
Overall aims of the study were:
1) To ascertain the pattern of change in selected speech and language variables between practically defined 'off' and 'on' states in PD.
2) To compare the pattern of changes in the practically defined 'on' and 'off' states in PD patients with stable versus fluctuating levodopa responses.
Study 1: Acoustic investigations
The control and PD groups differed significantly at all times on mean length of vocalisation of ‘ah’; ‘ah’ as high vs low as possible; variability of fundamental frequency (Fo, measured here in semitones to normalise for gender differences) in the reading passage; and pitch perturbation. There was no significant difference in amplitude perturbation, nor rate of speech.
The ‘stable’ group performed significantly better than the ‘fluctuating’ group on measures of ‘ah’ as long as possible, pitch perturbation, amplitude perturbation and variability of Fo. All other differences were non significant.
As groups overall neither showed any significant differences on acoustic measures between baseline and later time points. Inspection of individual performances indicated however that group outcomes may have been influenced by individual performances: some individuals improved significantly on selected variables, whilst others actually deteriorated.
Study 2: Intelligibility and speech naturalness
Speakers with PD were significantly less intelligible and perceived as significantly less natural than the control speakers. There were no significant group differences in intelligibility scores between baseline and later time points. Some individual speakers did show a significant difference (always improvement), and these tended to be the more impaired speakers. Generally all speakers with PD obtained only mildly impaired intelligibility ratings. This maybe created a ceiling effect making it difficult to detect/ produce a significant improvement.
Similar results were obtained for the naturalness ratings – there were changes across the cycle, but none statistically significantly better or worse.
Study 3: Language measures
We examined whether patients with PD but no depression or dementia show language deficits; whether performance on language tasks changed across the L-dopa cycle; whether patients with a stable vs fluctuating response to L-dopa performed differently.
Language tasks involved computerised tests of written rhyme judgment (do these pairs of written words rhyme or not: cow-know; mint-pint; cuff-rough), sentence verification, (is this a well formed sentence: admirals command baskets; oranges are fruit), category decision (is this a fruit or not – when presented with words on screen), synonym judgement (do these words mean more or less the same thing or not: ship-boat; bucket-water). We monitored response accuracy and reaction time to response.
Response accuracy did not differ significantly across time between PD and controls, nor stable and fluctuating patients. Reaction time in the PD participants fell significantly over the l-dopa cycle; but it did also for controls. There was no association between reaction time and motor scores (finger tapping; UPDRS). Only at 60 minutes was the PD group significantly slower at rhyme judgement.
|
Professor Nick Miller
|
|