Screening for Alzheimer's disease: Strength in numbers

Identifying patients at risk of developing a major neurocognitive disorder (also known as dementia) is a major public health challenge as the population ages. Certain populations are particularly at risk, especially patients with signs of cerebrovascular disease. ¹ Screening for cognitive disorders requires validated tests that are later supplemented by specialized examinations in a memory clinic. In clinical practice, two cognitive screening tests are widely used: the Mini-Mental State Examination (MMSE) ² and the Montreal Cognitive Assessment (MoCA). ³ Although these tests are similar in that they are both graded over 30 points and have common subtests, they also have differences. The MMSE was published first and is widely used in geriatric settings. This scale is also used to stratify major neurocognitive disorder: mild (>20 points), moderate (20–10 points), or severe (<10 points). However, the MMSE lacks sensitivity for screening prodromal patients and has a ceiling effect. ⁴ The MoCA, which was published more recently, is more often used with younger patients, especially in stroke units ⁵ or with patients who have Parkinson's disease. ⁶ The main difference between the MMSE and the MoCA is that the MoCA evaluates executive functions. The MoCA is considered more sensitive than the MMSE at the prodromal stage; however, its greater difficulty results in lower performance at the advanced stage, resulting in a floor effect. ⁷ It is not always easy for clinicians to decide which test to use in daily practice.

In their study, “ Utility of combination of Mini-Mental State Examination and Montreal Cognitive Assessment to predict incident Alzheimer's disease dementia in patients at high vascular risks”, ⁸ Kazuo Kitagawa and his colleagues propose a protocol to estimate the risk of converting to Alzheimer's disease (AD) dementia in a large cohort of patients at high vascular risk using a combination of the MMSE and the MoCA at baseline. The thresholds used have been validated in the literature for screening mild cognitive impairment (MCI): MMSE < 28 or MoCA < 26.

One caveat when administering two tests with some identical questions is the test-retest effect. For example, a patient who has already responded to questions about orientation or serial sevens during a previous test will find it easier to answer the same questions a second time. The authors addressed this issue by first administering the entire MMSE and then adding MoCA subtests that were not part of the MMSE.

In this study, the authors demonstrate that patients whose results from both tests were below the pathological threshold were at a much higher risk of converting to AD dementia, with a hazard ratio (HR) of over 20. In contrast, patients with results from only one test or no test below the pathological threshold for MCI remained at low risk of converting to dementia. This HR is unusually high. For example, according to a recent meta-analysis, ⁹ amyloid-positive subjects had HRs of 3.16 and 2.33 for converting from mild cognitive impairment (MCI) to dementia and from cognitively unimpaired to MCI/dementia, respectively. The lack of specificity in the clinical criteria could have contributed to the increased hazard ratio.

One way to improve the use of cognitive screening tests is to consider socio-educational level. ¹⁰ The authors incorporated years of education as a covariate in survival analyses. However, at the individual level, the chosen thresholds may misclassify healthy subjects with lower performance due to fewer years of education as having MCI or a major neurocognitive disorder. Therefore, we recommend adjusting the MMSE and MoCA thresholds with the years of education in clinical practice.

An important issue to consider is that the definition of AD used in this paper relies on clinical criteria, ¹¹ and diagnoses have not been confirmed using blood or cerebrospinal fluid biomarkers. Therefore, it is likely that some of the screened patients have a major neurocognitive disorder due to non-AD etiology. This could be due to vascular cognitive impairment or a combination of neurodegenerative and vascular etiologies (i.e., mixed dementia) or multiple neurodegenerative diseases. The lack of biomarker confirmation of AD is a major limitation of this study. The authors have likely identified the general risk of converting to dementia, rather than the specific risk of converting to AD dementia. This is why such a screening procedure cannot replace a more specialized evaluation in a memory clinic, such as neuropsychological testing, multimodal brain imaging, and biomarker dosing, to confirm the etiology. Many etiologies can present with the canonical “amnestic syndrome of hippocampal type”, which was previously considered specific to AD. Behavioral variant of frontotemporal dementia (bvFTD), limbic-predominant age-related TDP-43 encephalopathy (LATE), and dementia with Lewy bodies (DLB) may also present with progressive amnesia.^12–14

Another point to consider in defining major neurocognitive disorder is the alteration of functional independence, which is not addressed in screening tests such as the MMSE or MoCA. This can be evaluated using the Instrumental Activities of Daily Living (IADL) scale. ¹⁵ Four IADL items are particularly relevant: the ability to use a telephone, means of transportation, managing one's own medication, and handling finances. Thus, an alteration of any of these four IADL items may classify the patient as having dementia.

In conclusion, this original study contributes to the broader effort of screening patients with AD and other dementias sooner and more efficiently. Combining MMSE and MoCA may enable more precise screening of patients at risk for major neurocognitive disorders. With the advent of targeted anti-amyloid immunotherapies, screening of patients at risk for the disease is the first step towards faster, more targeted care, that best meets patients’ needs.