Alzheimer’s patients who took donepezil and memantine together had a measurably higher chance of surviving five years than those on either drug alone or no treatment at all, according to a large analysis of electronic health records. The study, led by Cyril Rakovski of Chapman University and published in Communications Medicine, applied causal inference methods to Oracle EHR real-world data and found that combined therapy raised the five-year survival probability by 0.050 compared with no pharmacological treatment. That modest-sounding number could translate to roughly 303,000 additional people crossing the five-year mark, given the size of the diagnosed population in the United States.
Why a five-point survival gain demands attention right now
Both donepezil and memantine have been prescribed for years as symptomatic treatments for Alzheimer’s disease. Donepezil, sold as Aricept, received its initial U.S. approval in 1996. The two drugs are already available together in a fixed-dose capsule called Namzaric, designed for patients stabilized on both medications. Yet neither drug’s label promises a survival benefit. The new findings suggest the combination does something beyond managing day-to-day symptoms.
One plausible explanation is that the paired drugs reduce systemic inflammation rather than simply slowing cognitive decline. Chronic neuroinflammation is a recognized driver of Alzheimer’s progression, and both cholinesterase inhibitors like donepezil and NMDA receptor antagonists like memantine have shown anti-inflammatory properties in preclinical work. If the survival gain stems from lower systemic inflammation, researchers could test the idea by tracking markers such as C-reactive protein and interleukin-6 over time in a prospective cohort of patients starting dual therapy. No such trial has been announced, but the hypothesis offers a concrete next step that goes beyond observational data.
The practical stakes are immediate. Millions of Americans already take one or both of these drugs. A survival difference of 0.050 in probability, applied across a large patient population, could shift clinical conversations about when and whether to prescribe both medications together rather than relying on monotherapy.
Oracle EHR data and doubly robust methods anchor the survival estimate
The Chapman University team drew on Oracle EHR Real-World Data, a large repository of de-identified patient records from U.S. health systems. They compared four exposure groups: patients receiving no Alzheimer’s drugs, those on memantine alone, those on donepezil alone, and those on both. To account for the many differences between patients who end up in each group, the researchers used a doubly robust causal inference framework, a statistical approach that combines outcome modeling with treatment-probability weighting so that the estimate holds even if one of the two models is somewhat misspecified. The result, published in Communications Medicine, was a 0.050 increase in five-year survival probability for the combination group relative to the untreated group.
Earlier clinical trial evidence had already shown short-term benefits when memantine was added to donepezil. A randomized controlled trial published in JAMA demonstrated gains in cognition and function for patients with moderate-to-severe Alzheimer’s disease who received memantine on top of stable donepezil therapy. Pooled analyses of two 24-week placebo-controlled trials later confirmed benefits across cognition, function, and global status, with adverse-event rates similar to those seen in control groups. Those trials, however, were designed to measure symptom management over months, not survival over years. The new observational study fills that gap by following patients across a much longer time horizon.
Chapman University’s news release estimated that approximately 303,000 people could potentially cross the five-year survival threshold if the combination were more widely adopted. That figure depends on assumptions about how many diagnosed patients would be eligible for and adherent to dual therapy, but it gives a sense of scale.
Dosage gaps, confounders, and the missing prospective trial
Several questions remain open. The Oracle EHR dataset provides aggregate survival estimates, but patient-level records with exact covariate balance tables have not been publicly released. That means outside researchers cannot yet verify how well the doubly robust method controlled for differences in disease severity, comorbidities, or socioeconomic factors that might independently affect survival. The study also does not break down five-year mortality by specific dosage levels or duration of combined exposure, so clinicians cannot determine whether a minimum treatment period is needed to see the benefit.
Another limitation is the lack of detailed adherence data. Electronic health records can show prescriptions written and sometimes refilled, but they rarely capture whether patients actually took the medications as directed. If patients who are healthier or have stronger support systems are more likely to stay on both drugs, some of the observed survival advantage could reflect those underlying factors rather than the pharmacology alone.
There is also the question of how the survival benefit might vary across subgroups. The analysis did not publicly report separate estimates by age, sex, race, or baseline disease severity. Those details matter, because physicians may be more cautious about adding a second drug in very frail patients or those with multiple comorbidities. Without subgroup results, it is difficult to know whether the five-point survival gain applies uniformly or is driven by particular segments of the population.
Prospective randomized trials would be the clearest way to answer these questions, but they are expensive and time-consuming. Designing a study powered to detect a 0.050 absolute difference in five-year survival would likely require thousands of participants and long-term follow-up. Until such a trial is launched, clinicians and regulators must weigh the real-world evidence against the inherent uncertainties of observational data.
Regulatory silence and how labels could evolve
So far, regulators have not moved to reinterpret existing prescribing information in light of the survival signal. The memantine–donepezil fixed-dose product is approved based on its ability to maintain or improve cognitive and functional measures, not to extend life. The current U.S. label for Namzaric, available through the FDA’s DailyMed listing, does not claim a mortality benefit or provide guidance on using the drug specifically to improve long-term survival.
For labels to change, regulators would typically look for converging lines of evidence: reproducible observational findings, mechanistic support, and ideally at least some randomized data pointing in the same direction. The new EHR analysis provides one of those pieces, but not all. It may nonetheless influence practice informally, as clinicians who are already comfortable prescribing both drugs weigh the possibility that they are not just easing symptoms but modestly extending life.
Professional societies and guideline panels could also play a role. If they judge the evidence strong enough, they might recommend earlier or more routine initiation of dual therapy in appropriate patients, while emphasizing the need to monitor for side effects and reassess benefits regularly. Such guidance would likely stress shared decision-making, since some patients and families may prioritize quality of life over a small average increase in survival probability.
What clinicians and families can do now
In the absence of definitive randomized data, the EHR findings are unlikely to produce a single, uniform standard of care. Instead, they add another factor to individualized treatment decisions. For a patient already tolerating donepezil, the potential survival advantage could tip the balance toward adding memantine, especially in moderate-to-severe disease. For those with borderline tolerability or complex comorbidities, clinicians may remain cautious.
Families can use the new data as a starting point for conversations with neurologists, geriatricians, or primary care physicians. Questions might include whether the patient is a candidate for dual therapy, how potential benefits compare with risks such as gastrointestinal side effects or dizziness, and what monitoring plan would be in place. Because the survival gain is an average effect across many patients, it cannot predict what will happen for any one individual, but it can help frame expectations.
For researchers, the Chapman analysis underscores the growing role of real-world data in neurodegenerative disease. Large EHR repositories can reveal patterns that traditional trials are not designed to capture, especially when sophisticated causal methods are applied. The challenge now is to build on those signals with targeted mechanistic studies and, where feasible, prospective trials that can either confirm or refine the estimated survival benefit of combining donepezil and memantine.
Alzheimer’s disease remains a condition with limited disease-modifying options and immense personal and societal costs. Against that backdrop, a five-point increase in five-year survival probability is not a cure, but it is not trivial either. If further work validates the signal, dual therapy could become a more deliberate tool not just for managing symptoms, but for modestly extending the lives of people living with the disease.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
