San Diego, CA—A quick and easy clinical prediction tool can delineate a 4-year risk for dementia in patients with Parkinson’s disease (PD). According to a study presented at the International Congress of Parkinson’s Disease and Movement Disorders, the tool, which can be applied in bedside assessments, predicted dementia with >80% accuracy in patients with PD.
“The 6- to 8-item clinical prediction tool was able to clearly delineate medium-term risk of dementia in PD,” said Julius B.M. Anang, MD, PhD, Neurologist, St. Boniface Clinic, Winnipeg, Canada. “Further confirmation of predictors and validation of tool is needed in larger cohort.”
Dr Anang told Value-Based Care in Neurology that dementia is among the most devastating nonmotor features of PD, causing severe decline in quality of life, increased caregiver burden, increased mortality, and often institutionalization. The prevalence of dementia in patients with PD ranges between 24% and 50%.
In Dr Anang’s opinion, it is necessary to develop a predictive algorithm for PD dementia so that only the highest-risk patients are treated with potentially harmful neuroprotective medications. Recognizing early cognitive impairment could also enable researchers to select better populations to test these drugs in clinical trials.
“It’s important to expose only the highest-risk patients to those side effects,” he said. “People who have the highest risk of developing dementia would be the best group to target with these medications.”
Using regression analysis, receiver-operating characteristic curves, and sensitivity analysis in the primary cohort, Dr Anang and colleagues chose 8 variables to generate a dementia scoring algorithm. They then tested these predictive variables in 2 external prospective cohorts in Montreal and Japan.
“We chose variables that predicted dementia on regression analysis and could be applied in bedside assessments,” Dr Anang reported. “This test is designed to be performed in a matter of minutes. Someone used to doing it should only take about 10 minutes.”
Screening for Dementia
The 8 variables included age >70 years, male sex, bilateral disease onset, rapid eye movement (REM) sleep behavior disorder, freezing and/or falls, orthostatic systolic blood pressure reduction >10 mm Hg, mild cognitive impairment (MCI), and having hallucinations.
Of these variables, 5 were significantly associated with dementia in this cohort. Bilateral disease onset, hallucinations, and freezing and/or falls were not significant, but all had odds ratios between 1.7 and 2.0.
The researchers then created 2 predictive scales. One scale combined all the variables (the primary outcome), and the second did not include variables directly related to cognition (MCI, hallucinations).
The scale predicted dementia with 82% accuracy, 71% sensitivity, and 85% specificity in 214 patients, Dr Anang reported.
“Using a 3-category delineation, dementia risk was 3% in the ‘low-risk’ patients, 31% in the ‘intermediate-risk,’ and 70% in the ‘high-risk’ group,” he said. “The triple combination of baseline REM sleep behavior disorder, MCI, and orthostatic systolic blood pressure drop conveyed the highest risk.”
Dr Anang added, “due to time constraints, physicians often fail to perform blood pressure of patients with PD, which is an important, nonmodal feature of cognitive dysfunction. Because most physicians that we contacted didn’t have this aspect of the scale, we couldn’t use their data.”
Dr Anang is continuing to verify this scoring system. “The best way to verify our findings is to have other physicians look at their patients and see how accurately this scale is able to predict dementia. If you see someone with a score of 5, this might be someone you want to follow more closely. In Canada, for example, patients are only seen once every 16 months by their neurologist, so those higher-risk patients should definitely be seen a bit more frequently,” he concluded.
Last modified: August 5, 2015