Depression and major depressive disorder in patients with Parkinson's disease†
Potential conflict of interest: None reported.
The prevalence of depression in Parkinson's disease (PD) varies greatly. In this study, we investigated major depressive disorder (MDD) and depressive symptoms without MDD in patients with PD. The psychopathological characteristics of depressive symptoms were assessed by a psychiatric interview. A total of 105 Japanese patients with PD without dementia were included. The Japanese version of the Beck Depression Inventory-II (BDI-II) with a cutoff score of 13/14 was used to screen for depression. Using a structured interview, a comprehensive psychiatric evaluation of patients with BDI-II scores >13 (high BDI patients) was completed using the criteria of the Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV-TR. Forty patients (38%) had a BDI-II >13, but 29 did not show any depressed mood. Five cases met the criteria for MDD (three current, two past) and one patient was diagnosed with minor depressive disorder. A slight depressed mood that was associated with worrying about PD was seen in 6 of 34 patients without any depressive disorder and fluctuated with aggravation of PD symptoms in two of these patients. For the diagnosis of MDD, the number of positive items from the DSM-IV-TR definition of MDD is most important and useful for differentiating MDD and non-MDD. The low-prevalence rate of MDD in our patient population suggests that PD may be a psychological stressor for MDD, but does not necessarily induce MDD. © 2009 Movement Disorder Society
Depression is a common symptom in persons with Parkinson's disease (PD), but its prevalence, as reported in a recent literature review, varies greatly, between 2.7 and 90%.1 Previous studies have reported that cognitive impairment and apathy are significantly associated with depression.2, 3 Structured interviews that identify the criteria of depression using the Diagnostic and Statistical Manual of Mental Disorders (DSM)4 and observer-rated scales (e.g., Hamilton Depression Rating Scale5 and Montgomery-Åsberg Depression Rating Scale6) have better psychometric properties, but interviewers must have clinical psychiatric training in the specification of symptoms. In studies using structured interviews and standardized diagnostic criteria, the prevalence of major depressive disorder (MDD) in patients with PD ranges from 2.3 to 55.6%.1
To elucidate the psychopathology of depression in patients with PD, the Beck Depression Inventory-II (BDI-II),7 a self-rating scale for depression, was used to screen for depression in patients with PD at a neurology hospital. BDI-II is the only Japanese version of the BDI that has been validated. A cutoff point of 13/14, which was originally suggested by Beck et al.7 to diagnose at least mild major depression, can distinguish between remitted versus mildly depressed Japanese patients with MDD.8 This study used a specialist mood disorder psychiatrist to diagnose MDD and related depressive disorders in patients with PD with BDI-II scores >13 and to assess the onset and psychopathological characteristics of depressive symptoms in patients with PD with and without MDD.
PATIENTS AND METHODS
All outpatients suffering from PD and treated in the Department of Neurology at Sapporo Azabu Neurosurgical Hospital were initially included in the study, with the exception of those with additional cerebrovascular disease, severe medical illness, or physical diseases associated with frailty. All had been diagnosed using the Core Assessment Protocol for Intracerebral Transplantation (CAPIT),9 and had been followed up for from 2 to 21 years. Their daily doses of medication were individually adjusted and included levodopa (carbidopa/L-dopa), dopamine agonists, amantadine, selegiline, or anticholinergic drugs. Because cognitive impairment makes it difficult to diagnose MDD correctly, patients with a Mini-Mental State Examination Scale (MMSE)10 score <21 were also excluded (eight patients). Four patients were excluded because of having cerebrovascular disease. The remaining 105 patients with PD (58 men, 47 women) were included in the study. The procedure and aims of the study were fully explained to subjects, and informed consent was obtained from each patient. This study was performed in accordance with the Declaration of Helsinki.
The Japanese version of the BDI-II with a cutoff score of 13/14 was used to screen for depression, because the cutoff point of 13/14 in BDI-II scores can distinguish between remitted versus mild depression in Japanese patients with MDD.8 Patients with a BDI-II score >13 were classified as depressed (high BDI) and were referred to a mood disorder specialist psychiatrist (T.I.) for a comprehensive psychiatric evaluation. Diagnoses of mood disorder (MDD, bipolar disorder, dysthymic disorder, cyclothymic disorder) were made in accordance with DSM-IV-TR.4 The Hamilton Depression Rating Scale (HDRS, 21 items)5 and the Montgomery-Åsberg Depression Rating Scale (MADRS, 10 items),6 which were evaluated by structured interviews, were administered to evaluate the severity of depression. We also used the global assessment of functioning (GAF) scale,4 which reflects the overall level of psychological, social, and occupational functioning of individuals. DSM-IV-TR indicates that the GAF score does not include impairment in functioning due to physical limitations, for example, PD. Based on these tests, the clinical variables and psychiatric characteristics for three nonoverlapping groups of patients were compared: (1) Those with a BDI-II score <14 (nondepressed group, low BDI); (2) those with a BDI-II score >13 and without current MDD (high BDI without MDD group); and (3) those with current MDD (MDD group). Using a standard psychiatric interview, a psychiatrist asked patients when and where each symptom occurred, what induced or worsened each symptom, and how each symptom changed over time. These interviews were recorded, and what patients were depressed or anxious about was analyzed.
The Mann-Whitney U-test was used to compare the symptom characteristics of low BDI and high BDI subjects without current MDD. The frequency results were analyzed using Fisher's exact test. Comparisons between high BDI with current MDD and without current MDD groups in BDI-II, HDRS, MADRS, and GAF scores were done using only 95% confidence intervals and data ranges.
Psychiatric Diagnosis of Mood Disorders in Patients with PD
Forty patients (38%) had BDI-II scores >13, but only three patients (7.9% of high BDI patients) met the DSM-IVTR criteria for current MDD, single episode. One patient diagnosed with minor depressive disorder was found to have significant major depressive symptoms, but did not fulfill the criteria for MDD. Two patients had had MDD, single episode, but were recovered at the time of the psychiatric examination, one patient by using antidepressants and the other spontaneously. No patient was diagnosed with bipolar disorder, dysthymic disorder, or cyclothymic disorder.
Table 1 shows the characteristics of patients with current and past depressive disorders as determined by the psychiatric examination. Depressive symptoms of MDD in these patients occurred immediately after or in the first year following the diagnosis of PD. Cases 1, 2, and 3 were concerned at the onset of PD and were worried by aggravation of PD symptoms. Depressed mood was worse when motor symptoms of PD worsened in Case 2. Two of three patients with current MDD (Cases 1 and 2) were treated with low doses of antidepressants at the time of the psychiatric examination. After the diagnosis of MDD, Cases 2 and 3 were treated with higher doses of selective serotonin reuptake inhibitors (SSRIs, fluvoxamine and sertraline) or a serotonin-noradrenaline reuptake inhibitor (SNRI, milnacipran) by a psychiatrist, and depressive symptoms (HDRS) and function (GAF > 70 indicating symptomatic and functional remission)11 were fully remitted within 16 weeks (Table 1).
Table 1. Characteristics of patients with PD with depressive disorders identified at psychiatric examination
|17-Items HDRS score||17||12||20||15||2||1|
|Age of PD onset||71||58||42||48||46||71|
|Age of PD diagnosis||74||61||45||48||46||73|
|Age of MDD onset||75||61||45||48||46||73|
|Diagnosis||MDD1||MDD1||MDD1||Minor D||Past 1||Past 1|
|Family history of mood disorder||–||–||–||–||–||–|
|Antidepressant at psychiatric examination||MAP (10)||FLV (50)||None||None||None||PAX (20)|
|Antidepressant therapy after the study|| ||FLV (100)||SER (100)|| || || |
|17-Items of HDRS score and GAF|
| At 8 weeks after Tx.|| ||1 and 75||7 and 65|| || || |
| At 16 weeks after Tx.|| || ||2 and 75|| || || |
Differences Between Low BDI Patients, High BDI Patients Without MDD, and Patients with MDD
Table 2 shows the clinical characteristics of the low BDI patients, the high BDI patients without current MDD, and the three patients with current MDD. All patients with MDD were women. Because of the small number (n = 3) of current patients with MDD, we did not analyze their data statistically. Low BDI patients and high BDI patients without current MDD did not differ significantly in terms of Hoehn and Yahr stage of PD,12 age, sex, duration of PD, L-dopa equivalent dose, and MMSE score.
Table 2. Characteristics of the patients with (i) low BDI, (ii) high BDI but without current MDD, and (iii) current MDD
|Number of patients||65||37||3|
| ||Mean||SD||Mean||SD||P value*||Mean||SD|
|Duration of PD (years)||7.4||4.6||7.8||4.2||0.53||5.3||0.6|
|Hoehn and Yahr||2.6||0.9||2.8||0.9||0.17||2||0.9|
|Levodopa equivalent dose (mg/day)||243||135||261||150||0.46||183||176|
|BDI-II total score||8.0||3.2||20.1||6.1||P < 0.001||29.0||0.6|
|17-Items HDRS score|| || ||3.3||2.5|| ||16.3||4.0|
|MADRS score|| || ||2.6||3.8|| ||26.0||7.6|
|GAF score|| || ||73.9||6.1|| ||56.3||6.1|
|Number of positive items of MDD criteria in DSM-IVTR|| || ||0.5||1.0|| ||6.0||1.0|
Table 3 shows the data range and 95% confidence intervals of scores of BDI-II, HDRS, MADRS, and GAF, and number of positive items on the DSM-IV-TR MDD criteria among high BDI patients without current MDD and current MDD patients. Scores on the 17-item HDRS and MADRS were much lower in high BDI patients without current MDD than in patients with current MDD, and these scores seem to distinguish these two groups better than BDI-II scores, as estimated according to 95% confidence intervals. HDRS, MADRS, and GAF scores indicated that the general functioning of patients with MDD was worse than that of patients with non-MDD due to depressive symptoms. Table 3 shows that the number of positive MDD items on the DSM-IV-TR is the most important and useful tool for distinguishing MDD and non-MDD, and that most high BDI patients had less than five of nine MDD symptoms on the DSM-IV-TR.
Table 3. Comparison of data range (95% confidence intervals) of scores of BDI-II, HDRS, MADRS, and GAF, and number of positive items of DSM-IV-TR MDD criteria between high BDI without current MDD (including or excluding a minor depressive disorder case; Case 4 in Table 1) and current MDD
|BDI-II score||14–40 (17.8–21.9)||14–40 (17.9–22.0)||27–30 (24.7–33.3)|
|HDRS score||0–15 (2.6–4.7)||0–10 (2.5–4.2)||12–20 (6.3–26.4)|
|MADRS score||0–17 (1.5–4.5)||0–13 (1.3–3.9)||19–34 (7.3–44.8)|
|GAF score||63–81 (71.9–76.0)||64–81 (72.2–76.2)||51–63 (41.2–71.5)|
|Number of positive items of DSM-IV-TR MDD criteria||0–4 (0.2–0.8)||0–3 (0.1–0.6)||5–7 (3.5–8.5)|
The individual patient interviews conducted by a psychiatrist were recorded and later analyzed. A structured interview of HDRS showed that 8 of 37 high BDI patients without current MDD were depressed. One patient was diagnosed with minor depressive disorder, and with depressive symptoms that were essentially the same as MDD symptoms and that were sustained for most of the day, nearly every day, for several months. In the other seven high BDI patients without any depressive disorders, depressive symptoms sometimes appeared but global function was not impaired continuously (e.g., they could watch TV and pursue their hobbies). According to the analysis of psychiatric interviews about time courses of severity changes in past depressive symptoms, in two of seven depressed patients without any depressive disorder, depressive symptoms tended to worsen in periods when PD symptoms worsened; similarly, in six of seven patients without any depressive disorders, depressive symptoms tended to worsen when they were worried about the progress of their PD. Depressive mood or loss of interest or pleasure in regular activities, which are core symptoms of MDD, did not meet the DSM-IV-TR criteria of MDD (the symptoms are present for most of the day, nearly every day during the same 2-week period).4 Their total symptoms did not fulfill another important DSM-IV-TR criteria of MDD, i.e., “the symptoms cause clinically significant distress or impairment in social, occupational, or other important areas of functioning.”4 In addition, based on DSM-IV-TR criteria for MDD, patients must experience at least five symptoms related to depression, including either (1) depressed mood or (2) loss of interest of pleasure. In this study, patients with depressed mood but without any depressive disorders experienced 0–2 symptoms. Hence, these patients were not diagnosed as having current MDD.
Approximately 40% of our patients with PD had mild to moderate depressive symptoms when screened with the BDI-II. Assessment by psychiatric interview by a mood disorder specialist psychiatrist showed that the prevalence rate of current MDD was only 7.9% of high BDI patients. Even if a clinically relevant depressive syndrome, such as minor depressive disorder and dysthymia, are included, the prevalence rate of clinically relevant depressive syndrome was 10% of high BDI patients. Nevertheless, it is important that screening by the BDI-II led to the diagnoses of four current depressive disorders and two past MDD, i.e., total 6 (15%) of 40 high BDI patients. On the other hand, the validity of the BDI-I as a screening and diagnostic instrument for depression in patients with PD was reported previously and BDI-I had a high positive predictive value (62%).13 There is a possibility that the differences in versions of the BDIs and cutoff points may influence the results of screening. If none of the low BDI patients had current MDD, the prevalence rate of current MDD in our 105 patients with PD was estimated at 2.8%, which is similar to the rate of 2.9% in the general population of Japan.14
Our results were similar to those reported in community-based studies in Norway and New Zealand.15, 16 Tandberg et al. reported that 24.1% of patients with PD had a BDI score ≥ 18, but only 7.7% met the criteria for MDD.16 Hantz et al. reported that 34.2% patients with PD scored higher than 5 on the 30-item General Health Questionnaire, but only 2.7% met the criteria for MDD.15 In particular, prevalence rates of MDD in community-based studies (2.7–7.7%)15, 16 were much lower than those in (university) hospital-based studies (12.2–23%).13, 17–19 The reason for this discrepancy remains unclear, but the presence of more referred patients in (university) hospital-based studies may contribute to a higher prevalence of MDD in this group. Cognitive impairment may be related to the prevalence rate of MDD in patients with PD. Tandberg et al. reported that impaired cognitive function and the presence of a thought disorder are significant predictors of MDD.2 In their study, 25.6% of patients with PD with a MMSE score < 20 had MDD, whereas 3.6% of patients with a MMSE score ≥ 20 had MDD.16 Patients with a MMSE score < 21 were excluded from this study, which may contribute to the lower prevalence rate of MDD.
The finding that so many patients with PD had high BDI-II scores may partly be explained by symptom overlap.20 Patients with PD are likely to have many nonmotor symptoms such as tiredness, fatigue, indecisiveness, loss of energy, concentration difficulty, or changes in sleeping patterns, all of which are also symptoms of depression. In this study, 29 of 40 high BDI patients did not show any depressed mood, and did not have any depressive problems psychiatrically. A structured interview of HDRS showed that seven high BDI patients without depressive disorders had slight depressed mood, but their depressed mood was not sustained. Accordingly, their depressed mood did not meet the criteria of MDD in DSM-IV-TR, because “the symptoms were not present for most of the day, nearly every day during the same 2-week period,”4 and they were not diagnosed with MDD or minor depressive disorder based on DSM-IV-TR because of less than four positive symptoms and lack of depressed mood or loss of interest or pleasure. In other words, part of these patients could be diagnosed with subsyndromal depression, which is defined as two or more depressive symptoms beneath the diagnostic threshold of MDD, as Marsh et al. highlighted the importance of subsyndromal depression and recommended the inclusion of subsyndromal depression in clinical research studies of depression of PD.20 The number of positive items of the MDD criteria is most important and critical for the MDD diagnosis as described earlier. On the other hand, HDRS and MADRS scores are not necessary for an MDD diagnosis and were originally developed for evaluating the severity and treatment effects in MDD in psychiatric practice.5, 6
Previous reports have indicated that depression in PD started before the onset of PD in some patients.21 In this study, however, past and current MDD in our patients with PD had first occurred in the year after the diagnosis of PD, suggesting that MDD may be triggered by object loss, i.e., loss of their health. The onset of MDD or depression in patients with PD may be variable. In addition, a psychiatric interview revealed that depressive symptoms tended to worsen in three patients with current MDD and six high BDI patients without current MDD when the patients had anxiety and worry about the progress of their PD. Confirmation of findings using a larger sample size will be required.
In this study, patients with current MDD were not treated for MDD or were treated with low doses of antidepressants at the time of the psychiatric examination. After the diagnosis of MDD, two patients were treated with higher dosages of selective serotonin reuptake inhibitors (SSRIs, fluvoxamine and sertraline) and serotonin-noradrenaline reuptake inhibitor (SNRI, milnacipran) by a psychiatrist (T.I.), and their symptoms and global function fully recovered within 16 weeks. This indicates the importance of early detection of MDD and early introduction of psychiatric intervention.
In conclusion, although MDD appears to occur in a small proportion of patients with PD, psychological understanding of the distress of patients with PD is an important issue. Early detection and definitive diagnosis of MDD enables patients with PD to have psychiatric intervention, including intensive antidepressant therapy.
Note Added in Proof
Author Roles: Takeshi Inoue: Conception and design, acquisition of data, analysis and interpretation of data, drafting all of the submitted publication material, critical revision of the submitted publication and material and statistics. Mayumi Kitagawa: Conception and design, acquisition of data, analysis and interpretation of data, drafting all of the submitted publication material, critical revision of the submitted publication and material and statistics. Teruaki Tanaka: Analysis and interpretation of data, critical revision of the submitted publication material and statistics. Shin Nakagawa: Interpretation of data, critical revision of the submitted publication material and supervision. Tsukasa Koyama: Interpretation of data, critical revision of the submitted publication material and supervision.
Financial Disclosures: Tsukasa Koyama has received honoraria from Pfizer, Astellas, and Eli Lilly; research grants from Dainippon Sumitomo Pharma, Pfizer, Astellas, and GSK; and has served on advisory boards for GSK.