A global cognitive test may miss early memory problems in smart, well-educated people
STUDY: Neves-Teles T et al, Scandinavian Journal of Psychology 2026;0:1–11
STUDY TYPE: Cross-sectional study
FUNDING: Independent
Background
The Montreal Cognitive Assessment (MoCA) is a popular test for cognitive decline. The 30-point test has a cut off is 25 or below, and it’s suggested to add one point for people with less than college education. But how do we norm the score for people with high education?
This paper points to a MoCA subscale: the Memory Index Score.
Memory Index Score
- Free Delayed Recall: The number of words (out of 5) recalled spontaneously.
- Category-Cued Recall: The number of words recalled after being given a category clue (e.g., “was it a type of fruit?”).
- Multiple Choice-Cued Recall: The number of words recalled after being given multiple-choice options.
The Study
- 78 Brazilian federal public managers (average age 48) screened remotely using the MMSE, the Montreal Cognitive Assessment (MoCA), and an abbreviated IQ estimate
- The MoCA Memory Index Score (MoCA-MIS) was compared to the global MoCA score
- Hierarchical regression identified predictors of cognitive performance
Results
Global MoCA scores looked mostly healthy: 82% of participants fell within the normal range.
The MoCA-MIS was more sensitive. Using the memory-specific index, only 56% scored in the healthy range. The MoCA-MIS flagged 44% as showing mild or more pronounced memory vulnerability. Higher IQ predicted better global MoCA scores.
Age predicted worse performance, but only after IQ was statistically controlled, consistent with the idea that intellectual reserve masks age-related decline until you factor it out.
Limitations
Small sample, cross-sectional design, and remote administration. The cognitive screening tools reflect normative risk, not clinical diagnoses.
How to Interpret the MoCA and Memory Index
You can find the MoCA-Memory Index here, but what is the cut off? For that we’ll need a 2022 paper by Roy Kessels that ran the test in 820 healthy people aged 18-91.
They normalized the scores by age, gender, and educational level. Once you know that, use the tables to find the percentile your patient falls in. Here’s how those percentiles correspond to levels of cognitive impairment in DSM-5:
- Major Neurocognitive Disorder: 3rd percentile or lower
- Mild Neurocognitive Disorders: Below the 16th percentile but above the 3rd
OK, that’s an oversimplification, as DSM really defines these levels by functioning. These cut-offs are for screening, not diagnosis.
In the tables, educational levels are divided as:
- Low: High school diploma or less (vocational training or technical certificates are included here, ie an ISCED level 0-4)
- Average (Av): Associates (2 year college) degree
- High: College (4 year) or higher
For example, a 64 year old woman with a college degree (high education) has a decent MoCA Total Score of 27 (57th percentile) but a Memory Index Score of 10 (5th percentile, suggesting a mild neurocognitive disorder).
These percentiles don’t tell us if she has dementia, but they do suggest some cognitive decline. The reason, and whether we refer to neurology, depends on the whole picture. If she has depression, we’d treat that first.
Trouble reading the tables below? Find them at the full paper.
MoCA Total Score Percentiles
MoCA Memory Index Score Percentiles
A Bell Curve View
You can also interpret the scores based on where they fall on the bell curve. Here’s that view:
— Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
