SAGE Test Scoring: How the Results Are Interpreted

Why SAGE Scoring Works Differently

Most people expect a test to come with an answer key. SAGE intentionally does not. The developers at The Ohio State University Wexner Medical Center explain that there is no self-scoring sheet because many questions have more than one acceptable answer — so the instruction is to take your completed answer sheet to your doctor, who scores it and talks through the results with you (Ohio State University Wexner Medical Center).

That design choice matters. It keeps the test easy to take on your own while keeping interpretation in trained hands, where a single number can be weighed against everything else a clinician knows about you. If you are new to reading any cognitive result, our broader guide to understanding your cognitive test results is a good place to start.

Key Facts at a Glance

• SAGE is scored out of a maximum of 22 points, with the total summed across the tasks it samples (Scharre et al., 2017).
• The number reflects seven domains of thinking: orientation, language, memory, executive function, abstraction, calculations, and visuospatial abilities.
• Scoring is done by a clinician, typically in about 30 to 60 seconds — there is no published self-scoring key.
• In validation studies, 17 or higher on the paper version was the cutoff that best separated normal cognition from impairment.
• A score is interpreted in context, including age and education, not read as a fixed pass-or-fail line.
• SAGE is a screening tool, not a diagnosis, and a lower score is a prompt for further conversation.

How Is the SAGE Test Scored?

SAGE is a four-page, pen-and-paper test you complete on your own. Ohio State reports that it takes most people 10 to 15 minutes, with no time limit, and that it comes in four interchangeable forms — you only need to take one (Ohio State University Wexner Medical Center). Once you hand in the answer sheet, a clinician scores it.

The developers' published description shows how the 22 points are distributed across domains (Scharre et al., 2017):

The form also asks about background details — such as education, past strokes or head injuries, family history, and current symptoms like changes in memory, mood, or daily activities. Those answers are not scored, but they give your clinician context for reading the number (Scharre et al., 2017).

This breakdown is about how the scale is built, not how to answer it. The questions themselves, and what counts as a correct response, are left to the clinician — which is exactly why the result has to be scored by one.

What the 22-Point Total Reflects

SAGE does not measure a single skill and turn it into a grade. Instead, it samples several areas of thinking and adds them into one total, so the number is a broad snapshot rather than a precise measure of any one ability (Scharre et al., 2017). A higher total reflects stronger performance across the sampled tasks; a lower total reflects more difficulty on them.

Because the score blends several domains, two people can arrive at the same number by different routes — one may find calculations harder, another may struggle more with memory. That is part of why the raw total is only a starting point for a clinician, who can look at which areas contributed to the score, not just the sum. For a look at how this compares with other instruments, see our explainer on what cognitive test scores mean.

What Do SAGE Score Ranges Mean?

The score ranges most often quoted for SAGE come from its validation research, and it helps to know exactly what they are:

• 17 to 22 (paper version): In the developers' studies, a score of 17 or higher was the cutoff that best separated people with normal cognition from those with cognitive impairment. The 2017 study found that same cutoff again, matching the original validation study (Scharre et al., 2017). A 2024 primary care study likewise treated 17 or above as the normal range (Scharre et al., Frontiers in Medicine, 2024).
• 16 or higher (digital version): People scored about one point lower on the tablet version (eSAGE) than on paper, so the best normal-range cutoff there was 16 or higher (Scharre et al., 2017).
• Lower scores: In the same research, a digital score of 14 or below best separated people with dementia from those with normal thinking. Mild cognitive impairment was harder to pin down: scores for people with MCI overlapped with the normal range, and no single cutoff separated the two groups cleanly (Scharre et al., 2017).

Two cautions keep these ranges in proportion. First, they are group-level research findings, not a verdict about any one person — some people with impairment score in the normal range, and some healthy people score below it. Second, the developers note that SAGE has age and education norms (Scharre et al., 2017), so the same total can mean something different for an 85-year-old than for a 60-year-old, or for someone with less formal schooling. Our guide to why age and education change how scores are read explains that adjustment in more depth.

Is There a SAGE Test Answer Key?

No — and that is deliberate. Ohio State does not provide a self-scoring sheet because many SAGE questions have several acceptable answers, and it asks that the test be scored by your physician (Ohio State University Wexner Medical Center). Answer keys you might find elsewhere online are not an official scoring tool.

There is also a practical reason to skip them. A screening test only means something if it captures how you think on an ordinary day. Looking up answers, or rehearsing the questions beforehand, turns the result into a measure of preparation rather than cognition — and it can hide the very changes a clinician is trying to catch. If you want a meaningful result, take the test cold, in ink, and let your doctor do the scoring. For background on the test itself, see our overview of what the SAGE test is.

How a Clinician Reads a SAGE Score

A physician does not interpret a SAGE total in isolation. In practice, several things shape what a given number means:

• Your history — a recent change from your own baseline often matters more than the absolute score
• Age and education — the same total can carry a different meaning depending on both
• Daily functioning — how thinking affects everyday activities like managing finances or medications
• Other findings — mood, sleep, medications, and physical health that can all move a score
• Follow-up over time — one screen is a single snapshot, not a trend

SAGE comes in four interchangeable versions, which lets a clinician repeat it later without reusing the same form (Ohio State University Wexner Medical Center). Ohio State also suggests that doctors keep a completed test on file as a baseline for the future. Comparing a later score with an earlier one is often more informative than any single result, because it shows direction rather than just position.

How SAGE Scores Change Over Time

The strongest evidence for tracking SAGE scores comes from a 2021 study of patients at Ohio State's memory disorders clinic who took SAGE and the MMSE repeatedly over several years (Scharre et al., Alzheimer's Research & Therapy, 2021). Among 424 people who met the study's criteria:

• Scores stayed stable for people with subjective memory concerns and for people with mild cognitive impairment (MCI) whose condition did not progress.
• Scores declined about 1.9 points per year for people with MCI who went on to develop Alzheimer's dementia.
• SAGE picked up a significant decline at least six months earlier than the MMSE in the groups whose thinking declined — for example, about 14 months from baseline versus about 20 months for people with MCI who converted to Alzheimer's dementia.

The same study found a limit worth knowing: once scores fall below about 5, SAGE stops showing useful change over time, so it is best suited to earlier stages (Scharre et al., 2021). The authors also note that their data come from a single clinic. These are group averages, so they cannot predict what will happen for any one person. A stable score over repeat tests is one useful data point, not an all-clear. If you or your family notice memory or thinking changes that persist or get worse, bring them to a clinician even when the score has not moved. A steady drop from your own baseline is also the kind of pattern a clinician will want to look into.

How Accurate Is the Score?

Validation research helps explain why clinicians take a SAGE result seriously while still treating it as a screen. In the original validation study, SAGE identified about 79% of people with cognitive impairment (its sensitivity) while correctly classifying about 95% of those without impairment (its specificity), and its scores correlated strongly with detailed neuropsychological testing (Scharre et al., 2010). The tablet version performed similarly, with 90% specificity and 71% sensitivity using a cutoff of 16 (Scharre et al., 2017).

Those numbers also show why no score should be read as the final word. A screen that catches most, but not all, cases is exactly the kind of tool meant to flag who needs a closer look — not to confirm or rule out a condition on its own.

Its real-world value is early detection. When SAGE was offered at 45 community events, 1,047 adults over 50 completed it, and 28% were identified as having cognitive impairment (Scharre et al., 2014). In a 2024 primary care study, providers who used SAGE documented new cognitive conditions in 9% of patients, compared with 1.5% among those who did not take it — about six times as often (Scharre et al., Frontiers in Medicine, 2024).

Is the SAGE Test a Test for Alzheimer's?

SAGE is often searched as a "test for Alzheimer's," but that is not quite what it does. It was designed to screen for mild cognitive impairment and early dementia in general (Scharre et al., 2010). A lower score says that thinking is weaker than expected in some areas; it cannot say why.

That distinction matters because many things besides Alzheimer's disease can lower a cognitive score. The Alzheimer's Association lists depression, untreated sleep apnea, delirium, medication side effects, thyroid problems, certain vitamin deficiencies, and excessive alcohol use as common causes of dementia-like symptoms — and it stresses that there is no single test that can determine whether someone is living with Alzheimer's or another dementia (Alzheimer's Association). Diagnosing Alzheimer's specifically relies on a fuller workup that may include medical history, neurological exams, cognitive and functional assessments, brain imaging, and blood or spinal fluid tests.

Where SAGE does help is in tracking: as described above, repeat scores can reveal the gradual decline that often accompanies progression from MCI to Alzheimer's dementia.

Paper vs. Digital SAGE Scores

SAGE also exists as a tablet-based version, known as eSAGE and offered commercially as BrainTest under a license from Ohio State. It uses the same questions as the paper test (Scharre et al., 2017). Because people score about one point lower on the digital version, a clinician will compare like with like rather than treat the two totals as interchangeable. For a broader view of how tablet and app-based tools fit into care, see our guide to how digital cognitive assessment works.

What a Lower-Than-Expected Score Means

A score below what a clinician expected does not, by itself, diagnose anything. It tells your provider that a fuller evaluation could help clarify the cause. Many contributors to a lower score are treatable, and no single test settles the question — clinicians combine cognitive assessments with medical history, physical and neurological exams, and sometimes imaging or lab work (Alzheimer's Association). Our guide to what happens after a low cognitive screening score walks through that process step by step.

For a sense of how other screens handle their own numbers, our guides on how MMSE scores are read and what a good MoCA score is walk through the same idea: a cutoff is a convention, not a biological boundary.

What Happens Next

After your clinician scores a SAGE, they will explain the result in the context of your overall health and decide, with you, what to do next. Sometimes that is reassurance and a plan to recheck later so there is a baseline to compare against; other times it is a referral for more detailed testing or a check for reversible causes. In the 2024 primary care study, follow-up after SAGE included extra visits focused on cognition and referrals for things like neuroimaging, lab work, neurology or psychiatry consultations, social work, and financial or legal planning (Scharre et al., Frontiers in Medicine, 2024). Either way, the goal is a clear, sensible next step rather than a label.

Taking the Next Step

To understand the exam behind the number, read our overview of what the SAGE test is.

To complement a clinician-scored paper screen with a structured check-in you can do from home and share with your doctor, learn how Orena's clinically validated at-home test works.