Among nonmotor symptoms in Parkinson’s disease (PD), neuropsychiatric ones (NPS; e.g., depression, anxiety, apathy, psychosis, and impulse control disorders) are particularly common and disabling.
1 Drug-naive patients with early PD commonly manifest mood changes and apathy segregating into two distinct neuropsychiatric syndromes that result from “both the psychological stress and brain changes.”
2 The contribution of psychosocial, demographic, clinical factors, complications of dopaminergic therapy and neurodegenerative process may differ by disease stage as recently suggested by Weintraub et al.
3 The authors reported apathy, anxiety, and depression—but not psychosis and impulse control disorders—as symptoms that differentiate patients with early, untreated PD from age-matched healthy subjects suggesting that some NPS might indeed relate to neurodegeneration in PD, whereas the other symptoms were more likely to be attributable to other factors such as complications of dopamine replacement therapy (DRT).
3Therefore, it is important to determine the pattern of NPS expression longitudinally from the initial stages of PD as well as effects of different variables. Various NPS, including also altered dreams and nocturnal sleep disturbances, are quite common in patients in the early stages of PD compared with matched healthy participants.
4–6 Depression, anxiety, impaired concentration, memory complaints, and insomnia have the greatest negative impact on health-related quality of life in patients with newly diagnosed PD.
7Methods
This study comprised 117 patients with newly diagnosed PD in the initial, hemiparkinsonian stage of the disease (Hoehn and Yahr stage 1). Patients were recruited at the Neurology Clinic, School of Medicine, University of Belgrade, from January 1, 2012 to December 31, 2013. Neurologists in Serbia were informed about the enrollment and encouraged to refer their newly diagnosed PD patients to us, even if they had already started antiparkinsonian therapy. The diagnosis of PD was (re)established in our institution by an experienced neurologist specialized in movement disorders according to step 1 of the UK Parkinson’s Disease Society Brain Bank Diagnostic Criteria.
9 Only those patients with a disease course shorter than 2 years from the first parkinsonian symptom(s) were included in the study. Exclusion criteria were atypical and secondary parkinsonism, some of the more prevalent genetic causes of parkinsonism (genetic testing excluded
parkin,
LRRK2,
DJ1,
PINK1, and
VPS-35 mutations), dementia (according to the Movement Disorder Society [MDS] Task Force recommendation criteria),
10 a history of psychosis requiring recent neuroleptic treatment, recent concomitant neurologic disorders, reported alcohol or drug/substance dependence, a history of significant head injury, and pacemaker or metallic implants.
One hundred and nine unrelated healthy controls matched for age, sex, and education were recruited in the same period (mainly spouses, friends, or hospital staff). Exclusion criteria for the healthy control group included the presence of a current or previous central or peripheral nervous system disorder, a current or previous psychiatric disease that requires specific therapy, reported alcohol or drug/substance abuse, a learning disability or mental retardation requiring special education, a history of significant head injury, and pacemaker or metallic implants. Both patients and healthy subjects underwent 1.5T MRI of the brain to exclude extensive brain vascular damage or other major brain lesions.
This study was approved by the Ethical Committee of the School of Medicine, University of Belgrade. Participants were enrolled in the study after they provided written informed consent. The authors declare that they acted in accordance with the ethical standards of the 1964 Declaration of Helsinki.
Clinical Assessment
Using a semistructured interview, all patients were carefully asked about demographic and clinical features. All tests used in this study were administered to both patients with PD and healthy controls (when appropriate).
The severity of PD was assessed using the MDS Unified Parkinson’s Disease Rating Scale (MDS-UPDRS).
11 Depressive and anxiety symptoms were assessed by the Hamilton Depression Rating Scale (HDRS)
12 and the Hamilton Anxiety Rating Scale (HARS),
13 and apathy was assessed by the Apathy Scale (AS).
14 The Neuropsychiatric Inventory (NPI)
15 was administered to all participants to assess the frequency (4-point scale) and severity (3-point scale) of 10 neuropsychiatric disturbances (delusions, hallucinations, agitation, dysphoria, anxiety, euphoria, apathy, disinhibition, irritability, and aberrant motor behavior), and a score for each disturbance was obtained by multiplying frequency by severity. The NPI total score was a sum of the subscale scores. Psychotic complications and some aspects of compulsive behavior were assessed by the Scales for Outcomes in Parkinson’s Disease–Psychiatric Complications (SCOPA-PC).
16 The Scales for Outcomes in Parkinson’s Disease–Sleep Scale (SCOPA-S)—consisting of the nighttime sleep (SCOPA-NS) and the daytime sleepiness (SCOPA-DS) subscales and SCOPA-S overall
17—and the REM Sleep Behavior Disorder Screening Questionnaire (RBD-SQ)
18 were used to define corresponding sleep problems.
Cognitive evaluation was done using the Addenbrookeʼs Cognitive Examination–Revised (ACE-R).
19 A comprehensive neuropsychological battery covering five cognitive domains and consisting of two tests for each domain was applied to all patients with PD and healthy controls, as previously described in the same cohort.
20 Diagnosis of PD with mild cognitive impairment (PD-MCI) was based on the MDS Task Force level 2 criteria
21 and required the following: (a) the presence of subjective cognitive complaints obtained from the patient or caregiver, (b) cognitive deficits that were insufficient to significantly interfere with everyday functioning, and (c) impaired performance (not attributable to causes other than PD) on at least two neuropsychological tests, represented by either two impaired tests in one cognitive domain or at least one impaired test in two or more cognitive domains.
Statistical Analysis
Demographic and clinical characteristics between groups were compared using analysis of variance for continuous variables and the chi-square test for categorical variables where appropriate. Discriminant function analysis (forward stepwise) was applied in order to test whether a set of variables (HDRS, HARS, AS, SCOPA-DS, and SCOPA-NS) was effective in predicting category membership (PD patients versus healthy controls). Results were considered statistically significant at p≤0.05. The analyses were performed using SPSS software (version 17).
Results
Of the 117 initially recruited PD patients, 111 were included in the analyses (two patients refused to complete the testing, two were not able to undergo MRI examination due to claustrophobia, one patient had extensive vascular MRI lesions, and one had meningioma). One hundred and five healthy control participants were included in direct comparisons (three participants were excluded because of extensive white matter MRI lesions, and one due to the accidental finding of a silent infarct) (
Table 1). Six patients (5.4%) were depressed and seven (6.3%) were anxious before the onset of PD. Among them, four (3.6%) and six (5.4%) patients were taking antidepressant and anxiolytic drugs, respectively, at the time of the assessment. None of the control subjects were currently or previously treated for depressive or anxiety disorder.
Even in the initial motor stage of the disease, patients with PD had significantly higher scores on all psychiatric scales and significantly lower scores on the ACE-R compared with healthy control subjects (
Table 1).
Discriminant Function Analysis
A forward stepwise statistical discriminant analysis was performed using five variables (HDRS, HARS, AS, SCOPA-DS, and SCOPA-NS) as predictors of membership in the two groups (patients with PD versus healthy controls). Values of 4.0 and 3.9 were considered for F-statistic to enter and to remove variables, respectively. The Apathy scale, SCOPA-DS, and SCOPA-NS scores were selected as the most important variables for differentiating between the two groups of participants. When these predictors were used, an F(3,204) value of 40.75 was significant (p<0.001). That is, there was statistically significant separation of the PD and healthy control groups based on three predictors combined: AS (F to remove=55.90, p<0.001), SCOPA-DS (F to remove=21.47, p<0.001), and SCOPA-NS score (F to remove=5.27, p<0.02).
Psychiatric Manifestations in Patients With Early PD
The presence of apathy, depression, and anxiety disorders was based on the cut-off scores for the AS (13/14),
14 HDRS (9/10),
22 and HARS (12/13).
23 A total of 37 patients with PD (33%) and four healthy controls (4%) had apathy (χ
2=30.93, df=1, p<0.001), 24 patients (22%) and five healthy controls (5%) were depressed (χ
2=13.20, df=1, p<0.001), and eight patients (7%) and one healthy participant (1%) had anxiety disorder (χ
2=5.35, df=1, p=0.036).
There were 47 patients with PD (42%) who had no scorable symptoms on the NPI. None of the patients had delusions or disinhibition. Regarding the presence of NPS, 22 patients (20%) had one, 14 patients (13%) had two, and 28 (25%) had three or more. The most often endorsed symptoms were anxiety (40 patients; 63%), depression (38 patients; 59%), and apathy (37 patients; 58%), as well as agitation (eight patients; 13%) and irritability (six patients; 9%) (
Table 2). The highest mean scores were found for apathy, depression, and anxiety. When severity scores were calculated only for those patients in whom the symptom was present, the highest scores were obtained for apathy, anxiety, and depression (
Table 2). These prevalence estimates were based on the proportion of patients with non-zero scores. However, since scores of 1 or 2 on the NPI items may be clinically trivial, the analysis also involved those with a score of ≥4 on each NPI item, as previously described.
24 There were 27 patients (24%) who had a score of ≥4 on at least one of the NPI items; among these, the most common items were anxiety (13 patients; 12%), apathy (13 patients; 12%), and depression (12 patients; 11%) (
Table 2).
A total of 23 patients with PD (20.7%) had REM sleep behavior disorder (RBD) according to the recommended RBD-SQ cut-off score (4/5).
18 None of the PD patients with RBD symptoms were treated with antidepressant medications. Patients with PD and RBD endorsed several psychiatric symptoms more commonly compared with those without RBD (
Table 3). Patients with RBD did not differ from PD patients without RBD in terms of nighttime [F
(1,111)=0.32, p=0.574] or daytime sleepiness [F
(1,111)=1.350, p=0.248]. Male patients suffered from daytime sleepiness more frequently than did female patients (χ
2=10.50, df=1, p=0.027).
Psychiatric Manifestations in Patients With Early PD With Versus Without MCI
It has been previously shown that 27 of our patients with PD (24%) fulfilled the MDS Task Force level 2 criteria for MCI in PD, when the z cut-off scores were settled on –1.5 SD.
21 Patients with MCI were more frequently depressed (χ
2=5.00, df=1, p=0.033) and anxious (χ
2=6.71, df=1, p=0.021) compared with patients with PD without MCI (PD-nonMCI). PD-MCI patients also had higher SCOPA-S overall [F
(1,111)=5.73, p=0.018] and SCOPA-NS [F
(1,111)=3.01, p=0.086] scores compared with PD-nonMCI patients. A higher number of symptoms on the NPI [F
(1,111)=6.76, p=0.011], higher NPI scores for depression [F
(1,111)=3.91, p=0.005] and sleep disturbances [F
(1,111)=4.59, p=0.034], as well as higher scores for depression distress [F
(1,111)=5.58, p=0.020] and sleep distress [F
(1,111)=4.04, p=0.046] were found in PD-MCI patients versus patients without MCI. However, no differences were seen for the AS, SCOPA-DS, and SCOPA-PC scores between these two groups of patients.
PD patients with MCI demonstrated a higher grade of altered dreams [F(1,111)=0.69, p=0.019; as assessed using the SCOPA-PC], sleep fragmentation [F(1,111)=4.50, p=0.04; as assessed with the SCOPA-S], and movements during sleep that woke the patient [item 7 on the RBD-SQ], compared with patients with PD without MCI. There were 16 patients with PD-MCI (24%) and 17 patients with PD without MCI (26%) who had RBD (χ2=0.59, df=1, p=0.428).
Psychiatric Manifestations in Drug-Naive Versus Treated Patients With Early PD
A total of 40 patients with PD (36%) were drug naive at the time of assessment. Among treated patients, 19 (17%) and 26 (23%) were receiving dopamine agonists and levodopa monotherapy, respectively, and the remaining 26 patients (24%) were treated with the combination of levodopa and dopamine agonists. Compared with drug-naive patients with PD, treated patients had higher scores on the total SCOPA-S [F(1,111)=4.07, p=0.046] and more vivid dreams on the RBD-SQ [F(1,111)=5.57, p=0.02]. Treated patients did not differ in the frequency of MCI (χ2=0.22, df=1, p=0.815) or in the mean ACE-R score [F(1,111)=2.41, p=0.123] compared with drug-naive patients.
Discussion
The main finding of our study was that, in addition to significantly higher scores on all psychiatric scales and significantly lower scores on cognitive measures (
Table 1), patients with PD in the initial motor stage of the disease compared to the matched healthy control subjects experienced more frequently apathy (33% versus 4%), depression (22% versus 5%) and anxiety (7% versus 1%). A novel observation was that both nighttime and daytime sleepiness, combined with apathy, were predictors for participants’ separation between PD and healthy controls. Apathy and daytime sleepiness can occur in patients with de novo PD before treatment initiation
25,26 and may help to distinguish persons with early PD from healthy controls. Patients with PD also had higher scores on the NPI items for depression, anxiety, apathy, and sleep disorders compared with healthy subjects. Among patients, a score of ≥4 was most frequently obtained on the NPI items for apathy, anxiety, and depression (12%, 12%, and 11% of patients, respectively). The distribution of NPS in our group was comparable to that reported in previous studies of patients with de novo PD.
3,27–30 Weintraub et al.
3 failed to find an association between severity of depression, anxiety, apathy, and cognitive performance. In our study, apathy, but not depression, appeared to be a significant predictor for the categorization of subjects into the PD group. However, depression remained frequent in the early stage of PD (22% of patients), in accordance with a previously estimated prevalence range (14%−45.7%).
3,31 Apathy affected 33% of our patients, which was in line with published estimates in drug-naive patients with PD, which ranged from 14.3%
27 to 33.4%,
32 with a further increase for coexisting apathy and depression from 22.9%
27 to 38.6%.
32Besides depression, apathy, and anxiety, sleep disturbances were also among the most common NPS in both drug-naive
4,31,33 and treated patients with early PD.
34,35 Clinically relevant daytime sleepiness, as the only gender-specific feature in patients with PD, was more prevalent among men than women. Diagnosis of RBD was established in 20% of our patients. This finding was substantially lower compared with other studies, in which RBD was diagnosed in almost one-half of the affected patients in a larger cohort with the use of the RBD-SQ
36 and in one-third of drug-naive patients with PD based on a polysomnography examination.
37 Mollenhauer et al.
4 found no difference on the RBD-SQ between untreated patients with de novo PD and healthy controls. However, such estimates should be analyzed with caution because the RBD-SQ (a self-reported questionnaire, which did not require input from the patients’ bed partner) might be insufficient compared with the more formal polysomnographic investigation.
4 In addition, patients receiving DRT exhibited different types of abnormal motor/nonmotor sleep manifestations that might be misinterpreted as RBD in clinical interviews.
38 Our patients with PD and RBD had more psychiatric comorbidities, including delusions and increased sexual preoccupation, compared with patients without RBD, similarly to the previous studies that recognized the associations between RBD and visual hallucinations as well as other psychiatric comorbidities.
36,38 Like Gjerstad et al.,
39 we failed to find a relationship between depression and RBD in our group, as observed in a recent study.
36 In addition, we did not confirm an association between daytime sleepiness and DRT in patients with early PD.
33 Compared with the drug-naive patients in our study, our treated patients had higher scores for overall sleep problems (SCOPA-S) and vivid dreams, suggesting a role of DRT even in early disruption of sleep continuity. Higher levodopa-equivalent dose intake before sleep was related to higher SCOPA-NS scores, greater awakening after the sleep onset, and the REM sleep deprivation, which might influence RBD in early-stage PD.
6 By contrast, Stefansdottir et al.
40 reported that restlessness, nighttime “off” periods, and sensory complaints were improved in drug-naive patients with PD at 1 year after initiation of DTR. In patients with advanced PD, DRT facilitated sleep, possibly through alleviating nocturnal motor fluctuations.
6 Finally, we did not observe any difference in depression, apathy, and anxiety scores between drug-naive and treated PD patients in the initial Hoehn and Yahr stage. No improvement in depressive symptoms was obtained after 3 months of DRT in early PD,
33 whereas the results on such efficacy after one year of treatment were inconsistent.
41Another important finding was that cognitive deterioration might be associated with a higher prevalence of depression and anxiety, but not apathy, in patients with early PD. Apathy was associated with frontal executive dysfunction in patients with early PD, and neuropsychological and imaging evidence implicated that apathy might be related to dysfunction of specific neural networks involved in cognitive inhibitory control.
42 Although apathy has been related to poor performance on some of the visuospatial and constructional tests,
27,42 a regression model rejected a link between apathy and cognitive impairment in the early stage drug-naive PD patients.
27 In addition, our PD-MCI patients had predominantly impaired nighttime sleep, with more complaints about altered dreams, movements, and awakenings during sleep compared with PD patients without cognitive impairment. Altered dreams were not related to the presence of RBD, implicating that altered dreams might be related to a cognitive impairment itself, and not the RBD phenomena, in the initial motor stage of PD.
5The main limitations of present study include its cross-sectional design, the relatively small number of drug-naive patients, the tertiary hospital setting, and consequent potential enrollment bias. However, we recruited a homogeneous group of patients in the initial motor stage of PD and compared them with age- and sex-matched healthy control participants, all tested with an extensive battery of motor, neuropsychological, and neuropsychiatric scales. Furthermore, we used the formal MDS Task Force level 2 criteria for diagnosing MCI.
21 Another advantage might be the exclusion of patients with known PD-related genetic mutations, in order to preclude cognitive and behavioral specificities related to monogenic PD.
43In conclusion, the results of this study confirm a high burden of NPS even in the initial motor stage of PD, with the combination of apathy, nighttime and daytime sleep problems being predictive for participants' separation into groups of patients with PD and healthy subjects. Cognitive deterioration might influence the frequency of NPS: MCI in PD was associated with a higher prevalence of depression, anxiety, and sleep impairment compared with cognitively preserved patients. The results of our study also implicate that the core psychiatric symptoms were not associated with DRT, highlighting a necessity for better understanding of their pathophysiological mechanisms. Finally, as Weintraub et al.
3 suggested, active screening for “a range of highly prevalent NPS” in early PD might be useful for planning an optimal therapeutic approach.