Deep brain stimulation (DBS) is an effective surgical treatment strategy for movement disorders including dystonia, essential tremor, and Parkinson’s disease (PD). The cardinal motor symptoms of PD (resting tremor, rigidity, and bradykinesia) respond well to DBS, whereas other motor and nonmotor symptoms have variable improvement (
1,
2). Rarely, DBS therapy can worsen nonmotor symptoms of PD or produce psychiatric complications, including mood dysregulation. Acute mood changes can occur during DBS programming sessions, ranging from uncontrollable crying to mirthful laughter (
3–
5). Longer lasting anxiety, anger, or depressive episodes can also manifest postoperatively, with both subthalamic nucleus (STN) and globus pallidus interna (GPi) targets (
2,
6–
15). The STN has been more frequently implicated than the GPi. Among patients treated with STN DBS, 4%−15% were reported to develop hypomania or mania (using the authors’ original terms), often in the first 3 months after surgery (
6). These mood states are typically transient and resolve with DBS setting adjustments; infrequently, they can persist, requiring pharmacological intervention (
11–
13).
There is no uniformly agreed upon term in the current psychiatric nomenclature to describe elevated mood states associated with DBS therapy. We propose the term “stimulation-induced elevated mood states” to describe behavioral changes consisting of elevated, expansive, or irritable mood and psychomotor agitation that occur during or shortly after DBS programming changes and may be associated with increased goal-directed activity, impulsivity, grandiosity, hypersexuality, pressured speech, flight of ideas, or decreased need for sleep. These episodes can also be associated with distractibility, inattention, and poor judgment, as the following case vignettes illustrate. Here we describe five patients who experienced stimulation-induced elevated mood states.
Results
Eighty-two patients were included in the chart review. Nine patients (11%) developed stimulation-induced elevated mood states. Of these, five illustrative cases are included in this study (all males with STN DBS; mean age=62.2 years [SD=10.5]; mean PD duration=8.6 years [SD=1.6]; follow-up period=40 months [SD=25.4]). Four patients were implanted with Medtronic DBS lead model 3389, and one patient (case 2) received a Boston Scientific Intrepid system. DBS leads were implanted under microelectrode recording guidance; postoperative imaging studies confirmed adequate placement in all cases. Elevated mood symptoms lasted 30 minutes to 6 months; in four of the patients, they recurred each time the same contacts were used.
Table 1 summarizes the five patients’ clinical and demographic characteristics and DBS settings associated with elevated mood.
Case 1
“Mr. A” was a 73-year-old man with no previous psychiatric history who underwent staged bilateral STN DBS for left-predominant tremor and rigidity with motor fluctuations. His right STN was implanted with excellent motoric response, followed by left STN 5 months later.
During a programming session 16 months after right STN surgery, the voltage on contact 1 was raised from 2.7 to 3.5 V. Prior to this, Mr. A had described his mood as calm and serene and had minimal spontaneous speech. Approximately 15 minutes after voltage was raised, Mr. A exhibited psychomotor agitation and became very talkative. Stimulation was lowered back to 2.7 V, and symptoms resolved. Two weeks later, when the same settings were trialed in clinic, elevated mood recurred: Mr. A reported feeling euphoric, paced around the office, crouched in a corner, and grabbed a female examiner’s leg. He walked backward and waved his arms as if conducting an orchestra during the neurological examination. Symptoms again resolved with DBS amplitude reduction.
Case 2
“Mr. B” was a 47-year-old man who underwent bilateral STN DBS for left-side predominant symptoms, motor fluctuations, freezing of gait, and dyskinesia. He had a history of dopamine dysregulation syndrome (manifested by compulsive use of dopaminergic medications), which had resolved 1 year previously, and a family history of bipolar disorder.
On postoperative day 2, Mr. B exhibited transient disinhibited behavior and lack of empathy. Stimulation was initiated after 3 weeks. Twelve weeks later, the patient reported he achieved best motor control in a monopolar configuration, although he was also more impulsive, irritable, and anxious, and talked faster with this setting. Medications were reduced and programming changes were made. At the 1-year follow-up, the patient’s family reported that he appeared to be “revved up” at times, particularly with one specific DBS configuration. He described episodes lasting 10–15 minutes when he became impulsive and irritable, talked faster than usual, and had trouble focusing. Two of these episodes were associated with panic attacks. These symptoms occurred when he temporarily increased DBS amplitude (usually before physical exercise) and resolved when he lowered it. Mr. B also shared he was taking a lot more medication than prescribed. Of note, the revved-up episodes only occurred when he increased stimulation.
Mr. B’s treatment included close psychiatric follow-up, with a behavioral intervention targeting dopamine dysregulation syndrome (keeping daily logs of the exact medication doses and times when he was taking them, taking medication no more than eight times daily, using a pill box, and involving his wife to ensure accountability) and supportive therapy. Dopaminergic medications were reduced. Over time, both the elevated mood and dopamine dysregulation syndrome symptoms resolved. Five years after surgery, the patient’s mood remained stable, and his motor symptoms were overall well controlled.
Case 3
“Mr. C” was a 53-year-old man with no past psychiatric history who underwent staged bilateral STN DBS surgery for left-side predominant motor symptoms. His right STN was implanted initially, followed by the left STN 1 year later. One month after left STN implantation, stimulation was initiated in monopolar mode; however, Mr. C was unable to tolerate higher voltages due to dyskinesia. A new bipolar group was created, with instructions to slowly increase voltage at home. Four months postoperatively, the patient reported significant stimulation-induced mood changes starting shortly after surgery: he felt “easily agitated, frustrated, snappy, short-tempered, and irritable.” For example, when his computer malfunctioned, he punched and broke the screen. These symptoms occurred in both monopolar and bipolar settings but were more severe in monopolar mode and with voltage >2.0 V; they resolved if the left DBS was turned off. Symptoms were addressed by using a double monopolar configuration and avoiding contact 1. Irritability persisted but ultimately improved with the addition of escitalopram 10 mg daily.
Case 4
“Mr. D” was a 65-year-old man who underwent staged bilateral STN DBS for primarily right-sided motor symptoms. He had a history of depression and dopamine agonist-induced impulse-control disorder (consisting of gambling, risky investments, and compulsive behavior surrounding household projects). Impulse-control disorder symptoms resolved after weaning off the dopamine agonist medication. The patient initially underwent left STN DBS surgery with excellent motoric response. His right STN was implanted 2 years later, after he developed left-sided symptoms.
Right STN stimulation was initiated 1 month later. Mr. D was given two monopolar groups (active contact 1− and contact 3−, respectively) to try at home. He reported best tremor control with contact 1; however, within a few hours, he developed dyskinesia and mood changes, described as “goofiness, restlessness, and acting giddy.” These symptoms did not occur when using contact 3; however, his tremor was insufficiently controlled. With slow upward amplitude titration, Mr. D was able to tolerate gradually higher voltages (up to 2.9 V) using contact 1 without recurrence of elevated mood.
Case 5
“Mr. E” was a 73-year-old man who underwent left STN DBS for right-sided motor symptoms and motor fluctuations. He had a history significant for major depressive disorder with onset in his youth, with a recent recurrence after receiving the PD diagnosis. He also had history of hallucinogen use, in full sustained remission, and a family history significant for mood and alcohol use disorders; four relatives had died by suicide.
Following initial programming in a monopolar configuration, Mr. E was instructed to slowly increase the amplitude (by 0.1 mA/day) at home. Ten days later, the patient reached 2.0 mA with good motor control. However, family members reported he became “hyper,” with reduced sleep and pressured speech. He was researching PD-related information online for 10–12 hours daily and believed he was going to “solve the problem of PD.” He also reported flight of ideas. The DBS device was turned off for 1 week and medications were adjusted.
Over the next 8 weeks, Mr. E slowly increased stimulation to 2.8 mA but became more disorganized, impulsive, and irritable. He was reprogrammed to a bipolar configuration and DBS amplitude was reduced to 2.5 mA. Elevated mood symptoms resolved, although motor control was insufficient. On his own, Mr. E increased the amplitude and again developed pressured speech, talkativeness, and increased goal-directed activity. These symptoms occurred at amplitudes >3.1 mA and improved when the DBS was turned off. Reducing amplitude to 2.5 mA led to lasting symptom resolution.
Table 2 highlights management strategies for stimulation-induced elevated mood, specifying those used in the five patients presented here (
12–
15,
17,
18).
Discussion
We described five men with PD treated with STN DBS who experienced stimulation-induced elevated mood states during or shortly after DBS programming changes, lasting minutes to weeks. Elevated or irritable mood was distinct from baseline and associated with psychomotor agitation, as well as a variable combination of additional symptoms. These episodes occurred after crossing a stimulation threshold, typically when using ventral contacts. In contrast to prior studies, where amplitudes >3.0 V have been noted, four of our patients experienced elevated mood symptoms at amplitudes as low as 1.0 V or 1.0 mA. Occurrence of elevated mood episodes at lower amplitudes in patients with adequately placed DBS leads suggests the presence of underlying risk factors that need to be better understood in order to anticipate and manage this potentially harmful complication of DBS therapy. Screening for risk factors for stimulation-induced elevated mood should be part of every DBS candidacy evaluation.
All four patients who developed elevated mood at lower amplitudes had psychiatric comorbidities that likely increased their sensitivity to the physiological effects of DBS, making them more prone to mood destabilization when stimulation was initiated or increased. One patient had a history of impulse-control disorder, one had concurrent dopamine dysregulation syndrome, and a third had a history of substance use; two (possibly three) also had a family history of bipolar disorder. Addictive disorders such as impulse-control disorder, dopamine dysregulation syndrome, and substance use disorders involve activation of dopaminergic reward pathways and may increase the risk of stimulation-induced elevated mood (
19,
20). Although DBS can be an effective treatment for impulse-control disorder, mainly by dopaminergic agent dose reduction, STN stimulation can impair decision making and increase impulsivity (
21). Three of our patients had mild cognitive deficits, which may additionally reflect dysfunction of the frontal subcortical networks. Recent studies have shown that both structural (prefrontal cortex atrophy) and functional (preoperative frontal function scores in patients undergoing STN DBS) indicators are associated with severity of neuropsychiatric symptoms in patients with PD (
22,
23).
Individuals with a family history of bipolar disorder may have a bipolar diathesis—that is, a genetic vulnerability that predisposes them to mood destabilization with antidepressant treatment (medications, electroconvulsive therapy, or phototherapy). Akiskal et al. (
24) coined the term “bipolar III” to describe antidepressant-induced hypomania in previously depressed individuals. In a study of approximately 500 patients, those classified as bipolar III were more likely to have a family history of bipolar disorder and completed suicide, a personal history of psychosis and suicidality, and greater chronicity and severity of depressive symptoms, compared with patients with bipolar II disorder (
24). Herzog et al. described a 65-year-old woman with PD and no past psychiatric history but a family history of bipolar disorder who underwent thalamotomy for tremor, followed by bilateral STN DBS (
13). This woman developed a sustained stimulation-induced episode of mania with psychosis, requiring treatment with clozapine and carbamazepine.
Risk factors for stimulation-induced elevated mood are summarized in
Table 3, based on literature review and the present case series (
7–
11,
14,
15,
17,
25).
The STN has been classically divided into three functionally distinct territories: the dorsolateral motor territory, the ventromedial associative part, and the medial limbic territory (
26). Although the precise pathophysiologic mechanism of STN stimulation-induced elevated mood is not fully elucidated, it has been postulated to result from the spread of current into neighboring areas such as the medial limbic territory or ventromedial associative region (particularly with monopolar configurations). This can lead to dysregulation of the mesolimbic dopamine circuits involved in emotional regulation, addiction, and learning (
14,
15,
20,
27). Using diffusion tensor imaging, Coenen et al. (
28) demonstrated recruitment of median forebrain bundle (MFB) fibers. The MFB is a crucial component of the mesolimbic dopamine system and therefore the reward circuitry; it has been implicated in mood disorders, addictive behavior, and learning (
15). Coenen et al. hypothesized that elevated mood states are caused by inadvertent MFB activation (
28).
In another original investigation, Schilbach et al. (
11) studied regional cerebral flow (rCBF) using positron emission tomography scan in one patient with STN DBS while activating a ventral contact (which had caused an elevated mood state) compared with using a more dorsal contact (mood-neutral). They found a differential increase of rCBF in the right dorsolateral prefrontal cortex, right middle temporal gyrus, and dorsal anterior cingulate cortex, suggesting that patients with underlying psychiatric vulnerabilities may develop a pattern of activation similar to that seen in bipolar disorder (
11).
Several case reports have also shown that leads placed ventral to the STN (in the substantia nigra) may result in acute depressive or elevated mood states (
3,
9). Correct lead location was confirmed in all our patients. It is important to note that the contacts that caused elevated mood were associated with optimal motor benefit in several cases. As such, multiple strategies may be necessary to address elevated mood while maintaining motor benefit (
Table 2). Newer directional leads, DBS systems with multiple independent current sources, and future adaptive DBS paradigms may also be useful in optimizing motor benefit without negatively influencing mood.
One of the patients we described (Mr. B) had severe co-occurring dopamine dysregulation syndrome. He experienced elevated mood symptoms on postoperative day 2, before his DBS device was activated. The high dopaminergic load in a patient with bipolar diathesis, perhaps combined with the lesional effect, were most likely contributors. Dopaminergic agents can independently destabilize mood or have additive effects to DBS, which is why it is paramount to address both comorbidities (
15).
Elevated mood symptoms recurred in four of our five patients when the same contacts were used. One patient (Mr. D) was able to tolerate gradually higher voltages with the same contact. This may indicate a habituation process, as suggested by Tommasi et al. (
5), allowing patients to safely use configurations that initially caused adverse effects if amplitude is slowly increased. Slow stimulation titration can be helpful with patients at higher risk of developing postoperative elevated mood.
None of the patients described here met DSM criteria for bipolar disorder (
29). The use of the terms hypomania or mania in previous reports was technically inaccurate, since duration criteria for hypomanic or manic episodes were rarely met and adjusting DBS settings or turning off the device usually led to symptom resolution (
7–
10). In DSM-5, criteria for hypomanic or manic episodes are met when symptoms occur in response to antidepressant treatment (e.g., medications, electroconvulsive treatment) but persist at a fully syndromal level beyond the physiological effect of that treatment (
29). As DBS therapy becomes more frequently used for the treatment of both movement disorders and psychiatric conditions, the inclusion of stimulation-induced elevated mood in the DSM warrants consideration. Proposed criteria for this new clinical entity are presented in
Table 4. This phenomenon has also been described in patients treated with DBS for treatment-refractory major depressive disorder or obsessive-compulsive disorder involving stimulation of the anterior limb of the internal capsule or the ventral internal capsule/ventral striatum (VC/VS) (
30–
32). Interestingly, in a series of 20 patients with VC/VS DBS, a personal or family history of bipolar disorder did not predict occurrence of stimulation-induced hypomania (
32).
Limitations of this study include selection bias; that is, only patients with severe behavioral symptoms are referred for psychiatric evaluations, and thus not all patients who had DBS during the study period were included. All patients in this case series had STN DBS, which has been linked to mood dysregulation more than the GPi target. Future studies should include systematic multisite reviews of all patients treated with DBS during a certain period, followed prospectively. Strengths of our study include the detailed documentation of programming settings, availability of psychiatric evaluations, and length of postoperative follow-up (up to 5 years), allowing the longitudinal tracking of symptoms and evaluation of the effectiveness of management strategies.