Most psychiatrists rely on the Folstein Mini Mental State Exam for a rapid, office-based neuropsychological assessment. It is what we were trained on, and we have given it to so many patients that we can pretty much reel off the items from memory.

But the MMSE has some serious limitations. While it is quite sensitive at picking up moderate to severe dementia, it is very poor at screening for patients with mild cognitive impairment (MCI), which is diagnosed in patients who have abnormally poor memory for their age but who do not show significant functional impairment (see TCPR May 2007). Patients with MCI progress to dementia at a rate of 10-15% per year, far higher than the baseline rate of 1-2% per year in normal elderly. In the typical scenario, an elderly patient will come in complaining of memory lapses, and will want to know if they have “normal” age related memory loss or something more serious.

The MMSE is generally unhelpful in such cases, because it’s a little too easy: both normal and MCI patients will typically score 26 or above, the usual cut-off point for cognitive impairment.

But a hot new test, called the Montreal Cognitive Assessment (MoCA) appears to work much better. The test, developed in Quebec with some help from UCLA’s Jeffrey Cummings, was developed specifically to better diagnose MCI. Think of the MoCA as a more challenging version of the MMSE; because it is more challenging, it will pick up a higher proportion
of patients with more subtle amounts of cognitive impairment. Like the MMSE, it is quick to administer (about 10 minutes) and has a maximum of 30 points. In a study comparing the MMSE with the MoCA, the MMSE had a sensitivity of only 18% to detect MCI (meaning it missed 82% of the people with MCI), while the MoCA detected 90% of MCI subjects (Nasreddine ZS et al., J Am Geriatr Soc 2005;53:695-699).

As always, though, high sensitivity comes at the price of somewhat lower specificity (false positives). In the study cited above, for example, the specificity for MCI was 87%, meaning that 13% of actually normal people will be falsely labeled as impaired -- still quite accurate. The really nice thing about the MoCA is that the test and detailed instructions are available (in
several languages) for free at www.mocatest.org. We think the MoCA should become the test of choice for fairly high functioning patients who come in worried about memory loss.

TCPR Verdict:

MoCA: Better than MMSE for mild cognitive impairment.