Parkinson’s disease has been regarded as a neurological condition mainly affecting motor function and arising from specific lesions in the brain stem. The recognition of dementia in Parkinson’s disease is of importance in management but the possibility that motor and cognitive functions may be located in the same region of the brain is of theoretical importance.

There have been numerous reports of the impairment of specific cognitive functions in patients with Parkinson’s disease. Mortimer and colleagues reported cognitive impairment in 93 per cent of a substantial group of patients with Parkinson’s disease.⁽¹⁾ Their data showed neither a clear distinction between impaired groups nor the presence of subtypes of Parkinson’s disease in which cognitive impairment is a more frequent occurrence. They proposed that cognitive impairment in Parkinson’s disease lay on a continuum of severity, rather than as a feature of particular subgroups. The impairments identified included deficits in memory, language, visuospatial functioning, abstract reasoning, slowness in intellectual tasks, and difficulty in shifting from task to task. These deficits are widespread among patients with Parkinson’s disease and can occur at an early stage of the disorder.

A proportion of patients with Parkinson’s disease show impairment of a range of cognitive functions akin to the global impairment seen in Alzheimer’s disease.⁽²⁾ However, the pattern of impairment is frequently less severe than in Alzheimer’s disease where the pathological changes in the brain are known to be widespread. Cognitive impairment in a range of disorders of movement where the main neuropathological changes reside in the subcortical region of the brain led to the concept of ‘subcortical dementia’, a form of intellectual impairment of lesser degree than in Alzheimer’s disease, but affecting several cognitive functions. Albert described a syndrome of which the main features were: emotional or personality changes, impaired memory, defective ability to manipulate acquired knowledge, and a striking slowness in the rate of information processing.⁽³⁾

Many issues arose as to the nature of ‘subcortical dementia’. Was it a clinical or a pathological concept? Was the difference between this and other forms of dementia simply one of degree? Did the pathological changes occur in the subcortical region of the brain alone? Was the syndrome of cognitive impairment distinctly different from other dementias or did the presence of motor features of the disorder simply give the intellectual impairment a distinct character? Was subcortical dementia a stable condition or a transitional state leading eventually to global dementia? Opinion has ranged from full acceptance of the concept to scepticism.^(4,5)

McHugh⁽⁶⁾ suggested that the subcortical region subserves functions not only in motor control and cognitive function but also in the control and display of mood. He suggested that these form a ‘subcortical triad’ of symptoms most convincingly seen in Huntington’s disease. A notable difference between this concept and that of subcortical dementia was that the pathological disturbance of mood is only intermittently present, whereas the motor and cognitive changes are persistent.

Cummings⁽⁷⁾ suggested that cognitive impairment in Parkinson’s disease takes three forms: one which is relatively mild and meets the criteria for subcortical dementia, a more severe form showing wider impairment of cognitive function but neuropathologically distinct from Alzheimer disease and a severe form which shows neuropathological changes in both the subcortical region of the brain, and in the cortex, the latter of Alzheimer type. This proposal provides a basis for viewing cognitive changes in Parkinson’s disease, albeit provisional.

Many reports have suggested that global dementia occurs in Parkinson’s disease. Whether such a severe change in cognitive function can be regarded as an intrinsic feature of this disease, whether it implies an extension of a neuropathological process more widely in the brain, or whether it suggests a different neuropathology from the outset is, as yet, uncertain. More recently the debate has shifted to whether dementia in Parkinson’s disease, dementia with Lewy bodies and Alzheimer’s disease should be viewed as a spectrum, or as separate conditions with varying degrees of clinical and pathological overlap.

Research to establish the status of dementia in Parkinson’s disease has confronted a range of methodological issues.⁽⁸⁾ A major problem is in the diagnosis of Parkinson’s disease itself. The original description of paralysis agitans by Parkinson was the identification of a syndrome rather than of a disease. The part played by such agents as heavy metals, infections, and vascular disease was subsequently recognized. In spite of the use of standardized methods, a substantial proportion of patients diagnosed as suffering from Parkinson’s disease in life do not show the expected findings in the brain post-mortem. In a follow-up study, 80 per cent of cases were shown to have neuropathological changes of Parkinson’s disease after death but over 20 per cent were diagnosed as having suffered from progressive supranuclear palsy, multiple system atrophy, or Alzheimer’s disease.⁽⁹⁾ Furthermore, some dementing illnesses may show disorder of movement as a clinical feature.

Cases of dementia in Parkinson’s disease have been reported for over a hundred years. The frequency of dementia reported in cross-sectional or prevalence studies ranges from 0 to over 80 per cent. A recent review found a prevalence of between 28 and 44 per cent in community studies but in older samples of subjects the prevalence was much higher.⁽¹⁰⁾

Follow-up studies have great advantages in studying the frequency of dementia in Parkinson’s disease as they allow the diagnosis of Parkinson’s disease to be checked, repeated assessment reduces errors in the recognition of dementia, the pattern of evolution of dementia may be followed, the underestimation of dementia by selective loss through death is avoided, and they reveal the incidence rather than the prevalence of the condition. The choice of methods of diagnosis and assessment that will remain appropriate throughout the period of the follow-up remains a problem.

A prospective, controlled study in the United Kingdom reported an incidence of dementia of 19 per cent after 4.5 years observation. A later report on the same cohort of subjects showed an incidence of dementia of 38 per cent after 10 years of observation. The control group showed cases of cognitive impairment but none amounting to dementia.^(11,12) A community based, prospective, controlled study, in Norway, showed the risk of dementia was 5.9 times greater than in the control group.⁽¹³⁾ A prospective, controlled study in the United States showed that dementia was 3.7 times greater in the Parkinson’s disease group with severely affected, elderly patients especially at risk (Table 4.1.6.1).^(14,15)

Those most likely to develop dementia are: older people, patients with Parkinson’s disease of longer duration, subjects who have a greater severity of motor symptoms and signs of Parkinson’s disease, and those who show greater physical disability.^{(11, 12, 13, 14 and 15)} Some studies have shown that male sex or late onset are associated with dementia. The apparent association between Parkinson’s disease treated with levodopa and dementia is probably due to improved survival.