Parkinson´s disease - speech articulation - dysarthria
doença de Parkinson - articulação da fala - disartria
Parkinson’s disease (PD) is a neurodegenerative disorder characterized by loss of
dopamine-producing cells and affects mainly individuals over 60 years but a young
onset subtype is also well recognized[1]. Motor limitations observed in PD are often attributed to basal ganglia dysfunction
associated with decreased dopaminergic input to the sensorimotor region of the striatum[2].
The progressive neuronal loss is associated with a variety of motor and non-motor
deficits in PD patients. In addition to the predominant symptoms such as muscular
rigidity, tremor, bradykinesia and postural instability, many patients develop a distinctive
alteration of voice and speech characterized as hypokinetic dysarthria[3]. Nearly 90% of people with PD will develop voice and speech disorders during the
course of the disease[4], which can have a negative impact on functional communication and result in poor
quality of life[5].
Dysarthria occurs as a result of disturbances in planning and execution of speech
and involves several neural mechanisms related to basal ganglia, cerebellum, supplementary
motor area and frontal circuits[6]. Voice and speech disorders associated with PD are most commonly characterized by
one or a combination of the following perceptual characteristics; reduced vocal loudness[7], breathy or hoarse/harsh voice quality[8], reduced voice pitch inflections or monotone voice[9], and speech impairment articulation[10]. Acoustic measurements revealed that PD patients produce undershooting of articulatory
gestures[11], which leads to imprecise articulation of consonants and vowels[12]. According to some authors, these changes may be present even before the onset of
more overt PD manifestations[13].
Symptoms of PD begin characteristically above the age of 50 years with a mean age
of onset around 60 years but some patients develop PD at a younger age. Young adults
confront problems that are fundamentally different from those faced by people in their
sixties or seventies, and the impact of PD in early-onset patients is therefore likely
to differ from that in older patients[14].
Age of onset of clinical symptoms in PD is one of several criteria employed for disease
classification[15]. Age cut-offs for early versus late onset subtypes vary but 40 or 50 years of age
is used as a defining threshold[16].
Although there is some evidence that age at onset may influence characteristics and
progression of PD[17], differences between early and late-onset PD have not been fully explored. Some
authors have suggested that late-onset PD is associated with more rapid progression,
whilst a slower and more benign course should be expected in early-onset patients[18].
Although speech articulation disturbances have been considered a symptom of disease
progression[19], little is known about its emergence in relation to disease stage and to global
motor function. Studies on the influence of age in the clinical manifestations of
PD, including speech, should contribute to a better understanding of specific symptoms
and to provide better diagnostic, therapeutic and rehabilitation strategies.
The aim of this study is to compare speech articulation between two groups of PD patients
according to age of onset and to correlate the degree of speech impairment with motor
symptoms and disease severity.
METHOD
Patients
Fifty patients with diagnosis of idiopathic PD and preserved speech intelligibility,
with or without speech complaints were included. Diagnosis was made according to the
United Kingdom Parkinson’s Disease Society Brain Bank Criteria[20]. Patients were classified and divided into two groups according to age at onset:
group I (onset between 40 and 55 years) and group II (onset ≥ 65 years). These categories
were termed middle-age onset and old-age groups, respectively. Both groups were matched
for duration of disease.
Exclusion criteria included diagnosis of dementia (DSM IV)[21] or depression (UPDRS-I)[22], any surgical procedure to treat PD, hearing or language disorders and current speech
treatment. Patients were asked to complete the assessment interview with the aid of
caregivers.
Procedures
All participants were receiving regular dopaminergic therapy and were evaluated during
the “on” phase.
Neurologic evaluation
Patients underwent a neurological examination, according to Hoehn and Yahr Scale (HY)[23] and Unified Parkinson´s Disease Rating Scale (UPDRS) part III[22]. Motor scale was divided into four domains: tremor, rigidity, bradykinesia and axial
impairment. Tremor evaluation was based on item 20 and 21; rigidity was based on item
22; bradykinesia was based on items 23, 24, 25, 26 and 31; and axial score was based
on items 18,19, 27, 28, 29 and 30. Motor examination was performed by a movement disorder
specialist.
Perceptual analysis
Patients were asked to emit the sustained vowel /a/, to count numbers from one to
20 and were induced to comment on their speech difficulties (spontaneous speech).
Samples were registered by one of the authors (AED) in a Sony™ MiniDisc, model MZ-R700
from Shure microphone, model SM-58, at a distance of 15 cm and inclined at a 45 degree
angle in relation to patient’s mouth. Samples were further analyzed by three assigned
speech therapists (judges) considered to have expertise in PD. In order to minimize
variability among evaluations, each judge was oriented to pause between analyses to
avoid tiredness. Evaluations were always performed early in the morning at approximately
the same time, blinded in relation to patient identification and at a threshold of
70 dB SPL. No communication among judges was allowed. The variable considered was
speech articulation, which was defined as efficacy of sound production for intelligibility
of oral communication and was considered precise (when all sounds were well defined and clear) or imprecise (when at least some sounds were not well defined and could not be fully understood).
Acoustic analysis
According to previous research[24], spontaneous speech (monologue) was used for acoustic analysis since it can be considered
suitable for the evaluation of speech articulation in PD. Speech samples were digitally
recorded and anonymized by a speech therapist (AED) in a quiet room using Shure microphone
SM-58, placed approximately 15 cm from the patient’s mouth. The formant frequency
values F1 and F2 were measured separately for each vowel for a 30-millisecond segment
at the temporal midpoint, determined by three blinded examiners, of each vowel using
a PRAAT software v5.3.30 [available at: www.praat.org (Phonetic Sciences, University
of Amsterdam)].
The vowel analysis was performed using the spontaneous speech in which the patients
were instructed to speak about their speech difficulties. For each patient, 10 occurrences
of the three vowels /a/, /i/ and /u/ were extracted from de monologue and was based
upon the established Vowel Articulation Index (VAI) develop by Sapir and coworkers[25]. The measurement of VAI can be expressed using the following formula: (F2/i/+F1/a//)/(F1/i/+F1/u/+F2/u/+F2/a/), which is a surrogate parameter of the first and second formant frequencies (F1
e F2) of the corner articulation vowels /a/, /i/ and /u/[26]
,
[27]. The vowel data of F1 and F2 in Hertz (Hz) were separately averaged for all corner
vowels of each patient.
Acoustic analysis was performed separately for both genders because mean VAI is related
to the speaker’s fundamental frequency of voice.
Statistical analysis
Analysis of group distribution showed that data from this study do not follow a normal
distribution and thus a non-parametric analysis was employed. The following non-parametric
tests were performed: the Mann-Whitney test was employed to compare stage of disease,
scores of UPDRS-III and speech articulation (acoustic and perceptual) between the
two groups. The correlation analysis of Spearman was performed for the correlation
of the same variables within each group.
Analysis reliability was estimated by the concordance degree among judges by Kappa
coefficient. Concordance was established by comparing evaluations obtained by each
jugde by Cronbach α correlation (p ≤ 0.05).
Ethics
The study was submitted to the Commission of Ethics for the Analysis of Research Projects
(CAPPesq) and approved. All participants were previously fully informed about the
research and signed consent forms before they were submitted to any evaluation.
RESULTS
[Table 1] shows sample characteristics and the comparisons between groups regarding neurological
and articulation features. Both groups were homogeneous in relation to disease duration,
clinical stage, motor status and perceptual and acoustic behavior of speech articulation.
Overall, these results indicate that PD patients with different ages at onset were
clinically similar.
Table 1
Demographics, clinical presentation and articulation characteristics.
|
Characteristics
|
Group I (n = 30)
|
Group II (n = 20)
|
Significance: p ≤ 0.05
|
|
Middle-age onset
|
Old-age onset
|
Mann-Whitney test
|
|
Mean ± SD
|
Mean ± SD
|
|
|
Age
|
59.96 ± 4.85
|
80.40 ± 7.37
|
p < 0.001
|
|
Age at onset
|
49.26 ± 3.41
|
70.85 ± 7.33
|
p < 0.001
|
|
Duration of disease
|
10.56 ± 4.46
|
9.73 ± 4.43
|
p = 0.538
|
|
Hoehn and Yahr
|
2.71 ± 0.71
|
3.18 ± 0.75
|
p = 0.736
|
|
UPDRS
|
|
|
|
|
Tremor
|
4.66 ± 2.68
|
4.15 ± 2.71
|
p = 0.611
|
|
Rigidity
|
9.80 ± 2.96
|
9.05 ± 3.15
|
p = 0.544
|
|
Bradykinesia
|
14.70 ± 4.41
|
15.45 ± 5.96
|
p = 0.506
|
|
Axial
|
11.43 ± 5.25
|
13.45 ± 5.32
|
p = 0.190
|
|
Perceptual analysis
|
|
|
|
|
Articulation
|
0.43 ± 0.50
|
0.50 ± 0.51
|
p = 0.528
|
|
Acoustic analysis
|
|
|
|
|
VAI – Male
|
0.75 ± 0.09
|
0.79 ± 0.05
|
p = 0.260
|
|
VAI – Female
|
0.81 ± 0.07
|
0.88 ± 0.08
|
p = 0.492
|
UPDRS: Unified Parkinson’s Disease Rating Scale; VAI: Vowel Articulation Index; SD:
standard deviation.
Group I included patients between 48 and 69 years of age and age of onset between
42 and 55 years group II included patients between 68 and 94 years of age and age
of onset between 65 and 84 years.
Both groups had similar disease duration (2-18 years) and HY staging (between 2 and
4). Motor scores according to the UPDRS scale and speech articulation behavior were
not statistically significant between groups.
[Table 2] shows perceptual characteristics of speech articulation.
Table 2
Perceptual characteristics of speech articulation.
|
Characteristics
|
Group I (n = 30)
|
Group II (n = 20)
|
Significance: p ≤ 0.05
|
|
Middle-age onset
|
Old-age onset
|
Mann-Whitney test
|
|
N (%)
|
N (%)
|
|
|
Articulation
|
|
|
|
|
Precise
|
17 (57)
|
6 (30)
|
p = 0.114
|
|
Imprecise
|
13 (43)
|
14 (70)
|
p = 0.528
|
[Table 3] shows correlations between neurological and perceptual and acoustic articulation
characteristics in both groups. Statistical analysis showed a positive relationship
between perceptual ratings and acoustical measures and severity of speech impairment
and axial and bradykinesia scores.
Table 3
Correlation between neurological symptoms and speech articulation.
|
Characteristic
|
Correlation coefficient
|
Significance: p ≤ 0.05
|
|
|
Spearman analysis
|
|
Speech articulation (perceptual ratings x acoustical measures)
|
+ 0.165
|
p < 0.001
|
|
Speech articulation x UPDRS-III Axial
|
+ 0.350
|
p = 0.013
|
|
Speech articulation x UPDRS-III Tremor
|
+ 0.041
|
p = 0.779
|
|
Speech articulation x UPDRS-III Rigidity
|
+ 0.263
|
p = 0.065
|
|
Speech articulation x UPDRS-III Bradykinesia
|
+ 0.277
|
p = 0.005
|
|
Speech articulation x Hoehn & Yahr
|
+ 0.246
|
p = 0.006
|
UPDRS: Unified Parkinson’s Disease Rating Scale.
Kappa correlation analysis indicated significant intra-judge concordance for perceptual
and acoustic evaluation (0.879; p ≤ 0.01). Cronbach α correlation analysis indicated
significant inter-judges concordances for perceptual and acoustic evaluation (0.957;
p ≤ 0.01).
DISCUSSION
Although aging itself is not a cause of PD, the disease is an age-related disorder
and affects all different aspects of speech and voice and leads to imprecise articulation
of consonants and vowels[28].
In order to identify a possible relationship between age at onset and PD speech impairment,
we analyzed the pattern of speech disturbances in individuals with middle-age and
old-age onset PD. Both groups were homogeneous and there was no significant difference
between groups for disease duration, HY stage, UPDRS-III scores and speech articulation
(perceptual and acoustic). This suggests that the influence of age at onset on demographics
data, clinical presentation and speech articulation was minimal or none.
Patients in our study were in the moderate or advanced stages of disease (mean disease
duration of 10 years, average HY = 3). Our data indicate that the degree of articulation
impairment is positively correlated with disease progression.
The specific pattern of the development of speech symptoms with disease progression
is still unknown. Although we could not find any differences in overall UPDRS part
III between groups, correlation analysis suggested that the severity of axial symptoms
and bradykinesia were associated with poor and imprecise speech articulation performance
in both groups.
In a study comparing acoustic speech characteristics and global motor performance
a similar correlation with bradykinesia was found[18].
These similarities might be explained by shared pathophysiological mechanisms responsible
for reduced articulation, bradykinesia and axial symptoms. Overall, these findings
support the hypothesis that speech articulation impairment could be the result of
axial dysfunction and bradykinesia.
We were able to document a speech articulation impairment not only based upon perceptual
judgment, but also substantiated by objective acoustic measures with a VAI, an acoustic
metric widely used to quantify articulatory function, as indication for a deterioration
of articulation. The combination of perceptual and acoustic analysis of speech articulation
as performed in the current study seemed to be appropriate to complimentarily obtain
clinical surrogate measures of the articulation, and to measure changes that would
be too subtle to be detected by perceptual judgment only. In our study, all patients
presented preserved speech intelligibility, but alterations of vowel articulation
were detected by a measurement of VAI with an average of less than 1.0 as indicator
of articulation impairment[29].
Comparison of acoustic variables in relation to gender was not a primary concern but
acoustic analysis was evaluated separately for both genders since mean VAI is related
to fundamental frequency[26].
The presence of tremor may negatively influence some functional abilities such as
writing and handling of common utensils but it does not appear to interfere with speech
articulation. Rigidity could also influence speech since increased muscle tone may
contribute to mobility limitation but in our study this correlation was not statistically
significant. However, other studies suggest that bradykinesia and rigidity may contribute
to the reduced mobility of speech-related structures and may play a role in the pathophysiology
of impaired articulatory function[30].
Our study was not designed to characterize specific types of articulatory dysfunction
or speech coarticulation. However it was possible to identify patterns of imprecise
articulation and phonetic alterations, which included omission (when certain sounds
were not produced), distortions (wrong emissions) and repetitions (when one or more
sounds were duplicated in the same word).
Admittedly, this study has some methodological limitations mainly derived by our attempt
to analyze only spontaneous speech in objective evaluation. In fact, according to
a previous study, complex tasks such as monologue are more likely to elicit articulatory
deficits in parkinsonian speech[24]. The impact of the age of onset on speech articulation cannot be captured by acoustic
and perceptual analysis alone and disability or functional changes perceived by the
patients themselves and/or their caregivers should also be taken into account.
In conclusion, our study suggests that global motor disability and speech articulation
impairment do not correlate with age at onset of PD symptoms or age of the patients
at evaluation. Moreover, speech impairment was associated with axial symptoms, bradykinesia
and stage of the disease.
