Why do only some people with Parkinson’s disease decline cognitively? Scientists have proposed various theories, including that coincident Alzheimer’s pathology is to blame. Researchers led by Joseph Winer and William Jagust at the University of California, Berkeley, tested this by measuring amyloid and tau levels in vivo with PET. In the December 11 JAMA Neurology, they reported that PD patients with mild cognitive impairment had no more tau tangles on average than did cognitively normal PD patients or healthy controls. “From this study and others, there is converging evidence that tau pathology does not relate to the cognitive decline in Parkinson’s,” Winer said. The study excluded people with Parkinson’s disease dementia.

  • Alzheimer’s pathology does not explain mild cognitive impairment in PD.
  • The culprit could be cortical α-synuclein, but testing this will require a PET tracer.
  • It remains possible that comorbid AD pathology causes PD dementia. 

Other researchers found the data intriguing. “The current thinking is that Alzheimer’s pathology, i.e. tau and amyloid, is the driving force behind cognitive deficits, even in people without clinical Alzheimer’s disease. This study… shows that the reality is more sophisticated and nuanced,” M. Arfan Ikram at Erasmus Medical Center in Rotterdam, the Netherlands, wrote to Alzforum.

Parkinson’s disease is primarily a movement disorder marked by aggregates of α-synuclein, but over time a majority of patients go on to develop cognitive problems or dementia as well (Aarsland et al., 2003). Because tau tangles correlate with cognitive decline in Alzheimer’s, researchers wonder if the same holds true in Parkinson’s. The evidence has been mixed. Two studies suggested this was the case for people with PD dementia or dementia with Lewy bodies, but another reported little tau pathology in PD patients with mild cognitive impairment, and no relationship between tau PET signal and cognition (Mar 2013 conference news; Sep 2016 newsHansen et al., 2017). None of these in vivo studies included amyloid imaging, leaving it unclear whether the PD patients also had AD.

Winer and colleagues measured amyloid using PiB and tau with AV-1451, and administered several cognitive and neuropsychological tests. They recruited 29 PD patients who were being treated at the University of California, San Francisco; 15 of them were cognitively healthy and 14 had mild cognitive impairment. The researchers compared them with 49 healthy controls enrolled in the Berkeley Aging Cohort Study.

The scans turned up little evidence of AD pathology in Parkinson’s. Only about 20 percent of PD patients had a positive amyloid scan, as compared to about half of the control group. Age may explain this difference, because amyloid pathology incidence increases with age and the members of the PD cohort were younger than controls, with an average age of 66 versus 74. Among six amyloid-positive PD patients, only one was cognitively impaired. At least for the other 13 PD patients with cognitive problems, something besides AD must be to blame, Winer noted.

In the whole cohort, tau pathology correlated with age and with amyloid status. Thus, among the PD patients, only the six with positive amyloid scans had elevated tau. As a group, PD patients with MCI bound the same average amount of tau tracer as did cognitively normal PD and amyloid-negative controls, while amyloid-positive controls had an overall elevated tau PET signal.

Importantly, tau accumulation followed the Braak pattern of cortical distribution seen in normal aging and AD. This implies it was not a direct consequence of α-synuclein pathology, which occurs predominantly in one location in early PD, the substantia nigra. “In our PD patients, AD pathology develops in the same way as it does in normal aging, and does not explain their cognitive impairment,” Winer concluded.

What does explain it? Possibilities include α-synuclein pathology that spreads into the cortex, cerebrovascular disease, or a loss of dopamine signaling. Lending weight to the first idea, Tiago Outeiro at the University Medical Center, Göttingen, Germany, recently reported that α-synuclein oligomers interact with cellular prion protein to harm synapses in mouse models of synucleinopathy (Ferreira et al., 2017). Other research has found that cognitive decline may even precede motor symptoms in PD, again hinting that cognitive problems could stem from the α-synuclein pathology intrinsic to the disease (Sep 2017 news). Testing this idea will require PET tracers for aggregated α-synuclein, which are under development (May 2017 conference news). “That’s the key missing piece to the puzzle,” Winer said.

At the same time, it remains possible that comorbid Alzheimer’s pathology leads to dementia in PD, Winer noted. He plans to follow this PD cohort to see if those who accumulate amyloid go on to develop dementia. Kejal Kantarci at the Mayo Clinic in Rochester, Minnesota, suggested that the AD pathology seen in the six PD patients in this study might not yet have reached the threshold required to affect cognition (see comment below).

Others said the jury is still out on the effect of tau pathology. Allan Hansen at Aarhus University, Denmark, noted that PD patients might have a lower tolerance for tau pathology than healthy controls do because they already carry a heavy pathological burden (see comment below). Animal studies have found that some strains of aggregated α-synuclein can seed tau deposits, suggesting there could be an interaction in some people (Jul 2013 news). Notably, the tau tracer used in this study, AV-1451, detects only the paired helical filaments that characterize AD tau tangles, leaving open the possibility that other types of tau pathology might be present in people with PD.—Madolyn Bowman Rogers


  1. Interpretation of the results of this study, as well as a similar study from our site, is difficult. It seems that the characteristic tau scans seen among patients with preclinical Alzheimer’s disease are not found among pre-dementia Parkinson’s disease (PD) patients. This could perhaps be explained by the negative synergy between pathological tau and α-synuclein: A PD patient, who already has pathological α-synuclein, would have lower tolerance for pathological tau, and quickly descend into dementia, effectively removing the patient from the (pre-dementia) sampling population of our and Winer et al.’s study.

    However, even at the later stages of cognitive impairment in PD, tau ligand uptake does not reach the extent we expected from postmortem studies. Exceedingly few subjects reach uptake values seen in typical AD cases. Gomperts and colleagues reported that tau ligand uptake in three out of eight cognitively impaired PD (PD-CI) patients exceeded that of controls, but reached nowhere near some of the uptake values seen in DLB patients in the same study (Gomperts et al,. 2016). Seung Ha Lee reported one out of 22 PD-CI patients with tau accumulation in the AD range (Lee et al., 2017). In contrast, up to 50 percent of PD with dementia patients are reported to have levels of tau in the AD range in postmortem studies (Irwin et al., 2013). This discrepancy could have multiple causes, best investigated using postmortem examination of scanned PD subjects.

    Regardless of the cause of this discrepancy, the present study demonstrates that the relationship between tau and cognition cannot immediately be extrapolated from AD to PD. Thus, further studies will be needed to examine the possible role of tau PET as a biomarker of progression or for enrichment of clinical study populations in intervention studies involving cognitive decline in PD.


    . Tau Positron Emission Tomographic Imaging in the Lewy Body Diseases. JAMA Neurol. 2016 Nov 1;73(11):1334-1341. PubMed.

    . Distinct patterns of amyloid-dependent tau accumulation in Lewy body diseases. Mov Disord. 2018 Feb;33(2):262-272. Epub 2017 Nov 23 PubMed.

    . Parkinson's disease dementia: convergence of α-synuclein, tau and amyloid-β pathologies. Nat Rev Neurosci. 2013 Sep;14(9):626-36. PubMed.

  2. The results of this study are compatible with prior reports. They reproduce prior findings in PD-MCI compared with cognitively normal PD, which are consistent with the prior observation that tau deposits accumulate in proportion to the extent of cognitive impairment in PD/PD dementia and in DLB. One implication of these results is that tau deposition, as indexed by [18F]AV-1451 binding, does not appear to be a sensitive marker of mild cognitive decline in PD. This is perhaps not unexpected given that the causes of cognitive impairment in PD appear to be multifactorial, including α-synuclein deposition and damage to multiple neuromodulator systems. These results also suggest that tau deposition may be a relatively late event in the course of cognitive decline in PD.

  3. Parkinson’s disease is traditionally known for its characteristic motor features. However, non-motor features such as hyposmia, autonomic dysfunction, REM sleep disturbances, and cognitive impairment are also highly debilitating in a significant fraction of PD patients. The various disease features seem to correlate with the accumulation of α-synuclein pathology in Lewy bodies and Lewy neurites, which progresses throughout the brain, and even through connected neuronal networks outside of the brain (e.g. gut).

    Interestingly, accumulation of other proteins such as tau, Aβ, and SOD1 is also common in the brains of PD patients, as postmortem studies reveal. While the relevance of those accumulations is unclear, it is tempting to speculate that they may also relate to the manifestation of features that are characteristic of other disorders such as AD (Aβ and tau) or ALS (SOD1). However, most studies have relied on the postmortem evaluation of the pathological protein deposits, since imaging studies have been limited by the lack of tools/tracers to specifically image each type of protein deposits.

    In this study, Winer and colleagues used PET tracers for tau and Aβ to investigate the contribution of these types of pathology to cognitive status in PD. The study was cross-sectional, involving cognitively normal and cognitively impaired PD patients, and also healthy control participants. While the study needs to be replicated in larger cohorts, it suggests that tau deposition may not be associated with cognitive deficits in PD patients, although it seems to be associated with age and Aβ deposition. It will also be important to conduct longitudinal imaging studies, and to eventually correlate the findings with postmortem evaluation of the brains.

    Investigating the molecular mechanisms underlying cognitive decline in PD and other synucleinopathies is extremely important, as this may open novel avenues for intervention in non-motor features of those disorders. In a recent study (Ferreira et al., 2017), we found that extracellular α-synuclein may interact with prion protein to cause synaptic dysfunction. The molecular pathway is complicated, and may also relate to the accumulation of other types of pathology, such as tau or Aβ. Therefore, the emergence of new PET tracers will be important to clarify unresolved issues in the field.


    . α-synuclein interacts with PrP(C) to induce cognitive impairment through mGluR5 and NMDAR2B. Nat Neurosci. 2017 Nov;20(11):1569-1579. Epub 2017 Sep 25 PubMed.

  4. This paper adds to the growing literature on amyloid and tau pathological processes in Lewy body diseases by focusing on patients with Parkinson’s disease with normal cognition or with mild cognitive impairment.

    Findings suggest that Alzheimer’s pathology in non-demented PD patients is not too different from older adults with no PD. Furthermore, similar to what was observed in the control group, cortical tau levels were associated with cortical amyloid levels.

    Patients with DLB tend to have more cortical α-synuclein pathology than non-demented PD patients. In DLB patients, both tau and amyloid pathologies tend to be higher than in their cognitively normal peers and similar to what was observed in the current study, and they correlate with each other (Kantarci et al., 2017). There may be a critical stage in the evolution of Lewy body diseases when a certain threshold of AD pathology is reached in the cerebral cortex to influence cognition. In the current study, perhaps the AD pathologic processes did not reach that critical level to have a significant impact on cognitive performance in PD patients with no dementia.


    . AV-1451 tau and β-amyloid positron emission tomography imaging in dementia with Lewy bodies. Ann Neurol. 2017 Jan;81(1):58-67. Epub 2016 Dec 19 PubMed.

  5. I found this paper very interesting for various reasons. 1) There is an increasing body of literature suggesting that the pathological underpinnings of various neurodegenerative diseases (i.e., Alzheimer’s, Parkinson’s, frontotemporal dementia, ALS) may overlap to some extent. 2) The current thinking is that Alzheimer’s pathology, i.e., tau and amyloid, is the driving force behind cognitive deficits, even in persons without clinical Alzheimer’s disease.

    This study investigates both these hypotheses and actually shows that the reality is more sophisticated and nuanced. The authors show that in persons with PD with or without cognitive deficits, the burden of tau does not differ and that this is also not different from healthy individuals. This suggests that the pathological substrate of cognitive deficits in PD is some other pathologic process. What exactly that is remains to be investigated. By extension, the study suggests that despite the co-occurrence of different pathological processes in the same persons, the clinical manifestation can differ. For instance, in this study there were amyloid-positive PD patients without cognitive deficits. Again, this suggests that there is also disease-specific pathology present in addition to overlapping pathologies.

    The implications of this work are in my opinion twofold: First, more research needs to be done to fully capture and understand the various pathologic processes occurring in various neurodegenerative diseases. Second, the currently available palette of biomarkers is useful to distinguish some, but not other, clinical diseases. Further biomarker research needs to be continued.

    There are two important caveats on the study: First, this was a cross-sectional study, which limits our conclusions with respect to causality and causal inference. Second, the healthy controls were not entirely community-representing but selected in one way or another. This might have introduced some bias resulting in comparisons with the “healthy” group that might not be entirely accurate.​

Make a Comment

To make a comment you must login or register.


News Citations

  1. Dementia in Movement Disorders: What Causes It?
  2. Tau Deepens Cognitive Trouble in Lewy Body Diseases
  3. Cognitive Decline an Early Warning of Parkinson’s Disease?
  4. α-Synuclein Antibodies Enter Phase 2, Sans Biomarker
  5. An Extra Strain on the Brain—α-Synuclein Seeds Tau Aggregation

Paper Citations

  1. . Prevalence and characteristics of dementia in Parkinson disease: an 8-year prospective study. Arch Neurol. 2003 Mar;60(3):387-92. PubMed.
  2. . In Vivo cortical tau in Parkinson's disease using 18F-AV-1451 positron emission tomography. Mov Disord. 2017 Jun;32(6):922-927. Epub 2017 Mar 3 PubMed.
  3. . α-synuclein interacts with PrP(C) to induce cognitive impairment through mGluR5 and NMDAR2B. Nat Neurosci. 2017 Nov;20(11):1569-1579. Epub 2017 Sep 25 PubMed.

Further Reading

Primary Papers

  1. . Associations Between Tau, β-Amyloid, and Cognition in Parkinson Disease. JAMA Neurol. 2018 Feb 1;75(2):227-235. PubMed.