Daily function measures are the primary regulatory currency for dementia drug approvals. The FDA uses standardized scales to assess whether a drug meaningfully slows cognitive or functional decline, and these measures directly determine whether a medication can be approved for market. Rather than requiring a drug to restore lost abilities, regulators now accept evidence that a drug can slow the rate of decline—and they measure that slowness using specific, quantifiable tools like the Clinical Dementia Rating-Sum of Boxes (CDR-SB) and the Alzheimer’s Disease Assessment Scale-Cognitive Subscale (ADAS-Cog). When Lecanemab was approved in July 2023, the landmark achievement wasn’t that it cured Alzheimer’s disease; it was that the drug demonstrated a 27% slowing of cognitive and functional decline over 18 months in the CLARITY-AD trial, measured on the ADAS-Cog14 scale.
The shift represents a fundamental change in how the FDA evaluates dementia therapeutics. Until roughly 2018-2019, regulators used a strict “dual-endpoint standard” that required drugs to show improvements in both cognitive and functional domains simultaneously. The new framework, introduced through FDA guidance updates, recognizes that slowing decline—particularly in early disease stages where patients have minimal functional deficits—constitutes a clinically meaningful benefit. This change opened the door for anti-amyloid monoclonal antibodies and other disease-modifying approaches that may not restore function but can meaningfully extend the timeline of cognitive stability.
Table of Contents
- What Are Functional Decline Measures and Why They Matter for Approvals
- The FDA’s 2018–2019 Shift Away from Dual-Endpoint Requirements
- How the 27% Benchmark Became the Efficacy Gold Standard
- The Role of Functional Measures in Trial Design and Patient Recruitment
- Amyloid-Related Imaging Abnormalities and the Hidden Cost of Efficacy
- Composite Scores and the Reshaping of Trial Efficiency
- Recent FDA Guidance and the Evolving Approval Pathway for Early Dementia
- Frequently Asked Questions
What Are Functional Decline Measures and Why They Matter for Approvals
The scales used to measure daily function in dementia trials are highly specific instruments, not subjective assessments. The ADAS-Cog includes 13 independent cognitive tests—word recognition, object naming, following commands, constructional praxis, ideational praxis, and orientation, among others—with scores ranging from 0 to 70, where higher scores indicate greater impairment. The CDR-SB (Clinical Dementia Rating-Sum of Boxes) evaluates six functional domains and is particularly sensitive to early-stage cognitive decline, whereas the original CDR scale was developed for moderate to severe dementia and misses subtle early changes. When a drug slows the rate of increase on these scales, that slowness translates into real clinical benefit: fewer months or years of progressive cognitive loss for patients.
The reason these measures are so critical to approval is that they provide objective, reproducible data about disease progression. A patient’s subjective report that “I’m forgetting less” or “I feel sharper” cannot support an FDA approval. Instead, regulatory reviewers examine data showing that treated patients scored 3 points higher on the ADAS-Cog at trial completion compared to untreated controls—meaning their cognitive decline was measurably slower. The FDA specified that these functional and cognitive measures must be the primary outcome of the trial, not secondary measures. This requirement ensures that a company cannot hide a marginal drug by focusing attention on biomarker changes or inflammatory markers instead of the actual clinical benefit that patients and caregivers can observe.
The FDA’s 2018–2019 Shift Away from Dual-Endpoint Requirements
Before 2018-2019, the FDA’s guidance for Alzheimer’s disease drugs required what was known as the “dual-endpoint standard”: a drug had to demonstrate meaningful slowing of decline in BOTH cognitive measures AND functional measures (typically assessed using scales like the ADCS-ADL-MCI, which measures activities of daily living in mild cognitive impairment). This dual requirement meant that a drug showing strong cognitive benefit but minimal functional change could not be approved. The rationale was sound—caregivers wanted assurance that cognitive preservation would translate into maintained ability to pay bills, manage medications, or perform household tasks. However, the dual-endpoint standard also meant that very few drugs could clear the regulatory bar, because functional deficits in early-stage disease are often minimal. The FDA’s updated guidance, formalized in draft and final forms between 2018-2019, fundamentally reoriented the approval framework.
For early-stage Alzheimer’s disease, regulators now recognize that meaningful cognitive benefit alone—without demonstrated functional benefit—can justify approval. The rationale is straightforward: in patients with mild cognitive impairment or early dementia, slowing cognitive decline can prevent downstream functional decline. A patient who maintains cognitive function longer will naturally maintain daily living abilities longer as well. This shift eliminated the requirement to wait until functional changes become large enough to measure. The downside is that critics argue regulators may now approve drugs with very modest cognitive benefits, since the evidence bar no longer requires proof of functional preservation.
How the 27% Benchmark Became the Efficacy Gold Standard
When the FDA approved lecanemab (Leqembi) in July 2023 based on CLARITY-AD trial data, the headline efficacy figure was 27% slowing of cognitive and functional decline at 18 months compared to placebo. This figure has since become the gold standard against which other anti-amyloid therapies are measured. The 27% figure does not mean that lecanemab reversed cognitive decline or halted it entirely; it means that the treated group’s rate of decline was 27% slower than the placebo group’s rate. In practical terms, patients in the treated group experienced a decline equivalent to roughly 5 months of additional stability compared to the placebo group over the 18-month trial period.
For dementia, where cognitive decline accelerates without intervention, this slowing effect is considered a substantial benefit. Donanemab (Kisunla), approved by the FDA in July 2024, also demonstrated statistically significant slowing of cognitive and functional decline in the TRAILBLAZER-ALZ 2 trial, though the precise efficacy percentage differs slightly from lecanemab’s published data depending on how outcomes are calculated and which endpoint is emphasized. The fact that both drugs show broadly similar efficacy (around 25-27% slowing) has become a reference point: regulators now view this magnitude of decline-slowing as clinically meaningful, and companies developing competing anti-amyloid therapies use this benchmark to plan their own trials and set statistical power assumptions. This creates a floor for approval—drugs showing substantially less than 27% slowing face skeptical regulatory review, while drugs matching or exceeding this threshold have a clearer path to approval.
The Role of Functional Measures in Trial Design and Patient Recruitment
Clinical trial protocols for dementia drugs must specify in advance how functional decline will be measured and what constitutes a meaningful difference between treated and placebo groups. Sponsors designing trials must choose between established scales like the CDR-SB, ADAS-Cog, ADCS-ADL-MCI, or newer composite measures like ADCOMS (Alzheimer’s Disease Composite Score). The choice of scale directly affects trial size and duration. For example, the ADCOMS scale, which combines cognitive and functional items from existing measures, has demonstrated greater sensitivity to early disease decline than the CDR-SB alone.
This improved sensitivity allows researchers to recruit smaller cohorts and complete trials in shorter timeframes because the scale can detect meaningful change with fewer enrolled patients. The practical consequence is that sponsors increasingly favor composite measures that combine elements from cognitive and functional assessments into a single score. This choice accelerates development timelines and reduces costs—advantages that translate into drugs reaching patients faster and at lower development expense. However, there is a regulatory tradeoff: composite measures are statistically more complex, and when a company reports a benefit on ADCOMS, regulators must then verify that the component scales (the specific functional tests within the composite) also show consistent benefit. A drug that improves cognitive items within ADCOMS but shows no change—or even slight worsening—on the functional items may face scrutiny during FDA review, even if the overall composite score is favorable.
Amyloid-Related Imaging Abnormalities and the Hidden Cost of Efficacy
The anti-amyloid monoclonal antibodies that demonstrated 27% decline-slowing benefits also carry a safety risk that regulators must weigh against efficacy. Amyloid-related imaging abnormalities (ARIA) occur in a subset of treated patients and manifest as either ARIA-E (amyloid-related imaging abnormalities—edema) or ARIA-H (microhemorrhages). In the CLARITY-AD trial of lecanemab, approximately 21% of treated patients experienced ARIA-E, compared to 9% in the placebo group. Most cases were asymptomatic and detected only through brain MRI, but symptomatic cases can cause cognitive decline, headaches, or confusion. This means that a drug approved for its 27% slowing benefit simultaneously introduces a new risk of brain swelling in roughly 1 in 5 treated patients.
The regulatory balance is contentious. The FDA approved lecanemab despite ARIA risk because the decline-slowing benefit was deemed to outweigh the safety concerns for patients with early-stage disease who are closely monitored. However, this decision created a medical management burden: patients on anti-amyloid therapies now require regular MRI surveillance to detect asymptomatic ARIA, and some must discontinue the drug if imaging abnormalities develop. This monitoring cost—both financial and in terms of patient burden—is often not captured in the efficacy metric of 27% slowing. A patient who achieves 5 months of additional cognitive stability but then develops ARIA-E and must discontinue treatment may not experience a net clinical benefit compared to the untreated path. Regulatory guidance increasingly notes that ARIA risk varies by APOE4 genotype, with APOE4 homozygotes facing substantially higher risk, yet routine APOE4 testing before treatment initiation is not yet standard practice.
Composite Scores and the Reshaping of Trial Efficiency
The FDA’s acceptance of composite measures like ADCOMS has accelerated drug development by reducing the sample sizes required for statistically powered trials. ADCOMS combines four cognitive subscales from the ADAS-Cog, two items from the Mini-Cog or Mini-Mental State Exam, and all six items from the CDR-SB functional assessment. The statistical sensitivity of this composite measure is substantially higher than any individual scale alone, meaning that a smaller number of enrolled patients can detect the same magnitude of treatment effect. Where a trial using CDR-SB as the primary outcome might require 1,200 participants, an identical trial using ADCOMS can achieve the same statistical power with perhaps 800-900 participants. This reduction in sample size shortens trial duration, reduces recruitment burdens on clinical sites, and decreases overall development costs.
However, composite measures introduce interpretability challenges. When a company reports that a drug improved ADCOMS by X units, regulators and clinicians must then look at the subcomponents to understand which functional domains actually improved and which did not. A drug could theoretically improve cognitive items sufficiently to yield a positive ADCOMS result while showing no benefit—or even harm—on the functional activity items. The FDA’s March 2024 draft guidance on Early Alzheimer’s Disease disease explicitly addressed this complexity, noting that functional assessment tools designed for moderate-to-severe dementia are often too insensitive to detect change in early-stage disease. This technical detail has real consequences: it means that “functional improvement” claims in early AD trials must be interpreted cautiously, and regulators now separately evaluate whether functional measures are even appropriate for the disease stage being studied.
Recent FDA Guidance and the Evolving Approval Pathway for Early Dementia
In March 2024, the FDA released a draft guidance titled “Early Alzheimer’s Disease: Developing Drugs for Treatment,” which clarified the regulatory expectations for drugs targeting mild cognitive impairment or mild dementia stages. The guidance explicitly states that functional assessment tools developed for moderate-to-severe dementia are often not sensitive enough for early disease; a patient with mild cognitive impairment may have zero functional deficits despite cognitive decline on formal testing. In response, the FDA now permits alternative primary endpoints for early AD trials, including cognitive assessments alone (without functional co-endpoints), amyloid biomarker outcomes, and time-to-event analyses (such as time to progression to moderate dementia). This represents a further evolution from the 2018-2019 shift: regulators are now comfortable approving early-stage dementia drugs based on cognitive benefit or biomarker change alone, without any functional measurement at all.
The practical implication is that the next generation of anti-amyloid or anti-tau therapies may be approved with primary endpoints that do not directly measure daily function, relying instead on cognitive scales or brain biomarker clearance. This accelerates approval timelines but raises a persistent question among patient advocates and clinical experts: does amyloid clearance or cognitive slowing truly translate into meaningful daily-life benefit? The expert commentary in medical literature from 2024-2026 reflects both optimism and caution. Optimism stems from the fact that lecanemab and donanemab clearly slow decline, and earlier intervention—catching disease before substantial amyloid accumulation—may produce larger clinical benefits. Caution stems from the recognition that 27% slowing represents a modest benefit, not a cure or reversal, and that ARIA and other safety concerns remain incompletely understood in long-term use. For patients and families, the approval of new drugs represents hope; for regulators, it represents a careful balance between extending treatment options and maintaining evidentiary standards for safety and efficacy.
- —
Frequently Asked Questions
What does “27% slowing of decline” actually mean in practical terms?
It means patients on lecanemab experienced roughly 5 additional months of cognitive stability compared to untreated patients over an 18-month trial—they declined more slowly, not that they were cured or improved.
Why did the FDA stop requiring both cognitive AND functional improvement?
In early-stage dementia, patients have minimal functional deficits, making functional change undetectable. The FDA now recognizes that slowing cognitive decline can prevent downstream functional decline.
What is ARIA and why does it matter to the approval decision?
ARIA (amyloid-related imaging abnormalities) is brain swelling or microhemorrhages that occur in roughly 1 in 5 patients on anti-amyloid therapies. The FDA approved lecanemab despite this risk because decline-slowing benefit was deemed to outweigh safety concerns for closely monitored early-stage patients.
How do composite measures like ADCOMS change drug development?
Composite measures are more statistically sensitive to change, allowing trials to enroll fewer patients and complete faster. This accelerates development but requires careful interpretation of which components actually improved.
Can a drug be approved for early dementia based only on biomarker change, not functional measures?
Yes—the FDA’s March 2024 guidance permits primary endpoints based on cognitive testing or amyloid biomarker clearance alone, without requiring functional assessment for early-stage disease.
How often do patients on anti-amyloid therapies need brain MRI to check for safety issues?
Lecanemab requires regular MRI monitoring to detect asymptomatic ARIA, though exact frequency varies by site and clinical protocol; roughly every 6-12 months is typical. —
