Clinical trials in advanced dementia face a fundamental problem: the population most likely to benefit from new treatments is often the least able to participate in the rigorous testing process. As dementia progresses to advanced stages, patients lose the cognitive ability to understand what a trial entails, communicate their experiences, or consent to participation. This creates a paradox that researchers have struggled with for decades. A patient with advanced Alzheimer’s disease—someone who may no longer recognize family members or speak in complete sentences—cannot reliably tell researchers if an experimental drug is helping, hurting, or doing nothing at all. Yet these same advanced-stage patients are the ones most desperately in need of effective interventions.
The challenges cascade through every phase of trial design and execution. Unlike trials for cancer or heart disease, where patients can report symptoms and side effects, dementia trials rely on proxy reporters (usually adult children or spouses) to assess cognitive and behavioral changes in people who cannot advocate for themselves. The disease itself is unpredictable—two patients with the same dementia subtype may progress at entirely different rates, making it nearly impossible to predict who will benefit or suffer harm. Researchers must also contend with high dropout rates, ethical concerns about informed consent, and the difficulty of isolating a drug’s effects from the natural decline of dementia. These obstacles do not make dementia trials impossible, but they make them substantially more complex and expensive than trials for many other conditions. Understanding these challenges matters for families considering participation, for clinicians counseling patients about experimental options, and for policymakers deciding how to accelerate research in an aging population.
Table of Contents
- Why Is Informed Consent So Difficult in Advanced Dementia?
- How Cognitive Decline Complicates Baseline Measurements and Study Design
- Recruitment and Retention Barriers in Dementia Trials
- Measuring Outcomes When Communication Breaks Down
- Safety Monitoring When Patients Cannot Report Symptoms
- The Role of Caregivers: Support and Potential Conflict
- What Dementia Trials Reveal About the Disease Itself
Why Is Informed Consent So Difficult in Advanced Dementia?
Informed consent—a person’s voluntary agreement to participate in a trial after understanding the risks and benefits—is a cornerstone of ethical research. In advanced dementia, this becomes nearly impossible to obtain from the patient themselves. Dementia gradually erodes the cognitive capacities required to give meaningful consent: memory (to retain information), reasoning (to weigh risks against benefits), and appreciation (to understand how the trial affects their own situation). A person in the moderate to advanced stages may consent to the trial one day and have no recollection of that conversation the next. Because the patient cannot truly consent, researchers typically obtain consent from a legal surrogate—a healthcare proxy or family member appointed through advance directives.
This creates a second layer of difficulty. The surrogate must decide based on what the patient would have wanted, information that is often unclear or absent. Did the patient ever express interest in experimental treatments? Would they accept the risks of an unproven drug? Many families agonize over this decision, and some disagreements between family members can become contentious. A study of surrogate decision-making in dementia trials found that about 30% of surrogates reported feeling unsure whether they were making the “right” choice for their relative. There is also the question of ongoing consent—whether a patient who deteriorates during a trial should be withdrawn, even if the surrogate originally agreed to participation. If a patient becomes increasingly agitated or physically resistant during study procedures, is that a sign they would withdraw if they could communicate? Researchers must navigate these questions without clear guidelines.
How Cognitive Decline Complicates Baseline Measurements and Study Design
One of the first steps in any clinical trial is establishing a baseline—a snapshot of the participant’s status before treatment begins. This baseline serves as the comparison point to measure whether the drug works. In advanced dementia, establishing a reliable baseline is extraordinarily difficult because cognitive decline is ongoing and variable. Two patients enrolled in the same trial on the same day will likely progress differently. One may decline sharply over six months while the other plateaus. This natural variability means that a large number of participants are needed to detect whether an experimental drug truly slows decline or whether observed differences are simply the result of random variation.
A trial of a Lewy body dementia treatment might require 400 or 500 participants to have enough statistical power to draw conclusions—compared to perhaps 100 participants for a drug affecting a single, measurable biomarker. This makes dementia trials extremely expensive and time-consuming. The phase III trial for the Alzheimer’s drug lecanemab involved over 1,700 participants and cost hundreds of millions of dollars to conduct over several years. The unpredictability also makes it hard to choose the right outcome measures. Should researchers measure cognitive decline using a standardized test like the Mini-Cog? But that test requires the patient to cooperate and follow instructions—something an advanced dementia patient may not be able to do consistently. Should they measure functional decline instead—the patient’s ability to dress, eat, or recognize people? That requires reliable information from the caregiver, who may not observe the patient at consistent times or may have different expectations for what constitutes “progress.”.
Recruitment and Retention Barriers in Dementia Trials
Finding eligible participants for dementia trials is far harder than it sounds. The disease affects millions of Americans, but many of them are in advanced stages and live in care facilities or depend entirely on family caregivers. Recruiting from nursing homes or assisted living requires coordination with facility administrators, staff training, and often, acceptance of the extra burden on already-stretched nursing staff. Many facilities are reluctant to enroll residents because trial participation adds administrative requirements and could disrupt the facility’s routines. Recruitment is also hampered by the underdiagnosis of dementia. Nationally, only about 50% of people with dementia have received a formal diagnosis. Many families attribute cognitive changes to normal aging or believe there is no point in seeking diagnosis if no cure exists. For researchers trying to identify potential participants, this means sifting through many appointments at memory clinics or primary care practices.
A trial site might screen 500 patients to enroll 50 who meet all inclusion and exclusion criteria. Retention—keeping participants in the trial until it ends—is another major challenge. People with advanced dementia are medically fragile. They may develop pneumonia, urinary tract infections, or acute illnesses that are unrelated to the trial but make continued participation impossible. Caregiver burden is also a factor. Some family members agree to trial participation thinking they can manage the time commitment but later find that attending monthly study visits, undergoing cognitive testing, and scheduling blood draws is too much. A meta-analysis of dementia trial retention found that rates averaged 60-70% over 12-month follow-up periods, meaning 30-40% of enrolled participants dropped out. This is substantially higher than dropout rates in trials for many other chronic conditions.
Measuring Outcomes When Communication Breaks Down
In most clinical trials, the primary outcome—the main thing researchers measure to determine if the drug works—is something the patient can report or a clinician can measure directly. In cancer trials, it might be tumor shrinkage. In diabetes trials, it might be blood glucose levels. In dementia trials, this becomes extraordinarily murky. The most important outcome—slowing cognitive decline—cannot be measured directly in a patient with advanced dementia who cannot cooperate with cognitive testing. Instead, researchers rely on caregiver reports or observational scales like the Clinical Dementia Rating or the Functional Assessment Staging. These scales are subjective. One caregiver might interpret subtle changes in behavior as meaningful decline while another sees the same changes as normal daily fluctuation.
A patient who becomes more withdrawn might be experiencing depression, medication side effects, or disease progression—and it may be impossible to distinguish which. This introduces noise into the data and makes it harder to detect whether a drug is actually working. Some trials attempt to use objective biomarkers—changes in the cerebrospinal fluid, brain imaging, or blood tests for tau and amyloid proteins. These are valuable but imperfect. A decline in a biomarker does not automatically mean the patient is getting worse clinically, and clinical improvement does not necessarily show up in biomarkers. Researchers must often measure multiple outcomes and hope that one of them shows a clear signal of benefit. The lecanemab trial, for instance, showed slowing of cognitive decline (a 35% reduction in decline rate over 18 months) but the absolute difference was modest—a delay of about five months in functional decline. Many families have questioned whether such a modest benefit justifies the risks of the drug (which includes a small risk of amyloid-related imaging abnormalities, a type of brain inflammation).
Safety Monitoring When Patients Cannot Report Symptoms
Every clinical trial includes safety monitoring—systematic tracking of adverse events (side effects or harmful reactions) to ensure participants are not being harmed. This is particularly critical in dementia trials because the participants are vulnerable and often cannot report problems themselves. If a patient in a typical drug trial develops dizziness, nausea, or confusion, they can report it. A patient with advanced dementia cannot. Researchers and caregivers must infer adverse events from behavioral changes or physical symptoms. A patient who becomes more agitated might be having a medication reaction, or they might be in pain, or they might be responding to a change in their environment. Distinguishing these possibilities is difficult and often requires additional investigation—blood tests, imaging, medical evaluation.
This adds complexity and cost to trial monitoring. There is also the risk of underreporting. A family caregiver who is exhausted might not report subtle changes, or they might attribute new symptoms to dementia rather than the study drug. A caregiver who believes strongly in the potential benefit of the experimental treatment might downplay side effects. To address this, responsible trials include regular check-ins with caregivers and, ideally, direct assessment by study staff. But this increases the burden on trial participants and can contribute to higher dropout rates. Trials also sometimes require hospitalizations or emergency department visits to be reported—a safety signal that is objective and hard to miss.
The Role of Caregivers: Support and Potential Conflict
Family caregivers are essential to conducting dementia trials, but their involvement also introduces new sources of difficulty. Caregivers provide the information researchers need, escort participants to appointments, and help monitor for side effects. Without caregivers, most advanced dementia trials would be impossible to conduct. However, caregiver interests may not always align perfectly with the trial’s scientific goals or even with the patient’s wishes.
Some caregivers are motivated by hope—they enroll a relative in a trial because they desperately want a cure and are willing to accept risk. Others are motivated by a sense of duty or altruism; they enroll to help research even if they do not expect direct benefit. Occasionally, caregivers may be motivated by other concerns: access to additional medical monitoring, relief from caregiving burden during trial visits, or even financial incentive payments. A caregiver might pressure a patient to continue in a trial despite obvious decline or distress, or might withdraw a patient prematurely because they are skeptical of benefit. Ethical oversight committees (Institutional Review Boards, or IRBs) screen for these conflicts, but they cannot monitor every interaction between caregiver and patient.
What Dementia Trials Reveal About the Disease Itself
Despite these obstacles, dementia trials have produced important insights—not only about whether specific drugs work, but about the disease itself. Many trials have revealed that dementia progresses more variably than researchers once believed. Some patients decline rapidly while others have long plateaus. Some show improvements in certain domains (memory, for instance) while declining in others (functional ability or behavior). This heterogeneity means that a drug that helps one person may be useless for another, which has led to growing interest in biomarker-driven trials that enroll only patients whose brain pathology matches the drug’s mechanism.
Trials have also shown that cognitive decline and functional decline do not always move together. A patient can maintain the ability to perform daily activities even while memory worsens, or vice versa. A medication that improves one domain might leave the other unchanged. This has practical implications: it suggests that treatment goals in advanced dementia might need to be more modest and specific than researchers once hoped. Rather than expecting a drug to reverse or stop all aspects of dementia, it might be realistic to aim for slowing decline in one specific area—memory, perhaps, or behavioral symptoms—while accepting that other domains will continue to deteriorate.
- —
