Abstract
Background:
Recognition of motor signs in the prodromal stage could help identify those at risk of developing Parkinson’s disease (PD).
Objective:
This study identified motor symptoms and signs in individuals suspected of having PD but who did not have a progressive reduction in the speed and amplitude of finger tapping or other physical signs indicative of bradykinesia.
Methods:
146 patients, who had symptoms or signs suggestive of PD, were serially evaluated by a movement disorder specialist, using the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) Part III and video recordings. If the patients ‘converted’ to PD during follow-up, they were categorized as cases and compared with those who did not meet PD criteria during follow-up (non-cases).
Results:
The 82 cases were more likely to have action dystonia or postural/action/rest tremor of a limb (OR 2.8; 95% CI 1.1–7.1; p = 0.02), a reduced blink rate at rest (OR 2.3; 95% CI 1.2–4.6; p = 0.01), anxiety (OR 8.9; 95% CI 2.6–31.1; p < 0.001), depression (OR 7.0; 95% CI 2.9–17.2; p < 0.001), or a frozen shoulder (OR 3.1; 95% CI 1.6–6.2) than the 64 ‘non-cases’.A reduction of the fast blink rate was common in patients who met the criteria for PD (p < 0.001).
Conclusions:
This study emphasizes that motor dysfunction is a component of the clinical prodrome seen in some patients with PD.
INTRODUCTION
The recognition of non-motor symptoms and early motor signs in the prodromal phase of Parkinson’s disease (PD) is an important academic initiative driven by the notion that, when neuroprotective treatments eventually become available, such individuals might be the best group to target for clinical trials [1].
The diagnosis of PD requires the presence of bradykinesia (reduction in speed and amplitude of finger tapping over 20 seconds) and the presence of rigidity and/or rest tremor [2–4]. These cardinal signs need to be distinguished from mild extrapyramidal signs including slowness in the elderly and psychomotor retardation in severe depression [5–7]. Subtle motor dysfunction, that may have been transient, can often be suspected in hindsight or after the review of historical copies of handwriting and family videos [8–11]. A strong family history of PD or tremor, loss of sense of smell, REM sleep behavior disorder, refractory constipation and a history of severe depression also increase the risk of developing PD [12–16].
Between 2007 and 2008, we examined newly-referred patients with motor and non-motor symptoms for objective signs of motor dysfunction. We included people with complaints of slowness, symptoms of depression or anxiety, disturbed temperature regulation with excessive sweating, asymmetric postural tremor, orthostatic hypotension, urinary incontinence, hyposmia, REM sleep disorder, and shoulder frozen [17–23]. Subtle objective motor abnormalities at baseline were identified and included ‘suggestive’ signs such as clumsiness, slowness or focal stiffness, and non-specific signs (see Table 1) [24, 25]. We then followed this group prospectively with the hope of better characterizing the pre-diagnostic motor phase of PD.
Inclusion criteria in patients with pre-diagnosis of Parkinson’s disease
MATERIALS AND METHODS
The follow-up phase of the study was conducted at the Movement Disorders Outpatient Unit of the Neurology Department, Eugenio Espejo Hospital (a national reference center) in Ecuador, between 1 January 2009 to 1 January 2016. In accordance with the Declaration of Helsinki, the local research ethics committee approved the study and all participants provided written informed consent.
One hundred and fifty patients referred mainly by neurologists and general physicians, but also some self-referrals, were assessment at baseline for signs of parkinsonism. Four patients were excluded; one of these had hemidystonia, one had probable vascular parkinsonism and two had symmetric postural/action tremor with dystonia [26, 27]. One hundred and forty-six patients were enrolled in the study because they fulfilled the following criteria: attended the baseline visit; free from overt parkinsonism and dementia at the baseline assessment; had one ‘suggestive sign of parkinsonism and at least one ‘non-specific’ feature, with one or more non-motor symptoms (see Table 1).
Assessment was undertaken at the first consultation and then every 3 months, and included a clinical examination by a neurologist specializing in movement disorders (FA). All patients were evaluated with the UPDRS Part I-II-III [28]. Dexterity was evaluated using fine and alternating movements (finger tapping, pronation and supination of the hand for one minute). Clinical examinations were recorded on video. The blink rate was observed (both at rest and fast voluntary blinking) [29]. The two test conditions were: (1) resting blink rate was counted and (2) the patient was asked to look straight ahead and blink as fast as possible for 1 minute. Blinks were counted as full blinks if at least 50% closure of the eye occurred. Participants were asked about a family history of PD and tremor [30, 31]. Information about early motor and non-motor manifestations for possible PD were also gathered using a questionnaire. The Hamilton depression and anxiety scales and the Mini-Mental State Examination (MMSE) [32–34] were applied with standardized scales in all patients by a neuropsychologist. Magnetic resonance imaging (MRI) with T2 FLAIR and diffusion-weighted imaging was performed on all patients at baseline, and did not suggest any structural causes for the presenting complaints. A sleep medicine expert asked during the interview if the patient had sleep disturbances including, REM sleep behavior disorder, excessive daytime sleepiness (EDS), insomnia and parasomnias. We asked during the interview if patients recall loss or reduction in their smell or taste and if they struggle with discrimination and identification of odors.
All patients were followed-up for a minimum of 1 year. If in the course of follow-up, the patients developed PD fulfilling clinical criteria [2, 35] then they were re-categorized as cases. Those participants who did not meet PD clinical criteria remained as non-cases for the analysis. The last follow-up visit was when a diagnosis PD was made for cases and the visit after which the observation period finished for non-cases
In six patients we confirmed the diagnosis of PD with a levodopa challenge test. One patient, who was diagnosed with PD had the diagnosis revoked during follow-up and dopaminergic treatment was stopped. The patient then worsened and reverted to take levodopa. During follow-up of the 82 patients who converted to PD, the diagnosis was revised to progressive supranuclear palsy in one case, multiple system atrophy in one, corticobasal degeneration in one and Lewy body dementia in two patients
Statistical analysis
The data collected were expressed as the frequency (percentage) for categorical variables and mean±standard deviation for continuous numerical variables. A comparison between cases and non-cases was done for demographic and clinical data at baseline as well as UPDRS score at baseline and at the last follow-up time. The categorical data were compared with z-test. Continuous data were compared by unpaired Student’s t-test. The tests were carried out with a two-tailed statistical significance level set at p = 0.05.
For exploratory purposes, a nested case-non case analysis (unmatched) was performed to estimate the association between some clinical features of interest, including potential risk factors for PD at baseline, and the development of PD.
Odds ratios (ORs), 95% CI) and p-values were calculated for these variables considering the clinical conditions as predictors and PD patients as cases. The statistical analysis was performed with SAS9.3 software.
RESULTS
One hundred and forty-six patients (48 females, 98 males) who met the inclusion criteria described in Table 1 participated in follow-up. All of them had subtle motor dysfunction or extrapyramidal signs at baseline (i.e., a combination of suggestive and non-specific motor signs on neurological examination). The patients had a mean follow-up duration of four years (range 1–7 years). The clinical features of these patients are summarized in Table 2.
Baseline and follow-up demographic data in cases and non-cases
Cases, Patients who met the criteria for PD. Non-cases, Patients who did not meet the criteria for PD.
Eighty-two patients (27 females and 55 males) were diagnosed with PD during the follow-up period. Their age at the onset of motor symptoms (59.4±14.7 vs. 56.0±17.1; p = 0.19) and at the baseline study visit (63.4±11.9 vs. 59.7±15.4; p = 0.10) tended to be greater. The average time between the date of registration with the study and the diagnosis of PD was 3.1 years. At the last follow-up visit, the age of converters was higher than in the non-converters (68.5±12.0 vs. 64.3±15.1; p = 0.06) (Table 2). The follow-up time during the study was similar in the PD patients (5.1±1.4 vs. 4.7±1.1; p = 0.03). When comparing the two groups, there was no difference in the frequency of a family history of PD (25.6% vs. 28.1%) or tremor (18.3% vs. 20.3%), and none of the patients had a family history of dystonia.
Baseline features were similar in the two groups, including clumsiness (43% vs. 39%), slowness of movement (28% vs. 25%), stiffness (29% vs. 22%), action dystonia or tremor of a limb (7% vs. 5%), slow blink rate at rest (28% vs. 25%), tremor (87% vs. 89%), UPDRS I (1.05±0.9 vs. 0.8±0.8), and UPDRS II (8.5±3.1 vs. 8.0±3.4) were no different (Table 3). Only UPDRS III 9.5±1.8 vs. 8.7±2.0) showed a small difference between groups at baseline (Table 3). RBD and other combined sleep disorders (46 vs. 19%) were different. Constipation (38% vs. 42%), hyposmia (28% vs. 31%), hypogeusia (17% vs. 17%), and frozen shoulder (17% vs. 9%) showed no clear group differences at baseline(Table 3).
Baseline assessment of UPDRS score and motor symptoms in cases and non-cases
Cases, Patients who met the criteria for PD. Non-cases, Patients who did not meet the criteria for PD.
At the last follow-up assessment tremor and dystonia triggered by action were more frequently seen in patients who were diagnosed with PD during that visit (26% vs. 11%; p = 0.02) (Table 4). Four patients showed dystonia of the foot after prolonged physical activity; one during swimming and another complained of dystonic cramping of the hand while playing the drums. Reduced mean blink rate at rest was also more frequent in the cases that were diagnosed with PD than in non-cases (55% vs. 34%; p = 0.01) (Table 4).
Association between signs and symptoms in last time assessment in cases and non-cases
Cases, Patients who met the criteria for PD. Non-cases, Patients who did not meet the criteria for PD.
In the last follow-up visit after which the observation period finished for non-cases, one hundred twenty-nine patients (88%) also had non-motor symptoms. Frozen shoulder (capsulitis) (62% vs. 34%; p < 0.001), depression (46% vs. 11%, p < 0.001), RBD and other combined sleep disorders (46% vs. 19% <0.001), and anxiety (31% vs. 5%; p < 0.001) were all more frequent in the PD cases, whereas constipation, hyposmia, hypogeusia did not show differences (Table 4).
When comparing UPDRS scores from the last assessment to the baseline assessment, greater changes were observed in the cases that were diagnosed with PD during follow-up; UPDRS (Parts I–III) (33.6±9.1 vs. 23.5±7.2; p < 0.001) and UPDRS motor (Part III) (19.1±5.3 vs. 11.7±2.8; p < 0.001).
Bradykinesia was, as expected, present in patients that converted to PD (100% vs. 33%; p < 0.001), and rest tremor (88% vs. 9%; p < 0.001) and rigidity (82% vs. 22%; p < 0.001) were also common, compared to the non-cases.
Comparison of fast blink rate between baseline (95.2±24.2 vs. 98.7±30.4, p = 0.43) and last assessment fast blink rate (80.2±24.6 vs. 96.0±30.6, p < 0.001), showed clear deterioration in rate in patients who met the criteria for PD during follow-up compared with non-cases.
In the last follow-up visit, an association was found between the presence of action dystonia of a limb /tremor postural/action/rest of a limb (OR 2.8; 95% CI 1.1–7.1; p = 0.02) and reduced blink rate at rest (OR 2.3; 95% CI 1.2–4.6; p = 0.01), and the development of PD (Table 4).
In the last follow-up visit non-motor symptoms, the probability of developing PD was higher when patients presented with anxiety (OR 8.9; 95% CI 2.6–31.1; p < 0.001), depression (OR 7.0; 95% CI 2.9–17.2; p < 0.001), and if they had symptoms of a frozen shoulder (OR 3.1; 95% CI 1.6–6.2; p < 0.001) (Table 4).
DISCUSSION
We followed a group of patients referred to a hospital neurological department with suspected parkinsonism. Patients who were subsequently diagnosed with PD during follow-up showed subtle motor signs for a mean 3.1 years before the diagnosis was made [2, 9–14].
The age at diagnosis of PD was similar to that reported in other studies and was similar to the participants who did not develop PD [28, 36]. The proportion with a positive family history of PD was similar in the two groups, but was higher than that reported in most other published studies [28, 29].
A frozen shoulder (capsulitis) is a well-recognized harbinger which can precede a diagnosis of PD by several months or even years [9, 39]. 62.2% of the cases in this series had pain and stiffness with limited movement at the shoulder, which is a higher percentage than that reported in other series and may relate to akinesia [9, 40].
Tremor was the commonest motor symptom, present in 87% of patients with PD, which is similar to that found in a previous study [28]. In 79 patients who met the criteria for PD, the tremor was postural/action and in three patients a monosymptomatic rest tremor occurred. There is some evidence to support a link between asymmetrical postural and action tremor and PD [19, 28]. In our series, all the patients with asymmetric tremor who developed PD also had additional subtle motor signs [19, 29]. Focal dystonia brought on by action and most frequently affecting the toes and foot (dystonic claudication) was also common at the time that patients were diagnosed with PD. This is recognized as presenting feature of young onset Parkinsonism (idiopathic and monogenetic forms) [21, 23], but was also noted in our study.
Reduction in resting blink rate is common in PD [22, 41]. It has been suggested that there is a link between the average spontaneous blink rate at rest and striatal dopaminergic function [22, 42]. We found a reduced blink rate at rest in 55% of the cases and a slowing of repetitive voluntary blinking performed over one minute [27]. Previous studies have suggested reduced spontaneous blinking as an early sign of PD [22, 42]. Further studies are required to validate the ‘fast blink test’ to confirm its utility in early PD.
UPDRS III is a poor discriminator between soft extrapyramidal signs in the elderly, depression and those individuals who meet diagnostic criteria for PD, but changes in motor scores over time may be a useful pointer. In our study action dystonia or tremor of a limb and reduced blink rate were other symptoms associated with a higher risk of developing PD [19–23, 40–42].
Many cohort studies have evaluated the progression of the prodromes of PD using scales that were designed for use in established PD [43, 44]. The design of our study allows a much broader capture of clinical observations of motor dysfunction, years before patients met the criteria for PD. This in turn might help the field develop better quantitative scales that better capture motor dysfunction in the prodrome. The limitations of the study include the short follow-up period, potential bias due to not using standardized scales for some motor and non-motor manifestations, as well as the lack of a control group that would allow a better understanding of the motor dysfunction trajectory.
More than half our cases converted to PD during follow-up indicating that the study group already had a very high risk of developing PD at the time of hospital referral. This study further emphasizes that subtle motor signs and symptoms may occur before overt bradykinesia can be identified clinically in PD and emphasizes that these need to be identified with similar care as non-motor symptoms in studies designed to identify at risk groups.
CONFLICT OF INTEREST
The authors have no conflict of interest to report.
