Introduction

Parkinson’s disease (PD) is the second most common neurodegenerative disease and its treatment continues to be a challenge [1]. Currently, deep-brain stimulation (DBS) is the most well-established and effective treatment for advanced PD patients presenting with motor complications as a result of standard pharmacological treatment. The most common target of DBS in PD patients is the bilateral subthalamic nucleus (STN) due to its alleviation of motor symptoms and dopaminergic treatment [2].

Prospective and retrospective studies addressing various aspects of suicidality after subthalamic nucleus deep-brain stimulation (STN DBS), including suicide attempts and suicides, as well as suicidal depression and suicidal ideation, have reported contrasting results, along with a range of suicidality rates (0 to 17%) [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]. In order to gain a comprehensive understanding of suicidality in PD patients after STN DBS treatment, we conducted a meta-analysis based on a systematic review of relevant literature. We calculated pooled rates of suicidal ideation, suicide attempts, and suicides.

Materials and methods

This meta-analysis was performed based on the recommendations outlined in the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement [21].

Search strategy

We performed a systematic review of literature indexed in electronic databases, including Medline, PubMed, Web of Science, and Embase. The following keywords were used for searches: “suicide” or “suicide ideation” or “suicide attempts” and “subthalamic nucleus stimulation” or “deep brain stimulation.” All eligible studies published before November 2019 were included, and no language restrictions were applied. All references cited in the retrieved articles were analyzed in order to identify potentially relevant studies.

Eligibility criteria

We included all observational studies (cross-sectional, case-control, or cohort) that analyzed and reported rates of completed or attempted suicide in patients with PD after STN DBS. All included studies followed the UK PD Society Brain Bank criteria to arrive at a confirmed PD diagnosis.

We excluded studies that did not focus on STN DBS in PD patients, as well as those that reported insufficient data and whose authors did not respond to our requests for additional information. Discussions, editorials, reviews, case reports, letters, commentaries, and critiques were also excluded.

Data extraction

Two independent investigators (YJX and BYY) systematically reviewed the literature and extracted relevant data. Disagreements were resolved after discussion with the corresponding author (XLY). The following data were extracted from each included study: surname of the first author, year of publication, study design, country of origin of cohort being studied, sample size, age of participants, and suicidality outcomes (rates of suicidal ideation, suicide attempts, suicide).

Statistical analysis

The meta-analysis was conducted using Stata 12.0 (StataCorp, TX, USA). We evaluated data on the rates of suicide ideation, suicide attempts, and suicides among PD patients who underwent STN DBS. The heterogeneity between studies was assessed using the Q test and quantified using the I2 statistic. I2 values between 25 and 50% were considered to indicate low heterogeneity, values between 50 and 75% indicated moderate heterogeneity, and values ≥ 75% indicated substantial heterogeneity. All I2 values below 25% indicated no heterogeneity. When no or low statistical heterogeneity was observed, we used a fixed-effect model for the meta-analysis. In the case of moderate or substantial heterogeneity, a random-effects model was used. Publication bias was assessed using the Egger’s and Begg’s tests.

Results

Literature screening and assessment

A total of 65 records were identified after examining the databases and removing duplicates (Fig. 1). After eliminating 23 records based on title and abstract, a full-text review was performed for 42 records. A total of 10 studies were automatically excluded by Stata because they reported no cases of suicidal ideation, suicide attempts, or suicides. We also excluded three review articles, three case reports, five studies that did not evaluate the prevalence of suicide in PD patients, and three studies that involved suicidality in patients who underwent DBS in the bilateral globus padillus interna (GPi).

Fig. 1
figure 1

Flow diagram of study selection

Full size image

After a detailed assessment based on eligibility criteria, the final meta-analysis included 18 unique studies involving 6652 PD patients who underwent STN DBS treatment (Table 1). Among these studies, three explored suicidal ideation, 14 investigated rates of suicide attempts, and eight studied rates of suicide in PD patients after STN DBS treatment.

Table 1 Characteristics of studies included in the meta-analysis
Full size table

Suicidal ideation among Parkinson’s patients after STN DBS

We observed moderate heterogeneity (I2 = 54.6%) among the three studies addressing suicidal ideation in PD patients after STN DBS. Using the DerSimonian and Laird method, we found that the pooled rate of suicidal ideation was 4% (95% CI 0.00 to %7.2, range 2–17%; Fig. 2). The sensitivity analysis showed that no individual study had a significant effect on the results (Supplementary Fig. 1). No significant publication bias was indicated by Begg’s (p = 0.296) or Egger’s test (p = 0.101) (Supplementary Fig. 2).

Fig. 2
figure 2

Forest plot of suicidal ideation among individuals after STN DBS (4%). The X axis indicates the 95% confidence interval. ES, effect size

Full size image

Suicide attempts among Parkinson’s patients after STN DBS

The rates of attempted suicides in PD patients who underwent STN DBS ranged from 1 to 7% across the 14 included studies. Moderate heterogeneity was observed among studies regarding this outcome (I2 = 66.6%, p = 0.004). Using a random-effects approach, we found that the pooled rate of suicide attempts was 1% (95% CI 1–2%; Fig. 3). Although Begg’s test indicated a potential publication bias (p = 0.037), Egger’s test did not (p = 0.056) (Supplementary Fig. 3). The sensitivity analysis showed that no individual study had a significant effect on the results (Supplementary Fig. 4).

Fig. 3
figure 3

Forest plot of suicide attempt among individuals after STN DBS (1%). The X axis indicates the 95% confidence interval. ES, effect size

Full size image

Suicides among Parkinson’s patients after STN DBS

The rate of suicide among PD patients who underwent STN DBS ranged from 0.6 to 7.1% (Table 1). Sensitivity analysis showed that only one study [21] significantly affected the results (Supplementary Fig. 5); after removing this study from the analysis, the pooled rate of suicide was 1% (95% CI 1–2%; Fig. 4). This result was based on a fixed-effect model since no heterogeneity was observed in the rate of completed suicide across the eight studies (I2 = 0, p = 0.973). The funnel plot was visually symmetrical (Supplementary Fig. 6), and Begg’s test indicated potential publication bias (p = 0.035), but not Egger’s test (p = 0.145).

Fig. 4
figure 4

Forest plot of complete suicide among individuals after STN DBS (1%). The X axis indicates the 95% confidence interval. ES, effect size

Full size image

Discussion

This meta-analysis examined data from 18 unique studies on suicidality in PD patients who underwent STN DBS. Our results show a high pooled rate of suicidal ideation in these PD patients (4%), while the pooled rates of attempted and completed suicide were both 1%.

Several studies have reported an association of STN DBS with suicidal ideation [5, 22], suicide attempts [3], and suicide [20]. In contrast, certain studies reported no suicides in their cohorts [23, 24], including one that collected follow-up data for 10 years [25]; these studies did not, however, report suicide attempts or suicidal ideation. A multicenter study addressing suicide after STN treatment among 5311 PD patients reported a suicide attempt rate of 0.4% and a suicide rate of 0.2% [12]. In a study comparing PD patients with the general population, the suicide rate among PD patients (0.08%) was 10 times lower than in the general population (0.8%) [26]. Our meta-analysis indicated much higher rates of suicide attempts (1%) and suicide (1%). Further studies of suicide attempts and suicidal ideation are critical to gain a better understanding of the association between STN DBS and suicidality.

An association between DBS and suicidality may be due the ability of STN DBS to increase the risk of impulse control disorders [27, 28]. DBS may affect STN structure, and striatal lesions can contribute to suicidal behavior and impulsivity [29]. In pallidal and thalamic DBS, disturbance of the basal ganglia circuitry, especially in the limbic component, may also induce mood disorders and/or suicidal ideas [30]. STN DBS may also interact with existing psychiatric comorbidities, leading to, for example, transient or chronic depression in 25% of PD patients after STN DBS [31]. Another study showed a 73% decrease in total levodopa equivalent doses immediately after DBS surgery, leading to an increase in apathy (3 vs. 13 patients before and after surgery), with two patients committing suicide [32].

The GPi is another target of DBS treatment in PD patients. A meta-analysis of randomized controlled trials indicated that stimulation of the STN or GPi is equally effective at improving motor symptoms and dyskinesia in PD patients [33]. Whether GPi stimulation exerts similar effects as STN stimulation on cognitive and psychiatric symptoms remains unclear [34]. GPi DBS may be less likely than STN DBS to lead to depression or suicidal ideation/behaviors because patients are more likely to continue their dopaminergic therapy after stimulation of the GPi than after stimulation of the STN [13]. Future studies should explore the effects of different types of DBS along with postoperative drug therapy on suicidality among PD patients.

To understand factors that may contribute to risk of suicide among PD patients, we evaluated reports of risk factors contributing to suicidality. We found that mood (e.g., hopelessness, depression, bipolar disorder), alcohol/substance-use disorders, and a history of self-harm or suicide attempts were associated with higher risk of suicide (Table 2) [26, 35,36,37,38]. PD patients who attempt or commit suicide appear to show more frequent suicidal ideation or suicide attempts, more severe psychotic symptoms, higher Beck Depression Inventory scores, and greater use of psychotropic medication than patients who do not attempt or complete suicide [4]. Patients who commit suicide may also show higher levels of sadness, discouragement, failure, guilt, self-disappointment, self-criticism, and lack of interest than those who do not [10]. Future work should clarify the effects of these various risk factors on frequency of suicidal ideation or behaviors and on how often these behaviors end in suicide.

Table 2 Assessment of potential risk factors for suicide in patients with Parkinson’s disease
Full size table

Our results must be interpreted with caution since most of our data were collected from observational studies with varied sample sizes and follow-up durations, leading to underestimation of suicidality. Another limitation that must be addressed in future work is the automatic exclusion of studies with a suicidality rate of zero. The publication bias observed in our study could be due to the fact that some large studies focused on the improvement of motor symptoms rather than suicide rates. Therefore, our results should be verified in large studies involving long follow-up.

In conclusion, the results of our meta-analysis indicate a relatively high suicide rate among PD patients after STN DBS. However, since there are a number of risk factors contributing to suicidality, our finding should not serve as a general contraindication against STN DBS as an effective treatment for PD. Instead, it emphasizes to clinicians the importance of careful monitoring of psychiatric disorders, especially suicidal ideation and suicide attempts, in PD patients before and after DBS [39]. Future studies using animal models should further explore the association between DBS and suicidality, as well as identify optimal regions in the brain that can be used as a target for DBS.