Introduction

Idiopathic Parkinson’s disease (PD) is the second most common neurodegenerative disorder after Alzheimer’s disease (AD), and motorically it is characterized by tremor, ridigity, bradykinesia, and postural instability. Whilst it was historically considered to be a movement disorder there are multiple non-motor symptoms, which often precede the motor symptoms by years or even decades.¹ These include dysautonomia, sleep disturbances, neuropsychiatric disturbances, pain, and sensory problems. These have a negative effect on quality of life and are associated with overall higher carer burden and, potentially, higher care costs whilst being frequently undeclared by patients.^{2, 3}

The motor syndrome is increasingly recognized as reflecting the ‘tip of the iceberg’ and, furthermore, Parkinson’s has been described as the ‘quintessential neuropsychiatric disorder’.⁴

Figure 8.1 The spectrum of neuropsychiatric disorders in Parkinson’s disease

This accurately reflects both commonality of neuropsychiatric symptoms (Figure 8.1) as well as the negative impact these disorders have on patients’ social functioning and ability to work.⁵ Alpha synuclein pathology and disruption of neurochemical pathways underlie some of the symptomatology, whereas adverse effects of treatments can precipitate other manifestations such as impulse control disorders and psychosis. Adequate recognition and a robust evidence base for management of these troublesome symptoms remain an urgent need; however, the literature in the field is expanding (Figure 8.2).

Figure 8.2 Graph showing number of articles published in each neuropsychiatric domain

Depression

In his original essay on the ‘Shaking Palsy’, James Parkinson noted that ‘A more melancholy object I never beheld’.⁶ Nowadays, depression is still recognized as a common non-motor feature of PD, with the prevalence of clinically significant symptoms being 35%.⁷ A multi-centre study demonstrated that half the study patients were depressed according to the Beck Depression Inventory (BDI) whilst only 1% had self-reported symptoms.^{8, 9} There are particular diagnostic challenges which arise from patients’ under-reporting of their depressive symptoms,¹⁰ hypomimia (lack of facial expression), concurrent mood disturbances such as anxiety or apathy,¹¹ psychomotor slowing, as well as from under-recognition by clinicians.¹² There are a number of validated screening tools available which may be used to identify depression in people with PD including the Hamilton Depression Rating Scale (HAM-D), BDI, Hospital Anxiety and Depression Scale (HADS), Montgomery Asberg Depression Rating Scale (MADRS), and the Geriatric Depression Scale (GDS). A working group on depression in Parkinson’s recommended that an inclusive approach be taken to symptoms (i.e. regardless of the aetiology) and that motoric ‘on’ or ‘off’ state should be considered when making a diagnosis.¹³

The underlying pathophysiology of depression in PD is poorly understood and likely to be multifactorial. A study of almost 1,500 patients concluded that higher rates of depression were seen in patients with increasing severity of motor symptoms.¹⁴ Whilst there may be a reactive element to living with a neurodegenerative condition, there is substantial evidence of a neurotransmitter deficit. This biochemical basis is likely mediated through alteration in serotonin, noradrenaline, and acetylcholine. These changes may have a cumulative effect with the hypodopaminergic state, with depressive symptoms being more prominent in the motorically ‘off’ state.¹⁵ Pharmacological trials have demonstrated an antidepressant effect with the dopamine agonists pramipexole and ropinirole.¹⁶

Treatment approaches to depression in PD should ideally be delivered within the framework of a multidisciplinary team. The choice of pharmacological therapy is often driven by determination of symptom severity and effect on a patient’s quality of life. The options in treating depression in PD are similar to that for depression in any chronic disease state and include selective serotonin reuptake inhibitors (SSRIs), serotonin noradrenaline uptake inhibitors (SNRI), monoamine oxidase type B inhibitors (MAOBIs), and tricyclic antidepressants (TCAs). However, there is a paucity of robust randomized controlled trial (RCT) evidence. Selective serotonin reuptake inhibitors are most commonly prescribed because of their more favourable side effect profile and lower potential for drug–drug and drug–disease interactions. The use of MAOBIs for the motor treatment of PD increases the risk of developing serotonin syndrome through the addition of an antidepressant, although this is rare in clinical practice.¹⁷ Whilst the use of TCAs have a potential advantage in promoting sleep in patients who suffer insomnia, the anticholinergic effects of this drug class and daytime somnolence should be considered.

Treatment of depression can be a complex matter, but there have been relatively few studies specifically on depression in PD and many of these have been of poor quality. A Cochrane systematic review ten years ago found ‘Insufficient data on the effectiveness and safety of any antidepressant therapies in Parkinson’s disease…’.¹⁸ A systematic review carried out in 2016 showed weak evidence in favour of SSRIs and TCAs; the positive results for the dopamine agonist pramipexole (mentioned earlier), but not for rotigotine, were noted; there were some promising results for CBT and conflicting results for transcranial magnetic stimulation (TMS); ECT produced good results in patients with severe depression, but no RCTs were available.¹⁹

Apathy

Apathy is one of the most common non-motor features of PD.²⁰ Increasing efforts have been made within the last decade to distinguish apathy as an isolated mood disorder from apathy as a by-product of depression, although the two symptoms frequently coexist.²⁰ Apathy may be defined as a ‘state of reduced motivation with decreased goal-directed behaviours’.²¹ This may result in low levels of activity, reduced interests, and a loss of socialization. The overlap in symptoms between apathy and depression can make diagnosing isolated apathy challenging. There are, however, certain characteristic thoughts and symptoms that are specific to apathy and the presence or absence of these help to confirm the diagnosis. These include emotional indifference and reactivity, reduced activity and interest in the world with a lack of concern for others.²² Depending on the diagnostic criteria used, and the concurrent symptoms of depression and cognitive impairment, the prevalence of apathy ranges from 17% to 70% in PD²³ and, similarly to depression, the symptoms may precede the motor symptoms of PD. The pathophysiology is poorly understood, but may be associated with a neuronal disruption in areas that regulate goal-directed behaviour: mainly dopaminergic projections between the frontal cortex and the ventral tegmentum.^{21, 24}

The development of diagnostic criteria for apathy has advanced significantly with the establishing of a task force, commissioned in 2008 by the International Parkinson’s and Movement Disorder Society, to assess the clinimetrics of the available rating scales.²⁵ Subsequent review of the clinimetric properties of 13 scales suggested that the five-item WHO Well-being Index (WHO-5) and Neurastenia Scale detect apathy severity, and the 33-item Lille Apathy Rating Scale (LARS) is valid in the diagnosis of 33 items, while the Starkstein Apathy Scale (SAS) has utility in the exclusion of apathy.²⁶

There is a range of therapeutic approaches to treating apathy including dopaminergic medications.^{13, 27} A double-blind RCT comparing the cholinesterase inhibitor rivastigmine with placebo showed a significant improvement in the symptoms of apathy at 6 months.²⁸

Anxiety

It is estimated that a third of people with PD experience symptoms of anxiety, which is considered to be a group of disorders consisting of generalized anxiety disorder (GAD), obsessive compulsive disorder, agoraphobia, social phobia, and panic disorder. It can be difficult to distinguish symptoms of anxiety from those of depression or even somatic PD (e.g. sleep disturbances, apathy), which can also be present simultaneously, so the identification of one should lower the threshold of clinical suspicion for the other. Classification of these conditions varies in studies, therefore the reported prevalence rates vary, ranging from 3.6% to 55%.^{29, 30} A systematic review reported the average prevalence of all anxiety disorders in PD to be 31%, with GAD, panic disorder, and social phobia being the most common.³⁰ This wide difference in prevalence may be accounted for by the under-reporting of symptoms by patients and under-recognition by clinicians.

Researchers will often rely on established criteria, for example from the DSM-IV, to make a diagnosis; however, diagnosis in a clinical setting can remain challenging. A study of 42 patients with PD in a university-based movement disorders clinic found that 29% had DSM-III-R anxiety disorder diagnoses, but an additional 40% had anxiety symptoms that did not meet the diagnostic threshold.³¹ Self-reported anxiety scales have been tested in PD, but their clinical utility is uncertain owing to the lack of consensus regarding appropriate cut-off scores.³² A detailed patient history and a collateral history are essential but often omitted because of time constraints in clinic. Furthermore, identification can be complicated by the presence of an underlying cognitive impairment. Specific symptoms of anxiety in PD include panic attacks during off periods, excessive worry, and increased subjective motor symptoms. The timing of symptoms can mirror motor fluctuations or manifest in an unrelated pattern.²³

Similarly to depression in PD, the aetiology of anxiety is multifactorial. There are neurobiological changes resulting in the development of anxiety and a proposed reactive model of anxiety secondary to the underlying burden of PD. Social anxiety is especially common in the context of the reactive model, with people with PD expressing concern over being negatively perceived in public leading to social withdrawal. Patients may also fear the progression of the disease, disability, institutionalization, and issues surrounding their death. With the high falls risk in the disease, an associated fear of falling can be very prominent.³³

The role of dopaminergic neurotransmission in the context of anxiety is poorly understood but has been linked to both social phobia and anxiety symptoms in animal models.³³ Seratonergic and noradrenergic systems also have a role in PD anxiety owing to their widespread distribution in the structures involved in emotional modulation. Research demonstrates that people with PD who score higher on anxiety questionnaires have a shorter serotonin transporter allele, highlighting the potential role of neuropsychiatric genetics in this group.³⁴

Anxiety in PD has a huge impact on quality of life, not only for the patient but also for their family and carers. Social anxiety can lead to isolation in the community which further exacerbates anxiety, creating a self-perpetuating cycle. It can also precipitate loneliness, particularly in older people, which can contribute to the subsequent development of depression. This can in turn affect treatment plans, hindering motivation and engagement in rehabilitation and healthcare services.³³

There is a paucity of evidence on the management of anxiety in PD. There are no robust clinical trials of either pharmacological or psychological strategies. Anxiety has been assessed as a secondary outcome in clinical trials for depression but because of the selection criteria used, any therapeutic effects on anxiety were diluted.³⁵

Guidance for managing anxiety in older adults with chronic physical health conditions³⁵ can be extrapolated, which places initial emphasis on lifestyle modification, focusing on optimal sleep, exercise, nutrition, and socializing in addition to eliminating any exacerbating medical causes for anxiety such as metabolic anomalies, nutritional deficiencies, and drug reactions. Pharmacological approaches are then advised with the introduction of SSRIs and SNRIs. The use of benzodiazepines is discouraged because of their side effect profile. Behavioural therapies in the form of cognitive behavioural therapy, mindfulness, and desensitization are also of benefit but are reliant on adequate motivation and adherence.³⁵

Psychosis

Psychosis in PD is a spectrum of neuropsychiatric manifestations consisting of ‘positive’ symptoms, namely illusions, hallucinations, and delusions.³⁶ Definitions of psychotic symptoms in the context of PD are consistent with the definitions in the wider psychiatric literature. The prevalence of psychosis also varies depending on the diagnostic tool used. In 2007, the combined National Institute of Neurological Disorders and Stroke (NINDS) and National Institute of Mental Health (NIMH) work group proposed a unifying diagnostic criterion for PD psychosis which includes the presence of at least one psychotic symptom.³⁶ Longitudinally, prevalence increases with disease duration,³⁷ and, at 12 years, 60% of patients from a community cohort reported hallucinations or delusions.³⁸

Proposed risk factors for the development of psychosis in PD include duration and severity of PD, and cognitive impairment. The development of these symptoms at a time when demands associated with caring for people with PD are already high results in a greater sense of burden for the carer in comparison with caring for those without psychosis and is also a predictive factor for nursing home placement.³⁹ Additional risk factors for psychosis include treatment with dopaminergic and anticholinergic medications. The relationship between the use of PD medication and development of PD psychosis remains controversial. Anecdotally, an association between PD psychosis and the use of dopaminergic therapies and duration of treatment has been observed. Psychotic symptoms can develop on initiation of medication with a subsequent improvement with reduction or withdrawal and this is most potently seen with the use of dopamine agonists. None the less, a causal effect of dopaminergic medications has not been established and psychotic symptoms have been described in recently diagnosed patients who have not yet started treatment.⁴⁰

The initial management of PD psychosis involves the identification and treatment of any precipitating factors. Psychosis may be the manifestation of an acute illness or change in medication and a thorough clinical evaluation is imperative. If the patient is tolerating the symptoms of psychosis with no adverse features, the best approach may be to watch and wait. Many patients with PD do not find hallucinations distressing and retain insight.

Where intervention is warranted, there are a number of approaches. In patients with significant cognitive impairment rivastigmine has good efficacy.⁴¹ Clozapine and pimavanserin (a 5-HT2A inverse agonist, not yet licensed for use in the UK)⁴² are efficacious. There are barriers for the routine use of clozapine in the clinical setting, namely the regular blood tests required to monitor for agranulocytosis and the registration required for prescribing the drug. Other atypical antipsychotics used in clinical practice include quetiapine, olanzapine, and risperidone, but these can cause worsening of motor symptoms and other side effects including QT prolongation on the electrocardiogram (ECG), sedative effects, metabolic syndrome, and a potential deterioration in cognition. The most recent NICE guidance advocates the use of low-dose quetiapine and clozapine as the most appropriate medications in PD psychosis,⁴³ and in clinical practice at present, quetiapine is commonly used.

Cognitive Impairment

Cognitive impairment is a common non-motor manifestation in PD and is heterogeneous in terms of its severity, rate of progression, and the cognitive domains affected. It has a significant negative impact on patients and their carers and is associated with increased disability, mortality,⁴⁴ carer burden,⁴⁵ and need for nursing home placement.⁴⁶

Cognitive impairment ranges from subjective cognitive decline, to mild cognitive impairment (MCI), to dementia, but does not necessarily progress linearly. Longitudinal studies suggest MCI in PD increases the risk of developing Parkinson’s disease dementia (PDD). Janvin et al. showed, whilst controlling for age, disease stage, education, and gender, that MCI was strongly associated with PDD development (odds ratio 5.1; 95% CI, 1.51–16.24, p = 0.005) over a 4-year follow-up.⁴⁷ However, this study included relatively small numbers. More recently, evidence suggests MCI may not always progress to PDD, with some patients remaining stable on longitudinal assessments or even reverting back to normal cognition.⁴⁸ The prevalence of MCI in PD ranges from 15% to 53%⁴⁹ but this large variation in prevalence is likely caused by differences in study settings (hospital versus community patients), variable clinical characteristics, and inconsistent definitions of MCI in PD. The Movement Disorders Society published a unifying set of diagnostic measures in 2012 to standardize practice across clinical trials.⁵⁰ From a clinical perspective, MCI and PDD can best be differentiated by determining whether the cognitive impairment significantly impacts activities of daily living.

Several longitudinal studies have demonstrated that approximately 50% of patients develop PDD ten years after initial PD diagnosis.^{51, 52} The Sydney Multicenter Study, the largest follow-up study of newly diagnosed patients, showed 83% of patients with PD developed PDD at 20 years post diagnosis and 75% developed PDD before death.⁵³ Additionally, not only does the progression of cognitive impairment in PD vary, so too does the pattern affected. Cognitive deficits may affect executive function, attention, processing speed, and/or visuospatial function.⁵⁴ The CamPaIGN (Cambridgeshire Parkinson’s Incidence from GP to Neurologist) cohort study assessed cognition at 3.5 and 5.5 years post diagnosis and found a decline in mini-mental state examination (MMSE) scores at a rate of 0.3±0.1 points per year over 5.2 years.⁵⁵ In addition, they showed that deficits in semantic fluency and visuospatial function at baseline were associated with a greater risk for developing PDD whereas deficits in executive function were not.^{54, 55} This finding, in addition to other studies, led to the ‘dual syndrome hypothesis’ which suggests that patients with primarily executive dysfunction, driven by changes in dopaminergic pathways, are less likely to develop PDD, whereas those with memory and visuospatial functional deficits, caused predominantly by deficits in acetylcholine (ACh), are more prone to rapid cognitive decline and progression to PDD.

The heterogeneity of both the presentation and progression of cognitive impairment in PD is not fully understood and is clearly a complex interplay between genetic and environmental factors. Risk factors associated with an increased risk for PDD include patient age, older age of disease onset, and the presence of hallucinations. In addition, the Parkinson’s phenotype with a more predominant gait dysfunction is associated with more rapid cognitive decline than those with a tremor-dominant phenotype.⁵⁵ Genetic factors also play a role with mutations in the alpha-synuclein (SCNA)⁵⁶ and glucocerebrosidase (GBA1) genes⁵⁷ being associated with more rapid cognitive decline, in addition to an earlier onset of PDD. One retrospective longitudinal study found that 56% of GBA1 mutation carriers had dementia at age 70, compared with 15% of sporadic PD patients.⁵⁷ Interestingly, patients with a PARKIN mutation, which accounts for 50% of autosomal recessive PD, are less likely to develop PDD, with one review finding PDD in fewer than 3% of cases.⁵⁸

The only drug currently licensed for PDD is rivastigmine, a cholinesterase inhibitor, with memantine, an NMDA receptor antagonist used as second treatment line if the patient is intolerant to rivastigmine.⁴³ However, a meta-analysis including three RCTs comparing memantine 20 mg daily to placebo concluded it only had a mildly beneficial effect on the global impression of change assessment, with no significant change on cognitive function assessed using the MMSE.⁵⁹ There is currently no proven treatment for MCI in PD, and given the limited pharmacological options, it is important to consider non-pharmacological options including physical exercise programmes, cognitive training and stimulation using pharmacological and non-pharmacological approaches.

A systematic review of eight studies including 158 patients suggested that physical exercise had beneficial effects on global cognition.⁶⁰ Physical activity intervention studies have historically recruited small numbers of patients and utilized heterogeneous interventions with varying degrees of exercise intensity, mode, and duration.

Cognitive training is a structured teaching programme designed to target specific cognitive domains. A meta-analysis involving seven studies (272 patients) showed a small but statistically significant improvement compared with controls.⁶¹ Overall, given the heterogeneity of cognitive impairment in PD and the impact it has on the patient’s quality of life and carer burden, the primary aim of management should be to apply an individualized approach whilst addressing patient and carer concerns and expectations.