Article Text

Download PDFPDF

Predicting Parkinson’s disease: worthwhile but are we there yet?
  1. Christopher H Hawkes1,
  2. Jacquie Deeb2
  1. 1Consultant Neurologist & Honorary, Senior Lecturer, Essex Neuroscience Centre, Essex, UK
  2. 2Neurology Research Fellow, Essex Neuroscience Centre, Essex, UK
  1. Correspondence to:
 Dr C H Hawkes, Essex Neuroscience Centre, Oldchurch, Hospital, Romford, Essex RM7 0BE, UK;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Current opinion suggests there is a preclinical phase in idiopathic Parkinson’s disease (PD) with a latent period somewhere between 6 and 40 years. 1,2 This prodrome, sometimes also termed the presymptomatic or pre-motor phase of PD, needs to be distinguished from non-motor features such as neuropsychiatric symptoms, dysautonomia, fatigue, and so on,3 only some of which start before the classical motor triad of PD (tremor, rigidity, and bradykinesia). Clearly, prediction of PD would be straight forward in those few parkinsonian patients with monogenetic disorder that in theory may be identified from birth, but most such mutations are very rare. A variety of predictive biochemical tests4 are proposed but as yet none is sufficiently reliable. Fluorodopa positron emission tomography (PET) or dopamine transporter imaging (DATScan) will detect deficiency of striatal dopamine but by this stage the motor phase–at least neurochemically–has probably begun. So might symptoms or signs foretell PD, and if so, which ones and how reliably? Such predictive research has many potential benefits as it enhances our understanding of the underlying disease mechanism, identifies subjects at risk of future PD, and provides an opportunity for neuroprotective therapy if this is eventually found to be effective.

Braak pathological staging5 represents a landmark in our understanding of PD and may provide a basis for any proposed clinically based predictive features of the motor stage (see fig 1 and table). The first two stages represent the pathological substrate of preclinical PD while stage III straddles the premotor and motor phases. It is estimated that when some 50% of nigral cells are lost the first motor symptoms will appear,6 heralding the end of the premotor phase; hence tests of structures involved at this juncture such as the amygdala and pedunculopontine nucleus are unlikely to be rewarding in the predictive sense.


Braak pathological staging of idiopathic Parkinson’s disease.5 Note that each stage is cumulative–for example, stage III has all the pathological features of stages I and II

Figure 1

Braak stages I–IV in idiopathic Parkinson’s disease. In stage I there is simultaneous attack on the brain via olfactory and medullary routes (coloured black). The two processes converge in stage III (coloured red) on the medial temporal lobes and then spread widely into the neocortex. (Reproduced with permission from Braak et al.5)

Figure 2

A page from the University of Pennsylvania Smell Identification Test. Microencapsulated odours are released on scratching the rectangular strip with a pencil as shown here.

Disorder of olfaction, autonomic function, and sleep are the best substantiated disorders in the premotor period.


Almost all researchers find olfactory impairment in PD, affecting 80% or more of patients.7,9 The question is does this precede the motor features or does it start at the same time? Apart from compelling neuropathological evidence of priority,5 three pivotal clinical studies address this question:

  • A study from Amsterdam10 evaluated prospectively 78 asymptomatic, first degree relatives of non-familial PD patients by olfactory tests and DATScan. Forty were hyposmic at baseline and when reviewed two years later, four had abnormal scans and PD was evident clinically. In the remaining 36 hyposmics who displayed no sign of PD the rate of decline of dopamine transporter binding was higher than in normosmic relatives.

  • The Honolulu-Asia Aging Study (HAAS) used the cross cultural Brief Smell Identification Test (BSIT) in 2263 healthy Japanese-American men aged 71–95 years.11 After 7 years’ follow up 19 developed PD, which gave a significantly raised relative odds for PD in the lowest BSIT tertile of 4.3 (95% CI 1.1 to 16.1) compared to those in the highest tertile. In the same cohort 163 died without signs of PD or dementia and subsequent postmortem revealed 17 with incidental Lewy bodies, the pathological hallmark of the presymptomatic phase.12 Once more those who scored in the lowest tertile of BSIT were significantly more likely to have Lewy body pathology (OR 11.0; 95% CI 1.3 to 526)

  • Transcranial sonography of the substania nigra was performed in 30 people with idiopathic anosmia but no motor deficit.13 Eleven displayed increased nigral echogenicity, typical of PD. Ten subjects agreed to have a DATScan, which was pathological in five and a further two were borderline, suggesting these seven might be in the pre-motor phase of parkinsonism. In fact, two of the five with abnormal DATScans have now developed clinical signs of PD (T Hummel, personal communication, 2006).

Gastrointestinal pathology in established PD is widespread

The last study basically demonstrates synchrony of olfaction with early biochemical and structural changes and awaits further follow up and replication. Nevertheless, the evidence provided by these three papers is consistent and, taken in conjunction with Braak staging, implies that smell impairment develops before the motor symptoms of PD.


There is clear evidence of marked and progressive α-synuclein deposition in the dorsal motor nuclear complex of cranial nerves IX and X in PD (Braak stage I). This is one reason for many of the autonomic features. In established PD several investigators find reduced cardiac uptake of the noradrenaline analogue, metaiodobenzyl guanidine.14 However, this evaluates postganglionic noradrenergic cardiac sympathetic function whereas the vagus provides parasympathetic adrenergic innervation to the cardiac plexus. Heart rate variability is abnormal in PD and thought to reflect both sympathetic and parasympathetic aspects of vagal function15 but there are no premotor studies of heart rate. Incidental Lewy bodies have been found in the cervical sympathetic ganglia and cardiac plexus in the absence of dorsal motor nuclear complex change,16 but it is unclear whether these incidental Lewy bodies represent precursors of classical PD or a parkinsonian syndrome such as dementia with Lewy bodies (DLB) or multiple system atrophy. Autonomic failure presaging a parkinsonian syndrome has been shown in two patients17 but the autopsy changes were more characteristic of DLB than PD.

Gastrointestinal pathology in established PD is widespread18 with Lewy bodies and Lewy neurites identified in the gastric submucosal plexus.19 Such pathology in those who died from non-neurological causes leads to the suggestion that the first change of all in PD (Braak stage zero?) is in Auerbach’s plexus and that the postulated “pathogen” ascends the motor vagal fibres–that is, retrogradely–to reach the medulla.19 There is just one prospective analysis of bowel habit in elderly patients who took part in the Honolulu Heart Program.20 This comprised 6790 males without extrapyramidal disease at enrolment, followed up for 24 years, of whom 96 developed PD. The adjusted risk of PD in those with less than one bowel movement per day compared to those with one or more per day was increased about threefold (OR 2.7; 95% CI 1.3 to 5.5) implying that constipation is a harbinger of PD.


A variety of sleep disorders are found in established PD21 of which the most studied are excessive daytime sleepiness and REM sleep behaviour disorder. The pathophysiology is not understood fully but most likely reflects cellular change found in Braak stage II and some of stage III, particularly involving the pedunculopontine nucleus and the hypothalamus. In the Honolulu-Asia Aging Study22 43 of 3078 men developed PD and those with prior excessive daytime sleepiness were approximately three times more likely to do so (OR 3.3; 95% CI 1.4 to 7.0). According to one source2311/29 subjects with unexplained REM sleep behaviour disorder subsequently developed parkinsonism, complementing the finding of decreased striatal transporter uptake in REM sleep behaviour disorder by others.24 In a further study of olfaction and sleep25 nearly all of 30 REM sleep behaviour disorder patients had significantly increased olfactory threshold and there was evidence of parkinsonism in eight, again implying that olfaction and REM sleep behaviour disorder are early features. A difficulty with most of these studies is the lack of pathological confirmation and, where this is available,26 the changes (at least in REM sleep behaviour disorder patients) were more in keeping with parkinsonism than PD.

Tests of olfaction are the simplest and probably the most valuable of the currently proposed predictive tests


The Honolulu-Asia Aging Study suggests that mid-life obesity and slow reaction time are premorbid indicators of PD.27,28 Slow reaction time is of interest as it could be explained by damage to the pontine reticular formation which activates spinal motor neurons, but premotor obesity, which probably indicates hypothalamic disorder as in Braak stage III, is less easily accounted for. Prior to the classical motor symptoms, patients may report non-specific pain, which cannot be attributed to joint disease.29,30 This could relate to dysregulation of the pain gating mechanism at pontine level as in stage II. Other candidates for the premotor phase include depression, phobia, and impaired colour discrimination, 31,33 and all are awaiting prospective and pathological confirmation.


The more convincing premotor features that concur with Braak staging are olfactory loss, constipation, and excessive daytime sleepiness. Although the Braak pathological classification is relatively new and awaits independent confirmation, for the most part it provides an explanation for many premotor symptoms. Opponents may argue that mild deficiency in neurotransmitter release, for example at cortical level, might induce premotor symptoms not reflected by α-synuclein staining. There are also obvious problems with specificity–that is, olfactory disorder may herald the onset of parkinsonian syndromes such as multiple system atrophy, Lewy body dementia, or even Alzheimer’s disease9 and the presence of incidental Lewy bodies may presage one of many types of parkinsonism. In terms of practicality, tests of olfaction are the simplest and probably the most valuable of the currently proposed predictive tests. Despite this, none of the suggested premotor features singly or combined has more than a 10–20% predictive power for future risk for PD (GW Ross, personal communication). In other words, the right question has still to be asked–and even if we discover a strongly predictive symptom or sign of impending PD, will we be able to do anything about it? There is no clear benefit from attempts at neuroprotection of damaged neurons in established PD34 and as yet no published trials in those at, for example, 50% risk. Discovery of new genetic markers may completely swamp pre-motor research but until then the quest for pre-motor characteristics should continue. In both cases, further prospective studies with very large cohorts of middle aged people will be essential. The UK Biobank35 represents a bold start in this direction.


Other content recommended for you