Share this article on:

Prospective Study Identifies New Predictors of Dementia in Parkinson's Disease

Hiscott, Rebecca

doi: 10.1097/01.NT.0000455667.59378.c8
Back to Top | Article Outline




A prospective study has demonstrated strong associations between orthostatic hypotension and REM sleep behavior disorder and dementia in Parkinson's disease.

Dementia is widely recognized as one of the most common non-motor features of Parkinson's disease (PD), occurring in as many as 50 percent of PD patients. But while past research has looked at motor features and disease subtype as predictors of PD dementia, fewer studies have investigated the link between non-motor symptoms and dementia in PD.

Now, a new prospective study published in the Aug. 29 online issue of Neurology has found that certain non-motor features of Parkinson's disease, including orthostatic hypotension, REM sleep behavior disorder, and color discrimination and gait dysfunction, could serve as predictors of dementia in PD patients.

“The most striking finding of our study is that the predictors of dementia are not necessarily what is already out there in the literature,” co-senior author Ronald Postuma, MD, MSc, professor of medicine in the department of neurology and neurosurgery at McGill University in Montreal, Canada, and researcher in neurosciences at the Research Institute of the McGill University Health Center, told Neurology Today.

“Typically, when we've looked at predictors of dementia there's been a lot of emphasis on the motor phenotype—but in this case it really does appear that the strongest predictors of dementia, at least in this cohort, were non-motor,” he said.

Back to Top | Article Outline


The study screened 80 dementia-free Parkinson's disease patients for a number of autonomic, sleep, psychiatric, visual, olfactory, and motor manifestations of the disease. The participants also underwent a baseline neuropsychological examination that measured the cognitive domains of executive function and attention, memory, and visuospatial ability. Patients were screened for dementia after 4.4 years of follow-up.

The authors reported that 27 of the 80 patients studied, or 34 percent, developed dementia at follow-up. Several variables were found to be strongly associated with PD dementia, most notably orthostatic blood pressure drop (p<0.001) and REM sleep behavior disorder, or RBD (p=0.001).

Patients with RBD at baseline had a 43 percent risk of developing dementia at follow-up, compared with 2.5 percent in those without RBD. In fact, only one of the 27 patients who developed dementia did not have RBD at baseline. In patients with orthostatic hypotension, having a systolic blood pressure drop of more than 10 mmHg increased dementia risk seven-fold (p=0.002). The authors noted that RBD has been linked with orthostatic blood pressure drop in past research, and could therefore be interpreted as a sign of autonomic dysfunction.

The strength of the association between orthostatic hypotension and dementia risk surprised the researchers, Dr. Postuma said. “We do know that patients who have Parkinson's disease and associated dementia are more likely to have orthostatic hypotension. I don't think I would have been surprised to see a small effect, but such a large effect surprised me. It's quite a striking relationship,” he said.

Other variables found to be associated with PD dementia included abnormal color vision (p=0.014), which tripled a patient's odds of developing dementia, as well as mild cognitive impairment, or MCI (p=0.002), and motor variables such as gait involvement (p=0.008), falls (p=0.042), and freezing (p=0.013). Older and male PD patients were more likely to develop dementia than younger and female patients (p=0.023).

While the study's findings are prospective, they could be helpful in identifying PD patients who are at risk of developing dementia in the future, Dr. Postuma said. “We're starting to get a sense of who is going to get dementia, and it's not just related to age. Having more surveillance for dementia symptoms as they emerge, or being more aware of the likelihood of cognitive side-effects of certain medications if a patient has risk factors for dementia, could be a useful clinical point,” he said.

Back to Top | Article Outline


The findings offer new evidence for non-motor predictors of dementia in Parkinson's disease and appear to confirm previously observed associations in PD dementia, experts who reviewed the study told Neurology Today. However, they cautioned that the research should be seen as exploratory, and the results will need to be replicated in future studies.

“The study gave nice insight into risks for dementia in Parkinson's disease that could be incorporated into clinical practice, in particular looking at changes in blood pressure and assessing for sleep disturbances and color vision,” said James E. Galvin, MD, MPH, professor of neurology and psychiatry at New York University Langone School of Medicine and director of clinical operations for the Center for Cognitive Neurology, director of the Pearl Barlow Center for Memory Evaluation and Treatment, and associate director of the Alzheimer Disease Center at NYU Langone Medical Center.

However, “it is a relatively small study and it's got limited power, and the authors didn't control for multiple comparisons. Many of these findings might not hold up to a multiple-comparison approach,” he said.

“I would emphasize that until we understand which of these variables are just correlating with each other, rather than independently correlating with dementia onset, their usefulness awaits confirmation,” said John N. Caviness, MD, FAAN, founder and director of the Movement Disorders Laboratory at the Mayo Clinic in Phoenix, AZ, and professor of neurology at the Mayo Clinic College of Medicine.

Indeed, the study authors noted that their findings are prospective and will need to be confirmed in future analyses. The link between RBD and dementia has been observed in past research, but the association between orthostatic hypotension and dementia has not yet been replicated, Dr. Postuma said.

Dr. Galvin also noted that including PD patients who had mild cognitive impairment at baseline may have influenced the study results. “Baseline cognitive status is a strong predictor of the causative pathway to dementia. It's a covariate or a confounder, in some senses, and the authors could have controlled for it in their analysis. It would be more interesting to look at the people who were not impaired at baseline, and see what happens to them,” he said.

He added that many of the findings do seem to confirm past research on predictors of dementia in PD—-including his own, which found that sleep disturbances, gait, falls, freezing, hallucinations, and depression could all serve as predictors of PD dementia.

The color vision finding is particularly novel, he said. “It's a fairly straightforward test to do, and I don't know that a lot of people have necessarily thought about abnormal color vision as a predictor.”

Dr. Caviness added that he appreciated that the study explored multiple variables, including some that had not been looked at in past research. “Maybe we've been too focused on narrow categories of biomarkers for Parkinson's disease dementia,” he said.



With this prospective study, “the researchers are starting to build a profile of patients who are non-demented and determine those that are most likely at high risk of progressing to dementia,” said Daniel Weintraub, MD, associate professor of psychology at the Perlman School of Medicine at the University of Pennsylvania, and a fellow in Penn's Institute on Aging. Such a profile could be useful in designing future clinical trials. For instance, “you may want to include people with these markers in your study focused on cognitive decline if you think they're at a high risk of declining cognitively over a short period of time. Or, if you want to have a non-demented population, you may want to exclude people with these problems because they might develop cognitive impairment over the lifespan of your study,” he said.

Dr. Weintraub added that confirming these variables as predictors of dementia, particularly strong risk factors in this study like RBD and orthostatic hypotension, could potentially have an impact on patient care. “The findings could be important for management reasons, to be able to give patients and their families more of an educated opinion about whether they're at an increased risk of developing dementia,” he said.

Back to Top | Article Outline


  • While REM sleep behavior disorder was associated with dementia in this study, neither daytime sleepiness or insomnia was significantly associated with Parkinson's disease dementia.
  • A baseline history of visual hallucinations (p=0.009), visual illusions (p=0.005), subjective cognitive complaints (p=0.03), thought disorders (p=0.015), and depression (p=0.047) predicted dementia in these patients.
  • Researchers also observed higher apathy scores (p=0.74) in patients who developed dementia.
  • Baseline occurrence of urinary symptoms, sexual dysfunction, and bowel dysfunction was higher in patients who developed dementia, but the findings did not reach statistical significance.
  • Although abnormal color vision was found to triple dementia risk, researchers did not find an association between olfactory dysfunction and dementia.
Back to Top | Article Outline








Back to Top | Article Outline


•. Anang JBM, Gagnon JF, Bertrand JA, et al. Predictors of dementia in Parkinson's disease: A prospective cohort study. Neurology 2014; E-pub 29 Aug. 2014.
    •. More about Parkinson's disease dementia
      © 2014 American Academy of Neurology