Doctors look beyond test scores to spot cognitive changes in Parkinson’s

Specialists evaluating cognitive impairment in people with Parkinson’s disease rely on much more than the score from a commonly used cognitive screening test. They also look for patterns in how patients performed on the test alongside their medical history and other health-related information, according to a new U.S. study.

“Extracting these patterns clinicians recognized provides deeper insights into how they interpret cognitive health creating a blueprint for future efforts to tailor the exam for detecting cognitive impairment in people with [Parkinson’s disease],” the researchers wrote.

The study, “A qualitative approach to extract diagnostic patterns of cognitive impairment in Parkinson’s disease,” was published in Scientific Reports.

Recommended Reading

July 8, 2026 News by Andrea Lobo

Parkinson’s advocates push for urgency as national plan begins

Cognitive changes can be difficult to detect in Parkinson’s

Parkinson’s disease is best known for its motor symptoms, but many people also develop cognitive problems ranging from mild difficulties with thinking, language, and memory to dementia. Mild cognitive impairment is an intermediate stage between normal cognition and dementia, but it does not progress the same way in everyone. Some people remain stable for years, some return temporarily to normal cognition, and others eventually develop dementia.

Because cognitive impairment can be subtle and vary considerably from person to person, detecting it early remains challenging. The Montreal Cognitive Assessment (MoCA) is one of the most widely recommended screening tests for cognitive impairment in people with Parkinson’s. The 30-point test measures skills such as memory, attention, language, and visuospatial abilities. A person’s total score is traditionally evaluated against a standard cutoff.

However, the specialists interviewed in this study described looking at patterns in a patient’s test performance alongside other health information rather than relying only on the total score to assess cognitive function.

“These health-related contextual dependencies are of particular interest, as understanding the patterns that clinicians identify as representative of [mild cognitive impairment] could lay the groundwork for developing a new method of interpreting the MoCA that improves the sensitivity and specificity for people with [Parkinson’s],” the researchers wrote.

To better understand how clinicians interpret the MoCA in real-world clinical settings, researchers reviewed medical records from nine people with Parkinson’s, including six with mild cognitive impairment and three with dementia.

For each person, the researchers compiled a patient profile containing one completed MoCA exam and one detailed neuropsychological report. The report included information about the person’s medical history, ability to perform daily activities, psychiatric symptoms, cognitive test results, and diagnosis.

Six clinicians who specialized in movement disorders each reviewed three patient profiles during individual interviews. As they assessed each person’s cognitive health, the clinicians explained how they reached their conclusions.

During one interview, findings from one patient profile were identified as more consistent with delirium than with Parkinson’s-related cognitive impairment. The researchers therefore excluded that profile from the analysis.

Researchers identified three patterns used to assess cognition

After analyzing the interviews, the researchers distilled three patterns from clinically meaningful features that recurred across the specialists’ assessments. Rather than relying only on the MoCA total score, the patterns combined observations from different parts of the cognitive test with information from each patient’s neuropsychological report.

The first pattern captured clinicians’ close examination of patients’ performance across different parts of the cognitive test, together with health factors that could influence cognitive function. For example, the pattern included errors on language, attention, and drawing tasks alongside factors such as education, hearing or vision problems, previous cognitive abilities, activity level, and signs of delirium.

The second pattern reflected clinicians’ interpretation of test performance in light of factors that could influence the expected results. For example, it included assessing language and attention performance in relation to a person’s education while also considering medications that could affect cognition, previous concussions, and a family history of dementia.

The third pattern illustrated how specialists considered explanations other than cognitive impairment for certain test difficulties. For example, difficulty repeating a spoken word could point to hearing problems rather than memory loss, while small or irregular drawings could be influenced by Parkinson’s motor symptoms or arthritis rather than cognitive worsening.

“We argue that by leveraging the innate human capacity for pattern recognition, which is the grounded authority of the differential diagnostic process, we can begin to decipher clinically relevant disease patterns within the complex symptomology of the cognitive impairment spectrum in [Parkinson’s],” the researchers wrote.

They added that “this insight could optimize cognitive exams for [Parkinson’s], thereby facilitating earlier identification and more timely introduction of interventions, such as exercise, to slow disease progression.”

Because the analysis involved data from only eight patients, however, the findings will need to be validated “in a large corpus of MoCA exams and patient health data,” the team concluded. The researchers also noted a limitation in the timing of the assessments: the MoCA exams and neuropsychological reports were completed a median of 1 year and 4 months apart, which could have influenced how the test performance aligned with the reported health information.