Website Edition: October/November 2012

Editor's Choice Article

Movement Disorders
DOI: 10.1002/mds.24978

Tricyclic antidepressants delay the need for dopaminergic therapy in early Parkinson's disease†‡
Katrina L. Paumier PhD1, Andrew D. Siderowf MD, MSCE2, Peggy Auinger MS3, David Oakes PhD3, Lalitha Madhavan MD, PhD1, Alberto J. Espay MD, MSc1,4, Fredy J. Revilla MD1,4, Timothy J. Collier PhD5, for the Parkinson Study Group Genetics Epidemiology Working Group

Article first published online: 3 MAY 2012

1 Department of Neurology, University of Cincinnati, Cincinnati, Ohio, USA
2 Department of Psychiatry, University of Pennsylvania, Philadelphia, Pennsylvania, USA
3 University of Rochester School of Medicine and Dentistry, Rochester, New York, USA
4 The Neuroscience Institute, Department of Neurology, Movement Disorders Center, University of Cincinnati, Cincinnati, Ohio, USA
5 Department of Translational Science and Molecular Medicine, Michigan State University, Grand Rapids, Michigan, USA

Funding agencies: This work was supported by the Parkinson's Disease Foundation's Advancing Parkinson's Treatments Innovations Grant to the Parkinson Study Group (PSG), the Davis Phinney Foundation for Parkinson's, the Morris K. Udall Center of Excellence for Parkinson's Disease Research at Michigan State University, and the University of Cincinnati (NS 058830) and the University of Pennsylvania Udall Center (NS 053488).

Relevant conflicts of interest/financial disclosures: All funding sources provided unrestricted support and had no role in the oversight or review of the research data or reporting. Full financial disclosures and author roles may be found in the online version of this article.

Abstract

This study examined whether antidepressants delay the need for dopaminergic therapy or change the degree of motor impairment and disability in a population of early Parkinson's disease (PD) patients. Preclinical studies have indicated that antidepressants modulate signaling pathways involved in cell survival and plasticity, suggesting they may serve to both treat PD-associated depression and slow disease progression. A patient-level meta-analysis included 2064 patients from the treatment and placebo arms of the following trials: FS1, FS-TOO, ELLDOPA, QE2, TEMPO, and PRECEPT. Depression severity was determined at baseline, and antidepressant use was reported in a medication log each visit. Kaplan-Meier curves and time-dependent Cox proportional hazards models determined associations between depression severity and antidepressant use with the primary outcome, time to initiation of dopaminergic therapy. ANCOVAs determined associations with the secondary outcome, degree of motor impairment and disability, reported as annualized change in UPDRS scores from baseline to final visit. When controlling for baseline depression, the initiation of dopaminergic therapy was delayed for subjects taking tricyclic antidepressants compared with those not taking antidepressants. No significant differences were found in UPDRS scores for subjects taking antidepressants compared with those not taking antidepressants. Tricyclic antidepressants are associated with a delay in reaching the end point of need to start dopaminergic therapy. The lack of change in overall UPDRS scores suggests the delay was not attributable to symptomatic effects.

© 2012 Movement Disorder Society

Volume 27, Issue 7, pages 880-887, June 2012

Article Text

Depression is a common comorbid disorder in Parkinson's disease (PD), affecting 40%-50% of patients.1, 2 Given its impact on disability and quality of life, recognition and treatment of depressive symptoms are important in disease management.3 Molecular mechanisms associated with the actions of antidepressant drugs suggest they have the potential to produce disease-modifying effects in PD.4-8 To date, clinical trials that have investigated antidepressant therapy in PD have focused primarily on safety and treatment of depression, and therefore lack the appropriate methodology and time course to determine whether antidepressants affect disease progression or disability. Therefore, the primary objective of this study was to assess whether antidepressant treatment modified the timing of need for dopamine pharmacotherapy in a population of early PD patients. Utilizing an integrated database compiled from 6 completed clinical trials from the Parkinson's Study Group (PSG) and the Neuroprotection Exploratory Trials in Parkinson's Disease Project (NET-PD), we examined whether antidepressant treatment alters the temporal course of disease progression, inferred from time to dopaminergic therapy, in early PD patients.

Materials and Methods

Study Population

A retrospective cohort study was conducted utilizing data from the following PSG and NET-PD clinical trials: ELLDOPA,9 QE2,10 TEMPO,11 PRECEPT,12 FS1,13 and FS-TOO.14 These trials included recently diagnosed PD patients not requiring dopaminergic therapy and assessed the efficacy of an adjunctive study drug or placebo over various durations of 6 months to 2 years. There were no enrollment exclusions for antidepressant use except for MAOIs, and subjects were required to maintain a stable regimen for 60 days prior to enrollment. One exception was the TEMPO study; it excluded antidepressant use except for amitriptyline, paroxetine, sertraline, fluvoxamine, and trazodone. Details of the study populations are provided in Supplemental Table 1.

End Point

The primary end point for the analyses in this study was time to initiation of dopaminergic therapy, defined as the number of days from baseline until the study investigator determined that the subject had reached a sufficient level of disability to warrant symptomatic therapy. This end point is a reliable measure of disease progression in early PD and has been utilized in a number of clinical trials to assess novel interventions.15 In addition, the percentage of subjects who had not reached the end point at 1 year was reported in the unadjusted Kaplan-Meier curves. This time point was chosen because it is the halfway point for the average duration of the longest study (approximately 2 years); therefore, any significant changes at this time may represent an important and perhaps clinically meaningful finding. As the treatment medication for the ELLDOPA study was levodopa, subjects from this study were excluded from the analyses of the primary outcome. The secondary outcome, degree of disability and impairment, was reported as annualized change in total (parts I-III), mental (part I), activities of daily living (ADL) (part II), motor (part III), and tremor (tremor items in parts II and III) UPDRS scores from baseline to final visit defined as (UPDRS score at last assessment – UPDRS score at baseline)/number of days between the 2 assessments x 365.25.

Exposure

Subjects taking antidepressants during the study were classified in 2 ways: those reporting in the medication log taking any antidepressant and those taking a particular class of antidepressant (SSRI, SNRI, tricyclic, amitriptyline hydrochloride specifically, atypical, >1 antidepressant). Days of exposure were determined based on reported start and stop dates of medication use. IN addition, subjects were classified by depression severity. Depression severity was classified as none, mild, or moderate/severe based on cutoffs from 3 depression scales used in the included studies. Mild depression was defined by a loss of interest in normal activities, unusual irritability, and reduced motivation in work, home, or social activities. Moderate/severe depression was defined by considerable distress or agitation, significant difficulties with work or domestic activities, loss of self-esteem, or feelings of uselessness and guilt. Depression severity was determined within each study as follows: the Geriatric Depression Scale (GDS)16 was administered to subjects in the FS1 and FS-Too trials, the Hamilton Depression Rating Scale (HDRS)17 was administered to subjects in the ELLDOPA and QE2 trials, and the Beck Depression Inventory (BDI)18 was administered to subjects in the TEMPO and PRECEPT trials.

Statistical Models

Kaplan-Meier curves were generated to assess the median time to therapy and the percentage of subjects who had not reached the end point at 1 year for baseline depression severity, overall antidepressant use, and specific antidepressant classes. To evaluate the independent role of antidepressants on the need for therapy, we used time-dependent Cox regression models. These models are an extension of the ordinary Cox proportional hazard regression, adapted for covariates that vary over time for the same individuals. Antidepressant use was included in the models as a cumulative proportion of study time the subject was taking medication each day of the study. Hazard ratios (HRs) and 95% confidence intervals (CIs) were generated for delay in dopaminergic therapy associated with depression severity and antidepressant use. Analyses were performed using ANCOVAs with annualized rate of change in total and subcomponent UPDRS scores as dependent variables and baseline exposure to antidepressants, depression severity, and potential confounders as predictor variables. Confounding variables that were accounted for in all adjusted analyses were: age, sex, race, treatment group, study, site, baseline UPDRS score, and prestudy antidepressant use. All analyses were performed using SAS 9.2.

Results

Baseline Characteristics and Depression Severity

A total of 2064 patients compiled from both treatment and placebo groups were included in the analyses. Of these, 451 were taking some form of antidepressant anytime during the study. Patient baseline characteristics were similar among the studies (Table 1). Approximately 15% of untreated subjects presented with mild or moderate/severe depression when they entered the study. Most subjects treated with an antidepressant at baseline fell into the "not depressed" or "mild depression" categories, with a small percentage being classified with "moderate/severe" depression. However, moderate/severe depression was more common among subjects taking an SSRI (16.87%) than a tricyclic (7.41%), SNRI (3.45%), or atypical antidepressant (5.71%) at baseline.

Table 1. Patient characteristics at baseline for each study

  Elldopa FS1 FSToo Precept QE2 Tempo Combined
Total per study (n) 361 200 213 806 80 404 2064
Age, mean (SD) 64.56 (10.91) 62.33 (10.38) 61.03 (10.36) 59.69 (10.26) 61.30 (11.26) 60.84 (10.8) 61.22 (10.67)
Male (%) 67.59 63.00 65.26 64.39 65.00 63.61 64.78
Race, white (%) 90.58 92.50 89.67 95.66 96.25 94.80 93.70
Prestudy AD use (%) 16.34 19.00 17.84 17.99 20.00 9.65 16.23
Total UPDRS, mean (SD) 25.89 (11.02) 23.68 (9.46) 22.34 (8.85) 24.62 (10.53) 23.35 (9.69) 25.03 (10.84) 24.54 (10.42)
Not depressed (%) 85.60 82.50 89.67 81.14 93.75 79.70 83.14
Mildly depressed (%) 13.85 10.5 7.98 12.66 6.25 12.87 11.97
Moderately/severely depressed (%) 0.55 7.00 2.35 6.20 0.00 7.43 4.89

Impact of Depression Severity and Antidepressants on Initiation of Dopaminergic Therapy

The unadjusted median time to therapy for subjects taking any antidepressant was significantly shorter than that for those not taking antidepressants at baseline (363 vs 449 days; P = .002; Fig. 1A). However, when antidepressant use was stratified by specific class of antidepressant at baseline, the median time to therapy was delayed, and the percentage of subjects not meeting the end point at 1 year was greater for subjects taking amitriptyline (P = .02) and atypical antidepressants (including bupropion, mirtazapine, trazodone, and wellbutrin; P = .01) compared with those not taking antidepressants (Fig. 1C).

Figure 1Figure 1. Kaplan-Meier curves showing median time to end point based on antidepressant treatment and depression severity. A: Survival curves for subjects taking any antidepressant (blue) compared with those who did not take antidepressants (black). B: Survival curves for subjects taking different classes of antidepressants: tricyclics (red), amitriptyline (orange), atypical (turquoise), SSRI (yellow), SNRI (blue), >1 antidepressant (pink) or no antidepressant (black) during the study. C: Survival curves for subjects based on depression severity at baseline: no depression (black), mild depression (blue), or moderate/severe depression (turquoise). Curves include raw data (days to end point) and are not adjusted for confounders.

When controlling for depression (and other confounders; see Materials and Methods section for a detailed list), subjects taking tricyclics, amitriptyline in particular, had a lower probability of initiating dopaminergic therapy (HR, 0.3; P = .004; HR, 0.3; P = .01; Table 2. When controlling for antidepressant use (any or each individual class), mild depression was associated with a higher probability of starting dopaminergic therapy (HR, 1.4; P < .004 for all models; Table 3). However, moderate/severe depression was not significantly associated with the same outcome. We explored the potential interaction between baseline depression severity and time-dependent tricyclic use by running a model with separate time-dependent indicator variables for tricyclic use among the 3 categories of depression. This model did not fit the data significantly better than the model with only a single variable for tricyclic use (chi-square [2 df] = 2.4; P = .30).

Table 2. Survival analysis for time to initiation of dopaminergic therapy depending on antidepressant use

All subjects HR 95% CI I value
  • Time-dependent use indicates cumulative proportion of study time subjects were taking medications. Models adjusted for baseline UPDRS score, baseline age, sex, race, treatment group, study, site, baseline depression status (none, mild, moderate/severe), and prestudy antidepressant use (yes or no; or prestudy tricyclic, amitriptyline, SSRI, SNRI, atypical use depending on main medication). Survival analyses exclude ELLDOPA because some patients were taking levodopa.
  • a
    Significant P < .05.
Time-dependent antidepressant use (any) 0.9 (0.6, 1.3) .63
Time-dependent tricyclic use 0.3 (0.1, 0.7) .004a
Time-dependent amitriptyline use 0.3 (0.1, 0.7) .01a
Time-dependent SSRI use 0.9 (0.6, 1.4) .64
Time-dependent SNRI use 0.9 (0.2, 3.1) .82
Time-dependent atypical use 2.3 (0.9, 5.4) .06

Table 3. Survival analysis for time to initiation of dopaminergic therapy depending on depression severity

Group HR 95% CI I value
  • Time-dependent use indicates cumulative proportion of study time subjects were taking medications. Models adjusted for baseline UPDRS score, baseline age, sex, race, treatment group, study, site, baseline depression status (none, mild, moderate/severe), and prestudy antidepressant use (yes or no; or prestudy tricyclic, amitriptyline, SSRI, SNRI, atypical use depending on main medicaiton). Survival analyses exclude ELLDOPA because some patients were taking levodopa.
  • a
    Significant P value compared with no baseline depression.
Baseline depression status (controlling for antidepressant use)
None (referent) 1.0 - -
Mild 1.4 (1.1, 1.7) .003a
Moderate/severe 1.0 (0.7, 1.3) .92
Baseline depression status (controlling for tricyclic antidepressant use)
None (referent) 1.0 - -
Mild 1.4 (1.1, 1.7) .002a
Moderate/severe 1.0 (0.7, 1.3) .93
Baseline depression status (controlling for amitriptyline use)
None (referent) 1.0 - -
Mild 1.4 (1.1, 1.7) .002a
Moderate/severe 1.0 (0.7, 1.3) .98
Baseline depression status (controlling for amitriptyline use)
None (referent) 1.0 - -
Mild 1.4 (1.1, 1.7) .004a
Moderate/severe 1.0 (0.7, 1.3) .86
Baseline depression status (controlling for amitriptyline use)
None (referent) 1.0 - -
Mild 1.4 (1.1, 1.7) .002a
Moderate/severe 1.0 (0.8, 1.3) .98
Baseline depression status (controlling for amitriptyline use)
None (referent) 1.0 - -
Mild 1.4 (1.1, 1.7) .002a
Moderate/severe 1.0 (0.7, 1.3) .93

Impact of Depression Severity and Antidepressant Use on Degree of Disability

A total of 2029 subjects (98%) had data available for UPDRS assessments. Baseline antidepressant use (any or by class) did not significantly affect the rate of change in any of the UPDRS outcomes when compared with subjects not taking antidepressants (Table 4). Although, not surprisingly, when stratified by depression severity, untreated subjects with moderate/severe depression had a significant worsening on the mental component of the UPDRS compared with untreated, nondepressed subjects (least-square mean, 2.2 vs 0.8; P = .04).

Table 4. Influence of antidepressant use on UPDRS scores (annualized rate of change)

  Total UPDRSMental UPDRSADL UPDRSMotor UPDRSTremor UPDRS
n LSM (SE) P value LSM (SE) P value LSM (SE) P value LSM (SE) P value LSM (SE) P value
  • Models adjusted for baseline UPDRS score, baseline age, sex, race, treatment group, study, site, baseline depression status (none, mild, moderate/severe), and prestudy antidepressant use (yes or no; or prestudy tricyclic, amitriptyline, SSRI, SNRI, atypical use depending on main medication).
  • a
    Significant P value (<.05) compared with no antidepressant use and not depressed. LSM, least squares mean; ADL, activities of daily living.
Baseline antidepressant use (any)
Yes 292 9.4 (3.5) .80 1.5 (0.8) .66 3.4 (1.3) .93 4.7 (2.7) .64 1.3 (0.7) .90
No 1737 10.4 (1.4) - 1.1 (0.3) - 3.2 (0.5) - 6.1 (1.1) - 1.2 (0.3) -
Baseline antidepressant use (by class)
Tricyclic 52 8.0 (4.6) .66 1.2 (1.1) .90 2.3 (1.7) .62 4.3 (3.6) .66 0.4 (0.9) .48
SSRI 162 11.3 (3.7) .78 1.5 (0.9) .69 4.4 (1.4) .43 5.7 (2.9) .94 1.7 (0.8) .52
SNRI 28 8.5 (5.4) .76 2.5 (1.2) .28 2.2 (2.0) .62 3.8 (4.2) .62 0.1 (1.1) .34
Atypical 34 1.6 (5.0) .10 0.6 (1.2) .69 1.8 (1.9) .46 -0.7 (3.9) .10 1.2 (1.0) .95
>1 antidepressant class 16 15.2 (6.6) .46 2.7 (1.5) .31 2.5 (2.5) .80 9.9 (5.1) .46 2.0 (1.4) .54
None 1737 10.2 (1.4) - 1.1 (0.3) - 3.2 (0.5) - 5.9 (1.1) - 1.1 (0.3) -
Baseline antidepressant use (by amitriptyline and other tricyclics)
Amitriptyline 42 7.6 (4.9) .61 1.3 (1.1) .86 2.2 (1.8) .59 3.9 (3.8) .60 0.7 (1.0) .64
Other tricyclic 10 9.8 (8.0) .97 0.9 (1.9) .93 2.9 (3.0) .93 6.0 (6.2) .99 -0.4 (1.6) .35
Baseline antidepressant use (yes/no)/depression severity (none, mild, moderate/severe)
Yes and none 193 9.7 (3.6) 0.83 1.2 (0.8) 0.65 3.2 (1.3) 0.97 5.2 (2.8) 0.64 1.2 (0.7) 0.86
Yes and mild 63 10.3 (4.3) 0.95 2.2 (1.0) 0.18 4.7 (1.6) 0.38 4.1 (3.4) 0.48 1.7 (0.9) 0.77
Yes and mod/severe 36 7.8 (4.9) 0.59 0.7 (1.1) 0.97 2.0 (1.8) 0.54 5.1 (3.8) 0.7 1.2 (1.0) 0.9
No and none 1493 10.6 (1.1) - 0.8 (0.3) - 3.2 (0.4) - 6.6 (0.9) - 1.4 (0.2) -
No and mild 179 10.6 (1.9) 0.99 0.7 (0.5) 0.98 3.5 (0.7) 0.65 6.4 (1.5) 0.87 0.8 (0.4) 0.13
No and moderate/severe 65 10.3 (3.0) 0.91 2.2 (0.7) .04a 3.2 (1.1) 0.96 4.8 (2.3) 0.43 1.1 (0.6) 0.71

Discussion

Our findings suggest that early PD patients treated with tricyclics, amitriptyline in particular, experience a significant delay in the initiation of dopaminergic therapy. Amelioration of depression in this PD population is likely less responsive to SSRIs than to tricyclics, as evidenced by baseline depression scores and previous findings in cohorts of early PD patients.19, 20 The lack of change in UPDRS scores suggests that the lower rate of reaching the end point is not associated with less worsening in motor function. This could be a result of better control of symptoms not captured by the UPDRS, such as depressive symptoms, although the probability of initiating dopaminergic therapy was adjusted for depression. It could also be a result of relatively better treatment of tremor compared with other motor symptoms, assuming that tremor is a target symptom in the decision to start dopaminergic treatment. Finally, although impossible to confirm in this study, the delay in the need of dopaminergic treatment may represent a disease-modifying effect of tricyclics, in line with preclinical studies.

Although antidepressants are safe and effective in treating depression in PD,19-22 we have a limited understanding of their potential effects on neurodegeneration. However, results from several preclinical studies show various classes of antidepressants modulate the signaling pathways involved in cell survival and plasticity.6-8, 23-25 More recently, both paroxetine23 and fluoxetine26 were shown to prevent the loss of dopaminergic neurons by inhibiting inflammation and oxidative stress in the 1-methyl-4-phenyl-1,2,3,4-tetrahydropyridine (MPTP) mouse model of PD. Furthermore, our laboratory found chronic amitriptyline treatment was neuroprotective for the dopamine system in the 6-hydroxydopamine rat model of parkinsonism.27 Finally, a recent open-label clinical trial demonstrated that treatment with select SSRIs restored brain-derived neurotrophic factor serum levels and attenuated motor deficits in PD patients already receiving dopamine therapy.22 Combined, these findings suggest that antidepressant treatment can evoke changes in the brain that may reduce the rate of DA cell degeneration in PD.

Prior clinical studies had not specifically addressed the neuroprotective potential of antidepressants. The data provided here suggest antidepressants may have neuroprotective properties and encourage prospective clinical studies. However, these results should be interpreted with caution, as depression and disability are closely intertwined, and it is difficult to separate the effects of antidepressant therapy from symptomatic relief of depression, which can contribute greatly to the quality of life.28 For example, Ravina and colleagues29 showed depressed PD patients were more likely to complain of symptoms, which resulted in a faster time to dopaminergic therapy than were nondepressed PD patients. Although we included both depression and antidepressant use as confounds in this study, there is a possibility that a putative disease-modifying effect elicited by chronic tricyclic use is the result of antidepressant efficacy and may not be reflective of a neuroprotective effect, especially considering the recent data by Mendez et al20 demonstrating tricyclics provide superior relief of depressive symptoms than do SSRIs or placebo. It is regrettable that our data could not be adjusted for differences in antidepressant efficacy among the various antidepressants, but several of the source studies did not include postintervention depression scales; therefore, the results may indeed represent greater antidepressant efficacy of tricyclics than a direct neuroprotective effect. The lack of changes in disability as measured by UPDRS among antidepressants and/or between depressed and nondepressed treated subjects supports this argument but may also suggest that a longer follow-up to determine differences in disability would have been necessary to examine whether the disability difference between groups could widen over time in favor of nondepressed or tricyclic-treated subjects.

Properly designed, well-controlled antidepressant studies are rare in PD patients; however, numerous open-label trials have been conducted to assess whether antidepressants are safe and effective in PD. These studies provided evidence that antidepressants from various classes are safe and effective at alleviating depressive symptoms in depressed PD patients.19, 21, 30 However, the end point/outcome measures of the aforementioned studies were not appropriate to determine whether antidepressants are capable of modifying the progression of PD. The short duration and lack of controls in these studies necessitate more research to ascertain whether antidepressants could provide disease-modifying effects in PD. To this end, we examined whether antidepressant treatment influenced the progression of symptoms over a more extensive time course (6-24 months). We found no differences in annualized UPDRS scores for subjects taking antidepressants compared with those who were not. In fact, both treated and nontreated subjects exhibited an expected rate of decline (between 8 and 11 points) for early PD patients,31 suggesting antidepressants do not have a symptomatic effect on motor disability. In addition, independent classes of antidepressants did not significantly affect UPDRS scores. However, subjects taking amitriptyline experienced the smallest degree of worsening (7.4 points/year) compared with those taking other classes of antidepressants (except atypical antidepressants) and those not taking antidepressants (Table 4). This lack of significant differences between groups may be a result of the limitations of our study. First, the population comprised patients early in the disease course (<5 years); therefore, the extent of symptoms experienced at this stage of disease may not be large enough to detect significant differences between groups. Second, we compiled data from multiple studies with varying durations, which may have diluted the magnitude of disability for studies with a longer time course. Third, we categorized subjects by the class of antidepressant taken, which may have made the sample size too small to detect significant differences between the groups. Fourth, the putative disease-modifying effects of amitriptyline or tricyclics in general may not be adequately captured through the UPDRS motor scale.32 Incidentally, tremor may have specifically influenced whether subjects received treatment with tricyclics, as its anticholinergic properties may theoretically attenuate tremor, although any such change was insufficient to drive differences in the UPDRS motor score. Because the UPDRS motor score is fairly insensitive to change in early PD,33 a scale that more directly examines a patient's quality of life may be a more sensitive measure of disease-modifying changes than the UPDRS motor score. It has been suggested by Harrison et al34 and others32, 35 that the ADL component of the UPDRS may serve as a better marker of disease progression than the other subscales, as it is most responsive to changes over time. Although not significant, we found the mean annualized change in ADL subscore for subjects taking tricyclics was less (2.3) compared with SSRIs (4.4) or subjects not taking antidepressants (3.2); see Table 4. Finally, we did not control for a specific clinical phenotype such as akinetic-rigid versus tremor dominant,31, 36 which could potentially influence the relationship between depression and initiation of dopaminergic therapy. Collectively, these limitations could be overcome in a prospective, well-controlled clinical trial. The data suggest that early tricyclic therapy affords patients more functional time in the early phases of PD (without dopaminergic treatment) and potentially a longer period of adequate disease management and good quality of life with standard dopaminergic therapy, although this remains to be determined by future studies.

Depression is one of the most debilitating and persistent nonmotor symptoms experienced by PD patients. In fact, it may have a greater negative impact on quality of life than motor disability.37 The reported prevalence and incidence rates for depression in PD are inconsistent because of overlapping symptoms, various assessment techniques, and different populations of patients.2 However, several clinical studies report that depression is often underrecognized and undertreated in patients with PD.28, 29 Similarly, we report that 348 subjects (17% of the total study population) were classified as having mild to severe depression at baseline. Of these, 248 subjects (71%) were not receiving treatment for depressive symptoms. In addition, mild, untreated depression was associated with an increased probability of beginning dopaminergic therapy. Furthermore, untreated subjects with moderate/severe depression exhibited a significant worsening on the mental component of the UPDRS. Because enrollment criteria included an established antidepressant regimen (>60 days), that many patients taking antidepressants were classified as depressed at baseline suggests their symptoms were not being adequately managed. However, results from a recent randomized trial showed that the tricyclic antidepressant nortriptyline was more efficacious in treating depression in PD patients than placebo, whereas the SSRI paroxetine was not, which suggests that, compared with the general population, different mechanisms may be involved in PD-associated depression.20 Likewise, our results indicate SSRIs may not be the optimal choice for treating depression in PD. For example, of the subjects taking an SSRI at baseline, more than 16% reported having moderate/severe depression compared with only 7% of those taking tricyclics. Therefore, depression in PD may result from different neurochemical abnormalities than major depression in the general population. Taken together, these findings highlight the impact of untreated depression in PD and suggest that depressive symptoms are not properly managed in early PD patients.

In summary, these findings support previous evidence that untreated depression contributes to disability in PD. Despite finding no change in UPDRS scores, the association of treated or untreated mild depression with a higher probability of beginning dopaminergic therapy suggests a role for depression in increasing overall disability and possibly accelerating disease progression. Similarly, the association of delayed initiation of dopaminergic therapy with tricyclic antidepressant treatment suggests that this class of drugs specifically may be the most effective in treating PD-related depression, altering the judgment of need for dopaminergic therapy and possibly slowing the course of disease progression. These results also illustrate the importance of treating both the motor and nonmotor symptoms associated with PD, as both types of symptoms contribute greatly to disability. In addition, nonmotor symptoms can exacerbate motor symptoms and decrease the quality of life for patients. Therefore, a therapy that can address both the nonmotor and motor symptoms of PD while slowing disease progression would provide the greatest benefit for PD patients. Prospective clinical trials and long-term follow-up of patients in extension or naturalistic studies are needed to determine the putative disease-modifying effects versus differential efficacy of antidepressant therapy on depressed patients who are optimally treated, depressed patients suboptimally treated or untreated, and nondepressed patients.

Acknowledgements

We acknowledge the contributions of the investigators and coordinators of the following PSG and NET-PD clinical trials (references given in text): ELLDOPA, QE2, TEMPO, PRECEPT, FS1, and FS-TOO. Special thanks to Roseanna Battista and Drs. Alberto Ascherio, Alok Sahay, Andrew Duker, Caryl Sortwell, Brian Terpstra, Kim Seroogy, James Herman, David Yurek, and JoEl Schultz and to the PSG Genetics and Epidemiology Working Group for their support and unwavering dedication to this project.

Author Roles:

Katrina L. Paumier contributed to the design and conceptualization of the study, interpretation of data, draft and revision of the manuscript. Andrew D. Siderowf contributed to the design and conceptualization of the study, interpretation of data, and revision of the manuscript. Peggy Auinger contributed to the design and conceptualization of the study, analysis and interpretation of data, and revision of the manuscript. David Oakes contributed to the analysis and interpretation of data and revision of the manuscript. Lalitha Madhavan contributed to the design and conceptualization of the study, interpretation of data, and revision of the manuscript. Alberto J. Espay contributed to the design and conceptualization of the study, interpretation of data, and revision of the manuscript. Fredy J. Revilla contributed to the design and conceptualization of the study, interpretation of data, and revision of the manuscript. Timothy J. Collier contributed to the design and conceptualization of the study, interpretation of data, and revision of the manuscript.

Financial Disclosures: Katrina Paumier is supported by funding from the Parkinson's Disease Foundation's Advancing Parkinson's Treatments Innovations Grant to the Parkinson Study Group (PSG), the Davis Phinney Foundation for Parkinson's, the Morris K. Udall Center of Excellence for Parkinson's Disease Research at Michigan State University, and the University of Cincinnati (NS 058830). Andrew Siderowf is supported by the University of Pennsylvania Udall Center (NS 053488). Peggy Auinger reports no disclosures. David Oakes is supported by grants from the NIH, the Michael J. Fox Foundation, and the Department of Defense. Lalitha Madhavan is supported by funding from the Morris K. Udall Center of Excellence for Parkinson's Disease Research at Michigan State University and the University of Cincinnati (NS 058830). Alberto Espay is supported by the KL2 Research Scholars mentored career development award through the NIH Institutional Clinical and Translational Science Award (RR026315-02), has received grant support from Medtronic, CleveMed, Davis Phinney Foundation, and the Michael J. Fox Foundation, personal compensation as a consultant for Solvay, and honoraria from Novartis, the American Academy of Neurology, and the Movement Disorders Society. Fredy Revilla is supported by funding from the NIH (subaward from Washington University on R01 NS058797) as a coinvestigator/site PI, the Davis Phinney Foundation, and the Gardner Family Center for Parkinson's Disease and Movement Disorders, personal compensation as a consultant for Lundbeck, and honoraria from Medtronic. Timothy Collier is supported by funding from the NIH (NS 058830) and the Michael J. Fox Foundation and also receives compensation as an associate editor for the European Journal of Neuroscience.

References

  1. Dooneief G, Mirabello E, Bell K, Marder K, Stern Y, Mayeux R. An estimate of the incidence of depression in idiopathic Parkinson's disease. Arch Neurol 1992; 49: 305-307.
  2. McDonald WM, Richard IH, DeLong MR. Prevalence, etiology, and treatment of depression in Parkinson's disease. Biol Psychiatry 2003; 54: 363-375.
  3. Papapetropoulos S, Ellul J, Argyriou AA, Chroni E, Lekka NP. The effect of depression on motor function and disease severity of Parkinson's disease. Clin Neurol Neurosurg 2006; 108: 465-469.
  4. Peng Q, Masuda N, Jiang M, et al. The antidepressant sertraline improves the phenotype, promotes neurogenesis and increases BDNF levels in the R6/2 Huntington's disease mouse model. Exp Neurol 2008; 210: 154-163.
  5. Sairanen M, O'Leary OF, Knuuttila JE, Castren E. Chronic antidepressant treatment selectively increases expression of plasticity-related proteins in the hippocampus and medial prefrontal cortex of the rat. Neuroscience 2007; 144: 368-374.
  6. Zschocke J, Zimmermann N, Berning B, Ganal V, Holsboer F, Rein T. Antidepressant drugs diversely affect autophagy pathways in astrocytes and neurons-dissociation from cholesterol homeostasis. Neuropsychopharmacology 2011; 36: 1754-1768.
  7. Zhao Z, Zhang HT, Bootzin E, Millan MJ, O'Donnell JM. Association of changes in norepinephrine and serotonin transporter expression with the long-term behavioral effects of antidepressant drugs. Neuropsychopharmacology 2009; 34: 1467-1481.
  8. Jang SW, Liu X, Chan CB, et al. Amitriptyline is a TrkA and TrkB receptor agonist that promotes TrkA/TrkB heterodimerization and has potent neurotrophic activity. Chem Biol 2009; 16: 644-656.
  9. Fahn S. Parkinson disease, the effect of levodopa, and the ELLDOPA trial. Earlier vs later L-DOPA. Arch Neurol 1999; 56: 529-535.
  10. Shults CW, Oakes D, Kieburtz K, et al. Effects of coenzyme Q10 in early Parkinson disease: evidence of slowing of the functional decline. Arch Neurol 2002; 59: 1541-1550.
  11. A controlled trial of rasagiline in early Parkinson disease: the TEMPO Study. Arch Neurol 2002; 59: 1937-1943
  12. Wang LH, Johnson EM,Jr. Mixed lineage kinase inhibitor CEP-1347 fails to delay disability in early Parkinson disease. Neurology 2008; 71: 462; author reply 462-463.
  13. A randomized, double-blind, futility clinical trial of creatine and minocycline in early Parkinson disease. Neurology 2006; 66: 664-671.
  14. NINDS NET-PD Investigators. A randomized clinical trial of coenzyme Q10 and GPI-1485 in early Parkinson disease. Neurology 2007; 68: 20-28.
  15. Marras C, Lang AE. Outcome measures for clinical trials in Parkinson's disease: achievements and shortcomings. Exp Rev Neurother 2004; 4: 985-993.
  16. Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982; 17: 37-49.
  17. Hamilton M. Rating depressive patients. J Clin Psychiatry 1980; 41: 21-24.
  18. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4: 561-571.
  19. Frisina PG, Tenenbaum HR, Borod JC, Foldi NS. The effects of antidepressants in Parkinson's disease: a meta-analysis. Int J Neurosci 2008; 118: 667-682.
  20. Menza M, Dobkin RD, Marin H, et al. A controlled trial of antidepressants in patients with Parkinson disease and depression. Neurology 2009; 72: 886-892.
  21. Bxarone P, Poewe W, Albrecht S, et al. Pramipexole for the treatment of depressive symptoms in patients with Parkinson's disease: a randomised, double-blind, placebo-controlled trial. Lancet Neurol 2010; 9: 573-580.
  22. Ricci V, Pomponi M, Martinotti G, et al. Antidepressant treatment restores brain-derived neurotrophic factor serum levels and ameliorates motor function in Parkinson disease patients. J Clin Psychopharmacol 2010; 30: 751-753.
  23. Chung YC, Kim SR, Jin BK. Paroxetine prevents loss of nigrostriatal dopaminergic neurons by inhibiting brain inflammation and oxidative stress in an experimental model of Parkinson's disease. J Immunol 2010; 185: 1230-1237.
  24. Jin Y, Lim CM, Kim SW, et al. Fluoxetine attenuates kainic acid-induced neuronal cell death in the mouse hippocampus. Brain Res 2009; 1281: 108-116.
  25. Chadwick W, Mitchell N, Caroll J, et al. Amitriptyline-mediated cognitive enhancement in aged 3xtg Alzheimer's disease mice is associated with neurogenesis and neurotrophic activity. PLoS One 2011; 6: e21660.
  26. Chung YC, Kim SR, Park JY, et al. Fluoxetine prevents MPTP-induced loss of dopaminergic neurons by inhibiting microglial activation. Neuropharmacology 2010; 60: 963-974.
  27. Paumier KL, Sortwell C, Daley BF, et al. Tricyclic antidepressant treatment protects dopamine neurons and increases neurotrophic factors within the nigrostriatal system. Program No. 45.13/E34. 2009 Neuroscience Meeting Planner. Chicago, IL: Society for Neuroscience, 2009. Online.
  28. Menza M, Dobkin RD, Marin H, et al. The impact of treatment of depression on quality of life, disability and relapse in patients with Parkinson's disease. Mov Disord 2009; 24: 1325-1332.
    Direct Link:
  29. Ravina B, Camicioli R, Como PG, et al. The impact of depressive symptoms in early Parkinson disease. Neurology 2007; 69: 342-347.
  30. Chen P, Kales HC, Weintraub D, Blow FC, Jiang L, Mellow AM. Antidepressant treatment of veterans with Parkinson's disease and depression: analysis of a national sample. J Geriatr Psychiatry Neurol 2007; 20: 161-165.
  31. Poewe WH, Wenning GK. The natural history of Parkinson's disease. Ann Neurol 1998; 44( 3 Suppl 1): S1-S9.
  32. Rascol O, Fitzer-Attas CJ, Hauser R, et al. A double-blind, delayed-start trial of rasagiline in Parkinson's disease (the ADAGIO study): prespecified and post-hoc analyses of the need for additional therapies, changes in UPDRS scores, and non-motor outcomes. Lancet Neurol 2011; 10: 415-423.
  33. The Unified Parkinson's Disease Rating Scale (UPDRS): status and recommendations. Mov Disord 2003; 18: 738-750.
  34. Harrison MB, Wylie SA, Frysinger RC, et al. UPDRS activity of daily living score as a marker of Parkinson's disease progression. Mov Disord 2009; 24: 224-230.
    Direct Link:
  35. Schrag A, Spottke A, Quinn NP, Dodel R. Comparative responsiveness of Parkinson's disease scales to change over time. Mov Disord 2009; 24: 813-818.
    Direct Link:
  36. Poewe WH, Wenning GK. The natural history of Parkinson's disease. Neurology 1996; 47( 6 Suppl 3): S146-S152.
  37. Schrag A. Quality of life and depression in Parkinson's disease. J Neurol Sci 2006; 248: 151-157.