ON-OFF Effects of Dopaminergic Therapy on Psychiatric Symptoms in Parkinson’s Disease

Parkinson’s disease (PD) is a neurodegenerative disease with prominent motor symptoms of tremor, bradykinesia, and rigidity affecting 3% of the population over age 65.¹ Motor symptoms of PD appear when degeneration of dopamine-producing cells of the substantia nigra (SN) is sufficient to seriously restrict dopamine supply to the dorsal aspect of the striatum (DS).² As would be expected given this pathophysiology, motor symptoms are dramatically improved by dopamine replacement medications, such as L-3,4-dihydroxyphenylalanine (L-Dopa) or dopamine receptor agonists. Traditionally, treatment of PD has centered on improving these motor aspects of the disease.

Nonmotor symptoms of Parkinson’s disease (PD) are an increasingly recognized cause of significant disability, emotional distress, caregiver burden, and diminished quality of life. They also are more predictive than motor symptoms of the need for hospitalization and therefore have important economic implications for the health care system.³ Nonmotor features of PD include psychiatric, cognitive, sleep, and autonomic symptoms. The pathophysiology of these symptoms and how they are affected by dopaminergic medications remain less clear than the motor abnormalities in PD. Here we focus on the psychiatric manifestations.

Anxiety, Depression, Apathy in PD

Symptoms of anxiety are common in PD. Significant anxiety occurs in 20% to 50% of PD patients and can manifest as any subtype of anxiety disorder, most commonly generalized anxiety, panic, or social anxiety disorders.⁴ Anxiety in PD appears to be related, at least in part, to the disease process itself, not simply as a response to the subjective experience of disease.^4,5 Anxiety frequently predates motor symptoms and hence diagnosis of PD.⁴ Epidemiological studies reveal that PD patients are at higher risk of developing anxiety than the general population even before the development of motor symptoms, ^6,7 and PD patients have higher rates of anxiety than patients with other disorders that are equally impairing.⁵ Patients with PD also have higher rates of reactive anxiety and anxiety disorders that meet the Diagnostic and Statistical Manual of Mental Disorders (DSM) criteria for generalized anxiety disorder, panic disorder, and social phobia.⁴ The effect of dopaminergic therapy on anxiety, however, is unclear. Some studies indicate that dopaminergic treatments can improve anxiety,^8,9 whereas others suggest that it exacerbates the symptom.^10–12

Similar to anxiety, depression is a common nonmotor feature of PD that often predates motor symptoms.^13,14 About 40%−50% of PD patients have clinically significant depression, which is thought to account for up to 60% of the decline in quality of life experienced by individuals with PD.¹⁵ Research on the effect of L-Dopa on depression has yielded mixed results. Maricle and colleagues⁸ demonstrated a rapid mood-enhancing effect of L-Dopa in a double-blinded randomized controlled trial, though most other studies suggest no effect or even exacerbation of depression owing to L-Dopa.^16–19 There is some evidence for the antidepressant activity of pramipexole in PD ^20,21 including a 12-week, randomized, placebo-controlled, double-blinded trial by Barone and colleagues, ²² although the efficacy does not appear to be as great as for tricyclic antidepressant therapy. ²³ There is also some data supporting the use of pramipexole in treating bipolar depression.²⁴

Apathy is a behavioral syndrome experienced by approximately 60% of PD patients. It commonly coexists with depression and cognitive decline,²⁵ and is associated with greater severity of illness and functional deficits. It is described as a reduction of goal-directed behavior from lack of feeling, interest, emotional reactivity, and motivation. ²⁶ Although it shares certain characteristics with depression, apathy is essentially a disorder of motivation rather than mood. ²⁷ Despite the high level of comorbidity between apathy, depression, and cognitive decline, apathy does appear to be an independent symptom with a subset of PD patient demonstrating apathy without clinically significant depression or cognitive decline.^28,29

Current Study

From reviewing the literature, the effect of dopaminergic therapy on psychiatric symptoms in PD is unclear. There is a paucity of studies and most are cross-sectional, observational studies that do not control for stage of disease or dosage of medication. The current study is a within-subject investigation of the effect of dopaminergic medication on psychiatric symptoms in PD, using standardized, validated measures of anxiety, depression, and apathy. This within-subject approach accounts for stage of disease and dosage of medication. Further, our findings in PD patients were compared with age- and education-matched controls that also completed standardized measures on two occasions with their data analyzed to correspond to the ON-OFF order of the PD patient to whom they were matched. In this way, we also controlled for effects of repeated testing and assessed retest reliability.

Methods

Participants

Thirty-six PD patients with an average Hoen and Yahr staging of 1.92 (SEM 0.10) participated in the study. All patients were evaluated in a general neurology clinic, were diagnosed by a licensed neurologist, and met 1) the core assessment program for surgical interventional therapy criteria for the diagnosis of idiopathic PD^30,31 and 2) the UK Brain Bank criteria for the diagnosis of Parkinson's disease.³² None were surgical candidates. Thirty-six age- and education-matched healthy control participants were also tested. Patients and controls diagnosed with dementia or mild cognitive impairment, abusing alcohol, prescription or street drugs, or taking medications such as donepezil, rivastigmine, galantamine, or memantine were excluded from participation. Further, if participants reported a change in function related to cognitive symptoms, performed below 100 on the Adult National Reading Test (ANART), or could not successfully draw a clock or copy a cube, they were excluded from the study. The data from three PD patients and six controls were not included in the analyses because they failed to complete all measures in both the ON and OFF sessions. This study was approved by the Ethics Review Board of the Sudbury Regional Hospital and all patients provided informed consent prior to testing according to the Declaration of Helsinki.³³

Severity and presence of disease were assessed for all patients both on and off dopaminergic medication using the motor subscale of the Unified Parkinson’s Disease Rating Scale (UPDRS) by a movement disorder neurologist. Motor symptoms were significantly improved on relative to off dopaminergic therapy [t (32)=8.05, p <0.001]. A screening neurological examination was performed on control participants, and none manifested signs of PD. All patients and no controls were treated with L-Dopa. Thirteen PD patients were also treated with pramipexole, a dopamine agonist medication. This medication constituted adjunctive therapy only, on average accounting for 25% of the daily L-Dopa medication equivalent. Mean group demographic information, UPDRS scores on and off medication, number of participants on antidepressants/anxiolytics, and daily doses of dopamine-replacement therapy in L-Dopa equivalents are presented in Table 1. Calculation of daily L-Dopa equivalent dose for each patient was based on theoretical equivalence to L-Dopa³⁴ as follows: L-Dopa dose+L-Dopa dose×1/3 if on entacapone+ bromocriptine (mg) × 10+cabergoline or pramipexole (mg)×67+ropinirole (mg)×20+pergolide (mg)×100+apomorphine (mg)×8. There were no statistically significant demographic differences between PD patients and controls.

Experimental Design and Procedure

On consecutive days, participants completed the Beck Anxiety Inventory (BAI),³⁵ Beck Depression Inventory II (BDI-II),³⁶ and Starkstein Apathy Scale (SAS).³⁷ The BAI is a 21-item self-report questionnaire that focuses on generalized anxiety disorder and panic disorder. The BAI was developed to minimize overlap with depression rating scales,³⁵ emphasizes somatic manifestations of anxiety,³⁸ and is commonly used in PD research.³⁹ The BDI-II has been shown to have good agreement with other valid and reliable depression rating scales and test-retest reliability is high (Pearson r=0.93). It also demonstrates high internal consistency (α=0.91).³⁶ The SAS has been shown to be the most valid and reliable measure of apathy in PD patients.³⁹

All patients performed these measures once on and once off dopamine-replacement therapy. The ON-OFF order was counterbalanced across participants, with 17 patients first performing the measures on and 16 first off treatment. During ON testing sessions, PD patients took their dopamine-replacement medication as prescribed by their treating neurologist. During OFF testing sessions, PD patients abstained from dopamine-replacement therapy for a minimum of 12 and a maximum of 18 hours prior to testing. Control participants also completed these measures on 2 consecutive days. Although control participants did not take dopamine-replacement medication during either testing session, their data were analyzed to parallel the ON-OFF order of the PD patient to whom they were matched prior to testing, based on age and education. The data of 14 controls were analyzed in the ON-OFF order, whereas the data of 16 controls were analyzed in the OFF-ON order. In this way, we controlled for order, fatigue, and possible practice effects.

Results

Table 2 presents the median and mean scores, as well as the standard error about the mean for PD patients and control participants on the BAI, BDI-II, and SAS in both the ON and OFF sessions.

Anxiety

A 2×2×2 mixed analysis of variance (ANOVA) was performed on the BAI scores with group (PD versus control) and order (ON-OFF versus OFF-ON) as the between-subject factors, and session (ON versus OFF) as the within-subject variable There was a main effect of group, F (1, 60)=14.01, mean squared error (MSe)=105.265, p <0.001 because of higher anxiety scores in PD patients than controls. There were also significant effects of session, F (1, 60)=8.086, MSe=15.558, p <0.025, reflecting higher BAI scores in ON relative to OFF sessions, and of session×order, F (1, 60)=7.02, MSe=15.558, p <0.025, resulting from highest anxiety scores in the ON session when it was performed first. The main effect of order, F (1, 60)=1.12, MSe=105.265, p >0.250 and the group×session interaction, F < 1, were not significant.

Planned, 2×2 ANOVAs were performed for PD and control participants separately, with order (ON-OFF versus OFF-ON) as the between-subject factor, and session (ON versus OFF) as the within-subject variable. For the PD group, the main effect of session was significant, F (1, 31)=5.29, MSe=20.178, p <0.050, reflecting higher anxiety scores on relative to off medication. The session×order interaction was significant, F (1, 31)=5.58, MSe=20.178, p <0.050 because of highest anxiety scores in the ON session when this session was performed first. The main effect of order did not reach significance, F (1, 31)=1.18, MSe=20.178, p >0.250. For the control group, the main effects of order (F < 1) and session [F (1, 28)=2.78, MSe=10.685, p >0.115], and the order×session interaction [F (1, 28)=2.39 MSe=10.685, p >0.125] were not significant.

For PD patients, a multiple regression was performed to predict the ON-OFF BAI difference score, including ON-OFF UPDRS difference score, daily L-Dopa equivalent medication dosage (LED), and years of disease as regressors. This model did not significantly predict ON-OFF BAI difference score, F (3, 29)=2.11, MSe=40.659, p >0.120. Though ON-OFF UPDRS difference score contributed significantly to the ON-OFF BAI score, t=2.33, p <0.050, LED, t=0.93, p >0.350, and years of disease, t=0.99, p >0.325, did not. ON-OFF UPDRS difference score also correlated significantly with LED, r=–0.392, p <0.025, but did not correlate significantly with years of disease, r=–0.214, p >0.225.

Depression

A 2×2×2 mixed ANOVA was performed on the BDI-II scores with group (PD versus control) and order (ON-OFF versus OFF-ON) as the between-subject factors, and session (ON versus OFF) as the within-subject variable. The main effect of group was significant, F (1, 60)=15.52, MSe=47.546, p <0.001 because of to higher depression scores in PD patients than controls. The main effects of order and session, both F < 1, and the interactions of group×session, F (1, 60)=1.78, MSe=6.15, p >0.150, and order×session, F < 1, were all nonsignificant. 2×2 ANOVAs were performed on BDI scores for PD and control participants separately, with order (ON-OFF versus OFF-ON) as the between-subject factor and session (ON versus OFF) as the within-subject variable. For both the PD and control groups, no main effects or interactions were significant, all F <1.

Apathy

A 2×2×2 mixed ANOVA was performed on SAS scores with group (PD versus control) and order (ON-OFF versus OFF-ON) as the between-subject factors, and session (ON versus OFF) as the within-subject variable. The main effect of group was significant, F (1, 60)=5.41, MSe =60.640, p <0.025 because of higher apathy scores for PD patients than controls. The main effect of order did not reach significance, F (1, 60)=2.77, MSe=60.640, p >0.100. The main effect of session, and the group×session, and order×session interactions, were also not significant, all F <1.

2×2 ANOVAs were performed on apathy scores for PD and control groups separately, with order (ON-OFF versus OFF-ON) as the between-subject factor and session (ON versus OFF) as the within-subject variable. For the PD group, the main effect of order was significant, F (1, 31)=4.92, MSe=72.400, p <0.050 because of higher apathy scores for those who participated OFF before ON. The order×session interaction was marginally significant, F (1, 31)=3.043, MSe=5.567, p=0.091, owing to the fact that SAS scores were higher in all patients during their second session (i.e., SAS ON>OFF for patients who performed OFF-ON and OFF>ON for patients who performed ON-OFF). The main effect of session, F (1, 31)=2.87, MSe=5.567, p >0.100, was not significant. For the control group, the effects of order and session, both F < 1, and the order×session interaction, F (1, 28)=1.64, MSe =5.017, p >0.200, were not significant.

Discussion

In summary, PD patients had higher anxiety, depression, and apathy scores compared with age- and education-matched controls as assessed by the BAI, BDI-II, and SAS scales, respectively. The novel finding is that PD patients had significantly higher anxiety scores on relative to off dopaminergic therapy. Further, this increase in anxiety on dopaminergic medication was significantly correlated with the extent to which patients’ motor symptoms improved from the OFF to the ON state as assessed by the UPDRS. Duration of disease and absolute daily LED were not significantly predictive of the ON-OFF BAI difference score, though LED and ON-OFF UPDRS scores were significantly correlated. Depression and apathy were not affected by dopamine-replacement treatment in PD patients. Control participants did not evidence any test-retest differences on their anxiety, depression, or apathy scores ruling out the possibility that our findings owed only to effects of repeated testing.

Effect of Dopaminergic Medication on Anxiety in PD

We found that dopaminergic medication increased anxiety in PD. This effect is significant across a group of patients who ranged broadly at baseline in terms of disease duration (1–19 years), daily dosage of dopaminergic medication (250–1300 mg), motor symptom severity (UPDRS 1–39.5), as well as in terms of their anxiety (BAI 0–36/63), depression (BDI-II 0–25/63), and apathy (SAS 2–26/42) symptomatology. The measurements of interest were taken across consecutive days and within the same individuals, thus, controlling for effects of disease progression, dosage of medication, or motor symptom severity. A further strength of this study was the use of a standardized, validated, and reliable anxiety measure.⁴⁰

Previous studies have yielded inconsistent results with respect to how dopamine replacement affects anxiety in PD. Maricle and colleagues⁸ demonstrated a dose-dependent improvement in anxiety for dopaminergic therapy in a placebo-controlled double-blinded trial. Only eight PD patients were tested using a nonstandardized visual analog measure for anxiety, however. Funkiewiez and colleagues⁹ also reported improved anxiety with dopamine replacement. This study used a subsection of the Addiction Research Center Inventory (ARCI) to quantify anxiety. The ARCI is a true or false questionnaire designed to assess subjective experiences of drug intake and personality traits and is not a primary, validated measure of anxiety.

Our results are also at odds with observations that PD patients are prone to develop symptoms of panic during ‘off’ periods when motor symptoms are exacerbated¹⁰ and particularly during episodes of freezing (i.e., temporary inability to move).⁴¹ This could relate to the facts that not all symptoms of anxiety are equal, are as effectively screened by standardized tools, and arise from the same mechanisms, particularly at different stages of illness. The panic experienced during off periods, characterized by tachycardia, diaphoresis, and tremulousness, is quite a different phenomenon than generalized anxiety, which includes more prominent cognitive symptoms such as worrying and anticipating disaster. Given this pervasive clinical understanding of anxiety as a nonmotor off symptom, our results are important in cautioning clinicians, however, not to equate all symptoms of anxiety to under-treatment of PD.

The notion that dopaminergic medication increases symptoms of anxiety is consistent with what is known about physiologic effects of stimulants, whose mechanism of action is largely related to increases in dopaminergic activity.⁴² Increased anxiety is considered a very common side effect of stimulant medications such as methylphenidate (methylphenidate hydrochloride product monograph, 2012)⁴³ and amphetamines (Adderall XR product monograph, 2013)⁴⁴, especially at higher doses. It is also seen in illicit stimulants such as cocaine.⁴⁵ Further, our results are consistent with some studies that have explicitly investigated the effect of dopaminergic medication on anxiety in PD.^10–12 Vazquez and colleagues¹⁰found an association between duration of L-Dopa use and panic attacks in a series of 31 PD patients whereas Lang and colleagues¹² showed a worsening of anxiety symptoms with the addition of pergolide to the treatment regimen of late-stage PD patients.

Finally, the finding that anxiety is increased in patients on relative to off dopaminergic medication could be analogous to the observation that certain cognitive functions−learning chief among them−actually worsen with dopaminergic therapy^46,47 using ON-OFF experimental designs in PD. These cognitive effects are attributed to dopamine overdose in brain regions that are relatively dopamine replete compared with the significantly dopamine-restricted DS.^46,48 Recall that a primary pathological change in PD is the degeneration of dopamine-producing cells in the SN, which project nearly exclusively to the DS.² This DS dopamine depletion is the genesis of the motor symptoms, and dopaminergic medications dramatically alleviate these movement abnormalities. The dopamine-producing cells in the ventral tegmental area (VTA) are relatively spared in PD and consequently, the brain regions receiving dopamine from VTA are comparatively dopamine replete. These brain regions include the ventral portions of the striatum (VS), limbic, and prefrontal cortices. It is postulated that dopaminergic medication dosages aimed at relieving motor symptoms, and hence, titrated to the dopamine deficiency in DS, actually overdose VTA-innervated brain regions.^46,47 This account is consistent with the finding that difference in BAI ON-OFF scores was proportional to the change in motor symptoms as a function dopaminergic medication dosage.

Effect of Dopaminergic Medication on Depression and Apathy in PD

This study did not demonstrate any changes in depression or apathy scores ON relative to OFF dopaminergic treatment in PD. Studies of pharmacotherapy for depression often reveal that several weeks of treatment are required to achieve improvements, and longer-term effects of dopaminergic therapy on depression were not assessed here. Our results are, however, at odds with the findings of Maricle and colleagues⁸ that mood and apathy were improved with L-Dopa. Maricle and colleagues’ study lacked baseline information on mood state of their participants that could potentially illuminate reasons for differences between their findings and ours, however. Though there was a relatively broad range of BDI-II and SAS scores in our PD patients, the mean baseline depression and apathy levels were relatively low (i.e., BDI-II 8.78/63 and SAS 11.41/42).³⁶ Further, Maricle and colleagues used a nonstandardized tool to quantify depression potentially leading to divergent findings. Future studies should aim to test the effects of dopaminergic therapy on mood in PD patients specifically chosen to represent a wider range of baseline depression and apathy scores. Effects of these medications might have differential effects depending on baseline mood levels.

Conclusions

We found that dopaminergic treatment can increase symptoms of anxiety in PD. These results further clarify the complex causes of anxiety in PD and potentially account, at least in part, for the finding of increasing anxiety at later stages of PD.⁴⁹ Given the significant impact of nonmotor features of PD on quality of life,^50–52 it is important that clinicians balance the side effects of dopaminergic treatment with the benefits on motor function.