Introduction

Although categorized as a movement disorder, Parkinson’s disease (PD) is a neurobehavioral disorder [1] and the importance of the behavioral components should not be underestimated [2]. The nonmotor behavior symptoms in PD are the major determinants of health-related quality of life in most reports from all parts of the world [2, 3], reflecting the fact that these behavioral problems are not culturally based. Studies have shown that the greatest stress for caregivers arises from behavioral rather than motor problems, with the corollary observation that the most important precipitants for nursing home placement are behavioral problems, not motor [4]. Yet the vast majority of treatment trials have focused on the motor aspects. The increasing number of research studies on behavior in PD reflects the increasing recognition of the importance of these problems.

Behavioral problems in PD are particularly difficult to study, more so than behavior problems in psychiatric disorders because there are obvious interrelationships between the motor problems and the behavioral. It has been well known since the disease was described that tremor and other motor problems vary with anxiety. In addition, there are definite interrelationships among the various emotional states themselves, such as sleep, fatigue, depression and anxiety. Complicating the picture even further is the overlap among the various disorders. For example, fatigue is associated with depression, i.e. people with fatigue are more likely to be depressed than nonfatigued PD patients, yet many fatigued PD patients are not depressed. However, fatigue is one of the defining features for major depression, as defined in the Diagnostic and Statistical Manual of Mental Disorders, 4th Edn, Text Revision (DSM-IV-TR) [5], the standard catalogue for defining mental illnesses. Fatigue may be associated with sleep dysfunction and vice versa, but they may be occurring as separate but overlapping syndromes. Apathy and sleep disorders are symptoms of depression, yet these are also syndromes in PD that may be quite distinct from depression. Understanding these relationships is crucial not only for an understanding the individual patient’s disorders but also for choosing treatment. Fatigue in the setting of depression may suggest the use of an antidepressant, whereas fatigue unrelated to a mood disorder may initiate the use of a stimulant medication. Patients have anxious, depressed, irritable or other unpleasant feelings when their motor condition is “off” but may also have these feelings when their motor condition is “on.” These symptoms are often called “nonmotor off” and have an uncertain relationship with dopamine activity [6].

There are two important underlying aspects for understanding the many behavioral problems in PD. Aside from the complexity of the phenomenology and their interrelationships, the various behavioral problems in PD are mostly independent of dopamine. For many years, PD has been simplistically considered a dopamine deficiency disorder, and the vast majority of treatment trials have focused on improving dopaminergic stimulation in the brain, or compensating for decreased dopamine, by altering glutamate or acetylcholine transmission. While these interventions have improved motor control, they have not been helpful in treating behavioral problems. The second is that, unlike the motor problems, the behavioral problems are often not progressive. This suggests different disease processes than are usually considered for the neurodegenerative disorders.

Personality

An intriguing but difficult question is whether there is a “Parkinson personality” [7] and, if so, which phenomena may be intrinsic to the disorder and which are due to the effects of the disease. There is a long but unknown duration of PD pathology present in the brain before the motor stigmata surface. The increasingly recognized “herald” symptom of rapid eye movement (REM) sleep behavior disorder being occasionally present for decades before motor symptoms makes it likely that some personality changes are part of the disease process. In addition, one of the most robust epidemiological associations for PD risk has been the inverse association with cigarette smoking [8]. Since cigarette smoking is likely one manifestation of personality, perhaps a measure of addiction potential, it is likely that this alone supports the existence of certain personality traits having an association with PD. Fatigue affects about one-third of patients by the time their PD is diagnosed [9], and both anxiety and depression are common. Olfactory impairment affects not only smell but also taste, thus perhaps reducing some enjoyable aspects of life, leading to subtle changes in habits and thus personality. Since these aspects of personality often appear before motor symptoms, and appear sufficiently often in the general population that they cannot be interpreted at the time as part of a disease process, they are seen as personality changes of aging. In the heyday of psychoanalysis, certain personality traits were thought to be risk factors for PD [10, 11], and some signs were interpreted as psychogenic rather than physiological.

Retrospective personality studies have shown differences between PD and non-PD, but these are obviously subject to recollection bias [12–14]. A large prospective general psychological study of patients in Minnesota followed almost 7000 patients for 40 years. Only anxiety present early was associated with later development of PD [15]. Depression and pessimism early on were not associated with an increased risk of PD. [15], and although several reports have described a reduction in novelty seeking in PD patients, a prospective study did not find this to be predictive of later development of PD [16].

Post-hoc studies of premorbid personality in PD patients have purported to find them more industrious, inflexible and cautious, with a low risk of impulsive decision making. Studies looking at contemporaneous personality differences between PD and controls generally found reduced novelty seeking, greater or equal levels of harm avoidance, and equal levels of reward dependence and perseverance [7]. Of note, however, is a small study of nine couples of monozygotic twins discordant for PD, which found no personality differences [17]. Poletti and Bonuccelli [7] concluded their review on the topic of personality in PD by noting that “empirical evidence does not robustly support the existence of a characteristic premorbid personality profile in PD patients.”

However, given the inverse association of PD with cigarette smoking, the lengthy duration of REM sleep behavior disorder and the increased premorbid development of fatigue, anxiety and depression, it seems likely that personality differences have not been identified due to methodological limitations, depending on how long before the onset of motor changes the personality changes are looked for.

Apathy

Apathy is a behavioral state that has variable definitions. In common parlance, as defined in dictionaries, it is “a state of indifference or the suppression of emotions” (Wikipedia), “the lack of interest or concern: lack of emotion or feeling” (Free Online Dictionary) and “a complete lack of emotion or motivation” (Urban Dictionary), but in psychiatric literature it is considered both a syndrome [18–20] and a symptom. In the DSM-IV-TR [5], there are two categories of major depression, melancholic, in which patients feel sad, and anhedonic, in which patients are unable to feel pleasure. The latter condition could be considered an apathetic depression.

Apathy is generally considered a disorder of motivation, therefore having a relationship with fatigue, as well as depression. Apathy is common in dementing disorders such as Alzheimer’s disease, dementia with Lewy bodies, PD and frontotemporal dementia, as well as in primary psychiatric disorders such as schizophrenia and major depression. Apathy is also a syndrome [20–22] (Box 13.1) Understanding the difference between a syndrome and a symptom is perhaps best appreciated by considering depression. One is thought to be depressed if sad; however, the syndrome of depression may involve several other symptoms such as fatigue, sleepiness, amotivation and irritability.

Data on apathy in PD has been inconsistent, presumably reflecting differences in assessment and populations. Prevalence rates in PD have been reported to be between 7 and 70% [23]. Similarly discrepant overlap rates for apathy, dementia and depression have been published. One report from the USA found that 29% of 80 PD patients had abnormal Apathy Scale scores but normal depression scale scores [24]. In another country, 52 of 164 PD patients were deemed apathetic, of whom 35% were depressed but not demented, 8% demented but not depressed, and 48% had both dementia and depression. Only five patients suffered from apathy alone [25]. In France, 56% of 159 PD patients were apathetic but only 9% had apathy without depression or dementia [26]. Apathy correlated, in general, with cognitive dysfunction and not motor severity, suggesting a nondopaminergic neurochemical basis. In a Scandinavian study of 2323 PD patients, apathy was noted in 38% but most had cognitive dysfunction or were on medications for depression or psychosis [27]

Apathy has been associated with reduced quality of life, depression, dementia and fatigue [18]. Specific criteria for a diagnosis of apathy in PD have been proposed [28], requiring a reduction in motivation, reduced goal-directed cognitive activity and reduced emotional display, both spontaneous and evoked. It is interesting that these medical criteria sometimes minimize the more general perception of the meaning of apathy as expressed in dictionaries, which likely reflect the nonmedical understanding of the term, which is a lack of emotions. Apathy is primarily a self-reported symptom that is partly but not completely reflected by outside perception. A person who shows no emotion and has restricted capability for movement may seem apathetic yet may not be.

In clinical practice, the major differential in classifying apathy is with depression. Since the two syndromes include reduced motivation, the clinician often cannot be confident of the diagnosis. One of the unusual features of apathy in PD, since it so often co-exists with dementia, is that it sometimes provides emotional insulation from a very restricting condition. In many cases, the apathetic patient is far less emotionally disturbed by his condition than the family. The patient often does not appear to care about being unable to do things that formerly gave pleasure and appears content, or at least not disturbed by, a very restricted existence.

The natural course of apathy has not been studied. It is likely that apathy related to dementia is progressive, whereas apathy associated with depression may not be. This is, however, quite speculative.

Apathy as a response to acute dopaminergic withdrawal has been described [29], and therefore may respond to reintroduction of the discontinued medication. Apathy may also be a potential side effect of psychoactive medications, including antidepressants, antipsychotics and benzodiazepines, so medication reviews should be undertaken with this in mind.

Treatment for apathy, in general, is unknown. Anecdotal reports and theories suggest that dopaminergic agents and stimulants may be helpful, but no convincing data exists. Cholinesterase inhibitors are thought to produce mild benefits in apathy in demented patients, and since the majority of PD patients with apathy are demented, this should be entertained. When apathy is considered a problem for the patient and depression is considered a possible concomitant syndrome, a trial of an antidepressant seems reasonable. In clinical practice, it may be extremely helpful simply to provide counseling to the family, explaining that apathy is a common feature of PD, has no clear treatment and may, in fact, be protective.

Anxiety

Anxiety is a common problem in PD [30, 31], affecting over 40% [30] and generally is underappreciated [32]. This underappreciation is reflected both in movement disorder specialist evaluations and in research publication numbers. There are only rare reports of treatment trials [30]. An interesting study that should be of great interest to doctors who care for large numbers of PD patients found that anxiety was associated with an increased rate of telephone calls to the treating neurologist [33].

Anxiety is defined as excessive nervousness or fear. There are several subcategories of anxiety. The classification system has been significantly altered from DSM-IV-TR to DSM-V [34], and is likely not widely understood by neurologists. Anxiety is common in the general population but more so in PD. Its epidemiology is also different. In the general population, anxiety generally appears in young adults and is highly skewed toward women. In the PD population, it arises a few years before or after the onset of the motor symptoms and affects the genders equally. The published prevalence estimates for anxiety in PD vary from 4 to 40% [23, 31, 35]

Until the diagnostic categories changed in 2013 with the publication of DSM-V [34], the most common anxiety disorders affecting PD patients were generalized anxiety, panic disorder and social phobia. In the new nosology, generalized anxiety disorder, a condition in which people were excessively nervous in general, does not exist, as the long-term outcome of this disorder indicated that generalized anxiety was a transient condition that evolved into another disorder, most commonly depression with anxiety. Social phobias are excessive fears related to social interactions. Of course, in PD, many patients shun a variety of social interactions because they may have dysarthria, which makes them very self-conscious about participating in conversation, or they may be self-conscious because of drooling, dyskinesias or tremors. They may avoid crowds because of the risk of falling or a distaste for the commotion associated with a large group. Panic attacks refer to acute episodes, generally not provoked by identifiable precipitants, during which people feel chest pain, shortness of breath, a sensation of immanent death, tingling, and numbness or other somatic symptoms. Data on anxiety in PD available for this chapter was published before the classification system changed, so the data presented in this chapter will require re-interpretation within a few years. For the purpose of most treating physicians and family, recognition of anxiety, panic or phobia and their associations with PD are the important issues, not classification.

In a 1-year cross-sectional study performed in Europe, the USA and Australia, 34% of the 340 subjects attending university referral centers were classified as having at least one DSM-IV anxiety disorder, and 12% had more than one [30]. Multiple anxiety disorders were more prevalent than any single anxiety category. Generalized anxiety disorder was the most common, followed by agoraphobia without panic or social phobia. Twenty-six percent of those with anxiety disorders also suffered from major depression, while 13% suffered from dysthymia, a milder form of depression. Ponton et al. [36], using very different evaluations in an American population, found equal levels of social phobia and multiple anxiety disorders, but more panic. An Australian study [31] of 79 PD patients diagnosed one-quarter with anxiety disorders, half of whom also suffered from depression. These numbers are significantly higher than the prevalence in the age-matched general population. Although the Australian study found that anxiety prevalence was greater in the younger patients, and that there was no relationship with PD medication use, these associations have not been adequately studied. Whether there is any relationship between anxiety and the side of worse motor symptoms, as reported in one study [37], is unclear [31].

Although “nonmotor-off” symptoms have frequently been associated with the “off” motor state, Leentjens et al. [38] studied 250 PD patients of whom 118 had motor fluctuations. They found that the fluctuators were more likely to have generalized anxiety, but the majority with anxiety did not have fluctuations. Those who reported fluctuations in levels of anxiety generally were more anxious when “off,” although some were more anxious when “on.”

There are no medications approved by the US Food and Drug Administration (FDA) for the treatment of anxiety in PD. There have been few studies of anxiety in PD. In some studies of antidepressants in PD, anxiety was also evaluated, but these studies were underpowered, meaning that the design of the trial was not centered on anxiety, so that insufficient numbers of anxious subjects might have been recruited to make definitive statistical inferences on drug effect. In these studies, the antidepressants, which are approved by the FDA for treating anxiety as well as depression in the general population, were not beneficial for anxiety in PD. The medications used for treating anxiety in the general population are typically benzodiazepines, buspirone, trazodone and selective serotonin reuptake inhibitors (SSRIs). Benzodiazepines are considered relatively contraindicated in the elderly due to the risk of confusion and falls. On the other hand, the SSRIs need to be used for 2 weeks or more before antianxiety effects begin to take effect, and they cannot be used on an as-needed basis. They also have the possible advantage of not causing sedation. The benzodiazepines may be used on an as-needed basis and may be particularly useful for insomnia associated with anxiety. In addition, clonazepam is commonly used to treat REM sleep behavior disorder.

Psychotic symptoms

Psychosis has generally been understood as a disorder characterized by the loss of reality testing. This classical definition has been revised in favor of characterizing disorders by the nature of their psychotic symptoms, such as hallucinations and delusions. Parkinson’s disease patients may develop these symptoms without medications, but typically they occur on medications for their motor problems. The relationship between medications and symptoms has been an area of interest because of the identical phenomenology seen in patients with dementia with Lewy bodies who are not taking anti-parkinsonism medications. In addition, in rare patients with long-standing PD that has never been treated, hallucinations have occurred [39]

Visual hallucinations affect approximately 20–30% of drug-treated patients; however, prevalence increases with the duration of follow-up [40]. In a study of 386 patients in Holland, 21% had hallucinations at baseline, but within 5 years, 45% of the ones who had not suffered hallucinations developed them [41]. Auditory hallucinations may occur as well but are generally about half as common, and usually occur in patients who are having visual hallucinations [40]. Unlike hallucinations that occur in primary psychiatric disorders such as schizophrenia, major depression or mania, the hallucinations are typically visual, whereas the primary psychiatric disorders are auditory, and, quite strikingly, mostly have no emotional import. Schizophrenic hallucinations are characterized by voices criticizing the patient, either directly or in the form of conversations about the patient. Parkinson’s disease patients generally see people or animals, which are not threatening and usually ignore the patient. For example, they may see children playing in the living room, two adults watching television, marching bands outside in the yard or bugs on the wall. The hallucinations may be in color or black and white, vivid or somewhat obscured, and may last from seconds to minutes. The hallucinations tend to be stereotyped, the patient generally seeing the same people doing the same thing each time. The auditory hallucinations may be voices, music or mechanical sounds, and tend to be indistinct, like a radio playing in another room. Tactile, olfactory and gustatory hallucinations may occur but are considerably less common. Two unusual types of hallucinations are notable [42], “presence hallucinations” and “passage hallucinations.” “Presence hallucinations” are not true hallucinations. They are a strong feeling that someone or something unseen is behind or to the side of the patient. These are not threatening. The patient feels that someone else, or possibly a pet, is in the room and turns to look. There are also “passage hallucinations,” the visual perception of people, animals or shadows passing quickly in the peripheral field, which always disappear on direct inspection. Some patients will also report reflections off their eyeglasses although there is no light source.

Delusions are irrational beliefs. These are not based on data or they are grossly distorted interpretations of real events. Unfortunately, in PD, the delusions tend to be paranoid [43]. The most common delusions are “jealous” delusions, with patients believing that their spouses are having sexual affairs. Other delusions also tend to be paranoid in nature, such as thinking people are living in the home or that the patient is being spied upon. Institutionalized patients will complain privately of being manhandled by staff, or poisoned, or that a staff member is selling drugs, supplying sex or selling body parts.

In Holland, about 30% of hallucinators also had delusions [41]. A retrospective chart review of 1453 patients in one clinic found that psychotic symptoms increased in frequency with increasing duration of disease, so that the average was 27% but by 12 years of disease the prevalence was 34% [44]. Other reports have provided similar estimates.

Psychotic symptoms of PD are quite different from schizophrenic symptoms. Parkinson’s disease patients do not experience delusions of grandeur. They do not believe themselves to be messengers of God. They do not hear voices commenting on them or believe that they are being controlled by aliens. In PD, the hallucinations have no emotional import, unlike the case with primary psychiatric disorders.

Several reports have mentioned conflicting risk factors for psychotic symptoms, but all agree that dementia is a risk factor, probably the most significant risk factor, and that the occurrence of visual hallucinations increases the risk of dementia developing within the next few years. Visual impairment is considered a risk factor for hallucinations [40]. A 2013 study reported many risk factors, including older age, depression, daytime sleepiness and others [41]. It must be noted that psychotic symptoms often develop without any change in medications or obvious change in medical status, so that a patient who had been stable and doing well for years may report new-onset hallucinations. This presumably reflects brain changes associated with disease progression, not a change in pharmacokinetics.

The evaluation of psychosis requires a review of all psychoactive medications to be certain that nonneurological medications, particularly anticholinergics for bladder control, are not contributors. Medical contributors such as occult infection and endocrine and metabolic disorders also need to be considered. When these have been excluded, then treatment requires a reduction, if possible, in PD medications, all of which may be contributory. There have been no studies comparing treatments for psychosis and no studies looking at dose reductions of PD medications. Double-blinded trials comparing levodopa (l-DOPA) and dopamine agonists have shown that dopamine agonists are more likely to cause hallucinations. It is therefore generally recommended that anticholinergics be tapered and stopped first, then amantadine and then dopamine agonists before levodopa. It is this author’s belief that one medication should be tapered and stopped rather than having all medications reduced, the notion being that medication side effects may be synergistic.

After PD medications have been lowered to their lowest tolerable dose, then, if psychotic symptoms persist, an antipsychotic should be added. Clozapine, quetiapine, olanzapine and pimavanserin are the only medications that have been subject to double-blind, placebo-controlled trials [40]. Olanzapine was clearly shown to worsen motor function and not to be helpful for controlling the psychosis. Clozapine, in two multicentered, double-blind, placebo-controlled trials, was shown to be extremely helpful for the psychosis, without motor worsening, and even with significant improvement in tremor. Quetiapine failed to show efficacy in three double-blind, placebo-controlled trials, although open-label trials have been quite successful. However, all of the quetiapine trials have shown no worsening of motor function. The only successful double-blind, placebo-controlled trial was too small to be interpretable. The other atypical antipsychotics have had variable results in terms of worsening parkinsonism, with at least some reports of worsened motor function with all, including aripiprazole, a partial dopamine agonist. In some countries, clozapine has been accepted as a government-approved medication for PD psychosis but not in the USA. The American Academy of Neurology task force recommended consideration of both quetiapine and clozapine. The Movement Disorder Society Task Force on nonmotor symptoms recommended clozapine [45]

Clozapine requires a complete blood count weekly for 6 months, every other week for 6 months and then every month thereafter, due to the risk of agranulocytosis, which is believed to be 1% in young adults and possibly higher in older healthy adults. Since this risk is independent of dose, the low doses used do not ameliorate the risk. Clozapine is generally effective at doses between 6.25 and 50mg/day. The time of greatest risk for agranulocytosis is the first 3 months.

Some experts have advocated the use of cholinesterase inhibitors to treat psychotic symptoms. In studies of PD dementia and dementia with Lewy bodies, these agents have been associated with a reduction in hallucinations; however, there have been no large studies of their effects when used directly to treat hallucinations or delusions. Their use is predicated on the association between dementia psychotic symptoms and the common observation that anticholinergics may induce psychotic symptoms in PD.

An entirely different pharmacological approach to treating psychosis was based in part on the observation that all of the second-generation antipsychotic drugs, in addition to blocking dopamine D2 receptors, also blocked the 5-HT2A serotonin receptor. Other observations have long linked serotonin to psychotic symptoms, so trials of pimavanserin, a 5-HT2A “inverse agonist” were undertaken. While the initial phase 3 and earlier phase 2 trials yielded results that were equivocal in terms of efficacy, all studies demonstrated that the drug was safe, without worsening of motor function, and had a side effect profile similar to placebo. Based on the initial trials, a very rigorous phase 3 trial was undertaken, with positive results [46]. At the time writing, the manufacturer is submitting data to the FDA for approval of this drug to specifically treat PD psychosis.

Depression

Depression is well known to be a common problem in PD, but the estimates for its prevalence vary significantly for a number of reasons. Primarily, the diagnosis of depression in the PD patient is a challenge. The markers generally used to recognize depression in the general population are facial expression, psychomotor retardation, soft voice, stooped posture and reduced gestures, all of which are intrinsic to the motor aspects of the disease. If one considers the requirements for the diagnosis of major depression in the general population using DSM-IV-TR criteria [5], fatigue, sleep disorders, psychomotor retardation and weight change, four of the required five symptoms, are already present in most PD patients, regardless of their mood.

The diagnosis of depression in PD rests primarily on mood, but useful symptoms include anxiety, which is, more often than not, associated with depression in PD, disability out of proportion to motor deficits [47], irritability and loss of interest in activities that should not be significantly compromised by motor dysfunction.

Depression is common in PD, regardless of the criteria used [48], with over 50% suffering from it in some form [49], with 17% having major depression and 35% having less severe problems [49]. Depression is associated with worse prognosis for both motor and cognitive function [50–52] and a greater effect on quality of life than motor dysfunction [53].

The course of depression is difficult to assess for many reasons. In a study that followed 413 early, mild, untreated PD patients who were participants in two observational (nontreatment) PD studies, 23% were rated as depressed on the Geriatric Depression Scale (GDS) at some point during the study [54]. Almost half were treated with antidepressants. Clinically significant depression remitted in half of the cases within 6 months. Those with mild depression were six times more likely to develop more severe depression than those who had not been depressed. Paradoxically, the use of antidepressants was associated with a reduced rate of remission of depression. In the general population, depression generally resolves spontaneously, although it may recur. It is unknown if this is true in PD.

The pathophysiology of depression remains unknown. For many years, the etiology of depression was debated, with one side advocating depression as a reaction to the diagnosis of PD, and the other maintaining that depression was an intrinsic disorder related to neurodegeneration in specific brain regions. Most authorities believe that depression occurs for both reasons. The reason for the opinion change was the discovery of the importance of basal ganglia structures in mood, uncovered by the results of the many deep-brain stimulation treatments for motor problems in PD. Abnormalities of serotonin and norepinephrine have been suggested but not confirmed [47]

Although depression is common in PD and had been the focus of discussions for many years, when a review article was published in 2013 [48], only six randomized controlled trials of antidepressants had been published, five of which involved placebo. This meta-analysis concluded that “antidepressant medications were not efficacious in depression in PD” [48]. This result supported a similar conclusion in 2005 [55], although the multicenter, placebo-controlled, three-arm study by Richard et al. [56] concluded that both venlafaxine and paroxetine were statistically more beneficial than placebo. However, the placebo response was so great in the study that the benefit of the active drug was diminished, as illustrated by the number needed to treat being 14 and 22 for the active drugs. None of the trials were longer than 12 weeks. A much larger double-blind, placebo-controlled study of pramipexole [57], a dopamine agonist used for treating the motor symptoms of PD, found a statistically significant but small antidepressant result, which, using the statistical technique of path analysis, was attributed to a direct antidepressant effect.

Although there are multiple statements in the literature decrying the undertreatment of depression in PD, it is unclear that treatment works well, and many patients treated long term with SSRI antidepressants remain depressed. There are no clear recommendations on treatment. The use of SSRI medications is generally preferred by authorities because of their more benign side-effect profile, but some data suggest that tricyclic antidepressants may work better. Most antidepressants have antianxiety properties so that the two often concurrent syndromes may be treated with a single agent. Often, the medication choice is dictated by the side-effect profile. Medications that cause sedation may be useful in the patient who sleeps poorly at night. Drugs that increase appetite and cause weight gain may be ideal if a patient is losing weight. Tricyclics generally have anticholinergic properties that may be useful in a patient with drooling or an overactive bladder. Antidepressants in other chemical families – noradrenergic, serotonergic reuptake inhibitors, bupropion (a dopamine reuptake inhibitor that does not enhance motor function), mirtazapine (a drug that increases noradrenergic and serotonergic transmission) – are also commonly used. Selective monoamine oxidase B (MAO-B) inhibitors approved for treating motor symptoms of PD do not have mood-altering properties unless used at high doses, which render the drugs nonselective. Nonselective MAO inhibitors are relatively contraindicated in PD because of the possibility of hypertensive crisis.

Electroconvulsive therapy has been reported in several small series to be helpful in treating refractory depression in PD [58] and has an added benefit of motor improvement, independent of improved mood. In general, motor improvement precedes improvement in mood but, unfortunately, does not last. The motor improvement lasts from hours to weeks but not more than that. There are few reports on maintenance electroconvulsive therapy, in which treatments are given once every few weeks, but sustained motor improvement does not seem to be obtained (J. H. Friedman, personal observation).

Psychosocial interventions, including cognitive behavioral therapy, education and psychodrama, have been studied in PD [59] and, although producing improvement, involved small numbers of subjects. Cognitive behavioral therapy is the best studied and has produced results that rival medication. The advantage of psychosocial intervention is the absence of additional medication. The disadvantage is the time and expense.