Abstract
INTRODUCTION
Mood and behavioral disorders are now considered among the main symptoms of Parkinson’s disease (PD) [1–4]. The prevalence of depression ranges from 20 to 40% [5, 6]. Anxiety and apathy, which are commonly associated with depression, can be also isolated. Their prevalence varies from 8 to 25% for anxiety [5, 6] and from 17 to 50% for apathy [5, 6]. Others disorders have been well described such as sleep disorders with inability to fall and/or stay asleep, disturbed motor activity during sleep, and excessive daytime somnolence [7], distressing fatigue, [8], hypomania and mania [9], psychosis (with or without associated dementia) [10, 11].
More recently, Impulse Control and Repetitive Behavior Disorders (ICRBs) have been described [12–14]. ICRBs are a range of neuropsychiatric disorders including impulse control disorders (ICD) such as pathological gambling, hypersexuality, binge eating, compulsive shopping, and repetitive behaviors, known as hobbyism, punding, aimless walking or driving behavior, and compulsive use of dopaminergic medication, known as dopamine dysregulation syndrome (DDS). ICRBs are frequent with a prevalence from 15% to 35% [13–17].
Behavioral symptoms often fluctuate according to motor fluctuations with depressive symptoms, anxiety, apathy and fatigue during “off” periods and, on the contrary, the occurrence of ICRBs, hyperactivity and hypomania during “on” periods, which show a strong link between a change of dopamine transmission and non-motor fluctuations (NMF) [18–19].
Behavioral disorders negatively influence the quality of life (Qol) of patients with PD. Cognitive impairment, sleep disorders and fatigue are related to poor quality of life in PD [8, 21]. Although the role of depression and apathy in QoL in PD and caregivers is well known [22–28], little is known about the impact of ICRBs and non-motor fluctuations.
The “Ardouin Scale of Behavior in Parkinson’s Disease” (ASBPD) is a recently validated scale specifically designed to detect and quantify (none to severe intensity) all mood and behavioral disorders in PD [29, 30]. The ASBPD classifies them as a function of the dopaminergic status of the patients, i.e. either hypodopaminergic (such as apathy, depression, anxiety, etc.) or hyperdopaminergic, such as ICD or addiction behaviors (pathological gambling, hypersexuality or compulsive shopping, etc.) or NM fluctuations.. The advantage of this scale is that the same tool is used to evaluate all mood and behavioral symptoms and to follow-up changes during the progression of the disease and as a function of changes made to dopaminergic treatment.
The aim of this study was to assess the impact of behavioral disorders using the ASBPD on QoL measured with the PDQ-39, a specific QoL scale widely used in PD. We hypothesized that behavioral symptoms and more specifically ICDs and related disorders could negatively influence QoL in PD.
MATERIALS AND METHODS
Participants
Consecutive patients with PD, older than 30 years at onset of PD and diagnosed according to UK PD Brain Bank criteria were selected by a neurologist competent in movement disorders [31, 32]. Exclusion criteria were a Hoehn & Yahr stage > IV in “on” state, dementia according to the Diagnostic and Statistical Manual of Mental Disorders IV (DSM IV), and the inability to read or understand questionnaires. Patients suffering from an atypical Parkinson syndrome were not included. 136 patients were recruited at 10 French centers and evaluated during one session.
The study was approved by the local Ethics Committee and was performed in accordance with the 1975 Helsinki declaration. All patients gave their informed consent prior to their inclusion in the study.
Assessments
In addition to sociodemographic data, the following measures were completed. Severity of PD motor symptoms were assessed using the Unified Parkinson’s Disease Rating Scale (UPDRS) in ON stage [33]. Overall disease severity was graded according to the modified Hoehn and Yahr scale [34]. Levodopa and dopamine agonist doses were recorded separately and dopamine agonist doses were expressed as levodopa equivalent doses, as described elsewhere [35].
Neuropsychiatric symptoms were assessed using a new international validated scale ASBPD [30]. This scale consists of 21 items grouped in 3 parts: Hypodopaminergic symptoms (Part I), non-motor fluctuations (Part II) and hyperdopaminergic behaviors (Part III). Part I successively evaluates depressed mood, anxiety, irritability and aggressiveness, hyperemotionality, and apathy. Part II evaluates the psychological state (mood and behavior) associated with the motor symptoms in the OFF or ON states among fluctuating patients. Part III assesses the presence and the severity of behavioral disorders induced by dopaminergic treatment, including nocturnal hyperactivity, diurnal somnolence, eating behavior, creativity, hobbyism, punding, risk-taking behavior, compulsive shopping, pathologic gambling, hypersexuality, dopaminergic addiction, and excess motivation.
Each item is rated on a five point scale (severe disorder- 4; marked disorder- 3; moderate disorder- 2; mild disorder- 1; absence of disorder- 0), by taking into account the severity and the frequency of the disorder and its impact. This semi-structured interview is completed by a neurologist, psychiatrist or a neuropsychologist familiar with PD.
Quality of life was assessed with the 39-item Parkinson’s Disease Questionnaire (PDQ-39) which produces eight subscores (mobility, activities of daily living, emotional well-being, stigma, social support, cognition, communication and bodily discomfort) and one summary index (the PDQ-39-SI) [36]. Higher scores reflect poorer Qol (range 0 to 100).
All the assessments were performed during a single visit.
Statistical analysis
Statistical analysis was performed using Stata software (version 13; Stata-Corp, College Station, Tex., US). Quantitative variables are expressed as means and standard deviations (SD) or as medians and interquartile ranges according to statistical distribution (assumption of normality studied using the Shapiro-Wilk test). Comparisons of patient’s quality of life scores between independent groups (gender, hypodopaminergic disorders yes/no, non-motor fluctuations yes/no, hyperdopaminergic behaviors yes/no) were performed using Student t-test or Mann-Whitney test if assumptions of t-test not met: (i) normality and (ii) homoscedasticity studied using the Fisher-Snedecor test. The relations between PDQ39 and dimensions of ASBPD were studied using correlation coefficients (Pearson or Spearman according to statistical distribution, noted r), considering an appropriate adjustment of type I error (Sidak’s inflation). Multivariate analyses (linear regression models) were performed according to univariate results and clinical relevance with adjustment on gender, age, treatment and UPDRS scores. Results were expressed as regression coefficients noted β and 95% confidence interval and were represented using forest-plots. The normality of residuals obtained for each model was studied as described previously. When appropriate, a logarithmic transformation was proposed to achieve normality of dependent variables.
RESULTS
Patient characteristics
136 PD patients (84% men) were included in this study. Mean age (±SD) was 61.3±8.5 years and average duration of disease was 8.8±5.4 years (Table 1). Median Hoehn & Yahr stage was 2 (IQ range: 1.5 –2.5). Patients had no significant cognitive deficits, as verified by the UPDRS part 1, item 1 score (0.46±0.55) (Table 1).
The mean UPDRS scores were as follows: UPDRS I (1.9±1.7), UPDRS II (7.1±4.8), UPDRS III (12.7±8.8) and UPDRS IV (5.0±3.6). 72% (n = 90) of patients presented motor fluctuations. The predominant treatment was the combination of levodopa with a dopamine agonist (n = 81, 59.6%) while 23.5% (n = 32) also received antidepressants, 13.2% (n = 18) anxiolytics and 5.9% hypnotics (n = 8) (Table 1).
Mean scores of hypodopaminergic disorders, non-motor fluctuations, and hyperdopaminergic behaviors using the ASBPD are given in the Table 2. 80.7% (n = 110) of patients had at least one hypodopaminergic disorder, 50.7% (n = 69) of patients exhibited non-motor fluctuations. 89.0% (n = 121) presented at least one hyperdopaminergic behavior (19.9% (n = 27) had 1 behavior, 16.2% (n = 22) had 2 behaviors; 14.0 % (n = 19) had 3 behaviors and 38.97% (n = 53) displayed more than 3 disorders)) (Table 3). Data on the PDQ39-SI and the eight subscores were described in the Supplementary Figure 1. 16.2% of patients (n = 22) had missing items in the PDQ-39 questionnaire and were excluded from the analyses. Of the PDQ 39 dimensions, the QoL of patients was best for social support (14.1±20.4) while higher values were recorded for bodily discomfort (45.4±20.0). Some sex-differences in PDQ 39 dimensions were noted: Scores of mobility, emotional well-being and bodily discomfort were significantly higher in women (p < 0.05) whereas men had significant higher scores for activities of daily living (p = 0.01). No difference for stigma, social support, cognitive impairment and communication was detected between men and women.
Correlations analyses
Clinical and neurological characteristics of patients (i.e. age, delay from diagnosis, Hoehn & Yahr, and Schwab and England score, UPDRS I, UPDRS II and UPDRS IV) were significantly correlated with the subscores of PDQ 39 and the PDQ39-SI subscore. (Table 4).
Patients who had at least one hypodopaminergic symptom presented a worse Qol for all the dimensions of PDQ-39 (except mobility and activities of daily living) compared with patients without hypodopaminergic symptoms (Table 5). Similarly, patients who had at least one or two NMF symptoms exhibited a worse Qol for all the dimensions of PDQ-39 compared with patients without NMF. On the contrary, having hyperdopaminergic behaviors did not modify Qol (except activities of daily living) (Table 5).
Univariate analyses of relationships between PDQ39 and items of ASBPD showed that PDQ39-SI were correlated with all items of hypodopaminergic disorders (except anxiety), NMF (in ON and OFF states), and only 4/14 items of hyperdopaminergic disorders (psychotic symptoms, diurnal somnolence, hobbyism and compulsive shopping) (Table 6). Hypodopaminergic symptoms correlated with the emotional well-being dimension and to a lesser extent with the cognitive impairment dimension of PDQ39 (except anxiety). Apathy was also correlated with activities of daily living and communication dimensions (Table 6). NMF in ON and OFF states were significantly correlated with mobility, activities of daily living, stigma and bodily discomfort dimensions of PDQ 39. NMF in OFF state were also correlated with emotional well-being and communication dimensions. Concerning hyperdopaminergic behaviors, hobbyism was slightly negatively correlated with mobility and emotional well-being, and excess in motivation with mobility. Psychotic symptoms and diurnal somnolence were correlated with activities of daily living and cognitive impairment. Compulsive shopping was slightly correlated with emotional well-being and social support.
PDQ39-SI were significantly correlated with the severity of symptoms and with the number of disorders for hypodopaminergic disorders (r = 0.38 and r = 0.41 respectively, p < 0.05), for NMF (r = 0.44 and r = 0.44, respectively) but not for hyperdopaminergic behaviors (Table 6).
Multivariate regression of relationship between PDQ39 and items of the ASBPD (adjusted for gender, age and UPDRS scores) showed slight differences with the results observed with univariate analyses (Supplementary Figure 2). Mobility was well correlated with the item “excess motivation” (β= –6.8 [–10.9; –2.6]) and with NMF in OFF state (β= 3.5 [0.1; 7.0]). Activities of daily living were only slightly correlated with diurnal somnolence (β= 5.1 [0.0; 10.3]). Major factors which impacted “Emotional well-being” were compulsive shopping (β= 6.0 [1.4; 10.6]) and hobbyism (β= –5.2 [–8.4; –2.1]), and to a lesser extent depressive mood (β= 4.7 [1.2; 8.2]), anxiety (β= 3.8 [0.8; 6.8]) and NMF in OFF state (β= 3.8 [0.8; 6.8]). Stigma and social support were only correlated with irritability and aggressiveness, respectively β= 7.3 [0.4; 14.3] and β= 6.7 [0.9; 12.4]. Psychotic symptoms (β= 9.6 [4.0; 15.3]) was the main factor having a negative impact on cognitive impairment; diurnal somnolence (β= 4.9 [0.4; 9.3]) and eating behavior (β= 4.1 [0.4; 7.8]) were also correlated to this dimension. Bodily discomfort was correlated with NMF in ON state (β= 6.3 [0.1; 1.6]) and depressed mood (β= 4.5 [0.1; 8.9]). Communication was correlated with none of the items. Finally, PDQ-SI showed significant correlation with NMF in OFF state (β= 3.1 [0.6; 5.6]) and hyperemotionality (β= 3.1 [0.1; 6.0]).
DISCUSSION
Both motor and non-motor factors may affect the Qol of PD patients. Among non-motor factors, our findings demonstrate a strong link between the alteration of patients’ Qol and the presence of mood disorders such as depressed mood, anxiety, irritability/aggressiveness, hyperemotionality and apathy, as well as non-motor fluctuations (ON or OFF). However, this link was weaker between hyperdopaminergic behavioral disorders (especially ICD) and Qol.
Mood and behavioral disorders are usually assessed with validated scales who evaluate only one or a few domains of psychic disorders [37–40]. In this study mood and behavioral disorders were assessed using the recently new validated ASBPD, a scale specifically designed to detect and quantify the main behavioral symptoms in PD in a single instrument [30] which limits the bias linked to the homogeneity of the population studied. This scale classifies mood and behavioral disorders as a function of the dopaminergic status of the patients, i.e. either hypodopaminergic (such as apathy, depression, anxiety, etc.) or hyperdopaminergic, such as ICD or addiction behaviors (pathological gambling, hypersexuality or compulsive shopping, etc.), although other neurotransmitters can be linked to pathophysiology [30, 41–44]. This scale, in addition, evaluates psychic NM fluctuations in ON and OFF conditions.
Qol was assessed using the PDQ-39 self-questionnaire [36]. This scale has been formally validated in several languages [45] and has been used in a large number of studies. It was recommended as one of the most appropriate Qol instruments in PD [46, 47].
Several factors may impact QoL. As published previously and according to our results, age, mean disease duration, motor disability (bradykinesia, rigidity, tremor, balance and gait disorders) and motor complications (such as dyskinesia and motor fluctuations) have been correlated to a poorer Qol [48–53]. However, as already suggested by previous studies, Qol depends not only on motor symptoms and complications but also on the occurrence of non-motor symptons (NMS) which could be more troublesome and disabling than motor disturbances, and contribute greatly to patients’ institutionalization in advanced forms of PD [54–56].
Among hypodopaminergic NMS, depressed mood, apathy, irritability / aggressiveness and hype-remotionality significantly affected Qol, as the scores of the PDQ39 SI and most of the PDQ39 items were significantly higher than those observed in patients without these symptoms. We also found a significant correlation between the number of symptoms and their severity, and PDQ39 scores (PDQ39 SI, emotional well-being, stigma, social support, cognitive impairment, communication and bodily discomfort), suggesting that these symptoms are major predictors of a poor quality of life. These results are in accordance with previous studies [23–35, 57–59]. For example, it has been demonstrated that depression could explain 60% of QoL deterioration in PD patients [22, 23]. The multivariate regression of relationships between PDQ39 and items of ASBPD (adjusted for gender, age and UPDRS scores) permitted determining that the main mood symptoms which influence Qol were depressed mood (on emotional well-being and bodily discomfort dimensions), anxiety (on emotional well-being), irritability, aggressiveness (on stigma and social support) and hyperemotionality (on PDQ-SI).
Moreover, as the disease progresses, most PD patients treated with dopaminergic drugs can develop NMF. NMF are non-motor symptoms that vary according to dopaminergic tone in a manner similar to motor fluctuations. NMF are relatively frequent: Around 50% of our patients presented NMFs. So mood and the behavioral state of fluctuating patients alternate between the “On” period (with the occurrence of ICD, hypomania, hyperactivity) and the “OFF” period (with the occurrence of depressed mood, anxiety, apathy). Patients with NMF reported a greater degree of disability than for motor fluctuations [19] and a deteriorated HRQol [60]. We demonstrated here that NMFs in OFF state are well correlated with most of the subscores of the PDQ39 and with a poorer Qol. NMFs in “ON state” were also correlated to deteriorated Qol but to a lesser extent (only with the bodily discomfort item).
Contrary to mood symptoms and NMF, the impact of behavioral symptoms mainly influenced by dopaminergic medication on Qol and disability has been poorly assessed. We showed here that there were only weak correlations between ICDs (compulsive buying, hobbyism, excess motivation, eating behavior) and some items of the Qol (social well-being, mobility, cognitive impairment and SI index), suggesting that they did not have a major impact on Qol. However, the multivariate regression between PDQ39 and items of ASBPD (adjusted for gender, age and UPDRS scores) showed that compulsive shopping has a negative impact on the emotional well-being dimension, likewise for eating behavior on the cognitive impairment dimension of the PDQ 39. Interestingly, the items “excess in motivation” and “hobbyism” had a positive impact on the mobility and emotional well-being dimensions respectively of the PDQ 39. This could be explained by the pleasure felt by the patient in performing their leisure activities. Recently, two studies reported a negative impact of ICD on disability and health Qol, especially the social well-being item of the PDQ39 emphasizing the importance of their diagnosis and management [15, 17]. These results are slightly out of line with ours as we did not demonstrate a major impact of ICD although we also showed a significant link with some ICDs and PQQ39, and specifically the emotional well-being dimension. These differences could be explained by the population studied. In their studies, the authors cited above compared QoL between patients with and without ICDs. In our patients with hyperdopaminergic symptoms, most of them (85%) had only mild symptoms which probably had no influence on QoL, and may even have had a positive impact on QoL, such as the “hobbyism” item in our series of patients (Supplementary Table 1). Only 41% of patients had hyperdopaminergic behavioral disorders with an inability to control their disorders which can be considered as pathological and only few had marked (21%) or severe (2%) symptoms (Supplementary Table 1). It is possible that a major deterioration of the QoL could be observed in patients with marked or severe ICD with deleterious social and familial consequences, which in our study represents only a small percentage of patients. Other studies on patients with severe ICDs are warranted to confirm this hypothesis.
Concerning the other behavioral disorders, psychotic symptoms (including hallucinations, delusions, or minor symptoms such as sense of presence, visual illusions, or passage hallucinations) were well correlated with the “cognitive impairment” dimension of the PDQ 39. The link between psychosis and dementia was already well documented, as cognitive impairment and dementia have been the most frequently reported risk factors for psychotic symptoms in PD [61]. In our study, although we excluded patients with dementia, this link can be explained by the presence of mild or moderate cognitive impairment in patients with a duration of disease close to 10 years. Excessive diurnal somnolence was also significantly correlated with the “activities of daily living” dimension of the PDQ 39, as could be expected [56].
NMS also include dysautonomic, neuropsychiatric and sensory/pain symptoms. This work only focused on mood and behavioral NMS. However, it is important to note that other factors, not explored in this study, could also contribute to a poor Qol and should be taken into account. Studies have shown, for example, a relationship between life-quality worsening and fatigue [53, 62], pain [25], cognitive status [23, 64], fall and fear of falling [65] and dysautonomic symptons [24, 58].
In conclusion, we found that PD patients’ Qol is not only affected by motor but also by non-motor symptoms, especially by the presence of hypo-dopaminergic symptoms and NMF. Some behavioral disorders also negatively affected patient’s Qol to a lesser extent, perhaps due to the fact that our patients did not present severe forms of ICD, which in fact reflected a non-selected PD population. In general, patients did not really express complaint about their ICDs (except when severe) due to the hedonistic effect of ICD in patients. Indeed, patients feel pleasure in performing their activities. However, it would be very interesting to explore the impact of behavioral ICDs on Qol of caregivers (spouse, family) as they express great disarray when confronted with them. Indeed, ICD may change the habits and the characters of the patients that could consequently alter the Qol of their caregivers by causing severe psychological problems (such as depression, anxiety, fatigue, stress...) or social troubles (such as lack of social contacts, reducing or giving up employment ...) or financial burden [66], while patients themselves have no impairment in quality of life.
CONFLICTS OF INTEREST
The authors declare that they have no conflict of interest.
Footnotes
ACKNOWLEDGMENTS
This work was supported by a PHRC (PHRC AOI 2008 DURIF) and by Novartis Pharma.