Reviewed by the Help Dementia Editorial Team — our editors review every article for accuracy against guidance from the National Institute on Aging, the Alzheimer’s Association, and peer-reviewed sources.
Drug companies measure Alzheimer’s treatment success primarily through three categories: cognitive testing scales, biomarkers of brain pathology, and clinical functional decline. The most widely used cognitive test is the Alzheimer’s Disease Assessment Scale–Cognitive Subscale (ADAS-cog), which evaluates memory, attention, and language across a 70-point scale. Companies track how slowly—or whether—test scores decline over time compared to placebo. A drug might be considered successful if it slows cognitive decline by 25 to 35 percent over 18 months, meaning patients treated with the medication lose cognitive function at a measurably slower rate than those receiving placebo.
The measurement landscape has shifted dramatically over the past five years with the rise of blood biomarkers. Previously, drug developers relied almost entirely on cognitive exams and brain imaging (PET scans for amyloid and tau). Now, companies measure phosphorylated tau (P-tau) and phosphorylated tau variants in the blood, along with plasma phospho-tau/amyloid-beta ratios. These blood tests are cheaper, less invasive, and more practical for monitoring thousands of patients. Eli Lilly’s donanemab and Biogen’s lecanemab—recent FDA-approved monoclonal antibodies—were both evaluated partly on their ability to slow cognitive decline as measured by the 11-point Clinical Dementia Rating Scale Sum of Boxes (CDR-SB) and their effects on brain amyloid levels.
Table of Contents
- What Are the Standard Cognitive and Functional Tests?
- How Have Blood Biomarkers Changed Drug Development?
- Why Is the Rate of Cognitive Decline the Primary Endpoint?
- How Do Companies Establish Clinical Significance Versus Statistical Significance?
- What Are the Gaps in How Alzheimer’s Drugs Are Currently Measured?
- How Are Patient-Reported and Caregiver-Reported Outcomes Being Incorporated?
- What Does FDA Approval Actually Require?
- Frequently Asked Questions
What Are the Standard Cognitive and Functional Tests?
The ADAS-cog remains the gold standard for regulatory approval, though it has limitations. It measures word recall, naming objects, following commands, and recognizing written words, but takes 40 to 45 minutes to administer and requires trained raters. The Mini-Cog or Montreal Cognitive Assessment (MoCA) are quicker alternatives used in some trials, taking 10 to 15 minutes, though they capture less granular decline. The Mini-Mental State Examination (MMSE) was once standard but is now considered too insensitive to detect early cognitive changes and has fallen out of favor for drug trials.
For functional decline—whether patients can manage activities like paying bills, shopping, or taking medications—companies use the Functional Activities Questionnaire (FAQ) or the Instrumental Activities of Daily Living Scale (IADL). Clinical reality often diverges from test results. A patient might score slightly better on the ADAS-cog because they memorized the test words from repeated administrations, or they might perform worse because they are fatigued or anxious during testing. The CDR-SB, which evaluates memory, orientation, judgment, and ability to function in the community, relies on information from both the patient and a caregiver, introducing subjectivity. Lecanemab’s pivotal trial showed an 18-month decline of 2.45 points on the CDR-SB in the treatment group versus 3.66 points in the placebo group—a 35 percent slowing of decline that was statistically significant but required approximately one in five patients to receive treatment to prevent one additional point of cognitive decline.
How Have Blood Biomarkers Changed Drug Development?
Blood biomarkers allow companies to identify disease presence and progression without brain imaging. Phosphorylated tau 217 (P-tau217) and phosphorylated tau 181 (P-tau181) correlate with amyloid and tau accumulation in the brain as seen on PET scans. This means a drug company can track whether their medication is actually reducing brain pathology, not just slowing cognitive decline symptoms. Eli Lilly used plasma P-tau217 reductions as a secondary endpoint in the donanemab trial, showing that the drug reduced this biomarker more than placebo, providing biological evidence that the mechanism of action was working.
The major limitation is that biomarker improvement does not always translate to meaningful cognitive or functional benefit. Some early-stage trials of investigational drugs showed dramatic reductions in amyloid or tau levels with no slowing of cognitive decline. Conversely, patients might show cognitive improvement without major shifts in blood biomarker levels. Additionally, the cost and standardization of blood biomarker testing remain inconsistent across laboratories, and most insurance systems do not yet cover these tests for routine Alzheimer’s monitoring in clinical practice. Drug companies running trials can afford centralized laboratory processing, but a patient’s doctor treating them in routine care may not have access to the same biomarker assays.
Why Is the Rate of Cognitive Decline the Primary Endpoint?
Measuring the rate of decline rather than absolute improvement reflects the biological reality of Alzheimer’s disease—current treatments slow progression but do not reverse existing cognitive loss. Slowing decline by even 25 to 30 percent can mean the difference between a patient retaining independence for two additional years versus losing it within six months. For caregivers, this translates to significantly more time before institutional care becomes necessary. When Biogen’s lecanemab slowed CDR-SB decline by 35 percent, that delay meant patients on average reached the threshold for more severe cognitive impairment roughly four months later than the untreated group over an 18-month period.
Disease progression is non-linear, which complicates measurement. Early-stage disease shows faster cognitive decline, while late-stage disease plateaus. A drug tested in early symptomatic Alzheimer’s may show impressive slowing of decline, but the same drug in moderate disease shows minimal benefit because the patients have already lost substantial cognitive reserve. This is why newer trials like those for lecanemab focus on asymptomatic amyloid positivity or mild cognitive impairment stages—the disease is progressing faster, so slowing that rate is more detectable. A company must choose carefully which patient population to test, because endpoint selection directly determines whether the trial will reach statistical significance.
How Do Companies Establish Clinical Significance Versus Statistical Significance?
A treatment can slow cognitive decline in a statistically significant way—meaning the difference is unlikely due to chance—yet provide minimal clinical benefit. Lecanemab showed a p-value of 0.001, indicating strong statistical significance, but the actual point difference on the CDR-SB (1.21 points) represents a difference of roughly 4 to 5 months of slowed decline. A 27-week clinical trial of an investigational drug for early Alzheimer’s showed statistically significant slowing of ADAS-cog decline (p = 0.03) but the effect was only 2.5 points on a 70-point scale—a 3.6 percent slowing, which many clinicians considered too small to justify medication side effects.
The FDA has created guidance documents allowing drugs to receive approval based on slowing decline by approximately 25 percent or greater in early-stage disease. However, this sets a relatively low bar; a drug slowing decline from 2 points per year to 1.5 points per year meets the threshold for approval but means a patient goes from expecting severe dementia in 10 years to expecting it in 13 years. Companies must balance regulatory approval likelihood against commercial viability—a drug showing modest effects can still receive approval and generate revenue, but physicians and patients must understand that the effects are incremental, not transformative.
What Are the Gaps in How Alzheimer’s Drugs Are Currently Measured?
Current measurement approaches undervalue quality of life, psychiatric symptoms, and caregiver burden. The ADAS-cog measures cognitive function but not depression, anxiety, or behavioral changes that dramatically affect daily life. A patient might score marginally better on cognitive tests while experiencing increased apathy or worsening mood—both common in Alzheimer’s disease and often triggered by amyloid-reducing therapies.
Lecanemab showed increased rates of amyloid-related imaging abnormalities (ARIA), a form of brain microhemorrhage or cerebral edema seen on MRI, occurring in roughly 25 percent of treated patients versus 10 percent of placebo, though most were asymptomatic. Caregiving capacity is rarely measured in trials, yet it is critical. A wife managing her husband’s condition may feel relief from delayed cognitive decline, but if that same drug increases behavioral disturbance or requires infusions every two weeks, her stress may actually increase. No major Alzheimer’s drug trial has included caregiver mental health or quality of life as a primary or secondary endpoint, despite growing evidence that caregiver burden predicts institutionalization of the patient more strongly than the patient’s cognitive score.
How Are Patient-Reported and Caregiver-Reported Outcomes Being Incorporated?
Newer trials increasingly include Patient-Reported Outcomes (PROs) such as the Alzheimer’s Disease Cooperative Study–Patient Reported Outcomes (ADCOMS), which asks patients about their own perception of cognitive decline, confidence in memory, and ability to manage daily tasks. Some companies are adding caregiver questionnaires measuring how much supervision is required for specific activities. Eli Lilly’s recent trials for donanemab included secondary measures of functional status through caregiver reports of the patient’s ability to manage finances, medications, and household tasks.
These subjective measures still rely on patient or caregiver insight, which can be unreliable in Alzheimer’s disease. Many patients lack awareness of their own cognitive decline (anosognosia), so their perception of benefit may diverge sharply from objective test results. Caregivers may overstate decline to justify their own stress levels or understate it because they are in denial. Despite these limitations, PROs are increasingly valuable for regulatory decision-making because they address the “so what”—does slowing cognitive decline actually matter to the people living with the disease?.
What Does FDA Approval Actually Require?
The FDA approved lecanemab for early-stage Alzheimer’s based on slowing CDR-SB decline by 35 percent over 18 months, requiring all treated patients to undergo regular MRI monitoring for ARIA. Donanemab was approved through a similar pathway—slowing decline plus biomarker evidence of target engagement. Regulatory approval is based on the balance of benefits and risks established in large randomized controlled trials, typically involving 1,500 to 3,000 patients. The FDA’s guidance specifies that a 25 percent slowing of decline in the early symptomatic stage is sufficient for approval, but companies pursuing approval may design trials to detect larger effects if they wish to differentiate their product.
Once a drug is approved, real-world effectiveness often diverges from trial results. Trial patients are highly screened, tend to be younger and healthier than typical clinic populations, and receive intensive monitoring and adherence support. A patient in routine practice taking lecanemab may miss infusions, have comorbidities not represented in trials, or experience side effects differently than the studied population. Post-marketing surveillance and real-world evidence studies—often conducted by health insurance companies or health systems—measure actual clinical outcomes, but these data are collected months or years after approval and are less rigorously controlled than pivotal trials.
- —
Frequently Asked Questions
What is the ADAS-cog, and why is it used in Alzheimer’s drug trials?
The Alzheimer’s Disease Assessment Scale–Cognitive Subscale is a 70-point test evaluating memory, attention, and language. It is the FDA’s preferred measure for Alzheimer’s drug approval because it is standardized, administered by trained raters, and sensitive to early cognitive changes.
What does a 25 to 35 percent slowing of cognitive decline actually mean for a patient?
It means patients treated with medication lose cognitive function at a slower rate than untreated patients—typically delaying the onset of moderate dementia by several months to a few years, depending on the drug’s effect size.
Can blood biomarkers replace cognitive testing in Alzheimer’s trials?
Not yet. Blood biomarkers show whether a drug engages its biological target (reducing amyloid or tau), but cognitive decline remains the primary measure of whether that biological change translates to meaningful clinical benefit.
Why is amyloid-related imaging abnormality (ARIA) important?
ARIA—brain microhemorrhage or swelling—is a safety concern in amyloid-reducing therapies like lecanemab and donanemab. It occurs in roughly 25 percent of treated patients and requires regular MRI monitoring, adding cost and complexity to treatment.
How do real-world results compare to clinical trial results for Alzheimer’s drugs?
Real-world effectiveness is typically lower than trial results because routine patients have more comorbidities, take medications irregularly, and are not intensively monitored like trial participants. However, real-world data is still emerging for recent approvals.
Do cognitive tests capture everything that matters to patients and caregivers?
No. Cognitive tests miss psychiatric symptoms (depression, anxiety), behavioral changes, and caregiver burden—all of which affect quality of life. Newer trials are beginning to include patient-reported and caregiver-reported outcomes, but these remain secondary endpoints. —
